ML18219A546
| ML18219A546 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 02/28/1976 |
| From: | Anderson K National Economic Research Associates |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| Download: ML18219A546 (279) | |
Text
{{#Wiki_filter:~ Il/8/I"/R NATIONALECONOMIC RESEARCII ASSOCIATES, INC. NEW YORK / WASIIINGTON/ PI IILADELPHIA/ LOS ANGELES
,~
SALES AND LOAD GROWTH OF PARTICIPATING COMPANIES ARIZONA .NUCLEAR POWER PROJECT 1974-1988 by
~ Kent P. Anderson
[
~ I February 1976
\
QS 1 I I O.
TABLE OF CONTENTS P~ae I. INTRODUCTION AND
SUMMARY
XX. RESIDENTIAL SALES XIX. COMMERCXAL SALES IV. MANUFACTURING AND MINING SALES 13 V INPUT ASSUMPTIONS I ADJUSTMENTS AND PEAK-LOAD CALCULATIONS 18 VI. FURTHER CONSIDERATIONS 26 A. Rate Structure 26 B. 'oluntary Conservation Measures 27 C. Mandatory Conservation Measures 30 D. Positive Incentives for Conservation TABLES 34 APPENDIX A APPENDIX B
0, Qo
SALES AND LOAD GROWTH OF PARTICIPATING COMPANIES ARIZONA NUCLEAR POWER PROJECT 1974-1988 I. INTRODUCTION AND
SUMMARY
This report presents projections of future sales and load growth for the five major participants in the Arizona Nuclear Power Project: Arizona Public Service Company (APS), the Salt River Project (SRP), Public Service Company of New Mexico (PNM), El Paso Electric Company (EPE) and Southern California Edison Company (SCE).'he projections are based upon econometric forecasting models developed at NERA espe-cially for this report. The report describes the important structural features and quantitative assumptions of the fore-casting models, and it compares the independent forecasts with those prepared by the companies.~ Finally, the report discusses the possible effects of phenomena. that may be im-portant, though difficult to model given the state of. the art. They include conservation, rate structure reform and load management. The Arizona Electric Power Cooperative is also a parti-cipant but accounts for only 3 percent of the sales. participants'ombined See, in particular, those reported in Supplement 4, Vol. VI of the Environmental Re ort, Construction Permit Sta e,
. Palo Verde Nuclear Generating Stations, Units 1, 2 and 3, submitted to the Nuclear Regulatory Commission on May 19, 1975; and that reported by Southern California Edison in S stem Forecasts 1975-1994, March 1975.
0, Qo e 0 '/
Table 1 summarizes the load projection results.~ Column (1) shows actual 1974 peak load; column'2) the com-pany forecasts for 1988; and column (3) the annual compound rates of growth implied by the company forecasts. Columns (4) and (5) give the growth rates for the low and high econo-metric projections obtained for each company on the assump-tion of no change in prices (after removing the effects of general inflation). This range also corresponds to the one that would be obtained if peak load growth were totally in-, sensitive to price variations. Columns (6) and (7) contain N the growth rates for the low and high 'econometric, projections obtained for cases where the real (i.e., deflated) prices of electricity and fuels are assumed to rise. These projections assume that, peak load growth is percentage-wise as sensitive to prices as total sales. It is likely that peak load is less sensitive to price than annual sales but not totally insensitive to price. If so, then the true range of outcomes for peak load ought to be somewhere between the ranges re-ported. Note that the effect of price variations is to widen the possible range of outcomes and with one exception--at both ends of the spectrum." The corresponding kilowatt-hour sales projections imply very similar growth rates and are not reported in the table. See Appendix A for projections of kilowatt-hour sales by user class, peak loads and corresponding growth rates. Results for other cases are also reported in this Appendix. 4 The SCE low-end projection remains almost unchanged.
0 Table 1 shows that the range of projections derived from the econometric models for the five-company group brackets the companies'roup forecast but is centered'bove it. This outcome strongly suggests (a) that there is no inconsistency between the company forecasts and projections obtainable from econometric models utilizing plausible assumptions about economic and demographic trends and about future gas and elec-tricity prices and (b) that the company forecasts are, if any-thing, too low on the whole. The range of variation in the econometric projec-tions arises from the application of different assumptions about several of the most important determinants of electri-city demand growth. These determinants, and the values assumed for them, are detailed in Tables 2 and 3. The figures for low growth and for high electric and low gas prices under-lie the "minimum" projections of Table 1. The figures for high growth and for low electric and high gas prices form the basis for the "maximum" projections. The growth assumptions are based upon regiona1 economic and demographic projections prepared by the Bureau of the Census, the Bureau of Economic Analysis and state agencies. The price assumptions are based upon historical company data, projected electricity generation costs and the prospective cost of natural or synthetic gas to distributing companies. Section V provides details on these and other model input assumptions. Sections II-IV des-cribe the features of the econometric models for the residential,
I o
commercial and industrial classes, respectively, for the t five companies. These models project kilowatt-hour sales. Section V describes how these are converted to peak-load pro-jections. Section VI focuses upon important, issues lying outside the, scope o'f the forecasting models. IX. RESIDENTIAL SALES To develop forecasting relationships for, residential sales, I relied upon statistical studies of variations in energy usage patterns across states. The forecasting rela-tionships thus obtained are of two different types: The first includes equations for estimating the saturations~ of electric ranges, electric clothes dryers, electric water heaters, built-in electric space heaters, a single room air conditioner, more than one room air conditioner, central air conditioning and evaporative coolers. The second is a single equation which estimates all other electricity usage per residential customer. The appliances noted above form what X call the major-usage category. The first four appliances in the group are particularly important where price effects h are concerned since they reflect electricity usage for which alternative energy sources are usually available.
'E For cooking P
and clothes drying, both natural gas and bottled gas represent Fraction of households owning.
)V dSyQgg P . I I G t ' t 1t ~ 0 ~ lil~4&A(tel+
- JV~J %Of ' t 0
Qi 0
competing fuels. For space and water heating, oil is poten-tially a competing fuel in addition to natural and bottled gas. Xn these categories, both the price of electricity and the prices and availability of competing fuels are likely to affect the level of saturation. For other home usage of electricity, such as air conditioning, lighting, food freezing and refrigeration, radio and television, electricity is for all practical purposes the only source of energy. Consequently, while electricity price may influence use, the prices of other fuels have almost no relevance. All such usage with the exception of air condi-tioning and evaporative coolers forms what one might. call the net-usage category. The econometric models project net usage per customer as a lump sum, which includes not, only the elec-tricity consumption of appliances in the net-usage category but also. increases or decreases in the average level of con-sumption per appliance in the major-usage category. This is R because net usage is derived by multiplying the saturations of the major appliances by estimated ave~acVe usage per appliance in the major-. usage category. Xf consumers in a particular area have larger units or use their appliances more inten-sively than the assumed usage rates indicate, this excess will be reflected in net usage. Estimates of average appliance usage and base-year (1974) saturation are shown in Appendix B. The base-year value for net usage is obtained by subtracting major usage from total usage per customer.
0, l I V f 0 0 0
The coefficients of the residential demand equations were estimated from data for the 48 contiguous states in 1970. (All 50 states were used for the air conditioning equations.) The statistical procedure known as "multiple re-gression analysis" was used to relate the saturation of each J of the major appliances and n'et electricity usage per custo-mer to a number of economic, demographic and climate-related variables, including income per household, percent'of house-holds with cash incomes of less than $ 3,000, the price of electricity, the price of natural gas, the price of fuel oil, the price of liquefied petroleum gas, average household size, percent of households in rural areas, percent, of households in central cities, percent of households built in the last 10 years, percent of households in buildings with five or more units and the average number of cooling or heating degree days. In developing the saturation equations other than those for ~ air conditioning, price data were calculated as averages for the period 1964-1970. In developing the air conditioning R saturation and net usage equations, 1970 prices were used. The variables and their estimated coefficients are reported in Appendix B. The only saturation relationship not directly esti-mated by multiple regression analysis was that for evaporative coolers. Lack of adequate data prevented its application in this case. I found, however, that the historical movement in the saturation of evaporative coolers in both APS's and SRP's
Qi Qi
service territories has been such as to maintain a fairly con-stant proportion between the fraction of customers not having room or central air conditioners and those having evaporative coolers. Table B-3 of Appendix B details this fact. The projections for evaporative coolers for these two companies assume that this historical relationship will continue to hold. histories for the other companies were not, available 'imilar although base-year figures were; the ratios indicated by the l974 figures are assumed to hold for the future. As indicated above, the relationships that. I used for the residential sector are based upon state data for.l970. This approach is known as a cross-sectional approach, because it uses data for various areas observed at a single point in time. Xn the most common alternative approach--time series analysis--the data are analyzed for a single area in successive time intervals. An important advantage in using cross-sectional data rather than time series data to estimate demand relationships arises from the fact that the adjustment of electricity usage to changes in prices and income may take a considerable period of time. The cross-sectional approach, inasmuch as it meas-ures the relation between electricity consumption and regional price differences that have prevailed over a long period, re-flects this long-run adjustment process. Moreover, regional
0 conditions vary considerably and thus provide a rich body of data with which to work. The time series approach, by con-trast, is restricted to year-to-year variations in conditions. These, until very recently, have tended to change smoothly over time, and it has been difficult if not impossible to obtain convincing evidence about long-term. or even short-term responses by time series analysis. h The relationships used to forecast the residential demand for electricity, and particularly the price respon-siveness of this demand, are quite similar to those reported in a study done in 1973 at the Rand Corporation'nd in a 1974 paper by Baughman and Joskow. In fact, equations ob-tained from the 1973 Rand study form the air conditioning part of the residential model. Several other studies have been published those by Fisher and Kaysen, Halvorsen, Kent P. Anderson, Residential Ener Use: An Econometric A~nal eia, prepared for the National science poundatxon (Santa Monica, California: The Rand Corporation, October 1973) . Martin L. Baughman and Paul L. Joskow, Interfuel Substi-tution in the Consum tion of Ener in the United States, Draft, Massachusetts Instr.tute of Technology, May 10, 1974. F.M. Fisher and C. Kaysen, The Demand for Electricit xn the United States (North Holland, 1962 Robert Halvorsen, The Demand for Electric Power in the United States, presented at The Wanter Meetl.ngs of the Econometrzc Society, New York, December 1973.
0 0
Mount, Chapman and Tyrrell,'nd Wilson,'or example. In some cases the results reported by those authors record dif-ferent elasticities than those reported here. In my view, these differences reflect errors in approach or technique. The single most important of these errors involves the failure to separ-ate electricity consumption into saturation-related and net-usage components. A second and related error is the failure in some cases to control'dequately for the influence of the prices of all important competing fuels. Finally, one or two of these studies include as part of the price effect, the effect of movement along declining block rate schedules as consumption has grown. This has occurred simultaneously with falling real electricity prices and raises the risk of error in estimation if not properly accounted for. III. COMMERCIAL SALES Few studies done to date have dealt adequately with the price elasticity of demand in the commercial sector. Demand in the United States: An Econometric A~nal sis Oak Ridge, Tennessee: Oak Ridge National Laboratory, June 1973); "Electricity Demand Growth and the Energy Crisis," Science 178, 1972, pp. 703-8; "More Resistance to Electricity," Environment, Vol. 18, Mo. 8, October 1973, pp. 18-.35; L.D. Chapman, et al., "Power Generation: Conservation, Health, and Fuel Supply," Draft, Task Force on Conservation and Fuel Supply, Technical Advisory Com-mittee on Conservation and Energy, 1973, National Power Survey, U.S. Federal Power Commission; and T.D. Mount, "Forecasting Energy Consumption: An Appraisal of Three Alternative Procedures," Draft, November 21, 1974. John Wilson, "Residential and Industrial Demand for Elec-tricity: An Empirical Analysis'-'Ph.D. diss., Cornell University, June 1969).
I 0
Mount, Chapman and Tyrrell obtain long-run electricity price elasticities of -1.36 to -1.60 and competing-fuel price elas-ticities of only 0.04 to 0.06.'he latter elasticities cannot be trusted, since analysis of the residential sector in which energy usage is similar (space conditioning, light-ing, cooking, etc.) shows that the formulation used by Mount et al. can result in too-low and inaccurate competing-fuel price elasticities. Even the own-price elasticities may be inaccurate, since Mount et al. make no effort to- control for differences in commercial sector composition from state to state or year to year or to account for the inverse re-lationship between average consumption and average electri-city price resulting from declining block rate schedules. Halvorsen corrects for the latter problem and obtains elec-tricity price elasticities of -0.56 to -1.21, but he estimates incorrectly-signed competing-fuel price elasticities.'4 Bonneville Power Administration has also attempted to estimate commercial-sector price elasticities, but the study, is plagued by lack of data to control for various factors other than the price of electricity. Mount, et al., ~o . cit. See Anderson, Residential, and Baughman and Joskow,
~os c it ~
Halvorsen, ~o . cit.
In recent work at NERA I have attempted to remedy the deficiencies encountered in earlier studies, by (a) using data for which the commercial sector is ~consistently defined from one state or utility district to another, (b) accounting for differences in commercial sector composition from one location to another, (c) employing electricity prices that control for the declining-block-rate-schedule problem mentioned above in Section II and (d) by developing a better measure of competing-fuel prices. The results of a NERA study incorporating the points mentioned above are shown in Table B-4 of Appendix B. The price and income coefficients are the elasticities of commer-cial demand with respect, to the corresponding variables. The uses 1971 data for the 48 contiguous states. Commercial 'tudy sales are defined as total commercial and industrial sales as reported by Edison Electric Xnstitute minus 1971 electri-city purchases by manufacturing firms as reported in the Census of Manufactures for'972. This helps to overcome the prob-lem of inconsistent definitions across different areas. The heterogeneity (or mixed-bag) problem is handled by including activity-level indices for the four non-commercial sectors included in the residual sales total: mining, transportation/ communications/utilities, contract construction and agriculture. The Census figure for Washington is adjusted to elimin-ate a sizeable error in the total for that state. Data for individual company sales provided the means for accomplishing this adjustment,
Use of a typical bill for a fixed level of consumption reduces the significance of the rate schedule problem, and construc-tion of a gas-oil fuel-price index to use in place of gas prices alone provides a superior measure of competing-fuel prices. Owing to the fact that the SRP and APS models are adaptations of models developed recently for these two com-panies for other purposes and prior to the models constructed de novo for the other three companies, there are some minor differences among the commercial sector models actually used in each case. Those for the latter use the full equation as reported in Appendix B.'hat for APS uses the price and income components of the equation, but. assumes demographic. and compositional effects to be negligible. (The latter are clearly far less important determinants; and, in any event, their omission is somewhat ameliorated by the separate treat-ment of mining as part of the industrial sector.) The model for SRP is based upon explicit activity-level projections for seven separate commercial subsectors. These projections obviate the need for the income and compositional indices. All other components of the equation are retained and are applied to a "base" projection representing the sum of the individual subsector projections prior to the computation of price and demographic effects. Since there is a separate mining sector in the PNM model, the mining index is not used.
IV. MANUFACTURING AND MINING SALES To project manufacturing and mining sales the models use a series of equations developed for individual SIC 2-digit manufacturing industry groups and one for mining activities. These equations project sales to manufacturing and mining customers as determined by industry output, the price of electricity and the price of competing fuels. The manufacturi'ng sector equations derive in part from 1971 energy price and purchase data reported in the 1972 Census of Manufactures by state for 2-digit industry groups. Twelve industries have been examined. They are identified in Table B-5 of Appendix B, which also, lists the number of states for which data were available and the re-gression coefficients obtained for the explanatory variables noted above. The coefficients for electricity and fuel prices measure the elasticity of demand with respect to these prices. In all cases the estimated equations require that sales be proportional to industry output, if prices and other factors do'not change. The procedure for estimating the manufacturing equa-tions was designed in such a way as to control for variations among states in the level and composition of output within each 2-digit industry group. Controlling for industrial com-position is of great importance when using regional, data to estimate the price responsiveness of industrial demand for electricity. Comparison of electricity usage per dollar of
output by industry groups across areas in the United States re-veals quite wide variations. The highest electricity consump-tion per dollar of output occurs in those areas where electri-city price is low, and the lowest consumption occurs in those areas where price is high. Several studies'" of the indus-trial demand for electricity have included in their measure of price elasticity a'omponent reflecting the fact, that in-dustries that are large electricity users tend to locate in areas where the price of electricity is low. Since a nationwide trend in the price of electricity is not. likely to influence the location of electricity-intensive industries, it is de-sirable that the equations used to predict the effects of price rises exclude the locational effect. If cross-sectional data are used to estimate these equations, variations in electricity consumption that reflect differences in industrial composition must be separated from those that, reflect variations in inten-sity of use by individual firms and industries. In an exploratory study of the primary metals indus-try'nd, later, at NERA a substantial effort has been made to examine the effect of price on electricity consumption in such a way as to minimize the influence of regional variations I in composition upon the estimated effect of prices. This 'has See, for example, Mount, et al., o~ cit. or Halvorsen,
~os cit ~
Kent P. Anderson, Toward Econometric Estimation of Indus-trial Ener Demand: An Ex erimental Athe lication to the Primar Metals Industr , prepared for National Science Foundation (Santa Monica, California: The Rand Corporation, December 1971). See also Wilson, o~ cit.
Qe been done in two ways. First, as already noted, demand elasticities are estimated separately for 2-digit manufac-turing industry groups. Second, the analysis of each industry group controls for variations in composition within the group. This is done by dividing the industry aggregates into sub-industries,. each of which may have a different pattern of electricity usage. For each state and industry group one can derive estimates of the level of electricity consumption that would have prevailed in the group if electricity consumption per dollar of output in each of the subindustries had been at the national average for that. subindustry. Xndustry consump-tion, so estimated, is then used in place of industry output as an explanatory factor in the industry sales equation to be estimated. This helps to ensure that the resulting price elasticities come closer to measuring the extent to which elec-tricity consumption would be affected by prices, holding in-dustry composition constant. The procedure described above is an attempt to con-trol for variations in industrial'omposition. But it is not completely successful. Census data on industrial composition are not available below the four-digit SIC code level and, in many cases, are not available at or below the three-digit SXC code level. Yet, even within three- and four-digit sub indus-tries, there is some regional variation in the mix of products and in the electricity required to produce these products.
0
-16 All one can say with any certainty is that controlling for industrial composition somewhat reduces the apparent elasti-city of demand for electricity with respect to its own price.
If it were possible to control for industrial composition perfectly, the estimated electricity price elasticities would probably be smaller ('.in absolute value) than those estimated to date. The low, or eve'n apparently, zero, elasticities of demand with respect to comp'eting fuels prices in several indus- . tries may be due to lack of a sufficiently refined model or of an adequately powerful procedure for statistical analysis. It is also possible that, inadequate accounting for variations in industry composition may work in the opposite direction when estimating competing-fuel price effects--that is, biasing them toward zero rather than away from zero. Interviews con-ducted recently by NERA at a number of manufacturing 'estab-lishments throughout the country indicate a significant poten-tial for substituting electricity for fuels in various manu-facturing operations, should there be a sufficient incentive to do so. No satis'factory study of the long-run price and output elasticities of the mining sector has yet come to light. The coefficients used in the model for this sector therefore have a larger "guess" component to them than those in other sectors. The output coefficient is assumed to be unity, indicating as for, manufacturing, a proportionality
i
between industry output and electricity sales prices and other factors (such as plant scale and industry composition) held constant. The estimate for the electricity price coeffi-cient, is based upon several indirect sources of evidence:
- 1. The short-run own-price elasticity of electri-city demand for copper mining in Arizona appears to be about.
-0.2, based upon a previous NERA study.
- 2. For industrial demand, the long-run own-price elasticity is probably at least six times the short-run elas-ticity.'~
- 3. For commercial demand (including mining) a long-run own-price elasticity of about -0.8 is obtained.
Based upon the relationship .between long- and short-run indus-trial own-price elasticities, a figure of perhaps -1.2 is plausible for mining. But"in view of the lower figures ob-tained for commercial and industrial uses (on the average), a value of -1.0 would seem.to be more reasonable, and this is the figure used in the model. For the competing-fuel price elasticity, even less information is available. The commercial sector results, which include mining, indicate a ratio of competing-fuel-price to own-price elasticity of nearly unity. The industrial sector results yield an average ratio of 0.24. Table B-5 indicates a mean electricity price elasticity of
-1.07 for the twelve manufacturing industries-studied. A recent NERA study of short-run demand obtained a short-run figure of -0.17 for the industrial class as a whole. These two estimates indicate a ratio of more than six to one.
Mount:, et al , o~ cit., obtain a ratio of more than seven to one using a dynamo.c lagged-adjustment model for the in-
~
dustrial sector.
0'
~
8
For mining I chose a value of 0.3, assuming that the ratio for mining alone would be much closer to the average for manufac-turing than that for commercial users. V. INPUT ASSUMPTIONS, ADJUSTMENTS'ND PEAK-LOAD CALCULATIONS This section outlines how the equations estimated for the three major user classes were formed into a forecast-ing model for each company. The first step was to obtain or calculate future values for the explanatory variables appear-ing in the sales equations. The OBERS projections were a principal source. Based upon the Census Series E population projection for the nation, they were developed jointly by the Departments of Agriculture and Commerce, and they include projections to 1990, and beyond, of population and personal income and earnings by sector of origin. The projections cover states, Business Economic Areas (BEA's) and Standard Metro-politan Statistical Areas (SMSA's). ,Those for the Phoenix SMSA and for the Los Angeles, Phoenix, Albuquerque and El Paso BEA's all provided important information for the models. The OBERS population forecasts for the above areas tend to be on the low side and are generally used to derive residential customer-growth projections for the low-growth cases. Population projections for the high-growth cases are based on various sources. (See Table 4.) The figures re-ported in the table are converted to the customer-growth projections shown in Table 2 on the basis of (a) the historical (1962-71) relationship between service territory growth and
0 0 I' 8
-19 that for the area to which the source projection pertains and'b) projected household size. The latter is, in turn, based upon the average of Census Series 1 and 2 projections for the nation and on the historic relationship between the national trend and that for the region (SMSA, BEA or county group) .in which the company is located.
Real income per household is calculated from the OBERS projections for real income per capita and the derived projections for household size. Projected growth in manufac-turing and mining is derived from the OBERS earnings projec-tions, on the assumption that real output growth will closely parallel real earnings growth. The OBERS projections are adjusted to correspond to different underlying population forecasts according to the following formula: adjusted output growth rate = (1 + a) . (1 + b) 1+ c) where: a = OBERS output growth rate b = different. population growth rate
\
c = OBERS-based population growth rate. Projections of household characteristics other than average household size (percent of families with incomes less P than $ 3,000 in 1969 dollars, percent of households in struc-tures of more than five units, percent of households located in rural areas, percent of households located in central cities, percent of households in structures built since 1960
I t 0 8
and fraction of households in single detached units) are based upon 1960 and 1970 Census data for counties. In most cases, the projections are either linear or exponential extrap-olations derived directly from changes observed between the two census years. In a few cases the projections are derived indirectly from extrapolations of underlying variables. For exampl'e, the projection of percent rural for SRP is based upon (a) an extrapolation of total rural households and (b) pro-jected household growth in the service territory; and for example, the projections of percent of households living in units built after 1960 are derived from the projected forma-tion of new households and demolition rates on old housing units estimated from Census data. Projections of two other demographic variables, frac-tion of population located in urban areas and population den-sity, are derived from projections for related variables: percent rural and total service-territory population, respec-tively. The price projections in 1974 dollars are obtained by the following procedure:
- 1. Calculate the average total cost of generation plus purchased power (for electric rates) or of purchased and manufactured gas (for gas rates) in the base year, 1974.
- 2. Estimate transmission and distribution loss-factors for each customer class and adjust average generation 1
or gas costs for losses.
0
- 3. Deduct the adjusted average cost figures from average revenue for each customer class in the base year to obtain an estimate of the net effect of (a) transmission, dis-tribution and other costs and (b) differences in rates due to differences in class load patterns.
- 4. Assume the above non-generation component of future rates to be constant after removing the effects of inflation.
- 5. Estimate the costs of future generation capacity or gas supplies in terms of base-year dollars and adjust. by customer class for losses.
- 6. Divide each cost, item into monetarily-fixed (non-inflatable) and inflatable categories. Adjust the real cost estimates of the non-inflatable components downward according to the projected rate of inflation.
- 7. Add up all cost items.
Table 5 s'ummarizes the cost calculations for electri-city and gas.. The low- and high-cost cases are distinguished from one another in the following ways: For electricity in the low case, future costs calculated as outlined above are averaged with base-year costs in accordance with projected, system growth to yield 1988 estimates corresponding roughly The costs of new generation capacity are inflated up to the, time of installation and then are treated as mone-tarily fixed. The rate of inflation is assumed to be six percent per year for the entire interval 1974-1988.
to average historical cost methods for setting rates. The high figure for electricity in 1988 is simply the estimate for future cost alone. For gas in the low case, future costs are estimated as indicated using $ 1.00/MCF as the cost of gas in 1988. The high estimate for gas assumes $ 2.00/MCF. These figures seem plausible in the light of (a).the likelihood of either the deregulation of natural .gas prices at the wellhead or substantial increases in regulated prices, (b) the already high costs of intra-state gas and imported LNG, (c) the pro-jected costs of synthetic gas and (d) the high cost of oil-the nearest fuel substitute in the event of gas curtailments. Rates for years between 1974 and 1988 are assumed to rise in straight-line fashion from beginning year to end year. The projections of industrial activity by sector for APS and of industrial and commercial activity by sector for SRP are adjusted to reflect the historical- tendency of firms in the Phoenix area to locate in the SRP 'or APS territories. This is done through the use of an elasticity coefficient reflecting the estimated historic relationship between each percentage point growth in activity for the surrounding area (BEA or SMSA) and that for the company. The effect of this procedure is to lower the APS industrial growth projections and raise the SRP commercial and industrial growth projections. A similar adjustment was not made for APS's commercial customers since no sector-by-sector activity-level pro-jections were used in that case.
The factors applied to the growth rate projections are re-ported in Table B-6 of Appendix B. Growth adjustment factors were also calculated for the industrial sectors for EPE, PNM and SCE but, perhaps owing to inadequate data, are suspiciously low.for,the former two companies. Moreover, the total demand projections for all three are not particularly sensitive to variations in these factors. The projections reported for these companies there-fore assume a one-to-one correspondance between industrial growth rates in .the pertinent BEA and those in the company service territories. Use of the growth adjustment factors reduces the low-end growth-rate projections for EPE and PNM to .0648 and .0670 respectively and raises that for SCE to .0984. The differences between these and the reported results are relatively small, because non-industrial elec-tricity demands are responsible for the bulk of sales growth. The effects of energy price changes (and indeed of I many other price changes) take time to achieve their full extent. This is attributable mainly to the fact that energy usage is heavily determined by the character of the existing stock of energy-using equipment and buildings. This stock turns over gradually; and while severe price changes may accelerate the process, considerable time is likely to be involved. This phenomenon is the basis for the distinction The reported rates are .0661, .0686 and .0958 for the three companies.
Qs l l
. between short-run and long-run, price elasticities of demand.
The former describe the immediate response; the latter the ultimate response. The forecasting models have a built-in lagged-adjust. ment mechanism that insures a gradual movement to the long-run equilibrium level. Price effects are expressed as proportional adjustment factors (Z's) and are governed by the following formula: log (Zt) = 0. 85 log (Zt 1) + 0. 15 ~ log (Z") where the subscript t refers to this year, t-1 to last year and Zt to the ultimate price effect indicated by current prices. Th'e weights of 0.85 and 0.15 in the formula imply that 70-80 percent of the ultimate effect of any price change - occurs within ten years and 85-90 percent within 15 years. (The exact percentage adjustment within any time period de-pends upon the size of the price change;) These weights may well over-state the rapidity of adjustment to prices--and hence the importance of price effects during the forecast period since many types of equipment and buildings have economic lives of more than 15 years. Still, they appear to I be consistent with the findings of dynamic lagged-adjustment models of demand. The final step in projecting future loads is to II sum the sales forecasts (where necessary) into class totals See for example, Mount, et al., 'o~ cit.
Qo and to convert these estimated annual sales figures to esti-mated loads at the time of system peak on the basis of his-torical load-conversion factors derived from company figures. Table B-7 of Appendix B reports the underlying data. Calculation of system peak loads by the above method raises a knotty issue, because price changes are likely to affect system peak loads and total sales differently. In particular, it is reasonable to expect that the price changes postulated for this study (which do not incorporate rate structure modification to accomodate peak-load pricing) will affect, peak load less strongly--either upward or downward than total sales. There are several reasons for this. First, a goodly portion of the effect of prices occurs in the form I of substitution of electricity for fuels or vice-versa in the functions of space and water heating, cooking, drying, etc. This substitution will affect sales much more heavily in off-peak periods (i.e.', during the cool season) than at the time of system peak. Second, the effect of price on the saturation of air conditioning, perhaps the principal con-tributor to system peak, is relatively weak, according to the evidence available. Third, price-induced conservation effects are, in the case of air conditioning, likely to occur on off-peak days--not on the hottest and most uncomfortable days when everyone will be inclined to use his cooling system. If this line of reasoning -is correct,. then the range of values obtained for peak-load growth in the no-price-change cases
may better indicate the likely spread of load growth possi-bilities than the range obtained in the price-change cases. (See Table 1.) In either event, the company forecasts are consistent. with the realm of alternative growth rates suggested by the projections of an econometric model. VI. FURTHER CONSIDERATIONS A. Rate,Structure The results obtained to date for price effects in-JI dicate that electricity sales exhibit a significant., degree of responsiveness to the level of electricity rates in the long run, despite some degree of variability from one type of customer to another and from one type of equipment to another. Where competing fuels are significant factors in the market, their price i=oo has been found to play a signifi-cant and measurable 'role in the formation of demand for elec-tricity. Far less has been learned about the responsiveness of electricity sales to changes in the 'structure of rates. Thi's is because there has been little variation in the char-aeter of rate structures over time in the United States or across regions within the United States. Although it cannot be easily established on the basis of historical data, there is every reason to believe that alterations in the rate structure that took the form of raising 'the tail blocks of electricity rate schedules would have price elasticity effects similar to those observed by changing the level of rates generally. It is difficult to predict the effect of the more radical rate
Qo structure revisions, such as adoption of time-of-day pricing or seasonal peak pricing. If the general level of rates were held constant, it is conceivable that energy sales could go up or go down under the imposition of such time-varying rate schemes. But it is not likely that there would be a great effect .upon total sales in either direction. By contrast, peak loads might respond quite. signi-ficantly to adoption of 'time-of-day or time-of-year rates. Unfortunately, at this time very little is known about the degree to which the users might shift loads from peak periods to off-peak periods as a result of the adoption of time vary-ing rates. Under sponsorship of the Federal Energy Adminis-tration a number of peak-load prie'ing experiments are now getting underway across the country. But it is unlikely usable output from these studies will be forthcoming much be-fore 1977. Almost equally little is known about the potential responsiveness of peak load to changes in the level of rates. As noted in the preceding section, there is some reason to believe that the price responsiveness of peak load is less than that for energy sales. B. Voluntar Conservation Measures Voluntary conservation is likely to be a useful supple-ment to broader programs of energy conservation, but it is unlikely that a voluntary approach to conservation would signi-ficantly reduce demand growth for an extended period of time. The stimulus for voluntary conservation must come either from
0 0
exhortation (by which individuals or businesses are encouraged to change their values or goals) or education (by which indi-viduals or businesses are provided the means to define and to achieve their goals more effectively). Exhortation could take the form of occasional jawboning by public officials, or it could take a far more broad-based form such as a political action movement supported by political parties, industry or public interest groups. The educational approach to voluntary conservation would depend heavily upon the cooperation of government and industry. Xt could take the form of advertising, new research and development programs and the provision of technical ser-vices to electricity users. Educational activities could include the dissemination of rules of thumb concerning best buys, installation methods, operating techniques and main-tenance standards. Xt could also include the dissemination of technical and performance data and of cost-benefit,eval-uations for all types of equipment. Experience to date has failed to indicate the extent, to which voluntary conservation on' continuing basis could be sustained by a program of exhortation or education. Past experience during the hydro shortage in the Pacific Northwest, for example, or during the oil embargo in Los Angeles, has shown that electricity users are capab'le of cutting back elec-tricity use considerably in a crisis situation. (The Los Angeles experience can be termed a voluntary one despite the Il/C/1 /cL
0 ~
-29<<
adoption initially of penalty measures, since within a short period of time after enactment, the penalty measures were suspended. Following the suspension, significant electricity conservation continued to take place without the threat of sanctions.). Nevertheless, it is uncertain whether the conser-vation achieved during a temporary crisis period could be sus-tained through a time of extended shortages. Moreover, any full evaluation of the costs and bene-fits of energy conservation programs must. recognize that the additional amount of conserva'tion achieved over time for a given level of outlays on a voluntary program is likely to diminish as users'pportunities for voluntary conservation diminish and as the costs of adopting any further voluntary conservation measures rise correspondingly. Similarly, the tures on education and exhortation are likely to diminish. That is to say, it will become increasingly costly to achieve higher levels of conservation through voluntary programs. At some point the additional expenditures required to expand any voluntary program would not be matched by the energy conser-vation benefits achieved. Just how effective voluntary energy conservation pro-. grams have been in the United States over the last one or two years is somewhat unclear. Statistical analyses of electri-city consumption during the last two years have failed to
0~ 0
provide any conclusive evidence about the role of voluntary conservation. Xt is, in any event, patently clear that the primary source of decline in the growth rate of electricity demand during the years 1974-l975 was the slow-down in econo-mic growth due to the recession. C. Mandatory Conservation Measures Some types of mandatory conservation measures might add to the effects of voluntary conservation or conservation induced by rate changes. These, could include the imposition of appliance efficiency or design standards upon manufacturers of electricity-using equipment and the adoption 'of tighter insulation standards or other building code changes governing materials or energy-using systems. ,Xn many instances, rising electricity rates would encourage these changes. Mandatory standards would slow the growth of loads only to the extent that price-.induced responses were further augmented. Stronger mandatory conservation measures, such as usage quotas, ought to be considered as second-best,alterna-tives to be adopted only in emergencies where rate increases,
'I voluntary conservation measures or weaker mandatory conserva-tion measures have failed to limit load growth to levels consistent with available capacity.
The effectiveness of strong mandatory conservation measures will depend, heavily upon the credibility of enforce-ment and the severif y of the penalties for infractions. Xf
enforcement is weak or the penalties inadequate, extensive violations of mandatory requirements may occur. Moreover, if the usage quotas are particularly severe, it is likely that enforcement cost would be high owing to the large num-ber of violations that, could be expected under severe cut-back conditions. D. Positive Incentives for Conservation Two major classes of positive incentives for conser-vation are noteworthy. The first includes techniques for "load management"; the second, various ways of subsidizing conservation activities. A few utilities are now beginning to experiment with the use of centrally-controlled mechanisms for cutting certain types of loads such as air conditioning or water .heating during. peak periods. Users who agree to in-stall automatic load cut-off devices receive reduced rates in return for allowing their equipment to be shut-off intermit-tently during peak periods. This approach appears to'offer considerable promise for reducing peak electricity loads. However, the limited operating experience with such schemes makes it difficult to estimate what the precise quantitative dimensions of load savings might be. The subsidy approach to conservation could take the form of low-cost financing or tax deductions for customers who make investments in energy-conserving devices, systems or improvements. Some utilities have already begun programs to provide low-cost financing for the installation of additional
0, 0
~ 32>>
insulation in residential dwelling units. Another scheme that has been considered is the use of rebates or the grant-ing of reduced 'rates to electricity consumers who cut usage to levels below their past levels of electricity consumption. These subsidy approaches raise serious questions of fairness and equity, since subsidies would help mainly those who were inefficient in their use of energy in the past and would do so at the expense of taxpayers or other electricity ratepayers. Moreover, the use of subsidies to encourage conservation may encourage the misuse of economic resources. During a period of inadequate electricity generating capacity, the problem is that electricity rates are too low in relationship to the market costs of supplying electricity.. The problem is not that investment costs for conservation steps are too high. To reduce investment costs for conservation because electri-city rates are too low, is in effect, trying to counter one distortion with another. A better approach would be to set. electricity at "market-clearing" levels--that is to say levels sufficient to restrain demand to the capacity available and to see to it, that small electricity customers have access to financing for conservation investments at rates that are neither markedly above nor markedly below the prevailing costs of capital. Xn any event, the potential for reduced load growth through subsidy schemes is likely to be of minor impor-tance owing to (a) the limited financial resources available to support such programs and (b) the limited array of possible
conservation steps that could receive the necessary political support o
e COMPARISON BETWEEN COMPANY FORECASTS AND NERA-Loads'rowth PROJECTED GRONTH RATES OF PEAK LOADS Peak Rates~ HERA Pro ectxons 1974 Company Company No Price C an e Price Chan e
~Co~man Actual Forecasts Forecasts Low-Growth Case Hx h-Growth Case Low-Growth Case Hx -Growth Case TIT 2,032 5,966 .0800 .0489 .0696 .0399 .0932 EPE 638 1,519 .0639 .0636 .0691 .0620 .1102 PNM 583 2,003 .0922 0651 .0798 .0636 .1208 9,997 19,120 .0474 .0550 ~ 0609 0552 .0959 SRP 1,845 5,090 ~ 0752 .0656 .0884 ~ 0579 .1075 Total 15,095 33,698 .0590 .0564 .0671 ~ 0543 ~ 0988 t Megawatts.
Average annual compound rates. of growth 1974-1988 expressed as fractions. Source: Supplement 4 to Environmental Report, Vol. VZ, May 19, 1975; Loads and Resources Summary, Southern California Edison Company; NERA estimates. 0
i II 0
~ ~ . ~: ~
YAJOR PORECASTZNG ASS~i TZONS Comoan Growth Assuao tion APS EPE PNiN SCE . SRP Case 75-80 1981-88 975-80 1981-88 975-80 198 -88 75-80 1981-88 97 -80 1981-8 (2) (3) (4) (5) (6) (7) (8) (9) (10) (ll) (12) Annual Growth Rate (0): Residential Customers 3.24 2.99 1.91 2. 04 2.68 2.51 2.66 2.25 5.15 4.45 Low 5.76 4.85 2.53 2.52 4.38 - 3.77 3.33 2.80 7.12 6.46 High Real Znco=e per Household 2.35 1.77 2.23 2.46 1.95 2.91 2.11 2.19 1.68
'.95 Output'.55 P~n .fac turing 5.15
~ 2.45-3.61 4.26 4.89 2.97 3.45 5.34 7.08 3.92 5.20 4 .22 4.90 2.67 3.22 ll.
14.89 36 7.96 11.59 Low High I I Average fo sector; figures vary for individual industries. Source: Company data and NEBA estimates.
o
- ~ ~
PROJECTED RATIO OP 1988 TO 1974 PRICE Com an APS EPE PNi4 SCE (4) Electricit Residential .97-1.40 .97-1.43 1.03-1.46 .83-1.19 97-1.36 Residential Htg. 1.00-1.45 1.06-1.62 1.15-1.66 84-1.22 97-1.36 Commercial 1.02-1.48 1.03-1.57 1.12-1.61 .86-1.26 .98-1.37 Industrial .91/1.22-1.29/1.83 P 1 06/1.15-1.61/1.80 1.05/1.30-1.49/1 93 .87/.93-1.26/1.37 ~ 97/1.12-1.35/1.64 So. Calif. Gas So. Union Gas (6 Gas Residential 1.06-1.66 1.16-1.91 1.30-2.16 Commercial 1.31-2.32 1.49-2.86 1.72-3.14 Industrial* 1.41/1.54-2.62/2.94 1.40/1.63-2.59/3.25 1.86/2.29-3.48/4.47 1988 price is in 1974 dollars. Where the ratio of 1988 to 1974 price is less than one, a value of unity is used in the projection. Alternative low/high values are shown to indicate the spread across indus-tries in each case. Source: Company data and NERA estimates. 0 hp
i I Q~ 1
~ ':0 POPULATION GROWTH--ASSUiMTZONS A'.ID SOURCES Low-Growth Hi h-Growth
~Con anv. 7 - 9 (2) (3) (4)
.0234 .0204 .0465 .0376 Phoenix BEA, OBERS projection Arizona, Arizona Department of Eco-nomic Security's high-end projection EPE -.0060 .0135 .0045 El Paso BEA, Texas, Bureau of Texas, Bureau of El Paso BEA, OBERS projec- the Census'er- the Census'er- OBERS projec-tion ies-E projection, ies-E projection, tion
'0- '0 migration zero migration PMP4 .0066 .0097 .0186 , ~ 0180 Albuaueraue B~, OBERS projection New i<:exico, Bureau of the projection, zero migration Census'eries-E SCE .0122 .0105 .0193 .0164 Los Angeles-Long Beach BEA, OBERS California, Bureau of the projection projection, '60-'70 migra-Census'eries-E tion
.0291 .0235 .0472 .0423 Phoenix S)ISA, OBERS projection Phoenix SYSA, Arizona Department of Economic Security's high-end projec-tion The apparent anomaly in the location of high and low growth rates arises from the characte of the factors used to adjust these figures to correspond to the service territory. When these historical factors are added, the result is a higher figure from the OBERS projection than. from the Census pro>>
jectlon 0
~ '
p
0'
~ ~ ~
ELECTRZCZTY AND GAS COST CALCULATZONS an Cost Co. oonent APS EPE PNH SCE (2) (3) 5) Electricit (C/b h) Generation:~ Base year Future 1.21 2.25 1.20
,2.25 1.05 2.25 1.56 2.25
- l. 42 2.25 Non-Generation:
Residential Residential Htg. 1.63 1.29 0.14-0.55'om 1.34 0.55 1.72 0.82
- 2. 24 1.72 1.03 Co.-.."..e "c-'al 1.12 0.72 0.98 1.21 0.97 Hanufacturing 0 '3-2.06$ 0.24-1.46~ 0.30-1.09s -0.04 -1.01s Hining 0.03-2.568 0.29 0 35 0 36s
-0.13'as So. Calif. So. Union APS Gas Gas (6) (7) (8)
($ /Hc ) Gas Supply: Base Yea- 0.58 0.59 0.38 Future 1.00"2.00 1.00-2.00 1.00-2.00 Non-Gas: Residential 1.13 0.78 0.87 Commercial 0.42 0.15 0.37 Hanufac uring 0.13-0.25s 0 03 0 26s 0.08-0.28$ Hining 0.21 0.10 0.12 Excluding losses. Not available. Figures differ according to industry. The negatives for manufacturing and mining reflect custom r load factors. Source: Company data and NERA estimates.
APPENDIX A
.0 f
~ - ~
~ ~
ARZZONA PUBLZC SERVZCE COMPA'.Pl ACTUAL AND PROJECTED SALES AND PE&K LOAD 1988 Pro ections w- rowt ases zen- row" ases Hag E ec. Low E ec. jixg E ec. Low E ec.
. 1974 No P ice 6 Low Gas 6 High Gas No Price 4 Low Gas a High Gas Sales Actual ~CllR'.I 8 Prices Prices Chanche Prices Prices (Thousand Megawatt-Hours)
(2) (3) (4) (5) (6) (7) Residential 2,540 5,153 4,497 6,218 6,982 6i068 8,410 Co ercial~ 3,131 6,619 6, 176 10,459 8,661 8,081 13,686
\
Zndustrial~ 1,780 2,453 1,891 2,752 2,906 2,241 3,262 Resale and Other 1,241 2,362 2,086 3,226 3,080 ~ 2I 722 4,211 Total~ 8,692 16,587 14,651 22,656 21,630. 19,112 29i569 Peak Load (Megawatts) Total 2, 032 3,966 3 j516 5,391 5,210. 4, 616. 7,079 Commercial sales consist of sales to other public authorities, irrigation pumping and sales to commercial or small light and power custo...ers. Zndustrial sales exclude sales to irriga-tion pumping. To al may not sum due to rounding. Source: Company data and NERA estimates. 0
e. ARIZONA PUBLIC SERVICE COMPEL ACTUAL AND PROJECTED GRONTH RATES 1974-1988 Pro 'ections w- rowt Cases Hx -Gzowt Cases Hxgh Elec. Low Elec. ~ High Elec. Low Elec. 1963-1974 No Price a Low Gas & High Gas No Price a Low Gas a High Gas Sales t~u Chan e Prices Prices Chan e Prices P ices
) (3) 5 6) (7)
Residential .1191 ~ 0518 0416 .0660 .0749 .0642 ~ 0893 Co~ercial~ 0727 0549 .0497 .0900 .0754 0701 .1111 Industrial~ 0617 .0232 .0043 .0316 ..0356 0166 ~ 0442 Resale and Other .0302 .0470 .0378 ~ 0706 .0671 ~ 0577 ~ 0912 Total .0726 .0472 0380 .0708 ~ 0673 ~ .0579 '.0914 Peak Load Total . 0914 ~ 0489 .0399 ..0722 .0696 .'0604 ~ 0932 Average annual compound zates:of growth, expressed as fractions.
~
Commercial sales consist of sales to other public authorities, irrigation pumping and sales to commercial or small light and power customers.
~
Xndustrial sales exclude sales to irrigation I h
~ II pumping.
Source: Company data and NERA estimates.
QO 0
EL PASO ELECTRIC COMPANY ACTUAL AND PROJECTED SALES AND PEAK LOAD 1988 Pro'ections Low-Growth Cases H2 h-Growth Cases Hag E ec. Low Elec. Hag E ec. Low Elec. 1974 No Price & Low Gas. & High Gas No Price & Low Gas & High Gas Sales Actual ~Chan a Prices Prices Chan e Prices Prices (Thousand Megawatt-Hours (2) (3) (4) (5) (6) (7) Residentialr 811 1,959 li789 2,629 2,134 1.948 2,847 Commercial2 1,436 3,701 3,887 7,244 3,950 4,149 . 7,743 Industrial3 388 585 394 637 634 427 694 Resale and Other 473 1,123 1,091 1,889 1,208 1,173 2,028 Total 3g108 7,368 7,162 12,400 7,926 7,697 13,313 Peak Load (Megawatts)- Total 638 1,512 1,482 2,569 lg626 1,592 2,758 Residential sales include sales to customers in commercial apartment buildings master metered. 2 Commercial sales consist of sales to other. public . authorities and the portion of industrial sales that includes non-manufacturing customers, as well as sales to commercial or small light and power customers, less the amount for commercial apart-ment buildings master metered. Industrial sales exclude sales to non-manufacturing customers. Total may not sum due to rounding. Source: Company data and NERA estimates. 0 5
0, QO
. ~
~
EL PASO ELECTRIC COHPANY'CTUAL AND PROJECTED GROWTH RATES Cases'974-1988 Pro'ections Low-Growt Hi -Growt Cases xg E ec. Low E ec. xg ec. w ec. 1962-1974 No Price a Low Gas 6 High Gas No Price a Low Gas a High Gas Sales Chan e Prices Prices Chan e Prices Prices
) 7)
Residential i .0711 .0650 .0581 .0876 .0715
'0646
.0938 Cor~rcial i .0708 ~ 0700 .0737 .1225 .0750 . 0787 .1279 Industrial i .0595 .0298 .0011 .0360 .0357 .0069 0424 Resale and Other .1575 .0637 .0615 .1040 .0693 .0670 .1096 Total 0779 ~ 0636 . 0614- ~
.1039 ;0692 " .0669 ~ .1095 Peak Load Total 0780 . 0636- .0620 .1046 .0691 - ..0675 1102 Average annual compound rates of growth, expressed as fractions.
Residential sales include sales to customers in commercial apartment buildings maste" metered. S Commercial sales consist of. sales- to other. public authorities and the portion of industrial sales that includes non-manufacturing customers, as well as sales to commercial or small light and power customers, less the amount for conmercial apart-ment buildings naster metered. i . Industrial sales exclude sales to non-manufacturing customers. Source: Company data and NERA estimates. 0
0 QO 0
~ -
~ ~ s
~ 0 ~ 0 PUBLIC SERVICE COMPANY OF NEH MEXICO ACTUAL AtlD:PROJECTEE SALES AND PEAK LOAD w
1988 Pro 'ections Low-Growth Cases Hi h-Growth Cases Hxgh Elec. Low Elec. Hxgh Elec. Low Elec. 1974 No Price 6 Low Gas & High Gas . No Price 6 Low Gas 6 High Gas Actual Prices Prices Sales ~Chan a
- (ghousansaPrices Hagawatt-Hours)
~Chan a Prices (2) (3) (4) (5) (6) (7)
Residential 828 1,939 1,777 2,752 2,418 3,380 Commercial 1,408 3,686 3,816 6,866 4,413 4,568 8,219 Industrial~ 362 559 410 668 8,463',220 682 501 813 Resale and Other 324 783 760 ~ 1,302 951 923 1,571 Total~ 6H967 6,763 11,587 - -8,212 13,984 Peak Load
-(Megawatts)-
Total 583 1,409 1,382 2,388 1,708 1,674 2,877 Commercial sales consist of sales to other public authorities and the portion of industrial sales that includes non-manufacturing customers, as well as sales to commercial or small light and power customers. Industrial sales exclude sales to non-manufacturing customers. Total may. not surL. due to rounding. The amount shown differs from the reported total Gwh because of a descrepancy between .the com-of',895 pany-reported total for large'ight and power sales
. and..the total given for industrial sales by type of industry.
Source: Company data and HERA estimates.
~ '
0 . ~ ~ PUBLIC SERVICE COMPANy OP NEW MEXICO ACTUAL AND PROJECTED GROWTH RATESr 1974-1988 Pro 'ections Low-Growth Cases Hi h-Growth Cases Hag E ec. Low E ec. Hag E ec. Low Elec. 1962-1974 No Price Low Gas High Gas No Price ,6 Low Gas High Gas
~)
5 S Sales Actual Chan e Prices Prices Chanche Prices Prices
) (3) (4) (5) (6) (7)
Residential .0914 . 062'I .0561 ;0896 ;0796 ~ 0730 .1057 Co~rcial .0948 .0712 .0738 .1198 .0850 .0877 .1343 Industrial 0259 .0315 ~ 0089 .0447 .0463 .0235 .0595 Resale and Other .1341 .0651 .0628 .1045 .0799 0776 .1194 Total .0794 .0640 .. 0618 .1034 .0789 ~ 0766 .1183 Peak Load Total ~ 0816 .0651 .0636 .1060 .0798 .0783 1208 Average annual compound rates of growth, expressed as fractions. Commercial sales consist of sales to other public ... authorities and the portion of industrial sales that includes non-manufactu ing customers, as well as sales to commercial or small light and power customers'ndustrial sales exclude sales to non-manufacturing customers. Source: Company data and NERA estimates.
0, 0
~ ~ ~ ~
SALT RZVER PROJECT ACTUAL AND PROJECTED SALES AND PEAK LOAD 1988 Pro ections Low-Growth Cases Hi h-Growth Cases Hag E ec. Low E ec. Hag E ec. Low E ec. 1974 No Price 6 Low Gas 8 High Gas No Price 6 Low Gas &- High Gas Sales Actual Change Prices Prices ~Chan e Prices Prices
-(Thousand Megawatt-Hours)
(2) (3) (4) (5) (6) (7) Residential. 2,752 6,681 5,806 7 g 733 8,750 7,607 10,092 I Commercial ~ 1, 810 5,076 ',955 7,850 6,818 6,657 10,538 Zndustrial~ 1, 691 4,069 3,399 4,577 6,023 5,053 6,768 Resale and Other 1,377 ~ 2,664 2,440 3,246 3,439 3,133 4,219 Totals 7,629 18,490 16,600 23,407 25,030 22,450 31,617 Peak Load
-(Megawatts)
Total 1,845 4,490 4,056 5,733 6,040 5~452 7i 702 Commercial sales include sales to commercial. and small industrial and agricultural pumping customers. Zndustrial sales include sales ta. large 'in-dustrial and mining customers and to a small portion of, commercial and..small. industrial users. Total may not sum due to rounding. Source: Company data and NERA estimates. Q ph
0, Qo
SALT RIVER PROJECT ACTUAL AND PROJECTED GROWTH RATES~ 1974-1988 Pro 'ections Low-Growth Cases Hi h-Growth Cases 1g ec w ec. n1gn L ec. w L ec ~ 1963-1974 No Price 6 Low Gas 6 High Gas No Price a'ow Gas & High Gas Sales Chan e Prices Prices Prices Prices
- 3) (4) (6) 7)
Residential . 1374 .0654 . 0548 .0766 .0861 .0753 .0973 Co=.ercial1 .1010 .0764 .0746 .1105 .0994 .0975 1341 Industrial~ .0461 .0647 .0511 .0737 .0950 0813 1041 Resale and Other .0463 0483 .0417 0632 ..0676. 0605 .0833 Total ~ 0845 .0653 .0571 ~
.0834 .0886 F 0801 ~ 1069 Peak Load Total .1115'0656 .0579 . 0844. .0884 .0805 ~ 1075
~
Average annual compound rates of growth, expressed as fractions. 1 Commercial sales include sales to commercial and small industrial and agricultural pumping customers. 3 Industrial sales include sales to large in-dustrial and mining customers and to a small portion of commercial and small industrial users.
~
The actual growth rate for peak load excludes sales for resale.
-.Source: Company data and NERA estimates. ~Q
(~
& r f
-/)
~
~,
~
~ r ~
r ~ cS.
I SOUTHERN CALZFORNZA EDZSON COMPANY ACTUAL A%)" PRMECTED SALES AND PEAK LOAD 1988 Pro 'ections Low-Growth Cases Hi h-Growth Cases Hag E ec. E ec. lg ec ~ w r. ec. No Price & Low Gas a High Gas . No Price 6 Low Gas S High'as Sales Actual ~Chan e Prices Pcicne ~Chan e Prices Prices
-(Thousand Megawatt-Hours)
(2) (3) (4) (5) (6) (7) Residential 13,060 27,027 25,733 36,711 .29,574 28,177 40,176 Total'974 Co~erci al Zndustrial~
~
20,517 12,583 52,893 18,506 57,946
.15,689 98,942 22,680 56,597 20,005 62,004 16,962 105,872 24,516 Resale and Other 4,931 10,430 10,530 .16,779 11,252 11,354 18,075 Sl p 090 108,857 109,899 175,112 117,428 ' ~
118'97 188;639 Peak Load
- (Megawatts)-
Total 9,997 21g167 21,206 33t391 22'71 22,903 36,026 Comanercial sales consist of sales to other public authorities and the portion of industrial sales that includes non-manufacturing customers, as well as sales to commercial or small light and power customers. Zndustrial sales exclude sales to non-manufacturing customers. Total may not sum due- to rounding. Source: Company data and NERA estimates.
L 0 ,E Ih 1
~
0 O ~ SOUTHERN CALIFORNIA EDISON CONPANY ACTUAL AND 2ROZECTED GROWZH RATES> 1974-1988 Pro 'ections Low-Growth Cases x n- rowrn ases Hag E ec. a Low E ec. Hag E ec. Low E ec. 1962-1974 No Price 6 Low Gas a High Gas No Price 4 Low Gas a High Gas Sales Actual Chanche Prices Prices Chan e Prices Prices (2) (3) (4) 5) ~6) (7) Residential .0907 ~ 0533 .0496 .0766 .0601 .0565 .0836 Commercial a.. .0711 0700 .0770 .1189 .0752 0822
~ ~ ~ 1244 Industrial .0706 0279 .0159 .0430. ~0337 .0216 .0488 Resale and Other .0650 .0550 .0557 .0914 0607 .0614 0972 Total .0749 .0555 .0562 0920 .0612 ~ 0619 ~ .0978 Peak Load Total 0699 +0550 .0552 .0900 ~ 0609 .0610 ~ 0959 Average annual compound rates of growth, expressed as fractions.
Commercial sales consist of sale to other public authorities and the portion of industrial sales that includes non-manufacturing customers, as well as sales to commercial or small light and power custor:.ers. S Industrial sales exclude sales to non-manufacturing customers. Source: Company data and NERA estimates.
8 h t ,l 1f, II K I k'l t t I o
t APPENDIX B
0 0
APPENDIX B The variables used to estimate the demand equations fall into five categories: income, price of electricity and alternate fuels, population data, housing characteristics, and weather. The description of these variables and sources are listed below. Except as noted, all explanatory variables are converted to logarithms 5.n the estimated equations. I'ATURATION DATA The variables used as the dependent variables in the appliance saturation equations except those for air conditioning are expressed as log Si 1-Si where S is the saturation of appliance i. This functional form has the desirable property that the estimated saturation will always be in the range of .0.0 to 1.0. The variables used -as the dependent variables in the'air conditioning saturation equations are expressed as S ~ log 1-gS 3 3 where S.3. or S is the saturation of air conditioning type i or j. This functional form implies that the estimated sum of air conditioning saturations will lie between 0.0 and 1.0. The saturation variables included in the study are as follows:
0, Electric Ran es, Electric Water Heatin , One Room Air Conditioner, Multi le Room Air Conditioners, and Central Axr Conducts.onin Housing units with these appliances as a percent of all occupied housing units. 2i S ace Heatin Penetration Housing units built with electric space heating from March 1960 to March 1970 as a percent of total housing units built during that time.
.3~ Clothes Dr ers a) Housing units with electric or gas dryers as a percent of all occupied housing units.
b) Ratio of housing units with electric dryers to housing units with electric, or gas dryers. Source: U.S. Department of Commerce, Bureau of the Census, Census of Housin , Detailed Housin Characteristics, 1960, 1970
'I~
INCOME The level of income imposes a constraint on the con-sumer's expenditures. The level of income will affect elec-tricity consumption in two ways'. 1) the mix of appliances the consumer will have; 2) the intensity of use of these appliances. 'Two income variables were included in the analysis: a) personal income per occupied housing unit, which serves as a measure for average level of income; and b) percent households with cash income less than $ 3,000, which provides a measure for income distributions. Source': U.S. Department of Commerce, Bureau of Economic Analysis, Surve of Curre'nt Business, June 1974; U.S; Department of Commerce; Bureau of the Census, Census of Po ulation, 1960, 1970; and Sales Management, Inc., The Marketin Ma azine, July 10, 1'960 and June 1'0, 1970.
I Qe I
B~3 III. PRICES for electricity is affected by the price t The demand of electricity as well as prices of competing fuels--natural gas, fuel oil and bottled gas. Thus, a rise in the price of electricity will dampen its use. On -the other hand, an in-crease in the price of alternative fuels will tend to increase j consumption as customers will switch to .electricity.
- a. Price of Electricity The consumer of electricity does not face a uniform electricity price but rather a structure of block rates with unit value usually declining as consumption increases. Thus, the use of average revenue as a measure of price is not satis-factory because, for example, a decline of average revenue in
~ e any particular state does not necessarily mean lower rates for all customers but could result from an increase in the number of all-electric homes which enjoy a lower rate. A better rep-resentative of electricity rate schedules that face the con-sumer is the '"typical bill" as reported in T 'i'cal Ele'c'tric Bills and All-Electric Homes published by the Federal Power Commission for classes of customers by average monthly con-sumption. These data are reported for all localities with populations of more than 2,500. State averages are also reported. In the study, typical bills were used to represent two classes of customers small users which do not have electric heat, and large users in all-electric homes. For each class, the typical bills were manipulated so as to rep'resent the "true"
market price, or the incremental cost. For the small users, the incremental cost was computed as the typical bill for 1,000 kilowatt-hours per month, less the typical bill for 500 kilowatt-hours divided by 500. Similarly, for large users, the incremental price was computed as the typical bill for 2,500 kilowatt-hours per month, less the typical bill for 1,250 kilowatt-hours divided by 1,250. Source: Federal Power Commission, T ical Electric B'il'ls: 1964-1970; Federal Power Commission, Al'1-L'1'ec'tric Homes: 1964-1970; and Federal Power Comm>aaron, Statistics of Electric Uti'1'it'ies'n the Unite'd Sta't'es,
"'rivate'1
- Own'ed, 1954-1960.
- b. Natural Gas Natural gas prices are also subject to declining block rates. Unlike electricity, however, no "typical bills" are published. The equivalent of typical bills were estimated
'using an empirical formula relating average revenue to usage: where AR is average revenue in dollars per therm Q is therms sold. per month per customer A is a parameter of the equation and therefore X = -log (AR )/log (Q
S 0'e
B-5 Using the same formula, the typical bills for 30 therms and 100 therms are obtained using the formulas: TGB = (AGB) 30 1 0 and TGB (AGB ) 100 1 " 100 00 Q where AGB. is the average monthly gas bill. Source: Amercian Gas Association, Inc., Department of Statistics, Gas Facts: data for 1954, 1960, 1964 and 1970.
- c. Fuel Oil The price of No. 2 fuel oil was computed from prices published month'ly in Fuel Oil and Oil Heat. for a number of cities across the country. The simple averages of the monthly prices were used.
.Source:. Industry Publications, Inc., Fuel Oil, and Oil Heat, monthly issues for the years 1954, 1960, 1964, 1970.
- d. Bottled Gas
\
Prices for liquefied petroleum gas (LPG) published in the U.S. Department, of Agriculture's A ricultural Prices were used. This publication lists prices paid by farmers for a large number of commodities. Since bottled gas 'is used mostly in rural areas, prices paid by farmers should be rep-resentative of the state average price. The 1954 price was estimate'd by appl'ying the average annual rate of growth for the prices of No. 2 fuel oil and
~ B 6 the gas price at. 30 therms from 1954 to 1960 to the 1960 price of liquefied petroleum gas. Source: U.S. Department of Agriculture, Crop Reporting Board, Agricultural Prices, Annual Summar : 1960, 1964 and 1970. IV. POPULATION DATA Po ulation Per Household Total population divided by total occupied housing units. Source: U.S. Department of Commerce, Bureau of- the Census, Census of Po ulation: 1960 and 1970; Census of House.ng, 1960; and Census of Housing, Detailed Housin Characteristics, 1970. V. HOUSING CHARACTERISTICS Five variables were included in this category:
- 1. Rural Occu ied Housin Units as Percent or Share of Total Occu ied Housin Units--
because natural gas is not available in rural areas, rural customers tend to use more electricity than urban residences.
- 2. Central Cit Housin Units as Percent of
.Total Occupied Housin Units.
- 3. Percent. Structures: 5 or More Units representing apartment residences that, on the average, have less appliancos because of space limitations.
- 4. Percent Structures: Built 10 Years Prior--
new houses are generally more spacious and better insulated.
- 5. Share Structures: Sin le Detached Housin Unz.1. s.
Source: U.S. Department 'of Commerce, Bureau of the Census Census of Housin , Detailed Housin Characteristics, 1970; Census of Housin , General Housin Characteris-tics, 1970; and Census of Housin , 1960.
0
~
I i
B 7 VI ~ WEATHER Average temperatures affect electricity consumption, both directly and indirectly: Directly, hot days in summer-time and cold days in winter will increase consumption for cooling and heating, respectively. Indirectly, average winter temperatures will affect, the decision on what heating system to install. In places with mild winters, the tendency is to have more electric units because of lower investment costs required to install a heating system. In cold areas, oil heat and gas heat are more common because lower fuel costs justify the larger investment in heating systems.
- a. Cooling Degree Days'his variable was included in the noncompetitive equation only to estimate the use by air conditioners. The 30 year average was used to compute 1980 demand.
- b. Heating Degree Days Thirty year averages of heating degree days were used to both estimate the appliance saturation equations and to project 1980 saturations.
- c. Mean July Temperature This variable was used only in the air conditioning saturation equations.
Source: U.S. Environmental Science, Services Administration, Climatolo ical Data, National Summar , Annual 1960, 1970; 1971 World Almanac.
BASE-YEAR INPUTS FOR RESIDENTIAL MODEL Com an APE EPE PNM SCE SRP (2) (3) (4) (5) Saturations Range .480 .400 ,470 .298 . 641 Clothes Dryer .280 .180 .450 .198 . 390 Water Heater .180 .225 .160 .105 . 340 Space Heating .180 .082 .025 .095 .419 One Room Air Conditioner .070 .038 .070 .218 .052 Multiple Room Air Conditioner .020 .007 .010 .036 .019 Central Air Conditioner .590 .090 .100 .149 .724 Evaporative Cooler .380 .790 .600 .135 .277 Usa e er A alliance Range 740 1,200 1,000 1,175 1,175 Clothes Dryer ~ 930 993 1,000 993 993 Water Heater 4,600 5,600 6,000 4,219 4,520 Space Heating 5,000 7,000 13,865 3,936 5,016 One Room Air Conditioner 2,000 1,200 1,165 720 2,267 Multiple Room Air Conditioner 4,000 2,400 2,330 1,440 4,533 Central Air Conditioner 6,050 9,100 4,418 2,360 6,800 Evaporative Cooler 1,500 1,400 390 360 1,000 t Net Usa e er. Customer 2,724 1,736 2,695 3,568 2,637
'Saturations are for 1973.
Source: Bureau of the Census; company data; and NERA estimates. C5 I
II TABLE 1 B-2'age of 2 FSTIHATED COEFFICIENTS OP THE REGRESSIOI( EQUATIONS USED TO PREDICT RESIDENTIAL USE PER CUSTOMER Saturation Coefficients Iilt10 O
. Electric Unit of All Dryors Space Kwh Usage Indepondcnt Clothes to Total Water Heating Coefficients )
Var)ohio Ran as Dr~era Iteat inc( Penetration PS ~Inde
~
endent Variables Constant -9.823 24.722 14.611 -3.826 0.088 -9.014 (-3 '05) (- 4.681) (, 1.2931 (-2 '19) ( 0.022) 3 '66) Income
)'crsonal Income por Oc-cupied Housing Unit 2.166 1.971 le 363
( 3.382) ( 1.777) ( 4.701) ~ Percentago of Households with Cash Income Less Than $ 3,000 (Percent) 0 '32 ( 4 '63) Prico Electricity (1,000 Kwh-
-0 568 500 Kwh)/500 ($ /Kwh)
("6.765) Average Electricity (1,000 Kwh-500 Kwh)/500 ($ /Xwh) 1 ~ 967 - 0.653 2 '59 -2. 221 (-Ge260) ( 3.481) (" 5.772) (-6 '09) Average Electricity (2,500 Kwh-l, 250 Kwh) /1, 250 ($ /Kwh) -3.639 (-7.091) Avorago Gas at 100 Therma ($ /Thorm) 1.768 ( 4.931) Average Gas at 30 Thcrms ($ /Thorm) 2.132 0 '97 2 '15 4 '42 ( 4.471) (- 2.577). ( 5.020) ( 7.901) age No. 2 Fuol Oil (C/Gal.) agc Liquefied Petro-
% curn Gas . (C/Gal.) )~*i .)'opulation per Occupiod Housing Unit 2.594 -4.102
(" 2o425) (-2. 103) Ratio of Rural Occupied Housing Units to Total Occupied Housing Units) 1.369 2.379 0.270 ( P.291) ' 2.884) (-1 ~ 504) Percent Center City Oc-cupied Housing Units of Units Units'Fraction) Total IOccupied Housing (Porcent) 0. 015 (- 1 397) Porcent Structuros Duilt ~ 10-Years Priori (Porcant) 0.031 - 0.019 0.039 0.027 ( 2.159) '(- 2.291) (- 1.545) ( 2.351) Percent Structures with 5 or Horo (Porcent) 0 '44 " 0.032 0.024 -0 '38 (-3.721) ( 3.522) (-1.293) (-2.90G) Weather Cooing Dcgrao Days 0. 240 30-Year Avarago ( 5.549) Heating Degrao Days 30-Year hvorago 0 '62 0.509 0.923 ( 2+697) ( 4.189) (-5.856) Fu r of Observations ~ 48 Coofficiont of Determination (R ) 0+660 Oo648 0 '19 0:796 0+688 0.845 Includos air conditioning usage. Piguros in parentheses aro t-ratiosi Not transformed to logarithm.
0 0 ESTIMATED COEFFICIENTS OF THE REGRESSION EQUATIONS USED TO PREDICT RESIDENTIAL USE PER CUSTOMER Unit of Saturation Coefficients Zndependent One Room More Tnan One : Centra Inde endent Variables Variables A/C Room A/C A/C (2 (3) ~4) Constant -37. 91 -50.64 -50.93 (- 3 07) .( ~ 2. 81) (- 2.70) Zncome Personal Income per Oc-cupied Housing Unit. ($ /Household) 3.54 4 ~ 69 5.21 ( 2.74) ( 2 ~ 49) ( 2.64) Price Electricity ($ /1000 Kwh) - 0.14 << 0.76 - (- 0.26) (- 0.96) 1.09'- 1.32)
~Housin Population per Occupied Housing Unit (Fraction) - 6.18 - 6.28 - 6.83
(-. 3.59) (- 2.50) (- 2.60) Share Single Detached Housing Units~ (Fraction) 1.74 - 1.35 3.57 ( 1.68) ( 0.89) ( 2.25) Share Rural Occupied Housing Units~ (Fraction) - 0.21 0.85 - 0.24 (- 0.25) ( 0.68) (- 0.18) Climate Mean July Temperature~ ('F) 0.1317 0 '262 0.1740 ( 7.56) ( 8 90) ( 6.54) Neer of Observations ~ 50 R 0 70 0 ~ 70 0.65
~Figures in parentheses are t-ratios.
~Not transformed to logarithm.
Source: Anderson, Residential, p. 40.
0 i I QO t
. ~
I t
TABLE B-3 Page 1 of 3 ARIZONA PUBLXC SERVXCE CO. EVAPORATIVE COOLER VS. AXR CONDITIONING SATUHATXON 1965-1974 Saturation of 1.0 Minus Evaporative Saturation of Saturation Ratio Year Coolers A/C Units of A/C Units (1)-:(3) (2) (3) (4) 1965 ~ 68 .36 .64 1.06 1966 .69 .38 .62 1.11 1967 .65 44 .56 l. 16 1968 .62
~
.46 .54 l. 15 1969 .58 .48 .52 l. 12 1970 .56 .53 .47 1.19 1971 .53 .58 .42 1.26 1972 ;46 .60 .40 1.15 1973 .43 .64 .36 1.19 1974 .37 .68 .32 1.16 Average 1970-1974 l. 19 Average 1965-1974 l. 16 Sources: Company data and NERA estimates.
l I'
TABLE 8-3 Page 2 of 3 SALT RIVER PROJECT EVAPORATIVE COOLER VS. AIR CONDITIONING SATURATION 1965 1974 Saturation of 1.0 Minus Evaporative Saturation of Saturation Ratio Year Coolers A/C Units of A/C Units - . (1) :(3) (2) (3) , (4) 1965 .61 .45 .55 1.11 1966 .60 .46 .54 1.11 1967 .56 .50 .50 1.12 1968 .50 .57 .43 1.16 1969 .43 .63 ~ 37 1.16 1970 . 414 .647 .353 l. 17 1971 .39 .71 .29 1.34 1972 . 344 .701 .299 1.15 1973 .311 .725 .275 1.13 1974 .265 .794 .206 1.29 Source: 1970, 1972-1974 data supplied by Salt River Project and 1965-1969, 1970 ~Re ublic aad Gazebbe surveys.
S 0)
TABLE B-3 Page 3 of 3 PROJECTED RATXO OF EVAPORATIVE COOLERS TO NUMBER OF HOUSEHOLDS LACKING REFRIGERATXON AXR CONDITXONING
~Com an Ratio APS 1.190 EPE 0.910 PNM 0.730 SCE 0.226 SRP 1.350 Source: Bureau of the Census; company data; and NERA estimates.
( 0
~
S
TABLE B-4 REGRESSXON RESULTS FOR COMMERCXAL SALES (48 STATES, 1971) r Ex lanator Variable Coefficient ~ . Electricity Price -0.819
~
{-2.81) Gas-Oil Price Xndex" 0.'810 ( 2. 04) Total Real State Personal Xncome 0.934 (16. 08) Real Xncome per Capita 0. 802 ( 1.09) Fraction of State Population in Urban Areas~ l. 51 ( 2.32) Population Density ;5. 54x10 4 (-1. 95) Agriculture Index l. 54 ( 1 ~ 18) Mining Xndex 4.79 ( 2.25) Transportation/Communications/Utilities Xndex 7 Contract Construction Xndex -6.11 {-1.42) Coefficient of Determination (R2) 0. 93 Standard Error 0. 31 All variables transformed to logarithms except as noted. FPC typical electric bill for 40 Kw/10,000 Kwh minus that for 30 Kw/6,000 Kwh. Figures in parentheses are t,-ratios.
"Index is a geometric weighted average of gas and oil prices, the weights being quantity shares computed on an equivalent-btu basis.
Not transformed to logarithm. Index is share of non-manufacturing earnings due to the sector in question; not transformed to logarithm.
~Not included in final equation owing to lack of importance and statistical significance.
4~
~ 'I I
t l 0 Cl
INDUSTRIAL SECTOR REGRESSXON RESULTS Ex lanator Variables~ E ectrxc ty Fuel Nage Ave. Farm Ave. Jan. No of Indus Grouo Price Price Rate Size Min. Temo. .Observations R
- 3) SEE'77788
-0.41 -0.16 42 0 16 0.20
(-1.93)3 (-2.15) Textiles -1. 54 0.49 23 0.53 0.29 (-4. 65) (1.28)
-0.97
(-1.90) 0 '0 (0.50) 1.79 (1.69)
-0.75
(-2.37) 1.77E-2 19 0 58 0.34 Lumber'aper (2 02)
-0.65 0.68 0.65 32 0.31 0.44
(-2. 00) (1.52) (0 95) Chemicals -1.39 1.33E-2 40 0.51 0.51 (-5.15) (1 ~ 73) Petroleum -2.14 -0.43 :.15 0.44 0.40 (-3.08) (-2.62) Stone, Clay -2.03 0.70 1.22 -0.55 35 0.60 0.27 (-4.54) (2.39) (2.79) (-4.67) P. Metals -1.19 0.28 1.15E-2 33 0.42 0.53 (-3.77) (0.70) (1.26) Fab. Metals -0. 75 -0.24 35 0 ~ 26 0.24 (-2.47) (-2.26) Machinery -0.85 0 ~ '30 -0. 17 35 0.50 0.19 (-4.00) (1.82) (-2. 62) Elec. Mach. -0.46 0.18 1.24E-2 33 0 ~ 29. - 0 28 (-1.52) (0.65) (2.33) Trans. Equip. -0.43 (-1.53) 0 '7 (1.04)
-0.08
(-1.06) 28 0.16 0 '9 All variables transformed to logarithms except temperature. Standard Error of Estimate. Figures in parentheses are t-ratios.
o. 0 ~
TABLE B-6 GROWTH ADJUSTMENT FACTORS'~om Row Model an Estimate Assum tion (2) APS 0. 59 0.59 EPE 0.40 1.00 PNM 0.00 l. 00 .SCE 1.46 l. 00 SRP 1.89 1.89 Ratio of industry growth rate in company service territory to that for corres-ponding BEA. 2 Actual figure was -0.72; but a negative number is highly implausible and must be rejected. Manufacturing only; mining figure is 1.20. Source: NERA estimates.
4 e E 8
Paqe 1 of 3 ARIZONA PUBLIC SERVICE CO. LOAD-CONVERSION AND LOSS FACTORS 1970-1974
,'Customer Class 1970 1971 1972 1973 1974 re (27 F33 AT (5)
Residential 2. 48 2. 39 2. 36" 2.30 2. 38 Commercial 1.76 1.76 1.76 1.75 l. 77
'ndustrial 1.23 1. 30 1.29 1.28 1. 28 Other" 1.88 1.74 1.76 1.45 l. 77 Losses 0.11 0. 11 0. 11 .0. 11 O.ll Notes: Load at system peak divided by average load.
Losses divided by total load ex losses. 3Includes irrigation.
"Includes public authorities and sales for resale.
Source: Company data.
s i
~
~ ll
8 RP-CONVERSION ACTUAL AND PROJECTED LOAD AND LOSS FACTORS 1965~1974 Total Large Large Including Distribution Average indus trial Average Mining Average Total ~ Transnission Year Peak Load Load Ratio Peak Load Load Ratio Peak Load Load Ratio Peak Load Losses Ratio (iviw) (Mw) (1) ": (2) (Mw) (Mw) 4) ~(5 Mw (Mw) I) -:(8) Mw Mw , (11) : (10) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) 1965 . 465 222 2.095 NA 30 NA 1966 522 243 2.148 50 33 1.S1S 57 1.188 629 652 1.037 1967 583 278 2.097 55 37 1.486 26 29 897 664 679 1.023 1968 626 287 2.181 57 1.239 59 1.439 742 762 1.027 1969 757 345 2.194 87 59 1.475 59 .966 901 944 1.048 1970 848 376 2.255 86 63 1.365 77 72 1.069 1011 1055 1.044 1971 957 438 2.185 103 67 1 537 77 66 1 167 1137 1120 985 1972 1143 498 2.295 83 69 1.203 92 73 1.260 1318 1360 1.032 1973 1226 541 2.266 102 1,378 109 79 1.380 1437 1448 1.008 1974 1363 591 2.306 115 77 li494 135 100 1+350 1613 1645 1.020 Projected 2 '0 1.50 1.35 1.020 Excludes sales for resale.. Source: Data supplied by Salt River Project.
8~ 0 QO
TABLE H-7 age o LOAD-CONVERSION FACTORS Com an Customer Class EPE PNM SCE (2) (3) Residential 1.91 1.43 2.24 Commercial ~ 1.92 1.98 1.48 Indus trial 1.31 1.49 1.56 Other 4 1.43 1.84 1.72
'Include:losses; adjusted to conform with actual 1974 peak loads.
~EPE and SCH based on 1973 data; PNM based on 1974 data.
'Includes sales to other public author-ities.
4Sales for resale. Source: Company data.
0 QO
jon2 531 ~..': . ~,l ~... ~ a definitive discussion of the studies which indicate 2 50 or more percent'f the installed capaci ty of the Applicants should be baseload facilities." Now, have you reviewed those responses to those questions? A Yes, I have. Q All right. And could you briefly and generically answer the question at this time for the Board? 10 I'Iithout actually retracing all the numerical steps and attempting to duplicate the analysis done, I 1.2 couldn't say with precision that the 50 percent criterion 13 is the one that should be met. But what I can say is that the -summaries that, 15 these companies have presented indicate that to me they I6 have .looked at the factors that are relevant and necessary
, l7 to be looked at in order to arrive at a correct judgment 18 about the fraction of capacity that is baseload.
I9 They have looked at the character of the load 20 duration'curve, they have looked at the relative costs p 21 the capacity, and relative cost of ooerations of each of 22 the different types of capacity they could use. 23 These are the factors that determine the 24 optimum mix of capacity and the optimum share of capacity 25 that should be baseload.
0 t
.0
, ~
'on3 532
. Has eel on an intuit i ve judgmental 1 eve l, 2 50 percent sounds reasonable to me.
3 As I say, that is something that depends for 4 precise demonstration on precise calculations. Q Mr. Anderson, when you say some thing like 50 percent should be baseload and you talk about precise calculations, you don~t mean it has got to be exactly 50 percent. It is a range around 50 percent. Isn't that correct? 10 A Yes, that's right. 4 MR. NORT()N: At this time we would ask that the
~ 12 multiple response to Question 3 be marked as Applicants'3
',Exhibit 7.
14 CHAIRMAN CLARK'- It wil1L be so marked. (The document referred to was 16 marked Ag1ol;i.cants'xhibit 17 Number 7 ffor identification.) 18 MR.. N()RT()N: Ne also move it be admitted in 19 evidence if there is no ob jection. 20 MR. LEI")IS: No objection,. 21 I (Applican'.tx'xhibit Number 7, 22 previously marked for identi fica-23 tion, was rieceived in evidence.) 24 l3Y MR. N()RT()N: A Mr. Anderson, I know you weren't here yesterday,
~ '( X
- 3. QUESTION Provide a definitive discussion of the studies which indicate 50 or more percent of'he installed capacity of the applicants should be baseload facilities.
ANSIER: Each of the participants, excluding AEPCO, has prepared a response to this inquiry. A. Arizona Public Service Com anv APS continually updates its plans for future generating units and other resources in order to supply its customers with the most economical energy available. These modifications to the Company's long range plans are made necessary by load forecast revisions, changes to the economic characteristics of new generating units, including the new fuel escalation rates, and the availability of new types of energy generation. The shape of the annual load duration curve for any specific electric system determines the proper relationship W of base load capacity, intermediate load capacity, and peaking capacity for the most 'economical operation of. that particular system. Base load units are defined as those units which have relatively high capital costs, relatively low fuel costs, and are designed to operate continuously for long periods of time. Intermediate units are those units with relatively lower capital costs and higher fuel costs than the base load units,
r. 0 L 0
.and are therefore utilized for shorter periods of operation. Peaking units are those units which generally have the lowest capital costs, but the highest fuel costs, and are usually designed and utilized for shorter periods of operation. Economic considerations (principally fuel and capital invest-. ment) dictate the relative positions of the different types of capacity under the load duration curve. Since base load units are those units wh'.ch prove most economical'f operated at near maximum capability for the maximum amount of time, the most efficient ratio of base load capacity to total capacity for any electric system is the one which has enough base load capacity to provide all of the energy under the "knee" of the system load duration curve, after allowances for capacity lost due to scheduled maintenance and forced outages of the units. Results of studies of the APS system show that fifty or more percent of the installed generation should be base load facilities. B. Salt River Project: Agricultural Xm rovement and Power District The Salt River Project conducted a 'study, which was completed on December 24, 1974, to determine the ideal mix of resources on the SRP system that would result in the lo>>est overall cost in meeting its generation commitment. Xn that report some general assumptions were made and are listed below:
- 1. Resources will be brought on line only when needed to meet load and will operate at the inost efficient generating point.
2 ~ Onc hundred percent of load will represent system sales plu firm sales to other utilities. 3-2
~
~ 0 0 S
- 3. Xnstalled reserves will be 154 of load.
- 4. Resource types and associated costs used were:
Investment Cost ( 0/KPf) Combustion Turbines 100 Combined Cycle 200 Coal (350 Hh'lass) 400 Nuclear (1270 Vite Class) 400 Pump Storage 200-280 Annual Fixed Cost Variable Cost ( $ /Kh'-Yea r ) (9/Imi>W) Combustion Turbines 17 30 Combined Cycle 34 20 Coal (350 IC" Class) '56 10 Nuclear (1270 IÃ Class) 53 3 Pump Storage 24-34 1-3 (or less';
- 5. The resource type makeup of installed reserves capability will be proportionally the same as for load carrying capability .
- 6. Sufficient excess energy will be available at
$ 10/Hh'II or less to supply the pump storage require-ments.
- 7. The eventual resource mix will be accomplished at such a distant time that current thermal resources will be retired and conventional hydro and purchase resources will be a minor part of the total picture.
By using the previous assumptions, the following methodology was used in determining the resource mix: A. $ /KNYr. cost points were calculated for 0 and 100"..load factors. B. Linear ~/1QlYr. cost v. load factor graphs were drawn for each resource type. C. Optimal load factors were determined by finding the intersection point between resource types on the cost-load factor graph. D~ Xn order to detormine the amount of required load carrying capability, the optimal load factor quantities were tran ferred to a load duration curve. 3-3
Q a 0
In order to deter...ine th total re.,ou<rcr mi>:, installed r=ser'e capacity rec:uire was com-bined ~;'th load-carrying capability. The results of thi stu.'.. indicated that the resource m~r. for SRP's generation capacit~ should be compr'ed o 20=. combustion turbines, 23.~ pump storage, anQ 57~ base lo'.d units. C. El Paso Eluctric Cornea.g~ The actu<<l mix of bas~:;, i".termediate and pea): loaQ generation that an individual s:stem has, dep nds directly upon the range ';nd duration of the s; stem load swing. This variatio::, a ch<.racteristic o~ each system is reflected in the load duration curve. 1'he annual load duration curves for El Paso Electric have genera'ly the same profile year after year but are offset reflecting annual system load growth. The economics of operation will tend to maximize the base load component in a system, since by design, base load units are for continuous operation at near maximum output. D. Southern California Edison Corn~an As stated in Supple-. ent Nn. 6 of the Palo Verde Vuclear Generating Station Environmental R~ port (page S6-Bl.2-10 all), base load capacity of approximately 40 to 50 percent of total system capacity is considered desirable for the Edison system. It is also stated that this range of desired ba e loaQ capacity is b<ascd on an evaluation of'he characterisLics of'he system load patter>>, and the magnitude and ratio of the fiicd and va l. ialblc cos ts of f Bpoc3. z.c rcsoul cos ~
- l. SCR System Toad Patter>>. Analyses 'of hi .toricil SC}";
systen load QaLa for. the period l96l through 1972 indicates that 3-g
kg approximately 40 percen'" a....ua'eak load was con t3.nuous ly demanded throug out cac yea - Since t.le oil embargo 0 1973 the ef ect of energy conservation has been to reduce energy sales more than pea'.". demand, resulting in slightly lower syste.( load factors an>. lower cont'nuo;-.s system demands. Although this trend is present ly expected to cont'ue through 1977, a gradu."-1 increase in both load factor and continuous demand is forecast af ter 1977. By 1986, it is p ojected'hat load factor and continuous demand will stabili.".e at value of approximately 60 percent and 40 percent, re pectively.
- 2. Fi>:ed and Vari.";ble Gc ~eration Cos'. s. The results of studies compa::.'>>g both the fi>:ed (capital. related) and variable t (fu~:1 and OG!') costs of specific resource types over a 30-35 year operating period indicate that it is more economical to serve the continuous syst: em demand with base load generation, i.e., nuclear and coal generating units; than with other resource types such as oil fired generating units. Because 'the variable costs for intermediate and pea};ing resources are substantially higher than for base load resources, intermediate and pea):ing are not economical if operated at high caoacity factors (65-80"=-).
For many years, Edison has performed studies, which evalu-ate, capital and operating costs, present worth of future revenue t recJuil. omen ts financing'onstraints and envi onmcnt tl factor to develop economically and environmentally acceptable generation resource expansion plans. A revie~. of these planning studies indicates that base load capacitv of approximately 40-50"., of the total installed gu>>eration capacity is desirable for the Ed3.son system ~ 3-5
8 E. PuI lic Service Co,.pan ~ of'.e;; !mexico Public Service Cor.".pany of tie~'".exico (P '.:!) is noi. adding base loaQ generatin:. units for a number o r~;asons. The only two proven fuels available to PYii at present for new generation are co=1 and nuclear. Fo economic reasons, utilization of eith..r of these fuels dictates the resulting generation be of the base load type. This is esoc cially true in. New Ilexico because of the large capital e':penditures needed for highly efficient sulfur dioxide removal equipment in the. case of coal-fired units. Using natural gas or oil for new interriediate load units is uneconomic and in the opinion of PHD's management, a poor use of limited resources, and can only be considered in the limited case of peaking units. ln fact, PNh is of .the opinion that, the fuel supply of its exis'.ing ga /oil fired units will be in serious jeopardy by the comi mercial operation date of PVNGS Unit Yo. l. PNH has no plans to expand its total use of gas anQ oil and is currently exploring ways to reduce its present consumption of these fuels. The key to reducing- this consumption is addition of more base load units. The most promising way of expanQing PNl'i's peaking capacity is by constructing a pumped storage hyd~-o facility. Because such a facility typically consumes one-and-a-half Kt"!1 or morc for every IU'h produced, it requires complcr>cntary base load capacity. Xf such bas>> load capacity is not available, thc pumped sto ago option is precluded. 3-G
V Cl
PVi'1 1s l.:: i11c pro ess 0= c.langing its Paul: Powe Resale ra' structu'-e to a "tim: of day" structure for its three largost customers un(ler T " juri..dict on. ThQse cu-.ltomers
~
each currently own a gas/oil fired cenerating station. The intent 0 the ri G structure in thes( ca. "s is to encoura(e these custo, rs to util ze the' gene": tion for peakin('urpose-1 rath.r than for base load. The result for the FPC customer is reducGQ us of gas <<nQ oil with no economi penalty. The result fo:-. P!il1 is le. s ccntribution to the P~i'.:1 system peak load anQ increased energy sales to these customers. ln fact, it is possible u:lder this rate structure for these custor,.ers to have load factors well in excess of l00"-o- when the customers contribution to the coincident PN!1 system peak load is used J thu . increasing PNN's overall load f'actor and utilization of base load capacity. At least one of these customers and possibly two will require energy, presumably from PN'1, to complement capacity it N is expecting to obtain from Colorado River Storage Project (CRSP) . The CRSP capacity will not have any energy sold with it and will require all energy receive . to be paid back. Such energy pay back. will not contribute to the Phi'l1 system 'peak but will contribu-:.G to the utilization of basQ load capacity. PNb1 )'las historic<<lip useQ itself or marketed to other utilitics ovor 80;-. of the energy available from its base load units. All of the above circum"tances reauire most of PY!1's nQw gQllel.a tio>> addi t 1.0:1, ovc'- thQ next QQca(.Q t 0 be of tllQ b<<;0 load typQ. 3-7
i Cl S
p jon4 533 1 but a gentleman by the name of Vr. Cable from Las Cruces, I 2 New Mexico, who I believe was an officer of an organization 3 called CAUSE, presented a statement for the Board that 4 while it was a rather technical statement, I believe 5 discussed many of the subjects that you have discussed 6 in your Ilera Report. rlave you had an opportunity to review the 8 trans cr i pt o f Mr. Cabl e's test imony? 9 A Yes, I have. 10 0 All right. Pfould you at this time feel 11 comfortable in addressing the questions raised by '1r. Cable? 12 A Yes. 13 9 Nould you please do so? 14 . A What I would like to do is respond to a 15 number of points and questions raised by as part of 16 Mr. Cable's statement. 17 The sections of the transcript to which I will 18 be referring begin on page 3 well, one section that 19 begins on page 304 and it runs through page 307. 20 ()n page 304 I have just one very minor comm nt. 21 ()n Item 6, which is discussed in lines 13 through 22 16, the effect of revising the rate structures, particulrly 23 in the form of say time of day rates or peak load pricing 24 'ould be more likely to reduce peak demand than to reduce 25 baseload or base demand.
t C f f l t II
jon5 534 So I would take issue .with the statement as it reads on lines 13 to 16. 3 The effect of rate structuring, rate 4 restructuring might be to if anything, should be in fact to increase the load factor of the company to smooth the load duration curve, to flatten it, and in effect t'ut to increase the relative proportion of capacity that would 8 be baseload, hence the increase baseload capacity relative to peaking or intermediate capacity. 10 9 Q 'o what you are saying is that factor raised 1 1 by Mr. Cable, instead of decreasing the demand for a 12 nuclear facility which would be baseload; would actually 13 not affect that at all? A Might actually increase the demand for base-
.load facilities at the expense of peaking facilities.'lould 16 reduc'e the demand for gas and oil fired
.17 energy?
18 Yes. 19 Q Please continue. 20 A ()n page 305 ~fr. Cable concludes the load 21 forecasting methodology employed in preparation of Section 8 22 should be checked against'lternative. methodologies, lines 23 23 through 25. In my judgment the study I performed is in fact 25 -an independent check by means 'of an el'ternative methodology
1' f H i Cl
jon6 1 on the, forecasts made by the company. 2 0 I assume Mr. Cable did not have the advantage 3 of'our study. I don~t believe we had it until about a 4 week ago, so I am sure he didn t. 5 A ')n page 306 Mr. Cable re fers to- a study by a 6 Mr. Stevens and Creek which adapts the methods used in an 7 ealrier,study by Mount, Chapman, Terrel. I have not seen the study by Stevens and Creek. 9 . I have seen the study by Mount, Chapman and Terrel. In 10 fact, I cite it in my report. 11 I believe there are some def iciencies in the 12 methodology employed by Mount, Chapman and Terrel. it was 13 a good study at the time it was done. I t, is already a 14 . coupl e o f years old.. 15 I think we have advanced the state of the art 16 demand estimation somewhat beyond the methodology employed 17 in the Mount, Chapman and Terrel Study. 18 . Q The'eficiencies in this previous study, 19 have you when you say there are deficiencies, are those 20 same deficiencies in your study? 21 A 'o. I have attempted I think successfully in 22 most cases to eliminate most of the serious weaknesses in 23 the Mount, Chapman and Terre 1 Study. 24 simple example- the Mount, Chapman and Terr el
~
A 25 Study treats industrial demand as one sir@le total. As in
1 .0 536 jon7eot I my. study, I break it down into a number of different 2 industries. I hope more accurately to reflect the 3 cfi ff er ing technologies o f those different industries. Similarly in the residential sector the 1'tount, Chapman and Terrel Study estimates residential e 19 demand as a single chunk total. I tried to break it down into a number of its essential components. Q Excuse me for interrupting. 10 A Those are just a couple of examples of the di f ferences. 12 Besides, in the study itself I do indicate
'13 some other technical differences.
14 16 17 18 19 20 21 22
~
23 24
mmifp20 537 you have any other comments to Mr. Cable~s ( ~ .- 2 1 Q. stat ements? Do A ()n page 307, Mr. Cable asks a number of questions. 4 Some of these questions are relevant to the study that I 5 per formed. 6 I understand that he was I guess he was really 7 asking not of my study, but of the company's studies. 8 Q As a matter of fact, I think he was asking for 9 the Staff's analysis. 10 A I see. All right ~ 12 I simply wanted to point out that the number of 13 the factors that he mentions in thse questions including 14 electricity prices and price elasticity were taken into 15 account in my study. And that my study does assume a signifi-16 cant long run prie'e elasticity in nearly all user sectors. 17 Q All right. 18 Would you just please continue through that, or 19 is that pretty well it? 20, 1 That is it. 21 Q- ()kayI . 22 In your report, did you have a footnote that dealt with Public Service Company of New Mexico, something that ws not included in your report?
~ 24 A I did want to add a brief footnote in respect to I'3
mm2 538 (
~ 2 Public Serv).ce of New Mexico s forecast.
As I mentioned earlier, I developed my forecast for the resale category purely on a proportional basis to estimated sales to the end user customer classes, commercial and 5 industrial. 6 In the case of Public Service of New Mexico, that particular, method of forecasting retail sales turns out to be not very accurate. It is independent,, but not very accurate 9 'ecause now projected contractual loads for the. Public Service 10 Company of New Mexico substantially exceed the estimates that I obtained. So my forecast, if adjusKed for the change in 12 already existing contractual commitments of the Public
. 13 Service of New Mexico, would have to be revised upward.
Q In other words, their demand for power in '82 as projected in your study, are even lower than you projected in 16 your, study?
.17 That is right.
18 MR. N()RT()N: I have no fvrther questions of this 19 wi tness at this time. 20 CHAI RMAN CLARK: Mr. Lewis?, 21 MR. LEWIS> Mr. Chairman,, one brief question. 22 C R() SS-EX AM I:Wil'I() N
'I 23 BY MR. LE1'1I S!
Q Mr. Anderson, page 21, yomri report, I believe, list of
'I
'I looking at number 25 there, which is one among a
S mm3 539 I,procedures used in deriving price projections and in number 2 5, you start out by referring to. the procedure of estimating 3 the cost of future generation capacity in erm of base year dollars. Then it goes on. 6 A Yes. 7 Q I am wondering if you could outline for us the 8 steps that you take in estimating the cost of the future gener-" 9 ation capacity? 10 A What I did to well, basically, it is a judgmental 11 process. There are no good, solid, fixed numbers that are 12 necessarily going to come true. 13 i'(hat I did was look at what the project itself was 14 projecting for the costs of the nuclear power plant, the Palo 15 Verde plant. 16 I also 'looked at cost 'estimates that ha've been 17 assembled by other researchers in my organization, concerning 18 the cost of new nuclear and other tyoes of generating units, 19 and then made a judgment, a subjective judgment as to what a 20 plausible, perhaps conservative on th'e high side, estimate of e 21 generating, average generating costs might be for new capacity 22 taken as a whole. 23 I assumed in this particular study a figure of 24 22 1/2 mills a kilowatt hour in '74 dollars for all types, of'5 new capacity averaged together.
1 2
'n 25
~
mills a 1975 dollars, that kilowatt hour. MR. LEillS: Thank- you. would already be about 540 CHAIRMAN CLARK: That concludes your questions? MR. LEWIS: Yes, sir. CHAI RM AN CLARK s Dr . l Mc Co 1 um? MR. MC C()LLUM: I think I would like to pursue a 8 more intuitive feel of what your model does so that we can get 9 an understanding of what it does as compared to what the 10 prediction either by the Staff or the Applicant has already 11 done, the econometric model. 12 It is sufficiently c'omplex in-here that it is not 13 readily apparent to a layman reader who has not really gone 14 . in depth in it okay, we have only received it recently. 15 If I were to ask you, how do you account for .abrupt 16 changes like the recent two-year input of energy problems that 17 we have had, how could you answer that? 18 The problem of the escalation of cost, and the oil 19 and the gas problem that has occurred? 20 THE 1'/ITNESS: The kind'of models well, the kind 21 of. models or submodels would go into a forecasting'odel are 22 designed really to try to accommodate the effects of abrupt 23, changes. If there'eren't any abrupt changes maybe you could 24 simply say the future was going to grow demand growth was 25 going to occur in the future at the same rate that other
mm5 1 1 'ccu-s in the past. It is only because ..things are going to 2 change that you worry about, doing a forecast of this kind. The way.you try to accommodate or account for the
'4 effects of abrupt changes, is to try to establish a cause 5 and effect relationship between the factors that constitute 6 that change,and the thing that you want to project.
7 In other words, you try to establish a cause 8 and effect relationship between prices and energy demand, 9 between income and energy demand, and so on, so that i f you 10 do have a sharp increase in prices, or i f yo~> go through a 11 period in which economic growth is slower than it has been 12 you can translate that into a revised -electricity demand 13 forecast. WR. WC C()LLUW: Nhat would your model predict now 15 that you have that abrupt change? 16 By the 'way, I presume that that abrupt 'change was 17 certainly included as fundamental data, even though you used IB 1970 information from as you stated earlier. 1'9 THE 1'(I TI1ESS: Yes. 20 Cause and effect relationship, or the structure of 21 the cause and effect relationship was estimated on 1970 data. 22 The way I actually developed the way I actually 23 developed the forecast then is to use those cause and effect 24 'elationships, taking 1974 as a base year and projecting l 25 future values for the causal factors out beyond 1974.
542 I 1 Now, actually I did do a little bit of a shortcut. J 4 2 I didn't attempt to precisely measure the exact price changes 3" that occurred in 1975, in the first year of my model. Rather, I made the 'assumption that prices would rise rather steadily over time, throughout the forecast period. So I may have actually understated a little bit the 7 effect of price rises recently, and perhaps overstated them a little bit later. But I might mention that one'hing that is that has to be borne in mind is that the price projec-10 tions that I make are real price projections.'hen you take out the effect of inflation generally in the economy in other 12 words what you have to do is measure the-price of electricity 13 as compared with the prices of all other goods and services in general ~ Nhen you do that you find that the price increase that we have had over the last couple of years hasneen 16 quite as abrupt as it looks like in money dollars. 17 In other words, what you are saying is that 18 electricity has gotten more expensive, but so has everything 19 else, so relative to other goods, the change hasn't been 20 qui te so abrupt. 21 MR. WC C()LLUM: That gets me to a point that might 22 be worth asking.
'23 .
I f I were to ask you what the pro jected no, 24 what the actual escalation rate was for the year 1975 in 25 construction costs for a nuclear plant like PVNGS, what
mm7eot 543 1 percent, would you say? I j 2 THE '1')ITNESS' couldn'. 3 I didn't specifically try to answer that question, 4 or even pose it. e 20 5 MR. MC C()LLUM - How about power plants generally 6 across the country? = Did you look at them generically? 7 THE NITNESS: I looked.,at construction co ts in 8 power plants as of 1975. In making my judgment about the 9 22 1/2, now 25 mills in ~75 dollars, per kilowatt hour, I ., 10 assumed there would be some additional escalation above and 11 beyond inflation, but farther on out in time, power plants 12 would not escalate any more rapidly than-goods and services 13 in general. 14 In other words, what I am predicting is basi cally a 15 leveling off of the rate of escalation of new power plant construction cost at about the rate o~ inflation, which could 16 17 'e 6 percent, 7 percent, 8 percent, d'epending upon how fast the 18 inflation in general is. Remember, t1hat is an average overall 4 19 type of power plant. Not just nuclear. 20 21 22 23
S , ~ () FP21 avl ,':,-... 2 1 the MR. MC C()LLUh't: elasticity of electrical How do you take power and the into consideration price in your 3 model?
'I'HE NITHESS s I take that into account first by 5 looking back at 1970 or '71, at these regions of the country, 6 the 48 or 50 states, and seeing how high electricity prices
~ 7 or fuel prices in one region affected demand in that area, 8 and how low prices for electricity or fuels affected demand 9 in that area.
10 I do this not on a 2-by-2 comparison basis, but 11 on a statistical basis with a statistical model. 12 In effect, what I try to do is find the best 13 causal relationship that I can to explain variations in 14 demand related to variations in prices: across these regions, 15 and by doing so, then I arrive at an elasticity which is 16 simply derived from the predicted effect of price on demand, 17 as derived from the statistical variation we observed in 1970 18 or -'71 across the United States. 19 It is basically in simylest terms, it is a 20 curve-fitting process, but it is a cemplicated kind of 21 curve-fitting process because there are multiple dimensions 22 involved and 23 MR. MC C()LLUM! That meam I can'0 find a single 24 figure or even a table of figures th~4; I can compare to for price elasticity, overall or for eadh area of this P[NGSl is
av2 545 1 that true? In your report. C 2 THE NI TNESS! No. There are tables in the report 3 that give the elasticities or the coefficients that determine the II elasticities. 5 MR. MC C()LLUM: 1'1ould you find the place in your 6 report and let~s go over that? 7 THE 1'1I I'NESS 'Yes. These are found+iin the appendix, 8 Appendix B. The first table I want to call your attention to C 10 is Table B-2. Now, if you look under the these are some equations for the saturations of a number of different types 13 of residential systems, electricity using systems, and for the "all other" category which is identif ied in Column 7 as 1 5 noncompetitive. 16 If you look at the column headed "Independent 17 Variables," you will see a number of variables identified as price. 19 Now, the price variables that were used in any 20 particular equation are determinable by taking a particular 21 column which represents an equation. 22 Say column 2 represents the equation for ranges. 23 I f you look down there, you will see there is a number minus 1,967. That corresponds to the effect of the average electricity price measured by 1,000 kilowatt hours 'minus 500
~ av3 546 kilowatt, hours, divided by 500. No'r(, it 'happens that in the p'articular mathematical 3, form of the saturation coefficients or the saturation equations let me restate that. That was not very clear.
You worked this in, I presume, by 24 your regression equation.
25 THE I'/ITNESS- Conservation is worked into the
forecast in two ways. The first way it is worked in is that prices go up, and because prices go up demand is reduced from what it otherwise would have been. In other words, the meaning of price elasticity is --- well, it has trio meanings. The meaning of price elasticity is, one, people can conserve more. Two, people switch to other fuels, and unless they also get more expensive in which case there may be a stand-off on that second aspect .of it. 10 And so conservation is reflected in the forecast insofar as they are measured by response to price. It is 12 also reflected in the forecast insofar as there was a 13 conservation in 1974, the base year. 14 For example, suppose I don't really know how much conservation there was in 1974, but let's say it. was 16 3 percent. ,17 This assumes that all future demand the way the 18 model is set up it assumes that all future demand will also display that 3 percent conservation that was in effect in 20 the base year plus any additional conservation that would come 21 about as a result of future price increases. 22 23 25
Ih FP T22 cam1 1 MR 1<C C()LLV!~1 'o I read that c orre ctl y 550 to be hat
'2 it woulcf be at the rate of 3 percent or empirical].y the next 3 year if it is more than that you implement it to a nevi con-4 ser vation figure?
5 THE r'i'ITNESS-'orgetting about the price increases 6 and talking about the 3 percent in the base year, it assumes 7 all increments to the demand are going to be also 3 percent 8 lower than they would have been. In other words, if in the 9 next year if you were going to add a hundred kilowatt hours 10 you would have added a hundred kilowatt hours to demand in 11 the absence of conservation. Now wi ll only'dd 97. 12 MR. /i<C C()LLU>',I'- This model, there is no rea "on 13 why you can~t plot year by year and so see demand, as I under-14 stand it ? 15 THE 1'(ITNESS: That's correct. In fact the'om-16 puter printout is -on a year by year basis. That is the way 17 the model works, on a year by year basis. 18 I for convenience summarize the results indicated. 19 1(R. j'tC C()LLUMi: I guess that may he one of the 20 'ifferences between economist and engineer. I like to see 21 c'urves and you like to see tables. 22 THE 1'1ITNESS- I think that. is true. Ne seem to 23 have a predilection for'ables. CHAIRHAN CLARK: Let me es'k this question, then. 25 Each of these questi.ons reflect som concern that has occurred
cam2 1 in various inputs to'his power plant and the need for i t. 2 'here has been several statements on the amount of electrical 3 house heating that might occur over the near future and then 4 later on. There has also been statements that Arizona is 5 particularly appropriate for having solar house heating. These are again inputs, as I understand after our 7 discussion here, which year by year you would feed the input
~
8 in as to what its effect is. )'Jhat if. I told you that 10 per-9 cent of all house heating in Phoenix by the year 1985, what 10 effect, does that have on you? How would you put that into I your model? Do you make tha t judgment, by the way, or do you 12 look and read and try to find out what is going on in the 13 technical field. that might affect these kinds of things? 4 ~ 14 THE NITNESS: Nell, let me answer your f irst 15 question first. ()ne could factor. in an assumption about the per-17 cent of households using solar heat simply by well, one 18 would have to do it by judgment, but one could do it. In 19 other words, the model itself predicts the saturation of 20 electric heating. 21 'o'w, you could make some assumption about what
.22 part of that 10 percent would come out of houses that other-23 wise would have used electrici ty, what part would have come 24 out of houses that otherwise would have used gas or some 25 other fuel and then reduce the saturation of electric heating
cam3 552 1 by that amount and just then recalculate the model and you
~
2 would 'come up with a, revised statement that would reflect
~
~
3 that factor so it could be done. It might be fair to say, although I think, 5 probably in all hope necessarily it probably would be 6 stretching things a little bit, but I think perhaps some of the response to price which represents, -as I mentioned, a swi tching phenomena, could be conceived as a switching not necessarily to fuels, but to something else like. solar 10 energy, so it is conceivable that already some of that is already some of that switching to solar is already 12 implicit in the prediction" of response to price. 13 In the old days it would have been'he response 14 to switch to gas, but today is the switch to solar. 15 I do try to keep to answer your third question, 16 'I do try to keep abreas't of what is being said about the 17 prospects for solar energy and to keep in mind how something like that might be factored into a model. I intended not to 19 do it within the formal framework. 20 For the forecasting model,. I have done, because 21 it~s still kind of a nebulous factor and if anything, I think should be handled perhaps as you suggested, after the fact 23 by means of simply doing a little extra- calculation on your 24 final answer to see what that does. I haven'0 done that in 25 this particular case.
cam4 MR. WC C()LLUM- ()pe of the ways I like to check some predicting models is to go back to a time where I have 3 data not only to the ooint where I want to predict, but also past that point. I realize on a statistical model you have to shelter yourself, I suppose, from knowing what already happened to keep you from perturbing your model to try out your model of the prediction and seeing if it will work. But have you done such work? THE lSITNESS:. I have done it. I have done it 10 not for this particular tudy, but I have done it using a very similar model recently for the Paci fic !northwest. I 12 would hesitate to go too much into that for two reasons. 13 ()ne is that a model that worked well, say between 14 the years of 1962 and '71, might not work well in the future
.and vice versa. Similarly, a model that fit well ove'r '2 and <<71 in Na hington and ()regon migint'; not similar model 17 might riot fit well between <<62 and <<7/Il here.
18 I don't know how it woulzif fit. I haven't done the
~
19 study. So I would hesitate to hazaznf1 a guess. 20 The only thing I can say a%at<<s factually true is 21 when I did fit the i'1ashington and ()roon models to <<62 and
'71 data using <<62 as a base year, l6he model predicted within 23 1 percent the actual '71 total. fheae were some sizable differences within some of the individual usage categories, 25 but the total came o'ut nicely..
0 S
can 5 WR. MC C()LLU!1: Yes. k I beli eve that's all. DR. STOBER: I rea lly only ha ve one question to follow -up this discussion. I,feel a little uncomfortable with taking your 5 model at face value. I would ask the question whether it would 6 be possible to present any curves that would at least make 7 me feel a little better. I like to look at curves, too. 8 Ãould this be possible to support your reoort?
, T13E WITNESS: I would be happy to supply curves 10'2 if you would help me a little fait by telling'me what sort of curves you have in mind?
DR. ST()HER: I would like to see some type of 13 curve that integrates some of the factors in your model, 14 but with the perturbation over the last couple years in . 15 accelerated prices and this type of thing, how good a fit you 16 get with a given set of assumptions and whether or not, as 17 you said, in your Pacific Northwest model, whether maybe some 18 of the individual submodels are way out in left field, but 19 the total comes out about right within.1 oercent. 20 1')ell; that~s the kind of thing I would like to 21 look at and be able to say, well, maybe he is way off here 22 and this other submodel is a compensating factor, but in the 23 end it, is a pyramid of judgments that you have to make, never-24 theless it comes out within reason depending upon where you 25 are headed, so that gives me a little more security in, trying
cam6 555 1 to evaluate your work. 2 TIIE l'/ITNESS> There is no reason why I coul do 't 3 do a at least do, a what you might call a backcast. mentioned '72 and '71 because those are years for which e 22 5 there is pretty good historical data for purposes of feeding 6 into the model. I could do that fairly easily. My'endency, I suppose, would be to summarize the I results in a table, but maybe I can do it graphically. The problem with graphs is we are always working in multiple 10 dimensions and it is so hard to squeeze everything down to 11 just two or three maybe. Haybe you can get three conveniently. 12 DB. ST()HER: I understand the -problem. 13 THE /IITNESS! I could do a retrofit, given time and 14 resources to do it. 15 DR. ST()HER: I am used to those things in 'fisheries. 16 1'1e have to make estimates ourselves. Ne get a better feel for 17 the data if we can look at it in that context. 18 4B. MC C()LLU/8: ()ne thing you might consider and 19 I don't know what arrangement" the Applicant has for doing 20 this, there are -predictions in the Applicant's and the 21 Staff's reports of the need of demand. Those are not, as I 22 recall, too multi-dimensional. Tney are more straightforward 23 and nonstatistical type models. 24 I f you could do something that reflects something 25 that could be looked at and compared with, in that selection
'()Tcam7 1
' THE 1'(ITNESS' could certainly present curves 2 showing the time path of demand growth as a result of the 3 various runs of 'the model and compare them with time paths 4 for demand growth predicted by each company. That would 5 be thB. MC C()LLUWs I guess that's what I had in
.'7 minds 8 THE l'(IT11ESS: I certainly "could do that. I would 9 be hapoy to do that, which wouldn't take much time at all.
10 12 13
,17 18 19 20 22 23 24 25
~ F jon1~P23 557' CHAIRNAN CLARK- I have a question or two. THE NITNESS: Could I respond to one other 3 - poin t? CHAIR/fAN CLARK: Qo ahead. THE l(ITNESS: I want to say that regardless 6 of how good or bad an ecnometric model is, it is always 7 a pyramided judgment. Judgment has to be exercsed all 8 the way along the line in developing an econometric 9 model. It is not, a purely objective, totally rational, 10 inviolable kind of structure..It is not. II From the beginning you have to judge whether 12 or not a particular body of data is suitable to be used 13 at all. Secondly, you have to judge what variables 15 are worth what variables to put in your model, hat 16 variables is it'orth to spend more'time to get; 17 information to put in your model. 18 You can~t put every variable in. You have to 19 figure out which ones are the best ones to put in and 20 the ones best left to the error term as we call it. 21 You have to make a judgment about the 22 mathematical form of the causal relationship you want 23 to find in the data. 24 You have to make judgments about all the 25 future values of all the factors that go into the judgments.
0 jon2 You have to make judgments all along the line. The only thing that is objective about it, it, is a process that is traceable. You can see how you get from A to B. You can see how you get from your assumptions to your forecast, step by step. That is the primary advantage. You see where things come from and how you get there but it is judgmental all along the way. CHAIRMAN CLARKE h'hat you just have been saying is along the lines of the questions. I wanted to ask you. 10 My learned colleagues would like enough informa-tion so that they can satisfy themselves that your approach 12 is at least sound. 13 I am too simple minded to do a thing like that, so I am taking a different approach. I assume in making your judgments and setting up this a thod of computation you used your best e ffo'rts to arrive at a logical projection
,17 whi ch may or may not he a ccura te.
18 Is that correct? THE NITNESS- Yes, sir. 20 'HAIRMAN CLARfi ! I believe you testified that your 21 projections roughly coincide with these of the Applicant. 22 THE ((ITNESS: Their projection.falls a little 23 bit below on the whole, falls a liittle bit below the midpoint of the range of projections 5 obtained. CHAIR'hAH CLARK: Did the Staff also, make
. ~jon3
'HE 559 pro jections?
]
~ 2 1
l'(ITNESS > I am 'sorry. You are asking me? CHAI RWAN CLARK: Ye s. Do you know? You d idn~ t 4 look at them i f they did. 5 THE NITNESS- No. I did not look at them. CHAIRMAN CLARKE That is enough for that.
'Now, I ask you this: if you were asked by the 8 management of one of these power companies for your advice 9 as to planning the future generating capacity on the basis of 10 some type of projection, would you say that you couldn' 11 advise them or would you say that you would advise them to 12 use these projections as the best available, or would you 13 give them some other way?
THE I'(ITNESS- I would advise first that they 15 use every bit of information that is available, inclu'ding 16 whatever information can be obtained from the projection. 17 I think in making a forecast or making a 18 planning decision, you have to make still at least one 19 more judgment beyond all the judgments that go to an 20 econometric forecasting model. You have to judge I 21 'you have to weigh that piece 22 of evidence against other piece., of evidence that don' 23 come directly from the forecasting model, from things that 24 you know, from your experience in running a company, from 25 things that you know from reading the newspaper, from
jon4 560 1 i f generally studying current affairs, from the whole variety
~
2 of information sources available to you. I think you have to view what the econometric 4 model can tell you as one part of the total input of information to the forecasting process.
.I think it is useful to use an econometric h 7 model to establish a possible range of outcomes, a high and a low, and get some feel for the possible to get a better subjective feel for the probability distribution 10 of outcome.
Then, once you have got that, you can make 12 perhaps a better planning decision because really a planning 13 decision on capacity investment is a decision between the risks you are trying to minimize the expected cost of 15 overbuilding and the expected cost of underbuilding ~ 16 That depends upon 'tt probability of distribution of load 17 growth. It is a fairly complicated kind of problem, 19 really, but I think you can make a better judgment i f you 20 have some feel for the probability distribution of growth 21 rates. CHAIN]AN CLARK: /fell, what class of citizens 23 or what group of people would in your judgment be the ones who would be most qualified to make the final decision on 25 growth of a system's generating capacity?
8 8 III
jon5 561
-'1 1 THE <<'ITNESS! My expertise is really not
X' h
~ ~ 'I 2 in the area of organizational management. I am not sure P
3 I could give the best managerial advice. CHAIRMAN CLARK: 1'(hat I am really asking you, 5 are you better capable of doing it, or is the is the 6 management of the power company better capable? . V 7 Between you .two, who would have the better 8 capacity? 9 THE WITNESS: ()h, it is the management of the 10 power company that has got to make the decision. There is 11 no question about it. At least between me and them. 12 CHAIRMAN CLARK t ~ I understand he does have to 13 make it. 14 THE NITNESS: And should'HAIRMAN CLARK: But I am asking whether 'you 16 think he is better qual'ified if you take these other 17 factors into consideration based on his experience as a 18 general thing. 19 THE NITNESS: I am an expert in one particular 20 area. I can provide some'etail and I hooe valuable informa; 21 tion in that particular area, but he has got to integrate that 22 with many other considerations. 23 CflAIRMAN CLARKE Some of them of a business 24 nature. THE NITNESS> Yes,'ome of them of a business
S jon6 562 nature, manyI of them of a business nature, that go beyond my competence and knowledge. He has to make the decision. I can perhaps e 23 4 help with some additional information but that is it. CHAIRMAN CLARK: Then f inally the 'decision 6 making has to be for those people that have the responsibility 7 for running the company. Tl.lE NITl<ESS: I think so. CHAI RWAN CLARK: And if we can test by the work 10 that you have done and by other means to see'that they have used, as far as this, a suitable basis for their judgment, 12 pelhaps that is about as far as we can ge. 13 THL= WITNESS: I think that's right. CHAI RMAN CLARK s Thank you very much. You may 15 be excused. 16 HR. N()RTOIl: Thank you, Mr. Anderson. 17 l'1itness excused. ) CHAI Ri~1AN CLARK - '/]e are quitting at 4:00 o'lock ~ 19 MR. N()RT()N: Mr. Bud York, please. 20 21 22
FP24 av I 563 1 Nhereupon, I
'2 R()LLAND Y()RK 3 was called as a witness on behalf of the Applicant, and 4 having first been duly sworn, was examined and testified as 5 fo llows-CHAIRMAN CLARK ! You may proceed.
DIRECT EXAMINATI()N BY MR. N()BT()N ! Mr. York, would you. please state your full name 10 and occupation? A Roll and, B-o-l-l-a-n-d, E. York. Senior vice 12 president, El Paso Electric Company, in charge of operations. 13 Q Mr. York, we have given you a copy of your 14 biography, supplied a sufficient number of copies to the 15 reporter. Is that a true and correct copy of your bi.'ography? 16 A Yes.
.17 WB. N()RTONc If there are a.o objections, we would 18 ask this be placed in the record as tihough read.
19 MR. LEl'iIS: No objection. 20 CHAIRMAN CLARK: Very well 21 It will be placed in the record as though reac{. 22 (Biography of Bolland E. Yasrk follows.) 23 24
8 'av2 564 BY MR. N()RT()Nt 1 9 I have also given you a copy of the Board's question 3 of January 9, 1976, number 4, which asks each of the 4 utilities who is a participant in this proceeding as an 5 Applicant to list the names of their facilities, which are 6 baseload, intermediate loads and peak loads, according to 7 the fuel, that is, gas, oil or coal> in some cases, cost oer 8 , kilowatt hour. 9 The Board has been furnished with copies of that and 10 the court reporter has, I believe, 30 copies, and we would ask, 11 . Mr. Chairman, that this be marked as Applicant's Exhibit 8. 12 Excuse me. l'1e ask number 4, the response thereto, 13 be marked as Applicant's Exhibit 8. 14 CHAIRMAN CLARK: It will be so marked. (The document referred to 'was 16 marked Applicant's Exhibit 1'lo. 8 17 fi for identi cation. ) 18 BY MR. N()RT()Ni 19 Mr. York, you have reviewed'his document. 20 To the best of your knowledge and inf'ormation, does 21 the information set forth there generally reflect the 22 response to the question requested hy the Board? 23 A I can~t speak for the other participants in this 24 application, but I can certainly speak for El Paso Electric, 25 and it is representative.
~~a A R:IciISSRRy 17' J IO BXOGRAPHY R, E, Yori 17TF. A"D April 21 '927 IIL7ICE OF B'frjlTH Anderson, Indf,ana POSXTYON. Senior Vice President MARITAL Harried (Ei Paso Divf.sk.on) STITUS: M ifc r 1<aoxi SUSIECSS Ei Paso Electric Corn<<ay 'RD.uSR ADDRESS: Past Office Box 9&2 OF C!!TIDAL~it furs Chrf.stine Hensen EK Paso, Icxss 79999 Stephen Yorh Diana HO~~ 6513 Soutbvind AJ!DIKES c Kl Paso, Iexas 79912 YorJ'ul!CATIOR!
Pucduc - SS CE - Juuuucy 27 ~ ASSI PB'L~~!RT MCOM1 DATGSx POSITION CWPAl!I IOSTIOR Feb,2 1974 Senfor Vice President El Paso Elcctrf.c Co. El Peso, Tees to present El Paso Dkv. 1971 1974 VP Paver Supply El Paso Eiectrj.c Co, El Paso, Texas 196S 1971 Supt+ of Produrtfon El Paso Eicetrfc Co, El Paso'exas Supt. ~ of Canstruction StOnc 6 iRebSter Engineering Corp u Vest Indies 1966 1966 Supt. of Constrvctf.on Stone 6 llebster EL Paso~ Icxas Engineering Corp, 1963 - 1964 ResM<<nt Engineer Stone 6 webster Haunt Star2., Engineering Corp, Most Virgfnie 1961 %963 ResMent Engfneer Stone 6 lR<<bster El Paso, 'Xexas Envinecrlnc; Corp, 1959 - 1961 Resident Engineer Stone. 6 webster Rockford, Engineering Corp, Illinois N58 - 1959 Cbfci Field Engineer Stone 5 4<<bster Parr, Engfneerf ng Corp South Carolina 7 1952 1958 c
'tone Chief Office Engineer 6 1R<<bster Enginccrf.ng Corp D,...
Port hecbes,
~ Ter~ ~
1956 << 1957 Chief Field Engineer Stone 6 webster S~~~.ny~ 'Xcxas ..
~engineering Corp I
1951 '- 1.956 ., Field Kngfneer ~ .'tone 6 lfebster Vorious Locations II Engineering Corp
8 Cl
ORCANZZATfOASc (aerherships in civic, business; relf pious end service organizations~ etc. end offices liel.d) Pest President - northeast Rotory Club t l~r 8evAcr-Vestern llH.3.s lfethndist Church; Chuircan of Aduinistrativc Board V~aher - EL ~xEdc Shrine Divan; C'xeetcrs Unit; Jesters, Worked on United Fund nnd 'P!CA Drives Y~~~er
- Texas Vnter Pollution
- VFST Co~itcce Control Association H~~er.- Operations Committee-western Systems Coordinating CauncE1 - Council V~~er thvber <<Coordination Arrnittae - Pour Corners timber K'L Paso Country Club; Xnteraational. Club 1&aber - MYRIO - Board?amber; V P Educatf.an 3~~r <<KX Peso Energy Ccnmisrion Nether - Rocky 3fountein Flectrical L'eeyre I'~er - Mministrative Comdttce - Arieone flucleer Paver Pro)cct h Jl-7 pt
i av2 '64 1 . BY MR. NORT()N: 2 0 I have also given you a copy of the Board~s question 3 of January 9, 1976, number 4, which asks each of the 4 utilities who is a participant in this proceeding as an I ' 5 Applicant to list the names of their facilities, which are 6 baseload, intermediate loads and peak loads, according to 7 the fuel, that is, gas, oil or coal in some cases, cost per 8 . kilowatt hour. The Board has been furnished with copies of that and 10 the court reporter has, I believe, 30 copies, and we would ask, 11 Mr. Chairman, that this be mar!~ed as Applicant's Exhibit 8. 12 Excuse me. Ne ask number 4, the response thereto, 13 be marked as Applicant's Exhibit 8. 14 CHAIRMAN CLARK: It will be so marked. (The document, referred to 'wa5
'Y marked Applicant~s Exhibit I'lo. 8 17 for identification.)
18 MR. N()RT()N: 19 Mr. York, you have reviewed'his document,. 20 To the best of your knowledge and information, does the information set forth there generally reflect the 22 response to the question requested by the Board? 23 A I can~t speak for the other participant's in this 24 application, but I can certainly speak for El Paso Electric, 25 and it is representative.
~
~ A ~
4, QUESTIONS'rovide a table of the total energy generated during ~ 1974 by each participant company for each of the base loads, intermediate loads, and peaking units according to fuel u . I, i.e., gas, oil, or coal. If cost per YNII is readily available, this should be included for each. ANSNER: A. Ari zona Pu.'>lie 'Service Company APS EIJERGY GENEP-,,ED DURI:.'.974 Units AVG. Cost 13ase Coal Oil Gas m/ 0'lit Four Corners 1, 2, 3 3,270,614 56,562 2. 3l 7 Four Corners 4, 5 1,258,311 7,341 1.781 Cholla 1 942, 691 1,416 3.720 navajo 1 442,754 3,370 2.022 5g914I 3703g37065g319 Intermediate Units Ocotillo 1, 2 446 I 310 457'43 ll 400
~
Saguaro 1, 2 442, 946 240, 886 14.253 Yucca 1 44,838 50,846 11.383 954,094 749;375 Peakinq Units Phoeni:: 4, 5, 6 94,~91 16,848 17.158 Ocotillo GT 1, 2 5,139 95,311 7,.777 Saguaro GT 1, 2 4,182 110,014 8.406 Phoenix GT 1, 2 3,008 46,852 7.990 Yuma GT 1, 2, 3, 4 51,594 29,514 18.542 Douglas GT 1 12,970 24.993 171,484 298,539
S Cl
ll. oalt I!iver Pro~ace Peaking Sources ~li Qro Gas Oil Coal osevclt 96,887,000 Iiormon Plat (conv.) 60~822I000 14ormon Plat (pmp. stor. ) 52,981,000 IIorse liesa (conv. ) 121,951,000 Iiorse Ilesa (pmo. stor. ) 123,147,000 Stewart Iit. (conv. ) 54,162,000 Crosscut EIydro (conv.) Crosscut: (fossil) 2,342,000 ICyrene j,'3 66I540,051 11,102,949 ICyr ne ', 4 63,708,148 1,767,852 ICyr<'..ie 65 24I922I460 8~7 6I540 ICyr<-..e >)6 30,036,552 11 I 222'48 Agua Pria i,'5 2 I 725 I 674 15,943,326 Agua Pria 66 804,780 '14,975,220 Peaking Subtotal 509I 950'00 191 I 079 I 665 63,768,335 Xntermediate Sources IIVdro Gas Oil Co a."~. Agua Fria J1 89,798,457 331,790,543 Agua Fria Pr2 118,579,086 335,746,914 A us Fria N3 355,415,140 579i887r860 rene Nl '39,084,373 51,598,627 rene '~2 55,635,688 184 ~ 173 ~312 Santan Nl 74,132,000 Santan N3 55,297,000 Xnt:ermediate Subtotal 658 I 512 I 744 1I 612 I 626 I 256 Base Load Sources Four Corners ><4 a 845,355,000 Nl N5'ohave 793I026IOOC Navajo Cl 691,175,000 Base Load Subtotal I329I556g000 Peaking Generation 764 g 798 I 000 IJ'HEI Xntermediat:e Generation 2 271 I 139I 000 Base Generation 2,329,556,000 Total SHP Generat:ion 5,365, 493,000
*Purchased and Xnterchanged 3,507,108,000 ICNII TOTAL 'RP GL'NI RATXON, PURCIIA"I";D AND Xi'TI'RCIIANGHD 8, 872, 601, 000 ICE'HI
- Xn addition, SRP received 3,507,108,000 I'I/II in purchased and interchange nergy in 1974. Some, but not all of this energy can be tied to type of enerating resource and fuel.
S C
C. Hl 1?nso Llacl:n.'.c Co oanv Hl Paso Hlectri,c Company Total Generation (Mi;:I) By Unit Load v e 1974 p ~ Base Load In'termedi.ate,Load Pea): Load 1 / 221 g 393 1,113,628 38,260 Oil 297,021 74,559 33,514 Coal 591,231 TOTi~~T 2, 109, 645 1,188,187 71,774 No cost dat;a (g/IJ ,) by unit type readily : vailable. D. Southern Californi.a Ldison Thermal Station Generation by Fuel Type For Base, Intermed'ate; and Pea): Loads 1974 Station Net Generation panel T oe Base Lo"d (1) Mohave Units 1 6 2 7 g 946 '90 g 000 Coal San Onofre 3 g 145~ 109 / 000 Nuclear Intermediate Load Alamitos Units 1 6 871 ~ 996 g 927 Gas Cool Nater Units 1 6 2 1 f 070 ~ 837 ~ 472 Gas Hl Segundo Units 1 4 Htisranda Units 1 4 lf162'50'71 940,939,432 Gas Gas Garden State 21,071,443 Gas EIighgrove Units 1 4 10,730,082 Gas IIuntington Beach Units 1,128,621,445 Gas Mandalay Units 1 6 2 449,559,133 Gas Ormond 13each Units 1 6 256,169,189 Gas Redondo Units 1 8 2g488~392~'332 Gas San Bernardino Units 1 455,684,247 Gas Alamitos Units 1 6 6g236g891~073 Oil Cool Ilater Units 1 6 2 32,041,528 Oil Hl Segundo Units 1 4 ~ 2,816,063,629 Oil Ltinanda Units 1 4 2*~629g446 568 Oil Garden St:ate 11,136,757 Oil Highgrove Units 1 - 4 10,319,918 Oil liuntington Beach Units 2,449,705,555 Oil Mandalay Units 1 6 2 1,177,961,867 Oil Ormond 13each Units' 3,935,130,811 Oil Redondo Unit:s 1 8 1,962,970,668 Oil San Bernardino Units 1 204,863,753 Oil
~'
i I 1 S I
Station Ilet Generation Fuel Tvne Peak Load (2) Alanitos Unit 7 lg676~417 Gas Hti~sanda Unit 5 2~ 676 / 465 Gas Huntington Beach Unit 5 2, 099, 840 Gas llandalay Unit 3 1,852,348 Gas Hllwood 64,524 Gas Alamitos Un' 7 3 / 905~ 583 Oil Hti~sanda Unit 5 3,645,535 Oil IIuntington:;each Unit 5 3f 139'60 Oil IIandal<". Unit 3 2,082,652 Oil Ellvood 2,266,696 Oil Includes total net generation f om station" rshero South'.rn California Edison is Project II<.!.age . E>:eludes generation from stations vshere Southern California Edison is a par-ticipant but not Project 14anager. (2) At stations srhere more than one fuel type is used, genera-tion is not metered by fuel type. hese f.'gures represent a ratio of the total gene:=ation from the tation based on the amount of each type of fuel used, taking into account the relative RATIONSefficiency of each fuel type. E. Public Service Company 'of New IIexico 1974 PIIH IIIisi GHHERiXTIOiV AND GH11E COST BY STATIOI1 Gas Base Load Units ID)II Mills/k!lh Reeves 1, 2, 3 1,'081,937 6.64
*Pour Corners 4,5
*San Juan 2 Intermediate Units Person 1,2,3,4 431,894 8.65 Peaking Units Prager 5 Santa Pe 23,405 22.01 Las Vegas 836 65.16 Oil Base Load Units IH'lII Hills/kIfh Reeves 1, 2, 3 59,201 11.56 Four Corner s 4 g 5 San Juan 2
i i I
Oil
) U'lJI Hi 1 1 s/):I Ih Internedi'ate Units Person lg2~3I4 il Prager Las Vega i
Pea):inc<
~
f,
~
',882 U.'ts Santa Fe 27,818 10,632
'rr!rr
'oal 17.47 31.82 23.08
!rills/l:!'rh Base Load Units Reeves 1, 2, 3
~Four Corners 4,
*San Juan 2 5 lr 098 ~ 870 3.75 605,841 5.65
'Xntszraerr'.:lte !!nite Per.son l, ',3,4 Pea):inq '.,inits Prager 6 Santa Fe Las Vegas TOTAL PLTPslT Base Load Units Hills/) I'lh Reeves 1, 2, 3 1,141,138 6.90
*Four Corners 4, 5 1,098,870 3.75
*San Fuan 2 605,841 5.65 Internediate Units Person 1, 2, 3, 4 459,712 9.18 Pea);ing Units Prager 6 Santa Fe 25,287 22. 74 Las Vegas 11,468 26.15 P. 'CPCO Unit, Type Fuel linet Energy Energy Cost (I 1);wlI) (Hills/) v.ll)
Base Oil, Gas 468. 3 12. 3 Pea)-ing Oil, Gas 18.9 24. 2
a ~' )1 0 I
Table 8.9. Average Price of Electricity in Cents Per kWh by Utility and Customer Class APS SRP SCE PSNH EPE Sma Large Coom. . All Cone. Coran. Year Resid. and Ind. Resid. Other Resid. Coen. Ind. Resid. Irrig. Cone. Ind. Resid. and Ind. and Ind. 1963 2.47 1.50 1.94 0.80 2.54 1 ~ 97 1.00 2.53 1.84 1.93 1.14 2.09 1.97 0.96 1964 2.39 1.47 1.88 1.00 2.36 1.80 .95 2.43 1.79 1.86 1.06 2.04 1.89 0.89 1965 2.34 1.44 1.83 1.15 2.22 1.69 .90 2.32 1.76 1.79 1.01 2.00 1.82 0.86 1966 2.22 1.40 1.73 1.09 2.10 1.59 .87 2.22 1.75 1.70 0.97 1.97 1. 73 0.83 1967 2.14 1.39 1.67 1.14 1.97 1.50 .84 2.31 1.68 1.62 0.90 1.88 1.88 0.81 1968 2.07 1.34 1.59 1.11 1.87 1.41 .80 2.01 1.68 1.55 0.88 1.80 1.56 0.76 CO I 1969 1.84 1:26 1.51 1.02 1.77 1.34 .76 1.89 1.54 1.46 0.84 1.71 1.44 0.71 1970 1.73 1.24 1;48 0.97 1.80 -1.32 .74 1.77 1 ~ 45 1.38 0.80 1.62 1.41 0.72 1971 1.64 1.18 1.38 0.85 1.75 1.30 .74 1.67 1.41 1.30 0.75 1.55 1.33 0.73 1972 1.65 1.18 1.46 1.01 1.85 1 ~ 39 .80 1. 72 1.54 '.35 0.76 1.49 1.27 0.77 1973 1.66 1.23 1.45 1.04 1.89 1.43 .91 1.69 1.08 1.35 0.76 1.44 1.21 0.73 All prices are divided by the GNP implicit price deflator, 1958 = 1.0. From G. A. Parsons, Ed., "Hoody's Public Utility Hanual," Hoody's Investor Service, Inc., N. Y. 1970 and 1974, and State of California (Docket tlo. 75-for-5) answers to Interrogatories (First Set) Propounded by State Energy Resources Conservation and Development Comnission to Southern California Edison Company, July 24, 1975.
'v 4 ~ S Qi
av3 0 And you also are a participant in many of the
,' r other units, or some of the other units, such as Four Corners 3 'nd so on, and these numbers are correct, to the best of your 4 knowledge; is that correct?
A Yes, they are. MR. N()RT()N- )1e would ask that the response to question no. 4, setting forth the megawatt hours, production, et cetera, for the various faciliti'es owned by the Applicants in this case as requested by the Board be admitted in evidence 10 if there are no objections. WR. I ERIS: No objection. CHAIRMAN CLARK: It will be r6ceived in evidence. 12 13 (The document referred to, 14 heretofore marked Applicant's Exhibit No. 8 for identification, was received in evidence.) 17 MR. N()RT()N: 'Ihank you. I understand Mr. Mc Collude has some questions he 19 would like to ask Mr. York. 20 MR. MC C()LLUM! Yes. My fx.rst question is where is 21 my copy? 22 MR. N()BT()N! 'I have one. 23, MR. MC C()LLUM ! Thank you,. 24 MB. N()BT()Ns Mr. Chairman., I might explain to the 25 people in the audience who are curiom: as to why we aren'
, ~ av4 566 giving a summary of what this response is, it is simply
~ ~ ~
a table of numbers, and there is no way to summarize it. CHAI BMAN CLAR1i < Thank you. MB. MC C()LLUM-'r. York, we really asked for this to sort of give us a base on which to clarify what procedures a power company, a typi ca 1 power company of the group that is here, would do in terms of the power that is generated in this nuclear generator that is shared between the five of you. 10 You have a certain baseload that you have expressed here, and it is made up of a total of about 2 million-plus 12 megawatt hours, and only about one fourth of that is coal. 13 The rest are gas and oil. 14 Now, as soon as you get your nuclear your part of the nuclear load, what will happen to that baseloa'd? Nhat 16 would you predict? 17 THE 1'/ITNESS: .Naturally, the nuclear would provide 18 the ma jor part of the baseload. However, we would have to supplement it with coal and this will replace some of our oil 20 and gas fired baseload generation we have today, because of 21 the depleting supply of gas and oil. 22 WB. MC C()LLUM: As I lool; at your figures here, you 23 have a relatively high baseload. 24 Am I correct in analyzing that compared to the 25 in termedia te?
567 THE WITNESS t Yes. htR. 11C C()LLUM~ I ntermediate peak. I 'presume just as an ordinary operating procedure you would expect to be using your full quota of the nuclear power plant, that is your part, all the time that is available? THE 1)ITNESS: That is correct. MR. MC C()LLUM: I am not sure I am asking the right person this, Wr. York, because this may be a legal question. Plould you, as one of the sharing persons, if someone 10 else did not use their nuclear power, be ab)e to purchase more than your share? 12 THE 1'IITNESS: At certain periods of the day, 13 yes, depending upon the load programmed each,day. It depends on the season. '(1e have seasonal load patterns, daily load patterns.. lie could utilize other participants'ortions 15 16 hours a day. .17 WR. MC C()LLUM: Would you say that at the end of 18 the year, would you come out in balance? I'<ould your share 19 the percentage of the plant that you own or would it be 20 possibly plus or minus? 21 THE 1'lITNESS: I am not sure I understand. 22 WR. MC C()LLU11! You said it varied with the time of 23 the day. 24 I wonder when you account for this at the end of 25 the year, whether you will have whatever percent as the total
av6 568 1 that you own as that amount of the power you received during 2 the year 7 THE 1'(ITNESS: I think as the nuclear units operate, 4 we will take the maximum share that we can, that we are 5 en ti tied to. MR. MC C()LLUM s Right ~ THE WITNESS: And if other participants have 8 overcapacity at some periods and they did not coincide with our 9 low load period, then we would purchase that from then. 10, WR. MC C()LLUM: I take it then you are in a position 11 of "I want," because 12 I THE P(ITNESS -'his would offset the oil and gas-13 fired, the higher cost. 14 MR. MC C()LLUM: So you are anxious to get your 15 full share or any other you can get to cut through your gas and 16 oil baseload you have here.
.17 THE 1'iITNESS: ()n an economy purchase basis, yes.
18 MR. MC C()LLUM: Tell me what an economy purchase 19 base is. 20 THE WITNESS't is not firm power. It is 2] purchased from some other utility on a share of the cost, 22 or difference in the cost of generation, incremental. 23 MR MC C()LLUM~ I know you don't give cost data on 24 your particular group of uni ts submitted here. I gather it 25 was not a convenient form to have it in, but I 'take it you have
/I av7 569 1 a feel for vihat you think the nuclear power is going to cost 2 versus what the gas and oil costs in your baseload. I 3 \ THE HITNESS: That is correct. 4 MR. WC C()LLUM: Can you expound on that a little 5 bit P THE HITNESS! The reason we didn'0 answer this, 7 sir, we didn't know whether this should include'he capital 8 investment together with the ()8h1, including fuel, or 9 whe'ther just ()8M including fuel. 10 1'fe have readily available the incr'cmental cost, 11 excluding the capital portion of it. The coal cost that we 12 would prhdict to replace nuclear in 1982 would be in the 13 neighborhood of 27 mills, and oil costs would be in the 14 neighborhood of 33 mills versus what we anticipate the nuclear 15 to be, about 22 mills per kwh. 16 MR. MC C()LLUl~f'- In your baseload units. This does
.17 not include your intermediate and peak load equipment, does it.?
18 THE WITNESS: Ho. That is baseload. 19 MR. WC C()LLUM: I believe I wi 11 depart from this 20 and I think it is all right to go ahead and'isit on how the 21 operation will be expected to ooerate. 22 Nhat I am trying to what I thought I would do'3 with Mr. York here is just to ask what his relationship will 24 be in this operating unit. This is more an administrative 25 rather than related to this question 4 table.
ave 570 1 Could you describe how you will interact vith the 2 people that will be operating the nuclear generator ayatea, 3 .,the everyday interaction to get popover decisions, this kind of 4 . thing, and then the overall administrati on, how do you participate in those considerations? THE 1'iiITflESS: After this plant is through? MR. MC C()LLUM - Yes. THE WITNESS: '"'e have various committees which will be established, administrative committee, operating 10 .committee, and practice and procedures will be established throughout these .committees. 12 Then our system operators will communicate with the system operators of the Palo Verde plant, where they 14 schedule the energy as it is expected to be available, on a daily ba "is. 16 WR. MC COLLU'F- Are you invol ved personally in any 17 o f thes e comm i t t ees? 18 THE WITNESS-' am a member of the administrative committee. 20 WR. MC C()LLUM~ Tell me aiPiaat the administrative 21 committee's responsibilities are? THE HITllESS: Presently, Sie executive committee 23 is an executive officer of each of She participants, as per the participation agreement, who have periocfic meetings to e discuss the progress, to discuss the procurement: of certain
- ~av9 571-1 types of material and equipment t and generally get the feel 2 of where we are, where have been and what it looks like as 3 to the future, in accordance with our plans and schedules.
WR. MC C()LLUM: I believe that is all. 5 MR. NOWT()N: ~PIe might ask Mr. York along the lines 6 of the last series of questions you asked to describe. the 7 administrative commi ttee, that many of the same participants 8 are already involved in and have been for a number of 9 years in a large fossil fueled plant. 10 MR. MC C()LLUM: Four. Corners? MR. N()RT()N t Yes. And have had that same kind of 12 experience. 13 MR. WC 'You C()LLUM- feel like that worked out
.14 well, this administrative procedure that has been handled with 15 the participating companies in the Four Corners 'project?
16 THE l')ITNESS: Yes. 17 I think the procedures have been well established. 18 They have been followed pretty strictly by the operating agent, 19 , which happens to be Arizona Public Service. 20 They welcome and relish input from the other 21 participants on their experiences and, of course, various 22 committees establish audits, both engineering and accounting audits, and anytime they have an additional expenditure such 24 as they might on environmental, they must call the participants 25 in together and decide upon what method and .chedules,
8 Cl I
,avl0 572 et cetera, but it has worked very well. MB. MC C()LLUM: ()kay. 3 l think that completes my qu. stions. 10. 12 13 14 I6
,17 18 19 20 21 22 24
o I C
mm I fp25 573 1 CHAIRMAN CLARKE It is our job, among others, to 2 determine whether there is a need for power, and the questions 3 asked a little while ago went somewhat beyond that with the 4 previous witness as to who makes the judgment as to whether 5' company proceeds with new facilities. 6 I would like to ask you another question along that 7 line for my educat ion. Do you feel that your company has any duty towards 9 your service area to have additive generating capacity? 10 THE WITNESS: Very definitely. That is one of our 11 prime responsibilities, to provide economical, reliable 12 service. 13 CHAIRMAN CLARK: Then, if the best projection you 14 are able to obtain indicated a need for additional facilities, 15 would you feel that i f you could f ind your way to do 'it, you 16 should provide those facilities, or provide the pawer from.
,17 some other source?
19 THE FITNESS:, If there were other sources that we 19 felt were less costly, yes. 20 CHAIBMAH CLARK: But you would feel an obligation 21 to provide the power from either your own facilities, or 22 purchasing other facilities ? 23 THE 'lNITNESS: Yes. 24 tetany time we have investigated purchasing from 25 other facilities. Ne do this almost constantly.
4l I
574 CHAIRMAN CLARK: Thank you very much. P
=
You're excused. MR. H()RT()N: Thank you, Mr. York. 4 (1'1itness excused. ) MR. N()RT()Nt At this time we would like to call Mr. Robert Lynn. Nhereupon, R()BERT LYNN was called as a witness on behalf. of the Applicants, and 10 having been first duly sworn, was examined and testified as follows-12 DIRECT EXAM IHATI ()N 13 BY MR. N()RT()H i 14 Mr. Lynn, would you state your name and occupation, 15 please? My name is Robert Lynn.
.17 I am Vice President of the Dechtel Power Corporation 18 of San Francisco.
19 0 Mr. Lynn, 'I have placed in front of you and 20 give to the members of the Board, HRC, Staff and the c'ourt 21 reporter, a copy of your personal resume. '22 Is that a full and complete cooy of your resume? 23 Yes, it is. 24 MR. H()RT()N! Mr. Chairman, we would ask this be 25 'laced in the record as though read at this time, if there e
8 f"
t mm3 2 1 are no objections. MR. LEPllS t No ob )ection. CHAIRMAN CLARK- Very well, 'it will be done. (Personal resume of Robert Lynn follows.) 10 12 13 16 17 18 19 20 21 22 23
o Q o
~ I
~
PEfISONAL f1 f SUMC ff~ NAMC I'tol>ar< (13>>l>) C:. I,, nn AI vi) DATE ~ CLASSIFICATION ~ice President - Services ORGANIZATION a I OCATION Tl>>:ri>>al ower Crpaniiation 1
. ~ Sa ti I l inci s<.o 2/i" x 2N" GLOSSY ORIGINAL DECHTEL EMPl.OYMFNT DATE Ma r cli 6 1 R E EMPLDYMcNT DATE Is)
SPOUSE'S NAME Jean CHILDRCN'S AGFS 20 ind 22 PHOTO DATE 1973 MILITARYSERVICE g RANK Elect ronic Technician 2/C US iNav PROF ESSIQNAL I.ICCNSES AND SOC;IETIES register('.>gine< r, State of Cal ifovv.ia Meri>bcr and pasl President ol't)ic San 1" rancisco Section of American Association of Cost Eiiginccrs. EDUCATION AND PERSONAL DEVELOPViENT PRQGRAViS DEGREE, CERTIF ICATE, CTC. SCI-IDOL MAJOR IOR SUBJECT) BS California Institute of Civil Engineering, 1948 7 c c )1 no log p QTIIEft, SIQNIF ICANT INFOIIMR'I'ION tftcfcr to instiiirlioiis )iL'foie co>>ililetiniJ) I)>>sinesi I)evelnliiiii:>>l Co>>ii>>itl>>c - 19i>8 DAG,I nl) S'it is I';i elioii an'd Motivation I')is c>>s sion roup - 1968 G DAG Cniiiliiiter '1 vcn()s Paitel Discussion .)969
, DAG I'lic Ta)eiit Iluiil Discussior Group - 1969 I:,xcc>>tiv< I'laii )'m)! va>>i - 1969 - 1970 - 197)
~
Coovdinatinlt Cu>>>>viitt>>c - 1970 - l 971 DAG - "Orf;aniration aiid Proccdurcs to Cope witli Growth" Discussion Groiip - 1973
Cl 0 4
, l(o.r)c rt L i J.iyilll
'OTIC I< ~IGNIF ICRNT INFOIIMRTION IC()rrl'rrrrre(I) lusc surr'LeMert TAL rhoc, IF nsouineol WORK I-IISTORY OATFS CQMiPANY, DIVISION OR POSITION HCLO, SUMiMARY QF
.DEPARTMF NT; R ESPONSIDI L I Tl C'S ANO FROM LOCATION ANO SUP I: R IOR SIGI IIF ICANT 'ACCOMPLISHMCNTS 1948'95'1 County Flood Control Civil Engineer r<<sponsible for assistatxe Disl riel in the design of hy(l'raulic struclurcs and Los A ngcles, Gal il'. appurtcnanccs.
1951 1954 industrial Division Worked on civil and structural design Los Angeles 4 Riverbanl~ artd sl,ructural cost estimates. Then Cali(or nia bccanic ficl(1 (ngincer on Lwo cartridge Li JI. II viii, I<. Ixoons, case plants and a lube oil packaging John Decich, 1. Paul II('. p'in nl.. It. Sl.c ven s 1954 1969 I.'.st,imill ) IIIi I)epir I'LIIIcrit l>r<)gr<<ssc(l frnr>> 1Irtir))atni Ln Senior i)all 1irrlcl ':cn, Gill rf, lw s l I I I I ir L (I I' S Il I) c I' i s 0 I' il n (1 Mi I n ir g e I' f
- 1. W. 13<<irn, I(. Ci. 1'"sti>n'ItinI,. A. M snag<<r of 1" sli>>)ating Wolfe arr(1 J. P. Yalcs hc]d Lop r(ispn>>sil)ility l'or 13cchtcl 1'till)atirrg III Sir ll 1atlcrsco al'Id D I ilnch 0 l'l'ic c s.
1969 l 972 'lirn:Ig<<>>l<<lit lnlo I'III(Ilroll Milllage r of Milli(II',<<I')Iclit Infort'Ilattotl S'III 1 rarrcisen, Calif. J. lV. 1<or>>c s 1972 10/72 Power 4 1>>(lustrtal. Manager of Divisior) Services San Fr;In<<isc(I, Calif.-
- 11. O. Rci>>selt 10/72 1/1/7S Popover. 6 ln(lustrial Div. Vice Prcsi(1<<nt II( Mgr. of Division Scrvic s San Francisco, Calif.
IrrsF sl/I'rI I h<FrrThl, rh( f, rr nrnrrrrrrnI
~
~
0 0 I
~ ~
OTIIEA SIGNIFICANT INFOIIMATION(CoIIIinUI Ii)
~
~
IUEC sUPPLEMENTAL PAGE, IF flEOUIAEOI
'I WORI< HISTORY COMPANY DIVISION QR POSITION HELD,
SUMMARY
OF DATES
'DEPARTMENT: R ESPONSI0 ILITIES AND FROM TO I.QCATION AND SUPERIOR SIGNIF ICANT ACCOMPLISHMENTS l./1/73 Pres. l3echtel Popover Corp, VP - Services The rrI'Ial Pow'er 0rg.
It. O. Reinsch IUSE SIIPrl <:.MEWrAL I AGE, IF AF.OUIIIEO)
~ 41 II Vw I
576 1 BY MB. N()BTON-2 Q Mr. Lynn, I have just given you a copy if the 3 Board s questions of January 9th, 1976, No. 2(1), which is as 4 follows: Give the bases .for choice of the escalation rates 6 for capital cost used by the Applicant in computing the cost 7 for the Palo Verde Nuclear Generating Station. 8 Mr. Chairman, we would f irst ask. that this be marked 9 as Applicants'xhibit I think we are at number' , No. 9. IO CHAIRMAN CLARK: It will be so marked. (The document referred to was 12 marked Applicant's Exhibit 13 No. 9 for identification.) 14 HY MB. NOBT()N: 15 Q Hr. Lynn, did you prepare or have prepared under 16 your direction, the response to this question? 17 A Yes. 18 0 And do you have any I know you have one, because 19 you pointed it out to me a couple of weeks ago any 20 corrections in response to that question? 21 Yes. 22 ()n page I, rather than showing individual percent-23 ages through the years <<75, '76 and <<77 and beyond, those 24 should have been bracketed as one number. 25 In other words, the bracket showing IO under 1975,
Cl 0
mm5, 577 1 and bracket under ~76 showing 8, and 7 under 1977 and ~
~ 2 beyond.
All right. And the idea there was, as opposed to for each of 5 the like nuclear steam supply system and so on instead of 6 using an individual escalation rate for each item, you 7 averaged them as 10, 8 and 7, and so forth? 8 A That is right. 9 Q Are there any other changes? 10 A I think on page 2, the SIL item up at the top 11 should have been SIC. BLF Index. 12 Q The initials SIC, as opposed to SIL? Is that 13 right? 14 Yes. 15 Now, Mr. Lynn, the Board has asked that you briefly 16 summarize your -response to this question so that the question 17 that was raised as a matter of fact, the resoonse that 18 this question goes to was raised by several people who made 19 limited appearances yesterday. We, therefore, would like you 20 to summarize your response for the benefit of those persons, 21 and the Board. 22 ()n page 1 we have noted the escalation rates that we 23 are predicting for the several years into the future. These 24 were developed by our organization and used in the development 25 of the estimates.
I 0 I
mm6. 578 1
~
In the preparation of developing these rates, we C 2 look at our past, we look at our present and we look at our 3 future.
,4 I would like to expand on that a little bit on 5 what is here, and go through it.
First of all, in the past we looked at all of our
~ 7 own construction operations. 1'1e .do have a great deal of 8 cost information on equipment, materials and on construction 9 labor in our historical data bank.
10 Secondy, we do have access and use a great deal of I 1 published information. ()n page 2 you have listed some samples, 12 but we do continuous monitoring of large-numbers of indices; 13 both the government indexes primarily in the industrial 14 commodities, and wage statistics. 15 4'e have an access to a large number of pri'vate 16 indi ces and so on. I am not sure of the total, bui
.17 it is a very large number we monitor continuously, and analyze.
18 As far as the present is concerned, we monitor 19 our present trends in escalation rates in several ways. 20 First we have a very large procurement department, 21 that I is in continuous contact witn large numbers of vendors 22 and suppli ers. He are constantly negotiating purchas orders, 23 contracts and in the course of this we do have access to the 24 attitudes and feelings of people in these companies. 1 25 Secondly, our senior management also has access
, ~ mm7eot rq79 1 through their contracts with the private sector. In that way 2 we also get their attitudes and expectations regarding 3 escalation rates. Also, I have noted that we have 60 active power
.e 25 projects involving some 100 generating uni ts. That data, 6 or the data from those projects is being fed continuously 7 into our cost analyst section and we continuously update it and 8 moni tor the trend.
Now, the future is another matter. 10 The future rates that we develop are based in 11 part upon the historical data and on the present trends that a 12 I have discussed above. 13 Ne operate, however, on the assumption that we 14 will have our economy will react and act in a stable way, 15 in a rational manner. 16 17 18 19 20 21 22 23
Q
. ~ 580 FFP T26 caml 1
~
Now, I
~
think
~
~
most of you are aware of what hap-2 pened in the 1973-74 area, so obviously our predictions are 3 not always precise. lie do have adverse conditions such as 4 . the oil embargo, price and wage control coming off in '73 and 5 ~
~74 and it did affect our escalation predictions considerably.
6 Also, for future escalation we take a look at our history 7 pretty carefully. l'le found in looking at our past history we can 9 identify certain bands of escalation through periods of time. 10 5 to 6 year periods over the past 25 years. I have noted the 11 band, we think I have identified some of them. I, 12 In the early 60s the escalation band was in the 13 2 to 3 percent rate. In the last 60s in the 5 to 7 percent 14 'nd it lookecf like we were going to stay in the 6 to 7 per-15 cent range until the problems of '73 and '74 did go where it 16 did go haywire on us. 17 ()ur assumptions for the future are that we will 18 go back to the 7 percent range of the early '70s ~ I think 19 almost all of us in the company feel that way and we
'I have 20 recently had good confirmation of that by our Federal Govern-21 ment, the Council of Economic Advisors for President Ford 22 'ndicated the same kind of a long range prediction. Also 23 based, in our minds, on having a relatively stable economic 24 condition.
In conclusion, we 'believe, based on what I
cam2 581 described, that our assumptions are reasonable and should be used. BY MR. N()RT()H: 9 Mr. Lynn, when was the when were the figures that appear on page 2 (I ) established? Nas that in early or P mid or late '74? 7 A They would have been prepared in late '73 or ear l y ~ 74. All right. So these numbers then, even though 10 we are looking at today in early 1976, were in effect those I I listed under 1975 or the number 10 percent was a prediction 12 of the future at the time it was made. Is that right? 13 A That>> s r ight. 14 . Q To steal Dr. McCollum's thunder, because I think 15 he is going to ask this question and ask you if you had a 16 chance to check that figure out. How that we are past 1975,
.17 how does it stack up, a year and half later?
18 A Nell, we just completed our analysis for 19/5 in 19 the last couple of weeks, and it's our'udgment, based on 20 the data that we have in hand in 1975, both in the labor 21 area, equipment and materials, that that number is more in the range of 38 percent. Very close to 8 .percent. 23 All right. And that~s the number we used. 25 Does that fact give you encouragement as to your
cam3 582 figures for 1976, ~77 and beyond? A Very definitely. MR ~ N()RT()N: That's all I had. CHAIRMAN CLARK: Mr. Lewis? 5, MR. LENIS: I have no questions. CHAIRMAN CLARK: Dr. Mc Collum? MR. MC C()LLUM: Mr. Lynn, do you remember o f fhand 8 what the average was for 1974? THE ilITNESS: For our power plant work it was in 10 the range of 15 plus percent. In this table you provided separate numbers for the various types, of equipment, material, 12 subcontracts, et cetera and you led me to believe that you 13 knew what those were already but after the statement that 14 you just made, I wonder if maybe that was a prediction as 15 we 11.
"16 Let me back 'off and say why I am asking. i'1e read
.17 in Economic News and Newsweek, et cetera, that the cost of 18 nuclear power plants have gone up 30 percent this year, and 19 this upsets the public and it makes me wonder and makes the 20 rest, of us wonder and then we come in and we find a report like this where we are saying we are going to power plant and 22 the escalation was only a very small part. of that 30 percent 23 that was printed in the media.
24 Nhat do you suppose is causing that? Why do we 25 actually have such smaller figures than we see predicted in
0 S '
cam4 1 the popular press? 2 THE WITNESS: I have no idea where the 30 percent number, comes from and I don~t subscribe to it, but I think if you will examine the escalation figures that others in the construction industry are using, like the Engineering News Record, Chemical Fleck and so forth, you will find that their numbers are so close to the kind of numbers we are 8 using. I know we are in the ball park wi th all of our 10 . peers in the construction industry. I really can't answer ll the 30 percent number. 12 MR. MC C()LLUM: I f you have now veri fied 1975 as 13 an 8 p rcent approximately, 8 percent escalation rate, and 14 in 1972 it was 6 or 7 percent, I take it you conclude that 15 there was a rapid expansion caused by the effects you men-16 tioned during just 2 years, and that this is what .verified
;17 your stability.
18 THE 1'1ITNESSt That's right. There was a comoensa-19 tion in that '74 period for the previous figure. CHAIR%AN CLARK: Thank you very much. 21 MR. iV()RT()N: Excuse me, Mr. Chairman. I think I 22 have a minor oversight. I don-'t think I think I had the 23 response marked, but, not put in evidence. I would like to at this time. That's Applicant's Exh'bit 9. 25 CHAIRMAN CLARK-'ny objection?
0E()Team 5 MR. LE1"1IS ! No ob jection. CHAIRMAN CLARK: Applicant's Exhibit 9 is received 3 in evidence. (The document referred to, hereto-fore marked Applicant's Exhibit Number 9, for identification, was received in evidence.) MR. N()RT()Ni Mr.'hairman, our next witnes is
.9 going to take longer than 10 minutes.
10 CHAIRMAN CLARK: I was about to raise that question 11 wi th you. I t is a q uar ter to 4:00 "and 10 to 4: 00. I f you 12 feel it is not very practical to split thi testimony over trio 13 days, I believe it would be wise to adjourn at, this time. 14 1'1R. N()RT()N: 1-{e very much wanted to go home tnis 15 evening, but I am sure he can stay over until tomorrow. 16 CHAIRMAN CLARK: 'lhat being the case, we will 17 adjourn until 9 o'lock tomorrow morning at th'is same place. 18 (1'i'hereupon, at 3-50 p.m., the hearing was adjourned, 19 to reconvene at 9 p.m., Wednesday, February 25, 1976.) 20 21 23 24 25}}