ML20080S645

Comparison of Beaver Valley Power Station Dose Projection Methodology Vs NRC Interactive Rapid Dose Assessment Model

ML20080S645
Person / Time
Site:	Beaver Valley
Issue date:	08/31/1983
From:	La Vie S DUQUESNE LIGHT CO.
To:
Shared Package
ML20080S640	List: ML20080S636 ML20080S645
References
NUDOCS 8310180434
Download: ML20080S645 (41)

Text

._ , . ..

i

.i l e O

i A Comparision of the Beaver Valley Power Station Dose Projection Methodology Versus NRC's Interactive Rapid Dose Assessment Model Prepared By: Stephea F. La Vie Duquesne Light Company Sr Health Physics Specialist Radiological n'gineering August 1983 ABSTRACT - A comparison has been perforcied between the US l Nuclear Regulatory Commission's Interactive Rapid Dose Assessment Method (IRDAM) and the dose projection methods in O place at the Beaver Valley Power Station and those planned for the future. This comparison identified nignificant differences between the results of the various IRDAM methods and the results obtained by the BVPS methods. The report concludes that the variances between results are due to differences between a site-specific method ircorporating specific plant and site parameter assumptions and a method intended to be generic for 81 power reactors.

'A@

Duquesne Li@t 8310180434 831003 PDR ADOCK 05000334 F PDR

~

. 1 8 4 A Comocrision of the Beaver Valley Power Station

/si Dose Proiection Methodology _

is_)

Versus NRC's Interactive Rapid Dose Assessment Model Prepared By: Stephen F. La Vie Duquesne Light Co,mpany Sr Health Physics Specialist Radiologicai Engineering

' August 1983 PURPOSE The purpsse of this analysis is to evaluate the Beaver Valley Power Station dose projection methodology in compar.4 son to the.

Interactive Rapid Dose Assessment Model (IRDAM) in use within the US Nuclear Regulatory Commission. This analysis was initiated in

("')'

N- response tc the NRC observations regarding an emergency exercise at the Davis-Besse site. In that exercise, the NRC reported dose projections, directly to the Governer of the State of Ohio, that were a factor of ten to twelve higher than those being reported by the licensee and/or the responsible state agencies.

Discrepancies of this nature have been reported following exercises at other power reactor sites.

The BVPS methodology has been coordinated with the approprioce agencies in the three states impacted by the ten mile emergency planning zone of the the Beaver Valley Power Station. The methods received extensive onsite review and were approved by the Station buperintendent upon the recommendation of the Onsite

.d 4

._y- - . - - - _ _ - - - - - , . . . , - _

~

2

, Safety Committee and the Radiation Safety Review Subcommittee 4

prior to being placed in use. Since the methods used by the offsite agencies are based on the BVPS methods, inconsistancy between the IRDAM code and the methods in place at BVPS and in the three states, could result in undesirable confusion durin8 an emergency. The objectives of this analysis is to, beforehand, assess any possible discrepancies, attempt to determine the reason for the discrepancy (s) and, if warranted, recommend appropriate corrective actions for the BVPS methods.

METHODOLOGY OVERVIEW

l. IRDAM The NRC contracted with the Pacific Northwest Laboratory for the development of a rapid assessment methodology that could be used by health physics per,onnel with little or no computer expertise. In May 1983, the NRC published

" Interactive Rapid Duse Assessment Model (IRDAM)"( ). The

( ) abstract provided on the NRC Form 335 (used to promulgate x.J the repert) provided that the code "...is intended to provide the user with a capability that is easily portable and user friendly so that it may be used at any location (HQ, Base Team, or Site Tean) while the staff is not fully augmented to scope the extent of the problem...." The abstract to volume 3 of the report explains "...In order to provide a rapid assessment capability consistent with the capability of the Osborne-1 computer, certain simplifying approximations and assumptions are made...."

The code contains minimal site specific data consisting of licensed power rating (MWe), reactor type, and containment design leakrate. The code also has built - in default information generic to all sites. As the code is largely generic to all licensed reactors, the results will differ

,s from site - specific assessment code results more likely l >

a 3

4

(~'N .'han t not. In the next section, these differences will be identified and quantified. -

Used for its intended purpose,'ie: a quick code for use

until. the staff experts are available, the code is appropriate. The code is not, however, a replccement for the site specific methods available at each reactor site.

2. BVPS Methodology The Beaver Valley Power Station dose projection methodology, currently in place, is described .in detail in the dose projection procedures EPP/IP - series 2.6.X, and in the supporting DLC calculation packages (2) . Several of the more significant procedures in this procedure series have been incorporated into an interactive TRS-80 micro-computer code (dubbed "RADOSE"). As the micro-computer code is ,

simply a mechanization of the hand calculational methods,

,-~

the code will not be discussed as a separate methodology.

~

The. BVPS methods are based on actual plant data and, in the absence of actual data, on postulated ana.1yses based on the actual plant configuration. Defaults, where tved, are also based- on the actual plant configuration. With few exceptions,~ the BVPS methods are site - specific. The underlying meteorological and dose assessn. eat methodology is based on NRC. guidance documents ( ~0) . In the inspection l report of the emergency plan site assessment review (conducted in September 1982), the NRC noted that the BVPS l dose assessment methods were technically adequate. The

( assessment team provided some items for improvement. These l items have been closed-out by the NRC as having been

satisfactorily resolved.

I (3

'N f

i l

. . - -, _ , - ,, ,.,,.m. .- _ , ~ , - . . . , . , _ , _ . - _ . . - , , - - - _,.m.- . . , . , - .

. 4

. v

/R. .

() 3. ARERAS/ MIDAS In response to NRC guidance in NUREG-0737(11) ,

~NUREG-0654(12) , and the then-draft Regulatory Guide 1.23 revision (subsequently retracted)(13) , Duquesne Light contracted with Canberra Industries for a computer - based dose assessment system with real time inputs. This system is being installed at this- time .and will be coapletely operational by Decemeber 1983. The ARERAS system utilizes the Pickard, Love and Garrick proprietary MIDAS software.

The software includes both class A aad class B models.

a. The class A model is based on the straight line gaussian model and is comparable to the straight line model in the BVPS procedures, and to that in IRDAM. In this class of model, the plume transport is assumed to continue in a straight line in the direction of the winds at the release point. The straight line model 7-s

\s_/ does not update the atmospheric transport for changes over time or for changes in the dispersion characteristics beyond.the immediate site environs. A class A model is conservative under most circumstances and is often over - conservative.

b. -The class B model is a particle-cell model using a modified potential flow windfield to assess transport beyond the site snvirons. The release is modeled as a series of discrete instantaneous releases. Each of these discrete releases is represented by a set of particles which are tracked in X, Y, and Z planes until

the model is stopped or the accumulation of data exceeds allowed storage space (about 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />). The dispersion of particles is updated periodically (typically 3 minute increments) with dose calculations

/~T updated every 15 minutes. The use of the modified

! \'.,)

potential flow windfield grid allows the model to

-- , w , --

. - , -wm , - ,+, y w y w w - - = .y- w-

, 5 y

hi

^

,-- explicitly address vertical wind shear and terrain 7

A/ s effects without: resort, to .the empirical corrections used in. the class A model. The model-iacorporates a digitized XYZ map of the site'and the environs out to i

ten miles. The class B model is far more I representative of. the deep valley meteorological regimes frequenting the-BVPS area.than is the class A model, i

The MIDAS software runs on a VAX/VMS11-750 mini-computer having 2 megabyte of core memory and 134 regabyte of on-line storage. There 'are two redur. dant computers with failover

-software to-assure availability. The' software incorporates i

4 numerous site - specific parameter data files to adapt the

~

MIDAS software. to the BVPS site. The data. in these

- parameter' files are based on actual plant configuration and site topography.- Effluent radiation monitors and

{

meteorological sensors are'iaput to the 'ARERAS system and

, - are available- to the MIDAS software on a r.eal-time basis; i Dial-up telecommnicaticas ports allow the NRC and authorized ~

offsite agencies to access meteorological, class A, and

- class B reports via a. remote terminal.

d

. METHODOLOGY FOR THIS REVIEW L This review will follow the general outline of Volume 2 of the t

IRDAM' manual (1).. For each major discussion point in IRDAM, a comparison will be made to the equivalent capability in the current BVPS methods ("BVPS") and the future ARERAS capability

("ARERAS"). Where' appropriate and/or feasible, a comparison calculation is provided. For each item, the discussion will be closed with an evaluation conclusion, p .; L

() .

• ""M"'"8'M*9*~#"""**T N t "

-ee-W_---- <wwg -t--Mid9w==p-9*=* R **-t'-4 e' E-"P+T"'P4*'T'*+48-t*T'*W1-'PM~F"7

• F'WTMF'WTW 'T9T'9NP""' 'N Nf- WW-#~"1'WF-fvvT"'T'7 7*V*"'"T'T"-f'"WT

. 6

, EVALUATION

i. ) 1. Release Options 1.1 Isotopic Release i

1.1.1 IRDAM

The code prompts tia: user to enter release data in terms of Ci/sec for each observed radionuclide. The dose model used is the semi-infinite cloud model -of Regulatory Guide

-1.109(5) . For the thyroid, IRDAM uses the RG 1.109 infant thyroid dose conversion factors and an infant breathing rate of 0.25 cubic meters per hour. It is noted that the radiciodine dose conversion factor listed on page 3 of Volae 2 are in error. These values are conversions for rem per curie inhaled, and not the factor indicated by the column heading. A review of the code modulc (IRDAMCAL.COM lines 1150, 7088) indicates that code properly uses this date along with the breathing rate to obtain the dose. commitment.

1.1.2 BVPS: The hand calculation procedure EPP/IP-2,6.6 provides p worksheets for the evaluation of the dose from a known isotopic V- release concentration, uCi/cc, or from a known isotopic quantity, C1. The release rate (Ci/sec) step is bypassed as the user uses the sample results (in uCi/cc or Ci) along with the release flow rate (in cfm) from plant flow instruments and X/Q to determine the dose. A single numerical conversion j factor for each nuclide incorporates the RG 1.109 semi-infinite model dose conversion factors and any necessary unit l

conversions. This bypassing of the intermediate Ci/sec result simplifies the calculation without sacrificing accuracy. The

! BVPS methods determine child thyroid dose rather than infant thyroid dose as the child thyroid dose is more restrictive.

l This choice is consistent with the methods in the offsite agencies. The child breathing rate is based on the standard assumption that the child inhales 50% of his daily intake of

[ air in an eight hour period, and is 0.63 cubic meters per hour.

Is

! v

7 j-~) 1.1.3 ARERAS: The . MIDAS software provides for the Optional entry of 5sl isotopic data from historical data files, from default files, or from manually entered data. Tn a sense, all of the MIDAS dose calculations are based on an isotopic mix. The dose model used in the class A model is based on an user choice of semi-infinite or finite cloud model. The semi-infinite model is directly based on the DR = 0.25(Ey-)(X) model of reference 7.

The finite model uses the correction factors developed in reference 7- and the algorithm of reference 14. The noble gas model does not include credit for 5 cm tissue attenuation as done in IRDAM and BVPS.- The thyroid dose is based on the child thyroid dose conversion factor of reference 5 and the annual average breathing rate of 0.42 cubic meters per hour.

J The dose model in the class B model is a complex algorithm L which calculates the plume, ground shine, and inhalution (WB, lur.g , thyroid) dosec. Further information on the assumptions inherent to this model is not available at this time.

(]N 1.1.4 ' COMPARISON: Volume 3 to the IRDAM manual (I) provides print-outs of IRDAM calculations for several different scenarios. There were no calculations performed for the ground level isotopic releases. Thus, no direct comparison of results

. is possible. Exhibit 1 provides a comparison of the dose conversion factors. Exhibit 2 provides the whole body doses calculated using the IRDAM, BVPS, and MIDAS dose conversion factors.

1.

1.5 CONCLUSION

The three methodologies provide comparable results for noble gases. The ARERAS results are likely to be higher since the MIDAS DCFs don't credit the 5 cm of attenuation by skin. For Xe-133, this equated to +20%. For the thyroid, the BVPS results will be slightly higher than the IRDAM results due to consideration of child versus infant. The ARERAS results A

+

L)

. ~

-e x --e---e -- ..%, -,---.m, m ,wr.,-m. ~,e----w . r- v v <r -=-4 -,&--- --ret- - - - . y- % ,,. = , - , ,v-

' 8

. r f-'s are 30 % lower than the IRDAM results due to the more realistic

\

'd breathing rate used.

1.2 Gross Release

1.2.1 -IRDAM

The code prompts for the input of gross Ci/sec of noble gas. The user can' identify an iodine to noble gas ratio and a filtration efficiency. The age of the material is used to determine the nuclide ratios. If the age is greater or equal to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, all of the noble gas is assumed to be Xe-133 and all of the iodine is assumed to be I-131. If the age is less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the dose conversion factors are based on a

"... ratio of doses received from the actual mix to those received from a single radionuclide (Xe-133 or I-131) of the same activity, as a function of time from shutdown (eg: " dose equivalent Xe-133, I-131)...." The dose before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is determined by first determining the dose beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and correcting with a first order linear expression.

(O The IRDAM calculations include corrections for decay and build-up in the containment. The. methodology for nuclides with an age less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, can not provide for site-specific radionuclide transport considerations (eg: containment sprays, in-containment filters, etc).

1.2.2 -BVPS

The BVPS methodology does not provide a method using gross Ci/sec release rates as an input, in as much as this is an intermediate result. The input data to the BVP3~ gross release model are radiation monitor reading (cpm), release path flow rate (cfm), and, accident scenario category (eg: LOCA,

fuel handling accident, etc). The accident cate8ories are based on the postulated radiological consequence analyses in the BVPS Final Safety Analysis Report, with some revision to moderate the overconservative nature of FSAR analyses. These a.ealyses, originally performed in accordance with NRC puidance O

-(O

~

9 h., with site-specific-and plant-specific assumptions and input data, are conservative. The total activity, by nuclide, released over the duration of the accident as provided by the FSAR, is used to determine the fraction of t he mix, as a whole, that ~each radionuclide comprises. Thus, in effect the fractions are averaged over the period of the postulated release.

These fractions are used along with the monitor response

-efficiency for each nuclide to normalize the irat rument response for.the different accidents. This methodology also provides specific fractions .for the radiofodine nuclides.

-Transparent to the user of the'worksheet, the method determines a' Ci/sec value for each nuclide and uses the RG 1.109 dose conversion factors to convert these to dose. The majority of this math was performed .during the development of the procedure, such that the user has a single conversion factor 3

for each accident sequence to equate monitor reeding to gross Ci/sec, and depending on the accident sequence a whole body and child thyroid dose conversion fcctor, specific to the accident.

There are worksheets specific to each type of radiation monitor.

The use of the FSAR source terms thus normalizes the monitor reading and dose calculation using radiation transport assumptions (eg: filtration, iodine partitioning) appropriate to the accident sequence. Note that only the nuclide fractions and not the total nuclide quantity are used in the above calculations. The actual radiation monitor reading is used to determine magnitude, and therefore, no correction for age is necessary. Although these source term mixes are postulated and may differ in magnitude and composition from an actual release nuclide mix, the methodology is more representative than the assumption that the mix is represented by a single

, O):

\

v L

~ 10 s

radionuclide. This methodology is implemented in the TRS-80 RADOSE code as well.

-To provide 'for the case where radiation monitors may be inoperative, the hand calculation methods provide a method based on etual X/Q and total FSAR released nuclide quantity.

This method bypasses the Ci/sec intermediate value. This

, method is also included in the TRS-80 RADOSE code.

1

1.2.3 ARERAS

The MIDAS sof*. ware uses a methodology similar to that

described above for the BVPS method and will uce the same source term ':ssumptions and monitor efficiencies used in.the development of the procedures currer.tly in use. However, the

. user will have the option to manually enter a source tera nix, other than those rasident in the code's parameter files, and to use that source term to normalize the instrument response and dose conversion factors. The dose conversion factors are the Q

, \s /

.same as described in section 1.1.3.

1.2.4 COMPARISON

As noted above, there are significant differences in the methodology between IRDAM cnd BVPS. The MIDAS methodology is similar to that of BVPS and should provide comparable results (see 1.1.3 above). Fxhibit 3 provides a comparison. of the accident sequence - specific dose rates determined by the BVPS methodology and the IRDAM gross results.

Exhibit 4 contains reproductions of the worksheets from the BVPS procedures. The worksheets were used instead of the RADOSE code since the worksheets allow entry of the intermediate Ci/sec result, whereas RADOSE operates on the raw monitor reading. The whole body results indicated that the ratio of the IRDAM results over the BVPS results ranged from 40.3 (IRDAM high) to 0.82 (BVPS high). Since the BVPS X/Q was about 2 times higher than the IRDAM X/Q, the difference between

,_ the two dose calculation methods (independent of X/Q) is twice V)

I

. 11

, that specified in Exhibit 3. Similarly, the thyroid doses V calculated by IRDAM were significantly higher.

Although the BVPS methods were verified prior to use, the fuel handling release calculation was performed raanually, using the source formulae and dose conversion factors from RG 1.109(5) in order to provide another means of comparison. This

' ~

calculation, exhibit 5, provides results that are identical to those obtained by the use of the worksheet, or that would be obtained by the RADOSE code.

A significant portion of the thyroid discrepancy can be assessed to the means of identifying the amount of iodine present. A user would probably not enter 100% as the iodine fraction. 'However, even the TID 14844(4) assumption of 25%

would lead to overconservative results in comparison to the Exhibit 5 iodine fractions. There is perhaps another problem.

f] Scenario 5 of IRDAM identifies itself as the 100% iodine source term case. The input summary verifies this with "... percentage noble gas = 0... ... iodine to noble gas ratio = 1.7E38..."

However, the output summary identifies whole body exposures identical to the whole body dose rates obtained using Scenario 2, which assumes 100% noble gas and no iodine. The code manual does not identify why the whole body doses are high in the absence of noble gases as identified by the input summary.

l The MID'.S code was not available to run test cases. However, as identified above, the MIDAS methodology is similar to that l of BVPS and will utilize site parameters identical to those in l BVPS. Thus, it is unlikely that the MIDAS results will ,

different significantly.

1.

2.5 CONCLUSION

The site - specific models provide comparable data which differs si8 nificantly from the data from the non-site D

l

~ . - -_- - - .-. .. -

• 12 4

-~

specific IRDAM model. Although the IRDAM_ inputs could be more y DN .1- advantageously selected, -the IRDAM results will not likely reflect the site -

specific radionuclide transport considerations that can be incorporated into a site - specific.

procedure. Further, it is unlikely, that in the early stages of an emergency for- which the code was intended, that sufficient NRC and licensee expertise will he available to identify these site-specific transport considerations,.

increasing the probability that the IRDAM generic defaults will be used.

1.3 Containment Leakage-1.3.1 .IRDAM: In this scenario, the release rate is generated by the code from input data on the containment leak rate and the Four cases can be input: fu21 melt, source -term assumptions.

gap release, coolant release, and containment monitor reading.

The method uses a default raionuclide mix based on the rated electrical power. The method utilizes one of four formulae:

Fuel melt (100% core inv) =

MWe(7.2E-2)(Leakrate_%)

a Gap release (1% core inv) =

MWe(7.2E-4)(Inakrate %)

Coolant Inv(.0001% core inv) =

MWe(7.2E-8)(Leakrate %)

l l

Monitor =

MWe(7.2E-4)(leakrate)(R/hr)/(Cal. Factor) i Although the IRDAM manual does not explicitly identify the ultimate use of the data obtained through the above, it is 4

expected that the code uses the gross release methodology identified in section 1.2. ,

1.3.2 BVPS: The BVPS methods do not calculate the Ci/sec from the

,, containment leakage as done above. The BVPS methods are based

< -%'/

13

.. -o

,e s on projecting- the amount of uncollected, and therefore

' ( )

\s / unmonitored, release by monitoring the collected release and incorporating appropriate . scaling factors to acccunt for the uncollected leakage.. This' monitor method can be used for either the gap activity case or the coolant inventory case.

The source ~ term used for this analysis is taken from the BVPS Final Safety Analysis report. The core inventory in the FSAR was calculated using parameters specific ta the core and . fuel.

configuration (n.utron flux, ' fuel density, etc). The gap activity is based on the core inventory and the fuel and nuclide specific diffusivity. The coolant inventory includes all- these considerations and additional plant specific parameters (coolant chemistry, coolant processing). Of these considerations, only the neutron flux is a function of core power (usually rated in terms of MWt rather than MWe to eliminate the need to consider steam plant conversion efficiency).

^

(n)

The fuel melt case is handled by using default FSAR results with actual X/Q values. BVPS does not currently have a formal procedure for equating the containment dome monitor reading to offsite dose, althou8h such a procedure is under development.

In establishing alarm setpoints for the dome monitors, the relationship between the monitor reading and containment activity was shown to be that a monitor reading of 31 R/hr equated to a site boundary whole body dose rate of about 1.2 l

mrem /hr and a child thyroid dose commitment of 170 mrem for l

every hour of inhalation.(15) In a related calculation (16) ,

it was shown that the air dose rate in the vicinity of the dome monitor was 1.4E6 R/hr for a source term based on fuel melt.

These calculations were based on shielding analyses performed r

l on the specific BVPS containment configuration (ie: large dry, l subatmospheric, annular crane wall).

I I

t l v i-

14 i- f-~3I 1.3.3 ARERAS: The MIDAS code analysis is similar to that described t

\s/ for BVPS and will use site parameters identical to those used in the current BVPS methodology. The MIDAS code, however, provides an additional capability to analyze the LOCA  ;

oransport.- This interactive program prompts the user for containment leakage, reactor coolant system leakage (or S/G tube leakrate), fraction of core inventory, and other similar plant parameters. An option provides for entering the dome monitor reading. This analysis is based on a generic algorithm driven by site parameter data files.

1.3.4 COMPARISON

The IRDAM scenarios did not address a ground level  !

release using containment leakage. Further, there is no identical method in the BVPS methodology, and the MIDAS software is not available for testing against IPDAM. Thus, no direct comparison is possible. However, it is possible to evaluate the ger.eric assumptions in IRDAM against the BVPS plant configuration.

()T IRDAM appears to assume that the relationship between gap activity and core inventory is 1:100. Table 14B-1 of the BVPS FSAR provides tabulations of core inventory and gap activity, based on the actual BVPS core and fuel configuration. Using data from that table, the fraction of core inventory in the gap ranges from 0.00014 (Kr89) to 0.17 (Kr85) with an average of l

0.0023, at equillibrium following full power operation for 650 days. These data compare to the 0.01 assumed by IRDAM.

IRDAM appears to assume that the coolant activity is 0.0001 of the gap activity. This appears to be low. A standard assumption used in determining coolant activity is that there is 1% failed fuel. This alone would infer a ratio of 0.01.

e The DVPS FSAR provides data that lead to a gap / coo? ant activity ratio of 0.004, which reflects the effect that coolant

.~-

i

F 15 (w - processing has on equilibrium activity. The site - specific

\~/ parameters which impact on the nuclide transport addressed by this scenario include coolant chemistry controls, degassification, demineralizers, and similar processes.

The derivation of the conversion factors used in .the formulae (7.2E-2,. etc) was not described in the IRDAM manual. However, it did not appear that consideration was given to ~ iodine mitigation features such as containment sprays or recirculation filters (not used at BVPS-1). The effectiveness of these features vary from plant to plant.

The containment dome monitor conversion of 1E4 R/hr for a containment full of gap activity, cannot be more than a rough estimate in that considerations signific3nt in such a l .

monitoring configuration, eg: geometry of detector / source configuration, response spectrum of the detector versus varying

[} source spectrum, can rcot be addressed in a generic solution.

Further, industry work into .the conteinment monitor, have

identified several aspects of such monitoring that make monitor-to-activity correlations tenuous at best. These topics were the subject of a Nuclear Safety Analysis Center (NSAC) workshop on Post-accident In-containment Radiation Assessments, conducted in Washington, DC.9/81.

l The BVPS procedure under development will use a radionuclide transport model that builds-up and decays radionuclides as a function of time following shutdown and of reactor coolant u system leakrate. The peak activity in this model will be l

reached sometime after shutdown. This differs from the FSAR assumption that the containment activity reaches the peak value l at time = 0.

l rh j Above, it was noted that a calculation performed using fuel i \

I \m/

l i

M.#aPM

. s- 16

,s melt- activity yielded 1.4E6 R/hr at the monitor location. If

'the FSAR provided ratio between the core melt activity and the gap activity of 0.0023 is multiplied against this reading, a rough estimate of the what the monitor might read for a gap activity situation can be obtained. The process yielded an estimate of 3200 R/hr, one third that assumed by IRDAM, Another numerical comparison can be made between' the IRDAM assumption for a fuel . melt and the results of the BVPS FSAR-analysis. Using the formula above with the 852 MWe rating of BVPS and 0.1% containment leakage:

Ci/sec = 852 x 7.2E-2 x 0.1 = 6.1 The recent re-analysis of the FSAR analysis for core melt provided that 21000 curies could be released over a time period of 3600 seconds, for an average release rate of 5.8 Ci/sec.

This appears to agree reasonably well with the IRDAM result.

'" Similarly, the re-analysis provided -a total release of 23.3 curies following an accident with a release of only gap activity to the containment. This. release equates to an average release rate of 0.0065 Ci/sec compared to the IRDAM result of 0.06 Ci/sec.

1. '> . 5 CONCLUSION: No direct comparison could be performed for the unmonitored cases. For the monitored case, the IRDAM result is high by a factor of 3, assuming maximum activity at time = 0.

The difference will be hi 8 her for the final BVPS method in which the maximum activity occurs sometime after shutdown, and for which the peak may be lower. Indirect comparisons for the unmonitored case showed an unexpected reasonable correlation for the fuel melt cases, but a difference of 10 between the gap cases. Due to the generic nature of the IRDAM method, this correlation may not be achieved at other sites.

v

17

(~~} 2. Source Term Adiustments

2.1 IRDAM

With the exception of the isotopic case, IRDAM provides a capability to modify the source terms determined by the methods described above in section 1. These methods provide the means to account for. relative iodine fraction present and effectiveness of filtration.

2.2 BVPS: Source term adjustments are inherently made within the method due to consideration of sice-specific source terms in the preparation of the methods. Also, real-time monitor data is used to drive the methods where appropriate.

2.3 ARERAS

Same as BVPS, except for the gro4s containment leakage model in which user specifies transport parameters during the analysis, and except that MIDAS provides an option to enter an I

actual source term to be used to normalize the monitor response, in lieu of the' file source terms.

O-t  ;-

v

2.4 COMPARISON

Addressed in section 1 above.

2.5 CONCLUSION

In as much as the source term modifications are l factored into the BVPS and MIDAS analyses during development, rather than during an emergency, it is unlikely that initial BVPS emergency response personnel will be able to discuss these considerations with NRC response center personnel. For this reason, initini NRC and licensee dose projections are likely to differ. Note that this is not the case with the offsite agencies,-in that these agencies generally use methods based directly on the BVPS methods. Further, when the ARERAS system is operational, these agencies will be able to access the results of the MIDAS methods by remote terminal, as will the NRC.

O

, ., '18 3.

-() Stebility Class Calculation N--? 3.1 IRDAM: IRDAM provides thrca'means to determine stability class.

These are tne lapse rate, sigma theta, and/or theta spread. As an alternative, the stability class may b% entered directly.

The lapse rate method requires the input .of height and temperature differences between two points of measurement and calculates tne lapse rate in degrees celsius per '100 meters.

The stability class is assigned using the lapse rate versus i stability class table in Regulatory Guide 1.23(13) ,

Sigma theta is based on the input of the wind variability and use of a conversion table in Regulatory Guide 1.23.

l The theta spread method simply determines sigma theta from the

input theta spread, and proceeds as above.

() 3.2 BVPS: The BVPS methods are based on lapse rate. However, that the height difference is set by site tower configuration, in the lapse rate is used in terms of degrees fahrenheit- per '115 feet. This protocol simplifies hand calcelations. In verbal comments provided during the 1981 NRC-site assessment of the

BVPS Emergency Preparedness Plan, the team meteorologist

-recommended several techniques to simplify the hand calculation procedures. One of these recommendations was to address only four stability classes: ABC, based on B; D; E; and FG, based on F. This technique was incorporated into the BVPS procedures.

3.3 ARERAS

The MIDAS software, at the-users options, uses either lapse rate or sigma theta, and in addition, can use split sigma I

l methods, in which the lapse rate is used for the vertical dispersion component and the sigma theta used for the r

horizontal component.

l 4

t

. . . ~ ~ .. . =. .

19 J

' /'") 3.4 ' COMPARISON: None required since . methods are selectable to 1 provide comparable results.

315 CONCLUSION: There can 'be differences in .the stability classifications if the different methods are used simultaneously. The BVPS methods are' based on the

' recommendations of Regulatory Guide 1.145, which provides that

' lapse rate (celta-T) method is acceptable without further 1 justification. The only . difficulty foreseeable is the misapplication of the site-calculated lapse rates in degrees

[ fahrenheit per .115 feet, against tables in different units. ]

Since the meteorological instruments read in fahrenheit and are

115 feet apart, the site's choice of units is appropriate in  !

that delta-t can be read directly with no further correction.

9

%./

j I

I 1

1 1

l l

l

20 es i 4. Default Values t

\_e' In that the ARERAS defaults were chosen on the basis of the

! ~BVPS. methods, there will not' be separate discussion of ARERAS defaults, except-for the cases where they do differ. The format of this section will differ in that the heaoing will identify the IRDAM default, and the following paragraph will discuss the default in light of the BVPS/ARERAS methods. S

4.1 Windspeed

2 m/sec There is no default windspeed specified for the BVPS procedures due to the availability of primary, secondary, and backup meteorological ~ sensors. In addition, 3VPS maintains periodic communications and a letter of agreement with the National Weather Service station located at the Greater Pittsburgh Airport. Two m/sec equates to 4.4 mph, which is high for~ a default windspeed for the nighttime inversion conditions which frequent the BVPS site.

$ (V)

4.2 Release Height: Ground Level This is consistent with the BVPS methods which treat all releases as ground level. ARERAS will treat the process 8as l- yent as an elevated release. However, this is- not a l significant accident pathway.

l l

4.3 Effective Stack Height: 50 meters l

In as much as all BVPS releases are ground level, this default l

j has no consequence.

1 i

l^  :.

l

@ It is noted that IRDAM does provide the user many opportunities to revise l- the def.*ults during a code run.

l

. . 21 4.4 Stability Class: F it w/s less than 5 m/s; E otherwise

{~]

V

'There is'no default stability class in the BVPS methods. Using a default of F or E for all . times of the day can be overconservative. A default of D would.be more appropriate for day time hours. Thia default could give rise to significant differences in dose projection results, if the site is using actual stability class and the NRC is using the default. The

'Xu/Q table on page 7 of volume 2 of the IRDAM manual shows a spread of 275 between the 1000 meter distance stability class A and stability class F.

4.5 Wind Direction: No predominant direction This default has no significance .in -IRDAM or BVPS in that direction dependent terrain corrections have not been included.

Terrain corrections are included in the MIDAS elevated models.

1

. U 4.6 Containment Inventory: Fuel Melt.

The BVPS and ARERAS methods use a site specific source term (by nuclide) based on the FSAR analysis. As noted above, this generic default can lead to significant differences in results.

4.7 Coolant Activity: 10 uCi/cc j The BVPS and ARERAS methods use a site specific source term (by l nuclide) based on the FSAR analysis. The total activity is about 45 uCi/cc. The MIDAS software will allow entry of a

?

recent coolant sample result in lieu of the default. It is likely the BVPS results will be hi her8 than those of IRDAM.

A

]

22 7~ 4.8 Iodine to Noble Gas Ratio: 0.02 N]

The BVPS and ARERAS methods have accident sequence-specific

-default nuclide files. Therefore, iodine / noble gas ratios are not explicity used as in IRDAM. As noted above, significant differences in results will be likely.

4.9 Time between shutdown and release: 0 hr The BVPS methods are based on the total activity postulated to be released over the duration of the incident, with the magnitude of the release being determined by an effluent radiation monitor. Thus, decay corrections are not used in the gross release (monitored) method. ARERAS provides for tima-dependent corrections to default source mixes and changes in instrument response due to time-dependent spectrum shifts, 1(} 4.10 Age of Material: Less than 1 day

\. /

See the response for item 4.9. Age correction is unnecessary 4.11 Iodine Filter Efficiency: 95%

The iodine filter efficiency assumed in the development of the BVPS methods was 99%. This value was chosen in light that it i

was below the technical specification efficiency of >99 % for h elemental and particulate, and in light of current research I which indicates that the iodine source term is too high.

i However, correction needs to be made in some accident scenarios for uncollected versus collected releases. (If 50% of the release is filtered with a 95% efficient filter, the effective l filtration is only 47.5%.

O

~s l

23 4.12 ' Containment Leak Rate: 0,17. for'PWR

(

-x/

This is consistent with the value used ir. the BVPS and ARERAS methods.

4.13 Duration of Release: 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> The BVPS FSAR provides that the containment will be -

subatmospheric in less than one hour with minimum engineered safeguards operating,'and much less in some sequences. In the event of a steam 8enerator tube rupture, the FSAR provides that the affected steam generator can be isolated in about 30 minutes (assuming no electric power available). The release stops shortly after isolation. The ma_n steam line break release lasts until the main steam stop valves go shut (in seconds) or the steam generator boils dry. The tank rupture sequences are limited by the physical capacity of the tanks.

The majority of the activity released in fuel handling accident

{}

'- in the fuel building will diffuse from the pool (and the building) within the first few minutes. The fuel building ventilation turnover rate is about 1 per hour. Thus for BVPS, the 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> assumption is simply too long.

4.?4 CONCLUSION: There are differences between the defsults used in the three methods. lnere are also differences between what is considered a default and what is considered to be quantifiable under most circunstances.

. v i

, .- _ _ _ , - . - e_ . . - . . , - . . . - - , . ,

.r . -,

24

(} y. Grcund Level Release

\~ '

5.1- IRDAM: The ground level calculations are performed with the following equation: ,

DR = (Q)(X/Q)(CF)

Where Q is the release . rate, CF is. the semi-infinite dose conversion factor and X/Q is the dispersion factor. The infant thyroid conversion factors incorporate a breathing rate of 0.25 m /hr. X/Q values are determined frcs a pre-determined matrix of XG/Q values. The matrix provides XG/Q values for all seven stability classes and six distances: 500, 1000, 2000, 3000, 8000, and 20,000 meters. The X/Q values are determined from the XU/Q values by dividing the X6/Q value by the windspeed in meters /sec.

Under the isotopic release cs se , . the X/Q values and the

[~ T se.ni-infinite. dose conversion factors are used directly in the L) above formula for each specific nuclide under consideration.

No whole body dose due to submersion in a cloud of iodine is assessed.

For the gross leakage, containment leakage, and/or coolant leakage cases, IRDAM calculates a total release rate. This release rate includes noble gases and iodines. The whole body Jose rate is determined by assuming this release rate to be all noble gas, thus indirectly addressing the whole body dose due to iodine. The thi roid dose is based on this total release, the iodine te noble gas ratio, and iodine filtration efficiency data. As noted abcve, the doses calculated are modified if the age of the release is less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> by an empirical, generic time-dependent correction.

,. 15 .:2 BVPS: The BVPS hand calculation methods use a matrix of XE/Q (s_- )-

f 4

r 79 ,e-, - - -

, , . p 4y = , - ~ . r --,-..,,-w ,

25

[}

values 'also. The matrix suppies XG/Q values for the four defined stability classes and 4 distances: Site boundary, 2 miles, 5 miles, and 10 miles - distances compatible with the protective action criteria and the offsite planning : zones. A

. graph ofXU/Q versus stability class and downwind distance provides a capability of determining XU/Q values for other distances. The XU/Q values are in units compatible with a wind speed in miles per hour. The RADOSE code calculates _the X/Q value using the formulae of Regulatory Guide 1.145(6)

(meander was not considered). The values of sigma-y and sigma-z are calculated by the code using the methodology of MESODIF II(0) . (A similar algorithm is in X0QD0')(") and PAVAN (18) , both NRC codes.)

The dose calculational model of the BVPS methods varies with the method. For the default FSAR case, the total FSAR-postulated activity released is used with the actual X/Q value. Semi-infinite dose conversion factors were multiplied

{A}

by the activity released during the development of the procedure, such that the user need only multiply a single factor by the current X/Q. This process was repeated for each accident sequence.

For the monitored dose calculation method, the FSAR analysis was used as a source of postulated source term mixes. In the development of the procedures, these source term mixes were used to normalize the instrument response (cpm and cfm to Ci/sec) and the dose conversion factors for each postulated accident. This process resulted in a set of conversion factors (monitor response, whole body, and thyroid DCF) for each accident and each monitor applicable to the accident sequence.

As the FSAR source terms were for the entire duration of the incident, the resulting source term is an average for the 1

(

~. __ _ . . _ _ _ __ _ __ _ . . . _ , - ._ . . , .

26 d.

je g accident duration. The magnitude is determined from the

\s) monitor reading.' There are no corrections for mix age or time since shutdcwn.

5.3 ARERAS

In the class A model, MIDAS calculates X/Q using the formulae of Regulatory Guide 1.145 including, when appropriate, meander. A new X/Q value is determined every 15 minutes, using averaged real-time meteorological sensor data collected over the immediately previous 15 minute period. There are options to use historic meteorological data (ie: a release that started earlier til an the actuation of the technical support center) or future data (ie: for a longer-term projection). In the class B model, the release is characterized as discrete puff releases each of which is represented by a set of particles. These particles are dispersed in X, Y, and Z planes by a modified potential field model established for the Beaver Valley site environs. It is beyond the scope of this report to discuss the meteorological bases of this model. This model is

_()/

\

site-specific.

The dose calculational methods of the MIDAS system are similar to those of the BVPS methods, with additional modifying parameters to fine tune the method to the release sequence / monitor configuration. These parameters include, for, example, monitor background subtract, monitor minimum reading, efficiency factor for the monitor, and nuclide - specific monitor efficiency (can be defined for different time periods to account for shifts in spectrum with time).

The class B model dose calculational method considers exposure from plume shine, ground deposition shine, and inhalation doses to the whole body, thyroid, and lungs.

5.4 COMPARISON

Comparisons of the dose calculation methodology O

V

--- ,.-y ,,n. , - , - , .--y_ . , , - . . - - -

.m-,.--._r.- --g+ - ---

27 p have been reported above. A comparison was made between tha X/Q methods. In this comparison, the IRDAM provided X/Q values (taken from .the scenario printouts) were plotted against downwind distance for B stability and F stability (exhibits 6 and 7).

X/Q values were calculated for the two cases using the methodology of Regulatory Guide 1.145(6) and the sigma-y and sigma-z graphs appended thereto. Meander was not considered.

The values calculated for several arbitrary distances were plotted on the same exhibits. To these, the values printed out by RADOSE for the same meteorological conditions were added.

Note that the RADOSE and DVPS cases incorporated building wake.

In this manner, Exhibits 6 and 7 provide a means to assess differences between the methods.

Exhibit 6 displays-the results for B stability and- 10 mph

[)

. v, windspeed. There is a maximum factor of 3 difference between the IRDAM and RADOSE results, with RADOSE higher. RADOSE appears to fit the Regulatory Guide 1.145 generated data better. Exhibit 7 displayed the results for F stability and 1

mph. Except at the close-in distances (due to building wake inccrporated in che RADOSE model), the IRDAM and RADOSE data agreed quite well witu the Regulatoty Guide 1.145 8enerated data (also incorporating the building wake correction).

No comparison can be made to the class B model, as it is not currently available. However, it is safe to assume that the class B model results will differ si8 nificantly for some

- combinationc of receptor distance and meteorological i

conditions, due to the higher degree of sophistication in the

! class B model.

5.5 CONCLUSION

There is reasonable agreement in the metaorological t (v l

)

f me+v a- e m at -.m. . + - yy, y -

, -- .yyor ww -, e ...---

y.1 y ------ rw - - . y. .=-=.gw.=s ---,-*,

3

.-. . . - - . .- _ - - - . - . . _ - . . - . . . _ . - _ - . . - ..._ - .-- ..- ... . . . .,.. =.. _.-.- .

4 i, .. .

28

'4 i .

i:-

~~

models utilized by IRDAM and BVPS. No general agreement can be

.= . expected between the class' B results-in MIDAS and the IRDAM -l 3 class A results. Refer to' previous conclusion sections for-I i

discussion of.the dose computation models.

?

i t

i i ,

f I

4 i

r .

6 1 .

s i

l .

l I

t i

l i

i l

t

' . - . - - . - , - - . . . . _ . - - - ~ . - . _ _ ~ , - . . , _ _ _ - , _ . _ _ _ _ _ . _ _ _

, - -,_ 29

, M., 6. Elevated Releases

'6.1 IRDAM: The code uses a cyl'ndrical source approximation in place of a more rigorous finite cloud model.- Unlike for~ the '

ground level cases, IRDAM uses a radionuclide breakdown for

dose' calculations. This mix is used with an age less than 24 l hours and greater than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This methodology offers significant advantages over that used in the ground level models. It- is not clear to this reviewer why emphasis was placed on.the elevated release over the ground level release, in as much as LOCA sequences at most reactors result in ,

significant ground level releases. Terrain corrections are not included.

6.2 BVPS: Not addressed in BVPS methodology.

6.3 ARERAS

The MIDAS saftware will handle elevated releases, using the formula of Regulatory Guide 1.145(0) . MIDAS uses a O

'v' finite plume model in lieu of the semi-infinite model, if selected by the user. The finite plume in MIDAS is based on

-the algorithm of Hamawi( ) and the methodology of reference 7, a more rigorous analysis than that of IRDAM. MIDAS incorporates site rpecific terrain correctiona into the class A elevated model.

6.4 CONCLUSION

ARERAS will provide results that are more representative of the actual release ccafiguration than will IRDAM, due to the incorporation of site specific terrain corrections. This can be envisoned by noting that IRDAM may treat a plume as elevated even after that plume has impacted on a hill side, making it become a ground level release. At the Beaver Valley site, a straight line plume will impact on terrain higher than the release point within 2 miles of the release point, in all directions, and much lower in some.

. 30 1

SUMMARY

x_-)

A review of the US Nuclear Regulatory Commission's Interactive Rapid Dose Assessment Model (IRDAM) against the current and future-(Atmospheric Radioactive Effluent Release Assessment System) models- at the Beaver Valley Power Station has-been completed. This review has identified differences in several areas between thn IRDAM code and the BVPS methods that can be

, attributed to the use of site-specific methods and assumptions in the BVPS methods as opposed to the generic methodologies and assumptions included in IRDAM. It is ncted that no attempt was made to assess code features, undocumented in the TRDAM manual, that could mitigate some of the descrepancies identified in this review. It was concluded that no revision was necessary in the BVPS methods currently in place and those planned for the future, as the methods are technically appropriate in the BVPS-specific setting. It was further concluded that the IRDAM

<- code cannot serve as a standard against which all other dose N- projection methods are compare <1, especially in licensing situations such as emergency exercises.

t J

l i

O

- 9- *~ - -e v-- , - - - , .- ,,---wv.m-- ,,y, on-- i.,,,_,,-, . , - - . - - .,--Tr r 'r w---7-r-----*'--*wm --~+-e--w'-w------

31 REFERENCES (D

V

1. R.W. Poeton, etal, Interactive Rapid Dese Assessment Model (IRDAM), NUREG/CR-3012, Pacific Northwest

-Laboratory, 1983.

2. DLC Calculations ERS-SFL-83-023 Dose Projection Source Terms ERS-AFM-83-024 EPP/IP-2.6.1, Issue 7 Rev. 1 ERS-SFL-83-025 RM-201 Reading that Corresponds to RM-219 ERS-SFL-83-028 EPP/IP-2.6.6, Dose Conversion Factors

3. USNRC, Assumptions Used for Evaluating the Potential Radiological Consequences of a Loss-of-Coolant Accident for Prespirized Water Reactors, Regulatory Guide 1.4, 1976

4. J.J. DiNunno, etal, Calculation of Distance Factors for Power and Test Reactor Sites, TID 14844, USAEC,

/ 1962

'(O_)

5. USNRC, Calculation of Annual Doses ' to Man froin Routine Releases of Reactor Effluents for the Purpose of Eyaluating Compliance with 10 CFR Part 50 Appendix I, Regulatory Guide 1.109, 1977

6. USNRC, Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plants, Regulatory Guide 1.145, 1983

7. D.H. Slade, Meteorology and Atomic Energy, 1968,-

TID-24190, Air Resources Laboratory, USDOC, 1968

8. D.C. Powell, etal, MES0DIF-II: A Variable Trajectory Plume Segment Model to Assess Ground Level Air Concentrations and Deposition of Routine-Effluent Releases from Nuclear Power Facilities, NUREG/CR-0523, Pacific Northwest Labs, 1979

9. Verbal comments presented by B. Zalcman of USNRC to Messrs Sieber, McIntire, and La Vie, September 1981

10. USNRC, Inspection Report 81-27, Emergency Plan Site Assessment Report for BVPS-1

32

>v

\~- 11. USNRC, Clarification of TMI Action Plan Requirements, NUREG-0737, 1980

12. USNRC, Criteria for Preparation and Evaluation of Radiological Emergengg_ Response Plans and Preparedness in Support of-Nuclear Power Plants, NUREG-0654/ FEMA-REP-1, 1980

13. USNRC, Onsite Meteorological Programs, Safety Guide 1.23, 1972

14. J.N. Hamawi, A method for Computing the Gamma Dose

• Integrals Il and I2 for the Finite-Cloud Sector-Average Model, Yankee Atomic Electric Company Report' YAEC-1105, 1976

15. DLC Calculation ERS-SFL-82-002, In-containment Hi-range ARM Alarm Set Points.

16. SWEC Calculation 12050 RP-081-0, LOCA DJses in'the Containment and Outside'the Personnel Hatch Door

17. J.F. Sagendorf, etal, X0QDOQ: Computer Piogram for the Meteorologica1' Evaluation of Routine Effluent Releases at Nuclear-Power Stations, NUREG/CR-2919, Pacific (x-/-) Northwest Labs,1982 .

18. T.J. Bander, PAVAN: An Atomospheric Dispersion Program for Evaluating Design Basis Accidental Releases of Radioactive Materials for Nuclear Power Stations, NUREG/CR-2858, Pacific Northwest Labs, 1982

~

l

.+

l l

"'e- d r -4 y,w,mo g g g g m- a- g 9--ey,. - -.+t-g-y 9e-- --%aa r - -- as- > = yn---r,nww-e- r -. fee g-

, . _. .. _ _ . - . = ~_ -. - . _

33 l EXHIBIT 1 Comparison of Iodine Dose Conversion Factors All are based on the Regulatory Guide 1.109 dose conversion factors,-but differ in the age group of interest.

RG 1.109_ Age Diff to Model ~DCF* BR** DCF Group IRDAM IRDAM 1.06E-2 0.25 0.00265 Infant MIDAS 4.39E-3 0.42 0.00184 Child -30%

BVPS 4.39E-3 0.63 0.00277 Child +4.5%

• mrem /pCi ** M /hr O

J IRDAM breathing rate is based on infant inhalation of 1400 M /yr.

3 (Reference 5) with the assumption that 50% of the day's inhalation occurs within eight hours.

MIDAS breathing rate is based on child inhalation of 3700 M /yr (Reference 5) averaged to one hour.

3 B_VP3 breathing rate is based on child inhalation of 3700 M /yr (Reference 5) with- the assumption that 50% of the day's inhalation occurs _within eight hours.

_. ~

34 EXHIBIT 2 7~

Comparison of Nobic Gas Doses (8v  ;

Assuming a X/Q of 1 sec/ meter cubed and a release of 1 Ci/sec, the following dose.1 are derived using the formulae of the three methods.

The BVPS methods are based on uCi/cc and <im as inputs. Thus, with a flow rate of 10000 cfm, the release concentration is 0.21 uCi/cc.

- DOSE, REM MIDAS / BVPS/

Nuclide IRDAM MIDAS BVPS IRDAM IRDAM Kr83u 8.620E-03 2.196E+00 8.547E-03 254.756 .992 Kr85m 1.330E+02 1.404E+02 1.323E+02 1.056 .995 Kr85 1.840E400 1.980E+00 1.821E+00 1.076 .990 Kr87 6.750E+02 7.056E+02 6.699E+02 1.045 .992 Kr88 1.680E+03 1.737E+03 1.661E+03 1.034 .989 Kr89 1.890E+03 1.971E+03 1.877E+03 1.043 .993 Kr90 .000E+00 1.863E+03 1.764E+03 .r49 .000 Xe131m 1.040E+01 1.782E+01 1.035E+01 1.713 .995 Xe133m 2.890E+01 3.744E+01 2.835E+01 1.296 .981 Xe133 3.360E+01 4.032E+C1 3.318E+01 1.200 .988 Xe135m 3.560E+02 3.852E+02 3.528E+02  !.082 .991 7..s Xe135 2.060E+02 2.196E+02 2.045E+02 1.066 .993

/ 's Xe137 1.6203202 1.728E+02 1.604E+02 1.067 .990 V Xe138 1.010E+03 1.053E+03 9.975E+02 1.043 .988 AVERAGE (Less Kr83m in MIDAS) 1.137 .920 g)

(

~s m._

j;;, ,

s 35 ,

e t

i i

. EXHIBIT ,,2A

Input Data to Exhibit 2 i

IRDAM MIDAS BVPS (rem-m3/ (May/ dis) (rem-cc-min /

Nuclide Ci-hr) hr-uCi-ft3) 4 Kr a3m 8.62E-03 .00244 4.07E-06 Kr85m 1.33E+02 .15600 6.30E-02 j Kr85 1.84E+00 .00220 8.67E-04 Kr87 6.75E+02 .78400 3.19E-01

Kr88 1.68E+03 1.93000 .7.91E-01 Kr89 1.89E+03 2.19000 8.94E-01 Kr90 .00E+00 2.07000 8.40E-01 Xe131m 1.04E+01 .01980 4.93E-03 l Xe133m 2.89E+01 .04160 1.35E-02 Xe133 3.36E+01 .04480 1.58E-02 i- .

Xe135m 3.56E+02 .42800 1.68E-01 2.06E+02 .24400 9.74E-02

. Xe135 Xe137 1.62E+02 .19200 7.64E-02 Xe138 1.01E+03 1.17000 4.75E-01

IRDAM

DR2Q x X/Q x DCF

DCFs from IRDAM

! MIDAS: DR=(9E-4) x X/Q x SUM (Q x Eg) '(Q in uCi/sec)

Ey from table 2.2.5.2.2-4 of MIDAS Reference Man.

t BVPS: DR=X/Q x FR x SUM (Q x DCFi) (Qin uCf /cc)

FR and DCF from EPP/IP-2.6.6 f

t

(

(])

. 36 e<%

( ) EXHIBIT 3

\_/

Comparison of IRDAM versus BVPS Methodology Cross Release Method Bases IRDAM: Scenario 2 (WB), Scenario 5 (Thyroid); 1 Ci/sec, 10 mph, B ,

stability, 1.9 mile receptor, 2.00E-7 sec/ meters cubed BVPS: Exhibit 4; 1 Ci/sec, 10 mph, ABC stability, 2 mile receptor, X/Q = 4.47E-7 sec/ meters cubed

-BVTS ----IRDAM* IRDAM/BVPS**

Sequence W. Body Thyroid W. Body Thyroid W.B Thy Generie 6.04E-5 1.97E-1 ---

VCT Rupture 4.8E-6 negl 12.5 Fuel Handling 1.3E-5 9.8E-6 4.7 2E4 T GST/WGDT Rupture 1.5E-6 negl 40.3 (G Main Steam Break 2.8E-5 2.1E-3 2.2 94 LOCA (Gap) 6.8E-5 1.1E-3 0.88 179 LOCA (RCS) 7.0E-5 2.4E-3 0.86 82 S/G Tube Rupture 7.4E-5 1.8E-4 0.82 1E3 Average 8.9 4.2E3

• IRDAM results are based on 100% noble gas or 100% iodine. The IRDAM scenarios only addressed cases with noble gas cu; iodine.

However, note that scenario 5 (the 100% iodine case) provided whole body doses equivalent to that of the 100% noble gas case.

There is some degree of uncertainty on the IRDAM code handling of 0% noble gas as indicated on the scenario input summary.

• • These ratios are based on the results obtained with the noted X/Q values and Ci/sec release rate. The actual magnitude of difference in the dose calculation method can be obtained by dividing the printed ratios by the ratio between X/Q values:

(2E-7)/(4.47E-7) = 0.45.

n  !

v

.* 37

, E xHi GIT 4-A.T

'"7 -

SPING -4 AVER /GE Ww.vstso,owtos a b e d X X -

g,i ,

/f. OC., Rate ILI Monitor CPM Flow Rete CFM g Cs/sec Mh 5.52 E-6 I 4 25 E-4 087 Awame 1.84 E-5 2.67 E-6 1.41 E-3 [ggQ Fuel Henemie 2.06 E-4 7ggg usL armen 1.11 E-6 8.54 E-5 WGDT Aseasee 1.84 E-5 1.41 E-3 $CgHsvtO 2 LOCA

. wra soos GAP Act. 4,37 E-7 3.37 E-5 hW'I4 #

"* fia'/cw f*aare 4.29 E-7 3.31 E-5 B s h b[#h Ch7 cng l0/kpk i , ,.

h g Does Factor Does Pro l.

Distance Rete Y/Q when aewy Rate Time DOSE AMes Cusse asc/m3 Chet Deyred N/w w ggy g q,q-) e 7 x l 08EI = 48 % x = CT f,w r_m,.-e l x

~

O .

x

d. '// 6/

=

l.3 E-5 x =

j x = x F,aon et.

q , t{. <Og-7

'" ~"'

~s, x .2./96/ = 9.'$ E-6 x =

x .g.< fo 60 = l.6 E-b a = Osr/

x

~

I' 'b W

[20PToR6 x

h /96[ = 2,$ 6-[ x ': N A lH x' [']

f. M b l .

x = 2.l b**$ x = @@

Other x = x =

- W ' x = x =

@ Frorm Attacrenent f l

<., t_

1' .' - g. ,;; ,

pg g gangs,,, LOCA g 2.91E1 {f.52E2 ,

( 2.19 E1 with 100% Gap Act. ( 2.52 E3 1 GSTN W W EO g g j1.54E2 i

! a.a.1 - e ,

w. a s.ss E3 l omre nue name  !

l l

w.

38 Ex. H i S t T 4- (cos 't)

A.

7K -

SPING -4 AVERAGE O M ww.etc.owres V . i, e ,

X X _

Row I 'J6O Rate Monitor CPM Flow Rate CFM g Cllsee M Ause== 5.52 E-6 4 25 E-4 N

. GET neere 1.84 E-5 1.41 E-3 0M lbD M Fuse Hanetne 2.67 E-6 2.06 E-4 $ppggno g usL armen 1.11 E-6 8.54 E-5

@ Ro6emme 1.84 E-5 1.41 E-3 O*8ff hfi .

LOCA 8

. wvth foot om Act. 4.37 E-7 3.37 E-5 ao fuel /cw f*4a* 4.29 E-7 3.31 E- 5, p k

?

Ch7 Cii9 d e f g h I h g Dooo Factor Dose Proj.

Distance Rate X/Q ~ wnese aser Rate Tirne DOSE

' Asises Cusse sec/m3 caer Triy,omt afw w agnr

ite x /,6d6 2. = (,.SE-5 x = Lcc A rm /

• y. (7c7 x 2.S2 E3 = /. /6-3 x -

s~'/ G tP Q

ac-r, . i

! x /.S6 E 2 = 7.CE-5 x = Locp i 2.0

/ x M '/7E-7 x

f.3863 .

= p.y E-3 x = y/.7fjy. , -

l o x /(& = 7.4 c-5 x = g,7g y G .

, l x 4. 47 E-7 . n, u

, x Spr, = /,9 E-y x = gueroit s t

I x = x =

l x

x = x =

c x = x! =l x

g W x = x =

@ Frorrr Attacewnerw r

= ~~- - - { :.",:;

{ 0."

M Handg 2.91 E1 LOCA 1.52 E2 2.19 E1 weth 1005 Gap Act. 2.52 E3 GSTN Rupture -i 3ACEO g .g [ 1.58 E2 a .a.1 - s r M. ' s.m a q DATE TIME NAuti 4 Q2cg ([J(>[g P- 2.b.l 4bCb(MPP(t b

4 v 39 EXHIBIT 5

- (p) . Verification of BVPS Result for Fuel Handling Accident -

v METHOD: DR = SUM1(Q

• EF 1 )

• X/Q ar Q 1 = S i

• Q t L DR = Q

• X/Q

• SUM1(S *NF) 1 where: DR = Dose rate, rem /hr Qg = Total release rate, Ci/sec X/Q = Dispersion, sec/m 3 DCF - . factor for nuclide i, 1 Dose cong/Ci-hr rem-m S - Fraction of nuclide 1 1

---Ratioed DCF i' Nuclide Curies Fraction DCF Whole Body Thyroid

~

. (] Kr85 -2.89E3 0.0263 1.61E-5 4.23E-7

v< Xe131m- 1.70E4 0.1546 9.15E-5 1.42E-5 Xe133m 1.00E3 0.0091 2.51E-4 2.28E-6 Xe133 8.91E4 0.81 2.94E-4 2.38E-4 Xe135 -5.06E0 4.6E-5 1.81E-3 8.33E-8 I131 8.54E-1 7.8E-6 2.90E-3 2.26E-8 I132 5.84E-1 5.3E 6 1.60E-2 8.48E-8

- 1133 2.18E-2 2.0E-? 4.90E-3 9.80E-10 l

l 1.10E5 Effective = 2.55E-4 1 -

l I131 7.8E-6 2.44E1 1.90E-4 I132 5.3E-6 2.91E-1 1.54E-6 l- I133 2.0E-7 5.78E0 1.16E-6 l

l Effective = 1.93E-4 3 3 Vnole body DCF = 2.55E-4 mrem-m /pCi-yr = 2.903E1 rem-m /Ci-hr 3 3 Thyroid.DCF = 1.93E-4 mrem-m /pCi-yr = 2.202E1 rem-m /Ci/hr W. body DR = 1 x 4.47E-7 E x 2.903E1 = 1.298E-5 rem /hr l

Thy,eie DR = 1 g x 4.47E-7 x 2.202E1 9.843E-6 rem /hr g =

'[

3gy Exhibit 6 B Stability; 10 mph l ."esults from RRDOCZ

- -Results f rom IRDRM N Results from RG1.145 (bide wake)

+ Results f rom RG1.145 (no wake) 1E-5 b L.

+ \

e -

9 e

+

y1E-.

.D y\-

" \ 1

\

OE. h.\

g e 1E-7 \ >

e + qN i s o 4N

\

X \s

's 1E-s l

l +

l

=~

1E-9 v in to m eee ae in m o - -

aaa n 1 DOWNWIND DISTRNCE, Mi

4 4o Exhib1t 7,

<m IE-2 ,

L)

F Stability; 1 mph

\' -

Results from REOSE

\ --

Results from IRDAH '

\ u Results from RG1.145 (bidg wake)

N + Results f rom RG1.145 (no make)

\\ '

1E-3 l

L

\ -

r

G N

o l .o s 3

o

+

N(

i 1E-4 t +  %

oJ o.

l 0 l 8 N

+ \

l c T

l \

1E-5 -

+

i 1E-6

'9 9

• 9""89 " 9 O - - mum ,m n. m i

DOWNWIND DISTRNCE, Mi a