ML20148L712

Notifies DOE of Several Issues Re Hgsystem/Uf Computer Code Suite Used for Consequence Analysis as Part of DOE SAR Upgrade Program & Plans for Resolution

ML20148L712
Person / Time
Site:	Portsmouth Gaseous Diffusion Plant, Paducah Gaseous Diffusion Plant
Issue date:	06/13/1997
From:	Woolley R UNITED STATES ENRICHMENT CORP. (USEC)
To:	Devault R ENERGY, DEPT. OF
References
NUDOCS 9706190349
Download: ML20148L712 (4)

Text

'

.e ie

.e United States EnncnmInt Corporation 2 Democracy Cemer 6903 Rockledge Dnve Bethesda, Mo 20817 Tel: (301) 5G4-3200 Fax: (301) 564-32ct United States Enrichment Corporation June 13,1997 i

Mr. Randall M. DeVault Regulatory Oversight Manager i

Oflice of the Assists Manager for Enrichment Facilities U.S. Department of Energy i

P.O. Box 2001 Oak Ridge, TN 37831-8651

Dear Mr. DeVault:

United States Enrichment Corporation (USEC) - Paducah Gaseous Diffusion Plant (PGDP) -

l Portsmouth Gaseous Diffusion Plant - Docket Nos. 70-7001/70-7002 - HGSYSTEM/UF 6

Issues and Plans for Resolution The purpose of this letter is to notify the Department of Energy (DOE) of several issues regarding the HGSYSTEM/UF computer code suite used for consequence analysis as part of 6

the DOE SAR Upgrade Program and our plans for resolution. This letter is in follow-up to a discussion between yourself and Mr. Stephen Routh on June 9,1997.

As required by Issue 2 of the Paducah and Portsmouth Compliance Plans, USEC is preparing amendments to its NRC Certification Applications to incorporate the results and l

analyses of the DOE SAR Upgrade Program presented in KY/EM-174 for Paducah and POEF-IAIES-89 for Portsmouth. During a recent verification and validation of the HGSYSTEM/UF6 suite, several issues affecting the DOE SARUP consequence analysis results were identified by j

USEC/LMUS. A summary of the issues follows:

Issue 1 l

,Y )'

t For scenarios involving building wake effects or plume lift off, the version of HGSYSTEM/UF used for the DOE SARUP calculated the consequences using a method 6

inconsistent with the documentation. Specifically, the " Briggs well mixed wake model" and " lift off correction factor" employed by the suite have been modified due to

(

reviewers' comments, but the results were not been updated.

9706190349 970613 Issue 2 PDR ADOCK 07007001 C

PDR The DOE DAC results are being used as baseline test cases for USEC/13fUS's code verification and validation (V&V). However, the V&V effort cannot be completed without documentation of the known variance in results using the previous calculation of the "well mixed wake model" or the " lift off correction factor."

M0054 g/m lllllll!!llllll,lllllll!.ll!llllli Offices in Pad ucah. Ke tucky Portsmouth. Ohio Washington. DC j

s 2

Mr.'Randall M. DeVault June 13,1997 Page Two Additional detail regarding the " lift off correction factor" and the " Briggs well mixed wake model" is provided in the enclosure.

M As previously discussed with members of the DOE /LMES SARUP team, we have concluded that the following tasks must be performed to satisfactorily resolve these issues for submittal of the certificate amendments required by Compliance Plan Issue 2:

1.

Develop a technical justification and basis for submitting to the NRC the older results presented in the current DOE SARUP DACs.

2.

Either (a) define and defend an acceptance criteria to use in the V&V [ allowing l

USEC/LMUS to establish the acceptable range of comparable results between old and new computer runs], or (b) provide additional cases using the new analysis techniques for use in the V&V.

As agreed with you in the June 9,1997 telephone conversation, USEC will work with LMES to complete the above tasks.

If you have any questions, please contact Steve Routh at (301) 564-3251 or J.D. Sohl at j

(502) 441-6607. There are no new commitments contained in this submittal.

i i-Sincerely, i

l obert L. Woolley Nuclear Regulatory Assurance and Policy Manager l

Enclosure:

HGSYSTEM/UF Use of " Briggs Well-Mixed Wake Model" 6

I l

cc:

DOE Site Safety Representative, PGDP DOE Site Safety Representative, PORTS R. C. Pierson, NRC NRC Resident Inspector - PGDP NRC Resident Inspector - PORTS l

i i

l l

l O

i Enclosure to June 13,1997 Leuer i

Page 1 of 2 HGSYSTEM/UF, Use of " Briggs Well-Mixed Wake Model" The well-mixed wake model (or Briggs model) is an empirical formula that provules a conservative best-fit to wind tunnel data. Two wind tunnel experiments focused on buoyam plumes affected by building wakes. The latter set of wind tunnel experiments involved i

complicated release simulations for a variety of vent configurations, building shapes, and wind angles. The equation for the model used in the DOE SARUP dispersion / consequence DACs is:

)

I

' Q'

-91.(F,,f

• C'.

v4 2

e u R; y

,g

,,g

, 3,,

0.0123 0.117 1 3.39 i F""

n

8 r R, g R, H,,

R* >

r l

where Q,is the mass flux (kg/s) of the plume constituent (either UO:F or HF) involved in the 2

building wake, un is the wind speed at the top of the building (m/s),

R2 is the scaling area in the wake (m ), with R being the representative scaling length of 2

the building (m),

F., is the nondimensional buoyancy flux term, j

x is the downwind distance from the source to the receptor (m),

W, is the width of the building (m),

H, is the height of the building (m), and a, and a, are the Gaussian horizontal and vertical dispersion parameters, respectively (m).

l l

In Equation (1), the numerator term, e# F, is defined as the " lift off correction

)

factor," or Bw, and describes the decrease in ground-level concentration due to buoyant lifting of the plume. This is shown as Bwe (2) i where F.. is the nondimensional buoyancy flux of the term. This term is calculated as F*

F,,.

(3)

,W ug g where F,is the buoyancy flux calculated using the standard Briggs plume rise equations (m'/s').

P Enclosure to June 13,1997 Letter Page 2 of 2 During the peer review! of the HGSYSTEM/UF suite, Dr. Briggs suggested that a slight modification to the well-mixed wake model was required to ensure that "better preditions would be produced for wider range of building shapes and release configurations." The Briggs model was modified to:

Qc W

e C' -

f

'3E 2

f i

uR,

. F*, ', '

4 x 'o' (4) 2 y

.o 0.037 0.03 H,

H, f,

g g

and the " lift off correction factor" was redefined as Bw*'

}

1 Equation (4) is consistent with the documentation presented in the HGSYSTEM/UF6 manuals. However, results from previous analyses performed using Equation (1) to model well-mixed wake effects and plume lift off have not been updated in the DOE SARUP with the new relationship presented by Equation (4).

Based on runs made by USEC/LMUS, for the cascade facilities scenarios in the DOE SARUP, the use of Equation (4) generates a 20 to 30 percent increase in consequences over the results calculated by Equation (1), in the near-field (100 to 500 meters). At the site boundary, the difference results in a 5 to 10 percent increase in consequences.

No determination of the impact on the results for UF handling and storage facilities 6

scenarios has been made by USEC/LMUS; however, this issue could affect scenario results that were calculated using the UF. MIXER, POSTMIX, POSTAP, or POSTHEG codes.

1Britter, R., G. Briggs,1. Sykes, HGSYSTEM/UF, Peer Review Panel Report, September 3,1996.