ML20031D670
| ML20031D670 | |
| Person / Time | |
|---|---|
| Site: | Allens Creek File:Houston Lighting and Power Company icon.png |
| Issue date: | 10/09/1981 |
| From: | Gotchy R Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20031D662 | List: |
| References | |
| NUDOCS 8110140007 | |
| Download: ML20031D670 (24) | |
Text
-
-a x - + ~
4
-_,---2--
- - - + - -
10/09/81 UNITED STATES OF meiau NUCLEAR REGULATORY C0t1MISS!0N BEFORE THE ATOMIC SAFETY AND LICENSING BOARD, In the Matter of
)
HOUSTON LIGHTING & POWER COMPANY Docket No. 50-466
)
(Allens Creek Nuclear Generating
)
Station, Unit 1)
)
NRC STAFF TESTIMONY OF DR. REGINALD L. GOTCHY CONCERNINr. THE INCREASED RISK OF CANCER AND NON-CANCEROUS EFFCCTS FROM APPENDIX I LEVELS OF RADIATION
[Cumings Contention 9]
~
Q.
Please state your name and position with.the NRC.
A.
My name is Reginald L. Gotchy.
I am employed at the Nuclear Regulatory Commission as a Senior Radiobiologist in the Radiological Assessment Branch. A statement of my professional qualifications has been prdviousi! submitted and I have testified in this proceeding in connection with the issue of Radioactivity in the Cooling Lake (Tr. 3242).
Q.
What is the purpose of this testimony?
A.
The purpose of this testimory is to respond to the Licensing l
Board's Order of September 1,1981 which ruled on summary disposition motions. That Order, inter alia, considered the Staff's motion for summary disposition of Cumings Contention 9 concerning the health effects of low level radiation. After assessing the supporting affidavits filed i
by myself for 1.he Staff, and by Dr. Leonard Hamilton for the Applicant,'
i as well as the opposing affidavits of Dr. Irwin D. J. Bross for the Inter-venors, the Board concluded that the following controverted issue remained to be litigated (Order, p. 70):
8110140007 811009-l PDR ADOCK 05000466 T
The increased risk of cancer and of noncancerous effects from Appendix I levels of radiation is considerably greater than the 0.2% value reached by the NRC, thus invalidating its favorable cost-benefit balance assessment.
In addition, I will present factual evidence regarding the potential impact of current reevaluations of the Hiroshima and Nagasaki radiation dose esti-mates on the validity of conventional risk estimates. Much of the following testimony is adopted from my previous affidavit submitted with the Staff's Motion for Summary Disposition on this issue dated November 26, 1980.
However, some of the information has been updated where necessary to respond to Dr. Bross or to reflect new information.
Potential Cancer Mortality and Genetic Risks Q.
Could you briefly explain what Appendix I guidelines are and what compliance by an applicant to these guidelines means?
A.
After a lengthy rulemaking proceeding initiated 'n 1971 and which included considerations of the potential health risks of low-level ionizing i
radia tion,1/ the Commission adopted Appendix I to 10 C.F.R. Part 50 in 1975.
Appendix I provides numerical guides for design objectives and limiting conditions for operation of LWRs to keep radioactivity in effluents as low as is reasonably achievable.
Design objectives and limiting conditions for operation conforming to the guidelines of Appendix I shall be deemed a
-1/
Final Environmental Statement Concerning Proposed Rule Making Action, Numerical Guides for Design Objectives and Limiting Conditions for Operation to Meet the Criterion "As low As Practicable" for Radio-active Material in Light-llater Cooled Nuclear Power Reactor Effluents,"
!? ASH-1248 (July,1973), pp.1-17 to 1-25, pp. 5-1 to 5-15, and pp. 58-1 to 58-19.
., = _
l conclusive sh'owing of compliance with' the "as low as is reasonably achievable" requirements of 10 C.F.R. Part 50.
In addition to the design objectives for annual doses for any individual in an unrestricted area from both liquid and gaseous effluents, a further requirement is imposed by Appendix I that the applicant include in the radwaste system all items of reasonably demonstrated technology that, when added to the system sequentially and in order of diminishing cost-benefit retten, can for a favorable cost-benefit ratio effect reductior.s in dose to the population reasonably expected to be within 50 miles of the reactor.
As an interim measure, the Commission accepted $1000 per total-body person-ren for making the necessary cost-benefit analysis and indi-cated that this represented a " conservative value" subject to modification at a later date.
1 NRC 277 at 284.
In sum, the Appendix I guidelines were designed specifically to limit the maximum exposure of radiation a person might receive from the operation of LWRs. Any fa'cility conforming to the guidelines would be considered t
a conclusive showing of compliance with the "as low as is reasonably achievable" requirements of 10 C.F.R. Part 50.
1 Q.
What is yoyr understanding of the above issue as it pertains to compliance with Appendix I guidelines?
A.
The thrust of the controverted issue is whether the health effects of such radiation meeting the Appendix I guidelines would, if included.in the cost-benefit analysis, result in an unfavorable NEPA balance for the Allens Creek facility.
_.-.,-,__,,---,2-.
._y-.
_--,.,y._.
. Q.
By way of background, how did the Staff originally assess the health effectsofexposuretoindividualsandthege[eralpublicfromAppendixI levels?
A.
The health effects of the NRC Staff':: p'roposed Appendix I rule was discussed in some detail in the Final Environmental Statement (f ES) prepared for that rulemaking.
(WASH-1258, July 1973). That assessment was based on the BEIR I Report.S However, only population dose comitments for people living within 50 miles of the power plants were considered. Subsequent evaluations by the Staff indicated that most of the population dose commitments from individual plants occurred outside the 50-m11e radius. The best examples for such generic calculations wera presented in the FES prepared for the rulemaking on the use of mixed oxide fuel in LWRs, the so-called GESMO proceeding.
" Final Generic Environmental Statement on the Use of Recycle Plutonium in Mixed Oxide Fuel in Light Water Cooled Reactors," NUREG-0002 (August, 1976). Those assessments, which included 1,000 MWe LWRs using a variety of fuels, indicated that LWRs meeting Appendix I design objective doses would result in population doce co'mmitments of about 100 person-rem to the entire population of the United States for each year of operation.
Thus, the health effects judgments generically hinge on two considerations:
(1) individual risk at doses of 5 mrem per year (or less) to the total body or 15 mrem per year (or less) to any organ of members of the public; and (2) collectPve risk to the U.S. population at annual population doses of 100 person-rem per year.S 2/"The Effects on Population of Exposure to Low Levels of Ionizing Radiation,"
BEIR Committee, National Academy of Sciences (November,1972).
3
..]liost estimates are now below 50 person-rem per year.
~
Q.
How could the NEPA balance for Allens Creek be potentially altered by health effects considerations?
A.
In order for the NEPA balancing of costs a1d benefits to be altered by health effects considerations, the NRC models used to calculate potential somatic effects (e.g., cancer) and genetic effects would have to seriously underestimate the risks. Therefore, the validity of the NRC health effects models will be analyzed in the following discussion followed by a generic assessment of potential risks using current health effects models.
Q.
How were the NRC health effects models developed?
A.
In development of its models, the NRC Staff has reli=J heavity on theBEIRReports(BEIRIReport(1972),andBEIRIIIReport(1980))andthe Reactor Safety Study health effects models which were essentially updated BEIR I models.
Since the 1972 report, the EPA, the NRC, and other federal agencies have generally used the BEIR guidance in assessing the risks of ionizing radiation.
BEIR is regarded as one of the most outstanding g'roups of experts on the medical and biological risks of radiation exposure.
The BEIR I Report and the NRC health effects'models have been subjected to considerable public exposure and debate and have withstood all challenges.
See,, eg., Tennessee Valley Authoritz (Hartsville Nuclear Plant, Units l A, 2A, 1B, and 2B), ALAB-463, 7 NRC 341 (1978). BEIR III deals with the scientific basis of effects of lud-dost radiation and encompasses a review ard evaluation of scientific knowledge developed since the BEIR I Repo'rt concerning radiation exposure of human populations.
Q.
Are these models generally regarded as conservative?
A.
Yes. These models produce estimates of risk that are generally characterized by most radi) biologists as tending to be upperbound (i.e.,
overestimates of the actual risk).
Indeed, both BEIR committees (1972 and 1980) noted that the lower bounds of the risk from exposure to low level and low LET radiation (the type emitted from LWRs) could include zero. S/
~
Q.
Ho>ever, these health effects models have been publicly criticized.
Could you respond to some of these criticisms?
A.
Most of the challenges to the conventional beliefs of the radiation bio-effects community have come from a few studies that have been roundly criticized for reasons ranging from dishonesty to poor statistical methods.
Following are my opinions as well as direct quotes from the BEIR I and BEIR III reports regarding some of these studies.
1.
Drs. Gofman and Tamplin:
The conclusion, therefore, is that the figures generated by Gofman et al. (20) are overestimates: The reasons for the'r overestimates are:
(i) An overestimation of the relative risk of solid tumor induction following irradiation of 0-9 year olds by a factor of 4-5, and by a factor of 10 for all other ages.
i (ii) The unreasonable assumption of a life-long plateau region following in utero irradiation.
(BEIRI,
- p. 188).
- / ee, BEIR I, p. 88; BEIR III, p.187.
S
2.
Dr. Sternglass:
The evidence assembled by Sternglass has been critically reviewed 'y Lindop and Rotblat (5) and by Tompkins and Brown '
It is clear that the correlations presented in sup; rt of the hypothesis depend on arbitrary selection of data su;jorting the hypothesis and the ignoring of those that do not.
(Emphasis added).
In several regards, t'
ata used by Sternglass appears to be in error.
u.e of the most vital assumptions in that model - that without the atomic tests the infant mortality rate would have continued to fall in a geometrically linear fashion -
is without basis either in theory or in observation of trends in other countries and other times. The dose of strontium-90 used in the experiments referred to by Sternglass were of the order of 100,000 times greater than those received by humans from all the atomic tests..."
(p. 178, BEIR I).
Part of Dr. Sternglass' presentation (to the BEIR III Committee) alleged that fallout from Chinese bomb-testing in 1976 led to an increased amount of radioactivity in milk in some areas of the United States.
He concluded that there was an increase in infant mortality in the eastern-seaboard states from Delaware to New England shortly after these events--an increase that he escribed to the radioactivity.
Although Dr. Sternglass stated that his analysis was incomplete, the Committee received no further data on this subject. We have concluded that the alleged association did not fit the time course for radioisotope movement into the cow-milk food chain, nor was there clear evidence of a universally applicable change in infant mortality rates.
Thus, the Committee did not believe that the allegation was substantiated.
(p. 561, BEIR III).
3.
Dr. Bross, et al:
The NRC Staff is aware of and recognizes the possibility that genetic differences among individuals may result in some, people being more sensitive to the effects of radiation than others.
However, it is also possible that other persons are less sensitive to radiation. The large population of Japanese bomb survivors involved all ages, sexes, and wide genetic heterogeneity. As a result, the risk estimators used by the NRC Staff (which are based heavily on those data) include susceptible and less-susceptible members of the population. The BEIR III Committee (1980), reviewing recent claims by Bross, et al., regarding the question of susceptibility concluded:
. The "sutceptible subgroup" model, although it may contain some grain of truth, nevertheless imposes so little structure on the inferences possible from analysis of dose-response data that it is unlikely that usable estimates can be obtained with it from available data.
The applications by Bross, et al have been clearly incorrect, and they provide no evidence that the risk of cancer from low-dose radiation is greater than indicated by conventional estimates.
(BEIR III, p. 559).
Q.
- id you also comment on Dr. Bross' "1981 Reassessment" which was sL f Intervenors on December 23, 1980 in opposition to the Staff's Moti;n ior Summary Disposition?
A.
In his "1981 Reassessment", dated October 9,1980, Dr. Eross purports i
to provide " reliable facts that can give direct answers to questions about low-level radiation hazard withoutguesswork"(emphasisadded). Therefore, he argues, "there can be no scientifically valid reason for bringing in obsolete, less rclavsnt data and for using extrapolations that are mostly guesswork." 1981 Reassessment, p. 2.
The Staff disagrees with Dr. Bross' position since it is based entirely on highly speculative epidemiological studies, and is not supported by scientifically defensible, repeatable controlled animal experiments which do support the current credible assessments of human risk such as BEIR I, BEIR III, UNSCEAR (1977), ICRP and NCRP.5/ His approach is that of a statistician, not that of a radiation biologist or other life scient'.st whose expertise is based on the results of measurable and repeatable events.
However, even an amateur statistician knows statistical correlations in non-repeatable situations can never be proven to be cause and effect unless such correlations can be corroborated in a reasonable manner 5/
UNSCEAR(1977): Sources and Effects of Ionizing Radiation, U.N. Scientific
~~
Committee on the Effects of Atomic Radiation.
ICRP: The International Commission on Radiological Protection.
NCRP: The National Council on Radiation Protection and Measurements.
9-i by controlled animal experiments, or ~by results from other populations.
A classic example of the pitfalls of such correlations was the correlation between the number of babies born in Germany during World War II atd the standing stork population.
It could have been postulated that the birth rate was declining because there weren't enough sto.ks to deliver babies. While the statistical correlation was strong, reasonable people know that the war was killing storks as well as expectant mothers, inducing stress related abortions, in additiori to keeping married couples away from each other for long periods of time. Thus, the Staff finds Dr. Bross' "1981 Reassessment" to be based on statistically weak epidemiological studies that have no support from the thousands of animal studies and studies of human populations exposed at high doses and dose rates.
4.
Dr. Mancuso et al. (The Hanford Study):
Mancuso, Stewart, and Kr.eale have reported finding dose-related excess cancer mortality among occupationally exposed workers, monitored with raciation badges, at the Haniord works in Richland, WA.
Their risk estimates are much higher than estimates derived from studies of the Japanese atomic-bomb survivors and the populations exposed to ionizing radiation for nedical reasons...
[T]he risk estimates'for multiple myeltma and pancreatic cancer were extremely high--so h'igh tha, they can be
~
discounted on ' logical grounds; such high estimates imply an imprcbably large causal role f or background radiation in the etiology of these diseases among the general populat!on.
r--
-,w y
,p-we-m-w e
-,e-n
-e w---
+s.-
s-
t i -,
d It is highly relevant to note that, if the risk estimates i
for multiple myeloma and pancreatic cancer were not-extremely high, they would not satisfy' conventional i
requirements for statistical significance. This necessary numerical relatfonship is a consequence of the limited sample size and low individual radiation doses of the Hanford workers.
Compared with the great majority of i
studies of irradiated populations, the Hanford study is distinctly lacking in statistical power. Th'at is, assuming the convantional estimates to be representative of the true risks of radiation-induced cancer, the Hanford study 4
could be expected to yield risk estimates that are 4
negative with probability around 40%, positive but i
statistically nonsignificent estimates with probability around 50%, and statistically significant but highly exaggerated estimates with probability around 10%. Thus, i
the low statistical power of the Hanford study, according to conventional studies of risk estimates, detracts considerably from the challenge posed by the study's results and from the validity of these estimates.
Other observations support the interpretation of the i
Hanford study results as small-sample phenomena. The emergence of multiple myeloma and pancreatic cancer I
(but not myeloid or lymphocytic leukemia) as the cancers i
most closely related to radiation, the observed (non-l significant) negative associations of dose with the lymphones, lymphocytic leukemia, and stomach cancer in the first '
analysis and with myeloid leukemia in the second, and the fact that the risk estimates obtained in the second, i
expanded analysis were lower than those obtained in the first are all consistent with great statistical instability.
Published criticisms of the Hanford study findings have i
suggested alternative explanations for the observed dose-associations, including confounding of radiation exposure with exposures to other carcinogens and inadequate dosimetry.
Only further study can determine the validity of these suggestions.
At present, however, there seems to be littic reason to abandon the body of epidemiologic evidence on radiation-induced cancer that, although based on greater exposures, yields consistent and statistically stable estimates.
l (BEIR III, pp. 553-556, footnotes omitted).
1 1
4 1
~ w-w ~ m a a swww env-mwe me,,-
.,m r~ m,--o,,w.,.m my n e e.w e,
_m.,-.,w,,.
,w.
.w, c,.,
,-n,~,.,,m
,,w._g,.,
-,.,.n
6.
Dr. Najarian and Dr. Colton (The Portrmouth Study):
The report by Najarian and Colton was based on interviews with next of kin for 525 (of 1,722) certified deaths at ages under 80 among fonner workers at tne Portsmouth Naval Shipyard in New Hampshire.... After the publication of their report, the authors were provided by the National Institute of Occupational Safety and Health with employment and radiation-exposure records from tha Portsmouth Naval Shipyard for the 1,722 names in the original collection of death certificates....
The analysis revealed that the decedents whose next of kin were contacted in the original study did not con-stitute a representative sample of those actually exposed.
In particular, it was more likely that the next of kin would be contacted, and the decedent would be correctly identified.as a nuclear worker, for exposed workers who died of cancer, compared with those who died of other causes.
[S]uccessive analyses of proportional mortality among Portsmouth Naval Shipyard workers [have contributed]
little to our understanding of health risks from low-level radiation.
However, they do provide a remarkable illustration of the dangers of response bias in epidemiologic studies.
(BEIRIII,pp. 559-560, footnote omittad).
The report was reevaluated by Colton in a July 9, 1980 report prepared for National Institute for Occupational Safety and Health (NIOSH). The Colton reevaluation was reviewed by a multidisciplinary team assembled by the National Institute for Occupationti Safety and Health (NIOSH). The Colton reevaluation was reviewed by a multidisciplinary team assembled by the National Institute for Occupational Safety and Health. The NIOSH team report was pub-lished in December,1980 ("Epidemiologic Study of Civilian Employees at the Portsmouth "aval Shipyard")and contrary to the claims of Dr. Bross, the final analysis shows no correlation between radiation dose (not at ALARA levels as claimed by Dr. Pross, but at 1,000 or more times greater doses) and risk of cancer or other non-cancerous diseases. As a matter of fact, NIOSH found " Statistically significant deficits in mortality occurred for
12 _
respiratory tuberculosis phxsa), scomacn cancer (SMR=71), diabetes mellitus (SMR=77), vascular lesions of the central nervous system (SMR=75), circulatory system diseases (SMR=86), diseases of the respiratory system (S!1R-75), diseases of the digestive system (SMR=78), plus other diseases.6_/
Q. What are some of the general arguments criticizing current health effects models?
A.
The following discussions are relevant to some of the arguments just discussed. However, they have been raised by several people (including sorne of the above) and deserve separate consideration.
The first argument is that as the dose rate declines, the health risk per unit dose increases.
BEIR III disagrees, concluding:
~
The available data relative to the effects of low-dose or low-dose-rate exposures on carcinogenesis in humans and experimental animals do not, in general, support the hypothesis of an increased probability of induction at low dose rates.
Increasing the duration over which a given dose of low-LET radiation is administered, either by decreasing the dose rate or by fractioning the dose, has been generally found to decrease oncogenic effects of ionizing radiation.
(BEIR III, p. 565).
The second argument is that the risks of radiation exposure increase in a multiplicative manner (i.e., inferring synergism) when accompanied by exposure to other carcinogens or toxic materials. The NRC Staff recognizes such a possibility.
However, (1) it should be noted that if such effects exist (e2.,
from cigarette smoking), they must first be identified before anything can be
~
done about them, and (2) the risk estimates from Hiroshima and Nagasaki and other human studies would include possiblu synergistic effects with other agents.
The BEIR III Committee, in reviewing the most recent data on possible synergism between cigarette smoking and exposure to radon-222 decay products, concluded the following:
-6/
SMR means standardized nortality ratio; the ratio of a particular effect in the study population standardized for age, sex, etc., to a control (unexposed) population (expressed here as percentage).
13 -
l it is of interest that emerging data in man indicate that cigarette-smoking does not contribute as strongly to the risk of bronchial cancer induced by radiation as had previously been tr.ought, although smoking shortens the latent period.
(BEIRIII,p.390).
t' Q.
Whac is the Staff conclusion with respect to the BEIR III Report?
l, A.
Based on the foregoing discussion, it is apparent that the BEIR Com-mittee has considerrd and discredited most of the studies that have criticized i
the conventional widsom regarding radiation bio-effects. Thus, the Staff believes that the BEIR III Report can be considered the latest authoritative guidance or the best scientific evidence available with respect to health effects from radiation exposure.
Therefore, the va'lidity of the health effects models used by the NRC can be determined by a comparison with the BEIR III Report.
Q.
What does this comparison with the BEIR III Report show insofar as the validity of the NRC health effects models?
A.
In such a comparison the Staff finds that the most probable risk values cited by the BEIR III Report are only slightly lower than those currently used 6
by the NRC Staff--120 cancer deaths per 10 person-rem versus 135. With 1
regard to estimates of genetic effects in future generations, the BEIR III Report yields a slightly lower risk than estimated by the NRC Staff--
6 220 effects per 10 person-rem versus 258.
Based on these comparisons with the BEIR III Report, the Staff believes that the validity of the NRC health effects models is confirmed.
Q.
Is it possible that new studies based on further evidence will dis-
- credit the existing models and contradict present conclusions with respect to the risk of low levels of radiation?
I r-e n-
,n.-
-.w,.,o
,,,..,,,.v.~,,,.
,,,,,,_,e...,..,,,.,,-._,-~.--.,n, e_
s.
-n.
.,,. -,, -, ~, =..,,, -
c
A.
Whether new evidence in the future will reveal unsuspected risks is unknown, but it is my belief that exercises in epidemiological mc.deling by expert statisticians who are amateur biologists will never result in signifi-cant advances in delineating the risks of low-level radiation exposure.
Clearly, joint cooperative efforts by teams of experts in the various areas needed for complex evaluations of epidemiologic are the only way such risks will ever be unravalled. Thus, groups of experts such as the BEIR Committee must be relied upon for the best guidance for responsible protection of the public from unwarranted risks, since no single expert can possibly answer such difficult questions.
In this respect, Dr. Bross is curiously alone in his claims that he has unravelled the ccmplex issue of human health risks from exposure to radiation at Appendix I dose levels (maximum of 0.005 rem per year to the total body and 0.015 rem to year to the thyroid or other organ),
and the Staff is unaware of any support for his claims by competent peers.
Q.
Since you have concluded that the health effects models used by the NRC Staff are in agreement with the latest authoritative guidance, how will they be applied to evaluate the potential health risks associated with LWRs operating in accordance with Appendix I design objectives?
A.
In the first case, exposures to individuals, the health risks are proportional to the Appendix I doses.
The following table shows the health risks to individuals per year of exposure to Appendix I doses.
,,-.e.,
n
-,--,n..
I.
Liquid Effluents App I Design Lifetime Risk per Year of Exposure
~
Ob.iectives (per Unit)
Et App. I Dose Levels Cancer Mortality Genetic Effects Total body 3 mrem /yr 0.4 chances in a million.
0.8 chances in a million Any organ
- 10 mrem /yr 3 chances in a million (gonad dose only) a.
Thyroid 10 mrem /yr 0.1 chances in a million b.
Total Bone 10 mrem /yr 0.3 chances in a million a
s II.
Gaseous Effluents Total body 5 mrem /yr U.7 chances in a million 1 chance in a million Any organ
- 15 mrem /yr 4 chances in a million (gonad dose only) a.
Thyroid 15 mrem /yr 0.2 chances in a million b.
Lung 15 mrem /yr 0.3 chances in a million c.
Total Bone 15 mrem /yr 0.4 chances in a million l
l j
All other organ risks are smaller than those shown.
. 4 Q.
How do these health risks compare to other sources of risk that individuals may be confronted with in their lifetime?
A.
In order to place the preceding risk values in perspective, the table below lists some lifetime risks of mortality that are numerically equivalent to one chance in a million:
~
Lifetime Mortality Risk Equivalent to one chance Source of Rig in a MillionJ/ ~
Cigarette smoking (primarily lung cancer and 1.5 cigarettes cardiovascular disease)
Drinking wine (chirrosis of the liver) 0.5 bottle Autorobile driving (accidental death) 50 miles Air travel (accidental death) 250 miles Rock climbing (accidental death) 1.5 minutes Canoeing (accidental death) 6 minutes Typical factory work (accidental death) 1-2 weeks Being a m, aged 60 (cancer and heart disease) 20 minutes Being pregnant (ages 20-34) 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> The average annual risk of mortality is on the order of 14 chances in 1,000; or about 10,000 times greater than the potential lifetime risk of annual radiation doses at Appendix I levels.
Q.
Does exposure to doses at Appendix I levels significantly increase the individual risk from exposure to natural background radiation?
-7/
Sir E. E. Pochin, "The Acceptance of Risk," Br. Med. Bull. 31, No. 3,
- p. 184 (1975).
a.
A.
No.. Background radiation varies from about 80 mrem / year for some low level areas (e_.g., Florida) to about 200 mrem / year in the high mountains of Colorado.
In other words, people are exposed to normal variations in background radiation exposure in the U.S. that are 1,000% or more in excess of the Appendix I design objective levels.
Since most experts believe baci.-
ground radiation accounts for a few percent or less of the total lifetime risk of cancer death (cancer accounts for about 20% of the total lifetime risk of mortality), the overall risk o'? cancer mortality would not be significantly changed by exposures at the Appendix I levels.
Q.
A second type of potential risk comes from cumulative population doses from LWRs operating at Appendix I design objectives. What is the mortality and genetic risk from that exposure?
A.
As mentioned previously, a reasonable estimate cf the dose to the U.S. population (delivered at average rates that are about 0.0005% of back-groundrates)isabout100 person-rem.N Current health effects models would predict on the order of 0.01 potential cancer death (_i_.e_., none) among the entire U.S. population during the remainder cf their lives.
In other words, there might be as much as one death in the years ahead for each year the U.S. operates 100 large LWRs. Thus, if the 220 million Americans now living 8]
Person-rem is the unit of collective population dose; it is the sum.
of the doses to each person in the exposed population.
For example, one millirem given to 1,000 people is one person-rem. Most assess-cents estimate less than 100 person-rem.
i i
18 -
were continuously exposed to the radioactivity released from 100 large LWRs for the life of those plants (about 30 years / plant), about 40 panole might die from cancer due to the operation of the LWRs among 4.4 million people who would_ die from cancer due to all other causes (i.e., about 0.0009%
of the total from all other causes).
In the case of genetic effects, the collective risk of a genetic defect occurring during the next 5 generations is about twice that of the risk cf cancer mortality.J Most of the genetic effects, while tragic, would be relatively insignificant compared to the current estimated risk of about 6% to 9% per generation 9I from all other causes (i.e., less than 0.01% of the total from all other causes).
Q.
What is your conclusion about the health risks to present day populations from cancer, and to future populations from genetic effects associated with normal operations (Appendix I levels) of large LWRs in the U.S.?
~
A.
Based on the foregoing discussion concerning the current health effects models, it is concluded that health risks to present day populations from cancer, and to future populations from genetic effects associated with normal operations (Appendix I levels) of large LWRs in the U.S. are in-significant relative to naturally occurring events. As a result of the de minimus nature of Appendix I health risks, the NEPA cost-benefit balance cannot be significantly affected by them and, thus, result in an unfavorable NEPA conclusion regarding the construction of the facility. Accordingly, this 9/ BEIR I, p. 56 ; BEIR III, p. 87.
--y 4-..-
-.-e y., ~
- ,+-*9y--g,oe
.,_- -.+
n.
-*---v---~ +
-r, y
contention that normally operating LWRs which meet the Appendix I design objectives can result in health costs that would alter the NEPA cost-benefit analysis must be rejected on the basis of the best scientific evidence available.
Potential Non-Cancerous Health Effects Q.
Dr. Bross states that it is now known that the immediate cause of radiation-induced cancers is genetic degradation, i.e., DNA damage in the genetic material of a human cell (1981 Reassessment, pp. 2-3).
Do you agree with this stateuent?
A.
It is not kn6wn to be the immediate cause. However, I would agree that the probable cause for most radiation-induced cancer is DNA damage in the genetic material of somatic cells (as opposed to reproductive cells).
Q.
What could happen if there was radiation damage to the DNA of repro-ductive cells?
A.
DNA damage of reproductive cells could lead to mutations expressed in the offspring of irradiated parents.
Radiation induced mutations may predispose offspring of irradiated parents to certain types of cancer and non-malignant diseases such as diabetes, schizophrenia and mental retardation.
Q.
What evidence demonstrates that radiation causes genetic damage?
A.
Although there is rp, direct evidence of human genetic effects of radiation even at high doses such as received at Hiroshima and Nagaraki (BEIR III, p. 79), experiments on both plants and animals cl.early demonstrate that such damage is done. Thus, it must be assumed that humans are affected in the same way.
_ __ _ _ ~ ~ _ _ _
__._-_ - = _ _. _ -..
l* ;_
9 i
i Q.
Dr. 8ross in his 1981 Reassessment advances what he calls the genetic degradation hypothesis which assumes that the leukemia risk per rad is greater at low doses than at high doses (p. 5). Do you agree with Dr. Bross' hypothesis?
l A.
No.
I strongly disagree with Dr. Bross' clairn that genetic de-gradation will lead to a higher risk per rad at low doses o; low dose rates.
Again, Dr. Bross has merely discarded the results cf many years of precisely l
1 controlled animal studies which clearly demonstrate, that the risk per rad declines as the dose and dose-rates decline (BEIR III, pp. 90 and 107).
As a result I must agree with the experts on the BEIR Committees (1972 and l
1 t
^
1980)that the assumptMn of a linear dose response relationship results in a reasonably conservative assessment of potential genetically determined health j
risks in future populations.
i Q.
Can direct exposure to radiation cause non-cancerous health effects l
i in humans?
~
i A.
In regard to non-cancerous health effects from damage to somatic cells in people directly exposed to radioactive effluents from nuclear power plants, the situation is more uncertain than for potential genetic defects i
in future populations. The BEIR III Consnittee examined the possibility for non-cancerous effects and concluded that the non-cancerous risks from low i
level radiation are small relative to the cancer risks:
... the data from the atomic-bomb survivors strongly suggest that the effects of ionization on mortality are l
specific, focal and principally carcinogenic.
For doses of less than approximately 300 rads of low i
1.ET radiation, the principal mechanism of life-shortening is the induction or acceleration of neoplastic diseases--
evidence of life-shortening from effects other than tumor induction is inconclusive and therefore carinot be used for quantitative risk estimates.
(Pace 505) i e
21 -
Similarly, the BEIR III Committee concluded that fertility is not affected below about 400 rads of acute low-LET radiation dose (p. 499),
while cataract induction does not occur for acute doses below about 200 inds and chronic doses below about 1,000 rads (p. 499). Furthermore, these con-j clusions are consistent with those of the UNSCEAR Committee (see for example, UNSCEAR-1977, para.18) and the ICRP (see for example, ICRP Report No. 26, pp. 8-13).
In the area of teratogenic effects following in, utero exposure, the BEIR III Committee concluded that doses in the range of 1,000 to 2,000 millirem (i.e.,
hundreds of
' times higher than Appendix I design ob-jective doses) can cause birth defects. However, the Committee warned:
... it is likely that there are threshold doses for most ma1 developments and that these are of a variety of magnitudes. Lowering the dose rate diminishes (emphasis added) the damage.
(p.493)
Thus, while there may be some risk of birth defects from radiation exposure at Appendix I levels. it is probab".e that thresholds for such effects exist at dose levels near or above those found in nature (i.e.,
20 to 40 times greater than Appendix I levels).
It is therefore my conclusion that the risks of non-cancerous diseases resulting from exposure to Appendix I radiation levels must be small (or zero) relative to the risk of eadiogenic cancer. That is also the principal reason I did not address this question in my 1980 affidavit supporting a motion for summary disposition of this contention.
Current Uncertainty in Cancer Risk Estimates Related to Hiroshima and' Nagasaki Dose Estimates Q.
Have any studies been completed since publication of the BEIR III Report which may change the risk estimates ccntained in that Report and the Staff'; reliance on it?
A.
In October,1980 scientists at the University of California Lawrence Livermore Laboratory (LLL) publislied revised dose estimates for Hiroshima and N
Nagasaki which indicated the Hiroshima neutron dose (rads) was greatly overestimated (factor of about 4 or 5) and the gamma dose (rads) was under-estimated by a factor of about 2.'
In Nagasaki, on the other hand, the previous neutron dose (rads) was also overestimated (factor of about 2) while the previous gamma dose was slightly overestimated.
In March,1981 a second LLL report was published which attempted to provide a preliminary assessmen. of what the LLL dose estimates might imply in terms of changes in current risk estimates (e.g.
BEIRIIi).E While t.he dose response curves for both cities were linear-quadratic (i.e., sublinear),
the authors concluded "The re-evaluated A-bomb-survivor data do not support the argument that low doses or low dose rates carry higher risk per rad than high doses or high dose-rates."
On May 31, 1981 this question was examined again in a special session of the annual meeting of the Radiation Research Society.E At that time, it 10/ Loewe, W. E. and E. Mendelsohn, " Revised Dose Estimates at HiroshiJna and Nagasaki," UCRL-85446 (Preprint), October 1,1980, 11/ Straume, T. and R. Lowry Dabson, " Implications of New Hiroshima and Nagasaki Dose Estimates: Cancer Risks and Neutron RBE," UCRL-85699 (Preprint), March 12, 1981.
12/ See "New A-Bomb Data Shown to Radiation Experts," Science, Vol. 212,
- p. 1364-365, June 19, 1981.
_ -~
i was generally concluded that if the LLL dose estimates are confirmed, the overall change in BEIR III risk estimates nggt result in small changes in cancer risk estimates based on Hiroshima and Nagasaki.
i In September,1981'a special symposium was sponsored by the Department j
of Energy to further examine the LLL work and its implications.13/ The general conclusion was that there is still considerable uncertainty in the air transport and shielding calculations and even the exact yield of the two bombs dropped in Japan.
Further, although no major changes in cancer risk estimates are expected, additional studies still are needed before a final conclusion regarding risk can be reached.
In the interim, it is the staff's position that the BEIR III cancer risk estimates are the most recent and reliable, and that future changes (if any) will not result in large changes in those estimates.
)
l CONCLUSI0h[
Q.
Considering your testimony above, what do you conclude as to whether the increased risk of cancer and non-cancerous effects from Appendix I levels of radiation is considerably greater than the 0.2% value reached by the NRC, thus invalidating its favorable cost-benefit balance assessment for the Allens Creek facility?
- l 1
13/ Reevaluations of Dosimetric Factors; Hiroshima and Nagasaki, Sympo'sium, U.S. Dept. of Energy, Germantown, MD, Sept. 15-16, 1981.
I i
1
..-,.,c
A.
The analyses of Drs. Bross, Sternglass, Gofman and Tamplin, Mancuso, el al_., Colton and Najarian, and others who allege that conventional health effects models underestimated the health risks from radiation have all been reviewed by accepted experts and found to be statistically weak.
In at least one case, data was selected to support conclusions reached before the analysis was done.
In addition, the 1981 NIOSH study of the Portsmouth Shipyard workers referred to by Dr. Bross found that the studies by Des. Najarian and Colton were flawed,
..d that a careful analysis of the total population of workers showed no correlation between cancer or blood diseases and radiation exposure for data collected to the present time, even though the study had a 99% probability of detecting the 5-Told excess claimed by Colton and Najarian. NIOSH also found that risks for many other noncancerous diseases and some cancers were lower than might be expected, liith respect to Dr. Bross' work, there is nothing in his "1981 reassess-ment" which would change the conclusions reachea above in my testimary. Even if. the present expert estimates of the risk of leukemia (the only case argued by Bross) were low by a factor of 20 as Bross argues, the overall estimate of cancer risk would only increase by a factor of 6.
Again, assuming that were true, an individual's lifetime risk from exposure to Appendix I Design Objectives Doses would still be less than four in 10-thousand from 30 years of continuous exposure at 5 millirem / year.
The normal lifetime risk of cancer mortality is currently about one in five. An incremental increase in the lifetime risk of cancer mortality of less than 0.2% (the most probable value would be about 0.009%) should not be a reasonable basis for challenging the NEPA cost benefit balance for Allens Creek, or any other nuclear power plant.
While the non-cancerous somatic effects cannot yet be quantified the evidence indicates such effects would be small relative to cancer, and would not significantly affect these conclusions.
-