ML20096D591

NRR Technical Newsletter.Volume 3,Number 1

ML20096D591
Person / Time
Issue date:	03/31/1992
From:	Office of Nuclear Reactor Regulation
To:
References
NUREG-BR-0125, NUREG-BR-0125-V03-N1, NUREG-BR-125, NUREG-BR-125-V3-N1, NUDOCS 9205180138
Download: ML20096D591 (9)

Text

-

1

)

_m m

.E

.E E

m.

e TECHNICAL nunmenos NEWSLETTER E "e Reviewing Advanced Light Water Reactor Designs

~

T. E. Murley The NRR staff has entered a period of intense activity in years has added to our understandmg of reactor safety as reviewing the designs of the next generation of light water well.

reactors. The resuhs of our reviews will set the standard for the safety design of nuclear plants to be built and oper.

Because we are usin, nat I d both a top-down review ated in the United States well into the middle of the next and a bottom-up resiew app mh, the result is expected century.

to be an integrated, balanced design review. The top-down review comes from our requirement that the designer feed Part 52 requires that the NRC make final safety judgments back the insights from his probabilistic risk assessment on a design at the time of cenificatica, before a plant has (PRA) into the design process. This has already led to a been ordered and constructed. The rule funher requires number of design changes to improve the plant's ability to that the NRC make final safety judgments on site-related cope with core melt accidents, for example.

maners before issuing a combined operating license (COL). These requirements place a greater level of disci.

The bottom-up re.iew deals with specific issues that hase

- pline upon the staff reviewers than was the case under the ansen with the current plant designs. Some of these spe-i two-step licensing process for Part 50, because the staff cific safety issues are:

cannot defer making a judgment on an issue in the expec-e Fire protection will be assured through better tation of making a final decision when the plant is under separation of safety systems.

construction.

j A further need for more discipline in the review process Intersystem LOCAs can be prevented by design.

e under Part 52 is that the staff must specify the inspectiors, tests and acceptance criteria (ITAAC) that are to be used The risk from accidents during shutdown operational e

to verify that the conditions of the COL are met during modes can be reduced by design.

construction. This is complicated in some areas in which e Defense m depth can be enhanced by considering practicallimitations have prevented the vendor from com-re melt accident mitigation features in the design.

pleting the design. In these areas, the staff will have to make final safety judgments by specifying design accer A reliability assurance program can integrate require-e tance criteria (DAC) to be verified by the staff after the gg g

COL is issued. The concept of design acceptance criteria spection and in-service testing with the technical is new to the staff, and it will require mnovative thir, king.

g g gg as well as discipline, to make it work' of equipment (assumed in the PRA during design) will in spite of these challenges, I believe the staff reviews will be maintained throughout the operating life of the be more thorough than those conducted in the past and plant.

will resuh in safer designs. We have a great deal more ex-perience now, because we have the benefit of some 1600 (Continued on next page) reactor-years of operating experience in the U.S. Many problems in the early designs have been revealed through operating experience and have been dealt with through Articles not appearing in this issue will appear in the backin regulations (e.g., TMI, ATWS and Station Black-April issue.

- out rules), A great deal of high quality research over the 9205180138 920331 PDP NUREO 1

BR-0125 R PDR

l 1

suppress 1GSCC. reactor coolant conductivity must be mairitained below 0.3 rnicrosiemens per centimeter, and IN THIS ISSUF~

sufhcient hydrogen must be added to the feedwater k. te-duce the ECP below -0.23V (standard hydrogen elec-h radiation field results in Reviewin8 Advanced Light Water Reactor Designs trode). Excess hydrogen in a hi{idrogen with radiolytic de-by T. E. Murley..............

I the catalytic recombination of h composition products (O and H 0 which cause IGSCC) 2 2 2 i ref rm wan BWR Hydrogen Water Chemistry and Zinc injection Passivation in the United States by Frank J. Witt..

2 The feedwater hydrogen injections required to estabbsh HWC ECP conditions vary between plants and between Industry Reports for License Renewal regions within a plant. At the Hope Creek Generating Sta-by P. T. Kuo........

4 tion and the Nme Mile Point Nuclear Station, the concen-tration of oxygen in the reactor coolant is reduced to 1 ppb Indian Point Unit 2 Steam Generator Girth by maintaining a feedwater hydrogen concentration of 0.6 Weld Cracking ppm and 19 ppm, respectively. The amount of feedwater by Herbert J. Kaplan and Alfred Lohmeier 5

hydrogen and resulung 0 in the recirculation water de-2 pend primanly on the tndmdual design characteristics of Reviews of Recent Epidemiological Studies of the reactor and can be correlated with the downcomer Radiation Risks dose rate and the mass flows in the core regions. A Gen-by Frank J. Congel and Charles A. Willis 6

eral Electric.Harwell BWR water radiol > sis model predicts the concentration of radiolysis products in various regiore, Electrical Distribution System Functional of the BWR primary system. Comparing Oa levels in the inspections Program Review and Lessons reactor plant recirculation system with this model indicates Learned s easonable agreement (References 2 and 3). The by Anil S. Gautam.

8 radiolysis model shows the different regions of the BWR respond differently to hydrogen additions and may require NEWSLETTER CONTACT:

increased concentrations of hydroter, in the reactor cool-Anna May Haycraft, NRR, 504-3075 ant to attain the -0.23V ECP needed to mitigate IGSCC in the reactor's internal structures. Utihties have measured the ECP in the reactor core to determine the quantity of hydrogen that should be added to suppress 1GSCC.

When we say that we expect the Advanced Light Water Reactors will be safer than current plants, there are three At the Edwin 1. Hatch Nuclear Plant Unit 1 in the U.S.

aspects to that assertion. First, the PRAs will show the pre-and Oskarshamn Unit in Sweden. copper sons in the cool-dicted core damage frequency to be lower for advanced ant prevented the HWC from yielding the results predicted plants because of less frequent challenges and a better by the BWR water radiolysis model These copper tons ability of the plant to cope with trie challenges. Second, reduce the effectiveness of the hydrogen recombinat:on the improved mitigation features included in the design with radiolytic decomposition products. These plants have will improve the containment performance under severe Admiralty brass condensers coupled with filter /

core damage accident conditions. Third, the dcsigns are demineralizer condensate cleanup systems which typically expected to reduce the number and complexity of actions rek, e into the coolant copper concentrations in the vequired of the operators, thereby reducing the chances of 15-30 ppb range and rine at concentrations in the 2-10 human errors and increasing our con'idence in the predic' ppb range. At Hatch Unit I, the licensee has replaced the tions of better prevention and mitigation of accidentt.

condenser Admirality brass tubes with titanium tubes. This action reduced the copper concentration in the feedwater.

BWR Hydrogen Water Chemistry which significantly decreased the hydrogen concentration and Zine Injection Passivation in in the feedwater which redved rectrulation caolant ECP '

t less than -0.2?v. Before the condenser tubes were re-the United States placed, sufhcient hydrogen could not be added to mitigate Frank J. Witt, DET IGSCC (-0,23V ECP) because of excessive N-16 radiation levels measured by the main steam line radiation monitor (MSLRM).

Hydrogen water chemistry (HWC) is a countermeasure to 4

mitigate intergranular stress corrosion cracking (IGSCC) in reactor recirculation piping and reactor internal compo-When hydrogen is added to the feedwa:er tc yield the nents in boiling water reactors (BWRs). Many U.S. unli-proper HWC, the N-16 steam activity increases, resulting ties are adopting HWC for their BWRs. with 8 units oper-in higher MSLRM levels. Because hydrogen dissolved in ating permanently with HWC,12 units having completed the reactor coolant reduces the oxidizing potential, the preimplementation test programs. and 9 c:her units install-N-16 is converud from a nitrate to a more volatile ammo-ing permanent equipment. The water environment con-nium ion. increasmg the amount of veam N-16 activny.

tributes in two primary ways to IGSCC: the oxiduing To address this problem, the licensees for some BWR power of the water indicated by the electrochemical po-plants have increased the MSLRM set points above the tential (ECP) of stainless steel and the concentrations of values specihed in the technical specifications to allow ionic impurities, particularly sulfate and chloride. To HWC.

2 i

r

7 The Electric Power Research Insutute (EPRI) and egen-The been:,ees can manage the increased occurrence of enced industry persormel hae deseloped generic guide-shutdown dose rates attnbuted to ilWC by performmg hnes for designing, it stalhng. and implementmg perma-normal radiauan protecuan procedures m;1udmg usmg I

nent HWC at BWR p, int sites. These guidehnes chscuss temporary shieldmg and chemically decontaminaung recir-

)

recommended safety is atures and address possiNe in -

culation pipmg. Although HWC may increase person rem, creased N-16 de e rates tnd fadure modes ci sptems that such an mcrease would be relatnely minor compared to store and handle hydroge a on the plant safety systems = lo the amount of person-rem exposure from the mspections present damage to safet/ related structures, the beensees and repaus needed to mamtam overla) welds and to rt-provide separation d:stanc:s of hydrogen storage fachues place the recirculanon p pmg associated wah p:pe cracking (hquid and gaseous Hg) to protect agamst a storage faciht) m BWR piants operating with normal water chemistry. The explocon. Separation d: stances are also prodded to pre-BWR Owners' Group has submitted a topical report to the vent flammable gas mixtures from form;ng at safety-NRC Jusufpng inservice inspecuon (ISh credit f or piants d

related air intakes if a pipe brecks at a s'orage f acthty The operating wTth HWC. If the NRC appros es this imualise, it NRC has approsed these guidehnes f or the beensecs to use would allow bcensees operaung ilWC pbna to reduce the in safely implementing HWC, frequency of 151 inspecuans, with an accompanpng re.

L ducuan in person-rem Therefore, ISI cted:t would offer adiuonal mt entive to implement HWC.

To manage the mcreased dose rate from the HWC-enhanced production of N 16, the hcerzees can perform The comrol of the budd up of rad:auan in reactor sys: ems an initial and conunuous radiauon sune) propam and has been of concern m BWR plants. GE renewed operat-change the procedures fer plant shieldmg and mamte~

ing plant correlaum and found that pbra h ntng 5 to 15 nance. N 16 has a half-hfe of ~.1 seconds and thus causes ppb of soluble tinc in the reacter water f at a resuh of the the raianon level to increase only upstream of the main uruque plant design parameters dacussed prevously for condenser. These changes to programmauc procedu:o m Hatch 1 and Oskarshamnn-1) had loaer pipmg dose rates addmon to the normal procedures for plant rad.auon pro-than piants that had onh trace amounts of nnc. I ahora-tection, are sufficient to ensure that. while the bcensee "

ton tests conbrmed that usmg soluNe zmc causes truch addmg hydrogen, the plant wall continue to meet the re-less Co-60 to be deposited in p: pes haung both normal quirements of Part 20 of Title 10 of the Code of Federal and h drogen water chemntnet These studies led GE to 3

Regulations (10 CFR part 20) and the recommendat ans deselop the General Electnc Zmc Inyecuan Paunauon of Regulatory Guide 5.S. "Information Relesant to Ensur-(GEZIP) process m wluch small quannues of zmc are m-ing that Occupauonal Radiation Exposure at Nuclear iected to the fmal feedwater. Currentt, eght U S. plants 3

Power Staucra Wu! Be As Lcm As Rea*onaN> Achte' huve implemented nnc mjecuan anJ four ether pbnts wdl able.

stan zmc injection in a few months. Dese rate measure-mems made at several of these plants indicate syruficantly lower lesets of Co+0 than are found mt plants not usmg in adition to reportmg increased amoums of N 16 in tne zmc. As a side ef fect of mjecung nnc, the activated iso-main steam hne, bcensees have recent!v reported m-tepe Zn45 (half-hfe 244 daysi has become a sigruhcant creases in shutdown dose rates m some U S. HWC plants contnhutor to the dose rates from reactor water and The increased shutdown dose rates pnmanh result from pipmg. To enhance the benefits of mjectmp eine, GE is Co-60 that is deposited m the corrosion fdm or p: ping and deselopm; methods to produce the Imc oxide that is component, and in hot spots in crud uaps Th:s increase depleted m the precursor Zn-64 Tht reference water varies among plants where an increase in soluNe Co-60 chemistry far the adsanced baihng water reactor is HWC and the spikm; of msoluNe crud (Co 60,1 D and 133 and GEZIP. The NRC conunues to rnonitor the imple-Mev: Mn-54, 0.54 Niev; Zn-6 5, l.h Mo) have been ob-mentauon of HM C and GEZiP and their effectneness n served These dose rates probaNy mcrease because HWC mm;aung IG5CC and reducing the ouildup of shutdown reduas the corrosion potenual. This reducuon resuhed m radiauon levels a restructunng of the oxMe corrosion deposits on surfaces inside and cuiside the core. Such a restructunng would RPIC"CCS result in the release of parttulate and soluble Co+0 The General Electric Company (GE) bebeves that the effect 1.

Electnc Pow er Re searc h

Insutute, EPRI may be temporary but may last for more than one cg:e S pg _ g g,.' B W R Hvd w n Water Cnematry To correct trus problem, the bcensee for the Edwin 1.

Omdehne: 19" Fev;sion," dctober lcSC Hatch Nuclear Plant, Umt 1, used adinonal temporary lead shielding during the last refuehng outage and n con-2 R. L Cowen. k, G. Head, M. E. In g, C. P. Ruu, sidenng chernically decontammau% the recirculation pip-and J. L. Simpson, "U S. Experience with Hydrogen ing to maintain an as-low -a s-re a sonaN y achievable Water Chem:stry m Boihng Water Reacton,' GE Nu-(ALARA) program Utihties that oprate Swedah HWC clear Energy Energy, San Jose, Cahtorma, Pmceed-plants have not observed increases m the shutdown dose mgs of Japanese Atomic Industnal forum. Water rate because these plams base mmirnmd the cobalt con' Chemntry Conference, Tokyo, Japan, Apnl 1L22.

tent in matenals that contact the reactor conlant To rwa-m, mize the accumulanon of radmactne matends that couH cause a shutdown condinon. the hcensees tot operating 3

C.P.Ruu C. C.Lm R. N. R@uxn, W. G Nrn, plants should mmimue the cobalt m replacement maten-A. R ; Curto "Wdel Calculaoun of Water Radiohsis ais, and the designers of advanced reacton should ehrn in BWR Pomary Cn!am," GE Nuclear Er:ergy, San nate cobalt where possMle.

Jom Cahforma, U K AE A Harweli Laborator), Umted 3

Kingdom, ARC Scientific Ltd., United Kingdom.

Industry Report SER Date Procee6ings of Water Chemistry of Nuclear Reactor Systems V. Bntish Nuclear Energy Society, Paper 33, Cable Inside Containmer t 1/93 London, England, October 1989.

Class 1 Structures 6/92 BWR Containment 8/92 4.

Electric Power Research Irntitute, EPRI-5283-SR-A, BWR Reactor Coolant System 11/92

• Ouidelmes for Pertnanent BWR Hydrogen Water BWR Reactor Vessel 11/92 Chemistry Installations-1987 Revision." September BWR Reactor Vessel internals 11/92 1987.

PWR Containment

$/92 5.

W. J. hiarble and R. L. Cowen, ** Status of Zine injec-hR C< ol. nt System 8/9 tion Passivation at U.S Boning Water Reactors." OE PWR Reactor Vessel Internals 3/93 Nuclear Energy. San Jose, Cahfornia, Proceedmgs of Screening hiethodoSgy 10/92 Japanese Atomic Industrial Fomm. %ater Chemistry Conference Tokyo, Japan, April 19-22, 1958, The staff found that the technical justificanons in these Industry Reports For License Renewal irs generally did not support concluuons on manarmg age-related degradation. These justafications were not sul-P. T. Kuo, PDLR ficient because the industry prepared the irs (1) before the preposed 10 CPR Part $4 was pubbshed, and (2) were based on an unsubstantiated assumption that existing pro-On Decerrber 13,1991, the Commission pubhshed the 10 grams were, by defimuon, c:fective in managing apng.

CFR Part 54 rule on heense renewal. Currently, the Com-Nevertheless. these irs contain valuable mformation con-mission issues bcenses to operate nuclear power plants for cermrig the effects of age related de gadation on SCs. The a fixed period of time not to exceed 40 years. However, staff is using this informatmn in developing the resiew and the Commission may renew these bcenses to extend the acceptance entena that will be incorporated into the regu-period of operation up to 20 years under the regulatory latory guide and the standard review plan on hcense re.

requirements set forth in 10 CFR Part $4.

newal (SRP-LR).

De regulatory philosophy ar# approach for the 10 CFR To mmtmize unnecessary revisions of the irs and to focus Part 54 rtle are based on two key principles. The first resource

• on the techmcal usues in each IR, NUhl ARC pnnciple is that the regulatory process is adequate to en.

and the staff agreed to initiate the followmg review process sure that the licensing bases c' all currently operatin:

in revismg the irs:

plants provide and maintain an acceptable level of safety 1.

NUNIARC submits the IR for operation. The second principle is that each plant's current licensing basis (CLB) must be maintained dunng 2.

The staff reviews the IR and provides written com-the renewal term. To maintain the CLB during the period me nts.

of extended opeiation, the licensee may need to imple-ment a program to manage the age-related degradation of 3.

NUNIARC reviews the staff's comments and submits a systems, structures, and components (SSCs) that are im-written response to each comment.

portant to license renewal.

4.

The staff reviews NUM ARC's responses and deter-To assist the NRC in renewing licenses, the nuclear indus-try is developing a series of technical reports (industry re-5.

The parties hold a pubhc meeting te discuss any com-ports (irs)) on generic resolution of, aad programs to ments which require additional resie w and response manage, the age-related degradation of a vanety of struc-and to discuss any issues regarding the IR under re-tures and components (SCs) irrportant to license renewal.

view. (This step may be repeated if necessary).

The irs are coordmatec by the Nuclear Management and

)

..UMARC submits the revised IR.

6.

a Resources Council (NUMARC) and sponsored by the U.S. Department of Energy (DOE) and the Electric Power This cooperative process has enabled the staff and Research Institute (EPRI). In the irs, the industry evalu-NUMARC to resolve many technical issues identified in ates the effects of age-related degradation on specific SCs, the irs. The remaimng steps in the IR review process will describes the bases for the manner in which exist'ng regu-be to review the revised irs after they are submitted by latory programs address the degradation concerns, and NUMARC. resolve any technical is.ues that remam and recommends specific corrective actions for specific SCs prepare an SER for each of the irs. The draft SERs will not presently addressed by effective age-related manage-be reviewed by the Committee to Review Generic Pequire-ment programs, ments (CRGR) and the Advisory Committee on Reactor Safeguards (ACRS). The staff will pubbsn the draft SERs By October 1990, NUMARC had submitted 111Rs for the for public comment before issuing the final SERs.

staff to review. The industry has written these irs on sub-jects affecting both boihng water reactors (BWRs) and The irs hase been reviewed by the staff from the Office pressunzed water reactors (PWRs). The staff expects to of Nuclear Reactor Regulation and the Ofhce of Nuclear complete its review and issue safety evaluation reports Regulatory Research and by expert consuhants from (SERs) on the irs as follows:

several national laboratones. The staff has provided an I

4

l average of 60 technical comments on each IR, NUMARC during pubhc meetings, An AIP is a general agreement has provided responses to staff's comments anc met with between NUMARC and the staff on the manner in which the staff to discuss the responses on all irs, except the 151 to resche a particular comment and in& cates that on PWR vessel internals. The staff and NUh1 ARC have NUMARC will revise the IR to be censistent with the reached an agreement in principle ( AIP) on approxi-agreement. The following hst provides the current disposi-mately 65 percent of the comments on the irs by review-tion of comments provided to NUMARC-ing of NUM ARC's responses and discussing these i sues Number of Number of Remeining Industry Report Ccmments Comments Cable Inside Containment 114 IB Class 1 Structures 99 5

BWR Containment 128 8

BWR Reactor Coolant System 50 BWM Reactor Vessel d7 13 BWR Reactor Vessel Internals 51 3

PWR Containment 95 14 PWR Reactor Coolant System 45 9

PWR Reactar Vessel 49 25 PWR Reactor Vessel Internals 65 No Meeting Yet Screening Methodology Reevaluaura per 10 CFR Part $4 Ahhough the numbers by themselves do not indicate the Indian Point Unit 2 Steam Generator extent of acceptabihty of the irs, they do indicate the in-Girth Mreld Cracking tensive effort expended by NUMARC and the staff m the IR review process.

Herbert J Kaplan and Alfred Lohme!er The staff is reviewmg the revned irs that NUM ARC recently submitted on PWR containment, DWR contain-Inserme inspection of the upper shell to transition cone ment, and Class I structures.

girth welds durmA the M9 spnng outage of Indian Point Unit 2, revealed circuinferential cracks in the weld and The IR review process enables NUMARC and staff *3 heat-affected rone material of four steam generatert.

focus their efforts on resching open issues m a ynematic When girth weld crack;ng was discovered in these steam and tecnnically sound manner. By employtr.g this reuew genernors dunng the 1937 outage, the licersee removed pmcess, the staff and NUMARC have identified several the defects by grinding arad did not repair any of the welds technical issues that must be resolved. These issues in-by reweldmg.

clude the evaluation of fatigue in metal components, the ennronmental qualification (EQ) of electncal equipment, The licensee removed metallurgit.al semples during the 1989 outage an! determmed that a corrosion fatigue and the thermal embnttlement of cast stainless tteel com-ponents. These are complex issues for which the design mechanism had caund the cracking The corrosion re-requirements for the currently hcansed plants vary consid.

suhed from cold ieedsater injection in the transition cone etably. By interacting tog 3ther, the indust y and the s aff gon during startup. plant operations, hot standby, and have more clearly defmed the npects of these issues thit other plant trensient modes of operation.

pertain to age elated degradation. This interaction as-Tlus time, the licensee used an automatic tungsten inert seted the staff in developing draft branch technical posi-gas (TIG) process to remove and repair weld the defects.

tions (ETPs) for fatigue and EQ to be considered for in-All girth welds were subjected to a 4-hour,1125' F local corporation into the SRP LR*

stress rebef treatment.

Whih. the staff has not completed its review of these irs. it f:ollowmc magnetic panicle inspection, the licensee found found that they contam valuable information on the ef-more indications in previously ground-out cavities, in weld facts of age-related depadauon on SCs. Once approvert repair areas, and within the original weld metal. The licen-by the staff, the irs will pronda the genenc techical Ms see ground out and rrtested these defects and returned for evaluatmg the effects that age-related degradation the plant to sernce.

couid have dunng the license renewal term. The irs pro-nde for a single review and aporoval with subsequent ref-In addition to the weld mpairs and heat treatment, the crencing, rather than repeutive rede us of the same sub-bcensee removed the utam generator downcomer resis-ject. By referencing the irs, heense renewal applicant will tance plates and inswiled a timer to delay feedwater by-enable itsell and the staff to focus their resources on re-pass valve closure to lesen the thermal shock effect dur-solving for specific plants the hcense renewal technical is-ing cold feedwater injection.

sues which are identihed in the irs as requiring enhanced j

plant-specific aging manag: ment and which are outside Dutmg the midcycle inspection after the 1989 startup, the the stone of the irs hcensee discovered more indications in the weld metal of 5

m--..

)

i

~

the four steam generator girth welds. The licensee again stores. The pubhe controsersy, legal actions, and volun-used the TIO process, for repair welding, but without the tary control measures combined to eliminate most of the 1125*F post weld heat treatment. Instead, the licensee gross misuses of radiation by the end of World War II.

epplied a 500'F post-weld treatment to produce a tem-pered heat-affected zone-The development of the atomic bomb prosided new impe-tus and funding for radmbiological research. In ?956, the The licensee, assisted by several consuhants and research National Academy of Sciences declared that radiation was organizations, concluded that the principle cause of crack the best understood environmental hazard, Research has inhiation and progression was related to the dissolved oxP continued and, today, radiation risks are very well under-gen :ontem in the secondary water, which acted on sus-stood, llouever, we have not yet determined the magni-ceptible materialin a high-tensile stress held. The hcensee tude, if any, of risks from exposure to low levels of radia-then machined a 360-degree 6-mch wide by 3/4-inch tion (such as less than about 10 rem per year).

deep groove around the girth welds of the steam genera-tors and fdled the grooves with a low sulfur weld metal' followed by an 1125,F heat treatment. The beensee als If the mechanisms of radiation injury were known, the i

M low level exposure could be an-wtll maintain low oxygen levels by using a nitrogen blanket in the condensate storage tank and will reduce the ternile swered with laboratory investigations. Smee the mecha-stress by prosidmg a smooth radius at the shell/ conc junc.

nisms are not known, epidemiological studies are con-tion, The licensee s actions should provide a basts for con-clutted to try to reduce the degree of uncertaint).

tinued safe operation of the steam generators.

Current risk estimates for low-lesel exposure are based i

Reviews of Recent Epidemiological pnmarily on the results of epidemiological studies of the sumvors of the nuclear weapon detonations at itiroshima Studies of Radiation Risks and Nagasaki. These resuits are supported by studiet of other highly irradiated groups such as the radium dial Frank J. Congel and Charles A. Willis painters, patients irradiated as a treatment for ank> losing spondylitis, and women irradiated as a treatment fv cersi-Introduction cal cancer. Where doses are high (above about 50 rem)

The NRC is responsible for protection of the pubhc and cancer rates are increase 1 For example, the cance rates workers from the ill effects of exposure to ionizmg radia, for the survivors of fliroshima and Nagasaki apparently tion. To meet this responsibihty effecovely, the NRC were increased about 5 percent.

needs to understand the magnitude of the nsks associated with radiation exposure. Thus, when new studies are re-Researchers have conducted numerous stuc'ies of groups ported that purport to cast new light in this issue, the staff receivmg lower doses, Fut the results are inconclusive.

examines them carefully. The staff recently reviewec' live Generally, they have found no increase in cancer rates, new epidemiological studies of radiation risks and the re-even in Gunapari, Brani, where 12,000 people receive sults of those reviews are summanred here.

doses of about 0.64 rem per year, which is about $ times the aserage backgrouno dose; in Kerala. India (0.35 rem

Background

per year); or in T anjiang County, China (0.3 to 0.4 rein Everyone is exposed to radiation at all umes. This has al-P F*'b ways been true, although no one knew about it until 1595 when x-rays were discovered Radiation injunes were re.

No radiation-induced genetic effects base been observed ported only a few months after radiation was discovered.

in any human populauon.

Since that time, tidiation and its biological effects have I

been the subject of intense world-wide scientific insestiga-Epidemiological Studie:, Reviewed tion. The important effects were soon identified. Even the possibility t.f genetic damage was reported in 1911. The The five studies reviewed were conducted by the Nanonal fundamentals of radiation protection also were identified Cancer Institute (NCl) [1], the Three Mile Island Public within a few years of the discovery of radiation. The first Heahh Fund (TMIPfi) 12l, the Mas,achusetts Depart-person known to be killed by man-made radiation was ment of Pubhc Health (MDPH) [3l, Steve Wing, et al. [4, Thomas Edison's assistant, Clarence Dally By the time of 5); and Sternglass and Gould [6]. The populations investi-Mr. Dally's death in 1904 Edison reported that pmper gated differed in most respects between each of the studies precautionary measures had been developed and that "I and the mvestigators reached markedly differing conclu-would continue the work myself but my wife won't let stons. The extremes were the NCI and the Stemglass-l me

Gould studies. The NCf study reported no detectable ill j'

effect in the populations around any nuclear power plant Radiation protection measures were not always apphd or Department of Energy (DOE) facihty in the U.S. How-and, as a result, hundreds of people died of radiauon.

ever, Sternglass and Gould contend that effluents from the induced cancer and others suffered radiation injuries.

Trojan nuclear plant aie killmg thousands of people annu-Early injuries initiated public controversy beroce 1900. Ra.

ally in Oregon,

, diation injuries caused by the use of x-rays to mvestigate wounds during World War I contributed to the contro-We reviewed these studies and ccncluded that none of versy. Despite the controversy, radiation was misused.

them conymcingly showed any discernible effect of low-Misuse is exemplified by Radium tonics being sold through level radiadon or provided any reason to beheve that the the mail and fluoroscopes being available in most shoe NRC should revise its effluent control practi-6

f The Sternglass-Gould Report The total calculated populauon dose from Ngnm effluents was only 260 person-rem. The Natiorn! Academy of Sci-E. Sternglass has long had the reputation of being one of ences (BEIR-V) estimates that this dose cot.ld cause a to-the most uninh;bited of the antinuclear actnists, and J. M.

tal of from 0 to 0.05 cancer deaths. Furthermore, the Gould is rapidly developmg a similar reputation. In con-doses from effluents were only a small fraction of doses ducting their work, which was funded by a political group from natural backgrour d radiation. Thus, even if sety trying to shut down the Trojan plant, they concluded that high values are assumed for the rad.ation risk factors, the radioactive effluents are killing over 8,000 people each effluent doses could not have caused a discermble in-year in Oregon. Their buis for this conclusion is that the crease in leukemia.

death rate in Oregon dechned in the 2 years preceding the startup of Trojan. If this "tre y!' had contmued, the death Oak Ridge Workers Stud) rate in Oregon would be far below tts current value' Ste*nglass and Gould contend that the effluents from Tro-Steve Wing proclaimed on national telension that he had jan caused the difference between the actual and the pro-shown that radiation risks are 10 times greater than the Jected death rus. However, if this trend had cont n-National Academy of Sciences' esumates. Before the 2

ued, the present death rate in Oregon would be far below proclamauon, there m httle mterest in this work because the nauonal average and, in another 75 years, the hfe ex-the sponsor, DOE, said no ill effects were being found pectancy in Oregon would reach 1,000 years. Clearly, the DOE rpidemiologists expressed surpnse at both the con-logic and the conclusion lack substance.

clusion and the public announcement.

3 Our review was compbcated by the omissions from the The content %n by Sternglass and Gould is made even pubhcations: they did not contain eitber the data or a full more dubious by the Trojan's outstandmg effluent control description of the methodology. We were told that neither record. Radioactive effluetas have been so hmited that the ts available.

total calculated population dose from all releases through 1986 was only 1.0 person tem (NUREG/CR-25!0, Vol.

The population studied consisted of the white males hired 3). By comparison. the dose from natural background ra.

to work at Oak Ridge between 1943 and 1972. This popu-diation to the population m the sicmit) of Trojan exceeds lation excluded people who worked there for less than a 1 person-rem every 4 minutes.

month, those who worked at otbr nuclear facihtss, and those for whom the dose or

,,ographic data were in-complete. The researchers drew their conclusions from The Massachusetts Department of Public Ilealth Study the following reported observations: (1) the hfetime doses The MDPH study also is seriously flawed. This was a case-were quite low, with the average being only 1.7 sem and control study of leukemia (other than chronic lymphatic with only 321 people (3.8 percent) getting more than 1 rem; (2) the average death rate of these people was sig-leukemia (CLL), which other investigatior s show to be nificantly less than expected (based on the data for white non radiogeme) in people oser 12 years of age in the 22 male Americans), (3) the cancer death rate also was less communiues within 25 miles of the Pilgrim Nuclear Power Station. The MDPH identified " cases," people for whom than expected; (4) the leukemia death rate was 63 percent leukemia had b en diagnosed, from medical records and higher than expected; and (5) with the arbitrary selection og 3 lag time" (la'e it period), the death rate could be selected matching " controls" The researchers estimated correlated with dose.

~

relative doses and assumed the extent to which the cases had doses higher than the control group to be a measure in our review, we noted the following. The observed ef-of the impact of radiation. The MDPH researchers inund fects could not be related to radiation dose because the one time period in which the estimated doses were hifher measured doses were only a small fraction of the total for the cases than for the controls. The MDPH concluded doses and because extmsure to other carcinogens was not that radiation had quadrupled the leukemia rate in that taken into account. Se'cond, the higher-than-expected leu-period for the more lughly exposed group-kemia rate is a common sariation in situations where ra-diation exposure cannot be the cause. Third, resoning to in our review, we found the MDPH conclusion untenable an anomalous " lag ume" tends to invahdate any correla-for several reasons. First, the short duration of the in, tion found between dose and effect. Fourth, the short creased incidence of leukemia is inconsistent with the in.

time between completion of the study and publication pre-crease being radiogenic; that is, the clerat-d incidence ciuded meaningful peer review; one reviewer was not disappeared just when it would have been approaching a given time for even a cursory review. Fifth, the report con-maximum if it had been caused by radiation. Second, the tains obnous errors in the few instances in which numbers distribution of doses from effluents assumed b, the MDPH can be checked. Fmally, the pubhcation contains an anti-is totally inconsistent with the actual calculated doses from nalear discourse that indicates a nonscientific agenda.

effluents; the doses from natural background and other Therefore, we concluded that the Oak Ridge study did not radiation are ignored. Third, the rnethod used for deter.

constitute a basis for changing regulatory pracuce.

mimng the location of the people is highly inaccurate:

quesuoning a surviving friend or relative by telephone to TMI Public Health Fund Study determme where the person hved and worked many years la this study researchers investigated the population hving ago. Fourth, the leukemia incidence of the low due group within 10 miles of the Three Mile Island Nuclear Stauen was well below the average for the state. Fi th, the pre-(TMI) for childhood cancers, leukemia,1 mphoma, lung 3

sumed consequences are tota!!y mconsistent with the car er, and all cancers. The researchers estimated the ra-doses, based on generally accepted methods.

di tion doses from the TMI accident, from the norrnal f

7

l effluents and from natural background radiation. The re-

5. Wing, Steve, et al., " Mortality Among Workers at searchers four d exposure rates to vary from 50 to 90 mil-Oak Ridge National Laboratory: Evidence of Rad:a-lirem per year from external sources of natural back-tion Effects in Follow-Up Through 1984-Supplemen-ground radiation. The maximum exposure from the TM1 tary Document," National Auxiliary Publication Serv-accidert was less than 100 rnillirem and the exposure from ice documen 04849 (P. O. Box 3513. Grand Central normal effluents was much less than that salue. The Station, New York, NY 10163-3513) no date, authors acknowledged being unable to imd any effect.

6.

Sternglass, E. J. and J. M. Oculd,

• A Study of Recent while admitting thur disappointment, llowever, they Rtses in Leukemia and Other Mortality Rates in Ore-chottld have been expected to find no effects of exposure gon Following Radioactive Releases from the Trojan to plant ef0uents because efnuent exposures were less Nuclear Plant," Unpublished Manusenpt October 25, than variauons in natural background.

1990.

The National Cancer Institute Study Electrical Distribution System his study was a heroic effort, covering 51 nuclear power Functional Inspections Prograrn plant sites and 10 DOE facilhies for a 35-year period. The Review And Lessons Learned researchers divided cancers into 15 categories rnd ana-lyzed the populations in 5 year age groups. He research-Anil S. Gautam, DRIS ers based the study pnmanly on mortaht) data but also included incidence data where they were available. They Previous NRC inspection teams had identified generic de-calculated both standard mortality and relative risk ratios ficiencies related to plant electrical systems which had the and produced over 23,000 statistical tests for eval" tion.

potential for comp.omising plant safety rnargms during postulated accidents. As a result, the Special Inspections One especially important feature of the NCI study was the Branch (RSIB), in conjunction with the Electrical Systems inclusion of calculabons for the years before the plants Branch and the regions, developed guidance and tech-tnnt into operation, his clearly showed the variation in niques for inspecting plant electncal distribution systems the data that could not be the result of plant effluents. For and cenducted five pilct electrical distribution system example, of the six standard mortahty ratios for childhood functionalinspections (EDSFis) during 1989-1990 to vah-leukemia that were significantly greater than one, four oc-date these techniques. Based on the finding

• of the pilot curred oefore startup. Similarly, in those instances in inspections, RSIB developed Temporary Irstruction which relative risks were significantly different from one, 2515/107 and conducted a training ccurse for regional in-only four were greater than 1 while 14 were less than 1.

spectors. As of January 1,1992, EDSFIs have been com-Thus, even though the study included old DOE facihties in pleted at 31 plants. The ref.as are scheduled to corrplete which releases were relatively high during World War II, EDSFis at all operating plants by early 1993.

the NCI found "no suggestion that nuclear facihties may be linked causally with deaths from leukemia or other EDSFIs evaluate the capability of the electrical system to cancers."

perform its intended functions during all plant operating and accident conditions, and the effectiveness of the h-In reviewing this document, we note that the results are censee's engineering and technical support for the system.

what would be expected from current knowledge of re.

Inspection teams generally consist of two electrical design leases and radiation risks.

engmeers, one mechanical design engineer, two electrical f eld inspectors, and the team leader. The teams perform References the inspections on site and, when necessary for the design portion of the inspection, at the licensee's corporate enge 1.

Jablon, Seymour, Zdenek Hrubec, John D. Boice, neering office. The inspection team; use a venical slice and B. S. Stone, " Cancer in Populations Living Near sampimg techrique to evaluate one or more electncalload Nuclear Facilities," National Institutes of Health Pub-flow paths between power sources and loads. The inspec-lication 90-874, July 1990, tors evaluate attnbutes such as postulated worst case elec-l trical loads dut,1g design basis events, the capacity and 2.

Hatch, Maureen C., Jan Beyes, Jeri W. Nieves, and op ration of power sources, the electrical protection and Mervyn Susser, " Cancer Nect the Three Mile Island coordination of protective devices, and the interfaces to l

Nuclear Plant: Radiation Emissions,"

Am.

J.

supporting mechanical systems. The inspectors also assess Epidemiology, Vol.132, No. 3, pp 397-412. Septem-the performance of tha heensee's organizabons regardmg ber 1990-engineering involvement in operations, root cause analysis 3.

Morris, Martha and Robert S. Knorr, " Southeastern and corrective actions, and self-assessment.

Massachusetts Health Study, 1978-1986," Massachu-setts Department of Public Health Report, October The results of the EDSFI program have confirmed the ex-1990.

istence of signif cant deficiencies, some of which required hcensecs to promptly determine operabihty and to take 4.

Wing, Steve, Carl M. Shy, Joy L. Wood, Susanne corrective action. The inspection teams identified defi-Wolf, Donna L. Cragle, and E. L. Frome, "Mortahty ciencies in the design basis that affected the capabibty of Among Workers at Oak Ridge National Laboratory:

electrical equipment to perform safety functions. Under-Evidence of Radiation Effects in Follow-Up through voltage relay settings were not adequate to provide 1984," JAMA, Vol. 265, No.11, pp 1397-1402, su!heient voltage to accident mitigating loads during de-March 20,1991.

graded grid conditions. Plant load studies did not corsider i

8

~. -. - ~. --

_ ~..

. ~. -.- - ~

conservative " worst case" loads and voltage drops for de-technical issues regarding electrical equipment and the

' graded voltage conditionsi Unanalyzed relay drift -in-safety significance of these issues. Engineering control of creased the time required to transfer power between plant modificatior.s appeared well established. Root cause sources. Medirm voltage circuit breakers had unconserva-analysis of failures of electrical equipment-appeared l

tive short circuit ratings. Neutral resistors were not large thorough.

1 enough to handle the current and dissipate the heat.

The largest number of findings identified during the ED-Fis involved emergency diesel generators and their me.

By conducting the EDSFis, the staff has helped licensees chanical interfaces. The inspection teams identified defi.

improve the functional capability of the electrical distribu-ciencies in the capacity of diesel generators to carry tran.

tion systems. For example EDSFI findings have focused sient loads; the multiple start capability; the setpoints for the licensees' engineering and technical support groups on fuel oil tank levels; the quantity, quality and transfer of (1) the necessity for controlhng loads relative to the fuel oil; and the air. start accumulator pressure in addi.

capacity of offsite and onsite power sources and (2) the tion, the diesel load sequencers had undersized relay con.

fault protection of electrical equipment. In addition to tacts and unanalyzed timer drift and transient loads. The correcting specific deficiencies, licensees have increased inspection teams determined that these deficiencies could their attention towa!d evaluating and improvmg the design have compromised the startup and operation of the diesels basis'of the electrical distribution system and also toward and could have affected the supply of emergency power to related engineering and technical support programs. Sev-i accident mitigating loads.

eral'icensees have conducted their own EDSFis to assess the electrical system design basis and its implementation.

The inspection teams identified certain weaknesses in the Varioas licensees have identified and corrected problems capabihty and performance of heensee engineering and before the NRC inspected their facilities. Some licensees technical support (E&TS) groups. Licensees had not re-have increased the scope of their design reconstitution viewed electrical design calculations performed by the ar.

programs 'to include problem areas, such as instrument 1

chitect engineer for accuracy and completeness and had setpcints, identified by EDSFls. EDSFis have also made I

not adequately performed self assessments of the design-licensees aware of the importance of retaining and updat-basis of the electrical system. Design basis documentation ing design basis documentation developed by the architect for the electrical system, such as calculations covering all engineer, vendor, and hcensee to demonstrate that the credible failures and modes of operation, was missing or electrical system operates properly and to justify modifica-incomp'ete, particularly at older plants. Test procedures tions to its components. The results of the EDSFis, in-did not include correct acceptance criteria and some enti-cluding a number of positive fmdings, have provided in-cal electrical equipment was not being tested.

creased assurance that the electrical equipment in nuclear plants will perform its safety functions. RSIB is developing

- Some inspection teams identified strengths with regard to a database of EDSF1 findings for further evaluation and E&TS groups. Personnelin the field understood numerous for inclusion in NRC information notices as appropriate.

i 1

4 f

9 T

,,w-

-w-,-,e--,

-re,-

.-.-y-+---

,,e e-e--

._ -= _ -, --yr---r, v-W-"

ww

~'*e.-w----u-w--

=~

e-"~~av

- +

v-

,