ML20138J348
| ML20138J348 | |
| Person / Time | |
|---|---|
| Issue date: | 05/02/1997 |
| From: | Collins S NRC (Affiliation Not Assigned) |
| To: | Ross D NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| References | |
| NUDOCS 9705080171 | |
| Download: ML20138J348 (31) | |
Text
.. _ -..
May 2, 1997
!T MEMORANDUM T0: Denwood F. Ross, Jr., Director Office for Analysis and Evaluation of Operational Data l
Original sianed by: Frank Miraglia for FROM:
Samuel J. Collins, Director Office of Nuclear Reactor Regulation
SUBJECT:
REVIEW 0F PRELIMINARY CASE STUDY, "NON-POWER REACTOR SURVEY" The NRR staff has completed its review of the preliminary AE00 case study "Non-Power Reactor Survey" (the Survey) as requested in your memorandum of December 9, 1996. Comments are attached for your consideration.
I hope you find these comments of value.
If you have any questions concerning our comments, please contact Marvin Mendonca of my staff at 415-2170.
Attachment:
As stated DISTRIBUTION:
CentralzFile w/ incoming, JMurphy.
AAdams BSheron PUBLIC PDND r/f KBrockman SCollins FMiraglia/AThadani RZimmerman DMatthews PDoyle JCallan AMohseni TMartin PIsaac HThompson CERossi SWeiss BSweeney EJordan RSpence EHylton TMichaels NRR Mail Room (YT#960231 w/ incoming)
WTravers HBell,IG l
PQualls, NCF0 TReidinger,.RIII.
PMilano TDragoun, RI SHolmes, RI BMurray, RIV TBurdick, RIII CBassett, RII WEresian glh l
- PREVIOUSLY CONCURRED gi y
PDND:PM* AII l
TECH ED*
PDND:LA*
PDND:D*
DRPM:D*
DONRR 5/* y BCalure MMendonca EHylton SWeiss TMartin SC p3 l) 1/21/97 4/11/97 4/11/97 4/23/97 4/30/97 OFFICIAL RECORD COPY DOCUMENT NAME:
G:\\SECY\\ACTN_ITM\\YT960231
%D-h in@
080013 y-go m mo a 9705000171 970502 PDR ORG N*******
W f**%
O
~ *a UNITED STATES l
. j NUCLEAR REGULATORY COMMISSION 4
o
' t WASHINGTON, D.C. 20555-0001 May 2, 1997 MEMORANDUM T0:
Denwood F. Ross, Jr., Director Office for Analysis and Evaluation of Operational Data Y-FROM:
Samuel J. Collins, Director Office of Nuclear Reactor Regulation
SUBJECT:
REVIEW 0F PRELIMINARY CASE STUDY, "NON-POWER REACTOR SURVEY" The NRR staff has completed its review of the preliminary AE0D case study "Non-Power Reactor Survey" (the Survey) as requested in your memorandum of December 9, 1996.
Comments are attached for your consideration.
I hope you find these comments of value.
If you have any questions concerning our comments, please contact Marvin Mendonca of my staff at 415-2170.
Attachment:
As stated
v May 2, 1997 f
f MEMORANDUM TO:
Denwood F. Ross, Jr., Director Office for Analysis and Evalua. ion of Operational Data original signed by: Frank Miraglia for FROM:
Samuel J. Collins, Director Office of Nuclear Reactor Regulation
SUBJECT:
REVIEW 0F PRELIMINARY CASE STUDY, "NON-POWER REACTOR SURVEY" The NRR staff has completed its review of the preliminary AE00 case study "Non-Power Reactor Survey" (the Survey) as requested in your memorandum of December 9, 1996.
Comments are attached for your consideration.
I hope you find these comments of value.
If you have any questions concerning our comments, please contact Marvin Mendonca of my staff at 415-2170.
Attachment:
As stated DISTRIBUTION:
Central File w/ incoming JMurphy AAdams BSheron PUBLIC PDND r/f KBrockman SCollins FMiraglia/AThadani RZimmerman DMatthews PDoyle JCallan AMohseni TMartin PIsaac HThompson CERossi SWeiss BSweeney EJordan RSpence EHylton TMichaels NRR Mail Room (YT#960231 w/ incoming)
WTravers HBell,IG PQualls, NCF0 Pflilano TDragoun, RI SHolmes, RI TReidinger, RIII BMurray, RIV TBurdick, RIII CBassett, RII WEresian glh
- PREVIOUSLY CONCURRED i l y
PDND:PM* AI TECH ED*
PDND:LA*
PDND:D*
DRPM:D*
DONRR BCalure MMendonca EHylton SWeiss TMartin SC d n,r/
1/21/97 4/11/97 4/11/97 4/23/97 4/30/97 S/
97 0FFICIAL RECORD COPY DOCUMENT NAME: G:\\SECY\\ACTN_,ITM\\YT960231 j
l 1
Y t
't NRR Comments on Preliminary Case Study NPR Survey General Comments The Survey should more clearly articulate the regulatory framework under which the non-power reactor (NPR) facilities operate.
These facilities are licensed under the provisions of Sections 104c of the Atomic Energy Act (AEA).
These provisions direct the Commission "to impose only such minimum amount of regulation of the licensee as the Commission finds will permit the Commission to fulfill its obligations under this Act to promote the common defense and security and to protect the health and safety of the public and will permit the conduct of widespread and diverse research and development."
The risk associated with these facilities should be more clearly articulated.
In several instances the Survey attempts to make comparisons between power reactors and non-power reactors.
If the Survey is not clear with respect to both the lower risk and the provisions of the Act regarding minimum regulation, a reader may not fully appreciate the findings of the Survey.
For example, readers may become concerned that non-power reactors are generally not housed in containments or lack redundancy and diversity in some instances although neither of these power reactor issues are required from either a safety or regulatory perspective.
The Survey provides a summary of information contained in the Safety Analysis Report for certain NPR facilities.
In several instances, the information has been superseded by newer information. The information is not placed into context with respect to the conservatism of the analysis and a discussion of the NRC Safety Evaluation.
Without this information, it will be difficult to understand the nexus between the safety analysis information, the NRC regulatory program and the issues and the recommendations identified in the Survey.
The purpose of the section on Core Damage Events and their relationship to the current vintage of NPRs should be made more clear.
The fact that some of this experience resulted from tests to establish criteria to ensure safe operations should be recognized.
Comments for each section are provided in the following pages to assist readers of the Survey in the understanding the regulatory requirements, and the application of these requirements in safety analyses, inspection and enforcement activities and licensee responses to reporting of events.
Executive Summary Page v, last paragraph, third sentence states that:
"It was also noted that the NPR class of reactors tend to have much higher individual rod worth, as contrasted with power reactors." NRR made some comparisons to power reactors.
The Grand Gulf Boiling Water Reactor (BWR) and the McGuire Pressurized Water Reactor (PWR) were selected to represent large relatively recently licensed power reactors in the U.S.
The Rhode Island Atomic Energy Commission (RIAEC) 2 megawatt (MW) material test reactor (MTR)-type fueled NPR was selected to represent pool-type 1
r f
V NRR Comments on Preliminary Case Study NPR Surve_y reactors in the Survey, and the National Institute of Standards and i
Technology (NIST) 20 MW test reactor was selected to represent the few larger NPRs.
For control rod withdrawal transients, power reactor reactivity insertion rates were about 0.02 and 0.075 percent Ak/k/second for the BWR and the PWR, respectively. This compares to about 0.0075 percent Ak/k/second for RIAEC.
NIST assumed an insertion rate of 0.05 percent Ak/k/second for the control rod withdrawal transient analyses.
It should be noted that a few NPRs (e.g., NIST) use ganged control rod withdrawal which result in similar reactivity insertion rates as the grouped and sequenced withdrawals of power reactors, but that many of the other NPRs have insertion rain considerably below those in power reactor design. Rod worth (ex the PWR rod ejection and the BWR rod drop accidents were abon ?>.90 percent Ak/k and 1.4 perCant Ak/k, respectivelj.
For the RIAEC NPR, the maximum rod worth from zero power is about 1.2 nercent Ak/k.
NIST maximum rod worth at zero power is about 3.75 percent Ak/k.
Although individual NPR rod worth can be higher than power reactor individual rod worth; this is not always the case and rod worth may not be the most important parameter for comparison of safety analyses (i.e.,
reactivity insertion rate for rod withdrawal transients seems to provide a more common comparison point between NPRs and power reactors).
Page vi, lines 2 through 4 states that:
"As could be expected considering the different licensing eras, large variations exist in the Safety Analyses Reports (SARs). Of significance was the fact that the reactivity potential was not uniformly considered." These sentences do not consider design variations and the appropriateness of different safety analyses for design variations.
The NPRs of 2 megawatts licensed power level, which were the primary focus of the survey, have considerable variations in design (e.g., their fuel, primary coolant system and reactor protective features differ considerably).
The only common d'esign feature among these NPRs is plate fuel, and even there the fuel elements and plates have considerable differences.
For all NRC-licensed NPRs, the maximum credible reactivity insertion was analyzed with no resultant fuel damage. These analyses include comparison of the specific fuel design and reactor conditions to applicable tests such as the SPERT and BORAX tests.
Page vi, lines 18 and 19 states that:
"These facilities should have a startup check sheet that instructs the operators to verify integrity of connections and operability of recorders." From observations during NPR licer. sing, operator licensing, and inspection activities, there is no facility that does not have a startup check sheet or list (usually they are multi-paged procedures and lists).
Some of the events discussed in the Survey have been found during startup checks.
For example, the event described on page 26 of the appendix of the Survey where equipment protection scrams had been disabled at the Georgia Tech Research Reactor on February 15, 1994.
2-
i
- 4 NRR Comments on Preliminary Case Study NPR Survey Also, the event on page 38 of the Appendix describes a power level deviation rod withdrawal interlock that was not operable during operation for the FNR on November 24, 1992.
These events were also attributed to problems during the performance of startup checklists, but were not due to inadequate startup checklists as discussed later.
Problems with startup checks discussed in the Survey appear to be mostly procedural compliance issues.
Page vi lines 19 and 20 states that:
"Also, an operator should know not to conduct fuel movements while the reactor is critical; yet it happened." This sentence refers to a University of Michigan event.
Inspection findings and licensee reports show that operators did know that fuel should not be moved when the reactor was critical but did not know the reactor was critical.
Page vi, lines 17 and 18 states that "[t]he Panel observed a pattern of indifference with respect to the proper operation of the reactor proteccion system," a'nd page vii, lines 1 and 2 states that "[h]owever, increased inspection and enforcement actions may be needed in order to ensure the necessary degree of excellence in operational performance."
Because the facts that support these conclusions are significantly altered by our comments, we believe these conclusions should be reexamined.
Introduction Section 1.1, Background One of the Survey categories of events on page 2, lines 14 and 15, is
" Reactivity related (some events involved degradation of the Reactor Protection System [RPS])." Review of the MTR Reactivity Control Events in the appendix finds that about one-third of those events are related to RPS-associated events.
Specifically, the University of Virginia event of April 28, 1993, the University of Missouri at Columbia Research Reactor (MURR) events of March 16, 1993, and September 7, 1993, the Georgia Tech event of February 15, 1994, and the Massachusetts Institute of Technology (MIT) events of January 25, 1995, and December 7, 1993, are RPS associated events.
Further, on more critical evaluation other events categorized as reactivity-related should be categorized in the RPS or another category (e.g., the MIT event of March 20, 1995, the NIST event of September 10, 1993, the MURR events of March 20, 1995, and of November 24, 1992, and the FNR events of March 24, 1993, and of November 24, 1992).
Although these later events do involve equipment that would monitor nuclear power or effect scram, the connection of the events and corrective actions to reactivity is not otherwise apparent.
Further, the University of Virginia event of June 13, 1994, and the Georgia Tech event of February 15, 1994, are equipment-protection-related/ procedural non-compliance events and should not be categorized as RPS or reactivity events.
With this sort of categorization, the frequency of reactivity-and RPS-related events becomes about the same at approximately 0.2 events / year / reactor, and the RPS-related events would be more frequent than other event categories considered (i.e.,
___O E
t' NRR Comments o_q Preliminary Case StudLl@LSurvey reactor coolant leak and fuel handling events). The new category would further support points made about RPS problems at various places in the Survey. Addition of text material in the Survey for RPS events should also be considered.
I With regard to page 3, Table 1, several questions and comments arise from comparing the appendix and this table.
For NIST, the unlatched fuel element event on September 10, 1993, was due to refueling and should be categorized as fuel handling although it did manifest itself in power fluctuations.
For the University of Michigan in 1993, only one event was found in the appendix rather than two (i.e., the March 24, event on pages 36 and 37 of the appendix).
For the University of Virginia in 1993, only one event was found in the appendix rather than two events (i.e., the April 28, event on pages 11 and 12 of the appendix). This April 28, event should be categorized as a RPS-related event.
For the University of Virginia in 1994, there was an event of June 13, 1994 that does not appear to be in the table.
For the University of Virginia in 1995, two reactor coolant system (RCS) leakage events were found in the appendix rather than one (i.e., the November 1995 event on page 27 cf the appendix and the August through October 1995 event on page 28).
For the University of Virginia in 1996, there was an event of March 13 in the appendix that does not appear to be in the table.
Also, there does not appear to be a reactor coolant leak event in the appendix for this year.
1 I
1 4
s 9
NRR Comments on Preliminary Case Study NPR Survey Section 1.2, " Selected Instances of Core Damage Ev..Js" Page 4, lines 8 through 11 states that " Recounting this historical
=
experience is intended to place into context the hazards of this class of reactors, relative to the large power reactors,.whose risk has been subject to much more exhaustive characterization (e.g., NUREG-1150,
" Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants," and THI-2)." The rationale as to why power reactors have been subject to more exhaustive characterization of risk than non-power reactors should be made here (e.g., the ccnsequences of potential accidents are much greater for power reactors than for NPRs).
The sentence from page 4, line 17 through line 20 states that:
"On the other hand, the non-power reactors have higher worth individual rods, i
are generally not housed in a robust containment (although it is suitable for the purpose); do not have the rigorous operator training programs as do power reactors; have much less of an exclusion radius; and, in some instances lack equipment redundancy and diversity." These attributes seem to be based on a comparison to power reactor standards, which are different because of the provisions of the AEA and regulations, and because of the great difference in potential consequences.
It should be recognized that all of these conditions meet criteria for NPRs.
Therefore, the phrase "(although it is suitable for the purpose)" applies to all these points.
The rest of the paragraph that starts on line 20 of page 4 states that:
"On balance the public risk associated with NPRs is still much less than that of large power reactors. The following discussion should illustrate this point." How the discussion, that follows these sentences, illustrates the lower risk of NPRs as compared to large power reactors should be better established. Also, the connection to rod worth, containment, operator training, exclusion radius, and equipment redundancy and diversity as discussed in the beginning of the paragraph should be clearly established.
Section 2, " Review of Selected Safety Analyses" Although much of the information contained in this section is factual in
=
nature, its context and purpose for being discussed is not always clearly established.
In some cases, as discussed in the comments below, the information is gathered from the original SAR, which has since been superseded or supplemented.
In addition, the SAR alone, without reference to the corresponding NRC Safety Evaluation, may not provide the reader with a full understanding of the design basis of the facility or the conservatism with which these analyses were performed.
Therefore, the comments in this section are intended to more clearly articulate the safety analyses assumptions, conservatisms and results.
Section 2.1, " Reactivity Related Analyses" Page 9, lines 8 through 10 states that:
=
reactor considered a sudden reactivity insertion of 1.5 percent with the i
reactor critical, and asserted it was a highly improbable event (Refs. 9 I
r r
NRR Comments on Preliminary Case Study NPR Survey and 10)."
The accident was analyzed in Section 5.9.2 of the SAR (Ref. 9 of the Survey) and was discussed in Section 2.14.4 of the safety evaluation of Ref.10 of the Survey.
Peak cladding surface temperatures of about 270 *F were calculated.
Page 9, lines 10 and 11 states that:
"A fuel loading event might produce as much as 2.5 percent addition, which could produce fuel melting." This accident was evaluated in Section 2.14.5 of the safety i
evaluation for the conversion from high enriched uranium (HEU) fuel to l
low enriched uranium (LEU) fuel (Ref.10 of the Survey).
This l
evaluation concluded that SPERY data and analysis in the licensee's, 1967 SAR showed no melting of HEU for the Georgia Tech design with this amount of reactivity insertion.
The evaluation also found that the LEU would be less susceptible to the same accident based on LEU physics characteristics.
Page 9, lines 11 through 13 states that:
"The internal containment pressure would be between 2 and 11 psig." The "2 and 11 psig" value is based on a typo in the SAR.
The correct value is specified in the SAR calculational appendix and in the Survey appendix as 2.11 psig.
Page 9, lines 14 through 16 state that:
"The University of Michigan (2 MW) reactor considered a 1.6 percent reactivity addition, which would take the fuel to 900 *F (Ref.11).
(Melting is above 1200 *F).
A 1.8 percent addition would result in incipient melting." Reference 11, the NRC staff safety evaluation Section 14.2, " Sten Nuclear Excursion" says:
The total reactivity worth of all experiments is limited by the Technical Specifications to 1.2% (1.58$).
To be conservative in its consideration of the effect on fuel element integrity of instantaneous reactivity insertion, the licensee has analyzed a step nuclear excursion in which 1.6%
l Ak/k (2.12$) reactivity is inserted into the core instantaneously.
Because of this limitation and the maximum rate of mechanically inserting reactivity into the core with I
the control and regulating rod drives, the staff has not been able to identify a credible method for instantaneously inserting 1.6% Ak/k (2.12$) reactivity; however, it is assumed for purposes of the analysis that it can occur.
The 1
reactor is assumed to be operating at a power level between 0 and 2 MW, at which time 1.6% ok/k (2.12$) reactivity is inserted instantaneously into the core.
The potentially significant consequence associated with the rapid insertion of reactivity accident are the melting of the fuel or cladding material.
Tests conducted by the Idaho National Engineering Laboratory on the SPERT-1 reactor containing fuel elements similar to those in the FNR indicates that an instantaneous reactivity 4
addition, ~1.6% Ak/k (2.12$), produces an energy release of l
~10 MW/s (9500 Btu).
The SPERT and BORAX tests (Miller et al., 1964; Zeisaler, 1983; Forbes et al., 1956; Edlund and
-.--=.-.--- -
s
't NRR Comments on Preliminary Case Study NPR Survey Norderer, 1957) concluded that no melting or fission product release occurred from that level of reactivity insertion.
Additionally, the licensee has reviewed the data of the BORAX test, which indicated that the instantaneous addition of about 1.8% ok/k (2.38$) to the FNR would melt the meat of the hottest fuel element if the ambient coolant temperature were as low as 70*F.
The minimum FNR coolant temperature is
~90*F making the peak energy less than at 70*F.
As indicated in Section 4.6.2, the instantaneous addition of 1.6% ok/k (2.125) available in the FNR, which is less than 1.8% ok/k (2.38$), was estimated to cause the fuel temperature in the hottest fuel element to increase to
~900*F, ~300*F below the melt!ng point.
a From the above analyses, the staff concludes that the nuclear excursion from a step insertion of 1.2% ok/k would not pose any significant hazard to the operatira pe-sonnel, the environment, or the public.
With regard to the licensee safety analysis, Section 14.1, " Excess Reactivity Addition," of the SAR contains the following:
The fuel elements used in the Ford Nuclear Reactor are of the same type as those used in the Oak Ridge Bulk Shielding Facility and the Materials Testing Reactor.
The transient behavior of water-moderated lattices utilizing this type of fuel element has been the object of intensive study in the Borax and Spert test. The control fuel elements, in which the control and safety rods travel, have been reinforced to prevent collapse of the safety and regulating rod guide channels in case of a nuclear excursion of the type intentionally promoted in the Borax and Spert test. These control fuel elements are designed to take a 70-psi differential pressure across the guide channels.
The data of the Borax Report (ANL 5211) indicates that the instantaneous addition of about.018 delta K/K to the barely critical FNR would melt the meat of the hottest fuel element if the ambient coolant-moderator temperature were as low as 70 F.
I The instantaneous addition of 0.016 delta K/K to the reactor is estimated to cause the center of the hottest fuel plate to rise to about 900 F, some 300 degrees below the melting i
point, and would result in the release of 34 megawatt-seconds of energy for an initial coolant-moderator temperature of 70 F.
If this energy were absorbed in the pool water to generate steam, it would give rise to a pressure increase in the building of 1.3 inches of water.
The building, as limited by the plumbing traps, is capable of containing a 2-inch increase without a breach of 7
7
.c i
NRR Comments on Preliminary Casj 5tudy NPR Survey integrity.
No significant release of radioactivity would occur under these conditions.
It is therefore concluded that, if.no experiment or group of experiments subject to instantaneous removal from the neutron flux affects reactivity by more than 0.016 delta X/K, and if the reactivity held down by the shim safety and regulating rods does not exceed 0.016 delta K/K, then no nuclear accident can cause any portion of the active lattice to melt.
The descriptions in the Survey and in the evaluation and analysis should be compared to ensure consistency, particularly with moderator-coolant temperature conditions and the results of test for rapid reactivity insertions.
Page 9, line 16 states with regard to th University of Michigan reactivity transients that:
"Oschtions were not postulated." This sentence may refer to the SPERT IV test results (Ref. 6 in the Survey),
which is quoted, in part, in the Appendix to these comments.
There is also some data related to other SPERT tests that may have been a reason for the statement in 1D0-16634, "SPERT Program Review" which is also quoted in part in the Appendix to these comments. As indicated in the staff safety evaluation and the licensee safety analyses, applicable BORAX and.SPERT tests were considered in establishing the acceptability of safety requirements. NPR safety analyses are preformed with computer models that have been verified by data, including SPERT and BORAX tests.
These analyses are used to establish various safety limits, limiting safety system settings, and limiting conditions of operation for NPR TS requirements.
Therefore, the sentence on the fact that oscillatory behavior was not postulated should be reconsidered unless there is additional data that is applicable to the University of Michigan NPR.
Page 11, lines 4 through 10, discusses a RIAEC analysis which assumes a ramp reactivity insertion and no reactor scram.
The results of the licensee analysis indicated that "the fuel would operate in nucleate boiling with no fuel damage until a trip was manually initiated." These results are under question based on independent, NRC-staff funded analyses at the Idaho National Engineering Laboratory.
The question relates to the assumption of no further reactivity addition once the point of nucleate boiling was reached. The adequacy of the licensing bases analysis, which is a different analysis wh n e a scram terminates the positive reactivity addition, is not questiored.
The licensee has been informed of the question and is evaluating the analyses with the Argonne National Laboratory (which did the original analyses).
It is suggested that these facts be added to the Survey to show the current status of this analysis.
Page 11, lines 11 through 14 states that:
"No reactivity insertion excursions for the 10 MW reactor at the University of Missouri (Columbia) were found (Ref. 19)."
The MURR licensee analyzed various reactivity transients in the SAR and associated supplements.
The licensee analyzed the maximum step reactivity insertion that the reactor 8-
i s
I NRR Comments on Preliminary Case Study NPR Survey could withstand without core damage using the Chic-Kin code (forerunner of PARET) developed at Bettis.
It was determined that burnout would occur at 25.23 MW. Calculat;ons show that the reactor could withstand a positive step of 0.006 Ak/k.
This analysis assumes that a scram occurs.
(SAR addendum 3, pg 67).
The licensee studied reactivity additions further as part of determining TS limits on experiment worth.
This work using PARET showed that an additbn of 0.003 Ak/k would result in a reactor power level of 14.48 MW and that 0.0025 Ak/k would result in a power level of 13.78 MW.
Both power levels are less than the power safety limit. The TS limit on unsecured experiments became 0.0025 Ak/k (SAR addendum 5, pg 16). The licensee also analyzed continuous rod withdrawal with scram from starting conditions of the reactor shutdown and at power.
The reactivity additions would be less than the 0.003 Ak/k found to be acceptable above. At the request of the Atomic Energy Commission (AEC) staff, the licensee also performed an anticipated transient without scram (ATWS) analysis that considered a continuous rod withdrawal from a s;1utdown condition, continuous withdrawal of one shim rod from full power, and continuous withdrawal of all shim rods from full power.
For continuous withdrawal from a shutdown condition, the licensee concluded that the operator has sufficient time to take action (e.g., 120 sec if the event started with the rods 2 inches below the critical position).
For continuous withdrawal of one shim rod, nucleate boiling is encountered 30 seconds inta the event.
The operator has 16 sec after receiving the first annunciation of a problem (high power run-in) to take action before a safety limit curve is exceeded. For continuous withdrawal of all shim rods, the operator has 12 seconds after receiving the first annunciator to take action to avoid core damage.
It should be noted that the MURR reactor protection system met most power reactor standards for j
redundancy and diversity when installed and that ATWS is not considered credible.
These analyses should be added to the Survey to describe MURR reactivity analyses.
j Page 11, line 17 through line 20, has results for the HEU fueled core i
=
reactivity transient analysis at the University of Virginia NPR.
Since February 1994, this reactor has been fueled with LEU; up to December 1993, it had been fueled with HEU.
Therefore, for a short period of time during the three-year Survey development period, the reactor was fueled with HEU.
However, the more germane results are for the LEU.
The results for the LEU analysis are a peak power level of 3.88 MW with an increase in total coolant flow to 837 gallons per minute. Therefore, just LEL results or results for both HEU and LEU should be presented.
Section 2.2, "Lcss-of-Coolant Accident (LOCA) and Loss of Flow Analyses" The Survey in this section did not discuss the NIST LOCA analysis.
For completeness this should be discussed.
Also, it should be indicated that keeping NPR cores cooled with water for even a relatively short period of time after shutdown prevents fuel damage and eliminates the need for a long-term ECCS.
For example, the NIST reactor is only required to have about a total of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> of cooling water after a LOCA to prevent fuel damage (" Safety Evaluation
.g.
O s
NRR Comments on Prelimis rv Case Study NPR'Survev i
Report Related to License Renewal and Power Increase for the National l
Bureau of Standards Reactor, NUREG-1007., September 1983, Section 14.4,
" Loss of Coolant").
Page 12, lines 4 and 5 states that:
" Georgia lech considered a LOCA i
which, absent operation of the Emergency Core Cooling System, would l
provide some fuel melting (Refs. 9 and 10)." Additionally, although the SAR section on this accident does not discuss this feature, the primary coolant section shows fast-acting isolation valves, one each on the outlet and inlet from the reactor vessel (approximately 4 feet of RCS
+
piping is exposed before the valves).. Therefore, another assumed i
conservatism which was not specifically stated is that the break occurs i
between these valves and the vessel or that a valve does not close.
The assumption of considerable operations at licensed power (which is rare for most NPR facilities, including this one) was also a conservatism in l
this analysis and others.
For completeness, these conservatisms should j
be included at this point in the Survey.
Page 12 lines 9 through 11 state that:
"The University of Michigan a
(2 MW) considered a core uncovery following a leak (162 gpm) with no makeup (Ref. 11). The core would uncover in 4 hrs; no core damage was calculated." Reference 11 in the Survey, the " Safety Evaluation Report for the Renewal of the Operating License," (NUREG-ll38), July 1985, states the following:
The licensee performed calculations to determine the maximum
~)
consequences expected from an accident in which a pneumatic tube was damaged and the entire tube below the bottom was l
sheared. The maximum flow rate from a pneumatic tube i
rupture, assuming an iritial pool level of approximately 28 ft above the opening where the drainage would occur, was determined to be approximately 210 gal / min, Additionally, it was assumed that no emergency makeup water was available and that a low water scram occurred.
The time required to j
drain the pool water to the top of the core was calculated l
to be about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />.
By the time the core started to uncover, the fission product decay heat would have decreased from approximately 7% of full power at shutdown to <l% of l
full power.
The flow rate from the I beam port rupture was significantly less than that resulting from the pneumatic l
tube rupture.
j From an analysis submitted by the licensee that accompanied the request to amend the license to permit-replacement of l
the core with LEU elements (J. B. Bullock, June 1952) the decay heat remaining in the fuel elements 10 min fc11owing a scram would be low enough to permit cooling of the fuel elements with air.
This analysis assumed all the decay heat i
to be absorbed in the core and that the after-heat pcwer density distribution remained constant.
The maximum temperature would be below 650 degrees F, which is 550 degrees below the melting temperature of the fuel elements.
i,
~.
4 NRR Comments on Preliminary Case Study NPR Survey Because of the concrete block shield wall surrounding the reactor tank, the staff also has not been able to identify a credible method for an instantar.eous loss of the reactor pool water.
Furthermore, the operator can take corrective action remotely without any radiation exposure.
There also would be sufficient time to evacuate the restricted areas of the building, thus minimizing the exposure to the occupants.
In the case of a reactor coolant accident, the reactor core is cooled mainly by natural convection airflow.
Additionally, the emergency makeup water to the reactor pool is capable of supplying emergency makeup coolant at a flow rate of approximately 600 gal / min, which is almost three times greater than the maximum rate of coolant loss.
On the basis of the above analysis, the staff concludes that the residual heat resulting from a loss of reactor pool water can be dissipated by the natural convection airflow in the reactor pool and no fuel melting would result.
The time needed to drain the tank (approximately 3 h) to the top of the core would allow for mitigating actions by the reactor operator, and the resulting consequences would not pose a significant threat to the health and safety of facility personnel or the public.
The descriptions in the Survey and in this analysis should be compared to ensure consistency, particularly with regard to leak rate and time to core uncovery.
i Page 12, lines 12 and 13, states that:
" Loss of flow was not a concern at MIT.
Additionally, no pool drainage event which would lead to a LOCA was identified as a concern (Refs. 12 and 13)."
For loss of primary flow, natural convection was demonstrated to adequately cool tha reactor core.
The licensee also analyzed a loss of secondary flow and the response of primary-system instrumentation.
This analysis concluded that the I&C system would scram the reactor before the primary coolant temperature limit was exceeded.
Because of the cooling system design with anti-syphon valves and a double reactor tank (a light water tank surrounded by a heavy-water tank), tank damage that would lead to a LOCA was considered unlikely.
However, the facility also has an ECCS system with redundant sprays.
These conservative analyses should be discussed.
Therefore, it is suggested that the sentences in question be changed to:
"The MIT analysis of the loss of primary-coolant flow showed that natural convection can provide acceptable core cooling.
The analysis of the loss of secondary cooling demonstrated that primary-coolant temperature limits would not be exceeded. Additionally, because of the cooling system design with anti-syphon valves and -a double reactor tank (a light water tank surrounded by a heavy water tank), tank damage that would lead to a LOCA is not credible.
The facility also has an ECCS system with redundant sprays as an additional protective feature (Refs. 12 and 13)."
Page 12, the last full ' paragraph on this page, refers to an accident at MURR that was analyzed for 10 MW when the reactor was initially licensed J
'NRR Commenti on Preliminary Case Study _ NPR Survey at 5 MW.
The boil dry accident (loss of flow without scram) that the Survey refers to is not part.of the current safety analysis licensing basis although it represents a ennservative accident scenario.
When MURR applied for 10 MW operattuo,1he AEC det+rmined that the
.nsequences of this accident shu61d.be mitigated. As a result, the I&C system was redesigned to meet General Design Criteria 20 through 25 and Institute of Electrical and Electronics Engineers Standard 279 eliminating the loss of flow without scram acciaent as a credible accident scenario for analysis.
The licensee analyzed a number of other issues.
Loss of cooling in the center test hole (the island) and other high heating areas cooled by the pool cooling system was examined.
Boiling within the island is a reactivity concern because the island has a positive temperature coefficient.
Boiling (or complete flow blockage) in other areas around the core cooled by the pool system would not cause a reactivity addition greater than that needed to cause the reactor to be prompt critical.
The licensee examined shearing of a beam port, which would cause a loss from the pool system.
Emergency pool fill can put up to 1000 gallons per minute into the pool, (compensating for a complete beam tube failure).
Pool system ruptures were examined with the loss of water bounded by the beam port failure analysis. Also, even if the entire pool system drains, the reactor core is still cooled because the reactor cooling system is a separate closed system independent of the pool system.
During the upgrade to 10 MW, the licensee modified the primary system, as it did the I&C system, in response to AEC questions to reduce the consequences of primary-system failures.
A 20 percent reduction in primary flow is needed to cause boiling in the core, which has a negative reactivity effect.
The licensee analyzed a doubled-ended rupture of the largest core cooling pipe at the worse location (between either isolation valve and the pool liner).
The drop in pressure will cause the reactor to scram, the primary coolant pumps to stop, and the core isolation valves to close.
The analysis shows that the amount of primary coolant lost from the system while these engineered safety features are actuating is such that the core will remain covered.
The decay heat will be transferred to the reactor pool without core damage.
The licensee also analyzed a loss of flow accident with conservative assumptions and concluded that no fuel damage would occur.
In the MURR ATWS analysis, the licensee looks at a loss of flow caused by accidental closure of the isolation valves without reactor scram.
The licensee concludes that the operator has 7.5 seconds to manually scram the reactor after the initial annunciation that the valves have left the full open position. The licensee did not attempt to determine what level of core damage would occur if the operator did not scram the reactor although other analyses mentioned in the Survey and these comments assumed several fuel melting scenarios.
The Survey should describe these analyses to be complete and accurate. _
e NRR Comments on Preliminary Case Study NPR Survey The Survey in the last two lines of page 12 and the first three lines of page 13 describes the University of Virginia loss of coolant analysis.
For the instantaneous LOCA with one core spray functioning, the shutdown mechanism for the reactor is loss of moderator as the water drains from the core. This conservatism should be mentioned in the Survey.
Secticn 2.3, " Inlet flow Blockage Analyses" The Survey in the second paragraph of this section does not specify the doses and degree of conservatisms for the MIT NPR flow blockage analysis. The doses from the flow blockage analysis are 0.28 rad for the 2-hour thyroid exposure, 0.9 rad for 2-nour gamma exposure at the closest edge of the facility fence, and 4 rad at the second and third floors of the building next to the reactor containment. The analysis contains other conservatisms, which should also be summarized in the Survey to present a complete description of the analyses.
The last paragraph on page 13 discusses associated NIST analyses for this section.
It should be pointed out that the NIST analysis contains many conservatisms that result in the calculated doses being higher than doses potentially would be.
For example, two of the conservatisms that impact doses from iodine are that the internal confinement charcoal filter recirculation system for iodine removal from the confinement is not assumed in the analysis and the emergency exhaust charcoal filters are assumed to have an efficiency of 95 percent instead of the TS-required 99 percent.
The same paragraph states that:
"At the site boundary, within the confinement building, the whole body dose was 0.012 rad for the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, 0.035 rad for 2-24 hours, and 0.058 rad for the next 29 days.
At the nearest site boundary outside of the confinement building, exposure from the exhaust plume would result in a whole body dose of 0.021 rad for the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, 0.038 rad for 2-24 hours, and 0.002 rad for the next 29 days." These doses are given for a person at the nearest NIST site boundary (about 400 m from the stack) with the assumption that the person stands at the site boundary for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day for 30 days following the accident. Additionally, the thyroid doses could also be provided.
Therefore, to present a more complete and accurate description of the analyses resclts, it is suggested that these sentences be changed to:
"The whole body doses at the nearest site boundary from radiation penetrating the confinement building are calculated as 0.012 rad for the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> following the accident, 0.035 rad for the next 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> and 0.058 rad for the next 29 days for a total dose of 0.105 rad for the 30 day exposure following the accident.
The whole body doses at the nearest site boundary from the plume are calculated as 0.021 rad for the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> following the accident, 0.038 rad for the next 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> and 0.002. rad for the next 29 days for a total dose of 0.061 rad for the 30 day exposure following the accident.
The thyroid dose at the nearest site boundary from the plume is calculated as 0.23 rem for the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> following the accident, 0.77 rem for the next 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> and 0.19 rem for the next 29 days for a total dose of 1.19 rem for the 30 day exposure following the accident."
J e
3 Comments on Prelimingylase Study NPR _ Survey The Survey in the first two paragraphs of page '14 should consider additional conservatisms and results of the MURR flow blockage analysis.
In addition to the. fuel element design at MURR., the coolant strainer and the preoperational inspection of the pressure vessel and core before closure reduce the possibili.ty of a flow blod.uge accident.
- Also, calculations indicate that a 75 percent blockage of the hot channel will not cause clad failure.
It was determined that thyroid doses would be limiting.
The dose for the accident is calculated as 322 millirem for a person 500 ft from the containment for an infinite period of time.
The total meltdown accident discussed in this section is not a flow blockage accident. The total meltdown accident evaluated the limiting radiological consequences of accidents. Although the results were just within the Part 100 values, the AEC staff stated that "the improbable nature of this type of accident and the conservative manner in which we have calculated the consequences lead us to the conclusion that in any real event, the doses would be at least one order of magnitude lower."
This analysis was also done before the staff reovired MURR to upgrade their I&C and primary systems to eliminate iM possibility or reduce the consequences of certain accidents, as pre %ously discussed.
Section 2.5, " Summary" Page 15, lines 8 through 11, Panel Recommendation #1.a states that:
"If the likelihood of an unanalyzed event is determined to be much higher than previously thought (viz, the inlet flow blockage at University of Virginia) then the licensee should demonstrate that this credible scenario does not have consequences which exceed those of the event previously considered bounding." This is the current NRC policy fer such events, with the addition of the policy of ensuring that the risk of such events is acceptably reduced. For example, in evaluating problems, such as the flow reduction at the University of-Virginia, licensees can establish corrective actions (e.g., such as changes to procedures, systems and training) which reduce the likelihood of recurrence of the problem.
Further, consideration should be given for the potential for indication and mitigation of such events (again, the example of the flow blockage event at the University of Virginia, where operators detected the problem and the reactor was scrammed to terminate the transient).
Therefore, the recommendation should be further amplified to include:
, or that the scenario is not credible or is readily mitigated by procedures, systems, and training."
Page 15, lines 12 through 14, Panel Recommendation #1.b, states that:
"When the facility is due for license renewal, the safety analysis should be reviewed in depth, taking into account the lessons of the last 10 years and updated as needed." With each NPR license renewal an indepth review is performed by NRC.
Lessons learned since the.last major licensing action (which can be more or less than 10 years) are considered.
Examples of such indepth review for this policy are described in the " Guidelines for Preparing and Reviewing Applications for the Licensing of Non-Power Reactors," NUREG-1537.
\\
's NRR Comments on Preliminary Case Study NPR Survey Section 3.1, " Reactivity Related Events" With regard to failures of the MURR regulating blade, page 17, lines 2 and 3, states that:
"While not affecting the ability to automatically shut down the reactor, these failures impaired the reactivity control function and led to a reactor trip." Each time, either the facility was manually shut down or power was reduced to investigate the event. Of the seven events, the reactor was manually scrammed three times.
The i
licensee did not lose the ability to manually control or scram the reactor nor to automatically scram the reactor as mentioned in the Survey.
However, the non-safety related, automatic control system functions were lost or degraded.
Page 17, lines 9 through 11 states that:
"There have been no known failures to scram on demand, although it appears likely that manual scram would have been required if an initiating event requiring scram had developed during the 5.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> the University of Virginia had its automatic scram disabled."
It should be recognized that other scrams i
were also available, although in all likelihood, based on alarms and indications, the manual scram would have been the first to be initiated, if necessary.
There were two automatic scrams associated with different
)
i temperatures for the mineral irradiation facility (these scrams are not TS-required, nor are they considered in the safety analyses).
Either or these two diverse, redundant automatic scrams would have automatically j
stopped reactivity additions without operator action (i.e., scram would probably have occurred at 53 megawatts, which is below the reactor safety limit for a large reactivity addition). A reactor bridge radiation monitor automatic scram at 30 millirem / hour was operable as required by TS (this scram in all likelihood would have occurred after p'ower level would have risen above the Safety Limit for a large reactivity addition). All other scrams were disabled by the problem (e.g., loss of primary-coolant flow rate., loss of poal water leve' increasing pool water temperature, increases in reactor power, o, decreases in reactor period). As previously mentioned, several indications of reactivity insertion were available to the operator (e.g., delta-T alarm, core gamma alarm, hot-thimble facility temperature alarm, automatic (servo) control mechanism tripping out at 7 percent deviation, and/or bottom or top out lights). Any one of these alarms could have alerted the operator to increasing power due to a reactivity addition or other problems.
These points should be described in the Survey to present a more complete and accurate characterization of the event.
Page 17, lines 12 through 13 states that:
"Those equipment failures that occurred should have been expected and covered by routine maintenance activities " An indication of which equipment failures should have been expected and the reasons they should have been covered by routine maintenance should be provided.
This would show licensees the Survey standards and criteria, and encourage improvements in this area.
Page 17, lines 13 through 15 states that "The basic problems appear to have derived from human error, inadequate training, complacency in O
I s*
NRR Comments of Preliminary Cate Study NPR Survey operation and management, and inattention to maintenance needs." As with our previous comment, the events in the appendix and associated analyses that demonstrate these attributes should be specified, again so that licensee improvement can be fostered and the Survey standards and criteria understood through examples.
Page 18, lines 5 and 6 states that:
"This is espe w ly important in light of accident potential from large reactivity tra.isients."
This sentence seems to be contradictory to the preceding paragraph which indicates that NPRs are very unlikely to have large, damaging reactivity transients.
Therefore, the sentence should be changed to:
"This is especially important in light of potential fuel damage consequences from large reactivity transients, although such events are highly unlikely based on considerations as previously discussed."
Page 18, lines 8 through 11, Panel Recommendation #2, states that:
"The NRC should review its requirements for startup checks, on a plant-specific basis, and implement license amendments where needed.
This same philosophy should apply to tests and maintenance on the RPS, in order to assure that the system is returned to operable status, after test or maintenance." Current requirements do establish that startup checks are to be performed. American National Standards Institute (ANSI)/American Nuclear Society (ANS) Standard 15.1-1990 gives NPR l
licensees guidance en the content of TSs and is supported by the NRC l
staff. ANSI /ANS 15.1-1990 has a requirement for procedures to startup, operate, and shut down the reactor.
Procedures that fulfill this TS requirement would include pre-startup checks.
The same is true for the reactor safety systems. ANSI /ANS 15.1 calls for operability tests following modification or repairs to the reactor control and safety systems. NUREG-1537 states that NRC recommends the practice of channel tests of all scram and power measuring channels required by TSs before each reactor startup and also states that NRC accepts the guidance of ANSI /ANS 15.1 for operability tests following modification or repairs to the reactor control and safety systems.
This guidance is implemented by l
facility-specific TSs.
Inspection program requirements currently are used to verify that startup, and post-maintenance, and post-test checks are done in accordance with TS and associated procedural requirements.
- Further, operator licensing often includes examination on such issues.
Page 18, from line 12 to the end of the page states that:
" Panel Recommendation #3: The NRC should ensure that a training course is developed by the NPR community or by individual licensees, for use by reactor operators and their supervisory chain.
The purpose of this course would be to instill an understanding of hazards of this family of reactors during postulated transients and accidents.
Inasmuch as AE00 has already developed a course along these lines for power reactors (R-800, " Perspectives on Reactor Safety"), AE00 should provide advice and assistance.
The course would be most effective if given by the NPR community itself.
Every operator and supervisor should, in the fullness of time, be required to take the course. The course should also be mandatory for NRC staff who work in the licensing and inspection areas." -
i
.g j
NRR Comments on Preliminary Case Study NPR Survey 9
Training of operators and managers in their organization currently i
involves preparation, conduct and participation in the operator licensing training and requalification programs.
This training is i
specific to each NPR and explains the associated hazards and protective 3
features. The regulations require training that is specific and applicable to the facility, as well as general reactor theory training and knowledge.
Current licensing, operator licensing, and inspection activities have shown that licensees understand the hazards of their reactors. A general course such as that recommended is unnecessary to ensure licensee personnel understanding of the hazards of their facilities, because this is already accomplished by the facilities.
For NRC licensing and inspection personnel, an NPR Technology Course was i
developed in cooperation with the NRC's Technical Training Division.
j This course is required in addition to other professional training for j
inspection personnel. The NPR Technology Course for all the different types of NPRs describes associated protective systems and designs, reactor physics, accident safety analyses, radiation protection, emergency preparedness, safeguards, licensing, TSs, and inspections.
This course provides an understanding of the hazards of and requirements to safely operate an NPR.
This recommendation should reflect current practice as described above.
Section 3.3, " Fuel Handling" The Survey refers to a " single event" when there were two or three fuel-handling-related events, depend'ing on the categorization (as previously l
discussed under the comments on Table 1 of the Survey).
The Survey commented that "the NRC should consider the safety significance of a fuel drop accident, but on a medium priority basis."
The staff currently considers _ fuel handling equipment and the potential i
fuel handling problems as a high priority issue.
Fuel handling issues are considered in the SERs for all licensed NPRs, i.e., whether fuel handling equipment is acceptably designed to minimize the potential for fuel handling problems.
Also, selected procedures are reviewed and activities are observed for potential fuel handling problems by inspectors.
Additionally, appropriate analyses of fuel handling accidents are required.
For example, NIST was required to analyze refueling accidents for a fuel element drop into the pool, a heavy i
object drop onto the fuel rack, a fuel cask drop, and failure of a fuel element.
Radiation doses at the site boundary were well below 10 CFR Part 20 limits.
Further, NUREG-1537 guidance indicates that licensees should analyze fuel damage accidents, which involve dropping or otherwise damaging fuel in any location or involve dropping, impact, or other malfunction of a non-fueled component.
Inspection activities Gre scheduled to ensure effective inspection, and on the average coincioe with fuel movement at about one NPR per year.
Additionally, inspection activities are planned and conducted for large movements of fuel (e.g., conversion to LEU or pool modifications), and these inspections avera~ge out to about one NPR a year.
Therefore, NRC
_y o
J 6'
NRR Comments on Preliminary Case Study NPR Survey inspectors on average observe fuel movement activities at least at two NPRs a year.
This comment.is adequately addressed by current NRC staff guidance, evaluations, and observations as described above.
Section 3.4, " Radiation Protection Events" Page 21, lines 14 through 16 states "Given that these are largely
=
institutions of higher learning, it is particularly important that a higher degree of management attention be given toward improving radiation protection practices. More guidance is not needed; rather, adherence to existing practices is needed." The Survey should not differentiate between licensees as a function of organizational purpose, 1
as the attributes, that establish a sound radiation protection program, are important for all licensees from an NRC mission point of view.
The Survey should consider that the last significant radiation protection enforcement at an NPR was prior to 1990, as evidenced by Table 2 of the Survey.
Some of the events and conditions in the Survey's Table 2 (e.g., the Georgia Institute of Technology August 1994, violation) did institute procedure changas as currective actions, and the Survey should not suggest such corrective actions are not needed since future t
circumstances may warrant them.
Section 3.5, " Design Basis Control" Page 22, first full paragraph discusses MURR regulating blade failures.
=
Of the seven problems that occurred with the regulating blade, analysis of the problems shows that only two of the failures (loose set screw on 1
input shaft) were related.
The other failures were unrelated, and only the first loose set screw problem established an opportunity for a lesson learned that could have prevented the subsequent problem (i.e., a maintenance procedure had not been updated properly).
These events are i
also discussed in more detail under related comments on the appendix.
The licensee recognized this relationship in the event report on the second failure.
Inadequate corrective action was the primary cause of i
only one of the events.
The statement that all of the seven failures necessitated manual reactor scrams is not consistent with the reports on these events.
In only three of the events was the reactor manually scrammed by the operator.
1 Also, these comments should be appropriately considered in Section 3.7 and in the appendix.
The first sentence of the first full paragraph and the associated footnote on page 23 are as follows:
"All reactors (power and non power) are licensed to operate as utilization facilities under Title 10 and in 4
accordance with the Atomic Energy Act (AEA) of 1954, as amended...'"
The AEA was written to promote the development and use of atomic energy for peaceful purposes and to control and limit its radiological hazards to the public. These purposes are expressed in paragraph 1104 of the AEA."
m A
- e NRR Comments on Preliminary Case Study NPR Survey The footnote says that the purposes of the AEA are expressed in that paragraph 1104, but the purposes of the AEA are specified in 5'ction 3 (The footnote appears to paraphrase Section 3d of the AEA).
There is no section or paragraph 1104 in the AEA.
If the reference in the footnote to paragraph 1104 is supposed to be Section 104, it also should be noted that other AEA sections also include regulatory authority for reactor licensing (most notably Section 103 for commercial power reactors).
The paragraph starting on page 25, line 14, states that:
"At least two
' events (those at the University of Virginia and the United States Geological Survey) directly indicate that current evaluations by non-power reactor licensees are not thorough enough.
Non-power reactor licensees are not relieved from the requirements of 10 CFR 50.59."
These two licensees improperly analyzed the situation for these two events.
NPR licensees must follow all applicable regulations, and based on regulatory interactions, NPR licensees understand this point.
In accordance with the enforcement policy, violations were written in both of these cases and escalated enforcement was takea against the University of Virginia.
This paragraph should be changed to:
"At least two events (those at the University of Virginia and the United States Geological Survey) indicate that non-power reactor licensee evaluations are not always thorough, Non-power reactor licensees know that they are to comply with the requirements, including 10 CFR 50.59, and violations for these two events were issued by the NRC."
Additionally, the description of the United States Geological Survey NPR event is not contained in the appendix and is not presented until page 29.
The description should be provided here so that the reader can understand the comment in the Survey. Also, the scooe of the Survey should be redefined since this is not one of the reactors previously mentioned.
A paragraoh in the Survey (page 25, line 18 through page 26, line 2) discusses the importance of design basis control and regulatory inspection and enforcement action.
In the case of the licensees noted in this paragraph, there was no indication of a widespread breakdown of the 50.59 review process; rather individual errors in a particular design modification occurred (as discussed in the previous comment).
The two events quoted were dealt with by actions consistent with the NRC Enforcement Policy.
NPR inspectors do review 10 CFR 50.59 evaluations for design changes as part of the NPR inspection program.
NPR Project Managers review the listings of design changes that are required to be reported in NPR annual reports to verify continued design control.
Further, NRC reemphasized the importance of design control, as well as other lessons learned from Millstone, at the NRC seminar at the TRTR annual meeting in 1996.
NRC lessons learned from Millstone and other events and conditions will continue to be incorporated in the NRC regulatory program.
The first full paragraph on page 26 refers to " repetitive occurrences."
=
The events and the reasons they are identified as repetitive should be.
J V
NRR Comments on Preliminary Case Study NPR Survey specified, so that the bases for the statement can be verified and so that licensees can learn specific areas for improvement from this analysis and can understand the criteria against which they are held by the Survey.
For the second and third full paragraphs on page 26, the Survey should specify any events or analyses which illustrate the points of the paragraphs.
Specific examples are needed to understand the Survey criteria for the statements and respond appropriately.
Section 3.6, " Operating Experience Feedback" Page 27, line 10, indicates that root causes and corrective actions were not detailed enough to be of use to others.
Examples and analyses to support the statement should be provided.
" Panel Recommendation #5.a states that:
"The NRC should review the reporting criteria for events and make such modifications as are needed to assure that important events are reported on a timely basis." The current reporting requirements are consistent with the consensus standards of the NPR community, ANSI /ANS 15.1.
For events that were not reported but should have been per requirements, appropriate enforcement action was taken.
The reporting criteria are established to ensure that events have an appropriate level of safety significance. As licenses are renewed or as licensees request license amendments in the area of reportable events, the definition of reportable events will compared to l
the ANSI /ANS 15.1 guidance. These practices ensure that reporting i
requirements and activities follcw NPR guidance, and that events of importance are appropriately shared among licensees and regulators-It should also be noted that the Survey seems to find the reporting and sharing of events generally comprehensive, and members of the NPR community have indicated that this is due, in large part, to the current regulatory program for NPRs.
The Survey should reflect these points.
Section 3.7. " Safety Culture" Page 28, starting on the fourth line reads as follows:
"At the University of Michigan (November 1992) during a routine maintenance period, the shim range-control rod interlock system was removed from the reactor control system for a modification that had been reviewed and approved by the facility Safety Review Committee.
Subsequent post modification testing was inadequate to identify that the power level i
deviation interlock was inoperable due to a wiring error. Additionally, the startup check list that was conducted after the modification was done by a trainee and there was no engineering or quality assurance oversight.
The review by the Safety Review Committee was inadequate and the decisions associated with the subsequent startup checklist were ineffective in demonstrating the operability of the system."
The Safety Review Committee (SRC) reviewed the planned modification, which was to remove the shim range-control rod interlock from the reactor control system.
The SRC evaluated this planned modification,
]
and cannot be held responsible for assuring that no wiring errors were q
+c NRR Comments on Preliminary Case Study NPR Survey made, nor for the misinterpretation of system response by the Senior Reactor Operator during the startup checklist.
The Senior Reactor Operator (SRO) monitoring the Control System Startup Checklist (CSSC) misinterpreted the system response to a CSSC procedure that hitherto had been adequate.
The CSSC was also monitored in part by the electrical engineer for the facility. As noted in the appendix to the Survey, the licensee did change the procedure but did not specifically state that there was unclear procedural guidance.
Based on the above, it is suggested that the quoted section beginning with the third sentence be rewritten as follows:
" Additionally, the post-modification check list was done by a trainee and monitored by a SRO.
The electrical engineer was present during the initial performance of the CSSC, but neither he nor any other member of the quality assurance team was observing the performance of the checklist step-by-step.
The SR0 misinterpreted a portion of the system response during the startup checklist.
The subsequent starte checklist, which had previously worked satisfactorily, was revise this problem to reduce the possibility of misinterpreting system i sunse.
Page 31, last paragraph states:
" Panel Recommendation #6:
The NRC should review its enforcement philosophy concerning non-power reactors.
Compliance with the regulations, licensing bases, and technical specifications is an essential component of safety regulations." NPRs are inspected in accordance with NRC Inspection Program.
NPR enforcement actions are in accordance with the NRC Enforcement Policy.
A key enforcement-related difference from the power reactor policy is that an NRR review of all violations is required by the NRC Enforcement Manual for NPRs whereas such review is only required for escalated enforcement on power reactors.
If there are particular examples of inspection or enforcement that were not in accordance with the requirements and NRC policy, these examples should be provided so that NRC can evaluate them and appropriately resolve the issue.
Sections 4.1, " Conclusions," and 4.2, " Recommendations" Page 34 refers to "AEC Chairman Thompson." According to the NRC's
=
Historian, T.J. Thompson was an AEC Commissioner from June 12, 1969, until his death in a plane crash on November 25, 1970.
He was never Chairman of the AEC.
He served on the Advisory Committee on Reactor Safeguard (ACRS) from 1959 to 1965 and was a Chairman of the ACRS.
Section 4.3, " Observations" This section basically quotes some conclusions and recommendations from Reference 2 of the Survey.
These conclusions and recommendations have been incorporated in NPR design and regulation.
The lessons have been learned (e.g., single failures are assumed for safety analyses; designs provide primary, secondary, and for some conditions tertiary protection for events; startup locks and interlocks are provided to prevent unpremeditated control rod withdrawal; shutdown margin is defined without taking credit for the most reactive rod; experiments and tests m
J v
l N_RR Comments _on Prit imi!Nrv Case Studv 1PR..survev l
l are controlled 'by procedures and -authorizatiions; instrumentation is required to provide accurate indications of reactor conditions; and startup procedures are designed.to verify shutdown functions). These observations are not lost on the current generation.
For example, the International Atomic Energy A_gency documents.354S1, " Code on the Safety of Nuclear Research Reactarru -Derdynf and SS?., " Code on the Safety of Nuclear Research Reactors: :0peratiorn,"' include such precepts. An NRC staff member ass'sted in the development of these safety standards.
NUREG-1537 also envelopes these principles. All of the quotations are important parts of the current safety program for reactors.
For the various described events and condition:, the Survey should identify problems in the implementation of the.,a principles and relate them to the root causes of the events.
Comments on References Reference 3 appears to be in Nuclear Safety, Vol. 5, No.1 (Fall 1963),
a not in " Nuclear Safety, Vol. 5, No. 2 (Winter 1963-64)."
Reference 6 refers to "PBE."
It should be " Power Burst Facility."
Reference 7 refers to "ID)."
It should be "IDO."
Comments on Appendix For the MIT event of March 20, 1995, on page 1 of the appendix, only one of six corrective actions is mentioned. All corrective actions should be describad.
For the fuel-handling event at the University of Michigan (2 MW) NPR on June 8, 1992, on page 8, under "Cause of Event," the value of "0.054 ak/k" should be "0.0054 4k/k."
The appendix on page 8 for this event should refer to the Augmented Inspection Team Inspection Summary Results, which state that:
The event had no immediate effect on the health and safety of the public.
However, the event was the result of personnel error involving two apparent violations of the facility's technical specifications:
(1) safety margin fcr fuel movements (Section 3.1.5), and (2) a failure to report the event within the time frame defir.ed by the technical specifications (Section 6.6.2.a).
Weaknesses identified included poor communications during the event, operators not reviewing or using procedures, a lack of a thorough review and full analysis of the event prior to further fuel moves, and the determination of reportability of the event.
Page 9 of the appendix states that:
"Two Severity Level III violations, I
for violating TS for the reactor to be subcritical during fuel movements l
and violating the procedure to have all rods fully inserted during fuel i
movements, and one Security Level I violation, for failure to report the event to the NRC within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, resulted in a $1,200 civil penalty
(
"A Severity being proposed."
Level III violation was issued for violating TS to have the reactor subcritical during fuel movements and violating the procedure to have This violation resulted all rods fully inserted during fuel movements.A Security Level IV violation, for failure in a $1,250 civil penalty.
Addi-to report the event to the NRC within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, was also issued."
tionally, it should be pointed out that corrective actions by the These corrective licensee were required prior to restart of operations.
actions include modification of procedures and installation of light indicators on the rod drive housing located on the bridge (the indica-Training and these tors are lit only when the rods are fully inserted).
procedural and equipment changes provided improved communication, control of fuel changes and visual indication on the status of the control rods. Additionally, the licensee requested and received an Organization of Test, Research and Training Reactors review (Review 92-02, Docket 50-2) to prevent recurrence and to strengthen the overall operation of the facility.
Page 10, under the event description, second paragraph, first sentence, should the phrase, "at the reactor" be "as the reactor"?
The description of the University of Virginia event of June 13, 1994, is not a reactivity control event or an RPS event.
The event involved an experimental material protection scram being inoperable and did not relate to reactivity control except that a scram signal was involved.
The scram is not required by the TSs and is not considered in the SAR for accident mitigation.
It is there to prevent minerals that are being neutron-irradiated for coloring purposes from self-annealing, which removes the coloration.
The inclusion of this as a reactivity control event does not seem appropriate.
Further, the event has no direct RPS implications either.
The event does have operator training and supervisory activity implications, which as reported in the Survey were dealt with by the licensee's corrective actions.
On page 11 and 12 for the University of Virginia April 28, 1993, event, the licensee's analysis and corrective actions should be discussed.
"The source range channel was not part of the Page 13, states that:
reactor safety system, but provide improved monitoring of low neutron flux levels at startup to ensure that subcritical multiplication and criticality can be observed." The sentence should be changed to:
"TSs do not require the source range channel for reactor scram protection, but TSs do require the source range channel to monitor low neutron flux levels at startup to ensure that sub:ritical multiplication and criticality can be observed."
On page 14 for the MURR March 16, 1993, event, all the licensee corrective actions should be presented.
On pages 14 and 15 for a MURR September 7, 1993, event, the second paragraph of " Licensee Corrective Actions" were not part of the event report or the licensee's corrective actions. _.
t e
NRR Comments _on Preliminary Case Study NPR Survey Page 16, first full sentence states that:
"A series of seven operational failures of the regulating blade that necessitated manual reactor scrams were reported from 1988 to 1996, because of deficient preventative maintenance and corrective actions."
As presented, inadequate corrective action is the primary cause of the April 26, 1994, event.
However, the other events are described as related to component failures, except for one of the events, that of January 23, 1996, which is attributed to an inadequate procedure. Most of the component failures are of different components.
In the September 21, 1988, event, a set screw was loose on the gearbox.
In the November 2", 1988, event, a gearbox output shaft sheared.
In the June 3, 1989, event, a chain was off the gear on a limit switch assembly.
In the November 24, 1992, event, a set screw was found loose in the gear box (something that should have been corrected after the September 21, 1988, event).
In the December 27, 1995, event, a dowel failed in the gearbox.
Although deficient preventive maintenance can be the cause of any component failure, the evidence does not show that the license should have foreseen these failures or tracked them differently over the 8 year period of occurrence.
Nor does the information provided suggest that the license should have had a different preventive maintenance program.
It should also be noted that this system is not safety-related nor is it part of the scram system. Accordingly, the sentence should be changed to:
"A series of seven operational failures of the regulating blade that necessitated reactor shutdown or power reduction were reported from 1988 to 1996.
The primary root cause was component failure for five of the events, an inadequate procedure for one of the events, and inadequate corrective action for another of the events."
page 17, under " Licensee Corrective Actions," "The licensee considered changing the Technical Specifications so that a failure of the regulating blade was not automatically a deviation from the LC0 and would allow for a timely reactor shutdown as an action statement, to alleviate the generation of a licensee event report." The corrective actions to date included replacing the component that failed, but has not included changing this TS.
Licensees should consider many different actions.
To date, MURR has not followed up on this alternative.
This should be recognized in the Survey.
Also, the Survey should include all licensee corrective actions.
A similar comment should be considered for page 21, after the first sentence of " Licensee corrective Actions" and for page 23, the last sentence in "Licensen Corrective Actions."
On page 19, the MURK event date should be November 4, 1992.
Page 26 of the appendix discussed a "MTR Reactivity Control Event" at Georgia Institute of Technology (5 MW) NPR on February 15, 1994.
The operator operated up to 0.5 MW without equipment protection scrams that were not required by TSs.
The operator failed to follow procedures.
It should be mentioned that these scrams don't provide mitigation of accidents in the safety analysis and that the event was reportable because of the procedural non-compliance.
Also, an enforcement action was issued for these violations of procedures.
Therefore, the last -
- ~ - - -. -
'a -
NRR Comments on Preliminary Case Study NPR Survey sentence under event description should be changed to:
"Although these scrams were not required by the facility TSs and provided no accident mitigation in the safety analysis, the event was considered reportable fo~r the inadequately implementei procedural controls (violations were issued for the associated problems)."
It also should be considered tha'. the operator was promptly removed from licensed duties and that the licensee subsequently determined that he should not return to licensed duties.
Further, this event was not an "MTR Reactivity Control Event" or an RPS-related event.
On page 33, under " Event Description," second paragraph, last sentence, the word "wa" should be "was."
Page 41, the first sentence of " Event Description," says "the reactor operated without a core low flow scram for a short period of time."
This implies that this scram was required at the time.
In fact the scram was not required as an adequate number of redundant low flow scram channels were operable.
The phrase should be changed to "an operator error resulted in an unplanned scram."
1 7
m 4
v NRR Comments on Preliminary Case Study NPR Survel Appendix SPERT IV test results (Ref. 6 in the Survey), which is quoted, in part:
' In a series of 10-msec transients with flow rates of 1.2, 2.4, 6, and 12 ft/sec, it was clearly shown that increasing the flow rate increased both the amplitude and the frequency of the oscillations.
However, both the fuel-plate temperatures at time of peak power and the maximum fuel-plate temperatures decreased with increasing flow rate....
A series of experiments for determining the threshold of instability was then begun in which an operational definition of the threshold was taken to be the appearance of power oscillations of amplitude 50% of the mean power.
The tests were performed for the following sets of initial conditions:
(1) ambient temperature,18-ft head, no forced coolant flow; (2) ambient temperature, 2-ft head, no forced coolant flow; (3) ambient temperature,18-ft head, and 1.2 ft/sec coolant flow forced through the core.
In these tests the reactivity was added to the core in small increments until the desired mean power or the desired total reactivity addition was attained.
For the first set of initial conditions, it took about 4 dollars and 10 cents of reactivity above low-power critical to obtain 50%
oscillations, which had a frequency of about 4 cps at a mean power of 7.5 Mw.
No evidence of incipient divergence in the amplitude of the oscillations was found, although the amplitudes varied somewhat, with a maximum of about 60%.
With the second set of initial conditions, a reactivity insertion of 4 dollars and 10 cents produced an oscillation of about 2 cps, which became divergent after a few seconds.
The amplitude of the last power peak just before the scram discontinued the test was about 110% above the mean power level of 3 Mw, and the natural-circulation inlet water temperature to the core had risen from ambient to about 70*C.
Under the third set of initial conditions no oscillations as large as ISO % were obtained, since before this point was reached the 4
temperature of the narrow-channel fuel-plate surfaces exceeded the predetarmined scram level of 300*C.
For example, during one attempt when the mean power was 10.5 Mw, after a total reactivity addition of I dollar and 60 cents, the temperature of one of the
)
thermocoupled narrow-channel fuel-plat surfaces increased from 140 to 385'C and caused a scram.
During the time the test was running, power oscillations averaged only about 15%.
i 1 I
~.
4 NRR Comments on Preliminary Case Study NPR Survey 100-16634, "SPERT Program Review", pages 11-12 states:
2.
Ramp Tests In addition to the step transients in Spert I, a number of ramp tests were performed in which reactivity was
- )
added continuously at a constant rate to the just-critical system, since this form of accident initiation is more typical of the types of accidents likely to occur in actual practice. The ramp tests were the first departure from the types performed in Borax.
In these tests the parameters of particular importance are the initial power and the rate of d
reactivity addition (ramp rate).
During a ramp transient the rate of logarithmic power rise, a, is initially zero, increases to a maximum, and then decreases to zero again at the time of maximum power.
Two readily determined indices of ramp transient behavior are the maximum power and the maximum in the rate of logarithmic power rise, a,.
The ramp experiments on the first core included ramp rates from 0.01% Ak/sec to 0.35% Ak/sec :nd initial powers from j
10"' watts to 105 watts with the following results.
~
j a.
The initial power is relatively unimportant...
b.
The peak power is approximately proportional to
~
the ramp rate, which makes the ramp rate the dominant factor in accidents of this type.
c.
If a ramp burst is characterized by the maximum value of o, it is essentially equivalent to a step burst having the same value of a...
d.
The highest ramp rate used, about 0.35% Ak/sec, was estimated to be an order of magnitude below that which would lead to core damage during the first burst.
e.
Although in every case the first burst during these ramps was safely self-limiting, continued withdrawal of the rods led to violent power oscillations which rapidly grew to destructive proportions necessitating reactor scram to prevent core damage.
3.
Instability Tests A series of instability tests was undertaken to investigate more fully the violent oscillations observed during the ramp test.
These instability tests were conducted by injecting a predetermined amount of reactivity at a modest rate and observing the reactor power behavior after the injection was completed.
For small reactivity injections the reactor operated stably at an equilibrium power determined by the reactivity added above the zero !
. - _ =
, =.
4 s
t i
o*
NRR Comments on Preliminary Case Study NPR Survey 4
d power critical conditions.
For larger reactivity injections unstable behavior developed and the reactor power went through a series of power bursts. About 50 instability i
tests were performed with the following results.
l a.
Large power oscillations appeared whenever the reactivity held in the form of moderator voids exceeded 4
1.5%ok.
i i
b.
These power oscillations were an order of 2
magnitude larger than those observed in Borax and frequently i
i approached the largest bursts observed in the step tests.
Thus, in some situations instability may constitute a serious hazard potential.
j 1
c.
The moderator voids collapsed rapidly, and essentially completely, prior to each power pulse.
The fractional collapse of voids in these tests aprare<1 to be much greater than that reported from Borax.
4-i d.
The interval between bursts was about 0.3 sec from test initiated from room temperature and about 0.9 sec for j
boiling test.
j e.
The tendency toward oscillation increased with the i
amount of reactivity above zero power critical and with the depth of water over the core.
f.
Reproducibility of detailed behavior was poor.
The sequence of oscillations appeared to be random although some regularities were observed.
In some cases sustained oscillations of roughly constant amplitude were observed.
In other cases the oscillations died out for as long as half a minute and then reappeared.
In several subcooled tests t
with 9 feet of water over the core the power peaks increased abruptly from sustained pulses of about 200 Mw peaks to i
erratic pulses exceeding 2500 Mw peaks.
g.
The mode of reactivity addition was not responsible for these oscillations since essentially the same results were obtained by incremental rod withdrawals to the final position in which the reactor power was allowed to come to equilibrium between increments.
h.
Long-term instability and the effect of the hydrodynamics of the system could not be satisfactorily investigated in Spert I.
l C
. _ -