ML18036A937

TVA 015 - TN5025 - TVA 2015 Integrated Resource Plan, Final Report

ML18036A937
Person / Time
Site:	Clinch River
Issue date:	02/05/2018
From:	Tennessee Valley Authority
To:	Office of New Reactors
Fetter A
References
Download: ML18036A937 (214)

Text

Integrated Resource Plan 2 015 F I N A L R E P O R T T E N N E S S E E VA L L E Y AU T H O R I T Y

Message from the CEO TVA is pleased to publish the 2015 Integrated Resource Plan (IRP) that provides direction for how TVA will meet the long-term energy needs of the Tennessee Valley region.

This document and the associated Supplemental Environmental Impact Statement evaluate scenarios that could unfold over the next 20 years. It discusses ways that TVA can meet future electricity demand economically while supporting TVAs equally important mandates for environmental stewardship and economic development across the Valley. And it reinforces that TVA is headed in the right direction. As in the 2011 version, this report indicates that a diverse portfolio is still the best way to deliver low-cost, reliable electricity to those we serve.

The 2015 IRP was developed in collaboration with TVA customers and consumers, business and industry leaders, community, academia, environmental advocates, and everyday people who depend on the public power TVA provides in partnership with local power companies. The involvement of so many passionate and informed people ensured thorough consideration of the differing views, priorities and issues that co-exist in the Valley. Special thanks go to the IRP Working Group and the Regional Energy Resource Council for their efforts to help create a more robust strategy that incorporates all resources to support load growth and serve our customers.

The 2015 IRP is an important document that will assist TVA in fulfilling its mission of serving the people of the Valley to make life better by delivering safe, clean, reliable and affordable electricity; a cleaner environment and sustainable economic opportunities.

We thank our many partners for their contribution to this effort.

William D. Johnson

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Executive Summary 1 Chapter 1 - TVAs Energy Future 7 1.1 TVA Overview 8 1.1.1 TVAs Mission 8 1.1.2 TVAs Customers 9 1.2 Integrated Resource Planning 9 1.2.1 IRP Objectives10 1.2.2 IRP Development10 1.3 Supplemental Environmental Impact Statement10 Chapter 2 - IRP Process11 2.1 Scoping12 2.2 Develop Study Inputs and Framework 12 2.3 Analyze and Evaluate 14 2.4 Present Initial Results and Gather Feedback14 2.5 Incorporate Feedback and Perform Additional Modeling14 2.6 Identify Target Power Supply Mix14 2.7 Approval of IRP Recommendations14 Chapter 3 - Public Participation15 3.1 Public Scoping Period16 3.1.1 Public Meetings and Webinars17 3.1.2 Written Comments17 3.1.3 Results of the Scoping Process17 3.1.4 Additional Comments18 3.2 Public Involvement in Developing Study Inputs and Framework 18 3.2.1 IRP Working Group Meetings18 3.2.2 Public Briefings19 3.3 Public Involvement in Review of the Draft IRP19 3.3.1 Public Meetings  20 Chapter 4 - Need for Power Analysis 25 4.1 Estimate Demand 26 4.1.1 Load Forecasting Methodology 26 4.1.2 Forecast Accuracy 28 4.1.3 Forecasts of Peak Load and Energy Requirements 28 4.2 Determine Reserve Capacity Needs 29 4.3 Estimate Supply 30 4.3.1.1 Baseload Resources30 4.3.1.2 Intermediate Resources 31 4.3.1.3 Peaking Resources 31 4.3.1.4 Storage Resources 31 4.3.2 Capacity and Energy31 4.3.3 Current TVA Capacity and Energy Supply 32 4.4 Calculate the Capacity Gap 33 i

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Chapter 5 - Energy Resource Options 35 5.1 Energy Resource Selection Criteria 36 5.1.1 Criteria for Considering Resource Options 36 5.1.2 Characteristics Required for Resource Options 36 5.2 Resource Options Included in IRP Evaluation 37 5.2.1 Existing Assets by Resource Category 37 5.2.2 New Assets by Resource Category 40 Chapter 6 - Resource Plan Development and Analysis 51 6.1 Development of Scenarios and Strategies  52 6.1.1 Development of Scenarios 52 6.1.2 Development of Planning Strategies  59 6.2 Resource Portfolio Optimization Modeling 62 6.2.1 Development of Optimized Capacity Expansion Plan  63 6.2.2 Evaluation of Detailed Financial Analysis  63 6.2.3 Uncertainty (Risk) Analysis  64 6.3 Portfolio Analysis and Scorecard Development  66 6.3.1 Selection of Metric Categories 66 6.3.2 Development of Scoring and Reporting Metrics 67 6.3.3 Scorecard design 72 6.4 Strategy Assessment Process 73 Chapter 7 - Study Results 75 7.1 Analysis Results76 7.1.1 Firm Requirements and Capacity Gap76 7.1.2 Expansion Plans  77 7.1.3 The Clean Power Plan and the IRP 89 7.2 Scorecard Results 90 7.3 Scoring Metric Comparisons 93 7.4 Preliminary Observations 94 Chapter 8 - Strategy Assessments 95 8.1 Strategy Assessments 96 8.1.1 Cost and Risk Assessment 96 8.1.2 Environmental Stewardship101 8.1.3 Flexibility102 8.1.4 Valley Economics103 8.1.5 Summary of Initial Observations104 8.2 Reporting Metrics Comparisons106 8.3 Sensitivity Analysis107 Chapter 9 - Recommendations 111 9.1 Study Objectives 112 9.2 Findings 112 9.3 Developing the Recommendation 114 9.4 Target Power Supply Mix 115 ii

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Chapter 10 - Implementation Challenges and Next Steps119 10.1 Overview of Next Steps120 10.2 Implementation Challenges120 10.3 Policy Considerations121 10.4 Process and Modeling Improvements 121 10.5 Conclusion122 Appendix A - Generating Resource Cost and Performance Estimates123 Appendix B - Assumptions for Renewables129 Appendix C - Distributed Generation Evaluation Methodology135 Appendix D - 2015 IRP: Modeling Energy Efficiency141 Appendix E - Capacity Plan Summary Charts159 Appendix F - Method for Computing Environmental Metrics171 Appendix G - Method for Computing Valley Economic Impacts185 iii

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents List of Figures Figure 1: 2014 TVA Portfolio........................................................................................................................ 2 Figure 2: Range of MW Additions by 2023 & 2033...................................................................................... 3 Figure 2-1: List of Scenarios and Strategies...............................................................................................13 Figure 31: Distribution of Scoping Comments by State.............................................................................17 Figure 32: Inputs and Framework Public Briefings....................................................................................19 Figure 33: Draft IRP Public Meetings....................................................................................................... 20 Figure 41: Comparison of Actual and Forecasted Annual Peak Demand.................................................. 28 Figure 42: Comparison of Actual and Forecasted Net System Requirements........................................... 28 Figure 43: Peak Demand Forecast ........................................................................................................ 29 Figure 44: Energy Forecast...................................................................................................................... 29 Figure 45: Illustration of Baseload, Intermediate and Peaking Resources................................................. 30 Figure 46: Baseline Capacity, Summer Net Dependable MW.................................................................. 33 Figure 47: Estimating the Capacity Gap................................................................................................... 33 Figure 48: Capacity Gap Range.............................................................................................................. 33 Figure 49: Energy Gap............................................................................................................................ 34 Figure 51: Coal Fleet Map........................................................................................................................ 38 Figure 52: Coal Fleet Portfolio Plans........................................................................................................ 39 Figure 53: List of New Assets...................................................................................................................41 Figure 54: Nuclear Expansion Options..................................................................................................... 42 Figure 55: Coal Expansion Options......................................................................................................... 43 Figure 56: Gas Expansion Options.......................................................................................................... 44 Figure 57: Hydro Expansion Options........................................................................................................ 45 Figure 58: Utility-Scale Storage Options.................................................................................................. 46 Figure 59: Wind Expansion Options.........................................................................................................47 Figure 510: Solar Expansion Options....................................................................................................... 48 Figure 511: Biomass Expansion Options.................................................................................................. 49 Figure 512: DR Expansion Options.......................................................................................................... 49 Figure 513: EE Expansion Options........................................................................................................... 50 Figure 61: Key Uncertainties.................................................................................................................... 53 Figure 62: Scenario Key Characteristics.................................................................................................. 54 Figure 63: Energy Demand Assumptions................................................................................................. 55 iv

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Figure 64: Gas Price Assumptions........................................................................................................... 56 Figure 65: Coal Price Assumptions.......................................................................................................... 57 Figure 66: CO2 Price Assumptions.......................................................................................................... 58 Figure 67: Key Planning Strategy Attributes............................................................................................. 60 Figure 68: Planning Strategies Key Characteristics.................................................................................. 61 Figure 69: Strategy Descriptions.............................................................................................................. 62 Figure 610: Sample Stochastic Result..................................................................................................... 64 Figure 611: Example Uncertainty Ranges................................................................................................. 65 Figure 612: Strategic Imperatives............................................................................................................. 66 Figure 613: Scoring Metrics..................................................................................................................... 68 Figure 614: Scoring Metric Formulas........................................................................................................ 69 Figure 615: Reporting Metrics.................................................................................................................. 70 Figure 616: Reporting Metric Formulas.....................................................................................................71 Figure 617: Scorecard Alignment............................................................................................................. 72 Figure 618: Scorecard Template.............................................................................................................. 72 Figure 71: Firm Requirements by Scenario................................................................................................76 Figure 72: Range of Capacity Gaps by Scenario...................................................................................... 77 Figure 7-3: Incremental Capacity Additions for All 25 Cases...................................................................... 78 Figure 74: Capacity (Summer Net Dependable Megawatts) for Strategy A by Scenario............................ 80 Figure 75: Energy (Terawatt Hours) for Strategy A by Scenario................................................................ 80 Figure 76: Capacity (Summer Net Dependable Megawatts) for Strategy B by Scenario........................... 82 Figure 77: Energy (Terawatt Hours) for Strategy B by Scenario................................................................. 82 Figure 78: Capacity (Summer Net Dependable Megawatts) for Strategy C by Scenario........................... 84 Figure 79: Energy (Terawatt Hours) for Strategy C by Scenario................................................................ 84 Figure 710: Comparison of Energy Efficiency Resources in Strategy D..................................................... 85 Figure 711: Capacity (Summer Net Dependable Megawatts) for Strategy D by Scenario.......................... 86 Figure 712: Energy (Terawatt Hours) for Strategy D by Scenario............................................................... 86 Figure 713: Comparison of Renewable Resources in Strategy E.............................................................. 87 Figure 714: Capacity (Summer Net Dependable Megawatts) for Strategy E by Scenario.......................... 88 Figure 715: Energy (Terawatt Hours) for Strategy E by Scenario .............................................................. 88 Figure 716: Strategy A Scorecard............................................................................................................ 90 Figure 717: Strategy B Scorecard............................................................................................................. 91 Figure 718: Strategy C Scorecard............................................................................................................ 91 v

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Figure 719: Strategy D Scorecard............................................................................................................ 92 Figure 720: Strategy E Scorecard............................................................................................................ 92 Figure 721: Scoring Metrics by Strategy & Scenario................................................................................. 94 Figure 81: System Average Cost.............................................................................................................. 96 Figure 82: Total Plan Cost (PVRR)............................................................................................................ 97 Figure 83: Risk/Benefit Ratio................................................................................................................... 98 Figure 84: Risk Exposure........................................................................................................................ 99 Figure 85: Cost/Risk Trade-Offs............................................................................................................ 100 Figure 86: Environmental Impacts..........................................................................................................101 Figure 87: System Regulating Capability.................................................................................................102 Figure 88: Valley Economics...................................................................................................................103 Figure 89: Reporting Metrics by Strategy & Scenario..............................................................................107 Figure 8-10: List of Sensitivity Cases........................................................................................................108 Figure 9-1: 2014 TVA Portfolio.................................................................................................................. 113 Figure 9-2: Evaluated MW Additions/Retirements by 2033 from IRP Base and Sensitivity Case Analysis......................................................................... 114 Figure 9-3: Range of MW Additions by 2023 & 2033............................................................................... 116 Figure 10.1: Remaining IRP Process Steps...............................................................................................120 Figure B1: Example of Wind Monthly-mean variability of net power capacity by 3TIER............................131 Figure B2: Sites across Tennessee Valley with historical solar irradiance data supplied by CPR .............132 Figure B3: Solar Fixed Axis and Utility Tracking Capacity Factors by Month............................................132 Figure B4: NDC by hour of the top 20 peak load days of Summer 1998-2013........................................132 Figure C1: DG Market Segments and Penetration Levels across IRP Scenarios......................................136 Figure C2: Correlation of IRP CO2 uncertainty values to EIA source data...............................................137 Figure C3: Development of National Renewable DG Penetration Levels..................................................138 Figure C4: National Renewable Energy Adoption Levels (Utility-led and DG)...........................................138 Figure C5: Residential/Commercial DG Adoption Levels (by 2040).........................................................139 Figure C6: Residential/Commercial DG Adoption Levels (Annual)...........................................................139 Figure D1: Energy Efficiency Performance on a Typical Peak Summer Day (2023)...................................143 Figure D2: Energy Efficiency Monthly Profile (2023).................................................................................144 Table D4: Tier Step Changes..................................................................................................................147 Figure D3: Levelized EE block cost comparison ($/MWh) compared to a greenfield combined cycle plant over time..........................................................................147 vi

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Figure D4: Example of a residential audit program modeled as part of a residential block......................152 Figure D5: Planning Factor Adjustment over time....................................................................................154 Figure D6: Uncertainty bands in $/MWh for each of the EE sector blocks as compared to a greenfield combined cycle plant................................................................156 Figure D7: Levelized Cost comparisons in 2015 (2015$/MWh)................................................................157 Figure D8: Levelized cost comparison ($/MWh) of EE tiers in 2015 and 2033........................................157 Figure D9: Levelized Cost Comparison by Sector through time..............................................................157 Figure F1: Scoring Metrics......................................................................................................................172 Figure F2: Scoring Metric Formulas........................................................................................................172 Figure F3: Reporting Metrics..................................................................................................................173 Figure F4: Reporting Metric Formulas.....................................................................................................173 Figure G1: Input and Output Impacts......................................................................................................186 Figure G2: Results..................................................................................................................................188 Figure G3: Current Outlook....................................................................................................................189 Figure G4: Stagnant Economy................................................................................................................191 Figure G5: Growth Economy..................................................................................................................193 Figure G6: De-Carbonized Future...........................................................................................................195 Figure G7: Distributed Marketplace.........................................................................................................197 Figure G8: Nonfarm Employment...........................................................................................................199 List of Tables Table 81: Summary of Observations by Metric Category104 Table 82: Summary of Observations by Strategy 105 Table D1: Tier, Sector, and Block Hierarchy145 Table D2: Weighting of EE Programs 145 Table D3: Net to Gross ratios and Lifespans for the EE programs within sectors146 Table D5: Block Characteristics for each sector148 Table D6: Resource Characteristic Comparison with EE150 Table D7: Design and Delivery Uncertainties151 Table D8: Indirect and Direct Stochastic Variables155 vii

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Table of Contents Acronym Index IRP Integrated Resource Plan TVA Tennessee Valley Authority EE Energy Efficiency IRPWG Integrated Resource Plan Working Group PPA Purchase Power Agreement SMR Small Modular Nuclear Reactor LPC Local Power Company EIS Environmental Impact Statement SEIS Supplemental Environmental Impact Statement NEPA National Environmental Policy Act EEDR Energy Efficiency and Demand Response FY Fiscal Year MW Megawatt MWh Megawatt hour GWh Gigawatt hour TWh Terawatt hour kWh Kilowatt Hour MAPE Mean Absolute Percent Error CC Combined Cycle gas plant CT Combustion Turbine CO2 Carbon Dioxide EPU Extended Power Uprates PWR Pressurized Water Reactor APWR Advanced Pressurized Water Reactor HVDC High Voltage Direct Current IGCC Integrated Gas Combined Cycle SCPC Supercritical Pulverized Coal CCS Carbon Capture and Sequestration PVRR Present Value of Revenue Requirements viii

Executive Summary Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 1

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Executive Summary Overview Developing the Recommendation The Tennessee Valley Authoritys 2015 Integrated The recommendation takes into account Resource Plan (IRP) will guide TVA in making decisions customer priorities around power cost and about the energy resources used to meet future reliability across different futures. Implementing demand for electricity through 2033. More like a the least-cost resource plan with these priorities compass than a GPS, it provides broad direction but in mind will help ensure TVA continues to fulfill not street-by-street instructions. its mission to serve the people of the Tennessee Valley.

The recommendation in this IRP provides strategic guidance on the resource mix to successfully respond In developing a recommendation from the study, to changing market conditions. The resource mix TVA has elected to establish guideline ranges will be: low cost, reliable, risk-informed, diverse, for key resource types (owned or contracted) environmentally responsible and flexible. that make up the target power supply mix. This general planning direction is expressed over the This study reinforces the importance that TVAs power twenty year study period while also including be reliable, affordable and sustainable into the future. more specific direction over the first ten year Our resource additions will build on TVAs existing period. In order to distill the considerable number diverse asset portfolio reflected in Figure 1. of cases evaluated through the original scenario and strategy analysis and the sensitivity cases, the recommendation uses ranges that are centered on results obtained under the Current Outlook scenario. See Section 2.2, Develop Study Figure 1: 2014 TVA Portfolio Inputs and Framework, for Our results show no immediate needs for new base a description and definition of scenarios. The load plants after Watts Bar Unit 2 comes online other scenario results provide a sense of how the and uprates are completed at Browns Ferry nuclear recommended mix might change as the future plant. Instead, we can rely on additional natural gas changes. The need to shift the resource mix will generation (combined cycle and combustion turbine), be based on these key variables:

greater levels of cost-effective energy efficiency, and increased contributions from competitively priced

• Changes in the load forecast renewable power. We also expect to have less coal-

• The price of natural gas and other based generation in our energy mix than we do today, commodities although it will continue to play an important role in the

• The pricing and performance of energy portfolio. efficiency and renewable resources

• Impacts from regulatory policy or breakthrough technologies 2

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Executive Summary The first three variables represent the fundamental Target Power Supply Mix drivers for most of the variation in the resource plans The recommendation for the power supply mix is produced across the strategy/scenario combinations.

presented in the form of ranges around boundaries Our planning direction, while initially focused around established by the IRP results. The solid bars represent the current view of the future, is flexible enough to the recommended range from the IRP scenario that indicate how that power supply mix would shift if one best represents our current estimation of the future.

or more of these key variables exhibits a material However, recognizing that a variety of future scenarios change from the forecasts used in the IRP. We also are possible, the wider range (shown in horizontal black recognize that impacts from breakthrough technologies lines) is provided to represent how the resource portfolio (like a significant advance in energy storage) would may respond in different future scenarios.

be a game-changer, and TVA will continue to monitor emerging technology as it develops. The recommended ranges represent incremental additions (or retirements) from the existing resource This approach provides a more robust recommendation fleet and may be owned or contracted assets. The than was developed in the 2011 IRP. While that results are bounded by the full range of the IRP cases approach provided a solid framework for the resource and sensitivity runs which reaffirm the merits of a decisions TVA has made since the TVA Board accepted diverse portfolio. TVA will closely monitor key input the IRP planning direction in the spring of 2011, the variables including changes in the load forecast, the changing utility marketplace requires a more flexible price of natural gas and other commodities, the pricing guide that provides decision-makers with a clear and performance of energy efficiency and renewable understanding of how the resource mix would evolve in resources, and impacts from regulatory policy or response to future uncertainties. The recommendation breakthrough technologies to help determine whether meets the dual objective of ensuring flexibility to respond adjustments should be made to the recommended to the future while providing guidance on how our ranges.

resource portfolio should change as the future unfolds.

Figure 2: Range of MW Additions by 2023 & 20331 1 MWs are incremental additions from 2014 forward to align to the IRP analysis base year. Board-approved coal retirements and natural gas additions as of August 2015 are excluded from the totals.

3

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Executive Summary Figure 2 shows the boundaries of resource additions Wind:

we are proposing by the end of the first 10 years of Add between 500 and 1,750 MW by 2033, depending the study (2023) and by the end year of the study on pricing, performance, and integration costs. Given (2033), shown in Megawatts (MW). The specific the variability of wind selections in the scenarios, recommendations by resource type are summarized evaluate accelerating wind deliveries into the first 10 below. years of the plan if operational characteristics and pricing result in lower-cost options.

Coal:

Continue with announced plans to retire units at Allen, Natural Gas Colbert, Johnsonville, Paradise and Widows Creek. (Combustion Turbine and Combined Cycle): add Evaluate the potential retirement of Shawnee Fossil between 700 and 2,300 MW by 2023, and between Plant in the mid-2020s if additional environmental 3,900 and 5,500 MW by 2033. The key determinants of controls are required. Consider retirements of fully future natural gas needs are trajectories on natural gas controlled units if cost-effective. pricing and energy efficiency and renewables pricing/

availability.

Nuclear:

Complete Watts Bar Unit 2 and pursue additional TVAs recommended planning direction reaffirms its power uprates at all three Browns Ferry units by 2023. commitment to a diverse resource portfolio guided Continue work on Small Modular Reactors as part of by the least-cost system planning mandate. The Technology Innovation efforts and look for opportunities ranges above provide a general guideline for resource for cost sharing to render these more cost-effective for selection, but the full case analysis studied in the IRP our ratepayers. and the SEIS includes ranges much broader than shown above driven by key drivers such as material Hydro: changes in economic conditions or regulations. We Pursue an additional 50 MW of hydro capacity at TVA believe meeting our future needs in accordance facilities and consider additional hydro opportunities with the resource technologies and ranges in this where feasible. recommendation will position TVA to continue to deliver reliable, low-cost, and cleaner power to the people of Demand Response:

the Tennessee Valley.

Add between 450 and 575 MWs of demand reduction products by 2023 and similar amounts by 2033, depending on availability and cost of this customer- How Did We Get Here?

owned resource.

We took the findings in the initial study and the Energy Efficiency: subsequent additional information gained from Achieve savings between 900 and 1,300 MW by 2023, sensitivity analysis to form the recommendation. We and between 2,000 and 2,800 MWs by 2033. Work with used five key measures to evaluate the performance of LPCs to refine delivery methods, program designs and the 25 initial plans created as part of the study. Those program efficiencies, with the goal of lowering total cost results told us:

and increasing deliveries of efficiency programs.

• Cost: Total costs are similar for many of the cases Solar: over the long-term, and strategies that allow for a Add between 150 and 800 MW of large-scale solar diverse mix of resource additions have a lower cost by 2023, and between 3,150 and 3,800 MW of large- than those that emphasize particular technologies.

scale solar by 2033. The trajectory and timing of Higher amounts of energy efficiency may create solar additions will be highly dependent on pricing, an upward pressure on rates in future years due to performance and integration costs. reduced sales.

4

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Executive Summary

• Financial Risk: Risks are minimized by maintaining and performance assumptions are critical to the final a diverse portfolio and not over-emphasizing any result, and lower costs or less uncertainty around specific resource type. this resource would increase its selection in the

• Environmental Stewardship: All strategies show portfolio. Our study results also highlight that higher significant improvement in TVAs environmental volumes of Energy Efficiency tend to increase system footprint and position the Tennessee Valley to have average costs. TVA and our LPC partners will need continued reductions in CO2 emissions. Strategies to balance Energy Efficiency amounts and programs that emphasize energy efficiency or renewables have to ensure that those who cannot participate in these the best environmental results. programs are not disproportionately impacted.

• Valley Economics: All strategies have a similar impact

• Renewable selection is highly dependent on gas on overall economic health and contribute to a price, load and cost and performance assumptions.

strong, vibrant economy across the region.

• Natural gas pricing and load levels remain key

• Flexibility: System flexibility is generally equivalent sensitivities for all resource decisions.

in most cases but is reduced when renewables are The results of these analyses supported the ranges strongly emphasized. established in the initial findings. The sensitivity cases, coupled with the original 25 case results, provide a robust set of 2,000 potential resource Reviewing these results led to questions from additions evaluated in the IRP from which the final stakeholders about how changes in assumptions or recommendation was derived.

resource choices might impact the findings. A series of sensitivity cases were evaluated with five main categories: nuclear additions, modified assumptions Policy Considerations for energy efficiency, alternative renewable resource In the process of developing the cases and reviewing costs, impacts associated with forcing resources the results with stakeholders, a number of policy-related not otherwise selected into the mix, and changes in issues were raised that are outside the scope of the fundamental drivers such as load growth and fuel IRP itself but will need to be considered as we move pricing.

toward implementation of any recommendations from Sensitivity Cases Key Findings the study.

• New nuclear or coal assets would offset gas builds For example, we recognize that a commitment to and renewable purchases. Nuclear additions significant levels of energy efficiency as part of the increase total cost but lower fuel risk. Small resource portfolio will likely put upward pressure on Modular Reactors are presently cost-prohibitive, but rates (absent a redesign) that could have negative cost-sharing would render them more financially consequences for low/fixed income customers as attractive. Subsequent IRPs will need to address the well as renters. The details of the approach we might expected expiration of licenses for TVAs operating take are outside the scope of the IRP report, but the nuclear units which may occur beyond the present study work we have completed will inform the follow-study window. on planning and evaluation of the Energy Efficiency portfolio. We also know that program design will be

• Energy Efficiency was successfully modeled as a key challenge to ensure that the broadest possible a selectable, supply-side equivalent resource.

Energy Efficiency portfolio can be offered through In general, energy efficiency programs eliminate the LPCs to minimize possible bill impacts on non-the need for natural gas units as well as some participants.

renewable purchases. As with any resource, cost 5

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Executive Summary We also realize that electric rates and job growth are critical concerns for Valley residents. While we have chosen to focus on two specific metrics to assess the macro-economic impacts of our resource choices, TVA remains committed to our least-cost mandate and our responsibility for regional economic development.

Although the IRP itself does not analyze either of these issues, the findings in this planning study do become key inputs in the financial planning cycle that helps TVA set rates and fund economic development activities.

There are several other policy issues that come into play when implementing recommendations from the IRP, especially if the target power supply mix relies on more load-side options, like Energy Efficiency programs, or resources that are more dispersed, like wind or solar facilities. Because of our unique business model, TVA and Local Power Companies (LPCs) will have to collaborate in new and innovative ways to ensure that this evolving resource portfolio remains reliable and provides the most value to all customers.

6

Chapter 1 TVAs Energy Future Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 7

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 1: TVAs Energy Future 1 TVAs Energy Future 1.1 TVA Overview The 2015 Integrated Resource Plan (IRP) will guide 1.1.1 TVAs Mission TVA in making decisions about the energy resources TVA was created by Congress in 1933, and charged we will use to meet future demand for electricity in the with a unique mission: to improve the quality of life in a Tennessee Valley. Having a long-range energy resource seven-state region through the integrated management plan enables us to provide affordable, reliable electricity of the regions resources. To help lift the Tennessee to the people we serve. It is crucial in a constantly Valley out of the Great Depression, TVA built dams for changing business and regulatory environment and will flood control, provided low-cost power and commercial better equip us to meet many of the challenges facing shipping, restored depleted lands, and raised the the electric utility industry. standard of living across the region. As times have changed, we have changed with them, meeting new A key challenge is projecting how much power we will challenges and bringing new opportunities. Today, TVA need, when and where, and identifying the optimum continues to serve the people of the Tennessee Valley mix of energy resources to meet future power demand. through its work in three areas: Energy, the Environment Electricity cant be stored economically in sufficient and Economic Development.

quantities, so electric utilities must constantly balance power supply and demand. Energy efficiency programs Energy can help reduce that demand, and various energy Safe, clean, reliable and affordable electricity powers resources can be used to supply future demand - from the economy of our region and enables greater constructing new generation facilities to contracting prosperity and a higher quality of life for everyone. After with others to provide needed electricity, including safety, our top priority is keeping our electric rates as renewable generation. But all of these options take low as feasible and our reliability as high as possible.

time to implement. Given the long lead times required to plan, permit and build generating facilities, demand TVA operates the nations largest public power system, forecasts involve 10- to 20-year outlooks. including 41 active coal-fired units, six nuclear units, 109 conventional hydroelectric units, four pumped-storage In addition, we must take steps to ensure TVA has the units, 87 simple-cycle combustion turbine units, 11 transmission infrastructure to get electricity to where it combined cycle units, five diesel generator units, one is needed. We currently operate and maintain 16,000 digester gas site and 16 solar energy sites.2 We also miles of transmission lines across the Tennessee Valley purchase a portion of our power supply from third-party region. As the population grows we must upgrade or operators under long-term power purchase agreements expand this system. TVA must allow adequate time to (PPAs).

properly study, engineer, site and plan environmental reviews to build additional transmission infrastructure. TVAs 16,000-mile-long transmission system is one of the largest in North America. For the past 14 years, the All of these activities entail varying levels of risk and system achieved 99.999 percent power reliability. It uncertainties which we try to account for in our IRP efficiently delivered more than 161 billion kilowatt-hours analyses and energy resource portfolio. In earlier IRPs, of electricity to customers in FY 2014.

we determined that a diversified energy portfolio is one of the best ways to reduce risks, and we have TVA makes annual investments in science and reconfirmed this in our 2015 IRP. It is important technology innovation that enable us to be at the that we maintain a mix of energy resource options, forefront of advances in the utility industry and help including nuclear, natural gas, coal, energy efficiency, us meet future business and operational challenges.

hydroelectric power and other renewables, to reduce Core research activities directly support improving our the risks associated with relying too much on a specific generation and delivery assets, air and water quality fuel type.

2 As of September 30, 2014 8

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 1: TVAs Energy Future and clean energy integration. Currently, we are involved the number and quality of jobs in the Valley and to in research activities related to emerging technological benefit the power system through smarter energy use.

advances in small modular nuclear reactors (SMRs),

grid modernization for transmission and distribution 1.1.2 TVAs Customers systems, energy utilization technologies and distributed Our relationships with our customers are crucial energy resources. to providing affordable electricity to residents and businesses. As largely a wholesaler of electricity, TVA Environmental Stewardship works in partnership with local power companies (LPCs)

TVA manages natural resources of the Tennessee Valley to deliver affordable, reliable electricity. We also deliver for the benefit of the regions people. We manage the electricity directly to some customers, large industries Tennessee River system and associated public lands and federal installations and exchange power with other to reduce flood damage, maintain navigation, support interconnected utility systems.

power production, enhance recreation, improve water quality and protect shoreline resources. LPCs make up most of TVAs customer base and are the backbone of the regions power distribution TVA manages its power system to provide clean energy system. Accounting for roughly 87 percent of total TVA and minimize environmental impacts from its operations. sales and 90 percent of total revenue, the LPCs are Today, air quality across the region is the best it has municipally-owned and consumer-owned (cooperative) been in more than 30 years. Since 1977, TVA has spent utilities. TVA generates and delivers electricity to the about $6 billion on air pollution controls and is investing LPCs, which deliver electricity to their residential, approximately $1 billion in more control equipment at commercial and industrial customers. Municipal LPCs our Gallatin Fossil Plant in middle Tennessee. Emissions comprise the largest block of TVA customers. Many of of nitrogen oxides (NOx) are 91 percent below peak the consumer-owned cooperative utilities were formed 1995 levels and emissions of sulfur dioxide (SO2) are 95 to bring electricity to once-sparsely populated rural, percent below 1977 levels through 2013. remote areas of the region.

TVAs emissions of carbon dioxide (CO2) were reduced Large industries and federal installations that buy 32 percent between 2005 and 2013. We project electricity directly from TVA, such as Oak Ridge approximately a 40 percent reduction in CO2 emissions National Laboratory, account for 13 percent of total by 2020 from 2005 levels. TVA is also reducing water sales and 10 percent of total revenue. TVAs electricity use and waste production from its operations as it exchanges with interconnected utilities also can retires coal plants, increases generation from natural produce revenue.

gas and renewable sources, and converts coal combustion residuals management to dry handling and TVA power contracts govern customer relationships, storage. including the pricing or rate structure under which power is sold. Our contracts with LPCs obligate TVA to Economic Development generate and deliver enough electricity to meet their full TVAs large power system, diverse fuel mix and robust electric load, including reserves, now and in the future.

transmission system allows us to provide high reliability and competitive rates to attract industry to our region. 1.2 Integrated Resource Planning During the past five years, TVA has helped attract or The purpose of integrated resource planning is to retain 240,000 jobs in our service territory and secure meet future power demand by identifying the need for more than $30 billion in capital investment for the region generating capacity and determining the best mix of through the Valley Investment Initiative program. This resources to meet the need on a least-cost, system-program, established in 2008, is designed to increase wide basis. The integrated approach considers a broad 9

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 1: TVAs Energy Future range of feasible supply-side and demand-side options objectives) were considered as different combinations and assesses them with respect to financial, economic of strategies and predictions of future conditions were and environmental impacts. The 2015 IRP will revise analyzed and evaluated.

our 2011 IRP. We are updating the 2011 IRP earlier than planned because several of the assumptions used in its We conducted the IRP process in a transparent, development changed. These include reduced demand inclusive manner that provided numerous opportunities for electricity and greater availability and lower cost of for the public to learn about the project and participate natural gas. in it. We met regularly with a wide range of stakeholders who served on the 2015 IRP Working Group. This 1.2.1 IRP Objectives group was composed of individuals representing state The following objectives guide the development of this agencies, distributors of TVA power, industry groups, IRP: environmental and energy advocates, academia and research institutions, and business and economic

• Deliver a plan aligned to mandated least-cost development professionals. (More information about planning principles the IRP Working Group is provided in Chapter 3, Public

• Manage risk by utilizing a diverse portfolio of supply Participation.) We believe this extensive outreach and demand-side resources produced a better IRP and are grateful for the questions

• Deliver cleaner energy and continue to reduce raised and the feedback and insights provided.

environmental impacts

• Evaluate increased use of renewables, energy 1.3 Supplemental Environmental efficiency, and demand response resources Impact Statement

• Ensure the portfolio delivers energy in a reliable manner As a federal agency, TVA must comply with the National

• Develop the ability to dynamically model energy Environmental Policy Act of 1970 (NEPA). This act efficiency in the study requires all federal agencies to consider the impact

• Provide flexibility to adapt to changing market of their proposed actions on the environment before conditions and identify significant sign posts making decisions. The NEPA process provides a

• Improve credibility and trust through a collaborative structured way for analyzing alternative actions and for and transparent process involving the public in the decision-making process.

• Integrate stakeholder perspectives throughout the For the development of this IRP, the primary product study from the NEPA process is a supplement to the 2011 environmental impact statement (EIS).

1.2.2 IRP Development The EIS focuses on the potential impacts of the various TVAs 2015 IRP was developed over a two-year period IRP strategies more closely and in greater detail than with extensive technical and economic analyses and do the environmental metrics presented in this IRP.

significant participation from our customers and other The impacts of actions to implement the IRP, such as stakeholders. building and operating a new generating facility, will be We used an integrated, least-cost system planning the subject of action- and site-specific NEPA reviews.

process that takes into account the demand for This study was prepared in accordance with NEPA, electricity, resource diversity, reliability, costs, risks, Council of Environmental Quality regulations for environmental impacts and the ability to dispatch implementing NEPA, and TVAs procedures for energy resources. Forecasts of inflation, commodity implementing NEPA.

prices and environmental regulations were evaluated simultaneously to provide needed information.

Constraints (corporate, strategic and environmental 10

Chapter 2 IRP Process Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 11

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 2: IRP Process 2 IRP Process results of these potential actions and potential future environments describe the portfolio in areas such as TVAs 2015 IRP process consists of seven distinct steps: operations, financials, environmental impact, macro-economics, and reliability.

1. Scoping Our goal is to identify an energy resource plan that

2. Develop Study Inputs and Framework performs well under a variety of future conditions

3. Analyze and Evaluate (e.g., a strong economy or a weak economy) thereby

4. Present Initial Results and Gather Feedback reducing the risk that a selected strategy or plan would

5. Incorporate Feedback and Perform Additional perform well under one set of future conditions, but Modeling poorly under a different set of conditions. This increases

6. Identify Preferred Target Supply Mix the likelihood that TVAs plan will provide least-cost

7. Approval of Recommended Plan solutions to future demands for electricity from its power system regardless of how the future plays out.

Public participation was integral to the process and is This decision making framework requires use of explained in more detail in Chapter 3. Steps 2 through 6 a scenario planning approach. Scenario planning are explained in more detail in Chapter 6. provides an understanding of how the results of near-term and future decisions would change under different 2.1 Scoping conditions.

The public scoping period for TVAs 2015 IRP began in October 2013. The objective in this initial step was Future decisions that produce similar results under to identify resource options, strategies and future different conditions may mean that these decisions conditions that merit evaluation in the IRP process. provide more predictable outcomes, whereas decisions Public scoping comments covered a wide range of that result in major differences are less predictable and issues, including the nature of the integrated resource therefore more risky.

planning process, preferences for various types of At the outset of our 2015 IRP process, we developed power generation, increased energy efficiency and a set of five resource planning strategies that would be demand response (EEDR) and the environmental analyzed as part of the IRP. These planning strategies impacts of TVAs power generation. The comments represent received helped to identify issues important to the decisions that Strategies represent future public and to lay the foundation for the supplement to TVA controls (e.g., business decisions over which the 2011 Environmental Impact Statement that supports asset additions, TVA has full control.

the 2015 IRP. idling coal plants, integration of Scenarios represent future 2.2 Develop Study Inputs and more flexible conditions that TVA cannot Framework resource options) control.

When developing a long-term plan for a power system, as opposed to A portfolio is the intersection utilities typically use a least-cost decision making the scenarios of a strategy and a scenario framework that focuses on a single view of the future. described below, and represents a multiyear At TVA, we also use a least-cost decision making which represent energy resource plan detailing framework. The Integrated Resource Plan informs TVA aspects of how TVA intends to meet on how potential resource portfolios could perform the future that future power needs.

given different market and external conditions. The TVA does not 12

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 2: IRP Process control (e.g., more stringent regulations, fuel prices, Each scenario can be thought of as a model of a construction costs). possible future. In one model, the economy might stagnate, fuel prices drop and electricity demand Different mixes of resource options (generating remains flat. In another, strong economic recovery could technologies and/or energy efficiency programs) formed lead to increased fuel prices and to rapid recovery in the framework for distinct planning strategies that were electricity sales and long-term demand growth and assessed over the 20-year IRP planning horizon. The increase the cost of building generating sources.

feasibility of each planning strategy was determined based on input from subject matter experts and To better assess the robustness of the strategies stakeholders. evaluated for this IRP, we purposely structured these scenarios to present different challenges to the Together we then developed a series of five scenarios resource planning strategies. The scenarios differ from representing alternative plausible futures to help us test each other in key areas, such as projected customer the performance of the resource planning strategies demand, fuel prices and future economic and regulatory under different conditions and, ultimately, to identify the conditions.

strategy that might represent the most flexible approach to ensuring the lowest cost, most reliable power for our The five scenarios and five strategies are shown in customers. Figure 2-1.

Figure 2-1: List of Scenarios and Strategies 13

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 2: IRP Process 2.3 Analyze and Evaluate 2.5 Incorporate Feedback and Perform After the resource planning strategies and scenarios Additional Modeling were developed, the performance of each planning After the public comment period ended, all comments strategy was analyzed in detail across all of the were reviewed and combined with other similar scenarios. This phase of the IRP used industry- comments as appropriate. We have responded to all standard capacity expansion planning and production substantive comments either by revising the IRP or cost-modeling software to estimate the total cost of associated supplemental EIS or by providing specific each combination of strategy and scenario. Other answers in the final supplemental EIS.

metrics, financial risks and environmental impacts, were developed from the cost-modeling results. 2.6 Identify Target Power Supply Mix Unique resource plans or portfolios were developed, After review of the public comments received and any one for each combination of scenario and strategy. additional analysis, TVA staff identified a target power Each of the 25 portfolios represented a long-term, least- supply mix based on one or more of the planning cost plan of different resource mixes that could be used strategies evaluated in the IRP. This general target to meet the regions power needs. reflects the mix of resources (supply and demand side) that best position the utility for success in a variety Every portfolio was ranked using metrics within a of alternative futures, while preserving the flexibility consistent, standard scorecard. Care was taken to necessary to respond to uncertainty.

note those portfolios that performed best overall, and those that performed well in most models of the future. 2.7 Approval of IRP Recommendations The metrics were chosen based on importance to No sooner than 30 days after the Notice of Availability, TVAs mission and captured financial, economic and the associated EIS will be published in the Federal environmental impacts. Portfolios were analyzed for Register and the TVA Board of Directors will be asked their robustness under stress across multiple scenarios, to approve the recommendations included in the study, as opposed to total overall performance since metrics including the target power supply mix. The Board will alone could signify good performance in one or two decide whether to approve the recommendations scenarios, but average or poor performance in all presented in the study, to modify them or to approve an others. alternative. The Boards decision will be described and explained in a Record of Decision.

2.4 Present Initial Results and Gather Feedback The draft 2015 IRP was released for public review and comment in March 2015. It presented a range of viable planning strategies for further consideration, and included scorecards and assessments using key metrics. As in the scoping period, TVA encouraged public comments on the draft IRP and associated supplemental EIS. These comments helped us identify public concerns and recommendations for the future operation of the TVA power system.

14

Chapter 3 Public Participation Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 15

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation 3 Public Participation The Tennessee Valley Renewable Information Exchange (TV-RIX) was established in September 2012, and Understanding the varying needs and priorities of our was actually a result of the 2011 IRP Next Steps to 9 million stakeholders and striking a balance can be further analyze renewable technologies, business challenging, but is a key to the IRP process. Gaining models and market trends. The group consisted of 17 that perspective is why we used a transparent and members representing renewable interest groups, state participatory approach in in developing this long-range governments, national/regional expertise and utility plan. Obtaining diverse input and support for the IRP industry representatives. TV-RIX provided inputs on was one of our goals. We wanted to make sure those biomass, hydro, solar and wind resources for the IRP who wanted to participate, could do so.

modeling process.

Our public involvement goals were to:

The Energy Efficiency Information Exchange (EEIX) was

• Engage numerous stakeholders with differing established in October 2013 to focus on the exchange viewpoints throughout the process. of ideas on naturally occurring adoption rates of energy

• Incorporate public opinions into the development efficiency. The group had 13 members representing of the IRP, including opportunities to review and local power companies, state energy offices and comment on various inputs, analyses and options non-government organizations. This group assisted in being considered. developing simple, flexible and cost-effective portfolios

• Encourage open and honest communication in order to be used in the IRP analysis and selection process.

to provide a sound understanding of the process Public involvement was a particular focus throughout

• Create public awareness and opportunities to the IRP process, including steps 1 and 2, Scoping and receive feedback.

Develop Study Inputs and Framework, and as part of

• Form an IRP Working Group made up of people step 4, Present Initial Results and Gather Feedback.

representing the broad perspectives of those who live and work in the Valley.

3.1 Public Scoping Period The formation of an IRP Working Group was a To begin the 2015 IRP process, TVA announced cornerstone of the public input process for the 2015 the start of a 33-day public scoping period on IRP, just as it was for the 2011 study. Working Group October 21, 2013.

members reviewed input assumptions and preliminary The Scoping period was publicized across results and provided feedback throughout the process. the Tennessee Valley through news releases, They provided their individual views to TVA, as well as advertisements and a notice on TVAs website. Notices representing and keeping their constituencies informed also were sent to people who participated in the regarding the IRP process. development of TVAs 2011 IRP.

The 2015 Working Group consisted of 18 In addition, on October 31, 2013, TVA published a representatives from business and industry, state notice in the Federal Register of its intent to prepare agencies, government, distributors of TVA power, a supplement to the 2011 IRP Environmental Impact academia, and energy and environmental non- Statement.

governmental organizations.

At the start of the public scoping period, we explained In addition to the IRP Working Group, individuals on why we were updating our IRP, its focus, how we would two stakeholder groups provided TVA guidance and conduct the planning process, and how the results expertise on renewable energy resources and energy would guide our future energy resource decisions.

efficiency and demand response.

16

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation Our goals in conducting public scoping were to ensure:

• Stakeholder issues and concerns were identified early and studied properly.

• Reasonable alternatives were considered.

• Uncertainties that could impact costs or performance of certain energy resources were identified.

• Input received was properly considered and would lead to a thorough and balanced IRP.

As part of scoping, we collected public input through public meetings, webinars and written comments.

3.1.1 Public Meetings and Webinars TVA held two public meetings as part of the scoping period. The first was in Knoxville, Tenn., on October Figure 31: Distribution of Scoping Comments by State 24, 2013, and the second was in Memphis, Tenn., on November 6, 2013. Both meetings were simulcast as 3.1.3 Results of the Scoping Process webinars, available online. The information collected during the public scoping period helped shape the initial framework of TVAs 2015 At each meeting, TVA staff described the process IRP and was used to determine which resource options of developing the IRP and associated SEIS and then should be considered.

responded to questions from meeting attendees both in person and online. The majority of the scoping comments were generated by the Sierra Club and Tennessee Environmental Participants included the public, congressional, state Council, consisting of email forms that thanked TVA for and local officials, representatives from local power recent coal plant retirement decisions and urged TVA companies, non-governmental organizations and other to prioritize the use of solar and wind energy, increase special interest groups, and TVA employees.

energy efficiency efforts, and work to reduce the local About 85 people attended the meetings in person or via economic impacts of coal plant retirements. A smaller webinar. campaign, promoted by organizations affiliated with the regional coal industry, submitted form letters citing the 3.1.2 Written Comments abundance and stable cost of coal, the high capacity TVA accepted scoping comments via mail, email, fax, factor of coal plants, the employment provided by the comment cards and online comment forms. About use of coal, and coals contribution to low and stable 85 percent of the scoping comments were submitted rates.

either as email forms or form letters promoted by two advocacy campaigns. Other comments, included:

We received a total of nearly 1,100 comments. Figure Energy Resource Options 3-1 shows the distribution of these comments by state. Most of the comments regarding potential energy resource options addressed the benefits and/or drawbacks of various energy options, including nuclear, coal, natural gas, solar and wind generation. Numerous comments encouraged increased energy efficiency 17

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation efforts while a small number of comments encouraged 3.1.4 Additional Comments increased use of other demand reduction options such After the close of the scoping period, TVA received as demand response and combined heat and power. comments related to the IRP from two advocacy Several commenters requested that TVA fully and fairly campaigns.

evaluate all potential energy resources In the spring of 2014, TVA received nearly 1,000 Environmental Impacts postcards through a Tennessee Sierra Club campaign.

of Power System Operations The message on these cards was similar to that of the Many of the comments addressed the negative and/or Sierra Club/Tennessee Environmental Council email beneficial environmental and economic impacts of the campaign during the public scoping period.

use of various energy resource options. These included air pollutants, greenhouse gas emissions and climate In the fall of 2014, TVA received about 4,500 form change, spent nuclear fuel and disposal of coal ash. emails through the takeactionTN campaign promoted Several comments also mentioned impacts resulting by the Tennessee Electric Cooperative Association from mining, particularly surface mining and the use of and Americas Electric Cooperatives. These emails coal and hydraulic fracturing to produce natural gas. advocated an all-of-the-above approach to energy Others encouraged TVA to assess the vulnerability of its generation, opposed greenhouse gas regulations power system to climate change, to assess the effects proposed by the EPA, expressed concern over reliance of climate change on TVAs power demand forecasts, on nuclear and natural gas generation and emphasized and to conduct more detailed analyses of local and low cost and reliability.

regional economic impacts, including employment.

3.2 Public Involvement in Developing IRP Process Study Inputs and Framework Several comments addressed aspects of the integrated In this step, we used the key themes and results resource planning process. Many of these supported identified from public scoping to help develop the study the use of least-cost analysis and asked TVA to be framework, including the range of strategies for IRP sensitive to the adopted plans impact on ratepayers.

analysis. During this phase, we worked closely with the Comments on planning scenarios included the IRP Working Group and held public briefings.

incorporation of the effects of climate change; varying 3.2.1 IRP Working Group Meetings approaches to incorporating regulation of greenhouse Beginning in November 2013, TVA met with the IRP gas emissions; the evaluation of future fuel prices, Working Group approximately every month. Twelve particularly for natural gas; and the impacts of current meetings were held prior to the release of the Draft IRP and anticipated environmental regulations.

and associated SEIS at various locations throughout the Comments on resource planning strategies included region. Two additional meetings were held during and maximizing renewable generation and energy efficiency, after the closure of the public comment period.

phasing out the use of fossil fuels, transmission grid The meetings were designed to encourage discussion upgrades and increasing distributed generation.

on all facets of the process and to facilitate information Other comments regarding the planning process sharing, collaboration and expectation setting for addressed the valuation of renewable energy resources, the IRP. IRP Working Group members reviewed and the removal of limits on quantities of renewable energy, commented on planning assumptions, analytical energy efficiency, demand response, and incorporating techniques and proposed energy resource options and the external health and environmental costs of all strategies. Specific topics included the energy efficiency energy resources. approach in the IRP models, TVAs power delivery 18

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation structure, load and commodity forecasts and supply Date Location resource options.

March 26, 2014 Chattanooga, Tenn.

Given the diverse makeup of the IRP Working Group, June 18, 2014 Knoxville, Tenn.

there was a wide range of views on specific issues, November 3, 2014 Knoxville, Tenn.

such as the value of energy efficiency programs, environmental concerns and the costs associated with Figure 32: Inputs and Framework Public Briefings various generation technologies. Open discussions supported by the best available data helped improve Participants had the option to attend in person or understanding of the specific issues.

participate by webinar. At each meeting, TVA staff made To increase public access to the IRP process, all non- a brief presentation, followed by a moderated Q&A confidential IRP Working Group meeting material was session.

posted on TVAs website, along with webinar recordings Topics discussed at the public briefings included and related presentation videos. We also developed an introduction to the integrated resource planning an IRP Working Group website for members to post process, resource options, development of scenarios information and to request data from our staff.

and strategies and evaluation metrics.

Date Location An average of 20 people attended each of the first three November 5, 2013 Knoxville, Tenn. public briefings in person, and approximately 50 people December 5, 2013 Knoxville, Tenn. participated via webinar. Recordings of the sessions January 13, 2014 Murfreesboro, Tenn. were posted on the IRP website.

February 19, 2014 Chattanooga, Tenn. TVA also briefed the public on the IRP process March 27 & 28, 2014 Chattanooga, Tenn. through presentations to local organizations, clubs and April 29 & 20, 2014 Knoxville, Tenn. associations.

May 29 & 30, 2014 Chattanooga, Tenn.

June 19 & 20, 2014 Knoxville, Tenn. 3.3 Public Involvement in Review of the October 7, 2014 Chattanooga, Tenn.

Draft IRP Dec 15 & 16, 2014 Knoxville, Tenn. TVA provided the draft IRP for public comment from March 13, 2015, through April 27, 2015. During this time January 26 & 27, 2015 Knoxville, Tenn.

TVA also held public meetings around the region to February 26, 2015 Webinar provide an opportunity for residents and stakeholders April 9 &10, 2015 Huntsville, AL. to learn more about the draft IRP, ask questions and June 16, 2015 Knoxville, Tenn. provide general feedback. These information and feedback opportunities are also consistent with our obligations under the National Environmental Protection 3.2.2 Public Briefings Act (NEPA).

In addition to the public scoping and IRP Working Group meetings, TVA held three briefings to update the Written comments were also accepted online, mail and public on the status of IRP development. Figure 3-2 email.

shows the dates and locations of these briefings.

19

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation 3.3.1 Public Meetings

• Many advocated earlier increased use of EE and Seven public meetings were held around the TVA region renewable energy (during first 5 years) and for a during the public comment period. larger reduction in fossil fuel generation.

• Greenhouse gas emissions and associated Date Location climate change are the most frequently mentioned March 19, 2015 Chattanooga, Tenn.* environmental concern.

April 6, 2015 Knoxville, Tenn.*

• The majority of detailed / technical comments are on modeling data inputs and assumptions.

April 9, 2015 Huntsville, AL

• Several comments point out that given the very April 14, 2015 Tupelo, Miss. small differences in most strategy metrics, TVA April 15, 2015 Memphis, Tenn. should select a strategy with high levels of EE April 21, 2015 Nashville, Tenn. and renewable energy given the differences in April 22, 2015 Bowling Green, Ky. environmental metrics and TVAs environmental stewardship mission.

• Also Webinar / Live Streamed Meeting

• Several comments emphasized the importance of power cost, reliability, and availability.

Figure 33: Draft IRP Public Meetings The approximately 125 unique comments submitted At each of these meetings, TVA presented an overview can be grouped into the following themes:

of the Draft IRP, followed by a moderated Q&A session supported by a panel of TVA subject-matter experts.

• Alignment with TVAs mission and least-cost Attendees were able to address comments or questions planning mandate to the panel. Attendees also had the option to submit

• Strategic direction and the final portfolio written comments or online comments at the meetings.

• Solar modeling Written and online comments were also accepted

• Modeling of wind options during the full public comment period in addition to at

• Treatment of energy efficiency as a resource the meetings. Approximately 400 people attended the

• Strategy evaluation metrics (scorecard) public meetings in person and on-line.

• Stakeholder engagement process Verbal and specific written comments enabled TVA Specific responses to these comments are provided staff to identify public concerns and recommendations in Volume 2 of the SEIS. In this section we have concerning the future operation of the TVA power summarized TVAs general response to each of these system.

broader themes:

3.4 Public Comment General Themes 1. Alignment with TVAs mission and least-cost planning mandate Reviewing comments received during the comment Comment Issue: The current IRP does not properly fulfill period yielded the following general observations:

the mission because it does not do least-cost planning

• There is widespread support for the overall IRP properly with regard to modeling/selection of energy development process, including the approach to efficiency. On the other hand, some methodology modeling energy efficiency and renewable energy improvements have enhanced TVAs ability to do least-as selectable resources and our extensive public cost resource planning.

involvement.

Response: The IRP study was conducted consistent

• A large majority of comments advocated increased with TVAs least-cost system planning mandate as use of EE and renewable energy.

contained in the TVA Act. The improvements we made 20

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation in our modeling methodology allowed TVA to analyze advocates in developing modeling inputs, and subjected energy efficiency and renewable energy as resources all assumptions to third party review and validation that the model could select. This treats these resources against industry references. The uncertainty ranges consistent with the way more traditional energy around the cost of solar used in the study encompass resources are modeled. In doing this, we also took into recent market trends. Sensitivity cases were also account necessary features of system operation such conducted to assess the impact of changes in these as resource diversity, reliability, resource dispatchability assumptions.

and risk factors.

4. Modeling of wind options

2. Strategic direction and the final portfolio Comment Issue: Wind was not properly modeled; Comment Issue: The IRP points in the right direction especially the assumptions around NDC (Net by demanding higher levels of alternative resources. Dependable Capacity) and capacity factors; and All strategies are too close to declare a winner so the characteristics of High Voltage Direct Current final portfolio should have higher levels of EE and (HVDC) wind were not consistent with the TVRIX renewables for broader benefits to Valley residents. recommendations. A learning curve (technology improvement) should be included.

Response: We appreciate that most people commenting on the IRP think its direction is correct. The Response: Assumptions for wind options were purpose of the IRP is to provide directional guidance to accepted from advocates and subjected to third party TVA and the TVA Board. If this guidance is approved validation. TVA elected to base modeling assumptions by the TVA Board, TVA would have the flexibility to on current technology and reasonable assumptions for increase levels of energy efficiency and renewable capacity value on peak and overall energy production.

resources. We agree that all of the strategies were Sensitivity cases were also conducted to assess the fairly close across IRP metrics, but the differences are impact of these assumptions on model results. TVA will important and should not be discounted. For example, continue to refine its modeling of wind resources.

all of the IRP strategies reduce the environmental impacts of the TVA system, but the higher levels of 5. Treating energy efficiency as a resource energy efficiency and renewables in Strategies D Comment Issue: Energy efficiency (EE) design and E have the best environmental results. Risks are parameters and risk assumptions are not well modeled.

minimized by maintaining a diverse portfolio which all of The IRP doesnt take into account all the EE benefits the strategies promote, but strategies that emphasize (line losses, avoided transmission & distribution specific resources more (Strategies D and E) have investment). The current model design makes EE higher risks. System flexibility is generally equivalent appear too costly and risky. Current assumptions across strategies, but is lower when renewable energy constrain deployment.

is emphasized. None of these differences are trivial. Response: TVA modeled EE as a selectable resource,

3. Solar modeling consistent with how other energy resources are Comment Issue: The study did not properly include traditionally modeled. This is innovative and was solar PV (rooftop) as an option. Initial cost assumptions widely endorsed by commenters. To do this, TVA had are too high based on recent market pricing. Costs for to create new techniques to capture risk and cost this option should decline at a faster rate. trends over time. While our modeling of EE for this IRP represents a significant improvement over how Response: TVA captured impacts from distributed EE traditionally is evaluated in IRP processes, we solar in the design of the scenarios; in addition, recognize that the methodology could still be improved, small commercial solar was included as an option in including accounting for the potential indirect benefits the cases. We considered assumptions from solar identified by the commenters. TVA will continue to 21

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation refine its modeling methodology. We do think that EE the public were afforded multiple opportunities to is inherently more risky than other energy resources provide feedback. Sensitive data, such as detailed because TVA does not directly serve most end users cost data, had to be handled more carefully because of the electricity it generates. Local power companies of the potential harm to TVAs business transactions.

have that relationship. To be successful, most energy However, TVA even opened up this kind of sensitive efficiency programs have to be embraced by these end data to stakeholder input and review. It was shared users and the LPC-end user interface adds another with members of the IRP Working Group on a level of uncertainty. Sensitivity cases were conducted confidential basis. TVA created sensitivity cases to to test several of the key assumptions and results are test specific parameters in order to better understand generally consistent with the original findings. In nearly the robustness of the IRP strategies. Many of these every case, additional deliveries from EE are envisioned. cases were created in response to comments that TVA received during the IRP process, particularly from

6. Strategy evaluation metrics (scorecard) members of the IRP Working Group. By necessity, Comment Issue: Cost and risk are reasonable metrics, this had to be done at the end of the planning effort.

but risk is over-estimated for EE and renewables. The Sensitivity cases were fully discussed with members system average cost metric should be eliminated. of the working group and the input we received was Flexibility metric does not provide any insight of value; considered in the preparation of the final report. The environmental metrics are not given sufficient weight in final report contains information about the sensitivity the assessment. cases.

Response: TVA developed the scorecard metrics with IRP Working Group input from stakeholders to capture key attributes of the planning cases. We believe all the metrics have value Jack Barkenbus, Senior Researcher and inform the decision around selecting a preferred Climate Change Research Network planning direction. Average system cost provides a Vanderbilt University key signal about rate pressure. The flexibility metric Nashville, TN provides an indication of general responsiveness of the system and is a first attempt at acknowledging and Lance Brown, Executive Director reporting this important metric. Stakeholders specifically Partnership for Affordable Clean Energy requested we not weight any of the scorecard metrics Montgomery, AL so that findings would not be unduly shaped by the assigned weights. Donald (Chip) Cherry, President and CEO Huntsville Madison County Chamber of Commerce

7. Stakeholder engagement process Huntsville, AL Comment Issue: Transparency and engagement efforts were good. Recommend sharing of detailed cost data Mary English, Senior Fellow with the public. Some resource modeling decisions University of Tennessee Baker Center for Public Policy made without stakeholder input. Consider providing Knoxville, TN sensitivity case results for public comment. Zachary Fabish, Staff Attorney Response: TVA has uniformly received high marks by Sierra Club commenters for the level of stakeholder engagement Washington, DC during this IRP process. Several commenters pointed Amanda Garcia, Staff Attorney out that this level of engagement is not seen in other Southern Environmental Law Center IRP processes. Stakeholder input was considered Nashville, TN in virtually all phases of the study. Stakeholders and 22

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 3: Public Participation Keith Hayward, General Manager and CEO Jack Simmons, President and CEO North East Mississippi Electric Power Association Tennessee Valley Public Power Association Oxford, MS Chattanooga, TN Richard Holland, Vice President Jay Stowe, Chief Executive Officer Tennessee Paper Council Huntsville Utilities Nashville, TN Huntsville, AL Don Huffman, Executive Director Michelle Walker, Assistant Commissioner Associated Valley Industries Office of Policy and Planning Chattanooga, TN Tennessee Department of Environment and Conservation Dana Jeanes, Vice President and Chief Financial Officer Nashville, TN Memphis Light, Gas and Water Memphis, TN Lloyd Webb, Strategic Planning Chair Tennessee Valley Industrial Committee Tom King, Director Cleveland, TN Energy Efficiency and Electricity Technologies Oak Ridge National Lab John Wilson, Research Director Oak Ridge, TN Southern Alliance for Clean Energy Washington, DC David Rumbarger, President and CEO Community Development Foundation of Tupelo Tupelo, MS Kate Shanks, Executive Director Office of Legislative and Intergovernmental Affairs, Energy and Environment Cabinet Commonwealth of Kentucky Frankfort, KY 23

Chapter 4 Need for Power Analysis Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 25

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis 4 Need for Power Analysis processes in each sector and expected changes in the stock and efficiency of equipment and appliances.

A primary purpose of this IRP is to accurately determine whether the energy resources TVA currently For example, in the residential sector, energy usage has available are sufficient to supply the power the was forecasted for space heating, air conditioning, Tennessee Valley region will need over the study period water heating and several other uses after accounting (2014-2033) and, if the anticipated demand exceeds for changes in efficiency over time, appliance saturation the current supply, to estimate the capacity gap and replacement rates, growth in average home size and determine what type and how much additional and other factors. In the commercial sector, we gave generating resources are needed. similar attention to changes in efficiency, saturation and other variables in a number of categories, including This chapter describes the four steps in the process lighting, cooling, refrigeration and space heating.

used to make this determination: compute demand, determine reserve capacity needs, estimate supply and Finally, working with TVA customer service estimate the capacity gap. representatives, we supplemented the historical data used in our modeling with industry analyses and 4.1 Estimate Demand feedback from our large, directly served customers The first step in forecasting future power needs is to regarding demand. This input helped us better estimate long-term growth in electricity sales and peak predict the magnitude and timing of changes in load demand. Peak demand, or peak load, is the highest attributable to plant closures and expansions.

one-hour power requirement placed on the system.

Key Forecast drivers In order to reliably serve customers, TVA must have sufficient resources to meet the peak hour demand.

Growth in Economic Activity The electricity sales and peak demand forecasts for this At least annually, TVA produces a forecast of regional IRP were developed from individual, detailed forecasts economic activity for budgeting and long-range of residential, commercial and industrial sales. We planning purposes. These forecasts are developed from checked the historical accuracy of these forecasts to county-level economic forecasts in order to accurately help ensure errors in data or methodology were quickly model the prevailing economic conditions in the region.

identified and resolved. We also generated a range Historically, the Tennessee Valley economy was more of forecasts (high, expected, and low) to ensure that dependent on manufacturing than the economies TVAs plans do not depend on the accuracy of a single of other regions. Industries such as pulp and paper, forecast.

aluminum, steel and chemicals were drawn to the Valley 4.1.1 Load Forecasting Methodology because of the availability of natural resources, access To forecast future electricity demand, TVA uses to a skilled workforce and the supply of reliable and statistical and mathematical models that link electricity affordable electricity. However, manufacturings share sales to several key drivers. These include the growth of non-farm employment has steadily declined in the in overall economic activity, the price of electricity, Valley, as it has across the region.

customer retention and the price of competing energy Our region is different from others in that sources such as natural gas.

manufacturings share of economic output in the We also apply an end-use forecasting model to capture Tennessee Valley actually increased since 1980, from the effect of underlying trends affecting residential, 17 percent to 18 percent. These trends speak to the commercial and industrial electricity sales such as changing nature of economic activity here. While changes in the use of various types of equipment or many labor-intensive manufacturing industries moved 26

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis overseas, a continued shift toward energy-efficient plants outside TVAs service area if TVAs rates become manufacturing processes in the Valley is helping to non-competitive. Additionally, large industrial operations preserve manufacturings contribution to total economic could generate their own power without distribution output. This is important to TVAs load forecasting or transmission line losses - an increasingly attractive in that it may indicate a weakening in the historical option to TVAs largest customers as hydraulic fracturing relationships between economic growth and load reduces the cost of natural gas to unexpected lows.

growth.

These risks are factored into TVAs load forecasts Because of this continued dependence on because they could affect future load, but we believe manufacturing, our regions economy tends to be more they will be offset by our commitment to keeping TVA sensitive to economic conditions impacting the demand rates competitive.

for manufactured goods. Near-term future growth in 2015 and 2016 is expected to benefit from positive Price of Competing Energy Sources cyclical economic conditions. After 2016, however, Changes in the price of electricity compared to the longer-term demographic pressures are expected to price of natural gas and other fuels also influences load hold average growth in Gross Regional Product near 1.6 growth.

percent as retiring baby boomers restrict the available If consumers can heat their homes and water cheaper labor supply. Population growth in the Tennessee Valley using natural gas or other energy sources, they may declined from an annual average of about 1.0 percent in move away from electricity in the long-term. The 1980 to 0.7 percent in 2014 and is expected to decline potential for this type of substitution depends on the to 0.5 percent by 2043, which will limit the demand for relative prices of other fuels and the ability of those all goods and services, including power.3 fuels to provide a comparable service. It also depends Price of Electricity on the physical capability to make the substitution. For Forecasts of the retail price for electricity are based example, while consumers can change out electric on long-term estimates of TVAs total costs to operate water heaters and replace electric heat pumps with and maintain the power system and are adjusted to natural gas furnaces, the ability to use another form include an estimate of the historical markups charged of energy to power consumer electronics, lighting by distributors of TVA power. These costs, known in and cooling is far more limited by current technology.

the industry as revenue requirements, are based on Changes in the price of TVA electricity compared to estimates of the key costs of generating and delivering the price of natural gas and other fuels also influence electricity, including fuel, variable operations and consumers choices of applianceseither electric, gas maintenance costs, capital investment and interest. or other fuels.

Customer Retention While other substitutions are possible, the price of Over the past 25 years, the electric utility industry has natural gas serves as the benchmark for the relative undergone a fundamental change in most parts of the competitiveness of electricity and the potential country. In many states an environment of regulated substitution impacts on load forecasts. Accounting for monopoly has been replaced with varying degrees the long-run impact of natural gas prices is especially of competition. Although TVA has contracts with the important in light of the increased competitive pressure 155 local power companies (LPCs), it is not immune to resulting from hydraulic fracturing and the shale gas competitive pressures. The contracts allow LPCs to give revolution. Although low gas prices make power TVA notice of contract cancellation after which they may production less expensive, it also tempts customers to buy power from other sources. Many large industrial shift from electricity to natural gas to meet their energy customers also have the option of shifting production to needs.

3 TVA population data from U.S. Census Bureau and Moodys Analytics 27

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis 4.1.2 Forecast Accuracy forecast, and the year-ahead forecast variances tend to Forecast accuracy is generally measured by how much be smaller.

the forecast deviates from the actual energy and peak demands, adjusted for abnormal weather. Figures 4-1 and 4-2 show annual forecasts for fiscal years (FY) 1999 through 2014 for peak load requirements and net system energy requirements compared to actual peak loads and actual energy use.

Figure 4-1 is a comparison of actual annual peak demands in megawatts (MW) to the peak demands forecasted one year earlier. The red Normalized Actual line represents what the annual peak would have been under normal weather conditions. The closer the blue dotted Forecast line is to the red Normalized Actual line, the more accurate the peak forecast. For example, Figure 42: Comparison of Actual and Forecasted Net System Requirements in FY14, the actual peak was only 1.3 percent greater than forecasted. Over-forecasts from FY11 to FY13 are related to the Great Recession, which resulted in a The mean absolute percent error (MAPE)4 of TVAs decline in weather-normalized peaks that continued well forecasts of net system energy and peak load after the recessions end. We are now seeing the return requirements FY 1999 through FY 2014 was 1.8 of modest growth in weather-normalized peaks. percent and 3.1 percent, respectively. While the nature of forecasting a single hours MWs inherently leads to elevated peak forecast errors, this includes an unusually large error (7.6 percent) in the FY09 peak forecast as the full severity of the Great Recessions impact was not yet fully realized.

Energy is less volatile so forecast errors tend to be a little smaller, but the Great Recession still adversely impacted the energy forecasts error rate. Ignoring the forecast for 2009 brings the energy error rate down from 1.8 to 1.4 percent, which is more in line with what we expect in a typical year. From informal conversations with peer utilities, TVAs energy forecast MAPE of around 1 to 2 percent is in line with other utilities.

Figure 41: Comparison of Actual and Forecasted Annual 4.1.3 Forecasts of Peak Load and Energy Peak Demand Requirements Over the next couple decades the Current Outlook Figure 4-2 is a comparison of actual and forecasted net Scenario5 anticipates net system energy and peak system requirements expressed in total annual energy, demand to grow about 1.0 percent and 1.1 respectively, measured in gigawatt-hours (GWh). Energy is somewhat which is somewhat slower than the 1.3 percent less volatile than peaks, which are based on a single experienced for both net system energy and peak hour of each year, because energy is the sum of all the 4 MAPE is the average absolute value of the error each year; it does not allow over-hours of the year. This makes energy a little easier to predictions and under-predictions to cancel each other out.

5 Refer to Chapter 6 for a discussion of the scenarios developed for this IRP.

28

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis demand over the FY1990 through FY2013 period. The use of ranges ensures that TVA considers a These lowered expectations from the long-term trend spectrum of electricity demand in its service territory are a function of both economics and energy efficiency and reduces the likelihood that its plans are overly projections. Slower economic growth, driven by the dependent on a single-point estimate of demand baby boomers retirement, and an ever-tightening growth. Alternative scenarios highlight the risk regulatory environment are both anticipated to inherent in forecasting and planning to a single point moderate future energy growth. estimate. The scenario-generated ranges are used to inform planning decisions beyond pure least-cost To deal with the inherent uncertainty in forecasting, considerations based on a specific demand in each TVA has developed a range of forecasts. Each forecast year.

corresponds to different load scenarios around the Current Outlook Scenarios forecast. The Current Outlook Scenario for the IRP is the forecast that TVA prepared for the FY2015 Long Range Financial Plan in fall of 2013. The range of forecasts for net system peak load and energy requirements used in the IRP are shown in Figures 4-3 and 4-4, respectively. Both include the Current Outlook Scenario and the highest and lowest growth scenarios that were modeled. They are the Growth Economy Scenario and the Distributed Marketplace Scenario, respectively. Annual peak load growth over the 2014 through 2033 time period varies from 0.3 percent in the lowest growth scenario to 1.3 percent in the highest growth scenario. Net system energy requirements grow at an annual rate Figure 44: Energy Forecast of 1.0 percent in the Current Outlook Scenario but growth dips as low as 0.0 percent in the lowest growth scenario and peaks at 1.1 percent in the highest growth 4.2 Determine Reserve Capacity Needs scenario. To maintain reliability, power providers must always have more generating capacity available than required to meet peak demand. This additional generation, called reserve capacity, must be large enough to cover the loss of the largest single operating unit (contingency reserves), be able to respond to moment-by-moment changes in system load (regulating reserves) and replace contingency resources should they fail (replacement reserves). Total reserves must also be sufficient to cover uncertainties such as unplanned unit outages, undelivered purchased capacity, severe weather events or load forecasting error.

TVA identified a planning reserve margin based on minimizing overall cost of reliability to the customer.

Figure 43: Peak Demand Forecast This reserve margin is based on a probabilistic analysis that considered the uncertainty of unit availability, 29

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis transmission capability, weather-dependent unit limited period even though it may be less economical to capabilities (e.g., hydro, wind and solar), economic do so. This may be due to transmission or other power growth and weather variations to compute expected system constraints. Similarly, some baseload units are reliability costs. Based on this analysis, we selected capable of operating at different power levels, giving a target reserve margin that minimized the cost of them some characteristics of an intermediate or peaking additional reserves plus the cost of reliability events to unit. This IRP considered strategies that take advantage the customer. The target or optimal reserve margin was of this range of operations.

adjusted based on TVAs risk tolerance. Based on this methodology, TVAs current planning reserve margin is 15 percent above peak load requirements and is applied during both the summer and winter seasons.

4.3 Estimate Supply The third step in the process of analyzing future power needs is to identify the supply- and demand-side resources currently available to meet future power demand. Our generation supply consists of a combination of existing TVA-owned resources; budgeted and approved projects such as new plant additions, updates to existing assets; and existing power purchase agreements (PPAs).

Generating assets can be categorized both by whether the power they produce is used to meet base, intermediate or peak demand or used for storage, and Figure 45: Illustration of Baseload, Intermediate and Peak-by capacity type or energy/fuel source. ing Resources 4.3.1.1 Baseload Resources 4.3.1 Baseload, Intermediate, Peaking and Due to their lower operating costs and high availability, Storage Resources baseload resources are used primarily to provide continuous, reliable power over long periods of uniform Figure 4-5 illustrates the uses of baseload, intermediate demand. Baseload resources typically have higher and peaking assets. Although these categories are construction costs than other alternatives, but also have useful, the distinction Power purchase agreements much lower fuel and variable costs, especially when between them is not (PPAs) refer to the energy fixed costs are expressed on a unit basis (e.g., dollars always clear-cut. For and/or capacity bought from per MWh). An example of a baseload resource is a example, a peaking other suppliers for use on the nuclear power plant.

unit, which is typically TVA system in place of TVA used to serve only Some energy providers also use natural gas-fired building and operating its own intermittent but combined cycle (CC) plants as incremental baseload resources. Power purchases short-lived spikes in generators. However, given the historical tendency provide additional diversity for demand, may be for natural gas prices to be higher than coal and TVAs portfolio. We are currently called on from time to nuclear fuel prices when expressed on a unit basis a party to numerous short-term time to run (i.e., dollar per million British Thermal Unit), a CC unit is and long-term PPAs. continuously for a generally considered a more expensive option for larger 30

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis continuous generation needs. Although natural gas-fired function as peaking units but use low-cost, off-peak CC plants could become more attractive for baseload electricity to store energy for generation at peak generation as the fundamentals of fuel supply and times. An example of a storage unit is a hydroelectric demand change and if access to shale gas continues to pumped-storage plant. These plants pump water to a grow. reservoir during periods of low demand and release it to generate electricity during periods of high demand.

4.3.1.2 Intermediate Resources Consequently, a storage unit is both a power supply Intermediate resources are used primarily to fill the gap source and an electricity user.

in generation between baseload and peaking needs.

They also provide back-up and balance the supply of 4.3.2 Capacity and Energy energy from intermittent wind and solar generation. Power system peaks are measured in terms of capacity, the instantaneous maximum amount of energy that can Intermediate units are required to produce more or be supplied by a generating plant and collectively by the less output as the energy demand increases and power system.

decreases over time, both during the course of a day and seasonally. Given current fuel prices and relative For long term planning purposes, capacity can be generating efficiencies, these units are more costly to defined in several ways:

operate than baseload units but cheaper than peaking units.

• Nameplate capacity is the theoretical design value or intended maximum megawatt output of a Intermediate generation typically comes from natural generator at the time of installation.

gas-fired CC plants and smaller coal units. However,

• Capability is the maximum dependable load-energy from wind and solar generation also can be carrying ability of units or the number of megawatts used as intermediate resources depending on the that can be delivered by a generating unit without energy production profile and the availability of energy restrictions (i.e., does not reflect temporary capacity storage technologies. restrictions caused by known fuel or mechanical derates) and less station power.

Hydro generating assets can generally be categorized

• Net dependable capacity is the maximum as intermediate resources, but their flexibility allows dependable output less all known adjustments (e.g.,

them to operate the full range from baseload to transmission restrictions, station service needs peaking. The limitation of hydro generation is restricted and fuel derates) and is dependent on the season.

more by water availability and the various needs of the This value, which is used by capacity planners, is river system such as navigation. typically determined by performance testing during the respective season. TVA uses the summer net 4.3.1.3 Peaking Resources dependable capacity of a unit because the capacity Peaking units are expected to operate infrequently of thermal generating units is reduced during the during short duration, high demand periods. They are heat of summer which is when the load on the TVA essential for maintaining system reliability requirements, system typically peaks.

as they can ramp up quickly to meet sudden changes in either demand or supply. Typical peaking resources are natural gas-fired combustion turbines (CTs), Overall power system production is measured in terms conventional hydroelectric generation and pumped- of energy (i.e., megawatt-hour). Energy is the total storage hydroelectric generation. amount of power that an asset delivers in a specified time frame. For example, one MW of power delivered for 4.3.1.4 Storage Resources one hour equals one megawatt-hour (MWh) of energy.

Storage units usually serve the same power supply The capacity factor of a power plant is a measure of 31

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis the actual energy delivered by a generator compared and renewables6. This resource plan will be used in to the maximum amount it could have produced at the the associated EIS report to represent the no action nameplate capacity. Assets that run constantly, such as alternative as required under NEPA.

nuclear plants, provide a significant amount of energy with capacity factors greater than 90 percent. Figure 4-6 includes both owned and purchased resources, in megawatts of summer net dependable Capacity Factor Examples Assets that capacity, and is divided into fuel-type (i.e., nuclear, are used hydro, coal, etc.). In this chart, the lower area in High capacity factor unit: infrequently, the figure contains the existing assets and the new A 1,200 MW unit could theoretically such as a expansion assets are shown at the upper portion of produce 10,512 GWh of power if it ran combustion the stacked area chart. EE is shown in the upper area every hour of the year. After planned turbine, of this chart to make comparison to the results from annual outages, the unit will typically provide the modeling discussed in Chapter 7 easier. There are produce 9,461 GWh or 90 percent of its relatively little existing contracts for demand response that need to theoretical capacity. energy with be accounted for and are shown in the existing assets.

capacity Similarly, the current renewable resources are shown in Low capacity factor unit: factors of the bottom of the chart.

less than A 250 MW natural gas-fired combustion five percent, Figure 4-6 shows how TVAs existing capacity portfolio turbine (CT) unit could theoretically although the is expected to change through 2033. The existing produce 2,190 GWh of power if it ran energy they assets only include resources that currently exist; assets every hour of the year. However, CT units produce is that are under contract; TVA Board-approved changes generally have a capacity factor less than crucial since to existing resources such as refurbishment projects; 5%, which means the unit would likely it is often and TVA Board-approved additions such as Watts operate about 438 hours0.00507 days <br />0.122 hours <br />7.242063e-4 weeks <br />1.66659e-4 months <br /> of the year and delivered at Bar Unit 2. Existing resources decrease through 2033 produce 110 GWh. peak times. primarily because of the retirement of coal-fired units and the expiration of existing contracts (power purchase Energy efficiency also can be measured in terms of agreements). The renewable component of the capacity and energy. Even though energy efficiency existing portfolio is primarily composed of wind PPAs.

does not input power into the system, the effect is Because the power generated from wind and other similar because it represents power that is not required renewable resources is intermittent, the firm capacity from another resource. Demand reduction also is (or the amount of capacity that can be applied to firm measured in capacity and energy. However, unlike requirements) for these assets is much lower than the energy efficiency, it does not offer a significant reduction nameplate capacity.

in total energy used.

Having a diverse portfolio of resource types - coal, 4.3.3 Current TVA Capacity and Energy nuclear, hydroelectric, natural gas and renewable Supply resources - and being able to use these resources in TVA uses a wide range of technologies to meet the different ways enables TVA to provide reliable, low-cost needs of Tennessee Valley residents, businesses and power while minimizing the risk of disproportionate industries. Figure 4-6 shows the generating assets that reliance on any one type of resource.

would be used to meet those needs over time for the baseline case, which reflects our current practice of optimizing the resource portfolio while scheduling the contribution of energy efficiency, demand response, 6 This IRP uses a new methodology to dynamically select these resources. See discussion in Chapter 6.

32

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis MW,SND Capacity for the Baseline Case Figure 4-7 shows TVAs estimated capacity gap or 45,000 shortfall based on the existing firm capacity and the 40,000 New DR annual firm requirement for the current outlook scenario.

New EE 35,000 30,000 New Gas CC New Gas CT Capacity Gap New Hydro 25,000 New Coal MW, SND New Nuclear 20,000 45,000 Existing DR Firm Requirement = Peak Demand + 15% Reserves - Interruptibles 15,000 Existing Gas CC 40,000 Existing Gas CT/Diesel 10,000 Existing Renewables 35,000 Capacity Gap Existing Coal 5,000 30,000 Existing Hydro Existing Nuclear 25,000 0

2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 20,000 Existing Firm Suply Figure 46: Baseline Capacity, Summer Net Dependable MW 15,000 10,000 5,000 Approximately 37 percent of TVAs capacity is currently -

2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 sourced from emission-free assets such as nuclear power, renewable resources including hydroelectricity, Figure 47: Estimating the Capacity Gap and interruptible load management. The renewable category shown throughout this document is based on modeled outputs of energy from renewable sources Figure 4-8 shows the range of capacity gaps such as wind, solar, and biomass. Therefore, this metric corresponding to the highest growth scenario, the is not intended to represent a quantity of certified Economic Growth scenario, and the lowest growth renewable energy credits. scenario, the Distributed Marketplace scenario. These and other scenarios are described in detail in Chapter 6.

Currently, 33 percent of TVAs electricity is produced from the nuclear fleet. Coal produces about 40 Capacity Gap Range percent of the generation, hydroelectric plants produce approximately 10 percent, and 3 percent was produced from renewable sources. The gas fleet produces about 13 percent with the majority of that generation being provided by combined cycle plants and the remaining generation results from energy efficiency efforts.

4.4 Calculate the Capacity Gap The need for power can be expressed either as a capacity gap or as an energy gap.

As noted previously, a capacity gap is the difference between total supply and total demand. More Figure 48: Capacity Gap Range specifically, it is the difference in megawatts between a power providers existing firm capacity and the forecast annual peak adjusted for any interruptible customer An energy gap is the amount of energy specified in loads plus 15 percent reserve requirements. GWh provided by the existing firm capacity resources minus the energy required to meet net system 33

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 4: Need for Power Analysis requirements (i.e., the energy needed to serve the load over the entire year). It includes the energy consumed by the end-users plus distribution and transmission losses.

Figure 4-9 shows the range of energy gaps TVA can expect under the net system requirements associated with the highest and lowest growth scenarios.

Energy Gap Range Figure 49: Energy Gap 34

Chapter 5 Energy Resource Options Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 35

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options 5 Energy Resource Options the agency is called upon to be a leader in Technology Innovation. Currently, we have identified three signature Maintaining the diversity of TVAs energy resources is technologies for special emphasis: small modular fundamental to our ability to provide low-cost, reliable reactors, grid modernization and energy utilization and clean electric power to Valley residents, businesses focusing on efficiency and demand response products.

and industries. For this reason, we considered the In support of our broader technology innovation efforts, addition of a wide range of supply-side generating we have included small modular reactors as an option resources, as well as energy efficiency and other in this IRP; and the technology advancements around demand-side resource options, to fill the forecasted energy utilization are also included in this study as part 20-year capacity and energy gaps identified through the of our modeling of EE and DR as selectable resource power needs analysis described in Chapter 4.

options.

The power needs analysis indicates that, under the Current Outlook scenario, TVA will require additional 5.1.2 Characteristics Required for Resource capacity and energy of 2,100 MW and almost 16,000 Options GWh by 2020, growing to 10,300 MW and more than To compare energy resource options available for new 58,000 GWh by 2033. generation fairly, it is important to have consistent data regarding the cost and operating characteristics of each 5.1 Energy Resource Selection Criteria option. A list of characteristics used in the 2015 IRP are identified and defined below. Section 5.2.2 will present During the scoping process, TVA identified a broad the numerical values for some of these parameters for range of energy resources that could be used to fill the new assets.

the predicted capacity and energy gaps. The next two sections explain the criteria that were used to reduce Cost characteristics:

this list to a manageable portfolio of expansion options.

• Unit capital costs: Each technology type must For a complete list of resource options considered, see have a representative $/kW, which is considered Chapter 5, Energy Resource Options, of the associated a total installed cost. Total installed cost includes EIS. equipment, engineering and interest during construction in present day dollars.

• Capital escalation rates: Since capital costs typically 5.1.1 Criteria for Considering Resource increase over time, a simplifying assumption could Options be that the capital costs escalate at the forecast Two criteria were used to ensure that only viable energy rate of inflation. However, some renewable energy resource options were considered in the IRP analysis.

technologies are forecast to decrease over time.

To be considered, resource options must:

• Construction spend schedule: Some technologies Use a proven technology, or one that has reasonable take a long time to build. Construction times for prospects of becoming commercially available in the nuclear units, for example, average about 10 years.

planning horizon To estimate the cash flow for the construction of a long-lead time build unit such as a nuclear unit, the Be available to TVA within the region or be available to percent of total capital dollars spent in each year be imported through market purchases is required. This metric is typically not needed for renewable assets which are smaller in scale and Technology is a key factor in TVAs ability to fulfill its generally built in less than a year.

mission in a balanced way. TVA continues to pursue

• Fixed operating and maintenance costs (FOM): FOM technological advances to become more efficient and costs are independent of the number of hours of sustainable. As part of its mission under the TVA Act, operation or amount of electricity produced and are 36

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options generally expressed in a dollar per kilowatt per year commercial availability, as well as permitting and

($/kW-yr). FOM includes operating and maintenance construction times. For example, if it takes four labor, plant support equipment, administrative years to build a combined cycle plant, then a new expenses and fees required by regulatory bodies. CC could not be selected prior to four years into the

• Variable operating and maintenance costs (VOM): planning horizon.

VOM costs are dependent on the number of hours

• Book life: The book life of a unit is the number of of operation and are generally expressed as a dollar years a resource is expected to be in service for per megawatt-hour ($/MWh). VOM costs include accounting purposes. Book life is the financial consumables like raw water, waste and water payback period which represents the amount of disposal expenses, chemicals and reagents. VOM time the asset is expected to be used and useful. A costs do not include fuel expenses. license extension, beyond the original asset life, is

• Fuel expenses: Fuel is the material that is consumed not assumed with any new generating option.

to generate electricity - for example, coal, natural gas, uranium and biomass. These costs are typically 5.2 Resource Options Included in IRP expressed in a dollar per million British thermal units Evaluation

($/mmBtu) and include the delivery charges.

• Transmission: A new generating resource has to TVAs existing assets, including existing TVA-owned be connected to the transmission system. Costs resources, as well as budgeted and approved projects, are typically expressed as a dollar per kilowatt ($/ and power purchase agreements, are considered fixed kilowatt). assets in the IRP evaluation. These assets are expected to continue operating through the duration of the Operating characteristics: planning period or through the terms of existing power

• Summer net dependable capacity: Each unit must purchase agreements and other contracts, where have a summer net dependable capacity rating in applicable.

megawatts.

Options for new generation to meet the forecast net

• Capacity credit: The capacity credit must be system requirements identified in Chapter 4 include:

estimated for variable units or non-dispatchable building new generating units, retro-fitting existing units resources. The capacity credit is the amount with controls to continue operations, development of of capacity immediately available at the highest energy efficiency and demand response programs, and demand times.

new power purchase agreements.

• Summer full load heat rate: A heat rate must be specified for each unit. A heat rate is a measure The next two sections describe existing and of the consumption of fuel necessary for a unit to potential new generation by resource category. For produce electricity. Heat rates are shown in British a comprehensive description of all resource option thermal units per kilowatt hour (Btu/kWh) and are attributes, characteristics and technologies, see based on a summer full-load heat rate. Heat rates Chapter 5, Energy Resource Options, of the associated are considered long-term planning assumptions and EIS.

include the expected degradation in the heat rate of a unit after the first two years. Although a heat rate is not typically associated with a nuclear unit, one is 5.2.1 Existing Assets by Resource Category necessary to model the fuel costs.

• Unit availability: A date when each unit would be Nuclear available for operation must be specified. Unit TVA currently operates six nuclear reactors: three at availability is restricted by technical feasibility or Browns Ferry Nuclear Plant, two at Sequoyah Nuclear 37

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Plant and one at Watts Bar Nuclear Plant. These plants a generating system to carry power for a specified time have a combined generating capacity of about 6,700 and does not include operational limitations such as fuel MW. On August 1, 2007, the TVA Board of Directors de-rates. We use a value lower than the capability of a approved the completion of a second reactor at Watts resource for the summer net dependable capacity. By Bar Nuclear Plant. This reactor will have a 1,150 MW 2016, the existing coal fleet will decrease to about 35 generating capacity. The new reactor is scheduled to active units with a total capability of 10,300 MW. Below become operational by the end of 2015 and is included is a snapshot of the planning assumptions for the coal as a current resource in TVAs generating portfolio. units.

Coal In addition to TVA-owned coaled fired units, TVA has TVA operates 10 coal-fired power plants consisting access to the output from a coal-fired power plant with of 41 active generating units with a total capability of a generating capacity of about 440 MW through a long-almost 11,900 MW. Capability is defined as the ability of term power purchase agreement.

Figure 51: Coal Fleet Map 38

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Total # of Coal Plant Current Operating Status Operational Plan Original Units Allen 3 Operational Retire all three units. Board approved plan to construct a 2x1 combined cycle plant adjacent to the site Bull Run 1 Operational Continue to operate Colbert 5 Unit 5 idled Board approved plan to retire all five Units 1-4 operational units Cumberland 2 Operational Continue to operate Gallatin 4 Operational Continue to operate with Board approved scrubbers and SCRs Johnsonville 10 Units 1-4 operational Retire all units Units 5-10 idled Kingston 9 Operational Continue to operate Paradise 3 Operational Board approved plans to construct a combined cycle plant on site, retire units 1 and 2, and continue operation of unit 3 Shawnee 10 Units 1-9 operational Board approved plans to control Unit 10 retired units 1 and 4. The remaining units will continue to operate until a long-term decision is made Widows Creek 8 Units 1-6 retired Board approved plan to retire all units Unit 8 idled Unit 7 operational Figure 52: Coal Fleet Portfolio Plans Natural Gas Hydroelectricity TVA operates 87 combustion turbines (CT) at nine TVA operates 109 conventional hydroelectric generating power plants with a combined generating capability of units at 29 dams. These units have the capability to about 5,400 MW and 11 combined cycle (CC) units at generate about 3,800 MW of electricity.

five plants with approximately 3,900 MW of capability.

TVA is also currently a party to a long-term lease of a In addition, TVA has a long-term power purchase 700 MW CC plant. agreement with the U.S. Army Corps of Engineers for eight dams on the Cumberland River system. These Petroleum Fuels facilities provide almost 400 MW of capability.

TVA currently owns five diesel generators and has a few other diesel generators under power purchase TVA anticipates about 60 percent of the capability to contracts. These resources provide a total capability of be available at the summer peak hour given all the about 120 MW. operational constraints.

39

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Energy Storage Energy Efficiency TVA operates one large energy storage facility. Our TVAs energy efficiency portfolio focuses on reduction Raccoon Mountain Pumped-Storage Plant has in peak demand and energy savings. From FY2007-four generating units with a SND capacity of 1,616 FY2014 these efforts contributed 535 MW of summer megawatts. Raccoon Mountain is TVAs largest peak demand reduction and save 1,506 GWh of energy hydroelectric facility and provides critical flexibility to the annually. Impacts are realized at the generator and TVA system by storing water at off-peak times for use include applicable transmission and distribution (T&D) when demand is high. losses, free rider/driver discounts, realization rates, and performance adjustments for actual weather. This Wind differs from other sources, such as the ERS Highlights TVA purchases all of the power produced by the Buffalo Report, which do not normally include these factors; Mountain wind farm in Anderson County, Tenn. Buffalo and are more reflective of end-user savings.

Mountain is the largest wind farm in the Southeast, with 18 turbines and 27 MW of nameplate capacity. Demand Response As defined in section 4.3.2, the nameplate capacity is Demand response programs also focus on reduction the maximum technical output of a generator, or the of peak demand. Under these programs, TVA industrial theoretical design value. and commercial customers can reduce their power bills by allowing TVA to suspend availability of power We also have long-term power purchase contracts in the event of a power system emergency. These with eight wind farms located in Illinois, Kansas and programs provide about 600 MWs of peak reduction.

Iowa. These facilities provide about 1,500 MW of Another program allows TVA to curtail power delivery nameplate capacity. TVA anticipates about 14 percent to participants for economic or reliability reasons. This of the nameplate to be available for peak summer program provides about 560 MWs of peak reduction. If requirements. TVA obtains the renewable energy credits needed, TVA also can reduce peak demand by about from seven of these farms. Renewable energy credits 85 MWs through in-house curtailments.

are a separate commodity formed from the production of energy at designated sites.

5.2.2 New Assets by Resource Category Solar A complete list of viable new resource options for IRP TVA owns 16 photovoltaic (PV) installations with a evaluation is provided below. A detailed discussion by combined capacity of about 300 kilowatts of nameplate resource category follows.

capacity. We also purchase solar power through several An independent third-party reviewed and compared programs and long-term power contracts totaling the parameters to proprietary and other industry nearly 72 MW of nameplate capacity with about 36 MW sources to ensure the modeled unit characteristics expected to be available at the summer peak hour.

and assumptions were representative of the respective Biomass generating technologies. See Appendix A for the letter TVA generates electricity at Allen Fossil Plant by co- summary of the benchmarking efforts of Navigant firing methane from a nearby sewage treatment plant Consulting, Inc. as well as a brief discussion of TVAs and by co-firing wood waste at Colbert Fossil Plant. internal benchmarking on resource costs ($/kW).

The co-firing is more like a fuel switch for coal and does not provide additional capacity to either of the coal plants. TVA purchases about 49 MW of biomass-fueled generation.

40

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Nuclear Utility-scale Storage

• Pressurized water reactor (PWR)

• Pumped-hydro storage

• Advanced pressurized water reactor (APWR)

• Compressed air energy storage (CAES)

• Small modular reactor (SMR) Wind Coal fired

• Midcontinent Independent System Operator

• Integrated gas combined cycle (IGCC) (MISO)

• Supercritical pulverized coal 1x8 (SCPC1x8)

• Southwest Power Pool (SPP)

• Supercritical pulverized coal 2x8 (SCPC2x8)

• In Valley

• Integrated gas combined cycle with carbon

• High voltage direct current (HVDC) capture and sequestration (IGCC CCS) Solar

• Supercritical pulverized coal 1x8 with carbon

• Utility-scale one-axis tracking photovoltaic capture and sequestration (SCPC1x8 CCS)

• Utility-scale fixed-axis photovoltaic

• Supercritical pulverized coal 2x8 with carbon

• Commercial-scale large photovoltaic capture and sequestration (SCPC2x8 CCS)

• Commercial-scale small photovoltaic Natural Gas fired Biomass

• Simple cycle combustion turbine 3x (CT 3x)

• New direct combustion

• Simple cycle combustion turbine 4x (CT 4x)

• Repowering

• Combined cycle two on one (CC 2 by 1)

• Combined cycle three on one (CC 3 by 1) Energy Efficiency (EE)

• Residential EE Hydro

• Commercial EE

• Hydro expansion project where spill permits

• Industrial EE

• Hydro expansion project where space permits

• Small-head or low-head (run of river) hydro project Demand Response Figure 53: List of New Assets Nuclear See Chapter 4, Section 4.3.2, for a discussion of the There are three nuclear expansion options available different types of capacity ratings.

to fill the expected capacity gap: a Pressurized Water Reactor (PWR), an Advanced Pressurized Water TVA could increase the electrical output of the three Reactor (APWR) and a Small Modular Reactor (SMR). Browns Ferry Nuclear units. This project could provide The PWR option is based on completion of the approximately 400 MWs of additional capacity and is Bellefonte brownfield site. The APWR and SMR options termed an extended power uprate (EPU). Figure 5-4 are not site specific. provides an example of the characteristics for one of these projects. The book life is based on the remaining Figure 5-4 shows some of operating characteristics life of the plant.

used to model each option. Summer net dependable capacity, summer full load heat rate, unit availability A nuclear PPA is also assumed to be available for model and book life are explained above. The annual outage selection. PPAs are available for selection based on rate percentage includes forced and planned outages. competitive information which cannot be disclosed. PPA 41

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options options are evaluated similar to build options with a few in section 5.1.2) versus the PPA options which have slight differences. One difference is when present value levelized cash flow payments based on the terms of the revenue requirements resulting from the expansion contract (similar to a mortgage). The other difference model selections are converted into cash flows, the for PPAs is if the asset is located outside of the TVA build options have significant capital expenditures transmission area then the necessary transmission that match the construction spend schedule (noted wheeling charges are included.

PWR APWR SMR* EPU 1 Unit Characteristics Summer Net Dependable Capacity (MW) 1,260 1,117 334 134 Summer Full Load Heat Rate (Btu/kWh) 9,853 9,715 10,046 9,558 Unit Availability (Yr) 2026 2026 2026 2018 Annual Outage Rate (%) 10% 10% 10% 10%

Book Life (Yrs) 40 40 40 29

• The SMR option is based on a twin pack, the minimum viable configuration.

Figure 54: Nuclear Expansion Options 42

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Coal BTUs of coal burned. The modeled CO2 emissions incur The 2015 IRP includes six coal expansion options, an emission penalty in the form of a dollar per ton of including two integrated gas combined cycle (IGCC) CO2 emitted.

options and four supercritical pulverized coal (SCPC) options. Two of the four SCPC options have one steam generator with a supercritical steam cycle. One of IGCC technology converts coal into gas. One IGCC these options includes CCS technology; the other does option has carbon capture and sequestration (CCS) and not. The other two SCPC options have two steam one does not. The CCS technology option is assumed generators with supercritical steam cycles. Again, one to be commercially available starting in 2028 and has of these options includes CCS technology, and one a 90 percent carbon dioxide (CO2) capture rate. Coal does not.

units typically have a CO2 emission rate of 205 pounds per million BTUs of coal burned so the CCS technology Three options to continue to operate the Shawnee coal would reduce the CO2 rate to 20.5 pounds per million plant with the addition of more environmental controls (on various units) were available for model selection.

SCPC SCPC IGCC* SCPC SCPC IGCC 1x8 2x8 CCS 1x8 2x8 CCS CCS Unit Characteristics Summer Net Dependable Capacity (MW) 500 469 800 1,600 600 1,200 Summer Full Load Heat Rate (Btu/kWh) 8,000 10,000 8,674 8,674 10,843 10,843 Unit Availability (Yr) 2022 2028 2025 2025 2028 2028 Annual Outage Rate (%) 17% 18% 10% 10% 11% 11%

Book Life (Yrs) 40 40 40 40 40 40

• The CCS technology is assumed to have a 25% penalty on a 625 MW IGCC plant.

Figure 55: Coal Expansion Options 43

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Natural Gas All options are based on a generic location. The CO2 The IRP evaluation includes two simple cycle emission rate for a typical gas unit is 117 pounds of combustion turbine (CT) options and two combined CO2 per million Btus of gas burned. The modeled gas cycle (CC) natural gas fueled options. The simple cycle units incur emission charges based on a dollar-per-ton CTs are available with either three or four turbines. The emission penalty.

CC options have either two turbines and one steam generator (CC 2 by 1) or three turbines and one steam In addition, the IRP evaluation includes options for generator (CC 3 by 1). CC units have supplemental purchasing power from existing merchant gas plants, capacity termed duct-firing capacity that adds acquiring merchant gas plants, and options in which approximately 100 MW to the base capacity shown. TVA would build additional gas-fueled units.

CT 3X CT 4X CC 2 by 1 CC 3 by 1 Unit Characteristics Summer Net Dependable Capacity (MW) 590 786 670 1,005 Summer Full Load Heat Rate (Btu/kWh) 10,132 10,132 6,946 6,598 Unit Availability (Yr) 2018 2018 2019 2019 Annual Outage Rate (%) 4% 4% 7% 7%

Book Life (Yrs) 30 30 30 30 Figure 56: Gas Expansion Options Petroleum Fuels TVA expects to phase out petroleum power purchases by 2028. There are no diesel fuels or other petroleum based resource options as a primary fuel source under consideration in the IRP because of emissions from these facilities.

44

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Hydroelectric River system also are required for municipal and Two new hydro projects are included in the IRP industrial uses, navigation, flood damage reduction, evaluation, developed in collaboration with the TVRIX recreation, water quality and other purposes. For this stakeholders. They include adding additional hydro reason, a fixed amount of monthly energy (provided turbines to existing dam facilities where there is space by River Operations) is entered into the model for the available with structural modifications. The other would conventional hydro stations. The model then uses the add turbines at existing dam facilities where water that hydro energy to level the load shape served by other is now spilled could be used to power more turbines. stations.

Both projects are similar to the larger TVA hydro Since hydro plants do not use fuel, a heat rate is not system and are energy-limited units. Energy-limited needed for modeling.

units are resources that cannot be dispatched (in the model) based on price ($/MWh) as are traditional Small- and low-head hydropower, called run of river, thermal generating resources, such as nuclear, coal also is included as an IRP resource option.

and gas. Hydropower cannot be dispatched based on A hydro PPA was also included in the IRP evaluation.

price alone because water releases in the Tennessee Dam Spill Dam Space Run of Addition Addition River Unit Characteristics Summer Net Dependable Capacity (MW) 40 30 25 Unit Availability (Yr) 2019 2018 2021 Annual Outage Rate (%) - - 4%

Book Life (Yrs) 40 40 40 Figure 57: Hydro Expansion Options 45

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Energy Storage underground cavern where it can be stored under The IRP evaluation includes a new hydroelectric pressure until electricity is required. The pressurized pumped-storage unit as a resource option. The air is then heated and directed through a conventional pumped-storage option would use three reversible generator to produce electricity.

turbine generators to either take electricity from the grid by pumping water into a higher altitude reservoir during Storage efficiency is included in modeling both these periods of excess power or add electricity to the grid energy storage options because of the energy losses by using the pumped water to power a turbine as it falls inherent to the energy conversion process and due from the upper to the lower reservoir. to the loss of water or air during storage. The storage efficiency percentage for these energy storage options A compressed air energy storage (CAES) option also represents the efficiency of one cycle (i.e., pumping is included as an energy storage option. A CAES plant water, then releasing).

is similar to a pumped-storage plant but, instead of pumping water from a lower to an upper reservoir, TVA did not evaluate any electric battery storage a gas turbine is used to compress air often into an options because of operational limitations.

Pump CAES Storage Unit Characteristics Summer Net Dependable Capacity (MW) 850 330 Summer Full Load Heat Rate (Btu/kWh) - 4,196 Unit Availability (Yr) 2023 2019 Annual Outage Rate (%) 7% 10%

Storage Efficiency (%) 81% 70%

Book Life (Yrs) 40 40 Figure 58: Utility-Scale Storage Options 46

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Wind The HVDC option would require a third-party to permit Because TVA cannot take direct advantage of the and build a new transmission line so the unit availability tax credits and other investment incentives offered is later than the other options. All unit availability dates by the federal government to encourage wind power were rounded to the next full year.

development, it has been more financially advantageous to acquire wind power resources through PPAs. This Wind resources are energy- and capacity-limited approach allows us to include wind as a resource resources. For this reason, we use an energy option in the IRP. The purchase of wind resources as production profile to dispatch wind energy rather a PPA, whether produced in or imported to the TVA than price. The method used for wind resources is region, lowers the costs of these resources to TVA and somewhat similar to hydro resources except that an its customers. TVA may evaluate the option of building hourly generation schedule (not a monthly amount) wind facilities in the future if investment incentives and/ is pre-loaded into the capacity expansion model. We or future federal or state renewable mandates change. also apply a capacity credit since the total nameplate capacity of a wind turbine cannot be expected at the Four wind options are included in the IRP evaluation, time of the system peak. To determine the capacity and the characteristics of these options were credit, we used historical data to estimate the typical developed with input from the TVRIX stakeholders. The wind power output at the time of the peak power Midcontinent Independent System Operator (MISO), the demand on the TVA system. This resulted in a 14 Southwest Power Pool (SPP) and the In Valley options percent capacity credit, meaning that 14 percent of represent various wind resources in different regional nameplate capacity is expected to be available at the transmission areas. The High Voltage Direct Current system peak. This reduced capacity is considered the (HVDC) option would use a direct current (DC) bulk summer net dependable capacity. Appendix B includes transmission system. The HVDC transmission system a more detailed discussion about the determination would reduce power losses that are typical of the more of the data assumptions for the modeling of the wind common alternating current (AC) transmission systems. options included in this IRP.

MISO SPP In valley HVDC Unit Characteristics Nameplate Capacity (MW) 200 200 120 250 Summer Net Dependable Capacity (MW) 28 28 17 35 Unit Availability (Yr) 2016 2016 2017 2020 Annual Outage Rate - - - -

Book Life (Yrs) 20 20 20 20 Figure 59: Wind Expansion Options 47

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Solar solar installations. The large and small scale commercial Similar to new wind generation, because TVA cannot options represent solar installations at different price take advantage of the current investment incentives points and with different generating characteristics.

offered to promote solar power development, it is more financially advantageous to acquire solar power Like wind resources, solar resources are energy-resources through PPAs. We may evaluate the option limited and therefore dispatched in the model using an of building solar facilities in the future if investment hourly energy production profile to ensure that solar incentives and/or federal or state renewable mandates generation is not utilized by the model when the sun change. is not available. Solar resources also are similar to the capacity-limited wind resources where the availability Four solar options, developed with input from the TVRIX of the unit at the time of the TVA system peak is less stakeholders, are included in the IRP evaluation at a than the full nameplate capacity. We applied a 68 minimum capacity block size of 25 MW nameplate percent capacity credit for the utility tracking unit and a capacity. All capacities are stated in alternating current 50 percent capacity credit for the fixed asset options.

(AC) terms. The unit availability date was rounded to the first full year. More details about the assumptions used in the The utility tracking option is a single-axis tracker that development of the unit characteristics for these solar allows the solar panels to follow the sun. The utility fixed options can be found in Appendix B.

option represents ground mounted fixed-axis/fixed-tilt Utility Utility Commercial Commercial tracking fixed small large Unit Characteristics Nameplate Capacity (MW) 25 25 25 25 Summer Net Dependable Capacity (MW) 18 13 13 13 Unit Availability (Yr) 2015 2015 2015 2015 Annual Outage Rate - - - -

Book Life (Yrs) 25 25 25 25 Figure 510: Solar Expansion Options 48

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Biomass the assumptions for hydro, wind, and solar, these Two new biomass options are included in the IRP options were also developed with input from the TVRIX evaluation: a new direct combustion biomass facility stakeholders. Because biomass co-firing is considered and a repower option, which is the conversion of a fuel switch opportunity, it was not included as a existing coal-fired units to biomass-fired units. Like capacity expansion option.

Direct Repower Combustion Unit Characteristics Summer Net Dependable Capacity (MW) 115 75 Summer Full Load Heat Rate (Btu/kWh) 13,500 12,243 Unit Availability (Yr) 2019 2015 Annual Outage Rate 5% 5%

Book Life (Yrs) 30 20 Figure 511: Biomass Expansion Options Demand Response similar to those used for natural gas combustion Demand response programs enable participating turbines (CT). Demand response programs are operated customers to reduce their power costs by allowing much like CTs, or peaker units, and focus on reduction TVA to limit their power during peak demand times. of peak demand. However, the terms of the demand Using a new innovative approach, these programs were response customer contracts are shorter than the modeled in the 2015 IRP based on unit characteristics expected book life of a CT unit.

Demand

Response

Unit Characteristics Summer Net Dependable Capacity (MW) 1 Summer Full Load Heat Rate (Btu/kWh) 10,132 Unit Availability (Yr) 2014 Annual Outage Rate -

Book Life (Yrs) 5 Figure 512: DR Expansion Options 49

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 5: Energy Resource Options Energy Efficiency EE generating units were developed to represent the The 2015 IRP reflects TVAs increased focus on energy residential (Res), commercial (Com) and industrial (Ind) efficiency (EE). A new, innovative modeling approach sectors. Then each sector was divided into three tiers, was used in this IRP to evaluate EE as a supply-side representing three distinct price points, for a total of resource, with characteristics and costs structured nine units. All of the tier 1 units are available beginning in similarly to conventional generating resources or power 2014, but the first year tier 2 and 3 units will be available plants. This allowed various EE generating units to varies by sector. These units are energy limited, similar be optimized against the other resource options. More to hydro, wind and solar units, and use annual hourly details about this modeling approach can be found in production profiles.

Appendix D.

Res Res Res Com Com Com Ind Ind Ind Tier Tier Tier Tier Tier Tier Tier Tier Tier 1 2 3 1 2 3 1 2 3 Unit Characteristics Nameplate Capacity (MW) 10 10 10 10 10 10 10 10 10 Summer Full Load Heat Rate (Btu/kWh)

Unit Availability (Yr) 2014 2022 2026 2014 2019 2022 2014 2018 2022 Annual Outage Rate - - - - - - - - -

Book Life (Yrs) 17 13 13 15 13 13 12 10 10 Figure 513: EE Expansion Options 50

Chapter 6 Resource Plan Development and Analysis Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 51

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis 6 Resource Plan Development and make (called planning strategies) in different sets of uncertain future conditions (called scenarios). The set Analysis of resource choices selected in any one future defines This chapter describes the process TVA used to how we would provide power to our customers under identify a target power supply mix that was based those conditions; we call that set of resource choices on the analysis done in the IRP. The process involves a portfolio, and it is created by modeling a planning choosing the types of resources that we could use strategy in a particular scenario. These portfolios are to meet the future power needs of our customers, then scored using some key factors (called metrics) that recognizing that the future is uncertain and our choices allow us to capture cost, risk, environmental footprint need to give us flexibility to adapt. So the approach and other aspects that should be considered when tests several options around resource choices we could deciding on the best target power supply mix.

Resource Identify Incorporation Identify Develop Develop Portfolio Preferred of Public Input Recommended START Scenarios and Evaluation Optimization Target Power and Additional Target Power END Strategies Scorecard Modeling Supply Mixes Modeling Supply Mix Process for identifying the recommended target power supply mix 6.1 Development of Scenarios and and a single scenario results in a resource portfolio7.

Strategies A portfolio is a 20-year capacity expansion plan that is unique to that strategy and scenario combination.

TVA uses a scenario planning approach in integrated resource planning, a common approach in the utility 6.1.1 Development of Scenarios industry. Scenario planning is useful for determining While most quantitative models used in long range how various business decisions will perform in an planning focus on what is statistically likely based on uncertain future. The goal is to develop a least-cost history, market data and projected future patterns, TVA strategy that is consistent with TVAs legislatively uses scenario analysis that allows for the possibility that mandated mission and also delivers our customers rate the future could evolve along paths not suggested solely stability over a variety of future environments. by historical trends.

Multiple strategies, which represent business decisions The scenarios used in the IRP analysis were developed that TVA can control, are modeled against multiple during the scoping phase of the study in 2013. The scenarios, which represent uncertain futures that TVA process used to develop these scenarios is described cannot control. The intersection of a single strategy below.

7 Portfolios are also referred to as capacity expansion plans or resource portfolios 52

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Identification of key uncertainties uncertainties. These uncertainties, shown in Figure 6-1, The first step in developing scenarios was to work with were used as building blocks to construct scenarios.

the individuals on the IRP Working Group to identify key Uncertainty Description TVA sales The load to be served by TVA Natural gas prices The price of natural gas ($/MMBtu), including transportation Wholesale electricity prices The hourly price of energy ($/MWh) at the TVA boundary (used as a proxy for TVA for market price of power)

Coal prices The price of coal ($/MMBtu), including transportation All regulatory and legislative actions, including applicable codes and Regulations standards, that impact the operation of electric utilities, excluding CO2 regulations The cost of compliance with possible CO2 related regulation and/or the price CO2 regulation/price of cap-and-trade legislation, represented as a $/Ton value National trending of distributed generation resources and potential regional Distributed Generation activity by customers or third-party developers (not TVA) See Appendix C for details on the method used to incorporate the effects of DG in the scenarios.

National Energy Efficiency An estimate of the willingness of customers nationally to adopt EE measures, (EE) adoption recognizing the impacts of both technology affordability and electricity price All aspects of the regional and national economy including general inflation, Economic outlook (national financing considerations, population growth, GDP and other economic and regional) drivers Figure 61: Key Uncertainties 53

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Construction of Scenarios

• Placed sufficient stress on the resource selection Scenarios were constructed using combinations of the process and provided a foundation for analyzing key uncertainties shown in Figure 6-1 and then refined the robustness, flexibility and adaptability of each to ensure that each scenario: combination of supply- and demand-side options

• Captured relevant key stakeholder interests.

• Represented a plausible, meaningful future in which TVA could find itself within the 20-year study period Figure 6-2 shows the key characteristics of the

• Was unique among the scenarios being considered scenarios selected for the IRP analysis.

for study Scenarios Key Characteristics The outlook for the future which TVA is currently using for resource planning 1 - Current Outlook studies Stagnant economy results in flat to negative growth, delaying the need for 2 - Stagnant Economy new generation Rapid economic growth translates into higher than forecasted energy sales 3 - Growth Economy and resource expansion Increasing climate-driven effects create strong federal push to curb greenhouse gas emissions; new legislation caps and penalizes CO2 4 - De-Carbonized Future emissions from the utility industry and incentivizes non-emitting technologies Customers awareness of growing competitive energy markets and the rapid advance in energy technologies produce unexpected high penetration 5 - Distributed rates in distributed generation and energy efficiency. TVA assumes Marketplace responsibility to serve the net customer load (no backup for any customer-owned resources)

Figure 62: Scenario Key Characteristics 54

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Determination of key scenario assumptions The Current Outlook scenario projects growth of The final step in scenario development was to forecast approximately 1.0 percent per year. Three scenarios key assumptions for each scenario. - Stagnant Economy, De-Carbonized Future, and Distributed Marketplace - project lower load growth Figure 6-3 shows the forecasted assumptions for than the Current Outlook scenario, while the Growth energy demand/load growth for each scenario. Economy scenario models a modest growth scenario.

Figure 63: Energy Demand Assumptions 55

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 6-4 shows the forecasted assumptions for gas scenarios, while both the Growth Economy and De-prices. Gas prices are similar for the Current Outlook, Carbonized Future scenarios assume a substantial Stagnant Economy and Distributed Marketplace increase in gas prices later this decade.

Figure 64: Gas Price Assumptions 56

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 6-5 shows the forecasted assumptions for coal Future scenario has the lowest price through the prices. Steadily increasing coal prices are forecasted planning period.

for all scenarios. Starting in 2019, the De-Carbonized Figure 65: Coal Price Assumptions 57

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 6-6 shows the forecasted assumptions for does not start until 2029. The Current Outlook and CO2 prices. All scenarios forecast a more stringent Distributed Marketplace scenarios share the same CO2 regulatory future. The highest CO2 prices are seen in price assumptions. Note that the CO2 cost curve for the the De-carbonized Future scenario. The CO2 penalty Distributed Marketplace is the same as the assumptions in the Stagnant Economy scenario is the lowest and used in the Current Outlook.

Figure 66: CO2 Price Assumptions8 8 The cost curve for the Current Outlook and Distributed Marketplace are identical.

58

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis 6.1.2 Development of Planning Strategies lowest possible financial cost. TVA has developed After the scenarios were developed, the next step in strategies in both categories for this IRP. The process the IRP process was to design planning strategies. used to develop planning strategies is described below.

Scenarios and strategies are very different. Whereas Identification of key strategy components scenarios describe plausible futures and include factors The first step in developing planning that TVA cannot control, strategies describe business Planning strategies was to identify the key decisions over which TVA has full control. strategies components, or attributes, to be represent Generally, strategies fall into two categories: included in each strategy. Ten distinct decisions and approaches that are intended to achieve a particular attributes were identified using input choices over goal, but dont restrict the energy resources that can from individuals on the IRP Working which TVA has be used to achieve that goal and approaches that Group and comments received control.

constrain how resources are used. In IRP modeling during the public scoping period.

terms, strategies that constrain resources are not fully optimized and may not produce plans that have the 59

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Attributes Description Constraints related to TVAs existing nuclear fleet, including Extended Existing nuclear Power Uprates (EPUs)

Limitations on technologies and timing related to the addition of new Nuclear additions nuclear capacity, including Watts Bar Unit 2, small modular reactors (SMRs), A/P 1000s and completion of TVAs Bellefonte Nuclear Plant Constraints related to TVAs existing coal fleet, including the current Existing coal schedule for idling coal units Limitations on technology and timing on new coal-fired plants, including New coal Carbon Capture & Sequestration (CCS) and Integrated Gasification Combined Cycle (IGCC) technologies Limitations on technologies and timing related to the expansion options Gas additions fueled by natural gas (CT, CC)

Energy Efficiency and Considers energy efficiency and demand response programs that are Demand Response incentivized by TVA and/or local power companies, excluding impacts (EEDR) from naturally occurring efficiency/ conservation Limitations on technologies and timing of renewable resources, including Renewables options that could be pursued by TVA or in collaboration with local power (utility scale) companies Purchased Power Level of market reliance allowed in each strategy; no limitation on the Agreements (PPAs) type of energy source (conventional or renewable)

Distributed Includes customer-driven resource options or third-party projects that Generation/Distributed are distributive in nature Energy Resources Type and level of transmission infrastructure required to support resource Transmission options in each strategy Figure 67: Key Planning Strategy Attributes 60

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Development of Strategies Using Attributes TVA combined these 10 components to create five distinct planning strategies for the IRP analysis. Figure 6-8 lists the five strategies and their key characteristics.

Strategies Key Characteristics A - Traditional Utility Planning Least cost optimization; EE/renewables selectable (Reference Plan)

Resources selected to create lower emitting portfolio based on an B - Meet an Emission Target emission rate target or level using CO2 as the emissions metric Most new capacity needs met using longer-term PPA or other C - Focus on Long-Term, Market-bilateral arrangements; TVA makes a minimal investment in owned Supplied Resources assets Majority of capacity needs are met by setting an annual energy D - Maximize Energy Efficiency target for EE (priority resource to fill the energy gap); other resources (EE) selected to serve remaining need Enforce near-term and long-term renewable energy targets; meet E - Maximize Renewables targets with lowest cost combination of renewables; hydro is included as a renewable option along with biomass, wind and solar Figure 68: Planning Strategies Key Characteristics Strategies A-C are strategies that achieve specific are getting more emphasis across the industry and outcomes without setting any constraints or targets we wanted to be able to answer questions about the on the resource mix required to achieve the outcome. benefits of increased EE and renewables in the mix. So By contrast, strategies D and E are approaches that these special purpose strategies will help inform TVAs specify the way the resource mix will be constrained understanding about how the system would perform if or how certain resource types must be prioritized in priority were given to either EE or renewable resources producing the resulting plan. Because strategies D to close the capacity gap.

and E are not fully optimized, they do not result in a plan that necessarily has the lowest financial cost. The Definition of Strategies latter two strategies were developed in collaboration After defining each strategys key characteristics, with our stakeholders because these resource types specific descriptions were developed for its strategy attribute as shown in Figure 6-9.

61

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Stategy C Focus on Strategy A The Strategy B Meet An Strategy D Maximize Strategy E Maximize STRATEGY ATTRIBUTES Long-Term, Market-ReferencePlan Emissions Target Energy Efficiency Renewables Supplied Resources Operate existing units Operate existing units Operate existing units Operate existing units Operate existing units Existing Nuclear through end of period through end of period through end of period through end of period through end of period New nuclear is available for New nuclear is available for Nuclear Additions selection selection Only nuclear PPAs allowed. No new nuclear No new nuclear Based on current fleet Based on current fleet Based on current fleet Based on current fleet Based on current fleet strategy; 1) All coal units strategy; 1) All coal units strategy; 1) All coal units strategy; 1) All coal units strategy; 1) All coal units Existing Coal can be selected for can be selected for can be selected for can be selected for can be selected for retirement 2) SHF controls retirement 2) SHF controls retirement 2) SHF controls retirement 2) SHF controls retirement 2) SHF controls available for selection available for selection available for selection available for selection available for selection New Coal New coal allowed with CCS New coal allowed with CCS PPA is allowed No additions No additions Gas Additions Expansion option allowed Expansion option allowed PPA is allowed Expansion option allowed Expansion option allowed EE and DR available for EE and DR available for EE and DR available for EE required to meet all EE and DR available for EEDR resource selection resource selection resource selection future energy needs first resource selection Aggressive renewable energy target enforced Expansion under current Expansion under current Expansion under current Expansion under current to promote growth in Renewables (Utility Scale) programs and new options programs and new options programs and new options programs and new options renewable resources first, available for selection available for selection available for selection available for selection through current programs or new options Expansion options Expansion options New energy storage not Expansion options Expansion options New Energy Storage selectable selectable allowed selectable selectable Expension allowed; Expension allowed; Expension allowed; Expension allowed; Expension allowed;

1) PPA available 2) Capacity 1) PPA available 2) Capacity 1) PPA available 2) Capacity 1) PPA available 2) Capacity 1) PPA available 2) Capacity Hydro projects to existing assests projects to existing assests projects to existing assests projects to existing assests projects to existing assests available available available available available Figure 69: Strategy Descriptions Strategy attributes were used in the modeling in 6.2 Resource Portfolio Optimization several different ways. For example, Strategy A has Modeling specific defined constraints such as new coal additions The generation of resource portfolios was a two-step only with carbon capture and sequestration. Other process. First, an optimized portfolio, or capacity components specified timing, such as allowing nuclear expansion plan, was generated, followed by a detailed additions to be started after 2022 in Strategies A and B.

financial analysis. This process was repeated for each strategy/scenario combination and for additional sensitivity runs.

62

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis 6.2.1 Development of Optimized Capacity calculated detailed production costs of each plan Expansion Plan including fuel and other variable operating costs. These TVA uses a capacity optimization model called System detailed cost simulations provided total strategy costs Optimizer.9 This model employs an optimization and financial metrics that were used in the strategy technique where an objective function (e.g., total assessment process.

resource plan cost) is minimized subject to a number of This analysis was accomplished using a strategic constraints.

planning software tool called MIDAS.10 MIDAS uses a Energy resources were selected by adding or chronological production costing algorithm with financial subtracting assets based on minimizing the present planning data to assess plan cost, system rate impacts value of revenue requirements (PVRR). PVRR and financial risk. It also uses a variant of Monte Carlo represents the cumulative present value of total revenue analysis,11 which is a sophisticated analytical technique requirements for the study period based on an 8 that allows for risk analysis by varying important drivers percent discount rate. In other words, PVRR is the in multiple runs to create a distribution of total costs present day value of all future costs for the study period, rather than a single point estimate.

discounted to reflect the time value of money and other The total cost for each resource plan (PVRR) was factors such as investment risk.

calculated taking into account additional considerations, In addition, the following constraints were applied in the including the cash flows associated with financing.

optimization runs: The model generated multiple combinations of the key assumptions for each year of the study period and

• Balance of supply and demand computed the costs of each combination. Capital costs

• Energy balance for supply-side options were amortized for investment

• Reserve margin recovery using a real economic carrying cost method

• Generation and transmission operating limits that accounted for unequal useful lives of generating

• Fuel purchase and utilization limits assets.

• Environmental stewardship In addition to computation of the total plan cost (PVRR)

The System Optimizer model uses a simplified over the full 20-year study period, a 10-year system dispatch algorithm to compute production costs and average cost metric was calculated. This metric a representative hours approach in which average provides an alternative view of the revenue requirements generation and load values in each representative for the 2014-2023 timeframe expressed per MWh.

period within a week are scaled up appropriately to It is not intended as a forecast of wholesale or retail span all hours of the week and days of the months. rates over the study period. Rather, it was developed to gauge the potential rate impact associated with a Year-to-year changes in the resource mix are then given portfolio and provides an indication of relative evaluated and infeasible states are eliminated. The least rate pressure across the strategies being studied. A cost path (based on lowest PVRR) from all possible second system average cost metric covering the period states in the study period is used in the IRP as the 2024-2033 also was computed. Reviewing these two optimized capacity expansion plan. metrics in combination with PVRR and the financial risk measures provides a clearer picture of the cost/risk 6.2.2 Evaluation of Detailed Financial balance for each resource plan.

Analysis Next, each capacity expansion plan was evaluated using an hourly production costing algorithm, which 10 MIDAS is also a Ventyx product.

9 System Optimizer is an industry standard software model developed by Ventyx. 11 Monte Carlo analysis is also referred to as stochastic analysis 63

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis 6.2.3 Uncertainty (Risk) Analysis Cost and risk metrics shown later in this report are Stochastic analysis of production cost and financials computed based on the expected values produced bound the uncertainty and identify the risk exposure from these stochastic iterations. The Midas tool allows that is inherent in long-range power supply planning, TVA to explicitly consider uncertainty and risk exposure because the fundamental forecasts used in those in the evaluation of the planning strategies. This analysis studies are inevitably wrong. Variability will result is based on applying probability distributions around the due to supply/demand disruptions, weather, market key variables used to frame the scenarios and define conditions, technology improvements, and economic assumptions used in the strategies. The Monte Carlo cycles. A Monte Carlo simulation allows for a better analysis in MIDAS includes 13 key variables:

understanding of the richness of possible futures, as

• Commodity prices: natural gas, coal, oil, CO2 well as their likelihoods so that plans can be made allowances, electricity price12 proactively as opposed to reactively. A stochastic model

• Financial parameters: interest rates, capital costs, is used to estimate probability distributions of potential Operation and Maintenance (O&M) costs outcomes by allowing for simultaneous random-walking

• Availability: hydro, fossil and nuclear variation in many inputs over time.

• Load forecast uncertainty: demand and At TVA, a representative Monte-Carlo distribution load-shape-year comprised of 72 stochastic iterations is developed for

• Planning parameters: reserve margin target each of the scenario/strategy combinations to more fully assess the likely plan costs for each portfolio. A sample The fundamental (expected value) forecasts for these stochastic result is shown in Figure 6-10:

key variables differ across the five scenarios, and so the uncertainty ranges (stochastic envelope) are also Example Stochastic Results different. So the evaluation of the uncertainty around the performance of the strategies considers both the variation across the scenarios (different plausible futures), as well as capturing the probability distribution around the expected forecasts represented by the stochastic envelope. As an example, Figure 6-11 shows these different uncertainty ranges around the TVA peak 5th Expected 95th load forecast.

Value Figure 610: Sample Stochastic Result 12 Stochastic electricity price was derived in MIDAS using stochastic variables as inputs 64

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 611: Example Uncertainty Ranges The figure shows the range of variation in the expected the EE resource, we consider two primary sources of forecast of peak demand across all five scenarios uncertainty: Design and Delivery Uncertainty. Design (represented by the gray shaded area); for orientation, uncertainty exists for the following reasons:

the Current Outlook scenarios fundamental forecast and its associated uncertainty range is shown in the

• Blocks are proxies for programs not yet developed black solid and dotted lines. The stochastic envelope, some of which represent as-yet undeveloped representing the uncertainty ranges from all five technologies scenarios, is shown as the blue dotted line and bounds

• Blocks are a blend of measures with different the uncertainty range evaluated in Midas. Each of lifespans and each with a different underlying load the 13 key variables has a set of scenario ranges and shape stochastic envelopes that ensure a more dynamic assessment of the variability in the performance of each Delivery uncertainty is driven by several factors:

planning strategy.

• The fact that TVA does not own the relationship with In addition to the uncertainty analysis based on most end-use customers in the valley the Monte Carlo modeling, in this IRP study we

• Experience in other jurisdictions around non-are including energy efficiency (EE) as a selectable performance (realization rate) for both energy and resource. TVA made this decision to allow full portfolio demand optimization to clearly demonstrate value proposition

• Uncertainty around the impact of future codes and and to allow flexible, nimble response to changing standards on program design and deliveries (are EE business environments. Uncertainty exists with all program deliveries as certain in 2033 as they are in resource types and is modeled in different ways. For 2015?)

65

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis A more complete discussion of the treatment of uncertainty for energy efficiency can be found in Strategic Imperatives Appendix D. The uncertainty around the EE resource, Maintain low rates combined with the Monte Carlo modeling of uncertainty, results in a robust evaluation of the planning strategies Rates and allows TVA to more confidently apply metrics using a scorecard framework as a way to assess overall performance. PEOPLE Asset Stewardship Portfolio PERFORMANCE 6.3 Portfolio Analysis and Scorecard Meet reliability EXCELLENCE Be responsible Development expectations & provide stewards a balanced portfolio Modeling multiple strategies within multiple scenarios resulted in a large number of portfolios. So, initially, our portfolio analysis focused on common characteristics Debt that strategies exhibited over multiple scenarios rather Live within our means than on specific outcomes in individual portfolios.

Strategies that behaved in a similar manner in most scenarios were considered to be robust - i.e., more Figure 612: Strategic Imperatives flexible, less risky over the long-term and able to lessen the impacts of uncertainty. Conversely, strategies that Optimizing TVAs asset portfolio is the primary purpose behaved differently or poorly in most scenarios were of integrated resource planning, but other imperatives considered more risky with a higher probability for also shape the process:

future regret.

• As part of the financial analysis, a balance sheet The first step in the portfolio evaluation process was and income statement are created for each portfolio to develop a scorecard to assess and compare the to capture the rate revenues required to fund each performance of planning strategies in each scenario. resource plan.

The process used to develop an evaluation scorecard is

• A coverage ratio method is used to ensure that the described below. overall debt limit is respected in each optimization run.

6.3.1 Selection of Metric Categories

• Stewardship obligations are considered in modeling TVAs mission and stakeholder concerns related of various compliance requirements, including to resource planning were key considerations in portfolio optimization which factors in a carbon developing a set of metrics for use in evaluating the penalty and includes key environmental metrics in performance of the portfolios generated in the IRP. the assessment of each resource plan (air, water and solid waste impacts).

To achieve our overall mission of providing low cost, reliable power to the people of the Tennessee Valley, TVA focuses on four strategic imperatives: balancing As part of the public involvement process, stakeholders rates and debt so that we maintain low power rates assigned priority to key concerns regarding the while living within our means; and recognizing the trade- development of a long-range power supply plan, off between optimizing the value of our asset portfolio and priority concerns were used in identifying metric and being responsible stewards of the Tennessee categories.

Valleys environment and natural resources.

66

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Based on TVAs strategic imperatives and feedback 6.3.2 Development of Scoring and from stakeholders, five metrics categories were selected Reporting Metrics for use in evaluating the performance of planning After establishing the metrics categories, the next step strategies: was to identify candidate metrics for each category.

These metrics can be grouped into two broad

• Cost, including both the long-range cost of the categories:

resource plan (present value of customer costs) as well as a look at a shorter term average system cost

• Scoring metrics to be used in the scorecard to (an indicator of possible rate pressure) assess the performance of each strategy in different

• Financial Risk, which measures the variation scenarios (uncertainty) around the cost of the resource plan

• Reporting metrics to be included in the IRP report by assessing a risk/benefit ratio and computing as supplemental information for purposes of the likely amount of cost at risk using data from explanation and clarification.

probability modeling

• Environmental Stewardship, which captures After considering the computational requirements and multiple measures related to the environmental likely predictive value of multiple candidate metrics, as footprint of the resource plans such as air emissions well as whether stakeholder groups would understand and water or waste impacts the purpose of each metric, TVA selected nine scoring

• Valley Economics, which computes the macro- metrics summarized in Figure 6-13.

economic effects of the resource plans by measuring the change in per capita income compared to a reference case

• Flexibility, which measures how responsive the generation portfolio of each resource plan is by evaluating the type/quantity of resources and the extent to which this mix can easily follow load swings.

67

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Scoring Metric Definition The total plan cost (capital and operating) expressed as the present value of revenue requirements over the 20-year study period (generated from the 20-year expected value PVRR stochastic analysis, or the expected value of the probability distribution of plan costs)

Average system cost ($/MWh), Average system cost for the first 10 years of the study, computed as the Year 1-10 levelized annual average system cost (revenue requirements in each year divided by sales in that year)

Area under the plan cost distribution curve between P(95)and expected value Risk/benefit ratio divided by the area between expected value and P(5)

The point on the plan cost distribution below which the likely plan costs will fall Risk exposure 95% of the time based on stochastic analysis CO2 annual average tons The annual average tons of CO2 emitted over the study period Water consumption The annual average gallons of water consumed over the study period The annual average quantity of coal ash, sludge and slag projected based on Waste energy production in each portfolio The annual system regulating capacity expressed as a percentage of peak Flexibility load; measures the ability of the system to respond to load swings The change in per capita personal income expressed as a change from a

% change in per capita income reference portfolio in each scenario Figure 613: Scoring Metrics 68

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 6-14 shows the formulas used to compute these scoring metrics.

Category Scoring Metric Formula PVRR ($Bn) = Present Value of Revenue Requirements over Planning Horizon Cost System Average Cost NPV Rev Reqs (2014-2023)

Years 1-10 =

($/MWh) NPV Sales (2014-2023) 95th (PVRR) - Expected (PVRR)

Risk/Benefit Ratio =

Expected (PVRR) - 5th (PVRR)

Risk Risk Exposure

= 95th Percentile (PVRR)

($Bn)

CO2 Average Annual Tons of CO2 Emitted

=

(MMTons) During Planning Period Environmental Water Consumption

= Average Annual Gallons of Water Consumed Stewardship (Million Gallons) During Planning Period Waste = Average Annual Tons of Coal Ash and Scrubber Residue (MMTons) During Planning Period System Regulating (Regulating Reserve + Demand Response + Quick Start)

Flexibility =

Capability Peak Load

= Difference in the Change in Per Capita Personal Income Valley Economics Per Capita Income Compared to Reference Case (for each scenario)

Figure 614: Scoring Metric Formulas 69

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis In addition to the nine scoring metrics, seven reporting metrics were chosen:

Reporting Metric Definition Average system cost for the second 10 years of the study, computed as the Average system cost ($/MWh),

levelized annual average system cost (revenue requirements in each year divided Year 11-20 by sales in that year)

The predicted variation in plan cost from the stochastic analysis, determined by Cost uncertainty using the difference between the tails of the distribution; the range in which plan costs will fall 90% of the time A measure of risk that the plan cost will exceed the expected value. This metric Risk ratio is developed by computing the ratio of the upper (higher cost) section of the cost distribution (between P(95) and the expected value) divided by the expected value The CO emissions expressed as an emission intensity; computed by dividing CO intensity 2 2 emissions by energy generated A measure of the quantity of spent nuclear fuel that is projected to be generated Spent Nuclear Fuel Index based on energy production in each portfolio Two measures were selected in this category: the variable energy resource penetration, which measures the amount of variable or intermittent energy Flexibility included in the plans; and a flexibility turn-down factor to measure the ability of the system to serve low load periods Employment The change in employment expressed relative to a baseline future Figure 615: Reporting Metrics 70

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis Figure 6-16 shows the formulas used to compute these scoring metrics.

Category Scoring Metric Formula System Average Cost NPV Rev Reqs (2024-2033)

Cost Years 11-20 =

($/MWh) NPV Sales (2024-2033)

Cost Uncertainty = 95th (PVRR) - 5th (PVRR)

Risk 95th (PVRR) - Expected (PVRR)

Risk Ratio =

Expected (PVRR)

CO2 Intensity Tons CO2 (2014-2033)

(Tons/GWh)

=

GWh Generated (2014-2033)

Environmental Stewardship Spent Nuclear Fuel Index Expected Spent Fuel Generated

=

(Tons) During Planning Period Variable Energy (Variable Resource Capacity) (2033)

=

Resource Penetration Peak Load (2033)

Flexibility Flexibility Turn Down Must run + Non-Dispatachable (Wind/Solar/Nuclear) (2033)

=

Factor Sales (2033)

Difference in the Change in Employment Compared Valley Economics Employment =

to Reference Strategy Figure 616: Reporting Metric Formulas 71

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis The scorecard metrics developed in collaboration with the IRP Working Group align with TVAs mission as shown in Figure 6-17.

TVA Mission Low-Cost Economic Environmental Technological River IRP Scorecard Metrics Reliable Power Development Stewardship Innovation Management Present Value of Revenue Requirements System Avg. Cost Risk/Benefit Ration Risk Exposure CO2 Emissions Water Usage Waste Flexibility Impact to Per Capita Income Figure 617: Scorecard Alignment 6.3.3 Scorecard design of the scenarios. Figure 6-18 shows the scorecard Once the scoring metrics were selected, the strategy template, which includes nine columns (one for each of scorecard could be designed. Using a format similar the scoring metrics, grouped by metric category) and to the 2011 IRP, the scorecard summarizes the five rows (one for each of the scenarios).

performance of an individual planning strategy in each Figure 618: Scorecard Template 72

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 6: Resource Plan Development and Analysis The scorecard serves as a summary tabulation of the performance of the planning strategy in each scenario.

To evaluate differences within a given scenario, all five scorecards should be reviewed. Interpretation of the performance of each strategy will be presented in Chapter 7.

6.4 Strategy Assessment Process Finally, scorecards were filled in based on an assessment of overall performance of each planning strategy in the five metric categories: cost, financial risk, stewardship, Valley economics and flexibility.

Each metric category was assessed individually based on the simple average of the strategys performance in each scenario (assumes each scenario was equally likely), and graphics were developed to facilitate interpretation of trends and to identify preliminary observations. These observations will guide the development of an action plan for further case analysis.

A cost/risk graphic was also prepared to enable an investigation of possible cost and risk trade-offs.

The strategy assessment graphics, along with information about observations from the IRP study and the action plan, can be found in Chapter 8.

73

Chapter 7 Study Results Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 75

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results 7 Study Results 7.1 Analysis Results This chapter describes the findings of the 2015 IRP. 7.1.1 Firm Requirements and Capacity Gap The results for 25 distinct portfolios are presented in The key components of each scenario were translated this chapter along with the scorecard measures as into a forecast of firm requirements (demand plus discussed in Chapter 6. reserves), which was used to identify the resulting capacity gap and need for power. This drove the selection of resources in the capacity planning model.

Figure 7-1 illustrates the firm requirements forecasts for the five scenarios studied in the IRP.

MW, SND Firm Requirements 45,000 40,000 35,000 30,000 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace Figure 71: Firm Requirements by Scenario Firm requirements were greatest in the Growth The shape of the firm requirement curves influenced the Economy scenario (highest load growth) and lowest in type and timing of resource additions in the strategies.

the Distributed Marketplace scenario (flat load growth The timing of additional resources was a function of until 2024). The remaining scenarios fell within this range the existing system capacity and the impact of the and generally displayed smooth but unique growth attributes used to define each strategy. Figure 7-2 trends, with the exception of the De-Carbonized Future shows the range of the capacity gaps across the cases.13 scenario; the discontinuity exhibited in that scenario is the result of the abrupt application of an aggressive CO2 penalty. 13 Strategy assumptions are discussed in Section 6.1 76

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results MW, SND Capacity Gap 15,000 10,000 5,000 0

(5,000) 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace Figure 72: Range of Capacity Gaps by Scenario 7.1.2 Expansion Plans grouped together with the scenarios on the horizontal The capacity expansion plans are presented below by axis. For example, the first bar on the left of the chart is strategy. Further information on the capacity expansion the incremental capacity results from the reference plan plans are presented in Appendix E - Expansion Plan under the Current Outlook scenario. The incremental Listing. capacity additions are grouped by resource type (i.e.,

nuclear, hydro, coal, etc.).

Figure 7-3 presents the incremental capacity additions for all 25 cases by 2033. The incremental capacity is The De-Carbonized Future and the distributed the selected results that fill the capacity gap referenced marketplace scenarios have the lowest demand above. The vertical axis is in summer net dependable forecasts and therefore have the least amount of (SND) megawatts, the capacity that can be applied to incremental capacity. Conversely, the Growth Economy firm requirements. The results for each strategy are had the highest demand and therefore results in the most incremental capacity.

77

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Incremental Capacity by 2033 MW, SND Strategy A: Strategy B: Strategy C: Strategy D: Strategy E:

Reference Plan Meet an Emission Target Market Supplied Resources Maximize EE Maximize Renewables 14,000 12,000 10,000 DR 8,000 EE 6,000 Gas CC Gas CT 4,000 Renewables Coal 2,000 Hydro Nuclear 0

Figure 7-3: Incremental Capacity Additions for All 25 Cases Capacity resource highlights are summarized below by Economic Growth scenario (high growth scenario). This resource type: includes utility and commercial scale renewables but does not include small distributed renewable assets.

Nuclear: The extended power uprate (EPU) capacity The assumptions on distributed renewables were expansion projects were selected in every case considered in the load demand projections for each providing approximately 400 MW. No new nuclear was scenario (see further discussion in Appendix C).

selected beyond the scheduled Watts Bar Unit 2.

Natural Gas: The addition of natural gas units vary Hydro: Two of the smaller hydro capacity projects were more significantly than other resources and depend on selected in 24 of the cases. An additional hydro asset the forecasted load in each scenario and the strategic is typically selected across the Growth Economy and focus. The maximum amount of additional CT capacity the de-carbonized future scenario, and also across the is approximately 5,000 MW in the high load world of maximize renewable strategy cases. the Growth Economy scenario. The lowest amount of additional CT capacity is about 200 MW in the Coal: No new coal plants were selected. In a few Distributed Marketplace scenario. The incremental Gas cases, additional coal units were retired beyond those CC capacity additions are similar across strategies A, currently planned.

B, D, and E with the Board-approved plans at Allen Renewables: Figure 7-3 shows the non-hydro and Paradise and grow over time with the extension of renewable assets (i.e., wind and solar) in summer net existing contracts, the acquisition of contracted assets, dependable megawatts which is the amount of firm or for new assets of about 800 MWs. In strategy C, a capacity that can be expected at the system peak. gas CC unit is replaced by a combination of controlled The renewable additions range from ~1,000 MW to coal and more renewables and EEDR.

~13,700 MW on a nameplate capacity basis. The EE: The amount of energy efficiency added in lowest selection of renewable assets occurs in the strategies A, B, C, and E is fairly consistent averaging Distributed Marketplace scenario (low growth scenario).

approximately 2,700 MW by 2033. The consistent The highest selection of renewables occurs in the 78

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results selection of energy efficiency is attributable to the solar assets are selected in the mid-2020 timeframe low price compared to other assets and the energy and wind assets are selected in the late 2030 time contributions from the energy efficiency blocks. The period. However, in the De-Carbonized Future scenario, one exception in this group is the Maximize Renewables wind assets are selected as early as 2020. The natural strategy/Distributed Marketplace scenario case which gas assets increase over time, with the first addition has a lower selection of energy efficiency at 1,900 occurring as early as 2020 in the Economic Growth MW by 2033 due to the combination of low load scenario and as late as 2032 in the De-Carbonized assumptions and the strategic focus on renewables. Future scenario. The TVA Board-approved Paradise The amount of EE in all of the strategy D/Maximize EE and Allen gas plants increase the gas portfolio by 2017 cases is also consistent at ~4,600 MW by 2033. and 2019 respectively, then the percentage decreases over time as existing third-party contracts expire. In DR: The incremental demand response averages out many scenarios existing contracts for gas combined about 460 MW across all 25 cases with a range of cycle resources are renewed or the underlying asset is almost 270 MW to 575 MW. acquired. Energy efficiency increases in all scenarios decreasing the need for new gas resources. Demand response maintains a consistent portion of the Summary by Strategy: capacity portfolio throughout the scenarios.

Strategy A: The reference plan is TVAs least-cost Figure 7-5 shows the energy portfolio which optimization plan and applies no special constraints or corresponds to the capacity charts in Figure 7-4.

targets. Nuclear energy increases over time due to the addition of Watts Bar Unit 2 and the extended power uprates.

Figure 7-4 presents the modeled capacity results for Hydro energy remains fairly constant. Coal generation the reference plan. The capacity portfolios show the decreases over the planning horizon as units are retired.

summer net dependable megawatts in 2033. The The renewable generation remains fairly constant in nuclear portfolio increases across all scenarios with the low demand scenarios (Stagnant Economy and the addition of Watts Bar Unit 2 and the extended Distributed Marketplace) and increases in the other power uprate projects. The hydro capacity increases three scenarios. Natural gas generation varies with slightly with the selection of projects that provide some load and strategic focus. Demand response, which additional capacity in all the cases. In the Growth produces low energy volumes, has been combined with Economy and the De-Carbonized Future scenarios, the energy efficiency into one group termed EEDR. The an additional hydro market asset is selected. The incremental energy efficiency contributes 9% to 11%

coal assets decrease in all scenarios by 2020 with of the energy portfolio by 2033. Case 1A (the Current announced retirements and decrease slightly more Outlook/Reference Plan case) results in 62% emission in the De-Carbonized Future and the Distributed free energy by 2033.

Marketplace scenarios where additional coal units are retired. For most of the reference plan case results, 79

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy A: The Reference Plan Capacity by 2033 Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbnized Future Scenario 5: Distributed Marketplace 0 10,000 20,000 30,000 40,000 50,000 MW, SND

• The nameplate capacity for the renewables category is as follows: Scenario 1: 5,050 MW, Scenario 2: 2,400 MW, Scenario 3: 7,500 MW, Scenario 4: 9,200 MW, Scenario 5: 2,200 MW.

Figure 74: Capacity (Summer Net Dependable Megawatts) for Strategy A by Scenario Strategy A: The Reference Plan Energy by 2033 Nuclear Hydro Coal Renewables Gas EEDR Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0 50 100 150 200 250 TWh Figure 75: Energy (Terawatt Hours) for Strategy A by Scenario 80

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy B: Meet an Emission Target focuses on This strategy was not formulated to reflect EPAs proposed achieving a system-wide CO2 emission rate target in Clean Power Plan (CPP) or rule. The proposed CCP was the least-cost manner. To set a target for the 20 year issued in June 2014 and its final form is uncertain. After EPA planning horizon ending in 2033, we used a glide slope issues the final rule, States will have one to two years to that reduces TVAs greenhouse gas emissions by 17 decide how to implement it. The CCP is also expected to be percent by 2020 and 80 percent by 2050 from a 2005 litigated by others. TVAs next update of its IRP will be able to baseline. Strategy B adopts the 2033 data point on that take into account these developments. See section 7.1.3 for glide slope of 557 pounds CO2 per MWh that translates future discussion of the CCP.

to a 50 percent reduction in TVAs system-wide CO2 emission rate from a 2005 baseline.

Figure 7-6 shows the capacity resources added by The significant contributions from the selected energy 2033 in strategy B across all five scenarios. The results efficiency and the renewable assets chosen in the from this strategy are very similar to the reference plan. reference plan result in reaching the CO2 emission The similarity of the case results was not anticipated target and therefore the two strategies are very similar.

during the development of the scenarios and strategies. Figure 7-7 shows the energy portfolio for strategy B.

81

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy B: Meet an Emission Target Capacity by 2033 Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace 0 10,000 20,000 30,000 40,000 50,000 MW, SND

• The nameplate capacity for the renewables category is as follows: Scenario 1: 5,925 MW, Scenario 2: 2,350 MW, Scenario 3: 8,300 MW, Scenario 4: 9,000 MW, Scenario 5: 2,200 MW.

Figure 76: Capacity (Summer Net Dependable Megawatts) for Strategy B by Scenario Strategy B: Meet an Emission Target Energy by 2033 Nuclear Hydro Coal Renewables Gas EEDR Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0 50 100 150 200 TWh Figure 77: Energy (Terawatt Hours) for Strategy B by Scenario 82

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy C: The Focus on Long-Term, Market Figure 7-8 presents the total capacity portfolios for Supplied Resources strategy is designed to constrain strategy C. The nuclear portfolio is similar to the TVA capital spending in the least-cost manner. In this reference plan strategy. The hydro assets increase case, new self-build assets (i.e., TVA constructed) were above the reference plan by an additional 40 MW restricted but improvements to existing owned assets project in the Growth Economy and the De-Carbonized and funds for energy efficiency and demand response Future scenarios of strategy C. The coal portfolio programs were allowed, as well as power purchase increases slightly above the reference plan because agreements. maintaining existing coal resources is more favorable than procuring market supply. Third-party renewable The original construct of strategy C in the Draft IRP assets increase above the reference plan since build allowed power purchase agreements of various lengths options arent available for selection. Gas assets to be selected with contract terms ranging from as compete across the scenarios and selection depends short as 10 years to as long as 20 years. However, on the scenario assumptions of load and commodity current market assessments indicate a lack of market prices. The volumes on the gas assets selected in this depth in the TVA territory and surrounding markets. To strategy are slightly below the reference plan. Energy attract investment and to secure a project, however, efficiency volumes remain similar across the scenarios is likely to require a longer-term commitment. Working as in the reference plan.

with our stakeholder group, Strategy C was revised to include only 20 year contract terms and the results were The energy portfolio for this strategy is shown in updated. For further information see Appendix E. Figure 7-9.

83

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy C: Market Supplied Resources Capacity by 2033 Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace 0 10,000 20,000 30,000 40,000 50,000 MW, SND

• The nameplate capacity for the renewables category is as follows: Scenario 1: 4,300 MW, Scenario 2: 2,950 MW, Scenario 3: 7,500 MW, Scenario 4: 9,100 MW, Scenario 5: 2,400 MW.

Figure 78: Capacity (Summer Net Dependable Megawatts) for Strategy C by Scenario Strategy C: Market Supplied Resources Energy by 2033 Nuclear Hydro Coal Renewables Gas EEDR Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0 50 100 150 200 TWh Figure 79: Energy (Terawatt Hours) for Strategy C by Scenario 84

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy D: The Maximize Energy Efficiency to the reference plan. The hydro differs in scenario strategy requires that future energy needs be met first 4 (the decarbonized future) from the reference plan with EE in the least-cost manner. where additional EE replaces a market hydro asset.

The coal portfolio varies in the Growth Economy and Figure 7-10 shows the energy efficiency capacity the De-Carbonized Future where the increase in EE additions in strategy D across all five scenarios. The results in additional coal unit retirements relative to the energy efficiency additions are categorized by end- reference plan strategy. Renewables are reduced by use sector (i.e., industrial, commercial and residential). about an average 400 MW SND in 2033 as compared The amount of EE in strategy D increases above to reference plan. Fewer natural gas units are selected the reference plan strategy starting in 2024 and is relative to the reference plan given the increased approximately 1,900 MW and 11,200 GWh higher than deliveries from EE. Figure 7-12 shows the corresponding the reference plan by 2033. energy portfolios. Energy efficiency displaces an average of 11 terawatt hours of energy from coal, gas, Figure 7-11 shows the complete capacity portfolio for all and renewables as compared to the reference plan.

of the strategy D cases. The nuclear assets are similar Strategy D: Maximize EE MW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Growth Economy Decarbonized Future Distributed Marketplace 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 Industrial 500 Commercial Residental 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 Figure 710: Comparison of Energy Efficiency Resources in Strategy D 85

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy D: Maximize EE Capacity by 2033 Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace 0 10,000 20,000 30,000 40,000 50,000 MW, SND

• The nameplate capacity for the renewables category is as follows: Scenario 1: 2,825 MW, Scenario 2: 2,700 MW, Scenario 3: 7,200 MW, Scenario 4: 7,175 MW, Scenario 5: 1,025 MW.

Figure 711: Capacity (Summer Net Dependable Megawatts) for Strategy D by Scenario Strategy D: Maximize EE Energy by 2033 Nuclear Hydro Coal Renewables Gas EEDR Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0 50 100 150 200 TWh Figure 712: Energy (Terawatt Hours) for Strategy D by Scenario 86

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy E: The Maximize Renewables strategy Figure 7-14 shows the capacity portfolios by 2033 enforces a renewable energy target of 20 percent by for the strategy E cases. The nuclear assets are 2020 and 35 percent by 2040. The renewable energy fairly similar to strategy A (the reference plan). Hydro target includes generation from new and existing hydro increases above the reference plan strategy with the sources. The renewable energy strategy objective is selection of a market asset in all 5 cases. The Maximize met in the least-cost manner. Renewables/Distributed Marketplace case with low loads, results in the most retired coal in the study.

Solar, wind and hydro resources were the renewable Renewables increase to more than 12 percent of the assets selected throughout the study. Figure 7-13 summer net dependable capacity portfolio by 2033 in shows the new renewable additions by technology all scenarios. The natural gas expansion is an average in five year increments for all five scenarios. The of 1,600 MW less than the reference plan across all megawatts shown are the nameplate capacities. In scenarios.

Strategy E, hydro assets are added in every scenario.

Wind is added by 2020 throughout the scenarios and Figure 7-15 shows the corresponding energy portfolios.

almost doubles by 2033. Solar is selected in the near- Renewable energy increases from an average of 17 term at smaller amounts in the scenarios with some terawatt hours in the reference plan strategies to 35 load growth. However, by 2025, the mix of renewables terawatt hours across strategy E.

averages across the scenarios to be 6 percent hydro, 47 percent wind and 47 percent solar on a nameplate capacity basis.

Strategy E: Maximize Renewables MW, Nameplate Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Growth Economy Decarbonized Future Distributed Marketplace 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 Hydro Solar Wind 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 Figure 713: Comparison of Renewable Resources in Strategy E 87

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results Strategy E: Maximize Renewables Capacity by 2033 Nuclear Hydro Coal Renewables Gas CT Gas CC EE DR Scenario 1: Current Outlook Scenaio 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: De-Carbonized Future Scenario 5: Distributed Marketplace 0 10,000 20,000 30,000 40,000 50,000 MW, SND

• The nameplate capacity for the renewables category is as follows: Scenario 1: 12,625 MW, Scenario 2: 12,125 MW, Scenario 3: 13,700 MW, Scenario 4: 11,450 MW, Scenario 5: 9,950 MW Figure 714: Capacity (Summer Net Dependable Megawatts) for Strategy E by Scenario Strategy E: Maximize Renewables Energy by 2033 Nuclear Hydro Coal Renewables Gas EEDR Scenario 1: Current Outlook Scenario 2: Stagnant Economy Scenario 3: Growth Economy Scenario 4: Decarbonized Future Scenario 5: Distributed Marketplace 0 50 100 150 200 TWh Figure 715: Energy (Terawatt Hours) for Strategy E by Scenario 88

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results 7.1.3 The Clean Power Plan and the IRP From a portfolio planning perspective, we think the On June 2, 2014, the EPA issued proposed standards TVAs carbon emission rate is a better customer-for carbon emissions from existing power plants, focused planning metric for use in the IRP.

referred to as the Proposed Clean Power Plan. At the The stringency of the Proposed Clean Power Plans time of this publication going to print, EPA indicates the state-by-state carbon emission numbers is established final rule will be released in August 2015. EPA received using four building blocks that together make up significant comments on the proposed rule, and the the best system of emission reduction or (BSER) for final rule is expected to change based on this input.

reducing carbon pollution. These are: (1) efficiency The Proposed Clean Power Plan sets state-specific improvements of the coal fired plants themselves, (2) emission guidelines for carbon dioxide (CO2) emissions increased dispatch of Natural Gas Combined Cycles from power plants, targeting a 30 percent nationwide (NGCC) , (3) increased utilization of renewable energy, reduction in CO2 emissions from 2005 levels by 2030.

at-risk nuclear, and nuclear under construction and (4)

Each states emission guideline is complex, and we increased demand-side energy efficiency. Each states refer you to EPAs website14 for a detailed explanation.

emission guideline is calculated by applying these four The Proposed Clean Power Plans emission guideline building blocks to 2012 historical fossil emissions and is in the form of a fossil energy, output-based, carbon generation. The EPA is proposing an interim goal dioxide (CO2) emissions rate for each state, which that a state must meet on average over the 10-year differs greatly from the system-wide carbon dioxide period from 2020-2029 and a final goal that a state emission rates discussed in this IRP. A system-wide must meet at the end of that period in 2030 (and carbon dioxide emission rate, such as those reported thereafter based on a three-year average). States must in the IRP, is the amount of CO2 (as measured in develop and submit plans to meet their goals and can pounds) that is emitted in the generation of a unit of comply individually or within a multi-state framework.

electrical energy (a megawatt), expressed in lbs CO2 As proposed, states would be required to submit their per MWh. As discussed above, the carbon emission plans to the EPA by June 30, 2016. The final form of rates in the proposed Clean Power Plan are primarily these standards is uncertain.

focused on fossil generation units only, and thus only While the IRP models the amount of carbon contained include a portion of a utilitys total energy generation.

in the delivered energy to our customers it does not In TVAs case, the Clean Power Plan emissions rates model a potential compliance strategy for TVA with do not include any of our hydro or most of our nuclear the Proposed Clean Power Plan. However, as a crude generation. Since collectively these non-CO2 emitting comparison, TVA has made a 30 percent reduction energy sources represent a large portion of the TVA in CO2 emissions from a 2005 baseline, the stated system, the difference in targets between a system-objective of the regulation. One might assume that wide CO2 rate and the Clean Power Plan targets for TVA would then have a low compliance hurdle with the fossil-fueled resource are very large.

CPP. However very much is unknown right now about TVA believes that the use of the overall carbon rate how the final rule might change when promulgated this is appropriate for the IRP because it simulates the summer. For instance, in the proposed rule Tennessees amount of carbon that is in the delivered energy to our emission guideline was made much more stringent customers. This rate can be expressed as: than most all other states by considering Watts Bar Nuclear Unit 2 as an existing unit even though it is not TVAs carbon emission rate = pounds of CO2 yet operational. TVA has objected to this exclusion and produced from power generation/total delivered we are awaiting the final regulation to see EPAs final energy MWh determination. Also, the final rule could change the timing of the regulation compliance period which would 14 http://www2.epa.gov/carbon-pollution-standards/clean-power-plan-proposed-rule have a significant change in the stringency.

89

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results While we think that it is both inappropriate and and trends that can be informative to the preparation premature to model the Clean Power Plan in the IRP, of the State Plans or assessing impacts from potential each of the five strategies significantly reduces carbon Federal Plans. The data, its underlying assumptions, emissions on the TVA system through the planning and associated information must be used appropriately period by retiring coal units and adding nuclear, in venues that extend beyond this IRP.

renewables and energy efficiency. Considering these reductions and that compliance with the Clean Power 7.2 Scorecard Results Plan is further complicated by calling for state-by- The fully populated scorecards for each of the five state emission reductions although TVAs integrated planning strategies are included in this section (see generation and transmission system encompasses Chapter 6 for a discussion about the development parts of several states, it is inappropriate and premature of the scorecard template). Each strategy scorecard to model the Clean Power Plan in this IRP. Regardless contains the metric values for that particular strategy of the final form of the rule and which strategy TVA in each of the five scenarios modeled in the IRP. The selects as a general planning direction from the IRP, metric values are based on the combination of the we will be bringing the nations first new nuclear portfolio optimization and uncertainty analysis work generation of the 21st Century online at Watts Bar 2 applied to each of the planning strategies under by 2016, retiring 13 units at two coal-fired power plants consideration.

in Tennessee (ALF, JOF) and 13 units at two coal-fired power plants in Alabama (COF, WCF) by 2018, and The scorecard for Strategy A is shown in Figure 7-16.

replacing some of this coal generation with lower- The highest PVRR is the Growth Economy due to the emitting natural gas, energy efficiency and renewables large build-out to meet firm requirements. The highest during the planning period. This will put TVA on a system average cost is the De-Carbonized Future. The trajectory toward complying with the Clean Power Plan Growth Economy has the highest risk exposure driven or regulatory requirements of another form in a carbon- by higher loads, and the Growth Economy has the constrained future. highest CO2 releases, water consumption, and solid waste production. Note that the scorecard presents the Neither Strategy B, nor any other strategy in this system regulating capability snapshot in 2033 (more IRP study is intended to be a compliance strategy values for this metric are discussed in Chapter 8). Since for the Clean Power Plan due to the aforementioned the Valley economics metric uses Strategy A as the differences in carbon emission metrics. However, the reference case in computing impacts, the change in per findings and recommendations from this IRP will no capita income is 0 percent for this strategy.

doubt be useful in providing generalized observations Figure 716: Strategy A Scorecard 90

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results The scorecard for Strategy B is shown in Figure strategy (and in particular the contribution from energy 7-17. These results are very similar to those shown efficiency and renewables) generally achieve the overall for Strategy A, since the portfolios developed in that system emission target designed for Strategy B.

Figure 717: Strategy B Scorecard The scorecard results for Strategy C are shown in highest environmental impacts for CO2 and water Figure 7-18. PVRR costs are slightly higher than consumption, primarily caused by the higher fraction of Strategy A reflecting the increased cost associated with fossil-fueled generation in this case. Flexibility scores the third-party PPA capacity additions in this strategy. are lower compared to the results for Strategy A due to Compared to Strategy A, system average cost metrics the lower percentage of regulating capacity added in are slightly better (actually lowest overall), and risk this plan (the portfolio has a higher contribution from exposure is just slightly higher. This strategy has the long-term PPAs and retained coal capacity).

Figure 718: Strategy C Scorecard 91

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results The Strategy D scorecard is shown in Figure 7-19. strategy, due to the requirement that resource needs PVRR cost rankings across the scenarios are similar to be met first with energy efficiency, thereby restricting strategy A but total costs are generally higher. System portfolio composition. However, this strategy has better average costs are similar in the first 10 years. The Risk/ performance in environmental metrics.

Benefit Ratio and Risk Exposure are higher for this Figure 719: Strategy D Scorecard Strategy E metric values are shown in Figure 7-20. renewables, which indicates that the enforced targets PVRR costs are higher than respective strategy A may be too high relative to the benefits derived from costs for all cases, the result of aggressive renewable adding renewable resources to the portfolio. This resource targets. Correspondingly, system average strategy has the best performance in all environmental costs are higher in all strategy E cases. The Risk/ metrics, driven by the higher concentration of renewable Benefit Ratio and Risk Exposure are higher with more resources in the cases.

Figure 720: Strategy E Scorecard 92

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Study Results 7.3 Scoring Metric Comparisons Figure 7-21 shows a comparison of how each strategy scored across all scenarios by metric Scenario 1 2 3 4 5 Alternative Current Stagnant Growth De-Carbonized Distributed Average Strategy Outlook Economy Economy Future Marketplace A 132.7 125.9 139.5 131.7 120.4 130.0 B 132.7 126.0 139.5 131.7 120.4 130.1 PVRR C 133.4 126.5 140.8 131.9 121.1 130.7

($ billion)

D 134.4 127.9 141.3 133.6 122.8 132.0 E 136.2 129.4 140.8 132.8 123.5 132.5 A 76.7 76.0 77.7 81.0 77.3 77.7 B 76.7 76.0 77.7 80.8 77.3 77.7 System Average Cost 2014-2023 C 76.4 75.7 77.8 80.6 76.8 77.5

($/MWh)

D 76.9 75.9 77.5 81.1 77.3 77.7 E 78.4 77.3 78.5 81.3 78.5 78.8 A 0.92 0.95 0.91 1.00 0.99 0.95 B 0.92 0.95 0.92 0.99 0.99 0.95 Risk Exposure C 0.90 0.95 0.91 0.98 0.99 0.95

($ billion)

D 0.94 0.98 0.92 1.03 1.00 0.98 E 1.03 1.04 1.03 1.01 1.05 1.03 A 140.4 132.8 147.5 140.3 127.1 137.6 B 140.4 133.0 147.6 140.3 127.1 137.7 Risk Exposure C 141.2 134.0 149.3 140.7 128.3 138.7

($ billion)

D 142.4 135.3 149.7 142.7 130.0 140.0 E 145.1 137.4 149.8 141.7 130.9 141.0 A 57.0 51.8 59.7 44.2 44.2 51.4 B 57.0 51.8 59.7 44.3 44.2 51.4 CO2 Emissions (million tons/year) C 58.2 52.9 59.1 44.2 45.2 51.9 D 56.2 50.7 57.6 41.8 43.5 50.0 E 52.2 45.6 54.2 41.6 39.9 46.7 A 72,952 73,429 79,489 65,890 67,536 71,859 B 72,988 73,438 79,508 66,001 67,533 71,894 Water Consumption (million gal-C 74,504 75,042 79,296 66,227 68,840 72,782 lons/year)

D 72,657 72,827 77,482 65,696 67,020 71,136 E 69,019 68,389 74,733 65,450 63,799 68,278 93

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 7: Name of Chapter Scenario 1 2 3 4 5 Alternative Current Stagnant Growth De-Carbonized Distributed Average Strategy Outlook Economy Economy Future Marketplace A 3.46 3.49 3.72 3.08 3.21 3.39 B 3.46 3.50 3.71 3.10 3.21 3.39 Waste (million tons/year) C 3.41 3.46 3.69 3.09 3.24 3.38 D 3.44 3.44 3.73 2.75 3.17 3.31 E 3.16 3.13 3.50 2.75 2.93 3.10 A 28.7% 28.0% 27.1% 18.9% 22.3% 25.0%

B 29.9% 27.9% 26.2% 19.7% 22.3% 25.2%

System Regulating Capability C 28.5% 26.0% 28.6% 20.4% 18.2% 24.4%

(2033)

D 27.7% 22.3% 26.4% 20.3% 25.0% 24.3%

E 20.9% 20.4% 23.5% 18.8% 16.0% 19.9%

A B 0.00% 0.01% -0.01% 0.00% 0.00% 0.00%

Percent Difference in Per Capita C 0.00% 0.01% 0.03% 0.01% 0.00% 0.01%

Income (Relative to Strategy A)

D 0.02% 0.02% 0.02% 0.02% 0.02% 0.02%

E -0.01% 0.00% 0.00% 0.00% -0.01% -0.01%

Figure 721: Scoring Metrics by Strategy & Scenario 7.4 Preliminary Observations

• Higher levels of energy efficiency and renewable resources are indicated in many cases over the 20 Based on the results of the modeling to date, TVA made year study period some observations about the case results:

• Changing environmental standards for CO2 will

• There is a need for new capacity in every scenario drive retire/control decisions on some coal-fired being modeled, even in the lower load futures generation in the mid-2020s

• There are no immediate needs for baseload

• Solar resources begin appearing in the resource resources beyond the completion of Watts Bar Unit plans in the mid 2020s; wind resources appear in 2 and the Browns Ferry extended power uprates the late 2020s in some scenarios, and generally the

• Most of the variation in expansion plans is around HVDC wind option is not selected until early 2030s natural gas and renewables and most of the resource plans show a tradeoff between EE and gas These observations are further explored in the resources assessments presented in Chapter 8.

94

Chapter 8 Strategy Assessments Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 95

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments 8 Strategy Assessments

• Expected Value PVRR, 20 Year - the total plan cost (capital and operating) expressed as the present This chapter explains the strategy assessments and value of revenue requirements (PVRR) over the 20-summarizes the results. Areas where additional study year study period.

may be needed and next steps in the IRP process are These metrics allowed us to compare the cost and also discussed.

financial risks associated with different planning strategies from both a short-term (10-year) and a long-8.1 Strategy Assessments term (20-year) perspective. (See Chapter 6, section To assess the performance of the five planning 6.2.2, for more information on scoring metrics, including strategies (explained in Chapter 6 and shown below), the formulas used to compute them.)

we used scorecard data to conduct four assessments:

Figure 8-1 shows the results for the 10-year system

• Cost and risk average cost metric. The blue bar represents the

• Environmental stewardship system average cost values for the first 10 years in the

• Flexibility study period (2014-2023), and the red bar represents

• Valley economics the second 10-year period (2024-2033).

We calculated the overall value of each strategy by averaging its performance over every scenario, since all of them are presumed to be equally likely.

8.1.1 Cost and Risk Assessment The cost and risk assessment was aimed at gaining a better understanding of the relative performance of different strategies in terms of total plan costs and financial risk.

The cost assessment was based on two scorecard metrics:

Planning Strategies

• System Strategy A: Traditional Utility Planning Average Cost Figure 81: System Average Cost Strategy B: Meet an Emissions Target ($MWh), Year 1 the During the first 10-year period, the system average Strategy C: Focus on Long-Term, average system Market Supplied Resources cost is essentially the same across all five strategies.

cost for the first However, in the second 10-year period, there is some Strategy D: Maximize Energy Efficiency 10 years of the variation, with Strategy D exhibiting the highest system study, computed average cost. This is likely the result of increased Strategy E: Maximize Renewables as the levelized costs for energy efficiency programs combined with annual system a resultant reduction in energy sales. These factors average cost combine to shrink sales and put upward pressure on (i.e., revenue requirements in each year divided by the system average cost.

sales in that year)

Figure 8-2 shows the results for the 20-year present value revenue requirement (PVRR) metric. The chart 96

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments shows the range of plan costs as well as the expected and B are virtually identical due to the selection of EE value for each strategy across all the scenarios. The in Strategy A which enables that case to essentially lower end of each bar is the best case (lowest cost) achieve the emission target set in Strategy B, while outcome from the uncertainty analysis; the upper Strategy C is slightly more expensive. Strategies D and end is the worst case (highest cost) outcome; the E, which constrain the selection of resource types used expected value is the point of transition between the in the plan, are projected to have a PVRR that is about two colored sections of each bar. Strategies A, B and $2 billion higher over the 20-year planning period, while C have roughly the same average PVRR results across the range of possible outcomes for all five strategies is all scenarios and the lowest set of total plan costs fairly consistent as shown by the height of the bars.

measured in terms of the 20-year PVRR. Strategies A Figure 82: Total Plan Cost (PVRR)

Two additional metrics were used to assess the risk of area between Expected Value and P(5) (when costs are each strategy: less than the expected value)

Risk/Benefit Ratio - the area under the plan cost Risk Exposure - the point on the plan cost distribution distribution curve between P(95) and Expected Value below which the likely plan costs will fall 95 percent of (when costs exceed the expected value) divided by the the time (this is also the worst-case outcome).

97

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Figure 83: Risk/Benefit Ratio Figure 8-3 shows the risk/benefit ratios for the five value, caused in part by the aggressive renewable planning strategies. In this metric, lower values indicate targets established in this case. We investigated better performance where the benefits outweigh the key assumptions in Strategy E in an effort to better risks. Risk/benefit scores less than 1.0 indicate that understand this result, and those results are discussed costs are more likely to be less than the expected value. in Section 8.3.

Strategies A-C have very similar scores, with Strategy Figure 8-4 shows TVAs risk exposure under the five C scoring just slightly lower (better performance) than strategies. This metric measures the worst-case Strategies A/B. Strategy E appears to be the most outcome as represented by the P(95) value of the PVRR risky from a financial perspective. It is the only strategy distribution and is useful in determining which strategies with a ratio greater than 1.0, indicating that plan costs present the higher financial risks.

in this strategy are more likely to exceed the expected 98

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Figure 84: Risk Exposure Strategies A and B have essentially the same risk analysis, leading to these higher financial risk scores.

exposure; the risk exposure for Strategy C is higher Strategy E has the highest risk exposure and is also by about $1 billion, reflecting the potential upside cost the only strategy with a risk/benefit ratio greater than associated with the long-term power agreements in 1.0. This indicates that this strategy may be the most that case. Strategies D and E have distinctly higher risky financially of those evaluated in the IRP. This result exposure values - as much as $3.5 billion higher. This is driven by the very aggressive targets for renewable indicates that strategies where resource selection is resources that are imposed in the strategy.

constrained, such as aggressive targets imposed for energy efficiency (Strategy D) and renewables (Strategy Another way to assess cost and financial risk is to E) carry higher financial risks than the other three combine the cost and risk scores so a trade-off analysis strategies. In both of these strategies, the required can be performed. Figure 8-5 shows cost/risk trade-resource contributions (EE or renewables) tend to limit offs based on total plan cost and system average cost.

the flexibility to optimize a portfolio in the uncertainty 99

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Figure 85: Cost/Risk Trade-Offs 100

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments These charts also reinforce the cost and risk

• CO2 Average Tons - the annual average tons of CO2 assessment results discussed about Strategies D and emitted over the study period E having somewhat higher plan costs and exhibiting

• Water consumption - the annual average gallons of higher financial risks, with Strategy E showing the water consumed over the study period highest cost and risk outcome. There is also a trade-

• Waste - the annual average quantity of coal ash, off between Strategies A/B and C in the upper chart in sludge and slag based on energy production in each Figure 8-5, which indicates that a somewhat improved portfolio.

(lower) risk/benefit ratio can be achieved for these strategies but at a slightly higher plan cost. Figure 8-6 shows the average environmental impact for each strategy for each of these three metrics.

8.1.2 Environmental Stewardship The graphic presents the impacts on a relative basis, As discussed in Chapter 6, strategy scorecards normalized to the highest impact for each metric. More include three measures for environmental stewardship information about the development of these metrics can performance:

be found in Appendix F.

Figure 86: Environmental Impacts Strategies A, B and C have almost the same all three metrics, with Strategy E showing the lowest environmental impacts across all three metrics, with impacts. The air and waste impacts in Strategy E are Strategy C having a slightly higher impact. Strategy significantly lower than the other strategies due to the D shows somewhat lower environmental impacts for emphasis on renewable resources.

101

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments 8.1.3 Flexibility This is the first time TVA has used annual system Annual system regulating capability, expressed as a regulating capability as a metric to assess the percentage of peak load, was used to measure the performance of a resource portfolio, and further work flexibility of the five planning strategies. TVA considers is planned after the completion of this IRP to determine flexibility - the ability of the system to respond to load what the minimum or optimum flexibility score should swings - as a key future consideration for long-range be for the TVA system.

resource planning.

Figure 8-7 shows flexibility scores for each strategy at This is especially true as the resource mix shifts from three points within the study window: 2014, 2024 and traditional, fully dispatchable central station units toward 2033 (higher is better).

more diverse and dispersed generating assets.

Figure 87: System Regulating Capability Strategy D has a higher flexibility score during the capacity additions due to the higher commitment to first ten years of the study period due to lower system long-term PPAs or energy efficiency resources in those load. However, during the second decade, the quick strategies. The results for Strategy E are significantly response units added in Strategies A and B result in different because this strategy has a higher percentage similar levels of regulating capability. By the end of the of non-dispatchable renewable resources and thus a study period, Strategies C and D have slightly lower reduced ability to respond to unexpected load swings.

flexibility scores, likely the result of fewer quick-start 102

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments 8.1.4 Valley Economics income level established by Strategy A in each scenario.

The impact of different planning strategies on the Valley More details about how TVA has computed this macro-economy was assessed based on the percent change economic impact can be found in Appendix G. The in per capita income, measured from the reference results are shown in Figure 8-8.

Figure 88: Valley Economics Strategy D consistently outperformed the reference relative to other resource options. However, the overall income level across all scenarios. This is likely due to variation in per capita income estimates is very small the retention of more investment in the Valley under this across the strategies. This indicates that the Valley strategy driven by the commitment to energy efficiency, Economics metric is unlikely to be a key consideration which results in increased investment in the Valley when selecting a preferred target power supply mix.

103

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments 8.1.5 Summary of Initial Observations The overall performance of the five planning strategies is summarized by metric category in Table 8-1 and by strategy in Table 8-2.

Metric Category Assessment Observations On the basis of average system costs, all five strategies are very similar over the first 10-years of the study period. Total plan costs over the 20-year study period are also Cost similar, with the strategies D and E that constrain the selection of resource types in the plan, more expensive.

Risk scores are lower for the strategies that emphasize significant investment in any one particular technology. For example, we see higher risk in portfolios that focus on Financial Risk higher levels of energy efficiency or renewables. (Note: sensitivity cases found similar risk profiles for portfolios that concentrated on nuclear technologies.)

All strategies show significant improvement in air (CO2), water and waste categories Environmental compared to the performance of the current resource portfolio, with the Maximize Stewardship Renewables strategy having the lowest environmental impact.

All strategies appear similar, but the ability of the system to respond to load uncer-Flexibility tainty is most limited in the Maximize Renewables strategy. The flexibility score for the Maximize Energy Efficiency strategy is likely a result of reduced loads.

All strategies seem to have comparable impact on the Valley economy as measured by per capita income. The Maximize Energy Efficiency strategy appears to have a slightly Valley Economics stronger economic impact due to a higher percentage of investments remaining in the Valley.

Table 81: Summary of Observations by Metric Category 104

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Strategy Assessment Observations

• Relatively low PVRR and System Average Cost during the first 10 years of the study period Strategy A:

• Lowest System Average Cost in the second 10 years of the study period Reference Plan

• Low financial risk (risk/benefit ratio less than one; second lowest risk exposure)

• Higher environmental impact compared to Strategies D and E

• Demonstrates flexibility Strategy B:

Meet an Emission

• Results are nearly identical to Strategy A Target Strategy C:

• Slightly higher PVRR, relatively low system average cost, and moderate financial Rely on Long-Term, risk Market-Based

• Higher environmental impact than other strategies Resources

• Somewhat lower system regulating capability than Strategies A or B

• Higher PVRR than Strategies A, B or C

• Relatively similar system average cost to other strategies during the first decade, Strategy D:

but high system average cost during the second decade due to increasing levels Maximize Energy of EE and lower power sales Efficiency (EE)

• Comparable to Strategy C on flexibility performance due to reduced sales

• Low environmental impact, second only to Strategy E

• Highest PVRR in all scenarios due to enforcement of renewable energy targets Strategy E:

• Highest risk/benefit ratio of any strategy (greater than 1.0)

Maximize

• Lower flexibility performance than other strategies Renewables

• Lowest environmental impact Table 82: Summary of Observations by Strategy 105

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments 8.2 Reporting Metrics Comparisons performance of the individual planning strategies in each of the results. Figure 8-9 shows a comparison As further described in Chapter 6, in addition to scoring of how each strategy scored across all scenarios by metrics, reporting metrics were selected to provide reporting metric.

further explanation and clarification in interpreting the Alternative Scenario Strategy 1 2 3 4 5 Average A 98.7 94.8 100.4 103.0 98.7 99.1 B 98.6 95.1 100.4 103.4 98.7 99.3 System Average Cost 2024-2033 C 100.5 96.7 104.2 104.6 101.4 101.5

($/MWh)

D 104.5 102.4 106.8 110.0 108.3 106.4 E 102.0 99.1 100.8 104.6 101.6 101.6 A 16,014 14,331 16,810 17,277 13,435 15,573 B 16,051 14,295 16,884 17,241 13,422 15,579 Cost Uncertainty C 16,538 15,318 17,910 17,940 14,628 16,467

($Bn)

D 16,477 15,008 17,420 17,919 14,296 16,224 E 17,527 15,677 17,751 17,664 14,589 16,642 A 0.058 0.055 0.057 0.065 0.056 0.058 B 0.058 0.055 0.058 0.065 0.055 0.058 Risk Ratio C 0.059 0.059 0.061 0.067 0.060 0.061 D 0.059 0.058 0.059 0.068 0.058 0.061 E 0.065 0.062 0.064 0.067 0.061 0.064 A 350.0 330.0 352.9 291.3 306.9 326.2 B 350.3 330.1 353.0 292.2 306.9 326.5 CO2 Intensity C 357.5 337.4 352.2 291.4 314.4 330.6 (Tons/GWh)

D 351.4 329.6 345.6 279.9 308.5 323.0 E 320.4 290.7 319.8 273.5 275.1 295.9 A 149.05 149.05 149.05 149.05 149.05 149.05 B 149.05 149.05 149.05 149.05 149.05 149.05 Spent Nuclear Fuel C 149.05 149.05 149.05 149.05 149.05 149.05 (Tons/Year)

D 149.05 149.05 149.05 149.05 149.05 149.05 E 149.05 149.05 149.05 149.05 149.05 149.05 A 24.9% 17.8% 31.2% 40.7% 19.2% 26.8%

B 24.7% 17.6% 33.5% 39.6% 19.2% 26.9%

Variable Resource Penetration C 22.9% 20.0% 31.8% 40.9% 20.0% 27.1%

2033 D 19.7% 19.6% 32.2% 36.5% 16.9% 25.0%

E 48.1% 48.1% 48.1% 48.7% 45.0% 47.6%

106

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Alternative Scenario Strategy 1 2 3 4 5 Average A 49.1% 45.2% 52.5% 65.0% 52.9% 53.0%

B 48.9% 45.2% 54.1% 64.6% 52.9% 53.2%

Flexibility Turndown Factor C 46.1% 45.9% 52.8% 64.9% 53.4% 52.6%

2033 D 46.7% 48.8% 55.5% 65.0% 56.4% 54.5%

E 62.2% 63.3% 60.6% 66.4% 68.0% 64.1%

A B 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

NonFarm Employment: % Changes from C 0.0% 0.0% 0.1% 0.0% 0.0% 0.0%

Reference Plan (A)

D 0.1% 0.1% 0.1% 0.1% 0.1% 0.1%

E 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%

Figure 89: Reporting Metrics by Strategy & Scenario 8.3 Sensitivity Analysis 2. Energy Efficiency and Demand Response During the course of developing the draft IRP, TVA sensitivities that tested the impact of key energy identified questions and findings that warranted further efficiency cost and performance assumptions evaluation prior to finalizing the study. In addition, we and assessed the impact of energy efficiency and received helpful stakeholder feedback from the IRP demand response on the overall portfolio.

Working Group (IRPWG), the Regional Energy Resource

3. Renewables sensitivities that evaluated the impact Council (RERC), and through our public meetings and of key pricing and performance assumptions formal comment period that helped identify key areas around renewable technologies.

that merited further analysis.

4. Resource sensitivities that tested the impact to the To address these questions and comments we case results of adding other resources not selected performed detailed sensitivity analyses which were in the initial runs.

reviewed with the IRPWG and the RERC in April of 5. Key Driver sensitivities that analyzed the impact 2015. The sensitivity cases generally fell into five primary to the case results if a specific combination of categories: assumptions was imposed on the optimization model, rather than using the correlated scenario

1. Nuclear sensitivities that tested the impact to the assumptions developed for the study. An example case results if different nuclear options not selected would be forcing in a high gas price forecast.

in the initial case runs were forced into the portfolio.

107

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments The sensitivity cases are listed below in Figure 8-10. Key findings are summarized below by sensitivity case category:

Nuclear Sensitivities

• Pressurized water reactor or Bellefonte Unit 1 Nuclear Sensitivity Results:

and Unit 2 This set of cases examined the impact of forcing in

• Advanced pressurized water reactor or AP 1000 Bellefonte unit 1 in 2026, Bellefonte 1 and 2 in 2028, an

• Small modular reactor AP 1000 in 2028, and a Small Modular Reactor into the resource plan in 2028. These resources were added in Energy Efficiency and Demand Response the year specified and the portfolio was re-optimized Sensitivities within the framework of Scenario 1 and Strategy A.

• No Energy Efficiency Planning Factor Adjustment Conclusions are as follows:

• Faster Energy Efficiency Ramp Rate

• New nuclear additions result in higher overall system

• No Demand Response costs than the reference plan but would deliver

• No Energy Efficiency or Demand Response value beyond the study window. Cost-sharing mechanisms that could be made available for Small Renewable Sensitivities Modular Reactors (SMRs) have not been included

• Extension of Solar Tax Credits but, if available, would render those options more

• Higher HVDC Wind Net Dependable Capacity &

financially attractive.

lower cost

• Short-term system average costs are higher with

• Slower Solar Cost De-escalation nuclear builds, but long-term average costs are

• Slower Wind Cost De-escalation similar to non-nuclear cases.

• New nuclear units eliminate natural gas builds and Resource Sensitivities some renewables which were the primary expansion

• Pumped-hydro storage units in the reference case. Energy Efficiency levels

• Compressed air energy storage are similar to the reference plan (case 1A).

• Integrated gas combined cycle with carbon

• System-wide CO2 emissions are lower as the capture and sequestration generation from the nuclear units replaces gas

• Supercritical pulverized coal 1x8 with carbon generation and displaces existing coal generation.

capture and sequestration

• New direct combustion Biomass Energy Efficiency & Demand Response Sensitivity Results:

Key Driver Sensitivities These cases examined several key Energy Efficiency

• Higher load inputs, including testing the impact of removing the

• No CO2 penalty planning factor adjustment (discussed in Appendix

• Lower gas price D) and accelerating the near-term ramp rates for

• Higher gas price Energy Efficiency. Cases were also run with no Energy Efficiency or Demand Response in the portfolio.

Conclusions are as follows:

• Removing the planning factor adjustment results in similar Energy Efficiency volumes as the reference Figure 8-10: List of Sensitivity Cases case (case 1A) through 2023, increasing thereafter to midway between the reference case and the Maximize Energy Efficiency strategy (case 1D) by 2033.

108

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments

• Increasing the ramp rate in the early years of the Resource Sensitivity Results:

study results in small increases in Energy Efficiency This group of cases examined the impact of forcing by 2033 with slightly more selections near to mid- in certain resource types not selected in the original term and little impact to overall system cost. case runs. These include forcing in pumped storage,

• Higher volumes of Energy Efficiency equate to higher compressed air energy storage, pulverized coal with system average costs because electricity sales are carbon capture and sequestration (CCS), integrated lower. There is a tradeoff between average system gasification combined cycle (IGCC) with CCS and cost and total system costs even in the reference biomass plants. These resources were added in 2028 case. and the portfolio was re-optimized. Conclusions are as

• Energy Efficiency continues to be perform as a follows:

resource in model results:

Energy Efficiency programs eliminate some

• Coal options generally displace demand response, of the need for natural gas units as well as natural gas and renewable generating assets. Each some renewable purchases. Energy Efficiency coal option increases total system costs.

volumes reduce generation from gas, coal, and

• Biomass options generally offset small amounts of renewable resources. demand response.

Demand Response programs eliminate some

• Pumped storage generation offsets future gas of the need for natural gas peaking units and generation and some renewables.

market purchases.

• Compressed Air Energy Storage generally offsets gas peaking and demand response assets.

Renewables Sensitivity Results:

These sensitivity cases examined the impact of Key Driver Sensitivity Results:

lower cost assumptions for wind and solar resources These cases analyzed the impact of changing one key and more favorable guaranteed capacity from High driver while leaving other case inputs unchanged. A Voltage Direct Current (HVDC) wind. Cases were also more aggressive load growth case was tested, as were analyzed that assumed solar and wind costs do not higher and lower gas prices. In addition, a case with no decline as quickly as assumed in the original scenarios. CO2 cost was analyzed. Conclusions are as follows:

Conclusions are as follows:

• If loads are materially higher than expected, resource

• Lowering costs and providing a higher guaranteed needs are primarily met with new natural gas builds net dependable capacity for HVDC wind results in and market purchases; renewables and EE remain selection as early as 2020. similar to reference case.

• Assuming lower prices driven by tax policy and

• In a low gas price case, more natural gas units availability of favorable solar sites, utility-scale solar are built and additional coal is retired. There are tracking is selected as early as 2020. also fewer renewable purchases and less energy

• Increasing solar escalation rates pushes out utility- efficiency.

scale solar selection to 2029 and halves the volume

• In a high gas price case there are fewer natural gas compared to reference case. units built, more renewable purchases and more coal

• Increasing wind escalation rates pushes out wind generation.

selection to beyond 2033.

• If the CO2 price penalty is removed there is additional

• As seen in other sensitivity cases, renewable coal generation and fewer renewable purchases.

selection is highly sensitive to gas prices.

109

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 8: Strategy Assessments Overall

Conclusions:

The sensitivity case runs were instrumental in informing TVA and the stakeholder working groups about key inputs affecting the results. In general, the sensitivity cases confirmed that our original case study results formed a reasonable boundary of future resource additions. Other high-level conclusions are summarized below:

• New nuclear or coal assets would offset gas builds and renewable purchases, increase total cost and lower fuel risk. Cost-sharing would render SMRs more attractive, and nuclear additions may prove more valuable in future IRPs given their long lives and the possible expiration of some of our existing nuclear licenses that may occur just beyond the study window.

• The original EE case results (strategy D) still form an effective boundary for EE results, and energy efficiency programs eliminate the need for most natural gas builds and some renewable purchases.

Removing the planning factor adjustment does not affect the near-term selection of energy efficiency and results in selections midway between case 1A and 1D at the end of the study period. Increasing near-term ramp rates does not materially change the overall trajectory or costs. Finally, higher volumes of EE result in higher system average costs, even in the Reference Plan (case 1A), driven by lower electric sales. In some cases, the impact to average cost is similar to the impact of adding new nuclear builds to the portfolio.

• Renewable selection is highly dependent on gas price assumptions, load, and unit cost and characteristics.

• Natural gas pricing remains a key sensitivity for all resource selections.

110

Chapter 9 Recommendations Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 111

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations 9 Introduction 9.1 Study Objectives The Tennessee Valley Authoritys 2015 Integrated The following objectives guided the development of the Resource Plan (IRP) will guide TVA in making decisions IRP:

about the energy resources used to meet future demand for electricity. Having a long-range resource 1. Deliver a plan aligned to least-cost planning plan enables us to provide affordable, reliable electricity principles.

to the people we serve. It is a crucial element for 2. Manage risk by utilizing a diverse portfolio of success in a constantly changing business and supply and demand-side resources.

regulatory environment and will better equip us to meet 3. Deliver cleaner energy and continue to reduce many of the challenges facing the electric utility industry environmental impacts.

in the coming years. 4. Evaluate increased use of renewables, energy We used an integrated, least-cost system planning efficiency, and demand response resources.

process that takes into account the demand for 5. Ensure the portfolio delivers energy in a reliable electricity, resource diversity, reliability, costs, risks, manner.

environmental impacts, and the unique attributes of 6. Develop the ability to dynamically model energy different energy resources. Various ranges of inflation, efficiency in the study.

commodity prices, and environmental regulations were 7. Provide flexibility to adapt to changing market evaluated to provide needed information. Constraints, conditions and identify significant sign posts.

including corporate, strategic, and environmental 8. Improve credibility and trust through a collaborative objectives, were considered as different combinations and transparent process.

of strategies and predictions of future conditions were 9. Integrate stakeholder perspectives throughout the analyzed and evaluated.

study.

We conducted the IRP process in a transparent, inclusive manner that provided numerous opportunities The analysis performed in the study is intended to for public education and participation. The analysis identify a resource mix that positions TVA for success performed in this IRP study relied on industry-standard regardless of how the future unfolds. The resulting models and incorporated best practices while using an power supply mix will meet these goals: low cost, innovative methodology to more fully evaluate the role of reliable, risk-informed, diverse, environmentally energy efficiency and renewable resources in the power responsible and flexible.

supply mix. Resource cost and performance input data were independently validated. 9.2 Findings The IRP study demonstrates that TVA power will continue to be reliable, affordable and sustainable into the future. Our resource additions will build on TVAs existing diverse asset portfolio reflected in Figure 9-1.

112

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations Figure 9-1: 2014 TVA Portfolio Study results show that there is no immediate need footprint and position the Tennessee Valley to have for new base load plants after Watts Bar Nuclear Unit significant reductions in CO2 emissions. Strategies 2 comes online and uprates are completed at Browns that emphasize energy efficiency or renewables have Ferry Nuclear Plant. Instead, we can rely on additional the best environmental results.

natural gas generation, greater levels of cost-effective

• Valley Economics: All strategies have a similar impact energy efficiency and increased contributions from on overall economic health and contribute to a competitively priced solar and wind power. We also strong, vibrant economy across the region.

expect to have less coal-based generation in our

• Flexibility: System flexibility is generally equivalent energy mix than we do today. In all cases, TVA will in most cases but is reduced when renewables are continue to provide for economic development in the strongly emphasized.

Tennessee Valley.

Reviewing these results led to questions from We identified five key measures to evaluate the stakeholders about how changes in assumptions performance of the plans created as part of the study.

or resource choices might impact the findings.

A review of the case results produced the following A series of sensitivity cases were evaluated with outcomes:

five key assumption categories: nuclear additions,

• Cost: Total costs are similar for many of the cases modified assumptions for energy efficiency, alternative over the long term, and strategies that allow for a renewable resource costs, impacts associated with diverse mix of resource additions have a lower cost forcing resources not otherwise selected into the mix, than those that emphasize particular technologies. and changes in fundamental drivers such as load Higher amounts of energy efficiency may create growth and fuel pricing. The results of these analyses an upward pressure on rates in future years due to supported the ranges established in the initial findings.

reduced sales. The sensitivity cases, coupled with the original 25

• Financial Risk: Risks are minimized by maintaining case results, provide a robust set of potential resource a diverse portfolio and not over-emphasizing any additions evaluated in the IRP from which the final specific resource type. recommendations were derived. Figure 9-2 provides the

• Environmental Stewardship: All strategies show range of capacity additions (by 2033, rounded) from the significant improvement in TVAs environmental IRP case and sensitivity analysis:

113

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations Figure 9-2: Evaluated MW Additions/Retirements by 2033 from IRP Base and Sensitivity Case Analysis Key findings from the sensitivity cases are summarized

• Renewable selection is highly dependent on gas below: price, load, and cost and performance assumptions.

• Natural gas pricing and load levels remain key

• New nuclear or coal assets would offset gas builds sensitivities for all resource decisions.

and renewable purchases. Nuclear additions increase total cost but lower fuel risk. Small Modular Reactors are presently cost-prohibitive, but 9.3 Developing the Recommendation cost-sharing would render them more financially The recommendation takes into account customer attractive. Subsequent IRPs will need to address the priorities around power cost and reliability across expected expiration of licenses for TVAs operating different futures. Implementing the least-cost resource nuclear units which may occur beyond the present plan with these priorities in mind will help ensure TVA study window. continues to fulfill its mission to serve the people of the

• Energy efficiency was successfully modeled as Tennessee Valley.

a selectable, supply-side equivalent resource.

In general, energy efficiency programs eliminate In developing a recommendation from the study, the need for natural gas units as well as some TVA has elected to establish guideline ranges for key renewable purchases. As with any resource, cost resource types (owned or contracted) that make up and performance assumptions are critical to the the target power supply mix. This general planning final result, and lower costs or uncertainty around direction is expressed over the 20-year study period this resource would increase its selection in the while also including more specific direction over the portfolio. Our study results also highlight that higher first 10-year period. In order to distill the considerable volumes of energy efficiency tend to increase system number of cases evaluated through the original scenario average costs. TVA and our local power company and strategy analysis and the sensitivity cases, the partners will need to balance energy efficiency recommendation uses ranges that are centered on volumes and programs to ensure that those who results obtained under the Current Outlook scenario.

cannot participate in these programs are not The other scenario results provide a sense of how the disproportionately impacted. recommended mix might change as the future changes.

114

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations The need to shift the resource mix will be based on the end year of the study (2033), shown in Megawatts these key variables: (MW). The results are drawn from strategies A, B and C which do not place specific targets on particular

• Changes in the load forecast. resource types. Strategies D and E were intentionally

• The price of natural gas and other commodities. designed to focus on meeting future resource needs

• The pricing and performance of energy efficiency with certain resource types only (Energy Efficiency and renewable resources. and Renewables). The results of these strategies

• Impacts from regulatory policy or breakthrough are included in the analysis but are not part of the technologies. recommended range because of their limited focus on particular resource types. These strategies provided The first three variables represent the fundamental valuable insight into the planning process and provide drivers for most of the variation in the resource plans outer bounds to which TVA could navigate if certain produced across the strategy/scenario combinations. developments or conditions unfold.

Our planning direction, while initially focused around the current view of the future, is flexible enough to The solid bars represent the range of results from indicate how that power supply mix would shift if one strategies A-C in the Current Outlook scenario, or more of these key variables exhibits a material which represents our best estimation of the future.

change from the forecasts used in the IRP. We also However, recognizing that a variety of future scenarios recognize that impacts from breakthrough technologies are possible, we provide a broader range (shown in (like a significant advance in energy storage) would horizontal black lines) to represent how the resource be a game-changer, and TVA will continue to monitor portfolio may respond in different future scenarios. The emerging technology as it develops. range for Energy Efficiency and Demand Response also incorporates TVAs current trajectory for these This approach provides a more robust recommendation resources to account for some of the implementation than was developed in the 2011 IRP. While that and policy uncertainties discussed in Appendix D and approach provided a solid framework for the resource Chapter 10.

decisions TVA has made since the TVA Board accepted the IRP planning direction in the spring of 2011, the The recommended ranges represent incremental changing utility marketplace requires a more flexible additions (or retirements) from the existing resource guide that provides decision-makers with a clear fleet and include contracted (market) positions that understanding of how the resource mix would evolve in can be sourced from resources that meet cost and response to future uncertainties. The recommendation performance requirements, providing flexibility for the meets the dual objective of ensuring flexibility to portfolio. The results are bounded by the full range respond to the future while providing guidance on of the IRP cases and sensitivity runs which affirm how our resource portfolio should change as the the merits of a diverse portfolio. TVA will closely future unfolds. monitor key input variables including changes in the load forecast, the price of natural gas and other 9.4 Target Power Supply Mix commodities, the pricing and performance of energy efficiency and renewable resources, and impacts from The recommendations for the power supply mix are regulatory policy or breakthrough technologies to help presented in the form of ranges around boundaries determine whether adjustments should be made to the established by the IRP results. Figure 9-3 shows the recommended ranges.

range of resource additions we are proposing by the end of the first 10 years of the study (2023) and by 115

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations Figure 9-3: Range of MW Additions by 2023 & 203315 Recommendations by resource type:

Coal: Continue with announced plans to retire units Demand Response: Add between 450 and 575 MWs at Allen, Colbert, Johnsonville, Paradise and Widows of demand reduction by 2023 and similar amounts Creek. Evaluate the potential retirement of Shawnee by 2033, dependent on availability and cost of this Fossil Plant in the mid-2020s if additional environmental customer-owned resource.

controls are required. Consider retirements of fully controlled units if cost effective. Energy Efficiency: Achieve savings between 900 and 1,300 MW by 2023 and between 2,000 and 2,800 MWs Nuclear: Complete Watts Bar Nuclear Unit 2 and by 2033. Work with our local power company partners pursue additional power uprates at all three Browns to refine delivery mechanisms, program designs and Ferry units by 2023. Continue work on Small Modular program efficiencies with the goal of lowering total cost Reactors as part of technology innovation efforts and and increasing deliveries of efficiency programs.

look for opportunities for cost sharing to render these more cost-effective for our ratepayers. Solar: add between 150 and 800 MW of large-scale solar by 2023 and between 3,150 and 3,800 MW of Hydro: Pursue an additional 50 MW of hydro capacity large-scale solar by 2033. The trajectory and timing at TVA facilities and consider additional hydro of solar additions will be highly dependent on pricing, opportunities where feasible. performance and integration costs.

15 MWs are incremental additions from 2014 forward to align to the IRP analysis base year. Board-approved coal retirements and natural gas additions as of August 2015 are excluded from the totals.

116

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 9: Recommendations Wind: Add between 500 and 1,750 MW by 2033, dependent on pricing, performance and integration costs. Given the variability of wind selections in the scenarios, evaluate accelerating wind deliveries into the first 10 years of the plan if operational characteristics and pricing result in lower-cost options.

Natural Gas (Combustion Turbine and Combined Cycle): Add between 700 and 2,300 MW by 2023 and between 3,900 and 5,500 MW by 2033. The key determinants of future natural gas needs are trajectories on natural gas pricing and energy efficiency and renewables pricing and availability.

TVAs recommended planning direction affirms its commitment to a diverse resource portfolio guided by the least-cost system planning mandate. The ranges above provide a general guideline for resource selection, but the full case analysis studied in the IRP and the SEIS includes ranges much broader than shown above driven by key drivers such as significant changes in economic conditions or regulations. We believe meeting our future needs in accordance with the resource technologies and ranges in this recommendation will position TVA to continue to deliver reliable, low-cost, and cleaner power to the people of the Tennessee Valley.

117

Chapter 10 Implementation Challenges and Next Steps Inputs & Analyze & Present Scoping Re-evaluate Recommend Framework Evaluate Findings 119

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 10: Implementation Challenges and Next Steps 10 Implementation Challenges and knowledge about technology advances and changing market conditions; (2) TVA should work closely with Next Steps local power companies as energy efficiency efforts This chapter outlines some of the challenges TVA and distribution-level resources are implemented; and faces in implementing the recommendations of the IRP (3) TVA and appropriate partners should investigate study and discusses key policy considerations and additional approaches in energy efficiencies and improvements to modeling and the study process. distributed resources, considering those who cannot afford the necessary investments and recognizing 10.1 Overview of Next Steps fairness and equity for all rate payers.

In the Draft Report, we provided a high-level schedule Implementing the recommendations from the IRP for next steps that included the release of the draft IRP/

will require close cooperation between TVA, local SEIS reports and the completion of a public comment stakeholders, our Local Power Company (LPC) period. Now that the comment period is concluded, we partners and Valley electric customers particularly have moved to the final two steps in our IRP process as around deployment of additional energy efficiency shown in Figure 10.1:

resources. The success of energy efficiency depends on end-use customer participation. TVA is primarily Spring/Summer 2015 Summer 2015 a wholesale power provider and the LPCs have the relationship with most end-use customers. TVA will Identify Incorporate Target Power need to partner with LPCs and others in the region to Input Supply Mix design additional delivery mechanisms to achieve the levels of penetration envisioned in the IRP. We have a

• Review comments

• Develop study history of successful collaboration around the design

• Complete additional recommendations analyses if needed

• Prepare final and delivery of EE programs and plan to build on that

• Revise the study report and post experience. There are a number of initiatives already report

• Request TVA Board action underway both internal to TVA and in cooperation with our LPC partners seeking more effective and innovative Figure 10.1: Remaining IRP Process Steps program designs, improved performance tracking and budgeting, and enhanced delivery mechanisms.

After TVA issues the final IRP and SEIS, there is a Similarly, TVAs status as a power wholesaler 30-day waiting period before the TVA Board of complicates deployment of cost-effective renewable Directors can be asked to make a decision about resources (primarily solar). The IRP envisioned using the IRP. After the Board makes a decision, the NEPA utility-scale solar resources which can be located process is completed by issuing a Record of Decision to provide the most value to the transmission or that documents the Boards action and its basis. distribution systems. While TVA owns and operates the high-voltage transmission grid, the distribution system 10.2 Implementation Challenges is actually a grid of grids belonging to the 155 LPCs, each with its own unique characteristics and operational The Regional Energy Resource Council (RERC), after challenges. Renewable resources installed on the reviewing the recommendations in the IRP, offered distribution grid necessitate the involvement of entities in the following advice to the TVA Board focused on addition to TVA, especially the LPCs. This is especially implementing those recommendations: (1) TVA should true for small-scale distributed (rooftop) solar resources.

consider all of the analyses in the IRP and continue Although TVA did not include small-scale rooftop solar to refine input assumptions based on relevant data, as a resource option in the IRP, we did include small-120

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 10: Implementation Challenges and Next Steps scale commercial solar as an option, and we analyzed consequences for low/fixed income customers as well significant levels of distributed generation penetration in as renters or other customers that do not participate in the scenarios to help us begin to understand how the the programs, which doesnt fit with our mission. IRP increasing use of distributed generation will affect the analyses we have completed will help inform the follow-TVA power system. on planning and evaluation of the EE portfolio. TVA recognizes that EE should be a key part of our power TVA is leading an initiative with the goal of determining supply mix consistent with the findings in the IRP. We the value of distributed resources on the system. also know that program design will be a key challenge Initial efforts are focused on small-scale distributed to ensure that the broadest possible EE portfolio can (rooftop) solar, but the method is general enough to be offered through the LPCs to minimize possible bill allow for value determination for other distributed impacts on non-participants.

options. Work is ongoing, led by a team that includes technical support from the Electric Power Research We also realize that the level of electric rates and job Institute (EPRI), to develop a methodology to identify growth are critical concerns for Valley governments, site preferences on the distribution systems of the businesses and residents. The IRP uses two specific LPCs. This work, along with locational analysis already metrics for the macro-economic impacts of resource completed by TVA, will help in placement of utility- strategies. These metrics and underlying analyses scale and distributed solar in support of the IRP provide important information about future revenue recommendations. requirements that affect future rate levels and will help inform the future direction of TVAs economic Finally, it is important to note that the recommendations development program. However, none of the strategies in the IRP also include more traditional resources, had a significantly different impact from the others on primarily gas-fired, that come with their own the Valley economy. Section 7.5.7 of the SEIS provides implementation challenges in the areas of siting and more information about socioeconomic effects.

permitting both for the units themselves and for the transmission lines and gas pipelines associated with There are several other policy issues that come into them. TVA has several teams working on various play when implementing recommendations from the aspects of the siting and permitting work necessary to IRP. For example, we know that the EPAs Clean Power ensure that when these resources are needed as part Plan will be finalized at virtually the same time this of the generation portfolio, we will be better positioned report is released. We will look at that final rule more to bring add them to the resource mix. specifically to understand how the IRP can inform TVAs compliance plans, but feel the study recommendations 10.3 Policy Considerations point us in a direction to meet whatever requirements are included in that rule. Because of our unique The IRP is a resource planning study focused on business model, TVA, stakeholders, electric customers identifying a target power supply mix for TVA. In the and its Local Power Company partners will have to process of developing the cases and reviewing the collaborate in new and innovative ways to ensure that results with stakeholders, a number of policy-related this evolving resource portfolio remains reliable and issues were raised that are outside the scope of the IRP provides maximum value to all customers.

itself but will need to be considered as we move toward implementation of recommendations from the study.

10.4 Process and Modeling For example, we recognize that a commitment to Improvements significant levels of energy efficiency as part of the As this IRP cycle winds down, we anticipate resource portfolio will likely put upward pressure on undertaking the next study within five years, sooner in rates (absent a redesign), and that could have negative this period if one or more of our indicators trigger the 121

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Chapter 10: Implementation Challenges and Next Steps analysis. As part of the after-action review of this study we have identified a number of improvements to either the study process or the models we use to conduct the planning study. These include continuing to develop the modeling approach to treating EE as a selectable resource, enhancing our process for developing the scenario/strategy framework and reviewing and refining the IRP process to ensure strong stakeholder feedback remains a key component of the study.

10.5 Conclusion TVA finds considerable value in undertaking an IRP and especially appreciates the input, review and insights of individuals on the IRP Working Group and the Regional Energy Resource Council. They spent considerable time helping us develop a robust plan that meets all the criteria outlined in our objectives. TVA values their involvement and expertise on behalf of all our stakeholders in making this a better IRP.

As with any long-term resource plan that attempts to prepare for the future, TVAs IRP reflects what we know today and can reasonably expect for the coming years. TVA, and our employees across the Valley, stand ready each and every day to continue our three-part mission around energy, the environment and economic development. We will do our best to continue to serve the people of the Tennessee Valley by providing low cost, reliable power in an environmentally responsible manner while promoting economic development across the Valley.

122

Appendix A Generating Resource Cost and Performance Estimates 123

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix A A wide array of new resource options were available in independently verified for either accuracy or validity, the capacity planning expansion models for selection and no assurances are offered with respect thereto.

to meet load growth or fill resource needs. Each new This Report does not represent any endorsement of resource option has a set of unique characteristics any particular resource type, nor a guarantee that any such as capacity, construction time, book life, heat resource type is viable or can be ultimately delivered.

rate, outage rate, capital cost, variable cost, and fixed This Report covers the TVA 2015 IRP only. We make costs. Chapter 5 includes a discuusion of the resource no representations, warranties or opinions concerning options considered in the IRP. An independent third- the enforceability or legality of the laws, regulations, party, Navigant Consulting, Inc. (Navigant), reviewed rules, agreements or other similar documents reviewed and compared the TVA planning parameters used in the as part of this work. Navigant and its employees are IRP to proprietary and other industry sources to ensure independent contractors providing professional services the modeled unit characteristics and assumptions to TVA and are not officers, employees, or agents of were representative of the respective generating TVA.

technologies. This appendix contains aletter report summarizing the benchmarking efforts of Navigant as Background and Scope well as TVAs internal benchmarking efforts. As part of the 2015 IRP effort, TVA is identifying and evaluating potential new power generating and storage resources necessary to serve future load. Estimated A.1 summary Letter: Navigant values for new resource cost and performance are Benchmarking Report necessary in order to perform generation capacity Summary Letter Report on Generating Resource Cost expansion and dispatch modeling. TVA requires and Performance Estimates estimated values that are internally consistent and representative of actual values to be observed in Developed for the 2015 TVA Integrated Resource Plan practice. Parameters include performance and cost September 12, 2014 for traditional, renewable, and alternative generation technologies, and also for power storage technologies.

Navigant Consulting, Inc., (Navigant) has reviewed Estimated values are obtained from several sources and recommended cost and performance parameters including the TVA business units, the Tennessee Valley for potential new power generation and storage Renewable Information Exchange, and the IRP project resource alternatives to be considered in the Tennessee staff itself.

Valley Authority (TVA) 2015 Integrated Resource Plan (IRP) (Resource Estimates). The work was Navigants task was to review the estimated values performed for TVA under contract work authorization provided by TVA for each resource type, and, as

1. 669468 and purchase order #709838 (revised). The necessary, develop alternative values, forming a set primary deliverable was a Microsoft Excel workbook of Resource Estimates that are indicative of what can summarizing the Resource Estimates and related be expected for each resource technology within the assumptions and notes. The preliminary draft workbook Tennessee Valley geographic area. The deliverable was delivered on April 25, 2014, and the final workbook was a spreadsheet workbook of tables - one for each was delivered on June 17, 2014. resource technology - that:

This report (Report) summarizes the work scope, the

• lists the parameters and associated values provided resources and parameters reviewed, and our primary by TVA, findings at a high level. In performance of this review

• lists alternative values as available and relevant, and and Report, we have in part relied on information

• recommends specific Resource Estimates for use in provided to us by TVA and third parties. While we IRP modeling.

believe this information to be reliable, it has not been 124

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix A Technologies and Parameters Reviewed outage rate, storage efficiency, storage input demand, Power generation and energy storage resources plant overnight capital cost, transmission upgrade considered in the review included the following, which cost, total overnight capital cost, variable operating &

represent alternatives for new capacity to serve future maintenance (O&M) cost, fixed operating & maintenance load: cost (both in $ and $/kW-year), firm gas charge, and book life.

• Natural gas-fired generation Single cycle combustion turbines When relevant and reliable industry values for specific Combined cycle combustion turbines (with and parameter values were available, they were utilized for without supplemental duct firing) comparison and as a basis for any Resource Estimate.

• Coal-fired generation Notes concerning the source and reconciliation of any Pulverized coal (with and without carbon material differences were provided in the workbook.

capture and sequestration)

Integrated gasification combined cycle High-Level Findings and Recommendations (coal) (with and without carbon capture and Navigant provided recommended parameter values sequestration) and performed direct comparisons with TVA estimates

• Nuclear generation for 264 values. For about two-thirds of these, the TVA BW205 design values were determined to be consistent with the AP1000 design recommended values (meaning within 10%, measured Small modular reactors relative to the original TVA estimate). The remaining

• Energy storage one-third of the values showed numerical differences of Pumped hydro-electric storage greater than 10%, characterized here as material. Of Compressed air energy storage (CAES) the materially different values, over half - representing

• Solar photovoltaic (PV) generation 62 of the 264 values reviewed - showed differences Utility scale (both fixed-panel and tracking) greater than 20%.

Commercial scale (both small and large)

• Wind energy generation Some parameters are correlated with others, and one Onshore within the Tennessee Valley key difference in interpretation or estimation sometimes Located in Midcontinent Integrated System led to a pattern of differences across parameters.

Operator (MISO) or Southwest Power Pool Additionally, variations in underlying classification (SPP) categories (cost allocation, for example) can mean Obtained via High Voltage Direct Current that there is some compensation or offsetting in net (HVDC) transmission effects when modeling. Overall, the substantial majority

• Biomass energy generation of TVA values were determined to be consistent with Co-firing recommended values, and otherwise reasonable.

Integrated gasification combined cycle (IGCC)

Regarding natural gas-fired generating resources, for (biomass) the 48 parameter values compared, 29 (59%) of the Direct combustion at new facility TVA values were consistent with values recommended Repowering of existing facility by Navigant. Roughly one-fifth of all parameters Cost and performance parameters vary somewhat showed differences of 20% or more. The only according to generating and storage technology, systematic material difference between TVA values but each technology generally has 8-12 applicable and recommended values was in annual outage rates, characteristics or parameters for which values were where the Navigant recommendations were higher reviewed. These include summer net dependable across the board. For a given resource, parameter value capacity, summer full-load heat rate, build time, annual differences vary in terms of impact, and a number of 125

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix A potentially offsetting differences are evident. located for biomass IGCC, and there are no such plants in service.) Where comparisons were possible, The vast majority (79%) of the 66 coal resource recommendations were materially higher for heat rate, parameters compared were in agreement. For the build time, outage rate, and plant overnight capital.

parameters with material differences, there was no Some differences were due to varying assumptions systematic pattern, although some differences were about plant sizing, however, and some potentially noted for plant overnight capital costs, build time, and offsetting differences were noted for variable and fixed variable O&M. O&M.

For nuclear generation, about half of the parameter On balance for all the generating and storage resources values (15 out of 31) were found to be consistent. Most examined, the substantial majority of the proposed of the remaining values were 20% or more different TVA parameter values for which comparisons were (12 values). Generally speaking, recommended outage performed were consistent with recommended values rates, plant and total overnight capital costs, and - about two-thirds of all compared values. For those variable O&M values were materially higher than TVA parameters with material differences in values of 10%

values. or more, a number of those were to some degree offsetting within a given resource/technology.

Regarding energy storage, two-thirds of the compared parameter values were materially consistent. Each value The TVA values reviewed were provided in spring 2014, with a material difference was at least 20% different. and the summary above relates to recommendations The parameters with such differences included variable and comparisons based on the values provided at that O&M, fixed O&M (both dollars per year and $/kw-year), time. Since then, TVA has modified numerous values and book life for pumped hydro; and annual outage to be used in its IRP modeling, in part reflecting the rate, storage efficiency, and plant and total overnight outcome of this review. TVA staff was extremely helpful costs for CAES. Some potentially offsetting differences and responsive both in providing supporting information were observed. needed in the review/comparison process, and in providing useful feedback and clarification on the draft Almost all of the solar PV parameter values compared workbook deliverable and the constituent parameter were consistent. Only a single material difference was values. It is clear that TVA is striving to fairly represent identified, where the recommended value for fixed O&M all of the potential new generating resources in its IRP (small commercial rooftop solar) was materially higher.

modeling, thus laying the basis for meaningful IRP For wind energy, 16 of the 29 parameter values modeling of resource expansion alternatives.

compared (or 55%) were consistent, with about half of the remaining values showing differences greater than 20%. Recommended outage rates were materially A.2 TVA Benchmarking summary:

higher than TVA values for all three technology Optimizing Asset Decisions alternatives. Other differences varied by technology, and When evaluating how to best meet future customer some potentially offsetting effects are seen. needs for electricity, TVA optimizes decisions using least-cost planning models. These models require Biomass options show consistent parameter values in inputs on variables such as capacity amounts, upfront about one-quarter of the comparisons, with material capital costs, and fuel usage parameters, and many differences in about three-quarters of the 29 values others. The models integrate all the variables for new compared. All of materially different values are at least resources under the various scenarios (i.e., various fuel 20% different. This applies to co-firing, new direct prices, demand projections, regulatory environments, combustion, and biomass repowering of existing etc) to select expansion units that best fit the portfolio coal. (No reliable source of industry information was needs and requirements in a total least-cost manner.

126

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix A One of the key assumptions that contributes to resource Supply Unit Characteristics selection is the cost to construct a particular unit. Option1 Overnight Capital Construction and capital costs are determined from Capacity Cost2 (MW) industry experience, vendor information, benchmarking, (2013$/kW)

CT 3x 590 $738 et cetera. These costs are presented as Overnight Natural CT 4x 786 $712 Gas Capital Costs in the table. This is the cost to build the CC 2 by 1 670 $1,097 CC 3 by 1 1,005 $1,030 asset and is computed as total dollars divided by the IGCC 500 $3,845 capacity of the unit in kilowatts ($/kW). PC 1x8 PC 2x8 800 1,600

$2,908

$2,722 Coal IGCC_CCS 469 $7,286 Depending on how an assets dispatch cost compares PC 1x8_CCS 600 $6,518 PC 2x8_CCS 1,200 $6,271 to other assets in the fleet, the amount of energy PWR 1,260 $5,460 sourced from an asset may vary greatly over time. Nuclear3 APWR 1,117 $5,856 SMR 334 $8,252 For example, when natural gas prices are low, those Pump Storage 850 $2,365 Storage assets powered with natural gas serve customers with CAES 330 $1,072 Utility_Tracking 25 $1080-$2353 more energy than when natural gas prices are high. A Utility_Fixed 25 $1080-$2059 Solar4 concept that is sometimes used to compare asset costs Commercial_Small 25 $3,529 Commercial_Large 25 $2,941 is LCOE or Levelized Cost of Energy. This measure MISO 200 $1,750 divides the total cost of an asset (i.e., construction Wind4 SPP 200 $1,750 In Valley 120 $1,875 and capital, ongoing maintenance and operating, and HVDC 250 $1421-$2242 dispatch costs which are primarily fuel) by expected Biomass Direct Combustion 115 $4,357 Repowering Existing 75 $1,092 output or generation. Spill Addition 40 $2,200 Hydro Space Addition 30 $1,800 Because dispatch costs and expected output vary Run of River 25 $2,550 Res Tier 1 10 $2,076 widely across all of the IRP scenarios, LCOE is not a Res Tier 2 10 $2,911 useful metric to benchmark resource costs. A better Res Tier 3 10 $3,817 Comm Tier 1 10 $1,168 comparison, and the standard for resource planning, is Energy Comm Tier 2 10 $1,931 Efficiency to compare $/kW installed capital costs. These are the Comm Tier 3 10 $3,341 Ind Tier 1 10 $1,154 actual inputs into the capacity expansion model and the Ind Tier 2 10 $1,908 costs benchmarked by TVAs independent third-party Ind Tier 3 10 $3,302 contractor. Footnotes:

1. Supply options represent generic site build costs, except the PWR resource which represents the Bellefonte site option.

2. Overnight capital costs do not include Allowance for Funds Used During Construction Benchmarking Capital Costs (AFUDC). All options include a generic transmission upgrade costs.

3. The PWR and APWR costs are for the first unit. The SMR cost is for a twin pack.

TVA engaged an independent third-party, Navigant 4. The capital costs for solar and wind assume that tax credits expire/decrease per Consulting (NCI), to review cost and performance current federal law. Sensitivity cases on utility tracking and HVDC wind test impact of extensions (range reflects capital cost range for sensitivity analysis). Solar capital assumptions proposed for use in the 2015 IRP. NCI costs are assumed to decline over time per recent trajectories and wind capital costs increase at less than the rate of inflation.

evaluated our assumptions for various unit types along with cost assumptions for renewable resources We have also prepared a comparison of our capital developed in a collaborative effort with stakeholders.

cost assumptions from the IRP study to a recent This independent assessment found that the majority Lazard report released in September 2014 to further of assumptions proposed for the study were consistent demonstrate the reasonableness of those assumptions.

with typical values used in the industry. Many of the The capital cost data from the summary table has been remaining assumptions were modified based on NCI adjusted to match assumptions used in this Lazard recommendations prior to running the IRP cases. The report (expressed in comparable year and including data in the table presented in the preceding section financing costs). This comparison chart shows how reflect adjustments recommended by NCI.

TVAs assumptions on capital cost compare to those 127

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix A recently published by Lazard. The cost comparisons are system. In addition, footnotes are provided to explain generally consistent given Lazards study is based on variations for each asset type.

nation averages and TVAs costs are specific to the TVA

• Source: Lazards Levelized Cost of Energy Analysis - Version 8.0. September 2014

• • Source: TVAs 2015 Integrated Resource Plan. Assumptions are in 2015$/kW capital costs and include financing costs.

1. The high end range for TVA represents a small commercial solar unit and the low end represents a large commercial solar installation.

2. Lazards high end represents a solar tracking unit and the low end represents a fixed-tilt solar unit. TVAs low end represents the solar sensitivity unit and the high end represents the solar tracking unit. Solar capital cost assumptions decline at 3.5% a year through 2020 and then remain flat through 2029.

3. TVAs high end represents the HVDC option; the low end represents the wind sensitivity analysis. These costs do not include transmission wheeling charges (similar to Lazards).

4. The TVA biomass unit assumptions are modeled after a recent utility-scale plant. The basis for the Lazard assumption is not specified.

5. The high end costs include carbon capture and storage.

6. The low end represents PWR (BLN), the midpoint is an AP1000, and the high end represents an SMR.

128

Appendix B Assumptions for Renewables 129

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix B Modeling Approach for Wind factors. For modeling purposes, TVA assumed the MISO and SPP wind had a 40 percent capacity factor,

& Solar Options the HVDC option originating from Oklahoma had a 55 Wind and solar resources have unique operating percent capacity factor, and the In-valley option had a characteristics that are different from thermal and 30 percent capacity factor.

other more traditional resources. To properly account for the contribution from these intermittent resources, The HVDC project has a 55 percent annual capacity the energy contribution is represented using hourly factor due to the availability of wind in Oklahoma and energy profiles that are imported into the model, and the newer technology of the wind turbines, which were the seasonal capacity of these resources is represented assumed to be GE 1.7-100 wind turbines at a height by a computed Net Dependable Capacity (NDC) value. of 80 meters. This capacity factor is much higher than The annual capacity factor of the hourly energy profiles TVAs existing wind contracts in other locations. The are also computed to ensure the total amount of energy chart below shows the range of capacity factors:

is comparable to industry benchmark sources. This appendix discusses the methodology TVA has used to determine both the energy profiles and NDC values for Wind Capacity Factors HVDC (future) 55%

wind and solar options that are considered in the IRP. 46%

42%

MISO (future) 40%

SPP (future) 40% TVA Wind Modeling 40%

38%

PJM 37% SPP Generation from wind is weather and location 37%

36% MISO dependent, and not dispatchable like more conventional 35%

32%

resources. Therefore, utilities need to develop a In Valley (future) 30%

reasonable representation of the output from wind for 20%

0% 10% 20% 30% 40% 50% 60%

use in long-range planning models. This wind shape is based on actual data collected from specific sites, or modeled data using wind turbine design assumptions.

TVA uses data from 3TIER to develop the planning Determining the Wind Net assumptions around wind shape and capacity factor Dependable Capacity (NDC) for use in the IRP. A typical week hourly shape for Planners must determine how much wind generation each month was developed by 3TIER for each wind is likely at the system peak hour so that appropriate option. Once a shape has been selected, the amount credit can be given to wind resources when computing of energy produced can be determined and a capacity the capacity/load balance to determine if the required factor computed (actual generation expressed as a reserve margin has been met in a given year. That percentage of maximum possible generation). capacity credit value is called the Net Dependable Capacity (NDC).

Determining Wind Capacity Factors The NDC is applied to the nameplate capacity and is TVA used actual results from its wind contracts (1500 used by the expansion model to meet the 15 percent MW in Oklahoma, Illinois, Kansas and Iowa), simulated reserve margin requirement. It is calculated in a six-step and actual data for the in-valley sites, and proposals for process and repeated for annual, summer and winter various projects to determine the capacity factors for periods for both the wind and solar resources.

the wind resources options included in the IRP. Since each of the options originates from different regions, TVA used a region-specific estimate for annual capacity 130

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix B

1. For each year of the sample period, select the 60 summer season (June-Sept). 50

• T VA focuses this process on the summer 40 Percent because the system peak occurs in that season.

30

2. Identify the top 20 load days of the summer.

20

• Using the top 20 days in the summer produces a 10 distribution of wind generation in the sample year.

3 Anomaly (oC)

3. Find the peak hour for each of those top 20 days. 2 1

4. Determine the wind generation for each of those 20 0 1

peak hours and convert to capacity factors. 2 3

• These generation values are converted to 1980 1984 1988 1992 1996 2000 2004 2008 2012 Time Mean capacity factor = 30.9%

capacity factors by dividing the hourly generation by the nameplate capacity of the wind resource.

Figure B1: Example of Wind Monthly-mean variability of net

5. Choose the 25th percentile of this capacity factor power capacity by 3TIER distribution.

• T VA selects the 25th percentile value to ensure that wind generation at the time of the system The Annual NDC was calculated as 14 percent based peak will exceed this value 75 percent of the time. on a portfolio view of all current wind contracts to

6. Then these 25th percentile annual capacity factor capture the diversity of location across the different values are averaged across all the years of the states of the region. This 14 percent NDC was used for all wind options. Specific sites of future wind in sample to produce the NDC used for planning MISO, SPP or in-valley is unknown, so it would be purposes.

inappropriate to assume a better or worse NDC at this time. A more specific NDC would be incorporated into For this IRP study, TVA repeated this calculation using the wind portfolio NDC calculation once specific sites 16 years of data ranging from 1998 to 2013.

are known. TVA did not consider over-subscription The simulated hourly wind generation was provided by contracts where transmission is limited to a level below 3TIER, a third-party company specializing in renewable the nameplate rating of the wind capacity which tends energy assessment and forecasting. The wind to improve both the annual capacity factor and the NDC generation was based on simulation of TVAs existing rating. The costs associated with the wind projects wind contracts in MISO, SPP, and PJM as well as a site modeled in the IRP do not reflect oversubscription; in in Kansas near where the HVDC site is proposed. 3TIER TVAs experience with several existing wind contracts, assessed the long-term variability of the wind for each this over-subscription provision is negotiated in the site in a retrospective analysis of historical wind speed terms and costs of a particular contract and is not easily and power. These data points were derived from a comparable to industry benchmarks.

mesoscale Numerical Weather Prediction (NWP) model that was statistically calibrated to match the observed Solar Modeling data during the measurement period at the height of the towers. An example of the variability of the wind net Similar to wind, solar resources are also weather and power is shown in Figure 1. location dependent. Modeling of solar options in the IRP proceeds in a similar fashion to wind, and requires determination of solar shapes, capacity factors and NDC values. Solar data was provided by members of the TVRIX stakeholder group who commissioned 131

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix B Clean Power Research (CPR) to provide TVA with the solar energy profiles for 26 sites across the Tennessee Valley shown in the map below. CPR provided SolarAnywhere data for 15 plus years of consistent, validated, time-series irradiance measurements that provided the historical basis for the NDC, capacity factors and hourly energy patterns.

Figure B3: Solar Fixed Axis and Utility Tracking Capacity Factors by Month Solar NDC values The determination of the NDC for solar resources utilizes the same process described for wind resources.

The figure below shows the range of NDC values for solar fixed-axis systems computed using data covering the period 1998-2013:

Figure B2: Sites across Tennessee Valley with historical solar irradiance data supplied by CPR Solar Capacity Factors Using the data supplied through CPR, TVA determined that annual capacity factors are 20 percent for the fixed axis and 23 percent for the single-axis tracking option.

The monthly capacity factors vary as shown in the following chart.

Figure B4: NDC by hour of the top 20 peak load days of Summer 1998-2013 132

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix B In the summer, TVA normally has a peak load at 5:00 p.m. EST, but can also see a peak load between the hours of 2:00 p.m. and 6:00 p.m. EST. The 25th percentile of solar generation of those hours would occur at 5:00 p.m. or 6:00 p.m. EST as the sun is setting. Therefore, the summer NDC was set at 50 percent for fixed axis, including utility scale, small and large-commercial. The utility tracking option has a 68 percent NDC.

All solar options have a 0 percent NDC during the winter, since TVAs winter peaks normally occur at 5:00 a.m. EST when solar is not available.

133

Appendix C Distributed Generation Evaluation Methodology 135

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix C Background commercial. To represent the load profiles associated with DG penetration in these customer groups, an Distributed generation (DG) is broadly defined as on-site natural gas plant was assigned to industrial generation that is produced on the distribution grid customers, while small solar was utilized for residential/

network. IRP strategies primarily focus on central commercial customers. Although an assortment of station or utility-scale resource planning options, DG technologies could realistically be deployed, these therefore the contributions from DG represented in this two technologies serve as useful proxies to represent IRP are primarily captured in scenario development.

DG across customer classes. Figure 1 shows how In the context of the selected IRP scenarios, DG is DG penetration, along with other uncertainties, was more narrowly defined as customer-driven, demand-represented across the various scenarios and how side generation which results in utility load reductions.

the customer classes discussed were applied to DG Additionally, DG was subdivided into two customer penetration.

categories: Industrial customers and residential/

Figure C1: DG Market Segments and Penetration Levels across IRP Scenarios 136

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix C Methodology The primary, or leading, driver was another IRP scenario uncertainty, CO2 regulation. CO2 regulation was viewed Different methodologies were applied to forecast DG as the most likely driving force to impact future levels penetration growth differences between customer-of renewable energy growth, both from a utility- and led Industrial and Residential/Commercial market customer-led perspective. Therefore, CO2 assumptions segments. Although the approaches differ, DG were first applied to determine utility-driven, national penetration levels across all scenarios directly impact renewable energy adoption rates. National renewable other scenario uncertainties, specifically commodities, energy adoption rates in turn drove customer-led DG electricity prices and loads.

renewable growth. Finally, national levels of DG growth Residential/Commercial Distributed were then appropriately scaled down to reflect regional Generation Penetration DG growth in the Tennessee Valley region.

To begin this sequential analysis, first, CO2 uncertainty Residential/Commercial DG penetration is defined as levels were correlated to traceable source data. The TVAs residential and commercial customers energy Reference case along with the GHG 10, GHG 15, and consumption that is self-generated by renewable GHG 25 cases of the 2013 U.S. Energy Information energy. Renewable energy encompasses all traditional Administration (EIA) Annual Energy Outlook were renewable resource types (solar, wind, hydro, biomass, chosen as the source material. EIA and TVA CO2 price geothermal). For the purposes of this analysis, all assumptions were correlated to interpolate reasonable Residential/Commercial DG is assumed to be solar PV.

national renewable adoption levels by 2040, EIAs end of To determine Residential/Commercial DG adoption analysis period.

rates, a sequential set of linear drivers were applied.

Figure C2: Correlation of IRP CO2 uncertainty values to EIA source data 137

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix C The national renewable energy adoption levels, applied to national DG growth rates to ensure relative adapted from EIA data, were then adjusted to develop consistency between IRP scenarios. The percentages corresponding national DG penetration levels. EIAs of renewable growth as a function of total renewables reference case and GHG 15 growth curves were growth were determined as described in Figure 3.

Figure C3: Development of National Renewable DG Penetration Levels Figure 4 charts national renewable energy growth rates marginal gap between each set of solid and dashed as a percentage of total generation for the electric sector lines indicates the quantities of DG penetration occurring (utility-led only) and including DG (customer-led). The across the various IRP scenarios.

Figure C4: National Renewable Energy Adoption Levels (Utility-led and DG) 138

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix C Finally, to translate national renewable DG adoption of Residential/Commercial DG penetration result in levels to TVA regional DG levels, a 75% multiplier was varying levels of TVA load loss as shown in Figures 5 &

applied to represent regional differences. As mentioned 6. Cumulative and annual capacity growth levels are also previously, Residential/Commercial DG penetration was shown to provide a sense of total and incremental growth assumed to be 100% solar PV to serve as a relative levels of renewable DG.

proxy for renewable DG growth. These growing levels Figure C5: Residential/Commercial DG Adoption Levels (by 2040)

Figure C6: Residential/Commercial DG Adoption Levels (Annual) 139

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix C Industrial Distributed Generation Penetration To accent the Residential/Commercial DG penetration analysis, industrial DG was also applied to reflect DG growth beyond renewable energy, namely from natural gas pursued by industrial customers. Industrial DG was applied across two scenarios: The Distributed Marketplace and Growth Economy. The following assumptions were applied across each scenario:

Distributed Marketplace Scenario: Assumed 50% of industrial customer load was lost to DG over the study period (representing 10% of total TVA load).

Growth Economy Scenario: Assumed 10% of industrial customer load with high steam needs were lost to DG over study period (0.6% of total TVA load).

Conclusion DG, both nationally and at the TVA level, is included in the 2015 IRP study as demand-side generation that is customer-driven (outside of utility involvement),

and results in a reduction to utility load. Industrial DG is load loss occurring from natural gas projects while Residential/Commercial DG is represented by solar PV projects. Residential/Commercial DG is driven by CO2 regulation and national renewable and DG growth rates. The resulting combination of both Industrial and Residential/Commercial DG growth rates are captured across the various IRP scenarios as load loss. The Distributed Marketplace scenario represents the most extreme load loss on the TVA system projected to be caused by DG.

140

Appendix D 2015 IRP: Modeling Energy Efficiency 141

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 1 Energy Efficiency in the IRP was developed in the 2015 IRP to employ blocks of energy efficiency impacts and costs that reflect the One of TVAs goals is to provide low-cost, clean, and characteristics of existing programs but do not require reliable electric power to consumers and it does this by the development of detailed program designs.

maintaining a diverse set of energy resource options.

Energy efficiency and demand-side management TVA energy efficiency programs typically address the programs have been part of TVAs energy portfolio major components of energy consumption in the areas since the late 1970s and include incentive programs, of lighting, building shell improvements, HVAC/control price structure changes and educational efforts upgrades, industrial process changes and a newly to encourage awareness and smart consumer identified approach, voltage regulation. Assumptions choices. TVA continues to offer programs under the on changes to load shapes and reductions in demand EnergyRight Solutions brand that include residential, and energy can be derived from the results of existing commercial, industrial, renewable (end-use-generation), programs and projected for blocks which serve as demand response and educational/outreach initiatives. proxies of yet-to-be-defined future programs, as well as continuation of existing efforts. This approach greatly TVA is currently engaged in evaluating new programs, reduces the staff time needed to develop modeling delivery and impacts as it continues to evolve the inputs and, if designed in small enough blocks, affords demand side management portfolio. These programs the opportunity for the model to select an optimum level help reduce reliance on power purchases from other of energy efficiency on an annual incremental basis to suppliers, reduce power production environmental match the given strategy and scenario inputs in each impacts and mitigate utility bill pressures by providing model run.

benefits to consumers and the TVA system. Refining the characterization of energy efficiency in models will enhance potential for success and assist in keeping 1.1 TVA Energy Efficiency Program electricity costs low. Characteristics A variety of delivery methods are used to deliver 1.0 Energy Efficiency Modeling programs to end-use consumers. Residential TVAs 2011 IRP used discrete energy efficiency programs are delivered through various channels, portfolios matched to specific strategies for the which include: up-stream incentives to manufacturers modeling effort. The portfolios consisted of detailed and installers; promotion and administration of TVA-program designs for individual energy efficiency and designed programs through local power companies demand response programs that outlined annual costs (LPCs); turnkey administration of TVA-designed and demand/energy reductions across a 30-year programs through third-party vendors; and design, planning horizon. In the 2011 IRP, energy efficiency promotion and administration of programs by LPCs.

consisted of over 20 individual program designs, and In the commercial and industrial sectors, programs the portfolios were considered must run components are offered to large customers directly served by TVA.

of their respective strategies. The majority of promotion and administration duties for LPC commercial and industrial customers are handled Two significant drawbacks to this approach were the by TVA field staff and a third-party administrator under lack of flexibility in the must run nature of the energy contract to TVA with the collaboration and coordination efficiency contribution for each strategy design and the of the LPCs. The Conservation Voltage Reduction (CVR) staff time required to develop program details for efforts program requires the participation of the individual that would not necessarily launch for several years. LPCs and does not involve promotion or participation To address these deficiencies, a different approach by individual end-use customers.

142

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Energy efficiency programs impact the system to the same system load shape drivers as the system reduce costs through demand reduction as well as load. The variable EE shape over the majority of the day energy savings. As can be seen in Figure 1, on a typical (Figure 1) and year round EE (Figure 2) demonstrates peak day, the energy efficiency resource provides load that EE resembles the cycling nature of an intermediate matching to TVAs overall load requirements for that day. resource like a natural gas combined cycle unit.

This is due to the EE resource portfolio design having Figure D1: Energy Efficiency Performance on a Typical Peak Summer Day (2023) 143

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Looking across a typical year (Figure 2), energy intermediate resources. The shapes differ by sector with efficiency resources provide fuel and operating cost the residential sector following weather patterns more savings by lowering demand across all months of closely than the commercial or industrial sectors.

the year and offsetting the need for base load and Figure D2: Energy Efficiency Monthly Profile (2023)

In the block designs used for the 2015 IRP, the The Block Concept residential sector has a defined capacity factor of 57%; Traditional supply side resources have the following the commercial sector has a capacity factor of 68%; characteristics:

and the industrial sector has a capacity factor of 80%.

These capacity factors are comparable to other base

• Capacity and energy - typically a known size in MW load and intermediate duty resources with capacity and MWh respectively factors typically greater than 40%.

• Install cost - typically a bus bar $/kW

• Construction lead time - years to build from initial project consideration 2.0 Model Inputs and Assumptions

• Operational characteristicsmust run number of For energy efficiency to be a selectable resource option hours per year, heat rate (fuel efficiency), capacity in the optimization model, energy efficiency block factor, etc.

characteristics must be developed that are conceptually

• Service Life - years comparable to other supply side resources.

144

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D TVA developed energy efficiency options in a similar use load shapes for current EE program within those fashion. Blocks of energy efficiency impacts and load sectors. For example, a residential EE block has a load shapes were constructed for three market sectors: shape similar to the weighted average of six residential residential, commercial and industrial. Each sector has customer programs annual load shapes (Table 1). Each a load shape similar to the weighted average of the end- of the sectors is comprised of three pricing tiers Table D1: Tier, Sector, and Block Hierarchy Residential Programs Block Weight Load shapes, contribution percentages and other program characteristics of the blocks are based on New Homes 12%

the detailed Program Design Templates developed Self Audit 2%

as part of the FY 2015 TVA budget. Cost and impact In Home Energy Evaluation 20%

estimates for the blocks use an average steady-state, Manufactured Homes 16%

fully-operational estimate of program designs rather than Heat Pump 10%

trying to reflect the variation of higher initial/end-of-life eScore 40%

program costs.

Industrial Programs Block Weight Blocks were grouped by sector based on commonality Tailored Solutions for Industry 54%

of market and similarity of load shape. Each sectors Custom Industrial 10%

block is composed of different TVA EE programs that Standard Rebate 36%

carry different weights. Weighting for each sector is found in Table 2 and is based on the past and projected contributions of the various programs. Commercial Programs Block Weight Custom Commercial 10%

Standard Rebate 90%

Table D2: Weighting of EE Programs 145

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Each block was developed to be 10MW and between associated set of modeled data including the on-peak 50-72 GWh in size. This size was chosen to provide capacity reduction and associated operational like flexibility for model selection by being a proxy for EE characteristics, which include an 8,760-hour load programs. Current programs each have a net-to-gross shape consistent with the sector end-use load shape.

(NTG) design assumption (Table 3) which accounts Since each EE block occurs at the end use level, the for free-ridership and other aspects of program characteristics are grossed up for transmission and efficacy and were weighted in the development of distribution losses to create a supply side equivalent the sector blocks. Each existing program also has an when modeled with other resource options.

Program Sub Program Lifespan NTG R1 New Homes 15 64%

R2 Kit & Self Audit 6 75%

R3 IHEE 18 80%

R4 Manufactured Homes (VHP) 15 80%

R5 Heat Pump Program 15 67%

R9 ESTAR Man. Homes 15 80%

R14 eScore 18 80%

C1 Tailored Solutions 10 70%

C2 Custom Industrial 10 70%

C3 Custom Commercial 15 76%

C10 Standard Rebate Commercial 15 69%

C11 Standard Rebate Industrial 15 74%

Table D3: Net to Gross ratios and Lifespans for the EE programs within sectors 2.0.1 Pricing accordance with the weighting table referenced above.

Once the operational characteristics of each sector Tier 2 and Tier 3 consist of programs yet-to-EE block was developed, pricing tiers were identified.

be-developed (some of which represent as-yet-Pricing tiers were developed to reach more deeply into undeveloped technologies) and pricing was based the pool of potential savings in the Valley; additional on the step function increase found in Table 4. The costs would need to be incurred to expand delivery breakpoints and step function increase for each of the system infrastructures and encourage greater sectors were developed through consultation with the participation. Blocks within Tier 1 were priced at managers of existing TVA programs and supporting the current portfolio of programs for each sector in consultants.

146

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Average Unweighted Increases Relative to Base Tier 2 Residential Industrial Commercial ERS Incentives 50% 70% 70%

ERS Variable Costs 26% 70% 70%

ERS Fixed and Low Variable 15% 10% 10%

ERS Other 19% 70% 70%

Tier 3 Residential Industrial Commercial ERS Incentives 100% 200% 200%

ERS Variable Costs 51% 200% 200%

ERS Fixed and Low Variable 25% 20% 20%

ERS Other 29% 200% 200%

Table D4: Tier Step Changes The steps in cost for tiers 2 and 3 are similar to a supply Levelized costs for each of the tiers within the sectors stack in which programs with the highest potential are can be found in Figure 4. Energy efficiency programs the lowest cost programs, programs with mid-potential are compared against a greenfield combined cycle are mid cost programs, and programs with lowest to plant, which energy efficiency tends to closely resemble mid potential are at a high program cost. As benefits are based on capacity factor. All block costs, including exhausted from of the lowest cost programs, it moves incentives, escalate at inflation (1.8% per year) so that down the supply stack to the next lowest cost program. energy efficiency becomes cheaper over time in real terms.

Figure D3: Levelized EE block cost comparison ($/MWh) compared to a greenfield combined cycle plant over time 147

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 2.0.2 Quantity to expand the delivery infrastructure from one year to Much like the supply side counterparts, EE programs the next and the expectation of increasing consumer/

also have operational-like limits on the ramp rate, or participant awareness.

year-over-year growth, based on startup time and development of infrastructure. The limits are driven by 2.0.3 Block Life For supply side resources, power contracts expire and program development, customer awareness, market power plants reach end of life and are retired. Similarly, penetration, participant acquisition and many other energy efficiency resources have useful lifespans (e.g.

customer and market factors.

light bulbs burn out, lighting systems must be upgraded Through 2018, TVA has a required energy efficiency and heating and cooling equipment must be replaced).

performance as part of a 2011 EPA settlement. These For the 2015 IRP, TVA has assumed each block of programs are embedded into the TVA annual business energy efficiency resource can be replaced with a planning cycle and are being modeled as must run similar block at the available price for that sectors resources for the IRP resource selection model. block.

For the selectable blocks, TVA assumed that growth The lifespans for each of the sector blocks were of the total delivered blocks each year cannot exceed developed based on program composition within 25% when compared to the previous year for years each sector, current program lifespan assumptions, 1-5. For years 6-15, the growth rate was limited to 20%, and measure lifespan assumptions used by industry and then is at 15% per year thereafter. These limits standards.

were based on the ability of TVA and program partners Block Design Parameters Final Residential Commercial Industrial MW per Block 10 10 10 GWh per Block 50 59 72 Growth Rate (Yr 1-5) 25% 25% 25%

Growth Rate (Yr 6 -15) 20% 20% 20%

Growth Rate (Yr > 16) 15% 15% 15%

Max Incremental Blocks per Year Tier 1 9 4 4 Max Incremental Blocks per Year Tier 2 7 4 2 Max Incremental Blocks per Year Tier 3 8 4 2 Max Incremental Blocks per Year Tier Total 22 12 8 Lifespan Tier 1 (Years per Block) 17 15 12 Lifespan Tier 2 (Years per Block) 13 13 10 Lifespan Tier 3 (Years per Block) 13 13 10 Table D5: Block Characteristics for each sector 148

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Each of the blocks in the different tiers and sectors There are two basic ways to incorporate the EE shapes has differing lifespans. Tier 1 block lifespans were into System Planning models:

determined using a weighted average based on existing programs. Tier 2 and 3 blocks are made up 1) As a load modifier: the energy efficiency shapes of programs yet to be developed as well as some are subtracted from the original system load and the potentially unknown technologies, therefore the same resulting net system load is fed into the models load estimates could not be applied. TVA instead used input.

an industry average lifespan for each of the sectors. 2) As a resource (selectable or non-selectable):

Residential and commercial tier 2 and 3 blocks have a consistent with how all other supply side resources 13 year lifespan and industrial tier 2 and 3 blocks a 10- are modeled (i.e. nuclear, coal, gas, hydro, etc.). EE year lifespan (Table 5). resources point to a defined energy pattern (i.e. the EE load shape) similar to a solar resource.

3.0 Energy Efficiency Methodology Each approach has pros and cons and the best within System Planning approach depends on modeling architecture and modeling objectives. For the 2015 IRP, TVA elected 3.1 Planning Approach to use the model-as-a-selectable-resource approach.

This allows TVA to model selectable EE resource Energy Efficiency (EE) programs have two basic impacts units for full optimization. Energy efficiency is non-that are relevant to planners:

dispatchable and operates similarly to a number of

1) A

 voided energy calculation - Energy not consumed other non-dispatchable generation resources in that means fuel not burned, resulting in savings in variable system operators cannot directly control it based on costs. Further, since program impacts are felt at the system needs. There are no variable operations and meter, they also avoid transmission and distribution maintenance (VOM) costs nor an emissions penalty (thermal) losses which can average 6.5% by the time (CO2 costs). Key input parameters are monthly avoided energy reaches an end user. capacity, $/kW (cost divided by summer peak kW) and an hourly energy pattern.

2) A

 voided capacity calculation - Capacity is avoided, because reduced electricity demand translates into reduced need for incremental capacity additions. 3.2 New Approach to Modeling from 2011 IRP Using EE program design parameters, hourly demand profiles are developed via engineering models, such TVA is taking a new approach to energy efficiency as eQuest, and then calibrated through program modeling to allow energy efficiency to compete evaluation. Inputs to the models include occupancy/ with other resources within each of the IRP cases.

utilization profiles and weather data. Each models This will create an opportunity to allow for full key output is an 8,760 hourly profile of a before end portfolio optimization, to better gauge the impacts use shape and an after efficient end use shape that of the programs in different situations, and to better are subtracted to get the net savings. The net savings demonstrate the value proposition for the resource.

shape is then regressed on weather and calendar The 2011 IRP study did not contain energy efficiency variables, revealing the relationship between savings as a selectable resource. Several different EE portfolios and temperature, day of week, season, etc. The model were scheduled as load modifiers in various scenarios.

is then forecast forward using TVA weather and load There was no supply stack concept in those portfolios, forecast as inputs. The final result is an hourly energy which in effect reduces model flexibility and limits efficiency savings forecast synched to the TVA load model outcome. TVAs new modeling approach for forecast.

149

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D energy efficiency as a competitive resource attempts 3.2.1 Comparability to Other Supply Side to enhance model visibility and potential impacts with Resources regards to least-cost optimization. Energy efficiency unit characteristics must be developed that are comparable to other supply side resources. Supply side characteristics that feed the capacity expansion model can be found in Table 6 and are compared against the energy efficiency power plant.

SUPPLY SIDE COMPARISON New Coal Comm EE Ind EE Res EE New CC New CT AP 1000 w/ CCS Year Available 2014 2014 2014 2019 2018 2028 2026 Outage Rate Heat Rate Fuel Costs Fuel CAGR CO2 Costs CO2 CAGR (starts in 2022)

O&M costs O&M Escalation Transmission Contingency Cost Project Contingency Cost Capital Costs Escalation of Capital Capacity Factor Technology Shifts Table D6: Resource Characteristic Comparison with EE 150

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D For supply side resources in the IRP, unit performance combined cycle plant constructed in 2033 possesses is not expected to be 100%. This delivery risk is the same heat rates, ramp rates, cost of construction captured in an outage rate for the unit. There is not (escalated for inflation), etc. as one constructed in 2015 a comparable outage rate for the modeled energy because we do not know what the future technology efficiency blocks; rather, the modeling approach will be. However, in EE blocks TVA allows for an assumes the block to be operationally available 100% assumption of technological improvements based on of the time. Efficiency is dependent on variables such the history of EE deliveries over the past 30 years.

as equipment reliability and service life, operating conditions, etc., that would impact operability similar to an outage rate. 3.3 Modeling Uncertainty The block design approach is novel and fits well with In addition to outage rates, Table 7 shows the potential model architecture, but introduces some uncertainties uncertainties that are captured in cost for supply around design and delivery that are unique relative to side resources. Examples include a carbon dioxide other resources. Design uncertainty is introduced by emission penalty, fuel cost uncertainty, project cost the creation of prescribed blocks of EE meant to reflect contingencies and cost escalation uncertainties. bundles of programs over time. Delivery uncertainty exists around claimed versus evaluated measures, the One item unique to TVAs modeling approach on ability to deliver and implement programs though TVAs EE blocks is related to technological improvements.

155 different local power companies, and risk around Traditional supply side resources do not reflect EE deliveries relative to future codes and standards.

advancements in technology over time. For example, a Uncertainty Design Deliver Proxy Programs in Blocks LPC Delivery Risk Measure Lifespand Blending Codes and Standards Unchanging Shapes Claimed vs. Evaluated Table D7: Design and Delivery Uncertainties Uncertainty of all types exists with supply side 3.3.1 Design Uncertainty resources and is modeled in different ways in the Since the modeled energy efficiency blocks are proxies analysis, but typically manifest itself as cost. For energy for technologies and programs not yet developed, efficiency, TVA considers the two primary categories there is uncertainty in their design and future of uncertainty mentioned above to remain comparable composition. Blocks in the study are modeled as 10 with other supply side resources. In addition, certain MW resources with a defined load shape by sector variables can be captured directly or indirectly in (residential, commercial, and industrial). The virtual the stochastic analysis performed in the study. Key nature of energy efficiency compared with the tangible, uncertainties are discussed in more detail below. physical attributes of supply side resources necessarily introduces a level of uncertainty around certain key 151

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D design attributes. for overestimation or underestimation of energy savings.

With respect to the energy shapes, the capacity 3.3.1.1 Measure Life Uncertainty expansion model uses a repeating annual energy Measure life or Effective Useful Life (EUL) is the median pattern for each block to the end of the lifespan. As number of years that the measure after installation is programs die off before the expected lifespan, they are expected to be in place and operable. This includes replaced with the same technology at no cost until the equipment life which is the number of years installed end of a defined block life.

equipment will be operational before it fails, and Figure 5 demonstrates how this applies to a 14-year measure persistence which takes into account residential audit program within a residential block that business turnover, failure or early retirement of the has a lifespan of 17 years. Several technologies die off installed equipment.

before the end of life, but the block assumes the energy Each of the energy efficiency blocks contains different is still there because the technology is replaced with like programs with different EULs. Tier 1 blocks contain kind (solid black line). Notice there is an overstatement currently developed TVA programs, and the block of energy for years 6 -17. For the technologies where lifespan was determined using weighted averages. contribution ends prior to end of block life, it is replaced Block lifespans for tiers 2 and 3 were approached with a similar block and contributes with the same differently. Because tiers 2 and 3 contain undeveloped energy pattern for the remaining block life. The risk in technologies and programs, industry average standards these cases is that we are overstating energy (by having were used for the different sectors lifespan. the same energy contribution every year) and underestimating costs (by assuming technology is Since the energy efficiency blocks are a mix of differing replaced at no cost to TVA). The blending of programs technologies with differing life measures, potential exists into blocks creates unique challenges for resource planning in that an average lifespan can create resource adequacy challenges in a particular year.

Figure D4: Example of a residential audit program modeled as part of a residential block 152

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 3.3.1.2 Fixed Shape Uncertainty accounts for attribution) in order to get the net Each of the energy efficiency sectors has a fixed end realization rate. In most studies reviewed by TVA, net use 8,760 load shape. For modeling convenience, realization rates tend to be less than one, although in the blocks are assumed to have an unchanging some jurisdictions realization rates reflecting actual composition over time, even though this is unlikely. performance exceeding planned savings have been TVAs stochastic modeling around the overall TVA load achieved.

shape partially addresses some of this risk but does not address the uncertainty around the shape of the Examples of lower realization rates (i.e. realized program block designs changing over time as programs and impacts) can be seen in more mature markets such technologies evolve and as the low hanging fruit of EE as California, Con Edison and Indiana where there are is picked off. extensive measurement and verification (M&V) data.

They illustrate that risk exists with regard to energy and 3.3.2 Delivery Risk Uncertainty capacity impacts even in these more mature markets. A lot of this is attributable to operational issues, calculation 3.3.2.1 Local Power Company Delivery methods, and inappropriate baselines. TVA does not expect to repeat industry experience with regards to Uncertainty claimed and evaluated measured discrepancies since Unlike conventional assets that can be constructed, TVA has different market drivers. However, TVA can operated and maintained directly by TVA, there is learn from their experience by noting that there is risk more uncertainty around the ability to implement EE around these future program assumptions. In the IRP resources in the Tennessee Valley because of the case, the risk is primarily around our ability to realize multiple parties involved and coordination around deliveries over the 20-year study period on programs end use customer adoption. The end use providers that as-yet have not been designed, undergone M&V are made up of the participants and the local power and been refined. This uncertainty increases over time.

companies. There are currently 155 Local Power Companies (LPCs) in the TVA region consisting of municipal utility companies and cooperatives. Since 3.3.2.3 Delivery Risk: Codes and Standards TVA is not the end-use provider there is risk in how the TVAs modeling approach assumes that selectable 155 local power companies would vary in their delivery EE resource deliveries are over and above any future of EE programs. Additionally, TVA and the LPCs need tightening of efficiency codes and standards. Currently, to establish delivery mechanisms to facilitate larger EE known codes and standards (C&S) are reflected in the deployment across the region and this takes time and load forecast in the IRP, and future increases in C&S are resources which may be different than a comparable, not assumed.

vertically integrated utility might experience. Treating EE as a supply side resource means that TVA believes delivery risk will diminish over time as it is available and deliverable in the same way that delivery mechanisms are developed and refined with a conventional resource is, and this creates a risk the LPC customers. A 10% adjustment is applied to around C&S tightening. A conventional gas turbine reflect delivery risk for years 1 through 5. At year 6, this for example, delivers MWs regardless of whether new adjustment begins declining at 2% per year. efficiency standards reduce TVAs sales in year 15 of the study. For EE, there is a risk that future tightening of C&S would reduce the amount of EE available to 3.3.2.2 Realization Rate Delivery Uncertainty deploy in the market or increase the cost of deploying The gross realization rate is the ratio of measured the EE resource in the future. As baseline efficiency energy reduction (actual) to claimed energy reduction requirements increase, then either the supply (volume)

(planned). The gross realization rate is typically of EE must decrease or the cost of the next series of multiplied by the Net to Gross ratio (a ratio which 153

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D measures must increase. TVAs current EE modeling our service territory and build the infrastructure with assumes that over the 20-year study period that TVA our LPC partners. This is represented as a 10% cost programs can be developed to exceed whatever the adder in years 1-5 that begins to decline in year 6. The then-current standards may be. other uncertainties around block design and delivery risk uncertainty are initially zero but begin to grow over time, starting in year 6. The total planning adjustment 3.4 Recognizing Design & Delivery is shown in Figure 6 and grows to 30 percent over the Uncertainty: Planning Factor out years of the study. This planning adjustment reflects Adjustment the fact that the further out in the future one goes, the more uncertain these proxy EE blocks are. The Why do all these performance issues and uncertainty planning adjustment is an approximation, not a precise matter? Dynamically modeling energy efficiency as a calculation, but is meant to reflect how uncertainties resource means that all variables, including resource increase over time.

costs, shapes and uncertainties, significantly influence the modeled needs for base load, intermediate and In this construct, the uncertainties manifest as cost peaking generation. There are several possible ways to in the model. The alternate approach was to restrict address this uncertainty analytically, including carrying volumes available in the out years, but TVA chose higher planning reserves, but each increases overall to keep the volumes consistent to test the model plan costs. To address these uncertainties and allow boundaries. Uncertainty manifesting as cost has certain energy efficiency to compete on the same playing field modeling advantages and also allows volumes to be as a supply side resource, a planning adjustment factor unconstrained. In many case results we can see full was made to reflect the two categories of design and selection of EE blocks occur, even in the out years with delivery uncertainty. the uncertainty adjustments, which allows for a more robust range of case results.

Initially, the primary risk TVA faces is delivery risk, largely around the ability to implement programs across Figure D5: Planning Factor Adjustment over time 154

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 3.5 Recognizing Uncertainty: varied by demand and weather pattern (i.e. load shape) distributions modeled in the analysis. Traditional supply Stochastic Analysis side resources have other factors that can change both While the planning adjustment captures design and their cost and generation levels: demand, fuel, O&M, delivery uncertainty, TVAs analytical approach also capital costs, CO2 emission penalties, etc. All such considers stochastic analysis on several key inputs. uncertainties manifest as cost in the model. Table 8 lists O&M cost escalations are stochastically varied in the the direct and indirect stochastic variables for several analysis using the same distributions as other O&M supply side resources as a comparison to energy costs. Resulting system cost impacts are indirectly efficiency.

direct indirect Stochastic Variables Diesels CT CC Coal Nuclear Hydro Solar Wind Energy Efficiency Gas price Coal price Oil price CO2 allowance price Electricity price Hydro generation Plant availability Load shape year Electricity demand O&M costs Interest rates Capital cost Table D8: Indirect and Direct Stochastic Variables Even after accounting for the planning factor cost uncertainty. The much narrower EE uncertainty uncertainty, EE blocks have a significantly lower range band is driven by the design and delivery uncertainties of uncertainty than a comparable combined cycle previously covered, stochastic variations on O&M cost plant as shown in Figure 7. The uncertainty bands and the indirect effects of the stochastic draws on the around combined cycle costs are much wider due overall system load shape.

to fuel, emissions, O&M, capacity factor and capital 155

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D Figure D6: Uncertainty bands in $/MWh for each of the EE sector blocks as compared to a greenfield combined cycle plant 156

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 3.6 Costs after Planning Adjustment Looking at the LCOE over time with the uncertainty adjustment, most of the EE blocks remain less Levelized Cost of Energy (LCOE) is a common metric expensive than a natural gas combined cycle unit to allow comparisons of total resource costs reflective through the IRP study period. Only Residential Tier 3 of capital costs, asset lives and expected fuel costs.

has block costs that are higher in the beginning and Looking at the comparison in Figure 8, EE compares end of the study period than a comparable combined favorably with other TVA resources in 2015.

cycle.

Figure D7: Levelized Cost comparisons in 2015 (2015$/

MWh)

Figure D8: Levelized cost comparison ($/MWh) of EE tiers in 2015 and 2033 Figure D9: Levelized Cost Comparison by Sector through time 157

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix D 4.0 Next Steps Energy efficiency modeling for the IRP was a collaborative effort across TVA and with stakeholders.

Modeling energy efficiency as a competitive resource introduces additional uncertainties around design and delivery that are unique from other traditional resources.

TVAs approach accounts for these uncertainties with a planning adjustment, which is hoped to refine over time as programs are developed, measured and verified. The modeling framework chosen for use in the 2015 IRP has produced a robust set of results that demonstrate the value energy efficiency brings to the portfolio, including an assessment of the outcome for cases that test the boundaries for EE. TVAs next step is to develop an internal business process to leverage this dynamic approach in resource planning and to revisit the assumptions behind some of the fundamental parameters.

158

Appendix E Capacity Plan Summary Charts 159

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Capacity & Energy Expansion over the planning horizon. The capacity is in gigawatts, which is 1,000 megawatts, and is based on the Results Appendix summer net dependable capacity value or the amount The capacity expansion plans are shown below by of capacity that TVA plans to have available to meet strategy. The capacity graphics show the total capacity summer peak firm requirements.

grouped by resource type (i.e., nuclear, hydro, coal, etc.)

Total Capacity Expansion Plans Strategy A: The Reference Plan GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

50 Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 1 1 1 2 2 1 1 2 2 2 EE 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 Gas 10 13 13 15 16 10 12 12 13 15 11 14 14 15 16 10 12 11 11 12 10 12 11 12 13 Renewables 0 1 1 2 3 0 1 1 2 2 0 1 2 3 4 0 1 2 3 4 0 1 1 1 2 Coal 11 8 8 7 7 11 8 8 7 7 11 8 8 8 8 11 8 7 5 5 11 8 7 5 5 Hydro 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 Strategy B: Meet an Emission Target GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 50 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 1 1 1 2 2 1 1 2 2 2 EE 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 Gas 10 13 13 15 16 10 12 12 13 15 11 13 14 15 16 10 12 11 11 13 10 12 11 12 13 Renewables 0 1 1 2 3 0 1 1 2 2 0 1 2 3 4 0 1 2 3 4 0 1 1 1 2 Coal 11 8 8 7 7 11 8 8 7 7 11 8 8 8 8 11 8 7 5 5 11 8 7 5 5 Hydro 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 160

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Strategy C: Market Supplied Resources GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 50 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 1 1 1 2 2 1 1 1 1 2 EE 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 Gas 10 13 12 13 15 10 12 11 12 13 11 13 13 13 16 10 12 11 11 12 10 12 11 11 11 Renewables 0 1 1 2 3 0 1 1 2 2 0 1 2 3 4 0 1 2 3 4 0 1 1 2 2 Coal 11 8 8 8 8 11 8 8 8 8 11 8 8 8 8 11 8 7 6 5 11 8 8 7 7 Hydro 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 Draft Strategy C: Market Supplied Resources GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 50 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 2 2 2 2 2 1 1 1 2 2 2 2 2 2 2 1 1 1 2 2 1 1 1 2 2 EE 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 0 1 2 3 3 Gas 10 13 13 16 10 10 12 12 13 15 11 13 13 14 15 10 12 11 11 12 10 12 11 11 11 Renewables 0 1 1 2 3 0 1 1 2 2 0 1 2 3 4 0 1 2 3 4 0 1 1 2 2 Coal 11 8 8 8 8 11 8 8 7 7 11 8 8 8 8 11 8 7 6 5 11 8 8 7 6 Hydro 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8

• The original construct of the draft strategy C contained shorter term power purchase agreements.

The revised strategy C only included 20-year commitments. See section 7.1.2 for further information.

161

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Strategy D: Maximize EE GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 50 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 EE 0 1 2 4 5 0 1 2 4 5 0 1 2 4 5 0 1 2 4 5 0 1 2 4 5 Gas 10 13 13 14 15 10 12 12 12 13 11 14 14 15 15 10 12 11 11 12 10 12 11 11 12 Renewables 0 1 1 2 2 0 1 1 2 2 0 1 2 3 3 0 1 2 3 3 0 1 1 1 1 Coal 11 8 8 7 7 11 8 8 7 7 11 8 8 7 7 11 8 6 5 4 11 8 7 5 5 Hydro 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 4 4 4 4 4 4 4 4 4 4 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 Strategy E: Maximize Renewables GW, SND Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 50 45 40 35 30 25 20 15 10 5

0 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 DR 1 1 1 2 2 1 1 2 2 2 2 1 2 2 2 1 1 1 2 2 1 1 1 1 2 EE 0 1 2 3 3 0 1 2 3 2 0 1 2 3 3 0 1 2 3 3 0 1 2 2 2 Gas CC 5 7 6 6 6 5 7 6 6 6 5 7 6 6 6 5 7 6 6 6 5 7 6 6 6 Gas CT 5 6 6 6 7 5 6 6 6 7 5 6 7 8 9 5 6 6 6 6 5 6 6 6 6 Renewables 0 1 3 4 5 0 1 3 4 5 0 2 3 5 6 0 1 3 4 5 0 1 2 4 4 Coal 11 8 8 7 7 11 8 7 6 5 11 8 8 7 7 11 8 6 5 4 11 7 6 5 4 Hydro 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Nuclear 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 7 8 8 8 8 162

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Below are the total energy charts that correspond nuclear, hydro, coal, etc.) over the planning horizon and to the capacity expansion plans above. The energy are in terawatt hours, which is a 1,000 gigawatt hours.

charts show total energy grouped by resource type (i.e.,

Total Energy Plans Strategy A: The Reference Plan TWh Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 12 17 17 1 5 12 17 17 1 5 12 17 18 1 5 12 17 18 1 5 12 16 16 Gas 23 31 30 36 37 23 26 24 30 38 25 25 28 31 29 21 13 12 17 18 20 22 17 19 25 Renewables 6 6 8 10 16 6 6 6 8 6 6 7 10 14 25 6 22 25 27 32 6 6 6 8 5 Coal 58 41 39 37 34 57 40 39 37 36 58 51 45 44 39 59 30 27 23 20 57 36 31 27 25 Hydro 16 17 17 17 17 16 16 17 17 17 17 18 18 18 18 16 17 17 17 17 16 16 16 17 16 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 Strategy B: Meet an Emission Target TWh Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 12 17 17 1 5 12 17 17 1 5 12 17 18 1 5 12 18 18 1 5 12 16 16 Gas 23 31 30 36 38 23 26 24 30 38 25 24 28 33 27 21 14 13 17 19 20 22 17 19 25 Renewables 6 6 8 10 15 6 6 6 8 5 6 7 10 12 28 6 21 24 27 31 6 6 6 8 5 Coal 58 41 39 37 34 57 40 39 37 36 58 51 45 45 38 59 30 27 23 20 57 36 31 27 25 Hydro 16 17 17 17 17 16 16 17 17 17 17 18 18 18 18 16 17 17 17 17 16 16 16 17 16 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 163

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Strategy C: Market Supplied Resources TWh Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 12 17 18 1 5 12 17 17 1 6 14 19 19 1 5 13 18 19 1 5 11 17 17 Gas 23 30 29 32 36 23 26 23 25 32 25 23 26 29 28 21 14 12 16 17 20 22 16 16 21 Renewables 6 6 9 11 10 6 6 7 9 7 6 7 10 14 25 6 22 24 26 31 6 6 6 8 6 Coal 58 41 39 40 40 57 40 39 40 40 58 51 45 44 39 59 30 27 24 21 57 36 33 29 29 Hydro 16 17 17 17 17 16 16 17 17 17 17 18 18 18 18 16 17 17 17 17 16 16 16 17 16 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 Draft Strategy C: Market Supplied Resources TWh Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 12 17 17 1 5 12 17 17 1 6 14 19 19 1 5 13 18 19 1 5 12 16 16 Gas 23 31 29 32 35 23 26 24 28 36 25 23 26 28 27 21 14 12 16 17 20 22 16 17 21 Renewables 6 6 8 10 12 6 6 7 9 6 6 8 10 15 25 6 22 24 27 31 6 6 6 8 6 Coal 58 41 40 40 40 57 40 39 37 37 58 51 45 44 39 59 30 27 23 21 57 36 33 29 28 Hydro 16 17 17 17 17 16 16 17 17 17 17 18 18 18 18 16 17 17 17 17 16 16 16 17 16 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68

• The original construct of the draft strategy C contained shorter term power purchase agreements. The revised strategy C only included 20-year commitments. See section 7.1.2 for further information.

164

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Strategy D: Maximize EE Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

TWh Current Outlook Stagnant Economy Economic Growth Decarbonized Future Distributed Marketplace 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 13 24 29 1 5 13 24 29 1 5 13 24 29 1 5 13 24 29 1 5 13 24 29 Gas 23 30 29 32 36 23 26 23 24 29 25 25 28 32 25 21 13 14 15 18 20 22 16 15 17 Renewables 6 6 7 9 6 6 6 6 8 6 6 6 9 11 25 6 22 25 27 25 6 6 6 6 3 Coal 58 42 39 35 34 57 40 38 35 33 58 51 45 39 33 59 30 23 19 17 57 36 31 26 23 Hydro 16 17 17 17 17 16 16 17 17 17 17 18 18 18 18 16 17 17 17 17 16 16 16 16 16 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 Strategy E: Maximize EE Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5:

TWh Distributed Marketplace Current Outlook Stagnant Economy Economic Growth Decarbonized Future 200 150 100 50 0

2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 2015 2020 2025 2030 2033 EEDR 1 5 12 17 17 1 5 12 16 16 1 5 12 17 17 1 5 12 17 18 1 5 11 13 12 Gas 23 21 17 18 19 23 19 14 15 17 25 16 17 20 22 21 14 14 18 18 20 17 9 9 11 Renewables 6 18 24 33 38 6 16 22 30 36 6 19 26 34 40 6 21 25 29 34 6 16 20 25 29 Coal 57 37 35 30 28 56 37 32 29 26 57 47 41 35 32 58 30 23 19 17 57 31 25 22 19 Hydro 17 18 18 18 18 17 18 18 18 18 17 18 18 18 18 17 18 18 18 18 17 18 18 18 17 Nuclear 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 55 67 69 67 68 165

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E Below are the total capacity additions on a year by year gigawatts and is grouped by resource type (i.e., nuclear, basis. The data is shown in summer net dependable hydro, coal, etc.) over the planning horizon.

1A, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.1 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.7 0.9 1.1 1.4 1.6 1.8 2.0 2.1 2.3 2.4 2.8 Gas 10.3 10.5 10.5 12.0 12.0 12.9 12.9 12.9 13.1 13.0 12.9 13.0 13.9 14.0 14.2 14.6 14.6 14.8 15.5 15.9 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.8 2.7 DR 1.5 1.6 1.6 1.6 1.6 1.6 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.8 35.1 35.5 35.9 36.5 36.3 36.7 37.1 37.6 38.0 38.4 38.9 39.4 39.9 40.6 40.9 41.5 41.9 42.6 1B, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.1 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.6 0.8 1.1 1.3 1.5 1.7 1.9 2.1 2.2 2.3 2.7 Gas 10.3 10.5 10.5 12.0 12.0 12.9 12.9 12.9 13.1 13.1 12.9 13.1 14.0 14.1 14.6 14.6 14.6 14.9 15.6 16.1 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.8 2.7 DR 1.5 1.6 1.6 1.6 1.6 1.6 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.8 35.1 35.5 35.9 36.5 36.3 36.7 37.1 37.6 37.9 38.4 38.9 39.4 40.2 40.6 40.9 41.5 41.9 42.8 1C, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.1 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.9 7.9 7.9 7.9 7.9 7.9 7.5 7.5 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.6 0.8 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.8 3.0 Gas 10.3 10.5 10.4 11.9 11.9 12.9 12.9 12.9 12.9 12.1 11.8 11.9 12.0 12.0 12.2 12.5 12.5 12.8 13.5 14.7 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.8 2.8 DR 1.5 1.6 1.7 1.7 1.7 1.6 1.6 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.8 35.1 35.5 35.9 36.5 36.3 36.7 37.1 36.8 37.1 37.7 38.2 38.6 39.2 39.8 40.2 40.7 41.2 42.6 1D, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.1 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.1 2.2 Gas 10.3 10.5 10.4 11.8 11.7 12.2 12.6 12.7 13.0 13.2 13.0 13.0 13.9 13.8 13.8 13.9 13.8 13.8 14.2 14.7 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.1 2.4 2.8 3.1 3.4 3.8 4.2 4.4 4.6 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.8 35.1 35.5 35.9 36.1 36.3 36.7 37.1 37.6 37.9 38.5 38.9 39.4 39.9 40.6 41.1 41.6 42.0 42.7 1E, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.1 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.8 4.8 4.8 4.8 4.8 4.8 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.7 1.1 1.4 1.6 1.9 2.2 2.5 2.8 3.2 3.5 3.8 4.1 4.4 4.8 5.0 5.4 Gas 10.3 10.2 10.2 11.8 11.6 12.2 12.2 12.2 12.2 11.8 11.5 11.5 12.3 12.2 12.2 12.3 12.2 12.2 12.7 13.1 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.7 2.7 DR 1.5 1.5 1.5 1.5 1.5 1.4 1.3 1.3 1.3 1.4 1.4 1.4 1.4 1.4 1.5 1.6 1.6 1.7 1.8 1.8 Subtotal 34.6 34.8 35.1 35.5 35.9 36.5 36.5 37.0 37.4 37.6 37.9 38.4 38.9 39.4 39.9 40.6 41.0 41.5 41.9 42.6 166

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E 2A, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.7 1.0 1.1 1.3 1.5 1.7 1.8 1.9 Gas 10.1 10.3 10.2 11.5 11.4 12.2 12.2 12.2 12.2 12.3 12.2 12.2 13.1 13.0 13.1 13.2 13.2 13.6 14.4 15.1 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.7 2.6 DR 1.5 1.5 1.5 1.5 1.5 1.3 1.3 1.3 1.4 1.4 1.4 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.8 1.7 Subtotal 34.4 34.5 34.7 34.9 35.0 35.5 35.3 35.5 35.7 36.0 36.3 36.7 37.1 37.6 38.0 38.7 39.1 39.5 39.9 40.6 2B, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1.9 2.0 Gas 10.2 10.3 10.2 11.5 11.4 12.2 12.2 12.2 12.2 12.3 12.2 12.2 13.1 13.0 13.1 13.3 13.4 13.8 14.6 15.3 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.7 2.7 DR 1.5 1.5 1.5 1.5 1.5 1.3 1.3 1.3 1.4 1.4 1.4 1.6 1.7 1.7 1.8 1.9 1.8 1.7 1.6 1.6 Subtotal 34.5 34.5 34.7 34.9 35.1 35.5 35.3 35.5 35.7 36.0 36.3 36.8 37.2 37.7 38.1 38.8 39.2 39.7 40.1 40.9 2C, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.9 7.9 7.9 7.9 7.9 7.9 7.5 7.5 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.2 2.3 Gas 10.1 10.3 10.2 11.5 11.4 12.2 12.2 12.2 12.0 11.9 11.5 11.5 11.5 11.5 11.7 12.0 12.0 12.2 13.0 13.5 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.7 2.7 DR 1.5 1.5 1.5 1.5 1.5 1.3 1.3 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.4 34.5 34.7 34.9 35.0 35.5 35.3 35.6 35.7 36.0 36.1 36.7 37.1 37.5 38.0 38.7 39.0 39.5 40.0 40.6 2D, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.0 Gas 10.1 10.2 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.7 11.5 11.7 12.5 12.3 12.4 12.5 12.2 12.2 12.6 13.0 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.1 2.4 2.8 3.1 3.4 3.8 4.2 4.4 4.6 DR 1.4 1.5 1.6 1.7 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 1.8 Subtotal 34.4 34.5 34.9 34.9 35.3 36.0 35.8 36.0 36.2 36.0 36.2 36.7 37.1 37.6 38.1 38.7 39.1 39.6 40.0 40.7 2E, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.7 5.7 5.7 5.7 5.7 5.7 5.2 5.2 Renewables 0.3 0.3 0.4 0.4 0.6 0.9 1.3 1.6 1.8 2.1 2.4 2.7 3.0 3.4 3.7 4.0 4.3 4.6 4.9 5.2 Gas 10.2 10.1 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.6 11.5 11.5 11.7 11.5 11.6 11.6 11.6 11.7 12.3 13.0 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.6 2.6 2.6 2.5 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.5 1.6 1.7 1.8 1.8 1.9 1.9 1.9 1.9 1.8 Subtotal 34.4 34.5 34.9 34.9 35.3 36.4 36.5 35.6 36.0 36.0 36.4 37.1 37.1 37.5 38.0 38.7 39.1 39.5 40.0 40.7 167

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E 3A, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.9 7.9 7.9 7.9 7.9 7.9 7.5 7.5 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.2 2.3 2.6 2.8 3.3 3.5 3.5 Gas CT 5.3 5.4 5.7 5.7 5.7 5.7 6.5 6.5 6.8 7.3 7.2 7.2 7.3 8.1 8.3 8.7 8.7 8.7 9.3 10.1 Gas CC 5.3 5.2 5.1 6.9 6.9 7.5 7.2 7.2 7.2 6.5 6.5 6.5 6.5 5.8 5.8 5.8 5.8 5.8 5.8 5.8 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.6 1.9 2.0 2.2 2.4 2.5 2.7 2.8 2.9 2.8 DR 1.5 1.6 1.7 1.7 1.9 1.9 1.7 1.7 1.8 1.8 1.8 1.9 1.9 1.9 1.9 1.8 1.8 1.8 1.8 1.8 Subtotal 34.9 35.2 35.9 36.6 37.1 37.5 37.8 38.2 38.9 39.2 39.6 40.2 40.7 41.1 41.7 42.4 42.8 43.3 43.8 44.5 3B, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.9 7.9 7.9 7.9 7.9 7.9 7.5 7.5 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.7 0.9 1.1 1.3 1.5 1.8 2.0 2.2 2.4 2.6 2.8 3.2 3.5 3.6 Gas CT 5.3 5.4 5.7 5.7 5.7 5.7 6.1 6.5 6.6 7.2 7.2 7.2 7.3 8.1 8.3 8.7 8.7 8.7 9.3 9.8 Gas CC 5.3 5.2 5.1 6.8 6.9 7.5 7.2 7.2 7.2 6.5 6.5 6.5 6.5 5.8 5.8 5.8 5.8 5.8 5.8 5.8 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.6 1.9 2.0 2.2 2.4 2.5 2.7 2.8 2.9 2.8 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.8 1.8 1.8 1.8 1.8 Subtotal 34.9 35.2 35.9 36.6 37.1 37.5 37.6 38.4 38.9 39.3 39.7 40.2 40.7 41.1 41.7 42.4 42.8 43.3 43.8 44.4 3C, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.8 4.8 4.8 4.8 4.8 4.8 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.9 7.9 7.9 7.9 7.9 7.9 7.5 7.5 Renewables 0.3 0.3 0.4 0.4 0.5 0.7 0.9 1.1 1.3 1.5 1.7 2.0 2.2 2.4 2.6 2.8 3.1 3.5 3.6 3.5 Gas CT 5.3 5.4 5.7 5.7 5.7 5.7 5.9 6.1 7.2 7.0 6.9 7.0 7.1 7.2 7.3 8.3 8.3 8.3 9.0 10.5 Gas CC 5.3 5.2 5.1 6.8 6.8 7.2 7.2 7.2 6.4 5.7 5.7 5.7 5.7 5.7 5.7 5.1 5.1 5.1 5.1 5.1 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 1.9 2.1 2.3 2.4 2.6 2.8 2.9 2.9 2.9 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.9 35.2 35.9 36.6 37.1 37.4 37.6 38.1 38.9 38.5 38.9 39.4 39.9 40.4 40.9 41.6 42.1 42.6 43.0 44.4 3D, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.4 2.5 2.8 3.3 3.3 Gas CT 5.3 5.4 5.7 5.7 5.7 5.7 6.5 6.5 6.8 7.4 7.2 7.2 8.1 8.0 8.7 8.9 8.7 8.7 8.7 9.3 Gas CC 5.3 5.2 5.1 6.8 6.8 7.5 7.2 7.2 7.2 6.5 6.5 6.5 6.5 6.5 5.8 5.8 5.8 5.8 5.8 5.8 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.1 2.4 2.8 3.1 3.4 3.8 4.2 4.4 4.6 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.9 35.2 35.9 36.6 37.1 37.5 37.8 38.2 38.9 39.2 39.6 40.2 40.7 41.2 41.7 42.4 42.9 43.5 43.8 44.6 3E, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.8 1.2 1.7 1.9 2.3 2.6 2.9 3.2 3.6 3.9 4.2 4.5 4.8 5.1 5.5 5.7 Gas CT 5.3 5.4 5.7 5.7 5.7 5.7 5.7 5.7 5.8 6.4 6.6 6.5 7.4 7.4 7.6 7.9 7.9 7.9 8.4 8.8 Gas CC 4.6 4.5 4.6 6.3 6.1 6.5 6.5 6.5 6.5 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 5.8 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.8 2.7 DR 1.5 1.5 1.6 1.6 1.6 1.5 1.4 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.8 1.6 1.6 1.6 1.7 1.8 Subtotal 34.2 34.5 35.2 35.9 36.4 36.8 36.9 37.5 38.2 38.8 39.6 40.2 40.7 41.2 41.7 42.4 42.9 43.3 43.8 44.5 168

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E 4A, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.4 5.4 5.4 5.4 5.4 5.4 5.0 5.0 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.3 3.7 Gas 10.4 10.1 10.2 11.4 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 11.6 11.5 11.4 11.5 11.4 11.5 12.2 12.3 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 1.9 2.1 2.3 2.4 2.6 2.8 2.9 2.9 2.9 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.7 1.8 1.8 1.8 1.8 1.8 Subtotal 34.6 34.5 34.9 35.0 35.2 35.9 36.2 35.2 35.6 35.3 35.8 36.3 35.8 36.1 36.5 37.1 37.4 37.8 38.2 38.7 4B, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.4 5.4 5.4 5.4 5.4 5.4 5.0 5.0 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.9 1.1 1.3 1.5 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.3 3.7 Gas 10.4 10.1 10.2 11.4 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 11.6 11.5 11.4 11.5 11.4 11.5 12.2 12.5 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 1.9 2.1 2.3 2.4 2.6 2.8 2.9 2.9 2.9 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.7 1.8 1.8 1.8 1.7 1.6 Subtotal 34.6 34.5 34.9 35.0 35.2 35.9 36.1 35.2 35.6 35.3 35.8 36.3 35.8 36.1 36.5 37.1 37.4 37.8 38.2 38.7 4C, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.8 4.8 4.8 4.8 4.8 4.8 4.8 4.8 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.7 5.7 5.7 5.7 5.7 5.7 5.2 5.2 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 1.1 1.3 1.5 1.7 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.5 3.7 Gas 10.4 10.1 10.2 11.4 11.2 12.2 12.2 12.2 11.5 10.7 10.7 10.7 10.9 10.7 10.7 10.8 10.7 10.8 11.4 11.8 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.0 2.2 2.4 2.5 2.7 2.9 3.0 3.0 3.0 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.5 34.9 35.0 35.2 35.9 36.3 35.4 35.0 34.7 35.3 35.9 35.8 36.1 36.6 37.1 37.4 37.8 38.2 38.7 4D, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 5.8 5.8 5.8 5.8 5.8 4.6 4.6 4.6 4.6 4.6 4.6 4.2 4.2 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.3 3.3 Gas 10.3 10.3 10.2 11.3 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 12.0 11.8 11.7 11.7 11.4 11.4 11.7 12.2 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.1 2.4 2.8 3.1 3.4 3.8 4.2 4.4 4.6 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.6 34.5 34.9 35.0 35.4 36.1 36.4 34.6 35.0 34.7 35.3 35.9 35.8 36.1 36.6 37.2 37.4 37.9 38.2 38.9 4E, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 5.8 5.8 5.8 5.8 5.8 4.6 4.6 4.6 4.6 4.6 4.6 4.2 4.2 Renewables 0.3 0.3 0.4 0.5 0.6 0.8 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3.1 3.4 3.7 4.0 4.3 4.6 4.9 Gas 10.4 10.1 10.2 11.3 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 11.7 11.5 11.4 11.4 11.4 11.4 11.8 12.2 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.6 1.9 2.1 2.3 2.4 2.5 2.7 2.8 2.9 2.8 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.7 1.7 1.9 1.7 1.7 1.7 1.8 Subtotal 34.6 34.5 34.9 35.0 35.3 36.2 36.2 34.6 35.0 34.9 35.4 36.1 35.8 36.1 36.6 37.1 37.5 37.9 38.2 38.8 169

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix E 5A, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.4 5.4 5.4 5.4 5.4 5.4 5.0 5.0 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.8 1.0 1.2 1.4 1.6 1.7 1.8 Gas 10.0 10.1 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.9 11.5 11.5 12.4 12.2 12.2 12.2 12.2 12.2 12.6 13.0 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.6 2.6 2.6 2.5 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.6 1.7 1.7 1.7 1.7 1.6 1.6 1.7 1.7 Subtotal 34.2 34.2 34.6 34.5 34.8 35.5 35.3 34.1 34.3 34.3 34.4 34.6 34.8 35.0 35.3 35.7 35.9 36.1 36.3 36.6 5B, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.4 5.4 5.4 5.4 5.4 5.4 5.0 5.0 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.8 1.0 1.2 1.4 1.6 1.7 1.8 Gas 10.0 10.1 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.9 11.5 11.5 12.4 12.2 12.2 12.2 12.2 12.2 12.6 13.0 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.6 2.6 2.6 2.5 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.6 1.6 1.7 1.7 1.7 1.7 1.6 1.6 1.7 1.7 Subtotal 34.2 34.2 34.6 34.5 34.8 35.5 35.3 34.1 34.3 34.3 34.4 34.6 34.8 35.0 35.3 35.7 35.9 36.1 36.3 36.6 5C, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 8.0 8.0 8.0 8.0 8.0 7.0 7.0 7.0 7.0 7.0 7.0 6.6 6.6 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.7 1.0 1.1 1.3 1.5 1.7 1.8 1.9 Gas 10.0 10.1 10.2 11.2 11.2 12.2 12.2 12.2 11.5 10.7 10.7 10.7 10.9 10.7 10.6 10.6 10.6 10.6 10.9 11.1 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.6 1.8 2.0 2.2 2.3 2.5 2.7 2.7 2.7 2.6 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.4 1.5 1.5 1.5 1.5 1.5 1.5 1.6 1.7 Subtotal 34.2 34.2 34.6 34.5 34.8 35.5 35.3 35.5 34.9 34.4 34.7 34.9 34.8 35.0 35.3 35.7 36.0 36.2 36.3 36.6 5D, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.2 4.2 4.2 4.3 4.2 4.3 4.3 4.3 4.3 4.3 4.3 4.4 4.4 4.4 4.4 4.4 4.4 4.4 4.4 Coal 11.7 11.3 10.3 9.1 9.1 8.4 8.0 6.6 6.6 6.6 6.6 6.6 5.4 5.4 5.4 5.4 5.4 5.4 5.0 5.0 Renewables 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.8 0.9 1.0 Gas 9.9 9.9 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 11.9 11.8 11.8 11.7 11.4 11.4 11.4 11.6 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.4 1.7 2.1 2.4 2.8 3.1 3.4 3.8 4.2 4.4 4.6 DR 1.5 1.6 1.7 1.8 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 Subtotal 34.2 34.2 34.9 35.0 35.4 36.1 35.9 34.7 34.9 34.4 34.8 35.1 34.8 35.1 35.4 35.7 35.8 36.3 36.3 36.8 5E, SND GW 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 Nuclear 6.7 6.7 7.9 7.9 8.0 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 8.3 Hydro 4.0 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 4.7 Coal 11.7 11.3 10.3 9.1 9.1 8.4 7.1 5.8 5.8 5.8 5.8 5.8 4.6 4.6 4.6 4.6 4.6 4.6 4.1 4.1 Renewables 0.3 0.3 0.4 0.4 0.6 0.7 0.8 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3.0 3.3 3.6 3.9 4.1 4.4 Gas 10.0 9.9 10.2 11.2 11.2 12.2 12.2 12.2 12.2 11.5 11.5 11.5 11.7 11.5 11.4 11.4 11.4 11.4 11.5 11.6 EE 0.1 0.1 0.2 0.3 0.4 0.6 0.7 0.9 1.1 1.3 1.5 1.7 1.8 2.0 2.1 2.1 2.1 2.1 2.0 1.9 DR 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.4 1.5 1.6 Subtotal 34.2 34.3 34.9 34.9 35.3 36.1 35.1 34.1 34.5 34.3 34.9 35.4 34.8 35.1 35.3 35.6 36.0 36.3 36.2 36.6 170

Appendix F Method for Computing Environmental Metrics 171

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F Process Method In developing the criteria for the environmental impact The environmental impact metrics can be grouped into metrics, TVA wanted to create a set of metrics two broad categories:

representative of the trade-offs between energy resources rather than identifying a single resource with Scoring metrics - these metrics will be used in the best environmental performance. By considering strategy scorecard to assess the performance of a air, water and waste in the IRP scorecard, coupled given set of portfolios created by modeling that strategy with the broader qualitative discussion of anticipated across the scenarios used in the study.

environmental impacts in the EIS, a robust comparison Reporting metrics - will be computed and included of the environmental footprint of the planning strategies in the IRP report as informational or supplemental better informed the selection of the recommended measures to help clarify or expand on the insights.

strategy.

Three environmental impact metrics for air, water and waste were selected for scoring and two, air and waste, for reporting metrics. The scoring metrics are shown in Figure 1.

Scoring Metric Definition CO2 Avg Tons The annual average tons of CO2 emitted over the study period Water Consumption The annual average gallons of water consumed over the study period The annual average quantity of coal ash, sludge & slag projected based on Waste energy production in each portfolio Figure F1: Scoring Metrics Category Scoring Metric Formula CO2 Average Annual Tons of CO2 Emitted

=

(MMTons) During Planning Period Environmental Water Consumption Average Annual Gallons of Water

=

Stewardship (Million Gallons) Consumed During Planning Period Waste Average Annual Tons of Coal Ash and Scrubber

=

(MMTons) Residue During Planning Period Figure F2: Scoring Metric Formulas 172

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F The two reporting metrics are shown in Figure 3.

Reporting Metric Definition The CO emissions expressed as an emission intensity; computed by CO Intensity 2 2 dividing emissions by energy generated A measure of the quantity of spent nuclear fuel that is projected to be Spent Nuclear Fuel Index generated based on energy production in each portfolio Figure F3: Reporting Metrics The formulas for the reporting metrics are shown in Figure 4.

Category Reporting Metric Formula CO2 Intensity Tons CO2 (2014-2033)

=

(Tons/GWh) GWh Generated (2014-2033)

Environmental Stewardship Spent Nuclear Fuel Index Expected Spent Fuel Generated

=

(Tons) During Planning Period Figure F4: Reporting Metric Formulas 173

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F Strategy Performance: Air Impact Metric CO2 Scoring metric results:

80,000

Annual CO2 Emissions in Thousand

70,000

60,000 1A

50,000 2A

Tons

40,000 3A

30,000 4A

20,000 5A

14

16

18

20

22 24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Year

Strategy A-The Reference Plan

80,000

Annual CO2 Emissions in Thousand

70,000

60,000

1B

50,000 2B

Tons

40,000 3B

30,000 4B

20,000 5B

14

16

18

20

22 24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Year

Strategy B-Meet an Emission Target 174

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F

80,000

Annual CO2 Emissions in

70,000

60,000 1C

Thousand Tons

50,000 2C

40,000 3C

30,000 4C

20,000

5C

14 16 18 20 22 24 26 28 30 32 20 20 20 20 20 20 20 20 20 20 Year

Strategy C-Focus on Long-term, Market Resources

80,000

Annual CO2 Emissions in Thousand

70,000

60,000

1D

50,000 2D

Tons

40,000 3D

30,000 4D

5D

20,000

4 6 8 0 2 4 6 8 0 2

20 1 20 1 20 1 20 2 20 2 20 2 20 2 20 2 20 3 20 3 Year

Strategy D-Maximize Energy Efficiency 175

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F

80,000

Annual CO2 Emissions in Thousand

70,000

60,000

1E

50,000 2E

Tons

40,000 3E

30,000 4E

20,000 5E

14 16

18 20 22 24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Year

Strategy E-Maximize Renewables CO2 Reporting metric results:

1,000

Annual CO2 Emissions Rate in

900

800 1A

700

600 2A

lbs/mwh

500 3A

400

300 4A

18 28 5A

4 6 0 2 4 6 0 2

20 1 20 1 20 20 2 20 2 20 2 20 2 20 20 3 20 3 Year

Strategy A - The Reference Plan 176

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F

1,000

Annual CO2 Emissions Rate in

900

800 1B

700

600 2B

lbs/mwh

500 3B

400

300 4B

2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 5B

Year

Strategy B-Meet an Emission Target

1,000

Annual CO2 Emissions Rate in

900

800 1C

700

600 2C

lbs/mwh

500 3C

400

300 4C

5C

14 16 18 20 22 24 26 28 30 32

20 20 20 20 20 20 20 20 20 20 Year

Strategy C-Focus on Long-term, Market Resources 177

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F

1,000

Annual CO2 Emissions Rate in

900

800 1D

700

600 2D

lbs/mwh

500 3D

400

300 4D

5D

14

16

18

20

22 24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Year

Strategy D-Maximize Energy Efficiency

1,000

Annual CO2 Emissions Rate in

900

800 1E

700

600 2E

lbs/mwh

500 3E

400

300 4E

5E

14

16

18

20

22 24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Year

Strategy E-Maximize Renewables Air Impact Metric Observations:

• CO2 emissions vary largely by scenario but decline over time for all strategies

• Strategies A,B and C have similar CO2 emission profiles across the scenarios, coming in about 3 percent above Strategy D and about 10 percent above Strategy E

• Strategy E achieves the lowest intensity at 296 tons/

GWh, which is about 10 percent lower than A, B and C and about 8 percent lower than D 178

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F Strategy Performance: Water Impact Metric Scoring metric results:

75,000

Annual Water

1A

65,000

2A

Consump0on in

55,000

45,000 3A

Million Gallons

14

16

18 20 22 24 26 28 30 32

20 20 20 20 20 20 20 20 20 20 4A

Year 5A

Strategy A-The Reference Plan

70,000

Annual Water Consump0on

65,000

1B

60,000

2B

55,000

in Million Gallons

50,000 3B

45,000 4B

14

16 18 20 22 24 26 28

30 32

20 20 20 20 20 20 20 20 20 20 5B

Year

Strategy B-Meet an Emission Target

70,000

Annual Water

65,000 1C

60,000

2C

Consump0on in Million

55,000

50,000 3C

45,000

Gallons

4C

2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 5C

Year

Strategy C-Focus on Long-term, Market Resources 179

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F

70,000

Annual Water

65,000

1D

60,000

Consump0on in Million

55,000 2D

50,000 3D

Gallons

45,000 4D

2014 2016 2018 2020 2022 2024 2026 2028 2030 2032

5D

Year

Strategy D-Maximize Energy Efficiency

70,000

Annual Water

65,000 1E

60,000

Consump0on in Million

55,000 2E

50,000 3E

Gallons

45,000

4E

14 16 18 20 22 24 26 28 30 32 20 20 20 20 20 20 20 20 20 20 5E

Year

Strategy E-Maximize Renewables Water Impact Metric Observations:

• Average water consumption declines over time in all strategies.

• Variation across scenarios within a particular strategy ranges from 10.5 percent for Strategies A/B to 13.8 percent for Strategy D. This is largely driven by the variation in load growth in the different scenarios.

• Average water consumption across the five strategies ranges from 56,960 for Strategy E to 59,210 for Strategy C or 2,250 million gallons. This represents a variation of about 4 percent.

180

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F Strategy Performance: Waste Impact Metric Scoring metric results:

6000000

Annual Produc4on in Tons

5000000

1A

4000000

2A

3000000

3A

2000000

4A

1000000

5A

0

14

16 18

20

22

24

26

28

30

32 20 20 20 20 20 20 20 20 20 20 Strategy A-The Reference Plan 6000000

Annual Produc4on in Tons

5000000

1B

4000000

2B

3000000

3B

2000000

4B

1000000

5B

0

14

16 18

20

22

24 26 28 30 32

20 20 20 20 20 20 20 20 20 20 Strategy B-Meet an Emission Target 181

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F 6000000

Annual Produc4on in Tons

5000000

1C

4000000

2C

3000000

3C

2000000

4C

1000000

5C

0

14 16 18 20 22 24 26 28 30 32

20 20 20 20 20 20 20 20 20 20 Strategy C-Focus on Long-term, Market Resources 6000000

Annual Produc4on in Tons

5000000

1D

4000000

2D

3000000

3D

2000000

4D

1000000

5D

0

20 20

18 20 20 20 20

28 20 20 14 16 20 20 22 24 26 20 30 32 Strategy D-Maximize Energy Efficiency 182

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix F 6000000

Annual Produc4on in Tons

5000000

1E

4000000

2E

3000000

3E

2000000

4E

1000000

5E

0

14

16

18

20

22

24

26

28 30

32 20 20 20 20 20 20 20 20 20 20 Strategy E-Maximize Renewables Waste Impact Metric Observations:

The trends in the production of high-level waste, which is primarily spent nuclear fuel and other fuel assembly components, parallel those of nuclear fuel requirements and are the same for all alternative strategies and average 149.05 Tons/Year.

183

Appendix G Method for Computing Valley Economic Impacts 185

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Background that dampen it. REMI is a general equilibrium model used by TVA for well over a decade and is currently Since the TVA Act promotes agricultural and industrial in use by over 100 universities, state and local development as a core TVA responsibility, the governments, utilities, and consulting firms throughout economic well-being of Tennessee Valley (hereafter the U.S. and Europe. TVAs model has been tailored Valley) residents has been part of the TVAs mission to the TVA Region by county and optimized to capture since 1933. In keeping with TVAs core mission, the the inter-industry and inter-regional linkages with IRP scoring process incorporates a single economic surrounding counties and the rest of the United States.

impact metric for each strategy of every scenario under As shown in Figure G-1, the direct effects, i.e. changes consideration. Per capita income is calculated in order in TVA expenditures and retail electricity bills, are input to assess the relative impact of each strategy on the into REMI, which capture any multiplier effects and general economic conditions of in the TVA Region.

interactions within the regional economy.

This metric is used as one input into the overall IRP Scorecard used to evaluate alternative strategies. As second metric, Valley employment is also included in this appendix but is not part of the scorecard.

Process Overview The U.S. Bureau of Economic Analysis provides a broad measure of per capita income that reflects not just wage income but total compensation, such as employer contributions to health insurance and retirement accounts. Additionally, it includes other income sources, such as dividends and transfer payments. Thus, per capita income provides a single metric that broadly reflects the general economic well-being of Valley residents and is readily understandable and relatable.

It is also one that will reflect the net effect of each strategys change in expenditures and electricity bills.

Increases in TVA expenditures on labor, equipment and construction materials stimulate the economy. At the same time, increases in consumers electricity bills required to fund those operations and construction activities, reduce consumers disposable income. Lower disposable income limits consumer purchases on goods and services in the TVA Region. Since strategies that involve increasing in-Valley expenditures tend to require higher electricity bills, their impacts tend to be Figure G1: Input and Output Impacts offsetting.

The PI+ Model by Regional Economic Models, Inc., Strategy A of each scenario serves as the Reference hereafter referred to as REMI, is used to model the Plan, so each strategy within each scenario is multiplier effects of each strategys expenditures that compared to Strategy A. Thus, increases in stimulate the regional economy and its electrical bills expenditures are only entered into REMI to the extent 186

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G that they exceed Strategy As expenses. In this way While there are ongoing national codes and standards REMIs outputs are the impact on per capita income that increase energy efficiency, TVA implements relative to the Reference Plan of each scenario. programs that expedite the adoption of energy efficient appliances and insulation that are over and above the minimum required. The economic impact Methodology of TVA investments in energy efficiency programs are Each strategy has a different annual revenue modeled as eight new jobs in the TVA Region for each requirement needed to fund its construction, $1 million spent. Of the jobs created 20 percent fall in generation, and energy efficiency programs. The the utility industry, 20 percent in construction industry, difference between the Reference Plan and the other and 60 percent in professional/scientific employment strategies revenue requirements are modeled as categories. All differences from the Reference Plan changes in the electricity bill for residential, commercial, are annual values, so changes in per capita income and industrial customers. Ultimately, rate payers must are generated by year. The per capita income output fund any increase in TVA expenditures. models the trajectory of economic impacts over time.

In order to rank and compare alternative strategies, While increases in a strategys revenue requirements the present value of the changes in per capita income tend to reduce consumers ability to purchase goods is evaluated with a 2 percent discount rate from 2014 and services, an increase in TVA expenditures to 2033. A low 2 percent discount rate is employed, stimulates economic activity, at least to the extent because the changes in per capita income were that they are purchased within the TVA Region. previously adjusted for inflation. Selecting a rate as high Expenditures that are almost exclusively sourced as 8 percent does not, however, materially impact the outside the TVA Region, such as fuel or purchased wind strategy rankings. The results are presented below for power from the Midwest, are excluded from TVA Region non-farm employment as well.

expenditures.

Since not all types of expenses have identical economic Overall Findings impacts, REMI was used to separately model the Figure G-2 provides changes in the TVA Regions per impact of renewable construction, non-renewable capita income caused by each strategy. The difference construction, non-fuel operation and maintenance in all scenarios for all strategies is quite small. From (O&M), and energy efficiency expenses. In this way 2014 to 2033 the average percentage change in per REMI identifies the ability of the TVA Regions economy capita income ranged from -0.01% to 0.03%. The to supply the necessary inputs and to what extent they results are expected to be small for several reasons.

must be sourced outside the region. Since most new First, TVAs revenue is a small percentage of the total construction expenses are likely to be natural gas-fired TVA Region economy. In 2015, TVAs revenues are power plants, REMIs custom construction industry expected to approach $11 billion, but the entire TVA for natural gas-fired power plants was incorporated Region economy is almost $430 billion. Second, all the into the analysis. Similarly, since most new renewable proposed strategies are similar approaches to supplying construction in the TVA Region will be solar installations, the regions power needs. Changing from one approach REMIs custom industry for solar plant construction was to another should not result in significant impacts on the used. This delineation between types of construction economy as a whole.

expenditures enhanced the accuracy of the results and followed directly from stakeholder feedback after completion of the 2011 Integrated Resource Plan.

187

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Per Capita Income*

2014-2033 Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Avg. of Annual % Changes Current Stagnant Growth De-Carbonized Distributed from Reference Plan Outlook Economy Economy Future Marketplace B - Meet Emission Target 0.00% 0.01% -0.01% 0.00% 0.00%

C - Focus on LT Market* 0.00% 0.01% 0.03% 0.01% 0.00%

D - Maximize EE 0.02% 0.02% 0.02% 0.02% 0.02%

E - Maximize Renewables -0.01% 0.00% 0.00% 0.00% -0.01%

Present Value of Per Capita Income (2013$)

A- Reference Plan $38,074 $36,206 $39,590 $37,502 $38,074 B - Meet Emission Target $38,074 $36,208 $39,588 $37,501 $38,074 C - Focus on LT Market* * $38,073 $36,209 $39,602 $37,505 $38,073 D - Maximize EE $38,080 $36,213 $39,597 $37,510 $38,081 E - Maximize Renewables $38,069 $36,204 $39,588 $37,502 $38,069

• U.S. Bureau of Economic Analysis definition reflects total compensation that includes wages and benefits and transfer payments, such as Medicare and Medicaid.

• • Full Name: Focus on Long-Term, Market-Supplied Resources Figure G2: Results Across the five scenarios, there are meaningfully of the changes in per capita income is $116 over and different assumptions about economic conditions above what would have been available in the Reference nationwide that impact the TVA Regions standard of Plan. To get a sense of what is driving the results, living. Per capita incomes are not, however, comparable the changes in in-Valley expenditures are graphed across scenarios because the varying scenario alongside the changes in revenue requirements.

assumptions generally overwhelm strategy-driven Increasing in-Valley expenditures provide an economic impacts. stimulus, while increases in the revenue requirements dampen economic growth.

Detailed Results - Current Outlook An important interpretation caveat is that REMI models Scenario different types of expenditures differently. Dollars Scenario 1 -Current Outlook reflects the expected spent on solar construction have a different impact case assumptions about the general state of the from dollars spent on gas plant construction or energy economy and power markets. In this scenario we find efficiency. Nonetheless, comparing aggregate changes that Strategy D, Maximize Energy Efficiency (EE), is the in in-Valley expenditures and revenue requirements can most beneficial. From 2014 to 2033 the present value provide insights into the models result.

188

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Figure G3: Current Outlook 189

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Strategy 1B - In terms of either in-Valley expenditures expensive and some revenues are spent on out-of-or revenue requirements, Strategy1B is little changed Valley wind generation. Compared to the Reference from the Reference Plan. Plan, revenue requirements increase more than in-Valley expenditures.

Strategy 1C - Strategy 1C experiences slightly higher revenue requirements than expenditures, which depresses per capita income. Detailed Results - Stagnant Economy Scenario Strategy 1D - Strategy 1D involves greater in-Valley expenditures, especially after 2024. Although relatively The Stagnant Economy scenario models a world in expensive, the in-Valley EE expenditures do lower which economic growth fails to materialize as expected.

revenue requirements relative to expenditures. Most strategies have a marginally positive impact, but the benefit of in-Valley EE expenditures gives Strategy D Strategy 1E - Renewable generation is relatively the largest gain.

190

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Figure G4: Stagnant Economy 191

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Strategy 2B - In-Valley expenditures increase, but Detailed Results - Growth Economy revenue requirements are only marginally higher.

Scenario Strategy 2C - Increases in in-Valley expenditures The Growth Economy Scenario models an environment are consistently greater than the increases in revenue of higher than expected economic growth and growing requirements. demand for power. This is the one scenario in which Strategy C, Focus on Long-Term, Market-Supplied Strategy 2D - Significant EE investments, modeled as Resources, provides the greatest benefit. This is all in-Valley jobs, result in higher in-Valley expenditures.

largely driven by TVAs presumed ability to secure Even though revenue requirements increase, in-Valley 10-year PPAs prior to the profitability of in-Valley expenditures more quickly increase.

solar generation. Strategy D that emphasizes energy Strategy 2E - Higher cost renewables are less cost efficiency programs is, however, the still second most effective with limited sales growth and lighter carbon beneficial.

regulation.

192

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Figure G5: Growth Economy 193

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Strategy 3B - In-Valley expenditures are generally lower residents paying for out-of-Valley generation. Through and revenue requirements are flat. 2024 revenue requirements are generally greater than in-Valley expenditures.

Strategy 3C - Revenue requirements are driven down because TVA is able to sign 10-year Purchased Power Agreements with gas-fired power plant operators but Detailed Results - De-Carbonized afterwards builds solar generation as technological Future Scenario improvements make solar power more efficient.

The De-Carbonized Future Scenario models a regulatory Strategy 3D - EE expenditures, modeled as in-Valley environment in which there are significant carbon jobs, ramp up dramatically beginning in 2023 and create taxes that impact the relative efficiency of alternative a significant stimulus effect that more than offsets the strategies. As in all but one scenario, the stimulus impact increased cost. of Strategy Ds in-Valley EE investments generates a marginally more positive economic impact than the other Strategy 3E - Since wind is primarily located outside strategies.

the Valley, focusing on renewables involves Valley 194

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Figure G6: De-Carbonized Future 195

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Strategy 4B - Strategy Bs revenue requirements and wind expenditures peak in 2019. After 2020 increases expenditures are little changed. in Non-Fuel O&M and in-Valley Solar Construction expenses provide a stimulus.

Strategy 4C - Revenue requirement growth is moderated by limited growth in the cost of purchased power. Expenses increase with in-Valley solar Detailed Results - Distributed construction after 2019 and significant EE in 2024 & Marketplace Scenario 2025.

The Distributed Marketplace Scenario models a world Strategy 4D - Significant in-Valley EE expenditures in which the economic and technological changes begin in 2024. facilitate a shift toward distributed power generation. In this scenario, Strategy D offers the only approach that Strategy 4E - Revenue requirements in 2016 through improves upon the Reference Plan.

2022 exceed in-Valley expenditures as out-of-Valley 196

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Figure G7: Distributed Marketplace 197

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G Strategy 5B - In-Valley expenditures and revenue Other Reportable Metric requirements are little changed.

Although not used in the analysis directly, percentage Strategy 5C - Changes in revenue requirements changes in Nonfarm employment from the Reference generally exceed in-Valley expenditures. Plan are presented in this Figure G-8. Like changes in per capita income, changes in nonfarm employment are Strategy 5D - EE investments lift in-Valley expenditures very small.

above increases in revenue requirements.

Strategy 5E - Focusing on expensive --and often out-of-the-Valley wind-- renewables increases revenue requirements faster than in-Valley expenditures.

Conclusion There are multiple approaches to meeting the TVA regions power needs. This analysis compared the economic impact of alternative strategies to that of the Reference Plan for every scenario. Each strategy involved changing the level of in-Valley expenditures and the magnitude of electricity bills required to satisfy each strategys funding needs. Using REMIs PI+ general equilibrium model tailored to the TVA service territory, the impact on per capita income of alternative strategies for meeting power demand was evaluated. By using custom industry models and base REMI capabilities, the impacts of different types of expenditures (e.g.,

renewable construction, non-renewable construction, non-fuel O&M, etc.) were modeled explicitly.

Under most scenarios Strategy D, Maximize EE, generated the largest gains in per capita income over and above the Reference Plan. EE expenditures disproportionately remain in the Valley and dampen future electricity costs. Both factors tend to improve the relative performance of Strategy D. That being said, the impact of all alternative strategies on per capita income was exceptionally small. Across all scenarios and strategies the average percentage change in per capita income from 2014 through 2033 ranged from -0.01%

to 0.03%. The present value of the stream of annual differences is small as well. Over a 20-year period, the Maximize EE strategy provides an additional benefit whose present value ranges from $116 to $158.

198

I N T E G R AT E D R E S O U R C E P L A N - 2 0 1 5 F I N A L R E P O R T Appendix G NonFarm Employment 2014-2033 Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Avg. of Annual % Changes Current Stagnant Growth De-Carbonized Distributed from Reference Plan Outlook Economy Economy Future Marketplace B - Meet Emission Target 0.00% 0.03% -0.01% 0.00% 0.00%

C - Focus on LT Market* 0.00% 0.04% 0.05% 0.02% 0.00%

D - Maximize EE 0.06% 0.11% 0.06% 0.07% 0.08%

E - Maximize Renewables -0.02% 0.02% -0.01% 0.00% -0.02%

Annual Average (Thousands)

A- Reference Plan 4,338 3,837 4,717 4,189 4,338 B - Meet Emission Target 4,338 3,839 4,717 4,188 4,338 C - Focus on LT Market* 4,338 3,839 4,720 4,189 4,338 D - Maximize EE 4,341 3,841 4,720 4,192 4,342 E - Maximize Renewables 4,337 3,838 4,717 4,188 4,337

• Full Name: Focus on Long-Term, Market-Supplied Resources Figure G8: Nonfarm Employment 199

tva.com/irp