ML20010H456
| ML20010H456 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 09/18/1981 |
| From: | Lauben N, Oreilly P Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20010H457 | List: |
| References | |
| ISSUANCES-OL, NUDOCS 8109240493 | |
| Download: ML20010H456 (10) | |
Text
___
UNITED STATES OF AMERICA s
NUCLEAR REGULATORY COPNISSION i
BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter o?
)
)
SOUTHERN CALIFORNIA EDISON COMPANY,) Docket Nos.: 50-331 OL ET.AL.
)
50-352 OL
)
(San Onofre Nuclear Generating
)
Station, Units 2 and 3)
)
-l l
_TMIMONY OF'G. NORMAN LAUBEN AND PATRICK D. O'REILLY l
l 1
(
Q.1. Please state your full name.
t A.
George Norman Lauben Q. 2. By whom are you employed, and describe the work you perform?
A.
I am employed by the Reactor Systems Branch,. Division of Systems Integration, Office of Nuclear Reactor Regulation, U. S. Nuclear Regulatory Commission.
~
i l
A copy of my professional qualifications is attached to this testimony.
Q. 3. Please state your f011 name.'
A.
Patrick Daniel O'Reilly Q. 4. By whom are you employed?
A.
I am employed by the U. S. Nuclear Regulatory Comission.
I am a Senior Reliability and Risk Assessment Analyst in the Reliability and Risk Assessment Branch, Division of Safety Technology, Office of Nuclear Reactor Regulation.
A copy of my professional qualifications is attached to this testimony.
Q. 5. Would vou describe the scope of the subject matter addressed in your testimony?
I A.
We have been asked to identify, relative to the risk associated with low power l
l testing, the significant postulated events which could occur at San Onofre Unit 2 that could potentially affect public health and safety.
i i.
810o24049381092}1 PDR ADOCK 05
- We have also been asked to address the safety significance of the low power testing program as it would affect the need to have in place an emergency plan which meets all the requirements of NU, REG-0654.
Q. 6.
What are the major factors affecting safe reactor operation during low
~
power testing?
~
A.
There are three major factors which contribute to a substantial reductit in risk for. low power testing as compared to continuous full power operation.
First, there is additional time available for the operators to correct the loss of important safety systems needed to mitigate relatively high risk events, or to take. alternate courses of action.
Secondly, the fission product inventory during this time would be very much less than during full power operation.
Third, there is a reduction in required capacity for mitigating systems at low power.
Q. 7.
What are the significant postulated events that could potentially affect the public health and safety?
A.
Since the publication of the Reactor Safety Study (WASH-1400), the WRC staff has continued to study the risk to the public from potential severe accidents at nuclear power plants.
This effort has confiam;d that the event teenarios dominating accident risks are generally the same for different PWR designs..
WASH-1400 concluded that.the dominant events for the PWR studied were (1) small break LOCAs with loss of the ermgency core cooling system, (2) transients in-volving total. loss of feedwater, and (3) failure of double check valves between the reactor coolant system (high pressure) and the residual heat removal e
system (low pressure), which results in a LOCA (inter-syst;m LOCA'y outside containment, i.e., the interior of the reactor vessrl communicates directly t
with the environment.
Based on our experience with risk studiet, it is our judgment that the dominant accident sequences for typical Combustion Engineer-ing-designed PWRs are similar to those identified in WASH-1400. This judgnent, therefore, applies to San Onofre Unit 2, with the exception of the inter-system LOCA.
Installation of leak testing equipment to periodically test the condition of the two check valves between the reactor coolant system and the residual heat removal system (called the shutdown cooling system in Combustion Engineering-designed PWRs) has essentially climinated the inter-system LOCA as a dominant risk contributor for San Onofre Unit 2 The NRC staff's evaluation of this matter is documented in Cection 5.4.3 of the staff's Safety Evaluation Report.
Q. 8.
.that was done specifically to address the question of risk due to Icw
~
power testing, and what was the conclusion of that effort?
A.
We have reexamined the dominant scenarios to estimate the reduction in l
the probability of the event because of the additional time available l
during low power operation for the reactor operators to correctLthe loss of important safety systems needed to mitigate the event or to take alternate courses of action.
Similarly, we have calculated the reduced fission product j
inventory for operation of an initially unarradiated core at 5% power for 6 months ind have determined the reduction in potential pnlic exposure via reduction in potential release magnitudes. Risk is roughly propor-tional to the prability of sevt:,e accidents (in which the heat sink is-lost) and to the fission product inventory in the core. From these factors we have estimt.ted that the overall reduction in risk to the public should be a factor of 500 to 10,000 if a plant is operated at 5% power from initial startup for 6 months compared to continuous full power operation.
l
. ]
. 4 Q. 9.
In your review of fuel load and low power test progra s what have you concluded about the potential risk of low power testing?
A.
Based on the actual power history of other reactors during their low power testing program, the actual power history expected at San Onofre Unit 2.
would result in even less available fission product inventory. The peak pcwer during this time period is only expected to be 3 or 4% of rate:$
capacity. Operation at this power level 'is only expected for a maximum of 20 days. This would' result in a further risk reduction by a factor of abcut 2.
It is therefore concluded that the public risk due to fuel loading and the proposed low power test progrm is less than public risk due to full power long-term operation by a factor cf about 1000 to 20000.
Q.10. How could risk to the public be affected by small break LOCA's?
A.
Risk to the public would occur only if there is release of substantial amounts of radioactive fission products outside the containment.
This could occur only if there'is a [a'ilure :o 2001 the core for an extended period of time. During this time the fuel element cladding would have to fail by overheating. The reactor coolant pressure boundary would have to be violated and the reactor building containment would have to Le violated. The important factor is that the core would have to remain uncooled for a significant le' gth of time.
(Approximately 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> during the lou 0wer test r.ogram).
e 4
u-
". r -
'T'-
V
. t However, NRC requires that all facilities licensed to operate are provided with reliable and redundant emergecc.y core cooling systems (ECCS). NRC regulations (10 C.F.R 50.46) require applicants to analyze a spectrum of pipe breaks and locations with various assumed equipment failures.
These analyses are performed with NRC-specified conservative assumptions and must demonstrate coolability of the core nr.d minimum generation of hydrogen. Specifically, ECCS evaluations must demonstrate that ECCS per-formance will result in a coolable geometry and less than li core-wide metal-water reaction even with the reactor at 102% power and worst-tase liccar heat rates. The San Onofre Unit 2 ECCS is required to conform to these requirements. Thus, for all power levels the requirements of NRC regula-tions provide adequate protection against. severe core damage. Thus, for small break LOCAs, substantiel risk to the haalth and safety of the public would occur only if the ECCS failed to operate as designed. Therefore, when performing risk assessment this conditica laust be evaluated.
Q. 11. With a reactor operating at a maximum of 5% of full power, coulr; these events lead to the significant amounts cf core damage?
A.
No. We have looked at these events and have concluded, as discussed herein, that at 5% power it is extremely unlikely that such esents would lead to significant amounts of core damage.
"Significant" is taken to be 5%
metal-water reaction.
Q. 12. Please provide the basis for your conclusions?
A.
LOCA analyses with severly degraded ECCS's weia performed which demonstrate the large c.nount of time that would be availabic at 5% power for diagnosis and corrective action to prevent significant core damage.
For these unalyses, it was conservatively assumed that none of the pumped ECCS systems was functioning.
Idaho National Engineering Laboratories (INEL), our 7
6-t consultant, ran calculations using the RELAP4 code to detennine the system behavior during the early portion of the transient. The T00DEE2 code was used for the remainder of the transient to calculate water bailoff, core uncovery, and the onset of core damage.
A bounding calculation was perfonned for a large break LOCA.
In such a
~
tese, with no pumped ECCS, refill by the accumulators is completed about 70 seconds after the break. The water level in this case is not above the top of.the core. At this time, the water in the reactor vessel would begin to heat up end boil away.. :owever, the core does not imediately start to heat up rapidly until a substantially greater amount of water has boiled off. The T00DEE2 analysis for this case shows that the fuel cladding temperature does not begin to rise rapidly to temperatures at which metal-water reaction (1800 F of higher) would occur for at least 2 1/2 hcors.
This -is the minimum time available for remedial action even for this highly unlikely event--large break LOCA coupled with ECCS failure.
For the more credible small breaks (but still assuming failure of the pumped ECCS) the margin is much larger.
For a small 2-inch cold-leg break LOCA, core uncovery would be delayed for about 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />.
Rapid increase in fuel clad temperature, leading to severe core damage, would not begin for about 21 hours2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br />.
Information was obtained from the NRC staff Project Manager for.the Sequoyah nuclear power plant concerning the actual maximum power level and test duration for the lovaower test program.
This information indi-cates that actual test power was about 4% of full power or less and lasted about 8 days.
e r
y
..e
---ss
. t The planned duration and power level for the low power test program at San Onofre Unit 2 will be almost the same as Sequoyah. Under such conditions, many more hours wauld be available for remedial action.
Q.13. Based upon this assessment, what is the likelihood of significant core damage at low power due to a LOCA?
A.
As indicated above, with the NRC's requirements for reliable ECCS performance, the potential for en*ere core damage and ' associated significant hydrogen generatien is very small even at full power.
The time available at low power for the operator to take corrective action is more than 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> in the event of a small LOCA.
In addition, the coolant flow required to dissipate decay heat at 10 hea,rs following a LOCA would be only about a gpm,which is within the capacity of the pumps used for the normal make-up systems. Because of the time available for the operators to correct malfunctions in the ECCS or to initiate cooling with the normal charging system, we believi that th'e pro bility of a small LOCA resulting in excessive idel damage ans significant radiological release is reduced by at least a factor of 400 to 8000 for low power operation as compared to operation at full power.
Q.14. What about the potential effects of ot. :r significant events, e.g.,
transients with total loss of feedwater, on the public risk?
A.
At 5% power the number of events that can result in failure to adequately cool the core is greatly reduced..All transients initiated by turbine trip are eliminated since the turbine is not on line. Total loss of feedwater caused by any other transient becomes negligible with respect e
r-
O i to core damage and public risk.
In such a case, core heat is transferred through the steam generators from the primary to secondary systems. After a scram from 5% power, we calculate that it would take about 21/2 days to boil the steam generators dry, conservatively assuming no feedwater makeup. During that period of time, diagnosis, corrective action or alternate heat removal methods could easily be acc$mplished. Moreover,
~
by that time fission product heat will have decayed sufficiently to that passive steam heat losses (radiant heat transfer) would be enough to keep the reactor cool, even if no corrective action were taken. As a consequence we believe that the total risk associated with these events is about 10,000 to 100,000 lower than_that at full power opera tion.
Q.15. If, as you say, feedwater transients are of almost no concern at low power, what about atlier scenarios? Could they not now become dominant at low power?
7-A.
Other transients (steam line break, steam generator tube rupture, rod ejecticn and ATWS) were also examined.
A similar reduction in risk was i
evident. Therefore, these transients did not become dominant.
Q.16. But an ATWS event could result in a significant primary to secondary heat irnbalance. Why isn't this of concern?
A.
We examined the risk potential for ATWS events at low power. The amber of ATWS events which can contribute to risk at low power is much reduced compared to full power. At low power, the only significant e
ATWS event is rod withdrawal followed by failure to scram. We examined the spectrum of ATWS events and factored this into the relative risk assessment.
4 l
l i
9-3 C. 17. What about ATW5 events involving loss of feedwater? Aren't these events significant at low power?
A.
At full power, these are the most significant ATWS event.
However at low power, they are not, unless coupled with rod withdrawal.
If one assumes that the highly unlikely sciasrio of total loss of feedwater is followed by a failure of the reactor system to scram, completed boil-off of the water in the steam generators would occur in about 30 minutes.
During this period there are a number of things the operator could do to bring the reactor to safe shutdown, including initiation of the boron in-jection system and diagnosis and correction of the failure to scram.
These would terminate the event before boil-off of significant reactor vessel inventory and thus, well before the onset of severe core damage. Moreover, at low power, significant overpressurization of the primary system does not occur because of the low integgqted reactor power.
Q. 18.
How does the above discussion relate to the need for a qualified emergency plan during low power operation?
l A.
The above discussion shows that abundant time (at least 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />) is available to take corrective ~ action to mitigate or terminate the most likely scenarios which could affect public risk during low power testing.
i For some sequences of concern at full power, no action would be required during low power operation to prevent public risk. Under these conditfons the risk is so small that there is virtually no need for a qualified emergency plan.
e
_,~r-
__.e.__
.-...v.
t STATEMENT OF PROFESSIONAL QUALIFICATIONS NORMAN LAUBEN My name is George Norman Lauben.
I am employed as a Nuclear Engineer the Reactor Systems Branch, Division of Systems Ihtegration, U.S. Nuclear i '.
~'
Regulatory Commission.
I have worked in the field of nuclear reactor safety for 19 years, and in nuclear activities for 23 years.
I have worked for the Comr.ission and its predecessor, the Atomic Energy Commission, since 1968.
During this time I have worked directly on reactor safety matters, including Emergency Core Cooling System (ECCS) performance review and Loss-of-Coolant Accident (LOCA) analysis.
I was a member of the 1971 AEC ECCS task force and the AEC Staff Panel for the ECCS Rulemaking Hearing.
I am the author of the T00DEE2 computer
'^
program used by the NRC and the nuclear industry for transient fuel pin thermal analysis during a LOCA.
I was a member of the technical team that accompanied Mr. Harold Denton to the Three Mile Island Reactor on March 30, 1979.
I have a B.S. and M.S. in Chemical Engineering from Case Institute of Technology (now Case Western Reserve University).
e e
-