ML092870215

Exhibit 3 - Affidavit of Joseph A. Jones and Dr. Nathan E. Bixler Concerning the State of New York'S Motion for Partial Summary Disposition of NYS Contention 16/16A

ML092870215
Person / Time
Site:	Indian Point
Issue date:	10/12/2009
From:	Bixler N, Jackie Jones Sandia
To:	Atomic Safety and Licensing Board Panel
SECY RAS
Shared Package
ML092870100	List: ML092870115 ML092870154 ML092870206 ML092870215 ML092870265 ML092870274 ML092870289 ML092870335 ML092870361 ML092870433 ML092870444 ML092870586 ML092870592 ML092870631 ML092870640 ML092870644 ML092870672 ML092870677 ML092870678 ML092870679 ML092870685
References
50-247-LR, 50-286-LR
Download: ML092870215 (25)

Text

Indian Point Nuclear Generating Units 2 and 3 Docket Nos. 50-247/ 50-286-LR NRC Staff's Response in Opposition to State of New York's Motion for Partial Summary Disposition of NYS Contention 16116A Exhibit 3

October 12, 2009 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of )

)

ENTERGY NUCLEAR OPERAI-IONS, INC. ) Docket Nos. 50-247150-286-LR 1

1 (Indian Point Nuclear Generating )

Units 2 and 3) )

AFFIDAVIT OF JOSEPH A. JONES AND DR. NATHAN E. BIXLER CONCERNING THE STATE OF NEW YORK'S MOTION FOR PARTIAL

SUMMARY

DISPOSITION OF NYS CONTENTION 16116A Joseph A. Jones (JAJ) and Nathan E. Bixler (NEB),' do hereby state as follows:

1. (JAJ) I am a Distinguished Member of the Technical Staff in Sandia National Laboratories' ("SNL") Radiological Consequence Management and Response Technologies Department. I have over 25 years experience in engineering and analysis, 20 years of which was at SNL. I have been primarily involved in emergency preparedness and consequence management activities as well as radioactive materials management and cleanup activities both nationally and internationally. I perform emergency plan reviews for the NRC Staff in support of new reactor license applications and perform evacuation time estimate reviews in support of new reactor license applications and Early Site Permits. I was the lead author on NUREGICR 6863, "A Development of Evacuation Time Estimate Studies for Nuclear Power Plants," published

' In this Affidavit, the identity of the affiant who supports each numbered paragraph is indicated by the notation of his initials in parentheses. Where both affiants support a numbered paragraph, no parenthetical notation of initials is provided.

in January 2005. 1 was also the lead au.thor on I\IUREG/CR 6953, Volume I "Review of NUREG 0654, Supplement 3, Criteria for Protective Action Recommendations for Severe Accidents," published in December 2007, and NUREGICR 6953, Volume II, "Review of NUREG 0654, Supplement 3, Criteria for Protective Action Recommendations for Severe Accidents, Focus Groups and Telephone Survey,"

published in October 2008. 1 serve as the SNL project manager on the NRC project, "Review of NUREG 0654, Supplement 3, Criteria for Protective Action Recommendations for Severe Accidents," which includes analysis of protective actions under varying evacuation time estimates. I am also the SNL emergency preparedness technical lead for the State of the Art Reactor Consequence Analysis ("SOARCA")

project. A statement of my professional qualifications is attached hereto as Exhibit

("Ex.") M.

2. (NEB) I am a Principal Member of the Technical Staff in SNL's Analysis and Modeling Department and have been at SNL for 27 years; my work during the past 17 years has involved work for the NRC. I have a Ph. D, in Chemical Engineering and have been primarily involved in computer modeling of fluid dynamics and nuclear accidents and consequences. I have led the development and application efforts on a variety of NRC codes, including VICTORIA, RADTRAD, MACCS2, MELMACCS, and SECPOP2000. I am the SNL project manager for development and application of the WinMACCS code suite, which the NRC Staff utilizes in performirrg consequence analysis for level-3 PRAs and other risk informed regulation activities. I am also currently working on consequence analyses for safety documentation of the Mars Science Laboratory Mission scheduled for 201 1. 1 also serve as the SNL consequence analysis technical lead for the SOARCA project. A statement of my professional qualifications is attached hereto as Ex. N.

3. In this declaration, we present our views wi'th respect to the issues addressed in the State of New York's ("NYS") Motion for Summary Disposition on Use of Straight Line Gaussian Air Dispersion Model for the Environmental Impact Analysis of Significant Radiological Accidents at Indian Point and NYS Contention 16116A, ("NYS' Motion"), the NYS Statement of Material Facts Not in Dispute ("NYS' Material Facts"),

and the Declaration of Bruce A. Egan, SC.D. ("Egan Declaration"), dated August 28,

4. NYS' Contention 16 states:

Entergy's Assertion, in its SAMA Analysis for [Indian Point],

that it "Conservatively" Estimated the Population Dose of Radiation in a Severe Accident, is Unsupported Because Entergy's Air Dispersion Model Will Not Accurately Predict the Geographic Dispersion of Radionuclides Released in a Severe Accident and Entergy's SAMA Will Not Present an Accurate Estimate of the Costs of Human Exposure.

5. We understand that the Atomic Safety and Licensing Board ("Board")

admitted NYS' Contention 16 on July 31, 2008. We also understand that the Board limited NYS' Contention 16 to three issues: (1) "whether the population projections used by Entergy are underestimated," (2) "whether the ATMOS module in MACCS2 is being used beyond its range of validity," and (3) "whether use of MACCS2 with the ATMOS module leads to non-conservative geographical distribution of radioactive dose within a fifty-mile radius of [Indian Point]." Entergy Nuclear Operations, Inc. (Indian Point, Units 2 and 3), LBP-08-13, 68 NRC 43, 112 (2008),

6. NYS' Contention 16A states:

The DSElS Improperly Accepted Entergy's Population Dose Estimates Of Radiation Released In A Severe Accident Despite The Licensing Board's Admission Of the State Of New York's Contention That The Air Dispersion Model Used By Entergy in its SAMA Analysis Will Not Accurately Predict the Geographic Dispersion of Released Radionuclides and Will Result in an Inaccurate Estimate of the Costs of Human Exposure.

7. We understand that the Board admitted and consolidated NYS' Contention 16A with NYS' Contention 16 on June 16, 2009.

8. (NEB) The MELCOR Accident Consequence Code System Version 2

("MACCS2 code") was developed to evaluate the potential impacts of severe accidents at nuclear power plants on the surrounding public. The MACCS2 code considers, among other things, phenomena related to atmospheric transport and deposition under time variant meteorology, short- and long-term mitigative actions, potential exposure pathways, deterministic and stochastic health effects, and economic costs. See NUREGICR-6613, Volume I, Code Manual for MACCS2: User's Guide ("MACCS2 User Guide") at iii (1998) (attached as Ex. K).

9. (NEB) The MACCS2 code consists of a number of modules that perform various functions including the modeling of atmospheric dispersion, exposure pathways, short- and long-term mitigation, economic costs, and deterministic and stochastic health effects. Id. at iii.

10. (NEB) The ATMOS module of the MACCS2 code performs all of the calculations for atmospheric transport, dispersion, and deposition as well as the radioactive decay that occurs prior to release and during transport through the atmosphere. Id. at 2-2.

11. (NEB) ATMOS is an integral module of the MACCS2 code and interacts with other modules of the code to predict costs and doses. Id.

12. Atmospheric transport and dispersion models similar to ATMOS attempt to predict and evaluate the concentration, deposition, and path of contaminants, utilizing mass balance principles. Atmospheric transport and dispersion models alone are not able to calculate radiological consequences or economic costs of an event, including severe accidents.

13. NYS' Motion, NYS' Facts, and the Egan Declaration reflect confusion between the modeling of atmospheric transport for emergency planning with modeling for severe accident mitigation alternatives ("SAMAs"), which use a statistical mean for each specific area within the modeled 50-mile radius.

14. (NEB) NYS and Dr. Egan attempt to compare the ATMOS module results for a single meteorological data trial to a single run under their alternative models. This is not a meaningful comparison. The issue of concern in a SAMA analysis is not the results of a single meteorological data trial but the results of numerous meteorological trials that provide the mean dispersion over the entire 50-mile radius. As will be discussed in more detail below, the ATMOS module has proven its acceptability for the purpose of conducting a SAMA analysis.

15. (NEB) The end result of conducting multiple meteorological trials is the calculation of a mean atmospheric transport, which describes the expected amount and timing of the contaminant release reaching any area within a 50-mile radius. This calculation then allows for the determination of the mean effect on dose and economic costs for each modeled event that may occur at some unknown time in the future under unknown weather conditions. It should be clear that a SAMA analysis is not meant to provide a prediction of the contamination for any specific weather event, but rather, provide a mean result for a type of event under the mean potential circumstances.

16. NYS' and Dr. Egan's concentration on single event analysis is inappropriate in light of the purpose and methodology for conducting SAMAs utilizing Probabilistic Risk Assessment ("PRA") techniques.

17. We are aware of no model other than the MACCS2 code that fully addresses (a) doses from cloudshine, groundshine, resuspension, and ingestion; (b) emergency and long-term phase response measures and their associated costs; and (c) economic values associated with farm and non-farm wealth in a manner consistent with

the requirements of SAMA analyses. These model components are integrated in the MACCS2 code along with the ATMOS module, and could not be replicated by other atmospheric transport models alone.

18. On the basis of our review of NYS' Motion, the documents attached thereto, and our knowledge of the MACCS2 processes, technical documentation regarding the AERMOD and CALPUFF models, and the modeling that is utilized in SAMA analyses, we conclude that NYS is incorrect in its challenge to the use of the Gaussian model in the ATMOS module of the MACCS2 code. Further, we believe that NYS' Statement of Material Facts Not in Dispute contains material assertions which cannot be supported and are factually incorrect. Based on our review of NYS' Statements of Material Facts Not in Dispute, we disagree, among other things, with the following NYS statements:

19. NYS' Material Fact No. 10 states that:

The DSEIS relies on the MACCS2 computer code output to calculate the economic cost of a hypothetical severe accident at Indian Point.

20. (WEB) NYS' assertion in Material Fact No. 10 incorrectly implies that the MACCS2 code is the sole basis for calculating the economic costs of a severe accident.

The MACCS2 code is used to calculate several, but not all, of the contributors to the economic cost of a severe accident. Specifically, population dose output values from the MACCS2 code are used to calculate Averted Public Exposure ("APE") costs, and offsite economic cost output values from the MACCS2 code are used to calculate Averted Offsite Property Damage Costs ("AOC"). NUREG-1437, "Generic Environmental Impact Statement for License Renewal of Nuclear Plants, Supp. 38, Regarding Indian Point Nuclear Generating Unit Nos. 2 and 3, ("DSEIS") at p. G-28 (2008). Other economic costs, however - including replacement power costs and onsite costs, among others -

are computed based on generic, regulatory guidance and would not be impacted by any change in the dispersion model.

21. NYS' Material Fact No. 12 states that:

Atmospheric dispersion modeling is the field of predicting the fate and consequences of releases of contaminants into the atmosphere.

22. NYS' Material Fact No. 12 is overly vague, in that the terms "fate" and "consequences" are undefined and meaningless with respect to atmospheric dispersion modelirlg. Atmospheric dispersion models do not calculate radiological dose or the economic consequences of a release. Although atmospheric dispersion models may estimate ambient air concentrations and deposition rates of particles at places of interest, they do not extend to calculating the mean radiological and economic consequences of a severe reactor accident. To obtain consequence results, a model needs to address doses from inhalation, cloudshine, groundshine, resuspension, and ingestion; intermediate- and long-term response measures and their associated costs; and economic values associated with loss of use and permanent loss of farm and non-farm properties. These factors are addressed in the MACCS2 code, but are beyond the scope of the alternative models proposed by NYS.

23. NYS' Material Fact No. 17 states that:

Where the purpose of the air dispersion model is to predict the actual exposure of individuals in the path of the pollutant plume in order to assign a monetary cost to the full extent of the potential health risk, and then to quantify in monetary terms the cost savings that can be achieved by mitigating that exposure, the air dispersion model must have a high degree of accuracy to avoid either understating or overstating the economic costs and benefits involved.

Dr. Egan repeats NYS' assertion. Egan Declaration at 25. NYS and Dr. Egan also assert that:

The need for accuracy in the predictive model is particularly important where the number of individuals who could be exposed to the pollutant, the level of such exposures and the duration of such exposures is greatly impacted by the actual path the pollutant plume follows once it is released from the source.

NYS' Material Fact No. 18; Egan's Declaration at 7 26.

24. Contrary to Dr. Egan's suggestion, it is well known that population dose and economic impacts resulting from a plume's actual path due to cloudshine and inhalation is small. In fact, the long-term phase of the accident accounts for the majority of the population dose and economic impact results. Approximately 80% of the population dose can occur during the long-term phase of the accident. See, e.g.

NUREGIBR-0184, Regulatory Analysis Technical Evaluation Handbook, Final Report

("Regulatory Analysis Handbook1')at 5.23 - 5.24, Table 5.3 (attached as Ex. J). The long-term phase of an accident follows the emergency phase, which is typically assigned a duration of 1 week for nuclear reactor accidents. MACCS2 User Guide, Ex. K, at 6-27.

All of the release and transport occur during the emergency phase. Dose and economic costs depend, to a large extent, on the mean amount of contamination deposited on the affected land. The specific path taken by a plume under any postulated meteorological condition has only a minor impact on this mean. The minute details of a particular plume's trajectory do not have a material impact on the overall SAMA analysis.

25. (NEB) NYS and Dr. Egan essentially assert that each meteorological run must be extremely accurate in order to produce valid results for a SAMA analysis. As previously discussed, the results of any single meteorological trial run have a negligible effect or impact on the overall SAMA analysis. NYS' and Dr. Egan's assertion that increased accuracy is required is wrong for many reasons.

26. First, the alternative models advanced by NYS and Dr. Egan have been shown in extensive testing by the Environmental Protection Agency ("EPA) to produce

results that are no more accurate than the ATMOS module for use in a SAMA analysis.

See EPA, AERMOD Implementation Guide, AERMOD Implementation Working Group

("AERMOD Guide") at 9 and14 (2008) (attached as Ex. 0 ) ; EPA, AERMOD: Latest Features and Evaluation Results ("EPA 2003"), EPA-454lR-03-003, at 30 (2003)

(attached as Ex. P); EPA, A Comparison of CALPUFF Modeling Results to Two Tracer Field Experiments ("CALPUFF Comparison"), EPA-454lR-98-009, at 26 (1998) (attached as Ex. Q). In this regard, calculation of a mean result over hundreds of weather trials, the ATMOS module generally performed as well as several more modern atmospheric transport models. NUREGICR-6853, Comparison of Average Transport and Dispersion Among a Gaussian, a Two-Dimensional, and a Three-Dimensional Model ("LLNL Study"), at 65-68 (2004) (attached as Ex. L). In the LLNL study, the ATMOS module produced results within about a factor of two of the expected values. Id.

27. (NEB) Second, increasing the accuracy of a portion of the model related to atmospheric dispersion and transport will have little impact on the overall results of a SAMA analysis, and will not result in materially different conclusions. A SAMA analysis produces a mean result over a large number of weather events, each sending the contamination in a different direction. The model needs only to identify the likelihood and in what concentrations, the contamination reaches any portion of the affected area.

As such, no single weather trial needs to accurately predict a plume track, as long as the average from all the weather trials adequately approximates the likelihood of a plume reaching a location and its concentration at that location.

28. In this regard, the Gaussian plume model utilized in the ATMOS module of the MACCS2 code is actually more conservative in estimating doses at larger distances from the point of release than the models suggested by NYS. The Gaussian model ensures that any radioactive contamination travels the shortest distance to each affected area and arrives at each affected area with a more concentrated plume. As a

result, the model predicts larger doses and economic impacts, because the contamination has not had additional time to decay or to be diluted by dispersion. In contrast, the primary model asserted by NYS as an alternative, AERMOD, has been shown in extensive testing to underestimate the concentration of contaminants when terrain effects are taken into account and is not well suited for use in the light wind conditions which NYS and Dr. Egan assert is particularly important to the analysis.

AERMOD Guide, Ex. 0 , at 9 and 14. If AERMOD were to be used in Indian Point's SAMA analysis, the underestimated concentrations would be expected to result in lower economic costs and population doses, and thus a lower consequence for a severe accident. In addition, the MACCS2 code has been compared to a LaGrangian particle tracking code, for estimating concentrations and deposition out to distances as great as 100 miles from the point of release (NUREGICR-6853, and produced mean results that agreed to within about a factor of two of those of the LaGrangian code within 50 miles.

LLNL Study, Ex. L, at 65-68.

29. NYS' Material Fact No. 19 states that:

The need for accuracy in the predictive model is particularly important where the number of individuals who could be exposed to the pollutant, the level of such exposures and the duration of such exposure is greatly impacted by the actual path the pollutant plume follows once it is released.

Dr. Egan further states that "[t'he need for accuracy in the predictive model is also particularly important where the economic costs of mitigation measures and the economic benefits of mitigation measures are fairly close, such as within a factor of 2 of each other." Egan Declaration at 7 27.

30. (JAJ) Neither NYS nor Dr. Egan provide any support for the view that a predictive model must be extremely accurate for the purpose of a SAMA analysis. A SAMA analysis is complex, and contains numerous uncertainties and approximations; as

a result, the use of allegedly "more accurate" models would have little effect given the uncertainties in the ultimate result, and would not provide meaningful increase in the accuracy of the conclusions. Further, AERMOD and CALPUFF are not more accurate than the A-TMOS module utilized in the MACCS2 code, for the purpose of conducting a SAMA analysis. For example, in a study of CALPUFF, statistical measures for simulated plumes were within a factor of 2 of the observed plumes; further, the results, at distances not significantly greater than required for SAMA analyses, were consistently found to be underestimated for lateral dispersion and overestimated for the central maximum and crosswind concentration. CALPUFF Comparison, Ex. Q, at 26.

31. In the LLNL paper cited by Dr. Egan and NYS, the paper's authors concluded that the MACCS2 code produced results at specific radial and compass locations that were within a factor of 2 of the expected results. This is the same uncertainty that has been found for CALPUFF as discussed above, and the same uncertainty that the EPA has documented for AERMOD results during tracer testing at the Millstone Nuclear Power Plant. EPA 2003, Ex. P, at 30; CALPUFF Comparison, Ex.

Q, at 26. The substitution of AERMOD or CALPUFF, with no demonstrated improvement in accuracy, over the ATMOS module is extremely unlikely to result in any substantial change in the predicted results or conclusions of Indian Point's SAMA analysis, particularly given the uncertainties, which are necessarily present in other portions of the SAMA analysis.

32. NYS' Material Fact No. 22 states that:

ATMOS is a steady state straight-line Gaussian plume model which assumes that any emissions from the Indian Point Station are imbedded (sic) in an air mass having a single wind speed that flows for each period of the simulation in a single straight line direction. The atmospheric stability classification is also assumed to be constant over that time period. 'Thus each simulation will result in a prediction that the pollutants will theoretically travel in a straight line to infinity or to the limits of the

computational domain, regardless of topographical features that might render such a trajectory impossible.

Dr. Egan makes an identical assertion. Egan Declaration at 35.

33. (NEB) In NYS' Material Fact No. 22, the cited statement regarding wind speed and stability class is in error. NYS and Dr. Egan assert that atmospheric stability and wind speed are maintained as a constant throughout the entire time period. In fact, within ATMOS, once a plume is formed, the wind speed, stability class, and precipitation rate can change hourly. LLNL Study, Ex. L at 8. Furthermore, when topographical features intervene between a source and an observation point, one of the significant effects is that the plume becomes more disperse (dilute) than it would have been otherwise. This means that neglecting the topographical feature leads to overestimation of peak concentrations in the plume; therefore, there is a tendency for the ATMOS module to predict larger costs because greater concentrations of contamination reach the more populated areas at the outer edges of the 50-mile region than would be predicted in a more topographically sensitive model.

34. (NEB) In situations where the plume is forced to rise over obstacles, the Gaussian model conservatively estimates the plume path as the shorter distance across, rather than around, the obstacle. The shorter path of travel ensures that the maximum amount of contaminant reaches the downwind areas, which then receive more accumulated radiological dose and greater economic consequences. Allowing the plume to travel around obstacles increases the path length and the travel time to downwind areas, during which time the plume experiences increased dispersion, deposition, and decay, which tend to minimize the irnpact on downwind areas located farther from the site.

35. (NEB) The meteorological data that is an input into the MACCS2 code, accounts for any down stream flow from the Hudson River based on the direction of wind

at the sampling station. This phenomenon is properly accounted for in the year's data set for weather at Indian Point.

36. (NEB) NYS' Material Fact No. 22, incorrectly suggests that the ATMOS module fails to account for changes in the plume path that might result from topographical features. The SAMA analysis does not rely on a single meteorological trial or set of data, but instead develops a mean of possible conditions that might exist in the event of a severe accident. MACCS2 User Guide, Ex. K, at 5-34 to 5-39. The MACCS2 code samples the meteorological data from an entire year and ensures that the plume travels through 16 compass sectors so that all of the potential plume paths are accounted for in the SAMA calculations. Id. at 6-6 and 6-7. This ensures that likely impacts for the entire area within a 50-mile radius have an appropriate statistical model for the likelihood of a plume reaching an area, as well as the plume's expected concentration. By using this statistically reliable model for the entire area, the mean consequences of any release may be developed and evaluated for all potentially affected areas.

37. NYS' Material Fact No. 24 states that:

High terrain in the potential path of the plume introduces several complicating factors into dispersion analyses:

a. The presence of high terrain distorts and changes the directions of approaching winds as the flow cannot pass through the terrain.

b. The distortion of the flow direction materially changes the downwind destination of pollutant material emitted into the airflow and also, for elevated emissions, changes the proximity of contaminants to the ground surface increasing the ground level concentrations.

c. The presence of valley sidewall together with the radiational cooling will cause drainage flows that further distort air flow directions.

d. High terrain may degrade the reliability of a single meteorological station[s] of being representative of the transport wind speed and direction needed by the model, especially for longer distance transport calculations, because wind directions measured near the surface will vary with location. The effect is most pronounced during the lighter wind and stable atmospheric conditions that occur at night.

Dr. Egan repeats this assertion. Egan Declaration at 7 20.

38. (NEB) In NYS' Material Fact No. 24, NYS and Dr. Egan continue to compare a single weather trial to the MACCS2 code's version of that same single weather trial. As previously stated, the more appropriate comparison is to compare the mean of all the weather trials in ATMOS to the mean of all weather trials conducted using an alternative model. Once the correct comparison is made, the MACCS2 code predicts conservative centerline doses (i.e., higher dose) at higher population locations farther from the site than would be predicted if terrain features were considered.

39. If NYS' assertion in Material Fact No. 24b were correct, the result would be greater deposition nearer the Indian Point site and less deposition at distances farther away from the site. Significantly, the areas with higher population densities and more dense land utilization are located at distances farther away from the site of potential release. See Egan Declaration, Exhibit 3. Thus, the more rapid deposition of contamination at locations nearer the Indian Point site would produce less conservative results (e.g., lower population dose) than is predicted by ATMOS. Accordingly ATMOS should produce a more conservative (higher) dose estimate at downwind locations and increased economic impacts. In contrast, the approach favored by NYS and Dr. Egan would essentially minimize the predicted population doses and economic costs from a severe accident.

40. (JAJ) NYS' Material Fact No. 24d asserts that the effects of high terrain and a single meteorological station are most pronounced under "light winds." However,

AERMOD would be unlikely to produce more accurate results under light wind conditions. Light wind conditions are addressed in the AERMOD Implementation Guide, which states that the approach that AERMOD uses to address plume meander has not been implemented for area sources. AERMOD Guide, Ex. 0 , at 14. As a result, concentration predictions for area sources using AERMOD may be overestimated under very light wind conditions. Id. Further, the AERMOD Working Group recognizes the uncertainties that affect the model for light wind conditions. Id. Thus, AERMOD does not provide greater accuracy than ATMOS and would not be an improvement over the ATMOS module used in the MACCS2 code under NYS' asserted conditions.

41. NYS' Material Fact No. 27 states that:

For the Indian Point site, in the case of terrain features across the river, the flow will either pass along the side or rise over the feature depending upon atmospheric stability conditions. Thus, air pollution imbedded in the air flow will not take the straight line trajectory across the valley that would be predicted by the ATMOS using data from the Indian Point meteorological tower.

Dr. Egan adds to this assertion that "[ulnder more stable atmospheric conditions associated with greater ground level impacts, the plume is likely to be turned down the river valley as it cannot pass through the terrain." Egan Declaration at 7 39.

42. (NEB) Unlike NYS' assertion in Material Fact No. 27, atmospheric stability is not the only factor affecting whether a plume released as a result of a severe accident will pass along the side of or rise over a topographical feature. Other additional factors, such as the heat of release (which provides a measurement of the plume's buoyancy), will also significantly affect the final path of a particular plume. MACCS2 User Guide, Ex. K, at 5-20 to 5-23. In situations where the plume is forced to rise over obstacles, ATMOS estimates the plume path through the obstacle; as stated above, the shorter path of travel over the obstacle ensures that maximum amount of contaminant

reaches the areas downwind with higher accumulated radiological dose and economic consequences.

43. (NEB) As stated above, when topographical features intervene between a source and an observation point, one of the significant effects is that the plume becomes more disperse (diluted) than it would have been otherwise. This means that neglecting the topographical feature leads to overestimation of peak concentrations in the plume at distances farther away. As a result, the ATMOS module disperses greater amounts of contamination in the more populated areas downwind of the Indian Point site, which correspondingly produces greater costs.

44. NYS' Material Fact No. 30 states that:

Under overall light wind conditions, even though the Hudson is still tidal at the Indian Point location, the net average downstream movement of the river water and the effects of drainage induced airflows will favor movement of air above and near the river surface to be down river.

45. (JAJ) As stated above, the AERMOD Implementation Guide states that predictions for area sources may be overestimated by AERMOD under light wind conditions. AERMOD Guide, Ex. 0 , at 14. In addition, the influence of the water's movement on airflow in the valley would have only a minor influence on airborne contaminants above the river.

46. NYS' Material Fact No. 45 states that:

AERMOD was developed for applications within 50 Km (about 31 miles) of a source.

47. (JAJ) NYS' Material Fact No. 45, identifies the limits of AERMOD which was not designed to provide predictions for the purpose of a PRA-type SAMA analysis for an area with a 50-mile radius. See Egan Declaration at 7 29 (citing Perry, "AERMOD: A Dispersion Model for Industrial Source Applications, Part II: Model Performance Against 17 Field Study Databases" ("AERMOD Performance"), Journal of

Applied Meteorology, Vol. 44, at 707 (2005) (attached as Ex. R)). See also EPA 2003, Ex. P, at 1. The AERMOD model's nominal design use, which is less than 31 miles, extends to an area that is less than 50 percent of the area modeled for purposes of a SAMA analysis, which extends to 50 miles. The uncertainties that arise with extending AERMOD to distances beyond its design basis would introduce unknown uncertainties into the SAMA analysis that would jeopardize the SAMA analysis' validity and conservative assumptions.

48. NYS' Material Fact No. 47 states that:

The AERMOD model was subjected to extensive statistical model evaluations in a variety of terrain settings. These efforts showed that AERMOD represented a major improvement over the ISC3ST and other models.

Dr. Egan makes similar conclusions.

49. (JAJ) The statistical model evaluations, identified by NYS and Dr. Egan in NYS' Material Fact No. 47, reveal significant uncertainties and limitations of the AERMOD model. In the EPA publication entitled "AERMOD: Latest Features and Evaluation Results," an evaluation of the AERMOD model was conducted. EPA 2003, Ex. P, at 5. Testing with developmental databases was conducted to assure that the AERMOD predictions were consistent with the ISC-PRIME predictions for stack-receptor combinations dominated by building downwash. Id. 20. The databases used in the comparative modeling included the Bowline Point site located in the Hudson River valley downstream of the Indian Point site and also included sites at the Millstone, Duane Arnold, and Lee Nuclear Power Plants. Id. The evaluations determined that AERMOD produced results in error by about a factor of two. Id. at 30. In other words, the results of AERMOD introduced substantially the same uncertainty as NYS asserts is introduced by the Gaussian model. These evaluations also showed that AERMOD under-predicted release at two sites and over-predicted the release at another site, further demonstrating

uncertainties within AERMOD. EPA 2003, Ex. P, at 29 - 31. AERMOD's inconsistent over-prediction and under-prediction of concentrations would introduce potentially non-conservative assumptions into a SAMA analysis, without decreasing the uncertainty.

50. (JAJ) Other studies have found similar errors in the AERMOD model with actual conditions varying by a factor of 2 from the predicted conditions. See AERMOD Performance, Ex. R, at 707. The authors evaluated AERMOD model performance against 17 field study databases. Id. at 695. They concluded that the AERMOD model provided sometimes conflicting results and overall was only able to predict high concentrations within a factor of 2, in building wake studies. Id. at 707.

51. NYS' Material Fact No. 48 states that:

The CALPUFF model is appropriate for simulating transport and dispersion in wind fields that change with space and time. It is often coupled to CALMET, a model that computes the needed wind and dispersion fields from meteorological data.

Dr. Egan essentially agrees with NYS' assertion. Egan Declaration 7 30.

52. (JAJ) In NYS' Material Fact No. 48, the assertion that CALPUFF is more appropriate than ATMOS for use in a SAMA analysis is incorrect. In an EPA study of CALPUFF, statistical measures for simulated plumes were found to be within a factor of 2 of the observed plumes, and the results were consistently underestimated for lateral dispersion and overestimated for the central maximum and crosswind concentration.

CALPUFF Comparison, Ex. Q, at 26. Also, NYS and Dr. Egan ignore the intended use of the meteorological model in the current context, which is to produce useful and reliable SAMA analysis. See Palla Affidavit at 25 - 28. As such, the comparison is not relevant to a SAMA analysis. The Gaussian plume model as implemented by the ATMOS module of the MACCS2 code, when properly applied, provides a level of accuracy that is acceptable for a SAMA analysis, which necessarily includes uncertainties given the unknown nature and timing of the actual event, among other

factors (i.e., estimates of Core Damage Frequency, SAMA implementation costs, and SAMA benefits).

53. NYS' Material Fact No. 51 states that:

The NRC, in Part 2 of a 2009 Presentation to the National Radiological Emergency Planning Conference ("NRC 2009 Presentation"), concluded that straight-line Gaussian plume models cannot accurately predict dispersion in a complex terrain such as the Indian Point site and are therefore scientifically defective for that purpose.

54. (JAJ) Our review of the NRC 2009 Presentation indicates that the Indian Point site was not identified or analyzed in the presentation. LaVie Declaration at 11-

12. Further the presentation addressed emergency planning and response and did not address PRAs or SAMA analyses. The actual plume path is significantly more important for emergency planning and response when decisions must be made regarding potential protective action recommendations in the event of a radiological release, than it is for PRA or SAMA purposes, where analyses seek to make broad probabilistic assessments comparing the change in the predicted impact based on implementing certain mitigating actions at the plant.

55. (JAJ) The NRC 2009 Presentation is clearly directed at emergency planning, which emphasizes a high-fidelity simulation of a real-time weather during an event or hypothetical event. As discussed above, a SAMA analysis is not designed to model a single event under specific meteorological conditions at a single moment in time. Instead a SAMA analysis models all the events of concern over a year of weather data, to develop a mean outcome of the likelihood of potential impacts for the entire 50-mile radius area of interest. For these purposes, the Gaussian model, as implemented in the ATMOS module to the MACCS2 code, provides acceptably accurate and useful results that would not be improved by the substitution of the meteorological models advanced by NYS and Dr. Egan. While the MACCS2 code includes a more simplified

Gaussian plume treatment for atmospheric transport and dispersion than might be selected as an emergency planning ("EP") tool, it provides an acceptable and conservative assessment for use in predicting the costs and benefits of mitigative actions.

56. NYS' Material Fact No. 53 states that:

The NRC 2009 Presentation also acknowledges the "gravity sink" phenomenon that could cause the plume to travel down river towards New York City from a valley site such as Indian Point. As Slide 46 explains, the air is not heated directly by the sun but by heat convection from the earth. At night the earth cools and because higher elevations cool faster, cool air flows toward warmer air in the valley. This flow is described by the NRC as "gravity drainage," and in the absence of other meteorological influences (such as high winds), the drainage will tend to flow down river.

Dr. Egan repeats this assertion in paragraph 45 of his declaration. He also states:

A second effect of mountainous terrain on sources located in river valleys, such as Indian Point, is the creation of drainage flows by the presence of the valley side walls.

For example, at night when the earth's surface cools by radiating its heat upward, the air in contact with the surface cools. Because it is heavier than other air at than elevation, it flows under the forces of gravity down the valley slopes toward the base of the valley. In the absence of other influences, the pooling of the heavier air at the low point of the valley cross section causes that air to tend to flow downriver following the valley contours.

Egan's Declaration at r[ 40.

57. The assertions made in NYS' Material Fact No. 53 and by Dr. Egan do not invalidate the use of the ATMOS module in the MACCS2 code for SAMA purposes.

As stated above, the ATMOS module produces conservative results in comparison to a model that would force the plume to follow the contours of the river valley. The ATMOS module's more direct plume path results in a prediction of increased deposition of contamination at greater distances from the site and in areas of higher population densities resulting in greater economic consequences as compared to a model that

would force the plume to follow the river valley. Modeling the plume in the manner suggested by NYS would result in a longer path length and deceased effective velocity, and would increase the deposition of contaminants in the river valley, closer to the release site. This would effectively decrease the impact to the areas with higher population densities, which are located at greater distances from the Indian Point site.

58. NYS' Material Fact IVo. 57 states that:

The study did not compare the results for a discrete event such as a postulated severe accident.

59. (NEB) The assertion in NYS' Material Fact No. 37 is not correct. On page 39 of the LLNL report, the authors describe the source term used in the analyses, identifying the radionuclides used, the height of release, the buoyant energy, and additional source term characteristics. Such factors are fundamentally similar to the description of a reactor source term. LLNL Study, Ex. L, at 39.

60. Dr. Egan states that:

The simplicity of the ATMOS model's assumptions are scientifically unreliable for use in the terrain in which Indian Point is embedded and the model therefore cannot accurately predict the geographic dispersion and concentration of a radionuclide release from that site.

Egan Declaration at 7 37.

61. (IVEB) Again, Dr. Egan misapprehends the use of ATMOS within the SAMA analysis, and makes incorrect comparisons of models used for unrelated purposes. For purpose of a SAMA analysis, ATMOS provides an acceptable degree of accuracy and precision for predicting economic and radiological effects based on the average of multiple weather trials. MACCS2 User Guide, Ex. K, at 5-31 to 5-42. The end result of running numerous weather trials, as is done with the ATMOS module in the MACCS2 code, is that the mean of deposition and concentration determined, so that meaningful comparisons of the projected costs and benefits of mitigative measures can be conducted. Id. at 7-45 to 7-50. Dr. Egan asserts without any evidence that

inaccuracies in the ATMOS treatment of the plume path would have a material effect on the SAMA conclusions.

62. Dr. Egan states that:

From a meteorological air flow perspective, the presence of the river, nearby terrain features and non-homogenous ground surface features all affect the overall air flow patterns, which affect the rates of vertical and horizontal mixing of any pollutants released from the plant. Because there are nearby mountains that are higher than the meteorological tower at the Indian Point Plant, ATMOS, which can only use one meteorological source at a time, cannot predict the wind speed and direction accurately from the meteorological data measured by that tower because wind speeds and directions in the valley are unlikely to be representative of the larger scale flow pattern that carry contaminates from the plant to the surrounding areas. It is important that atmospheric dispersion modeling of the effluents from the plant consider these factors in order to provide a reliable basis for estimating ground level concentrations and corresponding estimates of potential exposures to the surrounding population.

Egan Declaration at 7 38.

63. (NEB) Dr. Egan's challenge to the use of a single meteorological data source point for purposes of a SAMA analysis is incorrect. The Gaussian plume model utilized in the A-TMOS module of the MACCS2 code develops consequences based on multiple weather trials using a full year of weather data and multiple sequences to determine a mean of possible events. This approach provides more representative outcomes than would be generated from a single run, and is appropriate for the probabilistic modeling utilized in SAMA analyses. MACCS2 User Guide, Ex. K, at 5-22 to 5-27 and 5-37 to 5-40.

64. Dr. Egan states that:

[Tlhere is a consensus in the scientific community of meteorologists that create and use air dispersion models, and government agencies that rely on them, that a simple straight-line Gaussian plume model, such as ATMOS, is scientifically unreliable when applied to the complex terrain in which the Indian Point power station is located and

cannot accurately predict the dispersion and concentrations of radionuclides in a 50 mile radius of the Station. Because of these deficiencies, and because of the wide variations in population density within the 50 mile radius, the DSEIS's (sic) SAMA analysis could have grossly underestimated the number of people who would be exposed in a severe accident and the concentration of the doses they would receive. This would, in turn, underestimate the "cost" of a severe accident and thus the "benefit" of a proposed mitigation measure that would reduce the magnitude of the initial release of radiation from the plant or reduce the probability of the release occurring, or both.

Egan Declaration at 1160.

65. (NEB) Dr. Egan's statement is erroneous. The ATMOS module is utilized in the MACCS2 code to provide probabilistic assessments of potential outcomes resulting in a mean value over a large number of weather patterns. To adequately predict dispersion and concentration for SAMA purposes, the calculation needs to adequately predict the likelihood that a plume reaches a specific destination. This does not require that any single weather trial be highly accurate, only that on the average, the calculation adequately approximates the likelihood that a plume reaches a location.

ATMOS is used to develop consequences from multiple weather trials that are then averaged to provide a more representative result than would be obtained using a single weather trial. The Gaussian plume model was chosen for the MACCS2 code because it allows large numbers of realizations to be calculated. 'These realizations necessarily represent uncertainty in the weather data that might exist at the time of a postulated hypothetical accident. Large numbers (e.g. hundreds) of realizations are generally used in performing PRA and sensitivity analyses. LLNL Study, Ex. L, at 5. The use of large number of such realizations minimizes the importance of the outcomes in any specific weather trial.

66. Finally, in performing a SAMA analysis, if the path that a plume takes to reach a destination is circuitous because the wind direction changes along the way, the

path length is longer. lhe plume becomes more disperse and there would be more deposition than if the plume follows a direct path. In this regard, ATMOS is conservative in that it omits such dispersion and deposition effects. It is therefore not likely that the use of Al'MOS would signifrcanl.ly underestimate the number of people who would be exposed to a radiological release in the event of a severe accident, at the Indian Point site, or the concentration of doses, which they would receive.

Sworn and subscribed to Betore me this &th day of October, 2009.

My Commission expires Sworn and subscribed lo Before me this 1 2 t h day of October, 2009.

My Commission expires