ML20028A659
| ML20028A659 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 11/22/1982 |
| From: | Sheron B Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20028A640 | List: |
| References | |
| NUDOCS 8211240147 | |
| Download: ML20028A659 (19) | |
Text
UNITED STATES OF AMERICAN NUCLEAR PECilLATORY COMMISSION BEFORE THE AT011IC SAFETY AND LICENSING APPEAL BOARD In the Matter of METROPOLITAN EDISON COMPANY, ET AL.
Docket No. 50-289 (Three flile Island, Unit 1)
AFFIDAVIT OF BRIAN W. SHERON CONCERNING TPE APPEAL llEMORANDUll AND ORDER OF NOVEliBER 5, 1982 1.
I, Brian W. Sheron, being duly sworn, state as follows: I am Branch Chief, Reactor Systens Branch, Division of Systems Integration, Office of Nuclear Reactor Regulation. A copy of my professional qualifications is attached.
2.
In its Memorandum and order of November 5,1982 the Appeal Board stated its tentative views that neither the boiler-condenser mode of natural circulation nor feed and bleed have been demonstrated on the record as viabic means of recoving decay heat at TMI-1. The Appeal Board further stated that Semiscale tests S-SR-1 and S-SR-2 raise serious concerns about the viability of feed and bleed. The purpose of this affidavit is to address the Appeal Board's concerns regarding the significance of the Semiscale tests and to provide comments on the implication of the tests on the capability of TMI-1 to be cooled in the feed and bleed mode.
l
\\
3.
There are two Semiscale tests in question, S-SR-1 and S-SR-2.
They were performed by EG8G, Idaho, under contract to the Office of 8211240147 821122 PDR ADOCK 05000289 G
Nuclear Regulatory Research at the request of the Office of Nuclear Reactor Regulation. The purpose of these tests was to provide an experimental data base for a " feed and bleed" mode of operation following a small break LOCA. The data were intended for computer code verification. Before discussion of the test results, it is useful to discuss how Seniscale test data has been and will continue to be used in the licensing process.
4.
Use of Experimental Test Data in the Licensing Process No test facility, whether it be Semiscale, LOFT, FLECHT etc.,
exactly reproduces the behavior of a large PWR, Some aspect of the power plant facility is scaled in the test facility. For example, if the volume of the test facility is less than the volume of a large PWR, then the primary coolant system surface area will not scale in the same proportion as the volume. Similarly, if elevations in the scaled facility are not preserved, then cravity-dominated hydreulic behavior can be distorted. The scaling of the Semiscale facility has been selected as an optinization among such competing factors, including costs.
In general, and with some compromises among the competing scaling interests, Semiscale simulates most of the important phenorena associated with PWR behavior.
5.
However, the staff has nev_er taken Semiscale results, (or for that matter, any other test results including LOFT) and applied them directly to a large PWR. We have always maintained that the results from Semiscale and other test facilities are primarily for
code verification purposes. Our confidence in understanding large PWR behavior, including feed and bleed operation, is predicated on confidence in the computer codes which calculate the behavior. The main objectives of these scaled tests are to look for new or unique thermal-hydraulic phenomena associated with transient and accident scenarios and to assure that the computer codes are capable of predicting the observed behavior. By demonstrating that the computer codes can properly calculate and predict the behavior of scaled facilities, such as Semiscale and LOFT, under conditions similar to those that could occur in large PWRs, we believe that there is reasonable assurance that these sane computer codes can 1,e used to directly predict the behavior of the large PWRs.
6.
In summary, data from any test facility, such as fron Semiscale or LOFT, cannot be directly applied to a large PWR. Rather, it is used to demonstrate the ability of a computer code to predict the relevant thermal-hydraulic phenomena so that sufficient confidence can be gained that the code can be applied to predict the behavior of a large PWR.
7.
Theoretical Capability to Feed and Bleed The ability to successfully feed and bleed involves meeting two conditions; the first is being able to " bleed" primary coolant at a sufficient rate that the net energy renoved by the coolant discharg? is equal to or greater than the energy produced by decay heat. The second is being able to " feed" the primary systen with sufficient makeup coolant so that prior to the prinary coolant
inventory dropping to an unacceptable level below the top of the core, the net mass loss from the system is zero or negative.
Some plants such as THI-1 have high pressure HPI pumps that can operate at the pressurizer safety valve set pressure. The pressurizer safety valves provide a large capecity for heat removal so that they could cycle open and closed throughout an event in which feed and bleed cooling was being used. The accurate prediction of the safety valve flowrate at any time would not be significant since errors in predicted flow rate would only affect the cycling rate of the valves a'nd not the total mass loss. For plants with low pressure ECCS systems which cannot deliver coolant at the safety valve set pressure, the PORVS must be opened to depressurize the reactor system sufficiently to provide feed and bleed cooling. For example, analysis by Westinghouse of its plants with low head HPI punps have demonstrated that the PORYS must be I
opened before steam generator dryout occurs.
If the PORVs are l
opened early enough, a time is reached before inventory the drops to an unacceptable level below the top of the core when the steam production in the core becomes equal to on less than the PORY flow.
8.
To successfully cool the core, ECCS flow must also be greater or equal to PORY or safety v'alve flow (depending on the design) before core uncovery occurs. Safety valve or PORY flow and ECCS flow are both functions of pressure. To illustrate this pressure dependancy of plants with low pressure ECCS, EG&G, Idaho, its their recent reports, has graphically portrayed the ability of low pressure ECCS
plants and Semiscale to feed and bleed with curves similar to the one provided as figure 1.
These curves are for illustrative purposes only since decay heat actually varies with time, and the time of PORY opening is not accounted for. As can be seen from figure 1, point Number 1 shows the systen pressure at which the PORV energy removal rate equals the decay heat. For all system pressures above that associated with point 1, the PORY energy removal capability exceeds the decay heat generation rate and the firstcondition(abilitytobleed)issatisfied. Point 2 on figure 1 is the point where the mass injected by the HPI equals the mass lost out of the PORV. For pressures to the right of point 2, the mass loss by the PORV will exceed the HPI makeup rate and would result in eventual core uncovery. For pressures to the left of I
point 2, the HPI flow exceeds the PORY flow, and a net mass increase results. Thus, all pressures to the left of point 2 will satisfy the second condition (ability to feed).
In order to meet both conditions for successful feed and bleed, the plant must be i
within a finite pressure band in which both the PORY energy removal 1
rate exceeds the decay heat and the HPI mass addition rate exceeds the PORV mass depletion rate. This is shown as the " operating band" in Figure 1.
The capability of any sy5 ten tos feed and bleed, either a scaled test facility or a commerical PWR, is uniquely dependent upon the PORY or safety valve flow and the HPI pump flow characteristics.
l Although not shown on figure 1, system geometry also plays an l
inportant role. These curves are only illustrative of one point in
time. One curve of this type does not demonstrate conclusively whether a plant can be successfully cooled in feed and bleed or not. This can only be shown by a detailed computer analysis of the entire transient, such as can De obtained using the RELAP5 code.
Seniscale Behavior During Feed and Bleed 9.
Before describing the Semiscale tests in question, it is necessary to suminarize the primary coolant system behavior following a loss of all feedwater.
Assuming the event is initiated by a loss of all feedwater (both nain and auxiliary), the reactor will trip automatically, most likely on either a loss !
. sater or on low steam generator level. The RCS pressure will initially drop due to steam bubble j
expansion in the pressurizer which occurs due to hot leg coolant shrinkage following trip. This pressure drop is typically around a few hundred psi. The systen pressure will then increase as the hot leg coolant heats up to establish the necessary driving heads for natural circulation. The steam generator secondary coolant levels, initially at their operating levels, begin to drop due to boiloff without feedwater makeup. Assuming the condenser is not available, this boiloff will occur at the secondary safety valve setpoint (around1100 psi). Once'the secondary levels drop sufficiently, primary to secondary heat transfer will quickly decrease. After 1
the secondary side of the steam generators boils dry, heat produced in the core will no longer be removed by natural circulation to the secondary and will thus begin to raise the temperature of the l!.-
~
- w prinary coolant. As the prinary coolant heats up, it expands, 1
pushing. liquid into the pressurizer and in'.reasing the system pressure by compressing the stean in the pressurizer. Once the pressure reaches the PORY or safety valve setpoint, the steam in the pressurizer steam space will be expelled through these valves.
(Note that no operator actions have been assumed yet with regard to feed and bleed - These will be discussed later). Once the steam in the pressurizer is expelled, the system will be water-solid and the safety / relief valves will begin to discharge water.
- 10. After the prinary coolant reaches the saturation temper 6ture corresponding to the safety or relief valve set pressure, steam generation in the primary system will begin by boiling in the core.
Steam bubbles exiting the core will rise to and accumulat'e at the top of the reactor vessel. The liquid displaced by this growing steam bubbic will be forced into the pressurizer and out of the safety / relief valves. Once sufficient steam accumulates in the upper vessel region to uncover the hot leg nozzles, the ste!m will then flow to top of the hot leg U-bends in a B&W plant and accumulate. Again, water displaced in the steam generater and/or l
hot leg U-bends is forced into the pressurizer and out the l
t safety / relief valves. Once the steam bubble grows in the steam generators and the liquid drains don:ns, the hot legs, and in particular the surge line entrance, will be uncovered. Steam l
generated by core boiling can now enter the pressurizer. Note that the core has remained covered so far, and is in a pool boiling l
mode. Once steam can enter the pressurizer, it will discharge
through the safety / relief valves. At this point the rate of mass i
loss from the systen (ecreases significantly as'the safety relief valve flow transitions from liovid to steam.
If the system was allowed to continue in this mode with no primary makeup flow the i
continued mass loss from steam discharge would produce a very slow core uncovery.
I
- 11. The previous description was provided in order to show under what corditions the feed and bleed mode of cooling becomes effective.
Up until the time that the hot leg uncovered and the steam generated in the core was capable of passing out of the safety / relief valves, the mass loss rate from the system was high beceuse of the liquid discharge from the safety / relief valves.
This nass loss rate would be considerably in excess of the HPI mass nakeup rate.
Thus, even if the HPI punps were turned on early in the event and could inject at high pressures, a net mass inventory loss would be expected. However, once the net mass inventory loss was sufficient to uncover the hot legs and allow steam to enter the surge line and discharge from the safety / relief valves, the HPI mass addition rate will usually exceed the mass loss rate through the safety / relief l
l valves, and inventory recovery would proceed. Because the liquid I
to stean transition occurs with the vessel liquid level above the top of the core, no core uncovery wculd be expected and feed and *~
bleed would be considered successful. Therefore, the key to successful feed and bleed is that subsequent to hot leg uncovery,
9 the HPI r: ass flow should exceed the safety / relief valve steam flow, and the steam flow energy conte 7t out of the safety / relief valve should exceed the decay heat input.
If an analyses shows that subsequent to hot leg uncovery the safety / relief valve flow still continues to exceed the HPI flow, then core uncovery might evcntually occur.
In such a case, the analysis would have to be extended to verify that unacceptable core uncovery did not occur before the safety / relief valve flow decreased below HPI flow. As vill be discussed in the following sections, this latter condition occurred during Semiscale tests S-SR-1 and S-SR-2.
We will also shew that our analyses of a plant similar to THI-1 indicate that subsequent to hot leg uncovery, the HPI flow exceeds the PORY steam flow giving us reasonable confidence that feed and bleed has a good potential for success at TMI-1.
Semiscale Tests S-SR-1 and S-SR-2
- 12. The two Semiscale tests, S-SR-1 and S-SR 2 were designed to investigate feed and t,leed operation for plants with high and low head HPI pumps, respectively.
For both of these tests, the core simulator unexpectedly uncovered and caused the tests to be prcr:aturely terminated.
In the following section, the reason these f
tests produced core uncovery is discussed in detail. Also 1
discussed is the rationa.le which supports the conclusion that, had these tests been preanalyzed, their results would have been l
anticipated and predictable.
- 13. For Semiscele test S-SR-1, which had HPI flow scaled to a Westinghouse high head HPI plant (North Anna), mass and energy balance curves predicted the existence of a stable feed and bleed pressure band between approximately 1100 psi and 2235 psi. However, the facility experienced leakage from the primary system during the test which was not conpensated for. This led to a system mass loss which effectively incrersed the PORY mass removal curve so that the intersection of the PORY mass removal curve with the HPI flow curve occurred at an operating pressure to the left of point 1 on figure 1.
This would eliminate an operating pressure band for stable feed and bleed operation and would lead to the conclusion that feed and bleed was not a viable mode of decay heat renoval under the S-SR-1 conditions. This was indeed the case that was observed during test S-SR-1.
Had the leakage from the system been eliminated or compensated for, a stable feed and bleed condition should have been established.
14.
In Semiscale test S-SR-2, the HPI pumps were scaled to t'estinghouse plants with lower head pumps. For this test, the leakage experienced in test S-SR-1 was compensated for with additional HPI flow. The decay power was sinulated as a constant 2 percent, which is representative of decay power between 10 and 20 minutes after shutdown. This decay heat level was perhaps too high sir.ce steam l
l generator boiloff times for Westinghouse plants are around one l
hour, the time at which primary mass loss begins in the absence of any feedwater. The mass and energy balance curves for this test are shown in figure 2.
Also included on this figure are the l
uncertainties in each of the parameters. From this figure it is clear that there is a large uncertainty regardirig whether or not Semiscale, in the S-SR-2 test configuration, would be able to successfully feed and bleed. fluch of the uncertainty is attributed to uncertainties in environmental heat losses from the Semiscale system. EGAG reported that the environmental heat losses accounted for 25% uncertainty in both net core power and average PORY mass flow rate. The high uncertainty in environmantal heat losses is a direct scaling problem associated with Semiscale's high surface area to volume ratio compared to a PWR.
In both Semiscale tests, the actual core power was 78 Kw, while the net core power assumed for simulating core decay heat was 40 Kw. (The additional power was required to compensate for environmental heat losses).
l
- 15. A significant shortcoming of the test was the lack of a pretest calculation. We believe that had such a calculation been performed, it would have highlighted the possibility that a stable feed and bleed situation might not have been achievable for test 5-SR-2.
Subsequent to the test however, EG&G did a post test analysis using the RELAP5 computer code.
In figure 3, the predicted PORY flow is compared to the measured PORY flow. This comparison shows good agreement between the prediction and the data with respect to both the timing and the magnitude.
Figure 4 shows a comparison of the predicted to measured core liquid level. This comparison shows l
that while the RELAP code was able to predict the trends, the l
~
~.
~
actual levels did not compare too well throughout the course of the test. However, as can be seen at time zero, there is an offset between the two curves. Since this represents the time when the vessel is full of liquid, the offset is erroneous. This has been confirmed by EG&G, Idaho. Figure 5, provided by EG&G, Idaho, shows the same curves as figure 4, except the offset error has been corrected.
Frem this figure it can be seen that the code prediction compared well with the experinent out to about 1500 seconds. Beyond that tirm the experiment shews a continuously decreasing level whereas the RELAP calculation shows the level remaining constant. The reason for this difference has been attributed to the differences between the specified and actual flow versus pressure characteristic for the HPI pumps.
Figure 6 shows the comparison between the measured HPI flow and the predicted HPI flor using the higher HPI flow versus pressure curve. As can be seen, beyond about 1600 seconds the predicted flow exceeds the measured flow. Figure 7 shows the difference between the specified and actual HPI flow versus pressure characteristic. The higher HPI flow associated with the test specification was used in the RELAP calculation. This higher flow did not have a significant effect on the liquid level comparisons while de PORV discharge was liquid (up to t 1200 seconds) because the PORY discharge was significantly in excess of the HPI flow and dominated the inventory behavior. Once the PORY discharge transitioned to steam, the discrepancy between the HPI flow used in the RELAP calculation and the actual HPI flow used for the test had a significant effect on inventory, and thus on vessel level. The higher HPI flow used in f
the RELAP calculation resulted in the higher predicted vessel level than was observed in the test. EGaG, Idaho is in the process of rerunning the calculation of test S-SR-2 using the actual HPI flow from the test. These results are expected to confirm that the vessel level discrepancy observed beyond about 1500 seconds can be attributed to the use of a higher HPI flow in the RELAP calculation than was used in the test.
17.
In figures 8 and 9, the measured and predicted mass balances are shown. From these comparisons, it can be seen that the PELAP code Note predictions showed the same basic behavior as the test data.
that subsequent to PORY flow transition from liquid to steem, the steam flow in both the prediction and the test exceeded the HPI flow, resulting in a continuous mass loss. The excessive steam flow results from the core power being high for the test relative to that which would be experienced by an actual plant as discussed above. Had the calculation been perforned prior to the test being run, this continuous nass loss subsequent to PORV flow transition would have been noted and the possibility of core uncovery during S-SR-2 highlighted.
B&W Plant Calculation,
- 18. In order to show how the efficacy of feed and bleed in THI-1 is related to the previous discussions on feed and bleed and Semiscale Test S-SR-2, it is recalled that in the October 15, 1982 NRC Staff
Reponse to Appeal Board Order, Affidavit of Brian W. Sheron and Walton L. Jensen, Jr. ConcerningSemiscaletes't(S-SR-2)Results.
the results of a feed and bleed analysis were presented. This analysis was performed for the Midland plant, which is similar to Till-1, using the RELAPS computer code, the same code used by EG&G to predict the Semiscale test. The major difference between tiidland and THI-1 with respect to the feed and bleed analysis is that the Core power for THI-1 is about 5% higher than Midland. We believe that this difference is small enough so that the results obtained for the Midland calculation will representative of the THI-1 behavior.
In figure 10, the mass balance for the feed and bleed analysis is shown.
In this calculation, only one HPI pump was assumed operable. Moreover, no credit was taken for the PORV, with the bleed function being performed solely by the safety valves. The time at which the HPI flow into the system begins to exceed the steam flow out of the safety valve is indicated as point A.
This is the point of ninimum inventory in the system. At this point the coolant level is still approximately one foot above the top of the core.
If two HPI pumps were assumed, the level above the top of the core would be much higher. These results are j
similar to the results of an analysis preformed earlier by B&W for a plant with an initial core power 10% greater than that of THI-1.
Conclusions 19.
From the previous discussions, we believe that the following conclusions can be drawn:
a.
The ability of any system to feed and bleed, whether it is
Semiscale, LOFT, or a large PWR, is dependent upon the unique characteristics of that system.
In particular, the capability to feed and bleed depends upon the HPI flow characteristics, the safety / relief valve discharge capacity, and the decay heat levels under which feed and bleed must perform. Thus, it is not possible to directly extrapolate the ability or inability of one system to feed and bleed to another system, unless a detailed evaluation of the governing parameters warrant such an extrapolation, b.
Our evaluation of Semiscale test S-SR-2 leads us to conclude that excessive, unanticipated and uncompensated system leakage during the test rendered the test results atypical, and they do not provide any useful information regarding feed and bleed capability.
c.
Our evaluation of Semiscale test 5-SR-2 leads us to conclude that the system configuration for S-SR-2 would not result in acceptable feed and bleed behavior.
Following relief valve flow transition from liquid to steam, the steam discharge still exceeded the HPI flow, resulting in a continuing mass les.
i e
d.
The RELAP5 computer code was shown to be capable of predicting the Semiscele S-SR-2 test results. The observed discrepancy between measured and predicted vessel liquid level after 1500 seconds is attributed to the use of incorrect ilPI flow
characteristics in the RELAP analysis. An analysis is being performed with the correct HPI flow characteristics to confirm this conclusion, e.
The feed and bleed analysis performed by the staff for a plant sinilar to TMI-1 and considered to be applicable to TMI-1 shows that, even assuming only one HPI pump and nc credit for the PORV, successful feed and bleed was predicted with no core uncovery.
5 (b
Brian W. Sheron Subscribegandsworntobeforeme this J P day of November, 1982.
hth.(CLA).
col 44' Notary Publig
(
htths. ll O O' ' 'N[o My Commission expires:
o -
G
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION' BEFORE THE ATOMIC SAFETY AND LICENSING APPEAL BOARD In the Matter of METPOPOLITAN EDISON COMPANY, ET AL.
Docket No. 50-289 (Three Mile Island Nuclear Station, Unit No. 1)
CERTIFICATE OF SERVICE I hereby certify that copies of "NRC STAFF COMMENTS IN RESP 0h3E TO APPEAL BOARD MEMOPANDUM AND ORDER OF NOVEMBER 5,1982" in the above-captioned proceeding have been served on the following by deposit in the United States mail, first class, or, as indicated by an asterisk, by deposit in the Nuclear Regulatory Comission's internal mail system, or, as indicated by double asterisks, by hand-delivery, this 22nd day of November, 1982:
- Gary J. Edles, Chairman Dr. Linda W. Little Atomic Safety & Licensing Appeal Administrative Judge Board 5000 Hermitage Drive U.S. Nuclear Regulatory Comission Raleigh, North Carolina 27612 Washington, DC 20555 George F. Trowbridge, Esq.
- Christine N. Kohl Shaw, Pittman, Potts & Trowbridge Atomic Safety & Licensing Appeal 1800 M Street, NW i
Board Washington, DC 20036
(
U.S. Nuclear Regulatory Comission Washington, DC 20555 Robert Adler, Esq.
505 Executive House
- Dr. John H. Buck P. O. Box 2357 Atomic Safety & Licensing Appeal Harrisburg, PA 17120 Board U.S. Nuclear Regulatory Comission Honorable Mark Cohen Washington, DC 20555 512 D-3 Main Capital Building Harrisburg, PA 17120
- Ivan W. Smith Administrative Judge Ms. Marjorie Aamodt Atomic Safety & Licensing Board R.D. #5 i
l U.S. Nuclear Regulatory Comission Coatesville, PA 19320 Washington, DC 20555 Mr. Thomas Gertisky Dr. Walter H. Jordan Bureau of Radiation Protection l
Administrative Judge Dept. of Environmental Resources 881 W. Outer Drive P. O. Box 2063 Oak Ridge, Tennessee 37830 Harrisburg, PA 17120
Mr. Marvin I. Iewis William S. Jordan, III, Esq.
6504 Bradford Terrace Harmon & Weiss Philsdelphia, PA 19149 1725 I Street, NW Suite 506 Mr. C. W. Smyth, Supervisor Washington, DC 20006 Licensing TMI-1 Three Mile Island Nuclear Station John Levin. Esq.
P. O. Box 480 Pennsylvania Public Utilities Comm.
Middletown, PA 17057 Box 3265 Harrisburg, PA 17120 Ms. Jane Lee R.D. 3; Box 3521 Jordan D. Cunningham, Esq.
Etters, PA 17319 Fox, Farr and Cunningham 2320 North 2nd Street Gail Phelps Harrisburg, PA 17110 ANGRY /THI PIRC 1037 Maclay Street Louise Bradford Harrisburg, PA 17103 Three Mile Island Alert 1011 Green Street Allen R. Carter, Chairman Harrisburg, PA 17102 Joint Legislative Comittee on Energy Post Office Box 142 Ms. Ellyn R. Weiss Suite 513 Hamon & Weiss Senate Gressette Building 1725 I Street, NW Columbia, South Carolina 29202 Suite 506 Washington, DC 20006 Chauncey Kepford Judith Johnsrud Mr. Steven C. Sho11y Environmental Coalition on Nuclear Power Union of Concerned Scientists 433 Orlando Avenue 1346 Connecticut Avenue, NW State College, PA 16801 Dupont Circle Building, Suite 1101 Washington, DC 20036 l
Gary L. Milho111n, Esq.
4412 Greenwich Parkway, NW Ms. Frieda Berryhill, Chairman Washington, DC 20007 Coalition for Nuclear Power Plant Postponement Mr. Henry D. Hukill 2610 Grendon Drive Vice President Wilmington, Delaware 19808 GPU Nuclear Corporation Post Office Box 480
- Judge Reginald L. Gotchy Middletown, PA 17057 Atomic Safety & Licensing Appeal Board Michael McBride, Esq.
U.S. Nuclear Regulatory Comission LeBoeuf, Lamb, Leiby & McRae Washington, DC 20555 Suite 1100 1333 New Hampshire Avenue, NW
- Atomic Safety & Licensing Appeal Washington, DC 20036 Board Panel U.S. Nuclear Regulatory Comission David E. Cole, Esq.
Washington, DC 20555 l
Smith & Smith, P.L.
Riverside Law Center
- Atomic Safety & Licensing Board Panel 2931 N. Front Street U.S. Nuclear Regulatory Comission Harrisburg, PA 17110 Washington, DC 20555 l
1
- Docketing and Service Section Office of the Secretary of the Commissian U.S. Nuclear Regulatory Commission Washington, DC 20555
- Dr. Lawrence R. Quarles Atomic Safety & Licensing Appeal Board U.S. Nuclear Regulatory Commission Washington, DC 20555 Michael W. Maupin, Esquire Hunton & Williams 707 East Main Street P. O. Box 1535 Richmond, VA 23212
% -- =
%Q James M. Cutchin IV Counsel for NRC Staff 4
.