ML20071P173
| ML20071P173 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 11/01/1982 |
| From: | Clare G, Deitrich L, Strawbridge L JOINT APPLICANTS - CLINCH RIVER BREEDER REACTOR, WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20071P101 | List: |
| References | |
| NUDOCS 8211020374 | |
| Download: ML20071P173 (44) | |
Text
'
l l
,,3msuco DOCKETED UStiRC
'82 N0'l -1 P 3 :38 UNITED STATES OF AMERICA
_ _ , "r :2 0:S
NUCLEAR REGULATORY COMMISSION'6I' f '-
In the Matter of )
UNITED STATES DEPARTMENT OF ENERGY )
PROJECT MANAGEMENT CORPORATION ) Docket No. 50-537 TENNESSEE VALLEY AUTHORITY )
(Clinch River Breeder Reactor Plant) )
APPLICANTS' TESTIMONY CONCERNING NRDC CONTENTIONS 2d), 2 f) , 2g),
2h), 3c) and 3d) (Environmental Effects) and 5b)
Dated: November 1, 1982
~ r l
() .l5' i
' 1 8211020374 821101 PDR ADOCK 05000537
[ T PDR l
Q.l. Please state your names and affiliations.
A.1. George H. Clare, Manager, Licensing, Westinghouse Advanced Reactors Division. Lee E. Strawbridge, Manager, Nuclear Safety and Licensing, Westinghouse Advanced Reactors Division. L. Walter Deitrich, Associate Director, Reactor Analysis and Safety Division, Argonne National Laboratory.
I l
Q.2. Have you prepared statements of your professional qualifications?
A.2. Yes. Copies are attached to this testimony.
Q.3. What subject matter does your testimony address?
A.3. This testimony addresses the environmental effects of CRBRP accident analyses. This issue is defined in NRDC Contentions 2d), 2f) , 2g), 2h) , 3c) and 3d) (Environmental
. 1 Effects) and 5b). Specifically, NRDC alleges that
- 2. The analyses of CDAs and their consequences by Applicants and Staff are inadequate for purposes of licensing the CRBR, performing the NEPA i cost / benefit analysis, or demonstrating that the l radiological source term for CRBRP would result in potential hazards not exceeded by those from any accident considered credible, as required by ,
, 10 CFR 100.1(a), fn. 1.
d) Neither Applicants nor Staff have demonstrated that the design of the containment is adequate to reduce calculated offsite doses to an acceptable level.
1 This testimony addresses the basis for the selection of the core accident cases that are assessed in separate testimony addressing Contention 5b). See Q/A 40, 2
f) Applicants have not established that the computer models (including computer codes) referenced in Applicants' CDA safety analysis reports, including the PSAR, and referenced in the Staff CDA safety analyses are valid. The models and computer codes used in the PSAR and the Staff safety analyses of CDAs and their consequences have not been adequately documented, verified or validated by comparison with applicable experimental data. Applicants' and Staf f's safety analyses do not establish that the models accurately represent the physical phenomena and principles which control the response of CRBR to CDAs.
g) Neither Applicants nor Staff have established that the input data and assumptions for the computer models and codes are adequately documented or verified.
h) Since neither Applicants nor Staff have established that the models, computer codes, input data and assumptions are adequately documented, verified and validated, they have also been unable to establish the energetics of a CDA and thus have also not established the adequacy of the containment of the source term for post accident radiological analysis.
- 3. Ne ther Applicants nor Staff have given sufficient attention to CRBR accidents other than the DBAs for the following reasons:
c) Accidents associated with core meltthrough following loss of core geometry and sodium-concrete interactions have not been adequately analyzed.
d) Neither Applicants nor Staff have adequately identified and analyzed the ways in which human error can initiate, exacerbate, or interfere with the mitigation of CRBR accidents.
- 5. Neither Applicants nor Staff have established that the site selected for the CRBR provides
. adequate protection for public health and
! safety, the environment, national security, and i national energy supplies; and an alternative site would be preferable for the following reasons:
3
F b) Since the gaseous diffusion plant, other ,
proposed energy fuel cycle facilities, the j Y-12 plant and the Oak Ridge National Laboratory are in close proximity to the site an Accident at the CRBR could result in the long term evacuation of those facilities. Long term evacuation of those facilities would result in unacceptable risks to the national security and the national energy supply.
Q.4. What fundamental core conditions are most important in considering the environmental effects of accidents?
A.4. In the Applicants' Testimony on Contentions 1, 2 and 3 (Exhibit 1), it was shown that reactor accidents involve either:
o Excessive heat generation, or o Reduced heat removal.
O.S. What design, features are important to prevention of these two core conditions?
A.S. A discussion of design features which can prevent progression of these two conditions beyond the design base and preclude initiation of a hypothetical core
,' disruptive accident (HCDA) in a reactor of the general l
size and type of the CRBRP was presented in Applicants' Exhibit 1. These features include:
o Redundant, diverse reactor shutdown systems (RSS).
o Redundant, diverse shutdown heat removal systems (SHRS).
4 i
o Means to prevent inlet pipe rupture.
o Means to maintain a balance between individual subassembly heat generation and heat removal.
Q.6. Which of these features are of primary interest in the NRC Staff's estimates of the environmental effects of accidents in Appendix J of the Draft Supplement to the Final Environmental Statement (DSFES)?
A.6. The RSS and the SHRS are of primary interest. The Appendix J analysis makes estimates of the risks associated with HCDAs. The two systems which have the greatest influence on the Staff's Appendix J estimates of the frequency of progression to HCDA conditions are the RSS and the SHRS.
Q.7. What physical characteristics of LMFBRs are of primary importance in assessing the capability of the RSS to prevent excessive heat generation and progression to HCDA
- conditions?
A.7. The principal means of preventing HCDAs due to excessive
- r heat goneration is the RSS. The RSS must be able to provide a timely response to prevent excessive heat generation resulting from any credible reactivity incertion. The time response characteristics required of the RSS are strongly influenced by the kinetics of the l reactor, i.e., its response to reactivity insertions, i
l I
5
l e
Q.8. What are the reactor kinetics characteristics of an LMFBR and how do they compare to those of an LWR? )
A.8. Heat generation in a reactor is determined by the mass of l
fissile material present, the fission cross-section, and the neutron flux. For a reactor of the general size and type of CRBRP, control of the reactor power is accomplished by control of the neutron flux. The fundamental neutron balance states that:
[ Rate of change of neutron density] =
[ Net rate of neutron production in fission reactions]
-[ Rate of neutron loss by leakage and non-fission absorption]
For a critical reactor, the rate of change of neutron density is zero. Neutron production balances losses.
Withdrawal of a control rod from the reactor core will reduce neutron losses by non-fission absorption, so the neutron density (and reactor power) will increase, and vice versa. The rate at which changes in reactor power occur is determined by the rate and magnitude of change in non-fission absorption and by the kinetics parameters of the reactor under consideration.
A change in the balance between neutron production, losses, and absorption is manifest in a change in the effective multiplication factor, i.e., the ratio of the neutron density in one generation to that of the preceding generation. The reactivity, rho, is defined 6
I
\ . _ ___
in terms of the effective multiplication factor, k,ff, as rho = (k,ff - 1)/k,ff For a critical reactor, k,ff is one and reactivity is zero.
. Most neutrons are produced essentially instantaneously in the fission process. These neutrons are called " prompt neutrons." Prompt neutrons slow down from the energy at which they were produced to the energy at which they cause new fissions. This slowing down, along with diffusion to a fissile nucleus, takes a short time called the prompt neutron lifetime. Typical prompt neutron lifetimes are the order of 10 -5 seconds for LWRs and 10 -7 seconds for LMFBRs.
However a small, but important, fraction of the total number of neutrons resulting from fission appears as the result of radioactive decay of certain fission products, with half-lives ranging from a few tenths of a second to tens of seconds. These half-lives for " delayed neutrons"
~ r are nearly the same for LWRs and LMFBRs. It is these delayed neutrons which determine the reactor kinetics behavior under all credible operating and accident conditions. The effective fraction of total neutrons which appear as delayed neutrons depends on the material in which
, fissions occur (primarily 235 0 in an LWR, 'Pu and 238 U in an LMFBR), and to a minor extent on the reactor design.
i 7
_ - . _ _ . . - . _ - - _ . - _ _ . . . - _ . - - - - - - - - _ - - _ .- _ - - = . . . . . - .
Typical values are about 0.0065 for LWRs and about half that value for a reactor of the general size and type of CRBRP.
A critical reactor depends on both prompt and delayed neutrons to sustain the chain reaction. Thus, it is said to be " delayed critical." Should the reactivity become high enough that the reactor is critical on prompt neutrons alone, it is said to be " prompt critical." The latter condition is defined as rho = beta where beta is the effective delayed neutron fraction. It is convenient to normalize reactivity to the delayed neutron fraction, thereby introducing the " dollar" of reactivity, such that 15 of reactivity represents prompt criticality. One cent of reactivity is 0.01 dollar.
The equations relating reactivity and reactor power are
. well known. Figure 1 shows the approximate response of
/
reactor power to a step change in reactivity not close to IS, not considering any reactivity feedback (the conservatively estimated maximum design basis step reactivity insertion is 60g).
8
I I
j i I
g1 *-- A : +-- B ;!= C l
. s A. Stable initial Power g B. Region of Transient Reactor Period C. Power Rise on Stable Reactor Period I
Time
-Step Change in Reactivity
~
Figure 1.
It is seen that there is an initial rapid power increase which quickly slows to a power rise on a stable reactor 4
period. (The reactor period is the time for power to
~
increase by a factor of "e".) The transient reactor power and its r, ate of change are determined principally by reactivity and delayed neutron lifetime, and only in a secondary way by prompt neutron lifetime. The stable period and magnitude of power rise are essentially the same
- , ~
for LWRs and LMFBRs for reactivities not close to 1$.
Thus, the LMFBR, even with its shorter prompt neutron
! lifetime compared to the LWR, is not appreciably different in its control characteristics from an LWR.
Q.9. How do reactivity feedbacks affect LMFBR reactor kinetics?
A.9. The preceding discussion of reactor kinetics did not include any consideration of reactivity feedbacks 9
L_ . - _ - - - - - - _ _ . . - _ .
associated with change in reactor temperatures. In practice, such feedbacks are important in reactor control.
In a fast reactor, the most important of these feedback )
mechanisms in controlling the power rise associated with reactivity transients is the Doppler coefficient. The Doppler coefficient reflects a net increase in the proportion of neutrons absorbed without causing fission to those causing fission as the temperature increases. The decrease in reactivity due to Doppler feedback is a prompt effect; that is, no time delays associated with heat transfer or material motion are involved. Thus, Doppler l feedback is effective in attenuating power transients associated with large reactivity insertions, even including prompt critical conditions. The effectiveness of the Doppler coefficient in a fast reactor was demonstrated by experime ts in the SEFOR reactor.
Another important prompt feedback mechanism is fuel expansion. Fuel expansion decreases the fuel density which is reflected as a decrease in the fission cross-section and, consequently, a decrease in reactivity.
Other reactivity feedback mechanisms, such as coolant density changes, can influence the reactor heat generation.
However, these effects are not prompt in time, since a heat transfer delay is involved. Thus, such feedbacks are not i of primary importance in determining the speed of response 1
requirements for the RSS.
10
Q.10. What conclusions do you draw concerning the feasibility of designing the CRBRP RSS to prevent excessive heat generation?
A.10. Although an LMFBR of the general size and type of CRBRP will have a shorter prompt neutron lifetime and smaller delayed neutron fraction than would a typical LWR, the ;
~
control response requirements of the two reactor types are similar. This conclusion follows because the reactor kinetics for the range of reactivity insertions encountered in Design Basis Accidents are principally dependent on delayed neutron lifetimes and reactivity (normalized to the delayed neutron fraction). Thus, no extraordinary shutdown system response characteristics are required, and LWR technology is applicable. Furthermore, prompt reactivity
~
feedbacks from the Doppler ef fect and f uel expansion ;
I mitigate power transients associated with reactivity '
insertions. Thus, the short prompt neutron lifetime is of no practical significance in reactor control. As was demonstrated in Section 3.3 of Exhibit 1, it is feasible to
_' e provide shutdown systems, based on LWR technology, which assure a high likelihood of reactor shutdown. Such systems with adequate time response characteristics are clearly j within the state of technology.
4 11 l - - - - - - _ - _ _. - _ -
Q.11. What conclusions have you drawn concerning the NRC Staf f's estimates of the frequency of progression to HCDA conditions as a result of failure of the RSS on demand?
A.11. The NRC Staff's estimates of the frequency of failure of the RSS on demand and the resultant progression to HCDA conditions are based upon experience with LWR systems. The i Staff recognized that CRBRP has two shutdown systems, but gave only limited credit for the presence of the second system.
Based on the similarity of the shutdown system requirements, the CRBRP RSS can use technology similar to that used in LWRs, and the likelihood of failure of a l single shutdown system in CRBRP should be similar to that
. in an LWR. However, since two redundant, diverse, independent fast acting shutdown systems have been provided in CRBRP, rather than one such system as in an LWR, the likelihood of failure of the RSS should be substantially less in CRBRP than in an LWR. On this basis, the Staff's Appendix J estimates of shutdown system failure frequency
~ r and the resulting likelihood of HCDA conditions are conservative.
12
~
Q.12. What design features are of primary importance to prevention of reduced heat removal and progression to HCDA conditions?
A.12. The Reactor Shutdown Systems are designed to automatically shut down the reactor if reduced heat removal occurs while the reactor is at power (Exhibit 1, Section 3.3). The Shutdown Heat Removal Systems (SHRS) are designed to remove reactor decay heat and reestablish the balance between heat generation and heat removal (Exhibit 1, Section 3.3) .
Q.13. What SHRS general design characteristics and available experience support the NRC Staf f's Appendix J estimates of the frequency of SHRS failure?
A.13. The SHRS includes redundancy, diversity and independence to provide protection against random and common-cause failures. This is consistent with the approach used in the design of systems used to remove reactor decay heat in Light Water Reactor (LWR) plants. This supports the judgment by the NRC Staff that failure of the CRBRP SHRS
! would result in core degradation at a frequency similar to that estimated for Pressurized Water Reactor (PWR) plant systems (DSFES, Appendix J, Page J-3).
13
i Q.14. Are there additional characteristics which can enhance the capability of LMFBR SHRSs relative to LWRs?
A.14. Yes. There are several characteristics of sodium coolant that enhance the capability of LMFBRs for decay heat removal. Sodium has a high boiling temperature (1600 F) compared to the normal operating temperatures (1000 0F hot leg temperature) . The large margin to boiling assures that (1) the primary coolant system will not be pressurized by sodium vapor, and (2) a large temperature increase can be accommodated in the primary coolant without boiling in the core which could degrade heat transfer. Sodium has a high thermal conductivity: approximately 30 Btu /hr-ft OF rs 0.3 Btu /hr-ft OF for water. The high thermal conductivity assures effective heat transfer even at low sodium flow
~
rates. Although the specific heat of sodium is less than that of water (0.3 Btu /lb OF Es 1 Btu /lb OF for water), the large sodium inventory of the primary and intermediate heat transport systems (approximately 3 million pounds) provides 0 .
a large heat capacity (approximately 1 million Btu / F)
,- These thermal properties combine to enhance SHRS capability in three ways: (1) the high boiling temperature allows operation at atmospheric pressure and thus passive mitigation of primary coolant leaks; (2) the sodium coolant and systems characteristics facilitate shutdown heat ~
removal using only the thermal driving head to circulate coolant, i.e. , natural circulation; and (3) the large 14
system heat capacity and large margin to boiling provide a long time after reactor shutdown before shutdown heat removal is necessary.
Q.15. How does the boiling temperature of sodium enable passive maintenance of primary coolant inventory?
A.15. The large margin to boiling assures that the primary coolant system will not be pressurized by sodium vapor as a result of normal plant operation or a DBA. The only pressure sources in the primary coolant system are the static head and pump head. The primary coolant pump main motors are tripped when the Reactor Shutdown Systems (RSS) are tripped assuring that the normal pump head (approximately 150 psig) is relieved when the reactor is shut down. The only pressure sources during SHRS operation are the static head and the head from the primary coolant pumps operating on pony motors (approximately 5 feet maximum).
This low pressure allows the use of a totally passive
~
Guard
,- approach to maintaining primary coolant inventory.
vessels are provided around the primary coolant system components and elevated piping is used between the components. The upper lips of the guard vessels are high enough and the volume between each component and its guard vessel is small enough so that no leak from the primary coolant boundary could result in loss of so much sodium that the core or the reactor vessel outlet nozzles would be 15
uncovered. Thus, no active components (such as pumps or valves) are required to function to maintain primary coolant inventory. The guard vessel-elevated piping concept is illustrated in Figure 2.
~
Pump IHX Reactor Vessel D Flow- p
/ G
-q Guard Cb Guard Core /d Vessels Vesselj
- V N}
Figure 2. Guard Vessels and Elevated Piping Assure Primary Coolant inventory is Maintained.
Q.16. How does this approach to maintaining reactor coolant inventory enhance SHRS capability relative to LWRs?
/i.16. This passive approach to maintaining reactor coolant inventory in the event of a primary coolant leak can be functionally compared to the active Emergency Core Cooling Systems (ECCS) used in LWR plants. These passive features, which'take advantage of the physical characteristics of sodium, provide an inherently reliable means of enhancing the capability of the SHRS.
16
Q.17. How do the thermal characteristics of the sodium coolant and systems characteristics enable natural circulation?
A.17. The high thermal conductivity of sodium and the large ,
l margin to boiling are desirable thermal characteristics l
~
that allow the use of low flow rates (as low as 3 percent of normal full flow) to remove decay heat from the core
~
following coastdown of the primary coolant pumps. When heated in the core, sodium expands, becoming less dense; when cooled in an Intermediate Heat Exchanger (IHX), sodium contracts, becoming more dense. By locating the IHXs higher than the core, this expansion and contraction can be used to establish a natural thermal driving head which would circulate sodium through the core and primary coolant system, i.e., natural circulation. Natural circulation can remove all decay heat from the core even if all three primary pony motors fail to operate for decay heat removal.
Similarly, arrangement of the plant so that the steam generators are higher than the IHXs can provide sodium natural circulation in the Intermediate Heat Transport System (IHTS) to remove the decay heat from the primary
{ sodium coolant.
i l
The same principle can be used to take advantage of the fact that heating water yields steam which will rise from the steam generator forcing natural circulation between the steam generators and the steam drums. Similarly, rising steam and f alling condensate will transport heat to the i
17
Protected Air Cooled Condenser (PACCs) where heated air will rise to naturally cool the PACCs.
As shown in Figure 3, the components in CRBRP are arranged
- to provide natural circulation all the way from the core to the PACCs.
Protected
_ _ Air Cooled
-Condenser t
J.s I i
o
{.
A Steam Drum l I Reactor Vessel-- J- -Steam Generator 7
1HX -
Care = -
g_ v
- v. .
Figure 3. Elevation Differences in Major Components Provide a Natural Circulation Capability.
~
.- e l
The capability to remove heat by natural circulation to the PACCs supplements heat removal using power relief valves and a turbine-driven auxiliary feedwater pump (TDAFWP) which are also included in CRBRP.
18
(
Q.18. How does natural circulation enhance SHRS capability relative to LWRs?
A.18. Arrangement of the plant to take advantage of the desirable inherent heat removal characteristics of sodium provides for SHRS functioning with loss of offsite power concurrent with failure of all of the emergency diesel generators.
Further, the natural cooling capability of the PACCs provides the SHRS function even if the TDAFWP were to fail one hour after reactor shutdown. This gives CRBRP protection against SHRS failure due to loss of all electric power and loss of the TDAFWP. SHRS failure due to loss of all electric power and loss of the TDAFWP is a principal ,
failure mode considered by the NRC Staff in judging the reliability of SHRS based upon LWR experience (DSFES, Appendix J, Page J-4). Thus, natural circulation providet a passive, inherently reliable means for protection against SHRS failure and an enhanced SHRS capability relative to LWRs.
Q.19. How does the large system heat capacity enable maintenance of a large margin to sodium boiling?
A.19. The sodium coolant in the primary and intermediate heat transport systems has sufficient heat capacity to store 100 MW hr .of heat while increasing the bulk sodium temperature by only 300 F. Increasing the sodium temperature 300 F from its normal bulk temperature (approximately 850 0 F) would not result in sodium boiling and would not result in 19
inadequate core cooling or failure of the primary coolant boundary. As a result, a large amoJnt of reactor decay heat can be stored in the sodium coolant itself. Even if l one assumes a complete loss of heat sink (LOHS), all of the
. decay heat produced in the first 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after reactor shutdown (about 100 MW hr) could be stored this way. If the reactor has been shutdown for a day, all the decay heat produced in the next 4 days could be stored.
Q.20. How does the large system heat capacity enhance SHRS capability relative to LWRs?
A.20. Because heat can be stored in the primary and intermediate sodium, the assumed f ailure of the SHRS to transport heat to an ultimate heat sink (called Loss of Heat Sink - LOHS) would not result in rapid progression to BCDA conditions.
Plant operators would have a considerable period of time (at least several hours) to take corrective actions to establish or reestablish the SHRS function. In contrast, the NRC Staff's Appendix J analysis assumed that LOHS would always result in an BCDA (DSFES, Appendix J, Page J-3),
' without regard for the inherent margin provided by 'the heat l
transport system heat capacity. Consequently, this design characteristic provides enhanced SHRS capability which would make the Staff's estimate on the frequency of HCDAs due to LOHS conservative.
20
Q.21. What conclusion have you drawn concerning the NRC Staf f's estimates of the frequency of progression to HCDA conditions as a result of failure of the SHRS on demand?
A.21 The CRBRP SHRS uses the same design concepts--redundancy,
, diversity and independence--as are used in LWR plants.
This supports the NRC Staf f judgment (DSFES, Appendix J)
_ that the likelihood of failure of the SHRS would be no greater than that of similar LWR systems. However, there are three particular characteristics that enhance the capability of the SHRS: passive maintenance of primary coolant inventory, natural circulation, and large system heat capacity. The enhanced capability provided by these characteristics supports a conclusion that the NRC Staff's estimate of the frequency of HCDA initiation due to f ailure of the SHRS is conservative.
Q.22. Under Design Basis Accident conditions, how do the containment design characteristics limit the consequences and risks of accidents?
A.22. As shown in Applicants' Exhibit 1, Section 4, the Site
- / Suitability Source Term (SSST) release envelops the
- consequences of the spectrum of Design Basis Accidents and l
includes the effects of fission products, core materials and sodium under Design Basis Accidents conditions. The limiting Design Basis Accident results in a slow pressurization of containment to maximum pressures of less than 2 psig, as compared with a design pressure of 10 psig.
21
Even if the design pressure (10 psig) of containment is assumed throughout the release period, the containment can I
be designed to limit the radiological releases for the SSST (hence, for all Design Basis Accidents) well below the dose
- guideline values.
Q.23. Under conditions beyond the design base, how do the containment design characteristics limit the consequences and risks of HCDAs?
A.23. Applicants' Exhibit 1, Section 3.3 showed that CRBRP can be designed so that HCDAs are beyond the design basis.
Nevertheless, Applicants have included features in the design to provide additional margin for mitigation of these hypothetical accidents. As discussed in Exhibit 1, Section
. 5.2, these features are designed to meet the Structural Margin Beyond the Design Base (SMBDB) requirements in "Hypothe'tical Core Disruptive Accident Consideration in CRBRP" (CRBRP-3), Volume 1, Section 5.2 and the Thermal Margin Beyond the Design Base (TMBDB) requirements in CRBRP-3, Volume 2, Section 2.1. These features are designed to accommodate both the mechanical and thermal
! challenges resulting from HCDAs. As illustrated in Figure l
4 below, the SMBDB requirements provide design capability i to withstand an early mechanical challenge to the integrity j of the reactor coolant boundary. These requirements, in turn, are designed to prevent releases of radioactivity through the primary system, including the reactor closure 22
head, to the containment and an early (time periods on the order of seconds or minutes after initiation of an HCDA) challenge to the integrity of the Reactor Containment Building. The TMBDB requirements protect against both short term and longer term challenges to the integrity of the Reactor Containment Building resulting from the effects of whole core melting.
.. TMBDB (Core Melt
. Accommodation)
SMBDB (Energetic (s Accomrnodation)
\llllllllll-l
/
Figure 4.
~'
O.25. What is the significance of energetics to the risks and consequences of BCDAs?
A.24. Section 5 of Exhibit 1 showed that it is feasible to design CRBRP so that a realistic assessment of BCDA sequences, including best estimate analysis and a consideration of uncertainties, predicts a non-energetic outcome (no significant early mechanical challenge to the primary system integrity). Section 5 of Exhibit 1 also showed that 23 i _ __ __
l' i
pessimistic assumptions, well beyond those appropriate for a realistic assessment, must be invoked to predict an energetic outcome. Finally, Section 5 of Exhibit 1 showed that CRBRP can be designed to provide a structural margin which will accommodate the energetics predicted even in these pessimistic analyses. Significantly, substantial releases through the reactor closure head and an early challenge to containment integrity would not be predicted f or any of these cases.
Q.25. What is your opinion concerning the Staff's Appendix J estimates and assumptions regarding head releases?
A.25. In Appendix J of the DSFES, the assignment of relative probabilities and the selection of head release source terms for the primary coolant system response are judged to be conservative. The NRC estimates assume head release source terms that imply that all BCDAs would be energetic.
In fact, the likelihood of an energetic outcome is very low. In Table J.2, "CDA Class 1, 2, 3 and 4", consequences have been based on a source term corresponding to either Category III or IV for the primary coolant system response.
Both Categories III and IV imply an energetic HCDA (see p.
J-5) and substantial head releases due to mechanical challenges. This, in turn, has biased the analyses to overestimate the source terms released to containment and the consequences of HCDAs.
24
Q.26. Is an energetic BCDA a nuclear explosion?
A.26. No. Even f or those HCDA energetics analyses in which pessimistic assumptions have been made and an " energetic" outcome is predicted, the " energetic" result does not imply conditions at all similar to those resulting from either conventional (e.g., TNT) or nuclear explosives. A " nuclear 4 explosion" is physically impossible in an LMFBR, just as it is physically impossible in an LWR. This can be shown by comparing the basic physical characteristics of nuclear explosives, conventional explosives and HCDAs.
Q.27. What are the basic physical characteristics of nuclear explosives?
A.27. Nuclear explosives must be designed to minimize negative reactivity feedbacks while material motions are induced to provide a super-prompt-critical condition at reactivity insertion rates greater than a million dollars per second.
In that case, much of the energy release occurs in nano-seconds (billionths of a second) and results in peak pressures in the range of 5000 kilobars. Under such
' # conditions, much of the energy can be released in the f orm of shock waves 2 that can produce damaging impulse loadings on surrounding structures.
, 2 Shock waves are compression waves having a discontinuity at the wave f ront; they are f ormed, for example, when the speed of a body relative to a medium exceeds that at which the medium can transmit sound.
l I
25
~
l l
Q.28. What are the basic physical characteristics of conventional explosives?
A.28. Conventional explosives typically have initial pressures in the range of 300 kilobars. Much of the energy release occurs in micro-seconds (millionths of a second). Again, much of the energy can be released in the form of shock waves that can produce damaging impulse loadings on surrounCing structures.
Q.29. What are the basic physical characteristics of HCDAs?
A.29. An LMFBR, such as CRBRP, includes inherent prompt negative reactivity feedbacks that tend to limit any power excursions. As discussed in Q/A 9 above, the most important negative feedback mechanism is the Doppler coefficient which provides a negative feedback whenever the fuel is heated.
Although most BCDA sequences are predicted to terminate in a non-energetic manner (i.e., there is no sigr.ificant early
! mechanical challenge to primary system integrity), for some l
, pessimistic assumptions an energetic outcome could b,e predicted. In such energetic HCDAs, the reactivity insertion rates at prompt critical are typically in the range of tens of dollars per second. The energy release is limited by the inherent negative reactivity feedbacks and the movement of the fuel to regions of lower reactivity worth as a result of local pressurization. The peak pressures reached are typically less than 0.5 kilobars 26
l 1
4 (approximately 7000 psi). The energy of expansion of the pressurized materials is transmitted through the primary coolant system as pressure waves traveling at sonic velocity, J29.t as shock waves.
Q.30. How do the pnysical characteristics of nuclea explosions,,
. conventional explosions and en'ergetic HCDAs compare?
A.30. Table 1 provides representative values for characteristics of nuclear explosives, conventional explosives and energetic HCDAr.. Figure 5 illustrates the most important differences in regard to pressure and energy release.
~
Based on these comparisons, it is evident that the conditions associated with an HCDA are completely dif ferent from those associated with either conventional explosives
. Peak Prs.ssure (Kilobars) .
100,000 10,000 "^"-
, Nuclear
- Explosive '
1,000
~
, /l 100 Conventional -
Explosive 4
10 Energetic HCDA 1.0 0.1 0.001 0.01 0.1 1 10 100 1000 10,000 Tinw To Generate 50% Of Energy (Micro Seconds)
Figure 5. Nuclear and Conventional Explosive Comparison with Energetic HCDA.
27 i
4 6
NUCLEAR CONVENTIONAL ENERGETIC EXPLOSIVE EXPLOSIVE HCDA REACTIVITY INSERTION RATE GREATER WAN =
LESS W AN 100
($ / Sec) 1,000,000 MAXIMUM REACTIVITY $ 100 - 200 ------ Approx. 1 TERMINATION MECHANISM EXPANSION OF MATERIAL DEPLETION EXPANSION OF SOME WITH SHOCK WAVE OF REACTANT FUEL WITHOUT SHOCK NAVE LESS '111AN 0.010 3 > 1000 TIME 10 GENERA {E 50%
OF ENERGY (10 S)
TEMPERATURE (*K) 50,000,000 5000 5000 PEAK PRESSURE 5,000 300 0.5 (Kilobars)*
EXPANSION MUCH GREATER THAN SONIC GREATER THAN SONIC WITH SONIC WITH NO WITH FORMATION OF SHOCK FORMATION OP SHOCK WAVE SHOCK WAVE WAVE.
DAMAGE MECHANISM SHOCK WAVE LOADING SHOCK WAVE LOADING PRESSURE LOADING
- One Kilobar is approximately 15,000 psi.
TABLE 1 Nuclear and Conventional Explosive Comparison with Energetic HCDA
l or nuclear explosives and the use of the terms " nuclear explosion" or even " explosion" in relation to HCDA phenomena is simply incorrect.
Q.31. Do LMFBR accidents involve a risk associated with nuclear l
l explosion?
A.31. No.
Q.32. How can the risk associated with whole core melting be accommodated?
A.32. As shown in Exhibit 1, Section 5.3, whole core melting is a predicted outcome of some HCDA sequences. The effects of
, whole core melting on containment are characterized by a slow progression and there is considerable time (on the order of a day) before operation of the plant features provided to mitigate the consequences of such accidents is r equi re d,. Three types of TMBDB features are provided.
Instrumentation is provided to monitor the course of the accident and to assess the degree to which the containment
, is challenged (by measuring temperatures, pressure and
,' hydrogen concentration). To avoid unacceptable challenges to the containment, systems are provided to cool the containment, and to vent and purge containment to control hydrogen. In the event of the need to vent and purge, releases would be directed through a cleanup system that would remove a large fraction of the non-gaseous materials.
Since the accident sequence would proceed slowly and since these TMBDB features would be operator controlled, 29
flexibility exists to effectively manage the accident so as to minimize the accident consequences. Extensive sensitivity studies, which were summarized in Exhibit 1, Section 5.3, show that the TMBDB features can be designed for effective operation over a wide rcnge of conditions, including much more extensive sodium-concrete reactions than have been observed experimentally, variations in material properties, and variations in accident progression paths, while ensuring that radiological consequences are acceptably low.
0.33. What is your opinion concerning the NRC Staf f's Appendix J estimates and assumptions regarding containment failure i
under HCDA conditions involving whole core melting?
A.33. In Appendix J of the DSFES the NRC Staff estimated that the probability of containment failure as a result of the failure of containment mitigating systems (TMBDB features)
- could be as high as 10 -2 per demand. This is judged to be conservative. The criteria for and characteristics of these features are such that the Staf f's analysis ,
, e overestimates the likelihood of failure. In particular:
A. The TMBDB features are being designed to the specifications and requirements associated with Safety Class 3 components and systems (CRBRP-3, Volume 2, Section 2.1.1). Redundancy is being provided for the active components. Class lE power is being provided to these features.
30
B. The TMBDB features are being de.wigned so that appropriate testing and inspection can be performed after installation and periodically (CRBRP-3, Volume 2, Section 2.1.1) .
C. The active TMBDB components are located outside the Reactor Containment Building and as noted above the accident sequence is characterized by slow progression. This provides access and time for corrective actions, ensuring availability of TMBDB features when required. Maintenance could also be performed if needed af ter the features are brought into service.
Q.34. What is your opinion concerning the NRC Staff's Appendix J estimates and assumptions regarding releases from containment in the event of containment failure?
A.34. The predicted release of radioactive material in Appendix J of the DSFES is judged to be conservative for the following reasons: ,
~ r A. The overpressure failure of containment was assumed to occur at a pressure of about 20 psig.
This is considerably below the structural
. capability which can be provided. CRBRP-3, Volume 2, Table 3-10 shows representative analyses with failure pressures in the range of 45 to 55 psig.
31 i_. . _ _ _ . _ __ _ __ _ . _ . _ _
B. If containment failed by overpressure, it would likely be at a time in excess of the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> assumed by the NRC Staff. CRBRP-3, Volume 2, Section 3 shows representative analyses with times at which venting would be required of approximately 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />. If actions were not taken to vent, containment failure would occur at some time in excess of 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />.
Q.35. What conclusions have you drawn relative to the Staff's estimates of containment failure likelihood and releases from containment?
A.35. The Staff's estimates of release from containment are based on conservative estimates of the frequency of head releases. These estimates are conservative because they are based on assumptions which imply that all HCDAs are energetic. By contrast, an energetic HCDA is judged to be of low likelihood. In addition, the Staff has made a conservative estimate of the
. likelihood of containment failure by overpressure. .
/
Thus, the Staff's estimated frequencies of head releases and releases due to overpressure failure are conservative.
Q.36. What conclusion have you drawn concerning the consequences of beyond design basis events in CRBRP?
A.36. As indicated in Section 5.3 of Applicants' Exhibit 1, i
- atmospheric releases from HCDAs are characterized by l
I 32 i
i
radiological dose consequences that are acceptably low.
Moreover, these consequences are relatively insensitive to a range of initial releases of material through the reactor vessel closure head seals and because of the effectiveness of the cleanup system, these consequences are insensitive to containment vent times over a range of times between
~
about 10 and 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />. Furthermore, the analyses in Section 5.3 of Applicants' Exhibit 1 show -chat CRBRP can be designed so that the conservatively analyzed radioactivity releases compare favorably to WLSH-1400 values for similar beyond the design base events in LWRs.
Q.37. How does a more realistic calculation of the effects of CRBRP releases impact the resultant doses and the comparison of the CRBRP releases with LWR releases under similar beyond design basis conditions?
A.37. Repeating the calculations in Section 5.3 of Exhibit 1, but using meteorological data from PSAR Section 2.3 (Amendment 65), the current (heterogeneous) core design (PSAR Amendment 51), ICRP-30 models for bone surface (Endosteal
~ #
cells) and red bone marrow (NUREG/CR-0150, Vol. 3), and a more realistic calculation of gas sparging (carryout of fuel along with the gas that bubbles through the pool)3, 3This, considered a) a morg realistic temperature for the pool; 4500 F rather than 5000 F, and b) dilution of the Pu0 by the molten concrete. 2 33
. 1 the radiological consequences can be compared in the following tables:
DOSE
SUMMARY
FOR HYPOTHETICAL ACCIDENT SCENARIOS CONSIDERED (Rem)
Organ Case 1 Case 2 Case 3 Case 4 Bone Surface 0.027 0.19 6.47 27.0 Red Bone Marrow 0.026 0.040 0.56 2.18 Exclusion Boundary Liver 0.052 0.060 0.44 1.21
(~2 Hour)
Lung 0.021 0.032 0.72 1.77 Thyroid 0.014 0.020 23.4 19.6 W. Body 0.81 0.82 1.09 1.21
! , Organ Case 1 Case 2 Case 3 Case 4 Bone Surface 0.92 0.95 2.45 6.07 Red Bone
. Low Marrow 0.19 0.19 0.27 0.56 Population F
Zone Liver 0.36 0.36 0.18 0.32 (30 day)
Lung 1.54 1.55 0.82 1.00 Thyroid 85.3 85.4 8.13 5.43 W. Body 2.10 2.09 1.73 1.65 l
l l
I I
l l 34 l
l
l COMPARISON OF RADIONUCLIDE RELEASES TO ATMOSPHERE FOR CRBRP WITH LWRs FOR A COMPARABLE MELTDOWN SCENARIO Radioactivity Released (curies)
$1ement CRBRP PWR (3) BWR (3) 7 8 8
. Xe-Kr 3.6 x 10 1.0 x 10 2.1 x 10 5 6 6 I 2.1 x 10 2.0 x 10 1.1 x 10 3 4 4 Cs, Rb 5.2 x 10 1.2 x 10 7.6 x 10 5 5 4
Te, Sb 4.8 x 10 2.2 x 10 8.6 x 10 4 5 Ba, Sr 7.5 x 10 . x 10 2.2 x 10 3 4 3 Ru(II 2.8 x 10 3
.9 x 10 4
3.3 x 10 5
La(2) 4.1 x 10 2.9 x 10 2.9 x 10 II) Includes: Ru, Rh, Co, Mo, Tc (2} Includes: U, La, Zr, Nb, Ce, Pr, Nd, Np, Pu, Am, Cm (3)From WASH-1400, Appendix VI, Calculation of Reactor Accident Consequences, October 1975. The LWR scenarios used for comparison,here are PWR-6 and BWR-4 described in Section 2 of WASH-1400, Appendix VI.
~ r 35
Q.38. What conclusions have you drawn concerning the risks associated with beyond design basis events in CRBRP?
A.38. It is feasible to design CRBRP so that the risks of beyond design basis events are similar to those for LWRs.
Q.39. What conclusions have you drawn concerning the NRC Staff's analysis in Appendix J?
A.39. The Staff's analysis presented in Appendix J is conservative in three ways: First, the f requency of failure of both the RSS and SHRS are overestimated. Thus, the frequency of initiation of an HCDA is also over-estimated. Second, the radiological source associated with each of the HCDA classes (defined in Table J-2) is based on a head release (primary system failure category III or IV) .
This assumption, which implies that all HCDAs are energetic, leads to an overestimate of the frequency with which such releases would contribute to accident consequences. Third, the frequency of failure of containment due to overpressure is overestimated. Thus,
~ r the frequency of release due to HCDAs leading to overpressure failure is overestimated. Overall, the risk due to HCDAs as estimated by the Staff in Appendix J is conservative, with the greatest conservatism in HCDA classes 2, 3, and 4 which involve the larger releases.
36
Q.40. What accident conditions are appropriate for evaluation of the impacts of CRBRP accidents upon the Y-12 and Oak Ridge Gaseous Diffusion Plants?
A.40. To assess the potential impacts of accidents on the Y-12 l and Oak Ridge Gaseous Diffusion (K-25) plants, the Site Suitability Source Term (SSST) is the appropriate starting point since, as shown in Applicants' Exhibit 1, Section 4.1, this source term bounds all accidents considered credible.
Q/A 37 presented the results of Applicants' analyses for four HCDA cases which considered a wide range of releases of radioactive material through the reactor vessel closure head. All of those cases also considered whole core melting, reactor vessel and guard vessel penetration, sodium-concrete reactions and melting of the core materials into the concrete. Of the f our cases analyzed, the highest radiological releases were associated with Case 2, and this case has been selected for additional evaluation of the impacts of CRBRP accidents on
- < Y-12 and K-25.
In assessing the impacts on Y-12 and K-25, it is not appropriate to combine the already low likelihood HCDA sequence with other independent failures (such as failure of the containment isolation system or f ailure of the TMBDB mitigating features). Even if the combinations of such failures were considered, the risk from such cases would be 37
comparable to that from Applicants' Case 2, which has been used to assess the potential impacts of HCDAs on Y-12 and K-25. Although the consequences of the combined failures would be higher than Applicants' Case 2, this would be offset by the lower likelihood of such sequences. This can be seen by examining the results in Table J.2 of the DSFES.
Estimated probabilities and consequences are provided by the NRC Staff for CDA Classes 1 through 4. CDA Class 1 does not include the combination of other failures with the CDA. CDA Classes 2, 3 and 4 do include such combinations.
By multiplying the Staff's estimated probability for each Class by the Staff's calculated consequences (radiological i release) for that Class, a measure of relative risk of each of the Staff's four Classes of events is obtained. The
~
following table shows the products, normalized to the Staf f's CDA Class 1.
RELATIVE RISK FROM CDA CLASSES IN TABLE J.2 CINL Containment Isotooe Grouo
' Class Failure Mode Xe-Kr I Cs-Rb Te-Sb Ba-Sr Ru La 1 None 1.00 1.0 1.0 1.0 1.0 1.0 1.0 2 Overpressure 0.01 1.0 1.0 0.6 0.6 0.8 0.8 3 Isolation 0.01 1.3 1.3 0.8 0.8 0.6 0.6 4 Isolation 0.001 0.4 0.4 0.2 0.2 0.4 0.4 38
1 Based on this comparison, it is concluded that the NRC Staff's CDA Class 1, which has no containment failures combined with the CDA, provides a representative risk for all four of the Staf f's CDA classes. Applicants'
~
Case 2 involves containment conditions consistent with the Staff's CDA Class 1 and results in the greatest consequences of the four HCDA cases analyzed by the Applicants in Section 5.3 of Exhibit 1 and in 0/A 37 above. Consequently, the Applicants' Case 2 is an appropriate case, in terms of representative risk , to assess potential impacts of HCDAs on the Y-12 and K-25 plants.
6 i .
I 39
l l
STATEMENT OF QUALIFICATIONS ,
George H. Clare Westinghouse Advanced Reactors Division Oak Ridge, Tennessee 37830 From 1980 to the present I have served as Manager of Licensing at Westinghouse - Oak Ridge (CRBRP), with responsibility for managing assessment of CRBRP designs and the preparation of licensing material. These activities include consideration of features to prevent accidents, features to mitigate Design Basis Accidents, and margins to mitigate hypothetical core disruptive accidents.
I received a Bachelor of Science in Engineering Physics from Cornell University in 1972 and a Master of Engineering (Nuclear) from Cornell University in 1974.
After receiving my degrees I joined Westinghouse Electric Corporation as an Engineer at the Advanced Reactors Division.
Between 1974 and 1979 my position changed from Engineer to Senior Engineer. I was involved in licensing, safety analysis, and systems integration activities for the Clinch River Breeder Reactor Plant.
From 1979 to 1980, I served as Westinghouse Representative at the Fast Reactor Safety Technology Management Center at Argonne National Laboratory. There I participated in the management of activities in the Fast Reactor Safety Base Technology Program.
This included' monitoring and integration of safety research and development activities of DOE contractors throughout the US.
I am a member of the American Nuclear Society.
~
/
40
STATEMENT OF QUALIFICATIONS L. Walter Deitrich Associate Director Reactor Analysis and Safety Division Argonne National Laboratory Argonne, Illinois 60439 In 1980, I became Associate Director, Reactor Analysis and Safety Division, Argonne National Laboratory. My responsibility includes technical direction and administrative guidance of the fuel behavior and accident analysis activities, including phenomenology and code development related to LMFBR HCDAs. In addition, I have responsibility for analysis and phenomenology activities for LWRs.
I received a Bachelor of Mechanical Engineering degree from Cornell University in 1961, a Master of Science degree in
. Mechanical Engineering from Rensselaer Polytechnic Institute in 1963, and a Doctor of Philosophy degree in Mechanical Engineering f rom Stanford University in 1969.
Following graduation from Cornell, I joined the General Electric Company, Knolls Atomic Power Laboratory, as Engineer -- Thermal-Hydraulic Design, in which position I remained until 1964, when I left to enter graduate school at Stanford.
I joined Argonne National Laboratory in 1969 as an Assistant Mechanical Engineer in the Reactor Physics Division. I was assigned as a Lead Experimenter in the In-pile Experiments section, with responsibility for preparation, execution and analysis of TREAT experiments on behavior of fast reactor fuel under accident conditions. In 1970, this program was transferred to the newly formed Reactor Analysis and Safety Division (RAS).
In 1972, I was promoted to Mechanical Engineer and assigned as Group Leader -- Analysis, In-pile Experiments Section. My responsibilities included leading a group responsible for analysis and reporting of TREAT experiments simulating loss-of-flow and transient overpower HCDAs.
From 1974 to 1979, I served as Manager of the Fuel Behavior Section in RAS, with responsibility for modeling of fuel behavior and related phenomenological studies and code development.
From 1979 to 1980, I served as Special Assistant to the Associate Laboratory Director for Engineering Research and Development, providing technical assistance in management and direction of the reactor development programs at ANL.
41
. 1 I was promoted to Senior Mechanical Engineer in 1982.
I am a member of the American Socity of Mechanical Engineers, the i American Nuclear Society, and Sigma Xi.
,, e 42 1
i
I
\
STATEMENT OF QUALIFICATICNS ;
Lee E. Strawbridge Manager, Nuclear Safety and Licensing Westinghouse Advanced Reactors Division Madison, Pennsylvania 15663 Since 1980, I have been Manager, Nuclear Safety and Licensing with responsibility for directing safety analyses and licensing activities performed at the Westinghouse Advanced Reactors Division, Waltz Mill site for CRBRP and other nuclear projects.
I received a Bachelor of Science degree in Electrical Engineering from Pennsylvania State University in 1958 and a Master of Science degree in Nuclear Engineering from Massachusetts ,
Institute of Technology in 1959.
Following graduation from M.I.T., I joined Westinghouse Electric Corporation in 1959 as a Scientist in the Atomic Power Division and was in the position of Senior Scientist from 1962 to 1964.
In these positions, I performed nuclear design analysis for
, Pressurized Water Reactors and a wide range of advanced reactor concepts including thermal, epi-thermal and fast reactors.
From 1964 to 1966, I was Manager of Nuclear Development with responsibility for developing analytic techniques and applying them to the nuclear analysis of Pressurized Water Reactors and advanced reactors concepts. This included conceptual nuclear design analyses of a modular 1000 MNe LMFBR.
Upon formation of the Westinghouse Advanced Reactors Division in 1966, I was named Manager of Nuclear Development, with responsibility for all nuclear design analyses within the division. This consisted totally of work on sodium cooled fast reactors. I continued in this position until 1968.
3 From 1968 to 1971, I was Manager of FFTF Nuclear Design, with responsibility for the nuclear analysis and nuclear design of the Fast Flux Test Facility.
From 1971 to 1974, I was Manager of LMFBR Safety and Licensing, with responsibility for the safety and licensing activities associated with the LMFBR Project Definition Phase, which formed the basis for the Westinghouse proposal for CRBRP. The conceptual design activities for CRBRP were completed during this period and the initial specification of structural margin beyond the design base loads was made.
43
From 1974 to 1976, I was Manager of Safety Analysis with responsibility for directing many of the safety analyses reported in the CRBRP Environmental Report and the Preliminary Safety Analysis Report. In addition, safety analyses were performed and substantial input was provided to the FFTF Final Safety Analysis Report.
From 1976 to 1980, I was Manager of CRBRP Margin Analysis and Design, with responsibility for directing the analyses of hypothetical core disruptive accidents. This included the specification of structural and thermal margin requirements to mitigate the consequences of accidents beyond the design base and the preparation and submittal to NRC of the document CRBRP-3,
" Hypothetical Core Disruptive Accident Considerations in CRBRP."
I am a Professional Engineer, registered in the Commonwealth of Pennsylvania since 1967.
e ,
/
44
- - . _