ML20052G128
| ML20052G128 | |
| Person / Time | |
|---|---|
| Site: | Big Rock Point File:Consumers Energy icon.png |
| Issue date: | 05/07/1982 |
| From: | Yeon Kim CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.), NUS CORP. |
| To: | |
| Shared Package | |
| ML20052G119 | List: |
| References | |
| ISSUANCES-OLA, NUDOCS 8205140334 | |
| Download: ML20052G128 (16) | |
Text
,
5/7/82 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of
)
) Docket No. 50-155-OLA CONSUMERS POWER COMPANY
) (Spent Fuel Pool
)
Modification)
(Big Rock Point Nuclear Power Plant)
)
TESTIMONY OF YONG S.
KIM CONCERNING CRITICALITY ANALYSIS O'NEILL CONTENTION IIE-3 4
My name is Yong S. Kim.
I reside at 210 Hillsboro Drive, Silver Spring, Maryland.
Since January 1, 1963, I have been associated with NUS Corporation, an engineering consult-ing firm in Gaithersburg, Maryland.
I have primary responsi-bility with NUS Corporation for reactor physics analyses, including criticality analysis of spent fuel storage racks.
My resume, which is attached to this testimony, sets forth my educational background and work experience.
I Based on my educational background and work experi-ence, I am technically qualified to address the concerns expressed by the Atomic Safety and Licensing Board at page 5 l
l of its February 5, 1982, Memorandum and Order concerning potential criticality of the spent fuel pool (O'Neill Conten-tion IIE-3). I am also qualified to address the concern about l
i the potential supercriticality of the spent fuel pool under g;OSg[Oa PDR T
. certain conditions, expressed by the Board in its Memorandum and Order of February 19, 1982 (at pp. 48-49).
I.
INTRODUCTION In its Memorandum and Order of February 5, 1982, the Licensing Board questioned the adequacy of the criti-cality analysis of the Big Rock Point spent fuel pool performed on behalf of Licensee.
The Board adopted an argument advanced by Intervenors, as follows:
Criticality analysis performed by Dr. Kim is based on a water temperature of 212*F, assuming boiling of the spent fuel pool, with the containment at atmos-pheric pressure.
Even assuming that the containment is at atmospheric pressure (not neces-sarily conservative after a LOCA (loss of cooling accident)), the pressure at the bottom of-the spent fuel pool, due to the hydrostatic load is 28.14 psia.
The boiling temperature at that pressure is 247'F.
Since the effective activity coefficient K is not permitted to exceed 0.95, and since Dr. Kim's calculations reached this maximum, assuming 212*F, it is questionable if the calculations can be considered conservative.
The Board also expressed concern about the effect of possible rack distortions on k-effective (Memorandum at pp. 4-5).
In addition, in its February 19, 1982,
- Memorandum and Order at p.
48, the Board questioned whether the Big Rock Point spent-fuel pool might reach supercriticality if it were to begin boiling.
The purpose of this testimony is to respond to the Board's questions and concerns related to the magnitude of the pool water temperature that will be reached under the pool cooling system failure, the effect of higher water temperature on k-effective of the spent fuel storage racks, the effect of possible rack distortions on k-effective, and the potential of supercriticality through optimum moderation in nuclear fuel storage.
My testimony describes the methods and assumptions that were incorporated in my original analysis of the effect of steam voids (called steam bubbles elsewhere) and water temperature on k-effective.*
I compare the extent of conservatism in the earlier analysis with new calculations based on reduced effective steam voids and increased water temperature.
This revised analysis shows that the reduction in k-effective due to virtual In the nuclear industry, the term k-effective, as used in my testimony, is defined as " effective neutron multiplica-tion factor," and it is equivalent to the term K used in O'Neill Contention IIE-3.
. elimination of the steam voids more than offsets the increase in k-effective due to the higher water tempera-ture.
The thermal-hydraulic conditions discussed by Dr.
D. A. Prelewicz in his testimony serve as the basis for
~
this conclusion.
My testimony also shows that, based on the rack distortions discussed by Mr. R. F. Sacramo in his testimony, the k-effective value used in my analysis would not increase but, on the contrary, would decrease under such conditions.
In addition, my testimony demonstrates that the potential of supercriticality in the Big Rock spent fuel pool is insignificant because it would require boiling off of the. pool water below the level of the racks.
Since the Board states that my calculations reached the maximum limit k-effective of 0.95, I feel it is important to clarify the fact that the primary purpose of my original analysis was to determine the limiting fuel design without exceeding the maximum k-effective of 0.95.
Therefore, the k-effective value of 0.95 was used as the upper limit in seeking the highest enrichment (U-235 weight per cent in fuel) for the limit design.
In other words, the value of enrichment was chosen so that when the effects of dimensional and material tolerances, pool water temperature, calculational biases, and potential
. accident conditions on k-effective were added to the k-effective of the nominal condition, the final (or-maximum) k-effective of the storage system just reached 0.95.
The limiting fuel design is a modified version of G-3 fuel and different-from any of the existing fuel at the Big Rock Point reactor.
All the existing fuel at the Big Rock Point reactor is much less reactive than this limiting fuel design, as shown in Consumers Power Com-pany, Big Rock Point Plant, Spent Fuel Rack Addition, Consolidated Environmental Impact Evaluation and Descrip-tion and Safety Analysis, Section 2.4.4, p.
2-15.
Therefore, the " limiting design" fuel is more reactive than any of the fuel types currently stored in the Big.
Rock Point spent fuel pool.
Because the revised analysis presented in this testimony does not change any assump-tions about the limiting fuel design, but only assump-tions about water temperature and steam void fractions, this same conservatism is inherent in the revised k-effective calculated here.
II.
WATER BOILING TEMPERATURE In my original calculation of k-effective, a spent fuel pool water boiling temperature of 212*F was used for
. the accident condition of failure of the pool cooling system.
This water temperature was used because it was the saturation (boiling) temperature corresponding to the ambient pressure (14.7 psia) of the pool surface, and it had been an industry practice to use 212*F as the boiling temperature when we considered the formation of small
. steam voids in the pool ~.
For most spent fuel pools, the use of 212*F for the criticality analysis is conservative because k-effective decreases with increasing temperature.
My original analysis showed, however, that for the Big Rock Point pool, k-effective increases with increasing temperature.
When I performed that analysis, I had not been provided with Dr. Prelewicz's conclusion that the temperature in the Big Rock Point pool varies from 212*F at the bottom of the fuel rack to 237*F near the top of the rack.
Because of this result, the figure of 212*F was in fact non-conservative, -as noted by the Board.
Therefore, an analysis based upon more realistic thermal-hydraulic conditions provided in Dr. Prelewicz's testimony has been performed and is discussed below.
According to Dr. Prelewicz's testimony, the maximum boiling temperature of water at the top of the fuel assembly, assuming failure of the cooling system, is
. 237*F.
He states that the saturation temperature corresponding to the hydrostatic pressure at the bottom-of fuel is 243*F (instead of 247*F as indicated by O'Neill Contention IIE-3), but that this temperature cannot be attained due to the water flow patterns set up by natural circulation.
He also states that the inlet temperature of the coolant at the bottom of the fuel is 212*F.
For the calculation of k-effective, it is appropriate to take the average temperature along the length of the fuel, that is, the average between 237*F and 212*F, which is 224.5*F.
This increase in water temperature from 212*F, which was used in my original analysis, to 224.5'F (a difference of 12.5"F), results in an increase of k-effective of 0.0014 based on the analysis given in NUS File G-RA-12 Addendum No. 1.
III. STEAM VOID EFFECTIVE In the original calculation of k-effective, a steam void volume fraction of 0.206 was provided by Dr. Prelewicz.
I assumed that this fraction was uniformly distributed along the entire height of the fuel assembly.
This assumption produced the largest k-effective consistent with Dr. Prelewicz's maximum steam void volume fraction.
However, this assumption was excessively conservative l
l l
t
. in relation to the actual void distribution shown by Dr.
Prelewicz's analysis.
Based on a coolant inlet temperature of 212*F under the partial pool boiling condition, Dr. Prelewicz's original analysis, which is still valid, shows that the maximum steam void volume fraction is 0.206 at the top of
!~
the 70-inch active fuel height.
Thus, the steam voids exist only for the upper 0.276 inches of the fuel, and the average (effective) temperature over the active fuel height is 224.5'F.
Since the void fraction will vary along 0.276 inches of the boiling length from zero at the start of boiling position to 0.206 at the top, the average (effective) void fraction over the entire active fuel height is calculated to be only 0.00041.
This volume fraction (0.00041) yields an increase in k-effec-i tive of only 0.00001, which is effectively'zero, compared l
l to the increase of 0.0044 obtained in the original analysis.
Therefore, there is a net decrease in k-effec-tive of 0.0044 when the more realistic average void fraction is used.
IV.
COMBINED EFFECT OF STEAM VOID AND WATER TEMPERATURE The effects of the revised steam void volume frac-tion and the revised average water temperature yield the following net change in k-effective:
L
L.
Original k-effective:
+ 0.9500 Temperature:
+ 0.0014 Steam Voids:
- 0.0044-Net Effect:
- 0.0030
- 0.0030 Revised k-effective:
+ 0.9470 Thus, the revised k-effective, 0.9470, is less than 0.95.
V.
POSSIBLE DISTORTION OF RACKS According to Mr. Sacramo's testimony, the drop of a fuel assembly on top of a fuel rack is not expected to cause
..y_ distortion of the rack along the length of the stored fuel assembly.
Therefore, such an accident would not change k-effective.
With regard to temperature, Mr. Sacramo states that an increase in pool water temperature from 70*F to 237'F will produce an increase in the center-to-center spacing (pitch) of the racks of 0.015 inches over the nominal value of 9.0 inches.
This increase in the pitch will cause a decrease in k-effective of approximately 0.0020 based on the results of analysis given in NUS File G-RA-12 Addendum No.
1.
Interpolating this result to the
. average temperature of 224.5*F under the cooling system failure condition, we get a decrease of 0.0018 in k-effective.
Therefore, the heating of the racks wil' not result in any increase in the revised value of k-effective calculated above.
For conservatism, this reduction is neglected.
VI.
SUPERCRITICALITY OF SPENT FUEL STORAGE SYSTEN On page 48 of its February 19, 1982 Memorandum and Order, the Board questioned whether the Big Rock Point spent fuel pool might reach a supercritical condition if it were to begin boiling.
The Board based this question on its analysis of an article published.in Nuclear Technology (pp. 251-260, Volume 48, 1980).
The article shows that for unpoisoned stainless steel storage racks, k-effective increases sharply and may exceed 1.0 (super-critical condition) at very low water densities.
More-over, the authors state that the magnitude of this supercriticality depends in part on the computer codes and methodology used in the calculation.
The increase in k-effective at very low densities in unpoisoned stainless steel storage racks such as those for the Big Rock Point spent fuel pool has been recog-nized not only in this article (see figure 12 cf the L.
. article), but also in other articles, such as F.
M.
Alcorn, " Criticality Evaluation of Low-Density Moderation in PWR Fuel Storage," pp. 416-417, ANS Transactions (Nov., 1977).
Supercriticality is shown to be possible under conditions where the pool water is replaced by a mist, foam, or some other form of very low density water.
In a water storage pool, such as at Big Rock Point, this condition could only occur after nearly all of the pool water is boiled.away (i.e., at least below the level of i
the storage racks).
The exact water densities at which the system becomes supercritical depends on the material and geometry of storage racks, the type and enrichment of fuel to be stored, and the separation between the rack cell units.
For example, the stainless steel racks studied in the article cited by the Board used a 15 x 15 fuel assembly with 4.0 weight per cent U-235, different from the 11 x 11 fuel assembly with 3.80 per cent U-235 for the Big Rock Point racks.
Figure 13 of the article shows that for stainless steel racks of the Big Rock i
Point type (0.250-inch thickness), the supercritical condition never exists even for very low water densities, the maximum k-effective being less than 0.97.
l l
i l
l l
. No quantitative analysis with respect to super-criticality has been performed for the Big Rock Point spent fuel pool.
Normally, the supercriticality analysis under optimum moderation is required and performed for new fuel storage racks under dry storage conditions.
No such analysis is necessary for wet storage conditions such as at Big Rock Point.
It is established that the point of the maximum k-effective (whether above or below 1.0) occurs generally at water densities below 0.20 g/cc (about 20% of full density).
The water density for the Big Rock Point spent fuel pool is 0.953 g/cc-at 224.5*F.
This value is derived from Dr. Prelewicz's testimony which shows that the maximum effective (average) steam void fraction that will result from a cooling system failure is only 0.00041.
The void conditions stated above assume that the pool water level will be maintained.
As I pointed out earlier, supercriticality could occur in a water storago l
pool only after the pool water boiled away to at least-below the level of the storage racks. This condition is very unlikely at Big Rock Point in view of the ability to l
remotely supply make-up water and the long time period I
(approximately 700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br />) required to boil away all of the water in the pool.
(See Testimony of David P.
. Blanchard Concerning Christa-Maria Contention 8 and O'Neill Contention IIIE-2 at p.
8). Therefore, the supercritical condition will not occur in the Big Rock Point spent fuel pool under the assumed accident condition.
With regard to the computer codes and methodology employed in the supercriticality calculation, differences in calculated k-effective among various methods are normally much more pronounced in the very low water-density region, as the authors have shown.
- However, these differences are comparatively small near the full water-density region that would prevail in the Big Rock Point spent fuel pool, even after the cooling system failure.
Differences among the various methods are due-to differing levels of complexity among the methods.
The results of the simpler methods are normally corrected by comparing a limited number of cases with the results of the more sophisticated methods.
Furthermore, our Big Rock Point rack criticality analysis results are i
corrected in this manner.
Thus the problem of computer codes and methodology employed in criticality calculations does not exist in our analysis, i
L
- VII. CONCLUSIONS The maximum k-effective calculated for the same conditions as the prior analysis except for thermal-hydraulic input is 0.9470.
Any reduction in k-effective due to possible rack distortions under higher temperature is neglected in this result to be conservative.
The value 0.9470 is below 0.95 and satisfies the NRC criteria regarding k-effective for spent fuel pools.
The other conservative assumptions such as fresh fuel, radial infinity, axial infinity, and others used in the original analysis are still applicable to this result.
Also, the small steam voids which could be formed by the cooling system failure will not result in a supercritical condition.
I 1
i l
l l
j YONG S. KIM EDUCATION l
Catholic University of America, Ph.D.,1970 Massachusetts Institute of Technology, M.S., Nuclear Engineering,1961 University of Wisconsin,8.S., Chemical Engineering,1958 i
ll REGISTRATION Department of Defense Certified Fallout Shelter Analyst Registered Professional Engineer. State of Maryland,1977 Registered Professional Engirmar, State of California,1977 i
f EXPERIENCE l
NUS CORPORATION,1963-Present l
Internuclear Company, 1961-1962 M.I.T. Department of Nuclear Engineering, 1958-1960 NUS - Coordinates and participates in the nuclear criticality analysis activities for the design and licensing of poisoned and unpoisoned spent fuel storage racks. Performed nuclear analysis of advanced once through BWR design incorporating solid moderator as part of ACDA/ DOE uranium i
resource utilization program. Participated in nuclear criticality analysis and technical review of cavern growth related to human intrusion of nuclear waste repository in a domed salt formation.
i As Acting Manager of Nuclear Analysis, was responsible for technical work performed by nuclear analysis staff, including core design evaluation and analysis, core follow of operating reactors, 1
computer code development, and training relative to nuclear, thermal hydraulic, and mechanical behavior of fuelin nuclear reactors. In charge of reactor physics computer program development and applications. Adapted numerous nuclear and engineering computer programs to different types of digital computers ranging from the large scientific / engineering computers to minicompu-ters, and performed improvement and modification of existing large computer codes. Managed and participated in the training of nuclear utility engineers in the area of in-core nuclear and thermal-hydraulic analyses and the use and installation of related computer codes Involved with bid evaluation of various commercial power reactors with regard to nuclear design. Performed a complete nuclear analysis of the shuffled core of a U.S. nuclear merchant ship, NS Savannah.
Performed nuclear design analysis, shielding calculations, and safety analysis of military power reactors, including PM 1, PM-2A, PM-3A, SM-1, SM-1 A, and MH-1 A.
Previously developed computer programs for calculation of radiation doses due to radioactive release from nuclear power plants of nuclear rocket launch sites under normal or accident conditions. Analyzed physics design of the Japan Material Testing Reactor, and calculated shield-ing requirements for proposed radiation exposure facility at the AFRRl Reactor. Evaluated and compared n uclear fuel costs of proposed power reactors to assist electric utilities in the selection of reactor types. Instructed in NUS Fuel Management Workshop courses.
Developed and programmed NULOC-2 Control Data 6600 code for multicompartment loss-of-coolant accident analysis; NUTRIX and NUSIM, Control Data-6600 codes for three-dimensional physics analysis of operating reactors; NADAT-1 Honeywell DDP-516 assemblylanguage code for g
transfer of radiation data WINDIF, wind diffusion program for evaluation of reactor sites and air pollution for IBM-7090 and Control Data 3600; MOREDO, Control Data 3600 code to compute w
lii external whole body gamma dose in any population due to reentering nuclear rockets (for Space A
Nuclear Propulsion Office); NEEP, program for 18M-7090 and Control Data-3600 to compute l
l l
J YONG S. KIM Page Two effective energy of radioactive isotopes deposited in internal body organs (for SNPO); NURSE, nuclear rocket safety evaluation program for Control Data-3600 (for SNPO); FUELCCST-1, fuel cycle cost program for comparative study of fuel costs of different power reactors for 18M-7090 and Control Data-3600: EXGAM, code to predict integrated gamma dose for an airborne release of radioactivity for Control Data-3600.
Internuclear - Performed detailed nuclear calculations, including lifetime evaluation for a fully enriched boiling water reactor. Performed nuclear calculations for the University of Missouri Research Reactor. Analyzed and programmed one-dimensional transport problem with mono-directional sources and an isotropic scattering for radiative transfer applications for IBM-7090 (ISOLATE). Performed heat transfer analysis of small nuclear reactors for activation analysis i
applications.
M.I.T. - As laboratory instructor, supervised graduate students performing experiments on reactor physics, radiation detection, and shielding. Made experimental measurement of degrada-tion rate of lucite physical properties in a reactor core; correlated degrad9 tion rate with radiation dose as part of the organic loop pxperiment at the M.I.T. Reactor. Designed in-pile radiation monitor j
instruments for this experiment.
4 MEMBERSHIPS American Nuclear Society Society of Sigma Xi I
PUBUCATIONS
\\
"High Density Spent Fuel Storage Racks Design Analysis Report, Kewaunee Nuclear Plant, j
Criticality Analysis," NUS-1931, Part D, August 1977.
" Core Analysis Procedures Manual," CD-NA-76782, December 1976.
l l
"NUMICE 2 A Spectrum Dependent Non Spatial Cell Depletion Code," NUS-894, Revision 1, March 1976.
"NUSIM-3, A Digital Computer Program forThree Dimensional Nodal Reat. tor Simulation," March 1975.
"NUCELL - Cell Spectrum and Depletion Code Based on MUFT and THERMOS," May 1975.
"CYREP-lil, in-Core Fuel Management Code," May 1975.
"NULOC 2, NUS Multi Compartment Pressure-Temperature-Temperature History Program in Response to a Loss-of Coolant Accident," NUS-1160, March 1974.
"NUCONTEMPT, NUS Version of CONTEMPT-PS for Prediction of Pressure-Temperature Response l
to a Loss-of Coolant Accident," NUS-1164, March 1974.
i "NUTRIX, A Digital Computer Program for Three Dimensional Analysis of Time-Dependent Operat-ing Reactor," NUS 657, March 1970, and Revision 1, March 1972.
"FLYASH-il, A CDC-6600 Computer Code to Calculate Time-Dependent Isotopic inventory and i
l Radioactive Disintegration Rates in a Nuclear Reactor," NUS-878. February 1972.
l l
"NUFLOW-1, A Three Dimensional Nodal Core Analysis Computer Program with Internally Calcu-j lated Core Flow Distribution," NUS-857, January 1972.
l
-