ML20099L543
| ML20099L543 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 11/21/1984 |
| From: | Phyllis Clark GENERAL PUBLIC UTILITIES CORP. |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| 5211-84-2284, NUDOCS 8412010168 | |
| Download: ML20099L543 (81) | |
Text
- .
c 3
GPU Nuclear Corporation NUOIMI 100 Interpace Parkway Parsippany, New Jersey 07054-1149 (201)263-6500 TELEX 136-482 Writer's Direct Dial Number:
November 21, 1984 (201) 263-6797 5211-84-2284 Mr. Harold Denton, Director Office of Nuclear Reactor Regulation l
Nuclear Regulatory Commission Washington, DC 20555
Dear Mr. Denton:
SUBJECT:
DOCKET NO. 50-289 - BEYEA REPORT 1
As you know in August,1984, a report entitled "A Review of Dose Assess-ments at Three Mile Island and Recommendations for Future Research," pre-pared by Dr. Jan Beyea, under contract with the TMI Public Health Fund, was made available to us. The GPU companies had nothing to do with the undertaking and preparation of the Beyea Report; however, it was referred to during a Commission meeting on THI, and so I provided it to you.
We asked Drs. Jacob I. Fabrikant and Merril Eisenbud, two experts in the field, to advise us of their views with respect to the "Beyea Report."
I enclose a copy of letters dated October 6 and October 7,1984, to me and dated November 3,1984, to Mr. Heward, GPUN Vice President, and of the attachments to those letters in response to our request.
It would not do justice to the comments of Drs. Fabrikant and Eisenbud for me to attempt to summarize their views, and I shall not do so.
I do, however, suggest that they deserve careful study by you and your staff.
Sincerely, b$
P. R. Clark PRC/agh Enclosures 8412010168 841121 PDR ADOCK 05000289 P
PDR Ol GPU Nuclear Corporation is a subsidiary of General Pubhc Utihties Corporation l l
7.
l f'
JAcon I. FAHRIKANT. M.D PH.D.
138 AtWANAIMD ROAD IDEN KEI.EY. t*Al.tFUNNIA 9e?06 Y-I*irarmament or Naphatamar N Apt ATtoN ASID ttEAI.Til t*mvamatvv or CaLaroama lesmust.sv HIN hoeuM6 November 3rd, 1984 Mr. Richard W. Heward, Jr., Vice President Radiological and Environmental Controls G.P U Nuclear. Corporation 100 Interpace Parkway Parsippany, New Jersey 07054 Re: The 1984 Beyea Report
Dear Dick,
I am enclosing my written respo nse to the August 15th,1984 Beyea Report, as you requested. My response is both general and specific, but.I have not attempted to take on the Beyea Report on a point-by-point basis. That would require a committee of scientific experts and a great deal of time and effort.
As my two letters to Phil Clark (enclosed) point out, Merril Eisenbud and I are in full agreement as regards what should be con-sidered to be done, if anything.
It would appear that the estensive investigations conducted by and the reports of the Nuclear Regulatory Commission, the Department of Energy, the Environmental Protection Agency, the Food and Drug Administration, the General Public Utilities, the President's Comission, and others, concerning the dosimetry of the accident at Three Mile Island, together with the implications for potential delayed health effects, are being questioned by the current Beyea Report. Therefore, in view of the circumstances, a special expert scientific committee knowledgeable in the nuclear aciences-and engineer--
ing, radiation dosimetry, radiation epidemiology and statistics, and risk analysis and decision-making, should be assembled to examine all these key investigations, including the Beyea Report, and provide a comprehensive report and evaluation which will assess the credibility, validity, and degree of certainty associated with each of these investi-1 gations and reports.
There are at least three types of uncertainties which must be evaluated. (1) Uncertainties in data, arising from an inability to make very precise measurements, either because of inaccuracies in instruments or because of inherent variability.in processes. -(2)
. Uncertainties in assumptions and models used to analyze data. And (3) uncertainties that are intrinsically not estimable because import-ant phenomena.or principle have not yet been discovered.
It would appear that ' Governor Richard Thornburgh might best be in the position to request perhaps the National Academy of Sciences-
.j National Research Council to appoint members to a committee responsible -
for the study and the report chosen for their special competences and i
with regard for appropriate balance.
continued...
E
~l o
JACOM I. FABRIKANT. M.D PH.D.
135 ALVANAIMD NOAD ft EN K ELEY. CA R.IPOH N E A 94705 l*WoFEnsesem nr N Anfbtmv N ADI ATION AND IIKALTil t'wev,.antTV or CALarneNI4 BRMMELET
'43 8 "4 "O
Mr. Richard W. Heward, Jr., November 3rd,1984, page 2 Finally, Supplement No.1 of the Beyea Report, dated October 1984, has just arrived on my desk, kindly sent by Mr. Thomas Murphy. Constraints on my time imposed by a very demanding schedule simply does not permit me an extended review and comentary before I leave for Washington, D.C.
early on Tuesday morning, the 6th of November 1984, to join Messrs. Kuhns, Clark, Kintner, Fletcher, Rasmussen, et al at the Nuclear Regulatory Comission. However, a brief glance at Beyea's Supple. ment No. I provides no surprises, and suggests to me, at least, that my enclosed comentary on the August Beyea Report extends to the October Beyea Supplement.
I hope I have helped. Please keep me informed on the progress and the position of GPU concerning the Beyea Report. With all good wishes and with my warmest personal regards, I am Very sincerely yours, ack Jacob I. Fabrikant JIF:ib cc: Mr. Philip R. Clark Mr. E. E. Kintner Dr. James C. Fletcher Dr. Merril Eisenbud l
JACOs I. FAHNIKANT, M.D., Pn.D.
338 ALVAN AIM) NOAD IlEMMELEY. cal.1FUNNI A 94706 l'arcranann OF R ADiouM37 MADI ATION AND If EALTl4 L*nsvaasaryv or CAtJrunwIA Banustav solneM44 moos October 6th, 1984 Mr. Philip R. Clark, President G P U Nuclear Corporation 100 Interpace Parkway Parsippany, New Jersey 07054 Re: The Beyea Report Dear I have attached a brief statement concerning the findings and recommendations of the President's (0 mission on the Accident at Three Mile Island. The statement 's based on the report of the Public riealth and Safety Task Force; I have modified it from the original, which.I wrote, and which I prepared for testimony before Congress.
The President's Commission reviewed, in great detail, the nuclear accident radiation dosimetry. Its assessment found sufficient scient-ific evidence for estimating with considerable precision the collective dose equivalent and average or individual doses to the general popula-tion and the workers. Accordingly, using epidemiological and statistical methods evolved over more than two decades by national and international scientific bodies concerned with radiation and health, and based on current and conservative radiation protection philosophy, the President's Commission ~could estimate quite reliably the potential dalayed health effects in the general population and in the worker population of exposure to low levels of ionizing radiation released during the accident. The evidence compelled the conclusion that no detectable delayed health effects will occur in the general population, the worker population, or their progeny.
The recommendations of the President's Comission, therefore, were not directed to assessing potential delayed health effects in exposed populations in and around Three Mile Island, or in their progeny.
On the contrary, since it was concluded that no detectable health effects will occur, the recommendations emphasized those general conditions that would have wide application to all potential nuclear accidents, viz.,
research, education, radiation monitoring and surveillance, and improved emergency planning and response by Federal, State, and local agencies.
The recently released August 15, 1984 Beyea Report appears to be in direct conflict with the findings and recommendations of the President's Commission on the Accident at Three Mile Island.
Very; sincerely yours, f3%~ -
Jacob I.' Fabrikant JIF:ib
\\
.Fabrikant 1-BRIEF STATEMENT ON FINDINGS AND RECOMMENDATIONS OF THE PUBLIC HEALTH AND SAFETY TASK FORCE OF THE PRESIDEi:T'S COMMISSION ON THE ACCIDENT AT THREE MILE ISLAND The President's Commission on the Accident at Three Mile Island estimated
.that between March 28th and April 15th,1979, the collective dose of radiation resulting from the radioactivity released at TMI to more than 2 million people living within 50 miles of the nuclear plant was approximately 2,000 person-rem.
This represented an average increase of about 1% of the natural background radiation level each person living in that area normally receives each year.
Within 5 (10) miles, it was calculated to be an average increase of about 10%
(5%) of the annual background radiation.
On the basis of present scientific knowledge, the radiation doses received by the general population exposed during that period were so small that there will be no detectable additional cases of cancer, developmental abnormalities (i.e., birth defects) or genet '
ically-related ill-health (i.e., inherited disease) as a consequence of the accident at Three Mile Island.
During the period from March 28th to June 30t,h,1979 only, three out of approximately 1,000 workers were exposed to measurable low-level radiation received doses of 3 to 5 rem; these levels just exceeded the NRC maximum permissible quarterly dose of 3 rem.
The major health effect of the accident was on the mental health of the people living in the region of Three Mile Island and of the workers at the Three Mile Island nuclear plant.
High ' levels of mental distress occurred in household heads living within 5 miles of TMI; mothers with pre-school age children; teenagers living within 5 miles of TMI, with pre-school age brothers or sisters and whose families left the area; and the workers at TMI.
Fabrikant 2 The Commission recognized that although the radiation dose levels due to the accident 'were very low, nevertheless, not enough Nas known about the patential health effects of low-level radiation of a few rem or less.
It therefore recommended increased emphasis on better coordinated and expanded health-related radiation effects research, particularly on the biological effects of low-level radiation, and on the development of methods of monitor-ing and surveillance, and of mitigating adverse health effects due to radiation.
It further recommended educational programs for the public on how nuclear power plants operate, on radiation and its health effects, and on protective measures against radiation.
The Commission noted with concern that while Federal, State, and local agencies all responded to the emergency, there was, however, confusion over definition of responsibilities and a notable absence of designated authority responsible for protecting and insuring the public health and safety.
Emergency plans were either incomplete or were not designed to meet the demands of a protracted crisis.
Federal and State officials disagreed about the nature of the information on which to base emergency preparedness decisions, such as evacuation of vulnerable populations, and other protective actions during the emergency.
The Commission therefore recommended that there be significant involvement by Federal and State health agencies into emergency planning and response to a nuclear reactor accident, Emergency plans must detail clearly and consistently the actions public officials and utilities should take in the event of a radiological emergency to protect the public health and safety.
Specifically, they mcst insure the feasibility and effect-iveness of evacuation plans, requirements for protective measures against radiation, adequacy of plans for enviornmental radiological monitoring, and
.Fabrikant 3 l
adequacy and availability of health professionals and facilities for protecting pubMc health and worker health and safety.
wV hk Jac b I. Fabrikant, M.D., Ph.D.
formerly, Director, Public Health and. Safety The President's Commission on the Accident at Three Mile Island October 3, 1984 J
w w
w-g---
r
-+
r JAcon I. F'AHNIKANT M.D Pn.D.
ISS AtJAN AIMD Mt)AD alENKK1.EY.CAtJIM)HNI A 94708 I*noramason or RADmsour H AI3I ATION AND llEALTl!
L's awahaarv OF CAIJMastNia HsmessLav
'4 8 88 "4WMDS October 7th, 1984 Mr. Philip R. Clark, President G P U Nuclear Corporation 100 Interpace Parkway Parsippany, New Jersey 07054 Re: The Beyea Report Dear Mr ar,
I have enclosed a very rough working draft, prepared hurriedly for your review, that you may wish to use in part or completely to respond to the August 15, 1984 Beyea Report. It is written to serve as a framework only, but to have sufficient information for your staff to prepare a position paper for the public record. There is still much to do, particularly matters of verification, editing, corrections, and references cited.
The draft addresses three issues: (1) the findings of the President's Conunission on the Accident at Three Mile Island, concerning the radiation dosimetry of the accident and the potential delayed health effects in the general population and the workers; (2) the krypton-85 venting dosimetry in June-July 1980 and its potential health effects; (3) the assessment of the Safety Advisory Board, TMI-2, of the NRC Supplement No.1 to the PEIS concerning the worker collective dose equivalent during the TMI-2 recovery program. All three issues are a matter of record in the public sector.
These three areas are those addressed in the recently released August 15, 1984 Beyea Report. There are notable disagreements between the Beyea Report concerning dosimetry of the accident, the krypton venting to the atmosphere, and the worker collective dose equivalent during the recovery program, and the reports extant concerning these areas of investigation.
These disagreements are particularly evident in the findings, the implica-tions for potential delayed health effects, and the recommendations that flow from them.
At present, the Beyea Report appears at odds with the scientific evidence and the conclusions of investigations of recognized scientific bodies and groups who have been dealing with matters of radiation dosimetry and epidemiology at the nationa.1 and international levels. Accordingly, GPU Nuclear Corporation may wish to chart a course of action that places on the public record---perhaps prior.to the " Fund" meeting in Philadelphia in November---its own position concerning the Beyea Report.
Please keep.me informed of any decisions in this matter, where they appear appropri~ ate and I can assist. I am asking Dr. John Auxier to review the draft, to provide corrections and conrnents, and to pass them on to you.
With my best wishes and with my kindest personal regards, I am Very sincerely yours, f
' JIF:ib Jacob I. Fabrikant cc: Dr. John Auxier Dr. James Fletcher
s AN ASSESSMENT OF THE AUGUST 15, 1984 BEYEA REPORT, "A REVIEW 0F DOSE ASSESSMENTS AT THREE MILE ISLAND AND RECOMMENDATIONS FOR FUTURE RESEARCH", WITH SOME COMMENTS ON THE RADIATION DOSIMETRY OF THE ACCIDENT' AT THREE MILE ISLAND, THE KRYPTON-85 VENTING FROM CONTAINMENT TO THE ATMOSPHERE, AND THE PROJECTED WORKER COLLECTIVE DOSE EQUIVALENT DURING THE TMI-2 REC 0VERY PROGRAM mS
_<,-_---u Jacob I. Fabrikant November-1st, 1984
Fabrikant 1 The following report is in five parts:
(1) the radiation dosimetry of the accident at Three Mile Island derived from the 1979 President's Commission Report; (2)' the estimation of the potential delayed health effects'of the accident at Three Mile Island derived from the President's Commission Report; (3) the dosimetry and potential health effects of the releases of krypton-85 vented from the TMI-2 containment building in June-July, 1980; (4) the worke,' exposure experience during the clean-up of the damaged TMI-2 nuclear power plant; and (5) an assessment of the credibility, validity, and degree of certainty of the 1984 Beyea Report, "A Review of Dose Assessments at ihree Mile Island and Recommendations for Future Research," by Jan Beyea, Principal Investigator, August 15, 1984.
1.0 Radiation Dosimetry of the Accident at Three Mile Island 1.1 The Radiation Doses During Normal Operating Conditions Under Normal conditions, the 2,163,000 persons living in the 50-mile area surrounding Three Mile Island would receive an annual collective dose of about 440,000 person-rem; about 240,000 person-rem would come from natural background radiation, and the rest primarily from medical and dental radiation. The average dose-rate from natural background exposure to the individual living in the Harrisburg, Pennsylvania area-is about 116 mrem per year. This comes primarily from cosmic radiation from outer space, terrestial radioactivity in the soil and in building materials, and the radioactivity within the human body.
Under normal operating conditions of the Three Mile Island Nuclear Generating Plant, based on the Final Environmental Statement, for the
Fabrikan.t 2 almost.2 million persons living in the 50-mile area, the radiation dose was estimated to be 31 person-rem per year whole-body collective dose, and 0.017 mrem per year average whole-body dose to the individual.
From gaseous effluents from the Three Mile Island Plant these values were
~
estimated to be 2.05 person-rem per year whole-body collective dose, and 0.0011 mrem per year average whole-body dose to the individual, respectively. Over a 30-year operation, the total collective dose
.)
predicted was 930 person-rem, and the individual dose was 0.51 mrem.
1.2 The Radiation Doses During the Accident at Three Mile Isl-and During the accident at Three Mile Island, considerable effort by Federal Agencies of the United States, and by foreign groups, went into the accurate assessment of the radiation exposures received by the general population living in south central Pennsylvania.
There are quite accurate estimates of the collective radiation dose received by the approximately 2 million people residing within 50 miles of the Three Mile Island Nuclear Station resulting from the accident.
The initial estimates were mainly for the period from March 28 through April 15, 1979, during which accidental releases occurred that resulted in exposure to the offsite population. These measurements are continuing to the present.
Nuclear radiation doses were measured with instruments or detectors called thermoluminescent dosimeters (TLD).
The principal dose estimates were based upon ground-level radiation measurements from thermoluminescent dosimeters located within 15 miles of the TMI site. These estimates assumed that the accumulated exposure recorded by the dosimeters was from gamma radiation (that is, penetrating radiation that contributes dose to the internal body organs). The data were obtained from dosimeters
Fabrikant 3 pla'ced by Metropolitan Edison Company after the accident and covering the period to April 15, and from dosimeters placed by the Nuclear Regulatory Commission from noon of March 31 through the afternoon of April 7,1979.
Additional determinations provided by the Department of Energy using aerial monitoring that commenced about 4 p.m. on March 28, 1979 is also included.
Further data were collected by the Environmental Protection Agency and the United States Public Health Service.
There is also available equivalent accuracy on possible internal exposure doses received by ingestion or inhalation of radionuclides, particularly iodine-131.
1.2.1 TLD Measurements TLD measurements formed the basis for estimating the total external gamma radiation doses (due almost er lusively to the radioactive noble gas xenon-133 and a few other short-livei radioactive gases in the radioactive cloud) to the population during the TM1 accident.
The main TLD instruments were located within a 15-mile distance of the plant.
Radiation doses to -individuals living within a few miles of the nuclear plant were relatively low; some 260 people living mostly on the East bank of the Susquehanna River possibly each received between 20 and 70 mrem.
One person on a nearby island for 91/2 hours during the day of the accident received about 50 mrem. All other persons living outside a.1-mile radius and within 10 miles from the plant could have received an average dose of less than 20 mrem. Almost all recorded excess exposure above background levels occurred within a 10-mile radius.
There was no recordable radiation levels above natural background at a distance greater than 10 miles from the nuclear plant at any time during the accident.
- 1. 2. 2.
Radioactivity Released: Source Term The total amount of radioactivity released into the atmosphere from
]
Fabrikant 4 the damaged power plant durirl the period of March 28 to April 15, 1979 was calculated,to be about z.4 million curies, primarily consisting of the radioactive noble gas xenon-133. Approximately 10-15 curies of radioactive iodine as iodine-131 was released into the environment. The amount of radioactivity released into the environment has been estimated to be from 2.4 to 13 million curies, consisting almost entirely of xenon-133.
(Inrecentreportstherearereferencesto10,000curiesusedbyC.Berger in the 1980 ORNL Report. The number was " pulled out of the air" by W.K. Stratton as a source term to provide a value for comparing doses calculated by the relatively simple code used by C. Berger with the huge code used by J. Knox at LLNL. W.K. Stratton, and those associated with the effort, recognized that any value would do, but that a big number would provide good statistical results in less computer time. The 10,000 curies was never contemplated to be the real release. This is an example wherein the basis of the number can be misrepresented by those desirous of discrediting the TMI-II accident dosimetry.) This total release of.
radioactivity, known as the source term, was one way to determine the radiation dose to the entire population (collictive dose) and to the individual in the population (average dose), taking into account meterological conditions and population distribution of the population at the time of the nuclear accident. Another way to determine the collective dose was by use of the TLD radiation dose measurements.
1.2.3 Collective Dose and Average Dose in the General Population The collective dose to the population is a measure of the potential health impact resulting from the total radiation dose received by the entire population; for the Three Mile Island site, a 50-mile radius and approximately 2,163,000 persons were included in the calculation.
Since L'
Fabrikant 5
'this value is obtained by suming the estimated radiation dose's, measured in rem, received by each person in the affected area, the collective dose unit is the person-rem. The collective dose to all persons living within a 50-mile radius of TMI and outdoors' based on the TLD radiation dosimetry was estimated to be about 2800 person-rem.
Since most people spent a large amount of their tima indoors and were therefore partially shielded by buildings, and since the radiation dose indoors was about three-quarters of that outdoors, a more accurate collective dose to this exposed population was estimated to be about 2000 person-rem.
The average dose to any individual in the population living within 50 miles of the nuclear reactor, therefore, was estimated to be about 1 mrem. The average dose to an individual living within 10 miles of the plant was estimated to be about 1 mrem.
The average dose to an individual living within 10 miles of the plant was estimated to be about 6.5 mrem.
There were a number of ways to evaluate the magnitude of the radiation releases and the exposures to the general population.
If the maximum dose to any member of the public exposed within just a few miles of the reactor site was no more than 70 mrem, this could be considered to be equivalent to about one-half of the normal exposure the average American receives from natural background radiation each year; probably no more than 250 persons out of the entire population could have received this dose, and most of them received less. Another way of considering it was that this dose was equivalent to the difference between annual background radiation exposure in Harrisburg, Pennsylvania and Denver, Colorado. An average dose of 6.5 mrem is about 5 percent of the exposure from natural background annually in Harrisburg, and equivalent to the difference of living 2 weeks in Denver. The 2 mrem average exposure to persons living within 50 miles
Fabrikant 6 -
.of the nuclear reactor is far less than each person would receive over many years from coldr television, or about the exposure from cosmic radiation during a few round-trip transcontinental commercial jet airplane trips between San Francisco and Washington, D.C.
1.2.4 Internal Dose The radioactivity released during the accident entered the air, water, soil and food, and could ultimately have become incorporated into the human body by breathing it in, swallowing it, and absorbing it through the skin. This could result in an internal radiation dose to the tissues of the body.
Increases in the radionuclide concentrations of iodine-131 were reported in cows' and goats' milk, and in water and air; of cesium-137 in fish, and of xenon-133 and krypton-85 in air. The highest doses due to ingestion and inhalation of iodine-131 would occur in the thyroid gland, since iodine concentrates in that gland. However, whole body scanning of a large number of the general public living near TMI during i
the accident detected no radioactive iodine in this population; no radioisotopes related to the TMI accident were found.
The internal radiation dose due to ingestion of cesium-137 was negligible. The internal dose from inhalation of xenon-133 and krypton-85, primarily due to radiation exposure to the lung, was only a small fraction of that of the external dose.
Overall, the internal doses due to the radioisotopes released at Three Mile Island were negligible, and would have been only a minute fraction of the average annual dose received due to naturally-occurring internally-deposited radioisotopes in the body.
There has been criticism that the environmental monitoring done during the accident could have been inaccurate and incomplete.
The President's
v Fabrikant 7 Commission reviewed all the available dosimetry with great care and in great detail and there was little doubt that the extensive environmental monitoring _ based on thennoluminescent dosimetry measurements and food sampling were adequate to characterize the nature and extent of the radionuclides released and the concentrations of radionuclides in those media. The measurements performed by the Department of Energy (aerial surveys), by the Metropolitan Edison Company, by the Nuclcar Regulatory Commission (ground level dosimeters), by the Environmental Protection Agency, and by the Food and Drug Administration were extensive and suffic-ient to characterize the magnitude of the collective dose and therefore to assess and estimate any possible or potential long-term health effects (see below).
1.3 The Maximum Radiation Dose Received by an Individual The maximum dose that an individual locatee offsite in the populated area could receive was less than 70 mrem. This estimate was based on the cumulative dose (83 mrem) recorded by an offsite dosimeter at 0.5 mile east-northeast of the nuclear reactor site and assumed that the individual remained outdoors at that location for the entire period from March 28 through April 15.
The estimated dose could apply only to individuals in the immediate vicinity of the dosimeter site.
An individual was identified who had been on an island (Hill Island) 1.1 miles north-northwest of the site during a part of the period of higher exposure. The best estimate of the dose to this individual for the 91/2 hour period.he was on Hill Island (March 28 to March 29,1979) was about 50 mrem.
1.4 The Highest Radiation Doses Outside the Nuclear Plant Some of the Metropolitan Edison Company TLDs located on or near the Three Mile Island Nuclear Station site during the first day of the accident
Fabrikant 8 i
recorded. net cumulative doses as high as 1000 mrem.
These recorded readings did not apply directly to any persons located offsite.
1.5 The Principal Radionuclides Released to the Environment The principal radionuclides released'to the environment were the radioactive xenons and iodine-131. Aerial survey measurements made by the Department of Energy in the environment, measurement of the contents of the waste gas tanks, of the gases in the containment Building and the actual gas released to the environment confirmed that the principal radionuclide released was xenon-133.
Xenon-133 is a noble gas, is chemically nonreactive in the body, and does not persist in the environment after it disperses in the air.
It has a short half-life of 5.3 days and produces both gamma and beta radiation.
The risk to people from xenon-133 is-primarily from external exposure to the gamma radiation.
1.6 The Beta Radiation Dose from Xenon-133 Beta radiation contributes to radiation dose in an individual by inhalation and by skin absorption. The total beta plus gamma radiation dose to the skin from xenon-133 is estimated to be about 4 times the dose to the internal body organs from gamma radiation. This contribution would be considerably decreased by clothing. The total beta plus gamma radiation dose to the lungs from inhalation of xenon-133 increased the dose to the lungs by 6 percent over that received by external gamma exposure.
e 1.7 Radionuclides Found in Milk and Food Iodine-131 was detected in milk samples during the period March 31 through April 4,1979. The maximum concentration measured in milk (41 mci / liter in goat's milk, 35 pCi/ liter in cow's milk) was 300 times
. ~.
Fabrikant.9 lower than the level at which the Food and Drug Administration (FDA) would recommend that. cows be removed from contaminated pasture. Cesium-137 was also detected in milk, but'at concentrations expected from residual fallout from previous atmospheric atomic weapons testing, particularly the previous Chinese atomic weapons. testing.
No reactor-produced radioactivity was found in any of the 377 food samples collected between March 29 and April 30,1979 by the Food and Drug Administration.
1.8 The Important Biological Differences Among the Different Radionuclides of a Nuclear Reactor Accident 1.8.1 The Noble Gases: Xenon and Krypton Xenon-133 and krypton-85 are noble gases. Xenon-133 has a half-life of 5.3 days, and krypton-85, about 10.7 years. Other forms of these radio-nuclides include xenon-135 and krypton-84m, krypton-87, and krypton-88.
Xenon-133 is among the isotopes having the largest. radioactive inventories in the nuclear reactor core; krypton-85 is also present. in large amounts.
However, the noble gases are chemically and biologically inert.
The noble gases rapidly diffuse through the atmosphere worldwide, and have
.very limited health effects.
Xenon-133 and krypton-85 both emit beta.
and gamma radiation.
The only health effect to be expected would be primarily through direct exposure to high levels to the skin from beta radiation.
1.8.2 Iodine-Iodine-131 (half-life, 8.1 days), cesium-137 (half-life, 30 years) and strontium-90 (half-life, 28 years) are also radionuclides in the nuclear reactor core which can be released in the event of a nuclear reactor
Fabrikant 10,
~
accident. There are a number of radioisotopes of iodine, cesium, and strontium. These radionuclides, however, are all chemically active, and therefore biologically active, and they become incorporated in the body tissues at specific sites.
Iodine is a functional part of the thyroid gland; iodine-131 when ingested or inhaled into the body, is transported by the blood stream directly to the thyroid gland, and is incorporated into the cells of the gland. The gland cannot differentiate between-radioactive and non-radioactive iodine.
Iodine-131 has a short half-life, but when fixed in the gland, its beta and gamma radiation can injure the cells of the gland, leading to diseases of the thyroid, including cancer.
1.8.3 Cesium Cesium-137 disperses itself in the.various tissues of the entire body, entering the water inside and around the cells.
It is therefore most commonly found in the muscles of the body, which have the greatest amount of tissue bulk, as well as in the gonads.
It emits primarily gamma radiation.
1.8.4 Strontium Strontium-90 and strontium 89 (half-life, 52 days) have a special affinity for substituting for calcium in the body, and they are therefore readily found not only in the growing bones and teeth of young people, but also in the bones of adults. These primarily emit beta radiation.
Thus, the tissues or organs that are irradiated include both bone and bone marruw, and can cause diseases of these tissues, including cancer.
f a
2.0 Health Effects of the Accident at Three Mile Island
Fabrikant~11 2.1 The Health Effects on the General Population Due to the Radiation Released During the Three Mile Island Nuclear Accident Some release of low levels of radioactivity normally occurs into the environment during the routine operation of a nuclear reactor power plant The accident at Three Mile Island set off a series of events that raised the threat of risks to health of much higher levels of radiation exposure of the public to uncontrolled releases of radioactivity.
Low-level fonizing radiations (e.g., radiation doses of a.few rem or less) are thought to be able to contribute to three kinds of health effects.
First, some of the cells injured by radiation may occasionally transform into potential cancer cells, and after a period of time there may be an increased risk of cancer developing in the exposed individual. This health effect is called carcinogenesis. Second, if the embryo of fetus is exposed during pregnancy, sufficient radiation damage in developing cells and tissues may lead to developmental abnormalities of the newborn.
This health effect is called teratogenesis. Third, if radiation injures reproductive cells of the testis or ovary, the hereditary structure of the cells can be altered, and some of the injury can be expressed in the descendants of the exposed individual. This health effect is called mutagenesis or genetic effects.
There are other health effects of ionizing radiations, but these three important health effects---carcinogenesis, teratogenesis, and genetic mutagenesis---stand out because it is possible that low levels of radiation may increase the risk of these delayed health effects.
These observations have led to public confusion and fear about the possible health effects of low-level ionizing radiation from the radioactive releases during the nuclear accident at Three Mile Islande i
j
8 Fabrikant 12*
e 2.2-Cancer
~ The estimated number of cancer cases from all causes normally occurring
~
in.the population living;within 50 miles of Three Mile. Island of about 2,163,000 people over their remaining lifetime is 541,000 (325,000 fatial cancers and 216,000 non'-fatal cancers). The ~ estimated excess number of fatal and non-fata1' cancers associated with the increase in radiation
~
exposure due to the-TMI nuclear a'ccident based on.a collective dose'of:
2000 person-rem to the population was extremely low, and could be zero, and.it would not be' possible to detect or to distinguish this either in ~
the population or-in the individual. The number of' excess cancers, if any, would be so small, tilat it would not be possible to detect such an increase statistically.in over more than half a million cancers that would occur in the population even if the Three Mile Island accident had not happened.
Furthermore, cancers caused by radiation are no different.from any other cancers resulting from other causes; therefore, a particular cancer cannot be distinguished as having been caused by radiation. The-additional radiation-induced risk of cancer due to beta radiation and internally-deposited radioisotopes were estimated to be extremely small, and may be regarded as encompassed within the cancer risk values expressed.-
for whole-body radiation exposure. The conclusion, therefore, was'that there may be no additional cancers resulting from the ' radiation released during the accident.
If there are any additional cancer cases, however, I
the number will be so small that it will n'ot be possible to demonstrate this excess or to distinguish these cases among the 541,000. persons (in the 2 million population) living'within a 50-mile radius of Three Mile Island, l
L who would for other reasons develop cancer during the-course of their i
i lifetime.
h
Fabrikant 13
.4-a 2.3 Genetically-Related Ill-Health During the accident at Three Mile Island, the collective dose to the reproductive cells of the testes and the ovaries of the 2 million persons living within 50 miles of the plant was about 2,000 person-rem, with an average individual dose of 1 mrem.
In this population, assuming a 30 year generation time, there would be expected about 3,000 cases of genetically-related ill-health among the approximately 28,000 live children born each year; these are unrelated to the radiation from the nuclear power plant accident.
From an additional dose of 1 mrem above natural background radiation, there would be expected about 0.0001 to about 0.002 additional radiation-induced cases of genetically-related ill-health, representing less than 1 in 10 million live births.
This may result ultim-ately in no more than 1 additional case of genetically-related ill-health in liveborn children during all generations in the future. This number of " additional cases" is so small that it can never be detected or distinguished, if it does occur, among the spontaneously-occurring (in the absence of any added radiation exposure) cases of genetically-related ill-health in each generation during all future human existence.
The conclusion, therefore, was that it is probable that there will be no detectable cases of genetically-related ill-health resulting from the radiation exposure to the general population following the accident at Three Mile Island.
2.4 Developmental Abnormalities of the Newborn To the approximately 2 million people who live within a 50 mile radius of Three Mile Island, it was estimated that about 28,000 children woulri be born in 1979.
In this newborn population, about 300 children waad normally be expected to be born with developmental abnormalities in the absence of any added radiation exposure as a result of the accident
Fabrikant 14,
at TMI.
The estimated average individual radiation dose to the fetus of pregnant women exposed during the accident was below any threshold dose level known to cause detectable cases of developme. ital abnormality
- in the human embryo of fetus.
The conclusion was that no case of developmental abnormality may be expected to occur in a newborn child as a result of radiation exposure of a pregnant woman from the accident at Three Mile Island.
2.5 Behavioral Effects The most important health effect of concern that occurred at Three Mile Island was not directly due to radiation exposure.
This was the mental health andbehavioral effects on the general public and the nuclear-power plant workers.
Studies by the behavioral scientists of the President's Commission revealed significant mental health and behavioral effects both in the general population and in the workers. The Three Mile Island nuclear accident had a pronounced demoralizing effect on the general population living in the Three Mile Island area, including its teenagers and mothers of preschool-age children.
However, this effect proved transient _in all groups studied except the workers, who continued to show relatively high levels of demoralization four months after the accident.
Moreover, the groups in the general population and the workers, in their different ways, had continuing problems of trust of authorities that stem directly from the nuclear accident.
For both the workers and the general population, the mental health and behavioral effects were understandable in terms of the objective realities of the threats they faced during the accident at Three Mile Island.
3.0 Krypton-85 1
+
~
Fabrikant 15 3.1 Preparatory Studies The initial step of decontamination of the containment building was the removal of radioactive krypton-85 required for cleaning up after the accident.
It was necessary to remove 44,000 Ci of krypton-85 from the containment building atmosphere to pennit sufficiently safe conditions in the environment in which the recovery could proceed.
Metropolitan Edison Company proposed a plan in November 1979 to remove the krypton-85 by-venting it to the atmosphere under controlled conditions. This was followed by the Nuclear Regulatory Commission study submitted as an environmental impact study.
Subsequently, Governor Thornburg of Pennsylvania had two studies carried out, one by the Union of Concerned Scientists, and the second by the National Council of Radiation Protection and Measurements, assessing the radiation dosimetry of krypton venting into the atmosphere, and hence, the potential delayed health effects.
Hearings were held on methods and alternative approaches, such as
. cryogenic trapping. The extensive studies in preparation for the krypton venting to the atmosphere indicated that release under carefully defined and controlled conditions involving specific meterological conditions could be designed to achieve maximum atmospheric dispersion with minimum radiation exposure to the general popuP J
.5 living within the vicinity of Three Mile Island.
~ 3.2 Radiation Doses The krypton-85 venting from the Containment Building to the atmosphere occurred during a two-week period, from June 28 to July 11, 1980. Over this period, measured radiation doses at all defined locations were substan-tially below projected doses derived from computer-generated models. The radiation dose to the population living in the immediate area was estimated
-Fabrikant 16' to be 4.5 mrem to the skin and 0.0045 mrem whole-body dose.
From cryogenic samples taken at 0.5 mile from the nuclear plant an actual integrated beta dose to the skin of about 1.8 mrem was computed; an integrated population collective dose equivalent to the whole-body (gamma dose) of less than 0.03 person-rem was also calculated. These doses were quite small, with no potential health impact to the general population or the workers. The venting made possible the decrease in radiation dose concentration within the containment building with which the workers would come in contact as they began to remove the radioactive waste water, i.e.,
a decrease in dose rate by'about 200 rems per hour beta radiation to the skin, and about 1600 mrems per hour gamma radiation to the whole body.
4.0 Worker Exposure During the Clean-up of the Damaged Three Mile Island-2 Nuclear Power Plant The Safety Advisory Board of TMI-2, was constituted early in 1980 to provide expert scientific, engineering, and medical advice for guidance for the safe clean-up and recovery of the damaged TMI-2 nuclear power-plant.
The scientific advisors reviewed the Nuclear Regulatory Commission's December 1983 draft of Supplement No.1 to the Programmatic Environmental Impact Statement (PEIS) (NUREG 0683). The Safet'y Advisory Board of TMI-2 submitted a number of comments to the Nuclear Regulatory Commission concerning the Report in general, and Supplement No. 1 in particular.
4.1 Collective Dose Eauivalentfor TMI-2 Workers The range given in the PEIS Supplement No.1 of the estimate of the collective dose equivalent for workers expected to occur in the course of the TMI-2 recovery operations of 13,000 to 46,000 person-rem appeared to
Fabrikant 17 represent a more realistic assessment than the estimates proposed in the original PEIS, 'particularly since so much more data on the status of the damaged plant were available~.
As the clean-up progresses, the ranges of uncertainties will narrow depending on the engineering technologies developed and applied to the tasks, and as additional data become available to define subsequent' tasks.
These will impact the proposed collective dose equivalent assigned to each subsequent 'of concurrent major activity.
Thus, while the estimates proposed reflect the current status, it may be necessary to revise of at best narrow the range of estimates as the clean-up of the plant progresses safely to completion.
4.2 Potential Delayed Health Effects The conservative estimates of potential delayed health effects by the Nuclear Regulatory Commission Staff were in accord with current scientific and medical knowledge, and were consonant with the methods of risk assessment used by the International Commission on Radiological Protection, the United Nations Scientific Committee on the Effects of Atomic Radiation, the National Council on Radiation Protection and Measurements, and the National Academy of Sciences-National Research Council.
The Nuclear Regulatory Commission Staff estimates were statistically-derived numerical values and were conservative within the prudent philosophy of radiological protection of the workers and the general public.
Based on. current radiobiological knowledge and theory the numberical values could be considered as an upper bound, and the uncertainties associated with such risk estimates, derived by linear extrapolation from radioepidemiologic data at high doses, include the statistical probability that no delayed health effects could occur.
1
(
w-+
.n n.
.Fabrikant 18.,
LThis information can be used as a basis for radiation protection.
guidance in the special situation of the TMI-2 clean-up; the guidance or
. standard should be related to. risk. Whether the magnitude of the risk Lshould be considered acceptable or not depends largely on how avoidable it is, and to the extent not avoidable,-how it compares with the-risks of alternative options and those.nonnally accepted by the individual t-i-
- or by society in everyday life.
Evaluation of the adequacy of an occupational health standard, regulation, or guideline must ~ consider whether the g.
potential incremental risk imposed is regarded as acceptable _to the worker, both in the workplace'and in his way of life.
Such judgements are necessarily subjective; the currently proposed estimates of collective i
dose equivalent are believed to impose potential health risks to the
{
workforce that should be acceptable to them, and to society in general, since the risks, in perspective, are extremely small.in comparison to other risks that are now readily accepted.
4.3 Radiological Protection Data'for the Clean-up, 1979-1983 Recently available radiological protection data for the clean-up, 1979-1983, indicated that during the five-year period since the accident, i
approximately 16,750 worker-years were involved in the clean-up process resulting in a collective dose equivalent of less than 1700 person-rem.
Of the 16,750 worker-years, two-thirds recorded no measurable radiation i
exposure, and 85% involved doses of less than 0.1 rem per year, that is, less than the average annual whole-body dose received by all persons from natural sources of ionizing radiation. Moreover, a dose rate of 1
0.1 rem per year is considerably less than that received from all sources (including natural background radiation, medical'and dental radiation, commercial air travel, etc.) other than occupational exposure.
l.II~
_a.___,a-,
Febrikant 19 Occupational exposure levels in the range of natural background radiation are considered to represent negligible risks to individual workers.
For-
- example, a dose rate of 0.1 rem per year is only one-fiftieth of the ar.nual maximum permissible dose for occupational exposure recommended by national and international standard-setting bodies, including the Nuclear Regulatory Comission. The annual collective dose equivalent to the TMI-2 workers (1979-1983) consis'ted primarily of values considerably less than 0.1 cem. The risk of developing a delayed health effect, such as cancer, from a dose of 0.1 rem is considered to be about 1 in 100,000 (or about 10-4per rem) and this order of risk is generally considered by society as a negligible incremental risk to the individual.
The recorded radiation monitoring data demonstrated that approximately 96% of all TMI-2 clean-up workers received less than 0.5 rem per year, or less than 10% of the annual permissible dose. Of the remaining 4%
of the worker-years of exposure, no worker received more than the maximum permissible dose.
This record is an excellent achievement considering the immense engineering problems encountered and the unique nature of the work involved in.the clean-up process.
4.4 Precision of Estimates of Delayed Health Effects The Nuclear Regulatory Commission PEIS Supplement No. 1 determined that
.the revised estimates of worker exposure necessary for the clean-up process. (range 13,000 to 46,000 person-rem for a population of some 10,000 workers) would result in "from 2 to 6 additional deaths among these workers d'ue to cancer and from 3 to 12 additional genetic defects among their offspring.H Over the entire period of the clean-up process, the dose commitments associated with the recovery will be no greater than
Fabrikant. 20 -
those stated, and the numerical values for potential health risks estimated most likely represent an upper bound, and will be less.
The statistically-derived values presented by the Nuclear Regulatory Conmission Staff
- denoted a level or precision that was not warranted; while the estimates are conservative, they are also extremely small.
Furthermore, the estimates must not be taken to represent more than crude estimates of risk, based on the incomplete nature of the data at present available.
Several factors, not taker. into account in the calculation of those estimates, exist which compound the uncertainty of the members.
- First, the scientific evidence indicates, for experimental animal and human data, as well as theoretical considerations, that for exposure to low-LET radiation at low doses, the linear dose-response model probably leads to overestimates of the risk of most cancers, but can be used to define the upper limits of risk.
Second, in these calculations, no allowance was made for the likelihood that the carcinogenic ormutagenic effectiveness of low-LET radiation was reduced at low dose rates through the action of biological repair processes.
Third, the individual cancer risks used in the derivation of these numbers may rise of fall as the follow-up of the radioepidemiological study groups from which they are ultimately derived is extended to longer periods.
Fourth, the risks have been derived for the most part at high total doses (which may have been sufficient to inactivate potentially susceptible cells from which a cancer 2
might result), and linear extrapolation could tend to overestimate risk of low-LET radiation.
Fifth, the numerical values of the risk estimates derived from radioepidemiological surveys are themselves crude and uncertain and often have wide statistical confidence limits.
These uncertainties are made even wider by uncertainty about the dose-response relationship and the risk projection model.
7 Fabrikant 21 However, the uncertainties tend in the main to emphasize the conservatism of the risk est'imates as presented by the Nuclear Regulatory Commission Staff. This is clearly the situation where the linear hypothesis is applied.and no allowance is made for biological repair processes; where age-distribution relative to potential reproductive performance is not considered; and where upper-bound uncertainties derived from high-dose and high dose-rate data and extrapolated to the region of low doses and low dose-rates tend to a multiplicative effect in the calculation of risk estimates. These overestimates may serve to offset any calculations that argue that these numbers reflect cancer deaths, and do not therefore represent the number of individuals affected, or that they are based on absolute risk projection models rather than relative risk projection models for predicting future risks to an exposed worker population.
If expressed in terms of cancer incidence, including non-fatal cancers, estimates of risk could be higher by a factor of roughly 1.5 considering the predominance of men in the workforce. And whereas within a particular homogenous population the protection of future risk may probably best be done on a relative risk basis, as yet no firm conclusions can be drawn as to the appropriatene'ss of either model for projection forward in time wihtout further years of observation of irradiated populations.
- However, the current evidence indicates that estimates of lifetime excess cancer risk may vary only by a factor of 2 or 3, depending on which projection model is chosen.
l 4.5 Other Reports of Record
)
Differing viewpoints may exist to those of the Nuclearr Regulatory Commission which oppose the PEIS Supplement No.1 in an effort to challenge the range of the calculated estimates of the worker collective dose equiv-alents or the potential delayed health effects that could occur.
These I
Fabrikant 22
. positions are not unique to the clean-up of TMI-2, but rather tend to apply to many of the societal activities involving ~the use of ionizing radiation.
Frequently these viewpoints are not predicated on sound scientific evidence, but rather on controversial or incomplete reports or personal statements, either 'that are in conflict with the preponderance of scientific evidence on radiation dosimetry or on existing methods for estimating th' delayed-health effects on populations of exposure to low e'
levels of. ionizing radiation.
Several such reports have been published some recently, seeming to claim degrees of carcinogenic radiation effects at low doses _ in humans that would be incompatible with the linear hypothesis being conservative, and may even underestimate the effects at low doses and dose-rates. Many of these studies are limited due to in::omplete data bases, inadequate dosimetry, confounding factors, unconventional statistical methods, or unconfirmet results.
The situations individually or collectively are not convincing enough to argue against the conservatism associated with the linear hypothesis, nor do they provide evidence that the risk of cancer from low-dose radiation is greater than indicated by conventional estimates. These claims compel no scientific reason for national and international standard-setting groups to abandon the body of epidemiologic evidence on radiation-induced cancer that, although based on greater exposures, yields consistent and statistically stable risk estimates.
Fabrikant 23 5.0 An Assessment of the 1984 Beyea Report 5.1 Introduction Any evaluation which will assess the credibility, validity, and degree of certainty associated with the findings and conclusions of the 1984 Beyea Report can be misinterpreted depending, in large measure, on the position of the reader. Although undoubtedly other. interpretations of these three concepts are reasonable, we see them as representing the following questions:
1.
Credibility:
Does the report as a whole portray to the scientific community a consistent, believable picture of the dosimetry of the accident and of our understanding of the risks of radiogenic cancer? Do the various assumptions required seem to fit together and relate easily to plausible mechanisms of the radiation releases, of the radiation exposures impacting the general population and workers, and of radiogenic cancer induction in exposed human populations?
2.
Validity:
Does the analysis of the radiation dosimetry conform with recorded data where it is possible to observe them? Do the cancer risk estimates in the Beyea Report conform.with observed risks of radic-genic and nonradiogenic cancer where it is possible to observe them? Do any properties of the methods used seem to violate fundamental principles or empirical observations?
3.
Degree of Certainty: How good are the numbers in the risk estimates in the Beyea Report? Are they biased in one direction or the other? What are the consequences of the likely differences between the true estimates and thoseestimated in the Beyea Report? Where do the uncertainties come from, and how could they be reduced? What are the uncertainties in the information needed to assess potential cancer risk estimates in the Beyea
Fabrikant 24 Report?
The folloWing is ~an overview assessment of the 1984 Beyea Report L
_ from the point of view of the radiation dosimetry of the accident at Three Mile Island, and the potential delayed or late health effects i
which could result in the general population and the worker population from exposure to the low level radiation releases at the time of the accident.
5.2 The Radiation Dosimetry of the Accident at Three Mile Island Section 1 of this report provides a synopsis of the dosimetry of the accident at Three Mile Island.
It is derived completely from the Staff l
Reports (see summaries, Appendix A and Appendix B) of the Health Physics Task Group of the Public Health and Safety Task Force Report of the Staff Reports to The President's Commission on the Accident at Three Mile Island. The " Summary of the Health Physics and Dosimetry Task Group.
Report" taken from the Reports of the Public Health andSafety Task Force isattached(AppendixA). On pages 8, 9, and 10, a number of important statements are made, e.g.,
'TLD measurements formed the basis for estimating the total external gamma radiation doses (due almost exclusively to the radioactive noble gas xenon-133 and a few other short-lived radioactive gases in the radioactive cloud) to the population during the TMI accident.
The total release of radioactivity into the atmosphere from the damaged nuclear power plant during the period of March 28 to April 15, 1979, was calculated to be about 2.4 million curies, primarily consisting of radioactive noble gases.
Approximately 10-15 curies of radioactive iodine were released into the environment. This total release of radio-activity, known as the source term, was one way to determine the radiation doses to the entire population (collective dose) and to the individual
Fabrikant 25 in the population (average dose), taking into account meterological weather conditi'ons and population distribution demographic data at the time of the accident. Another way to determine the collective dose was by use of the TLD radiation dose measurements.
- A more accurate collective dose to this (2,163,000 persons living within a 50-mile radius of the TMI stie) exposed population is estimated to be about 2,000 person-rems above normal background levels.
- 0verall, the internal (radiation ) doses due to radioisotopes j
released at TMI were negligible, and would only have been a minute l
fraction of the average annual dose received due to naturally occurring, internally deposited radioisotopes in the body.
- The collective dose for these 1,00L (TMI) workers from the time of i
the accident on March 28, 1979, through June 30, 1979, was about 1,000 person-rems.
These findings were based on analyses of all available relevant and reliable data of record from the public and private sectors by August 31, 1979. The analyses were carried out by a team of six internationally known radiation physicists from national laboratories (from the j
United States and Canada) and University centers, in conjunction with 25 scientific colleagues, consultants, and advisors from the major national laboratories and universities in the United States and Great Britain.
5.3 The 1984 Beyea Report Criticisms of Reports of Record The 1984 Beyea Report, with a team of six unknown (and with no university or research laboratory affiliations cited) persons attempts to criticize the President's Commission Report, and reports by the Nuclear Regulatory Conunission, Department of Energy, Environmental Protection Agency, and scientific papers of record on the dosimetry of the accident at TMI. The Beyea Report states repeatedly that there are gaps in the
7 Fabrikant 26 published data and large uncertainties, and concludes that the missing data results in serious underestimates of the collective dose to the population.
'.4hile these uncertainties could very well result in the conclusion that the-doses were, as all studies concluded, extremely low, the Beyea Report prefers to conclude the contrary, viz., large radiation releases went undetected. By compounding uncertainties, preliminary assertions, and examples of data admittedly uncorroborated or unsubstantiated, Beyea.
concludes that all the low-dose estimates are incorrect, and all the high-dose estimates appear reasonable, if not correct.
The Beyea Report is not a scientific report on the dosimetry of the accident.
On the contrary, it is an overt attempt to find-the limitations---
both ommissions and comissions---which lead to " official" estimates of doses of radiation released during the TMI accident.
It demonstrates profound bias in failing to address the strengths of the methods, and the reliability and credibility of the observed or calculated data.
The Report lacks scholarship, is poorly bdlanced, and is frequently pejorative.
Statements made are incompletely referenced, and frequently not supported at all by the literature.
Conclusions are drawn that are frequently uncritical.
The Report does not identify the faults in dosimetry estimates of the President's Commission Report, the Regovin Report, or others, but prefers to argue that the gaps in. data were incorrectly reconstructed.
It is clear, the Beye' group has difficulties with dosimetry terminology a
and units:
it refers to" population doses" rather than the correct unit,
" collective dote", and such uncritical terms and units "a 300-rem thyroid population dose" and a " delayed radiation dose" are used.
The President's Commission assessment of tt e radiation dosimetry of the accident at Three Mile Island is among.the rost complete and scholarly
Fabrikant 27
.of-all the reports extant.
Its staff of scientists and expert consultants with access to.the most extensive network of reliable scientific groups, methods and technologies in radiation dosimetry, demographic and dosimetric data, computational mathematics and computer sciences, both in the United States and abroad, insured the completion of a comprehensive and scholarly report to the Comission in a timely fashion. The 1984 Beyea Report places emphasis on uncritical remarks of the Comission's report, as it does on other official (NRC, DOE, Rogovin) reports, with unfounded and.
frequently pejorative statements, e.g., in its Introduction, p.1:
"Because the major studies on the subject were undertaken in the months soon after the March 28, 1979 accident, and completed under considerable pressure for immediate findings and reassurances, it is not surprising that these official studies cannot provide complete, scientifically justifiable answers.
Subsequent studies in the scientific and engineering literature have not resolved the residual uncertainties."
And further, on p.2:
1 "On the contrary, the investigators reviewed in this (Beyea) study were found to have been extremely clever in using a combination of inference and science to extract information l
from limited data. Problems remain because a great deal of crucial data does not exist, or is unreliable.
Researchers have been forced to replace the missing information with assumptions and to manipulate, as best they can, the unrel-iable data."
Such statements are unsupported, uncritical, unsubstantiated, and un-corroborated. They cannot, and should not, be considered scientific, credible, valid or reliable.
5.4 The Beyea Report and " Doses to the Whole Body" The Beyea Report, in its Tables 1, 2, and 3 lists a number of reports, and concludes, "The TMI literature contains a substantial range of whole-body population dose estimates from the noble gases released in the initial accident---from 276 to 63,000 person-rem delivered to the general l
population within 50 miles (see Table 1, column 1).
Such a divergence
r Fabrikant 28 -
. is sufficient to indicate the degree of uncertainty on the question."
This statement-is misleading and leads the reader to assume that there is a very wide spectrum of collective dose estimates, all troubled by great uncertainty. However, Beyea is uncritical and biased: he fails to point out that all the" official" estimates---the President's Commission, Lawrence Livennore National Laboratory, Oak Ridge National Laboratory, Environmental Protection Agency, Nuclear Regulatory Commission, Nuclear Safety Analysis Center (EPRI), and Wood (ard (Pickard, Lowe, and Garrick, Inc.1---all fall roughly within 1 order of magnitude,afactor of 10.
In view of the circumstances and data, this must be considered reliable and reproducible. However, Beyea dismisses this observation resulting from the nation's (and international) outstanding scientific comunities, and scientific resources, as uncertain and displaying discordant results, in pite of the extensive scientific documentation extant.
On the other hand, two whole-body collective dose estimates, one by Takeshi and one by Kepford,16,200 person-rems and 63,000 person-rems respectively, are not only given equal weight by Beyea, but are applauded as being more reliable, and are subsequently used to project health consequences or as a basis for future scientific research.
Each estimate was derived by one man, whose scientific credentials are suspect, whose material has not been critically reviewed, nor published in the peer-reviewed literature, and who did not have access to the original data.
Mr. Seo Takeshi is listed, on p. A45 as " associated with the Kyoto Nuclear Reactor Laboratory", and his reference is cited as: "S.Takeshi,
" Excerpts from the author's review published in the Japanese journal Nuclear Engineering, Vol 26, No.
3," (unpublished mimeographed notes, Kyoto University Nuclear Reactor Laboratory, Kyoto, Japan, not dated)."
~
c-Fabrikant 29 Mr. Chauncey Kepford is ~ listed on the same page, as, "a nuclear critic, associated ~at the time with the Environmental Coalition on Nuclear Power,"
and his reference is cited, on the same page, as, "Chauncey Kepford,
" Testimony.before the NRC Atomic Safety and Licensing Board, August 20, 1979, in the matter of Public Service Electric & Gas Co., Salem Generating Station Unit #1,'Dockett #50-272,"(1979)."
The' conclusions are evident; Beyea has placed undue emphasis on two single-authored personal statements, uncritically reviewed, unsubstantiated and unpublished; one is undated and the other calculated and stated in August 1979, within perhaps 1 month of the availability of data on the dosimetry of the accident, and months before the " official" eports on the accident were completed and available. He has rejected the " official" reports as incomplete and unreliable, and therefore suspect.
5.5 The Beyea Report and the Radiciodine Releases Perhaps no section of the Beyea Report is as confusing and redundant as is the inordinate emphasis placed on the radioiodine releases during the accident, and the unaccounted or " missing" radiciodine.
Dr. Merril Eisenbud addresses the question directly and compels the conclusion that (1) if the excessive radiolodine~was released, it would have been readily detected and accurately measured by the experienced governmental agency teams, (2) that whatever remained isnot " missing", but was contained and not released from the damaged core and has subsequently decayed, and (3)the11,000,000 curies (still unaccounted for) was calculated from source term data and codes developed prior to the 1975 WASH-1400 Report, and which may very well be inaccurate and a considerable overestimate.
Dr. Eisenbud's conclusions need no additional supportive evidence; his analysis is cogent and compelling (see Appendix C).
i Fabrikant 30'-
There is additional information now available that makes much of Beyea's; analyses of the radioiodine releases irrelevant, and therefore
-spurious. This is the " Review of Recent Source Term Investigations",
presented 'by William R. Stratton, Ph'.D., formerly of the Los Alamos National Laboratory and chairman of the NRC Advisory Comittee on Reactor Safeguards. The " Review" was presented'in July 1984, and announces the forthcoming report of the American Nuclear Society's Special Comittee.
on Source Terms.
It is presently before~ the Nuclear Regulatory Comission-
~
for evaluation, and a committee of The American Physical Society is presently assessing its methodology and calculations.
In veiw of the present August 15, 1984 release of the 1984'Beyea Report,.it is appropriate to quote directly from sections of the July 1984 Stratton ANS Report.
From page 1 of the Stratton ANS Report, concerning the complete ANS study:
"This review is based largely on the study recently completed by the American Nuclear Society's Special Comittee on Source
- Terms, committee members are: M Christian Devillers, France; M. Sergio Finzi, CEC (alternates, M. William Vinck, M. Anesto Della Loggia, M. Brian Tolley); Dr. Mario Fontana, U.S.A.;
Mr. Michael Hayns, United -Kingdom; Dr. Hans H. Hennies, F.R.G.
(alternative, Mr. Deter Hosman); Dr. Herbert J.C. Kouts, U.S.A.;
Mr. Saul Levine, U.S. A. ; Dr. A.P. Malir tuskas, U.S. A. ; Mr. James F. Mallay, U.S.A.; Mr. Andrew Millunzi, U.S.A.; Mr. Masao Nozawa, 4
{
Japan (alternate, Dr. Ryohei Kiyose); Dr. Walter Pasedag, U.S.A.;
Mr. A. Schuerenkaemper, JRC-Euratom; Dr. Robert L. Seale, U.S.A.
l (Vice Chairman); Dr. William R. Stratton, U.S. A. (Chairman);
i l
Fabrikant 31 Dr. Richard _C. Vogel, U.S.A.; Mr. Edward A. Warman, U.S.A.
Individuals who contributed significantly to the report are:
Mr. An. drew Pressesky, U.S.A.: Dr. Walton Rodger, U.S.A.; Dr.
Thomas Kress, U.S.A.: Dr. Robert Burns, U.S.A."
From p.1, ABSTRACT:
"The' state of knowledge relative to the evaluation of source terms subsequent to a severe reactor accident is examined.
The following matters are assessed:
the methods and assumptions used to describe fission product behaveior and retention associated with various phenomena, response of plant systems and structures, and a summary of source term results obtained by various investigators.
These are compared to results quoted in WASH-1400."
l From pp. I and 2, INTRODUCTION:
"The source tenn means that amount and type of radioactive materials which would be available for escape to the environment l
from a reactor which has undergone a severe accident.
This is an accident in which fuel is damaged by overheating to the point of allowing substantial escape of fission products l
to the containment from the fuel and the containment may not have functioned adequately to prevent the escape of significant amounts of radioactivity to the environment.
l Source terms have been recognized form the early days of nuclear energy development as the important factor of risk.
Because the technology for making accurate and valid estimates of the source term was not available at that time, the conser-vative, non-mechanistic assumption was m.de that essentially all of the fission products could be released from a severely
Fabrikan't 32 -
damaged reactor. This conservative assumption was later slightly modified and incorporated into regulations which are still in force at this time.
This early assumption and the subsequent regulations focussed on radioiodine as the principal substance of concern.
This was because of its relative abundance, its high biological activity (iodine is known to concentrate in the thyroid),
and its assumed elemental gaseous form, which provided ready transportability.
During the Three Mile Island accident in 1979, a surprisingly small amount of iodine escaped to the environment, contrary to expectations based on regulatory prescriptions.
It was then theorized that the iodine, escaping from the fuel into a chemically reducing atmosphere (due to the presence of water and hydrogen) became an iodide, was readily dissolved in the water, and so became unavailable for escape.
Thus, chemistry, which previously had been largely neglected, was seen to play an important role in severe accidents.
Other aspects of severe accident considerations were identified at that time. As a result, large programs to investigate source terms, with the objective of providing a more realistic and accurate estimate, were undertaken by government agencies and industry, both in the U.S.and abroad.
The principal focus of this work was the analysis of severe accident sequences chosen because they represented the upper range of consequences and/or exemplified phenomena believed to be important in understanding the chemical and L
Fabrikant 33 physicalprocesses that detennine fission product behavior'in
~
severe accidents. This work is an extension of the methodology brought to a considerable stage of maturity by WASH-1400 (The Reactor Safety Study,1975), an earlier effort to quantify the risk from nuclear energy.
The American Nuclear Society chartered the Special Committee on Source Terms to examine the state of knowledge relative to the source term, and the methods and assumptions used to describe fission product behavior and retention associated with various phenomena, plant systems, and structures in a severe reactor accident. The Committee was also to provide a summary of source term results obtained by various investigators, and to compare these data to those presented in WASH-1400.
The Comittee recognized that both probability and consequences are intrinsic elements of risk; however, the Committee's charge included only an examination of consequences as predicted by analyses, and these.only up to the point of potential escape of radioactivity to the environment.
The probability of occurrence was examined in a general.way to show that severe accidents are predicted to be exceedingly rare."
I From p. 3, IMPORTANT RADIONUCLIDES:
" Typically, a large number of fission product species exist in the fuel in a nuclear reactor. Radionuclides escaping into the environment in the unlikely event of a severe reactor accident vary in their importance as to potential consequences. The factors determining the importance of a radionuclide in this regard are: 1) its total inventory in the reactor; 2) its physical and chemical properties which n_.-
e~-
..nm,
L Fabrikant 34.
determine its behavior in the plant and the environment; and
~
I
- 3) its biological characteristics.
Some of these factors are inherent, and others depend on feqtures of the accident and plant design; thus, the importance of a radionuclide depends to L significant extent on specific aspects of the hypothetical accident sequence being considered.
Radioiodine has long been and still is considered to be a very important radionuclide. However, it is clear that its treatment has been significantly over-conservative, and even historically incorrect. Other important radionuclides include cesium, tellurium, and, of much lesser importance, some of the alkaline earths and noble metals.
Like iodine, the importance of cesium also has been previously overstated.
The noble gases, though very volatile, are chemically inert, and thus have a low importance in severe accidents."
From pp. 12 and 13 FINDINGS, OBSERVATIONS, AND RECOMMENDATIONS OF THE COMMITTEE
" Major Finding The Comittee has concluded that the state of knowledge and the analytical methods and assumptions on which current calculations of the source term are based have progressed far beyond those on which WASH-1400 (The Reactor Safety Study, 1975) was based.
In general, an ample foundation has been provided to warrant reductions of the source term estimates in WASH-1400 by more than an order of magnitude to as much as several orders of magnitude. This major conclu-sion is based on reviews of chemical and physical processes relevant to severe accident analysis; severe accident L
7 Fabrikant 35 sequences which bound risk from nuclear power plants and represent the ranges of phenomena involved; the status of severe accident modeling and calculational codes; containment capability; and the results of a number of source term studies performed both here and abroad.
In addition, the comittee has considered studies performed on its behalf of a number of important parameters and phenomena which had not previously been given adequate emphasis. The noble gases are exceptions because of-their chemically inert character, and because they do not underge the wide range of chemical and physical interactions which are the fundamental cause of the reduced release of most fission products; however, the very fact that they are inert also leads to low radiological consequences.
Findings Supporting or Qualifying the Major Finding a)
Iodine will be released and transported predominantly as cesium iodide and cesium as cesium hydroxide.
These species will form aerosols and be subject to aerosol depletion processes, are highly soluble in water, which will be present, and can be irreversibly adsorbed onto metal survaces, resulting in greatly reduced releases compared to WASH-1400.
This finding holds for all light water reactors and all accident sequences, b) The more severe accident sequences developed in WASH-1400 6d more recent Probabilistic Risk Assessment studies provide a sufficiently complete basis for in-depth analyses of source terms. These sequences cover the high end of the release spectrum and involve the phenomena and processes that are considered to affect the escape and transport of u. -. - -
Fabrikant 36.
fission products.
c) Sequences ani plant details are important in estimating plant-specific source terms.
d)
If there is no breach of containment, there is essentially no release of fission products; if containment breach is delayed more than a few hours after core degradation, the source term is greatly reduced, independent of the final size of containment breach. Containment is less susceptible to early breaching than previously believed.
e) A substantial basis exists for knowledgeable analysts to calculate LWR source terms with a high degree of confidence in the results."
In conclusion, the July 1984 Stratton ANS Report (American Nuclear Society's Special Committee on Source Tenns" makes the August 1984 Bayea Report and its Appendices A through E obsolete, inaccurate, and irrelevant, thereby vitiaijits credibility, its validity, andtcertainties.
5.6 The 1984 Beyea Report and Health Impacts On p. 2 of the Introduction, Beyea states: "It should be noted that this re. port does not critically examine the quantitative coni.cction that is made in the TMI literature between radiation doses and projected health effects." Then, Section 6.9, pp 32 to 34, the Report provides a naive "A Sumary of Health IImpacts Described or Implicit in the Literature." Here, Beyea makes error after error in his approach, takes liberties with the established and conservative approaches of radiation protection philosophy and risk estimation, and tries to simplify, as he sees fit, a very complex scientific literature of cancer-induction in human populations exposed to low-level ionizing radiation,
r
(
Fabrikant 37 Without justification, Beyea chooses 'as the upper-bound collective r
dose estimate of 63,000 person-rems (based soley on the Kepford
-the testimony) to general population of over 2 million persons living within l
50 miles of Three Mile Island at the time of the accident, and based r
l on the National Academy of Sciences-National Research Council's BEIR
(
l III Report, projects'a maximum life-time cancer risk of 12.6 excess l
cancer deaths. He cites the President's Commission Report, the Rogovin and Report, Secretary Califano's press conference statement.
In two-and-a-half pages he makes numerous errors of fact, and by compounding uncertainties, preliminary assertions, and examples of data that Beyea admits to be uncorroborated and unsubstantiated, Beyea l
hypothesizes a very large estimate of excess cancer deaths from the radiation releases of the accident.
Some obvious errors are worth citing; however, the greatest error is his simplistic approach to a very complex science.
For example, Beyea states: "Although uncertainty exists about such low-level radiation risks, the (National) Academy (of Sciences 1980 BEIR III Report) projects 1
0.6 to 2.0 delayed cancer deaths per 10,000 person-rem. That is incorrect.
First, the cancer risk estimates for the President's Connission Report, the Rogovin Report, and Secretary Califano's statentent were derived from the NAS 1972 BEIR-I Report, not from the 1980 BEIR III Report. The risk coefficients in the two reports are different.
Second, the BEIR-!!! Report does not present absolute risk estimates i
as probabilities per rad; rather, from p.194 of the BEIR-III Report:
"The final estimates are expressed as the numbers of excess cancers or of excess cancer deaths in an exposed population of 1 million people followed from the onset of exposure to the end of life.
These numbers i
may also be expressed as percentages of the numbers of cancers normally l
1 Fabrikant 38 expected for a, population cohort of that size over the period under consideration and in the absence of the additional radiation exposure.
Their expression per rad is generally avoided in the final tables, because it would suggest a commitment to the linear hypothesis that some members of the Committee wished to avoid, believing that the effect per rad is most probably variable, an increasing function of dose in the region from zero rads up to a point where cell-killing becomes important."
Third, Beyea cites" official dose-response coefficients" and " con-ventional dose / response coefficients" as the basis of his calculation.
This is wrong. He probably means " age-and sex-specific regression coefficients or risk coefficients."
Fourth, even if he choses a collective dose of 63,000 person-rems for 2.3 million persons, the individual doses (average of 27 mrems) would still be too small to justify calculation. The BEIR-III Committee chose whole-body doses of 10 rads administered acutely, or 13 to 14 rads administered continuously (at I rad /yr for males and females, ages 50 to 65 years) as the lowest doses because "Below these doses, the uncertainties of extrapolation of risk were believed by some members of the (BEIR-III) Committee to be too great to justify calculation."
(p.144,BEIR-IIIReport)Furthermore,theCommitteestated(p.139)that "It is by no me'ans clear whether dose rates of gama or x-radiation of about 100 mrads/yr (of background radiation levels) are in any way detrimental to exposed people."
Fifth,- the BEIR-!!! Report chose specific dose-increments for l
computation of excess cancer risk as follows:
" Selection of dose increments for which cancer risk estimates are made was guided by existing maximal pemissible dose limits, information on occupational L
Fabrikant 39 exposure recorded in recent surveys (cf. Chapter III) concern for a hypothetical situation in which some part of the general population might be exposed to a single dose of 10 rads, and uncertainty as to l
whether a total dose of, say, I rad would have any effect at all."
(p.193, the 1980 BEIR-III Report.)
l Therefore, the Beyea calculations based on BEIR-III risk l
coefficients that are not considered either age-or sex-specific l
regression coefficients, are spurious and irrelevant.
l i
l
Fabrikant 40 5.6.1.
The Reasons Why the 1984 Beyea Report Calculations on Health Impacts are Spurious and Irrelevant-l The 1984 Beyea Report attempts the estimation of carcinogenic risk from whole-body exposure to ionizing radiation released during^the accident at Three Mile Island.
In doing so, the report makes a number of assumptions concerning the latent period (or induction period),
selection of a projection model (e.g., absolute risk or relative risk),
and the need for adjusting for competing causes, as by life-table methods.
However, the single assumption that weakens the position taken is the l
l tissue doses absorbed resulting from the accident.
In any scientific endeavor that attempts to organize information for a practical purpose, there are at least three types of uncertainties which the authors have failed to recognize: uncertainties in data; uncertainties in assumptions and models; and uncertainties that are intrinsically not estimable.
1.
Uncertainties in data; these uncertainties arise from an i.
inability to make very precise measurements, either because of inaccurac-les in instruments or because of inherent variability in processes.
The measurement or calculation of doses is an example of the former; counting the number of cancers in a cohort population exemplifies both uncertainties.
2.
Uncertainties in assumptions and models used to analyze data.
A model may fit the observed data in a narrow range, but could be substantially in error elsewhere, either because of inability to estimate risk coefficients precisely or misunderstandings about the nature of the physical, chemical, and biological processes involved.
The models for i
l projecting and extrapolating risks associated with radiation are all l
l subject to such uncertainties.
I
Fabrikant 41 3.
Uncertainties that are intrinsically not estimable; these uncertainties Krise because important phenomena or principles have not yet been discovered.
For example, before the discovery that high-LET radiation was more damaging per unit dose than low-LET radiation, it might easily have been assumed without question that only the dose was
~
important.
The 1984 Beyea Report assumes that the calculation of probability of cancer induction at low doses and low dose rates is a straight forward exercise of the application of simple formulae, taken, for example, from such reports as the National Research Council's 1980 BEIR-III Report.
This is simply not the case.
There is a great deal more to the estimation of risk than size of the population at risk, time since exposure or latency interval, dose, and dose-response function.
Other influential factors include demographic characteristics such as age and sex, quality of radiation, perhaps dose-rate, perhaps host factors that are yet to be identified, e.g., hormonal state, genetic make-up, immune competence, etc., and other enviornmental factors also yet to be identified, e.g., chemical carcinogens.
In listing these factorswe cannot ignore the problem of bias, such as may arise in the comparison of exposed and controls when ascertainment is incomplete or differs in its completeness.
If ascertainment, i.e., the gathering of information on the events of interest, is equally incomplete in both the exposed and the centrol samples (or other source of expected values),
relative risk estimates will not be affected, but absolute risk estimates will be reduced and by the degree of incompleteness.
But if ascertainment is differentially incomplete, then both the relative and the absolute
Fabrikan't 42 -
risk estimates wil.1-be biased. This we know is precisely the case at Three Mile'Islatnd, where large numbers of persons left the area during l
the first week of the^ accident. Thus, ascertainment is incomplete, the f
population 'at risk' is smaller than previously considered, perhaps by a
" half or more, and the absolute risk estimates will be increased by the
~substantial degree of. incompleteness.
i' i
- It is important that estimates be sex-specific not only for tumors peculiar to one sex or the other, but also for leukemia, thyroid cancer,-
probably lung cancer, and perhaps other sites as well. Until we know more about the comparative performance of relative and absolute risk estimates veshould think that any site of cancer for which male and female l
incidence rates differ. would require that estimates of risk be made in 1.
sex-specific fashion.
[
Age at exposure is being recognized as having a major influence on risk, but without our understanding why.
In part its influence may J
l_
reflect hormonal status or other physiological state dependent on age,
{
as in breast cancer, or the time-dependent accumulation of tissue changes j
induced by cocarcinogens.
1 i.
Quality of radiation has long been recognized as a major factor in i.
j the risk or radiogenic disease, and the related concepts of quality factor and relative biological effectiveness (RBE) are used in expressing risk f
estimates in rems rather than rads. Unfortunately, there are only very I
approximate ' estimates of the RBE ratios for radiation of various 3
i qualities and for different end-points.
Dose-rate becomes important if we are using high dose and high v < ep-p we Wvp-*
FT*yi+my-q==g---*-ywe-g er *w i-re gvM'-
g'eW**fw-et'7* V M--
"Pk'
?
W
-=1-tF
't PMf'-%*1+'
9't-g-'*Et'er 1y f*3T"
-Tt"T*Py 1weeei 9 gr --up*1W
-67'17-9---=-*P^=9e-t' F*mk*e+y m' v9W?g #Nq'r e
Fabrikant 43
.~
dose-rate data to estimate low dose and low dose-rate effects, as the BEIR III Committee has done in relying on the experience of the A-Bomb survivors. At Three Mile Island, the ameliorating effect of dose-reduction due to protection of dose administered at low dose rate would serve to reduce the risk estimates substantially. Although the BEIR III Committee reached the conclusion that the dose-rate effect on human tumor incidence was too uncertain to justify a quantitative adjustment for its magnitude, committee 40, working concurrently on NCRP Report #64, decided to recommend a reduction factor of between 2 and 10 when high dose and high dose-rate observations are used with a linear dose-response function to estimate low dose and low dose-rate effects of low-LET radiation. The UNSCEAR 1977 Report used a factor of 2.
With respect to host factors other than age and sex, and carcinogens other than radiation, little can be said except that investigators reporting their results, and those who depend on those results, should be aware of the possibility that such factors may be present and may influence the results in some unexpected way.
Thismightcomeabout because of host or environmental factors that interact with radiation to exaggerate or minimize the effect of radiation, or because of some characteristic associated with exposure to radiation that independently affects the normal expectation of teh effect under study.
Other factors influencing estimates of radiogenic risk are inherent in the various limitations of the underlying observations, e.g., their precision as to diagnosis and dose, but such tactors are not peculiar to i
the estimation of radiogenic risks. A related point concerns the trans-lation of external dose to tissue dose.
If comparable estimates are to
Fabrikan.t 44,
be obtained from different human series, reliable tissue-dose estimates must be availatrie.
During the Three Mile Island Nuclear accident, estimates of external doses.ndicated that only extremely low-level exposure could have impacted a small population during the initial days of the accident.
When transformed to tissue doses, for the majority of the population that could have been potentially exposed, the actual absorbed doses would prove to be negligible tissue doses.
Some reference should be made to specific statistical methods.
In statistics we regard risk estimates as members of a larger set, perhaps an infinite set, of similar estimates and we have ways of placing them within a specific range of values with a pre-determined level of confidence that, it we could repeat indefinitely the experiment or survey leading to the estimate, the proportion of such estimates lying within the specified range, or confidence interval, would correspond to the pre-determined level of confidence. Thus we calculate, for example, 95 percent confidence intervals, or 80 percent confidence intervals, at will, and these estimates are often the most useful ones we can make, far more informative than estimates that do not carry a measure of their inherent variability.
When factors other than numbers of subjects and events, time following exposure and dose must be taken into account, as is usually the case, they must be adjusted for in some fashion or other.
t I
Fabrikant 45 The 1984 Beyea Report fails to recognize that its concern with uncertainties regarding very low doses following the Three Mile Island Accident precludes estimation of the carcinogenic risk to the exposed populations.
In studies of animal or human populations, the shape of a dose-response relationship at low doses may be practically impossible to ascertain statistically. This is because the sample sizes required to estimate or test a small absolute cancer excess are extremely large; specifically, the required sample sizes are approximately inversely proportional to the square of the excesses.
For example, if the excess is truly proportional to dose and if 1,000 exposed and 1,000 control subjects are required to test the cancer excess adequately for 100 rads, then about 100,000 in each group are required for 10 rads; and about 10,000,000 in each group are required for 1 rad.
Thus, risk coefficients based on a knowledge of dose-response relationships can never be estimated for the dose ranges of concern at Three Mile Island.
The BEIR Connittee of the National Research Council concluded that it is not known whether dose rates of gamma or x-rays of about 100 mrads/yr are detrimental to man. Any somatic effects at these dose rates would be masked by environmental or other factors that produce the same types of health effects as does ionizing radiation.
It is unlikely that carcinogenic effects of doses of low LET radiation administered at this dose rate will be demonstrable in the foreseeable future.
For higher dose rates, e.g., a few rads per year over a long period i.e., far in excess of the levels determined during the Three Mile Island accident-a discernible carcinogenic effect could become manifest.
Furthermore, the 1984 Beyea Report assumes that the precision of
Fabrikant 46 '
such cancer risk coefficients as estimated in the BEIR-III Report are precise and cer'tain. However, it failed to recognize that the BEIR-III Committee's most difficult task has been to estimate the carcinogenic risk of low-dose, low-LET, whole-body radiation. BNdddgnized that the scientific basis for making such estimates is inadequate, but it also recognized that policy decisions and the exercise of regulatory authority require a position on the probable cancer risk from low-dose, low-LET radiation. Accordingly, the Committee decided that emphasis should be placed on the assumptions, procedures, and uncertainties involved in the estimation process, and not on specific numerical estimates.
In other words, the BEIR-III Committee recognizes that policy decisions cannot be reached or regulatory authority exercised without someone's taking a position on the probable cancer risk associated with such radiation.
Because critical analysis of the different data bases disclosed major inadequacies, however, the Committee decided to emphasize the assumptions, procedures, and uncertainties involved in the estimation process, and not specific numerical estimates. The variety of mathematical functions that could be used to express dose-response relationships reflects additional uncertainty. Therefore, the Committee concluded that the best method of expressing the range of uncertainty associated with these' problems would be to present an envelope of risk estimates.
The probabilities of cancer induction following exposure to low-dose radiation presents fon111dable problems.
Even for its illustrative computations of the lifetime risk from whole-body exposure, the Comittee
Fabrikant 47 chose three situations: a single exposure of a representative (life-
~
table) population to 10 rads; a continuous, lifetime exposure of a rep-resentative (life-table) population to 1 rad /yr, and an exposure to
-1 rad /yr over several age intervals exemplifying conditions of occupational exposure. The three exposure situations do not reflect any circumstances that would normally occur, but embrace the areas of concern-general population and occupational exposure and single and continuous exposure.
Below these doses, the uncertainties of extrapolation of risk were believed by some members of the Committee to be too great to justify calculation.
Thus, the uncertainties were considered too great to justify calculation of risk below dose levels of 10 rads (whole body) administered acutely or about 75 rads (whole body) administered chronically over a lifetime.
And finally, any attempt to use epidemiological surveys that challenge the conservatism of the linear hypothesis for low-LET radiation exposure would be fraught with failure.
Studies by a number of scientists who have claimed a greater carcinogenic effect due to exposure to low-dose ionizing radiation than generally accepted are reviewed in detail in Appendix B of the BEIR-III Report. None of these studies was considered by the Committee to constitute reliable evidence at present for use in risk estimation, for various reasons, including inadequate sample size in some instances, inadequate statistical analysis, and unconfirmed results.
Published criticisms'of these various study findings have suggested alternative explanations for the observed dose associations, including confounding.
of radiation exposure with exposures to other carcinogens and inadequate dosimetry.
In some instances, only further study can determine the validity of these suggestions.
Further followup of the studies of
Fabrikant 48.
nuclear workens for example, workers' and of other groups occupationally exposed to simi'lar quantities of highly fractionated radiation may eventually tell us whether the risks and the spectrum of affected cancer -
I sites differ markedly from what wotid be expected from studies of more heavily exposed populations. ~At:present, however, there seems to be no reason to abandon the body of epidemiologic evidence on radiation-induced cancer that, although based on greater exposures, yields consistent-and statistically stable estimates.
?
4
?
j..
Fabrikant 49 5.7 The Krypton-85 Releases The Beyea-Report addresses the June-July 1980 venting of noble gases from TMI-2 containment only occasionally and indirectly.
Section 3.0 of this analysis provides the currently known information correctly.
5.8 The Beyea Report Appendix F: A Review of the Cleanup of Three Mile Island Unit 2.
The Beyea Report, in this section, demonstrates a naive approach to a very complex problem. This section, prepared by one private consultant and a part-time graduate student provides "too little, too late." Not only was there an interim NRC NUREG-0683 Supplement No. I draft report PEIS dealing with the occupational radiation dose available in February 1984 before the release of the August 15, 1984 Beyea Report, but the Final Report, NUREG-9683, Supplement No.1, PEIS is available at the beginning of October 1984. While the Beyea I
Report recognized that this made its Appendix F outdated and irrelevant it nevertheless published it as part of its own report.
The response of this review to the Draft Report, and hence to the Beyea Report is included as Section 4.0 (supra), and is included (on pp. A26-A29) of NRC's NUREG-9683, Supplement No. 1, Final Report.
~
Fabrikant 50 -
CONCLUSIONS' It would b,e redundant and' time-consuming to rev'iew and criticize in detail.the 1984 Beyea Report 'section-by-section.
Tht.c is not the purpose of this review. The illustrations outlined-are sufficient to
~
respond to my introductory remarks concerning an assessment of the credibility, validity, and-the degree of certainty associated with the findings and conclusions of the 1984 Beyea Report.
- 1. Credibility. The evidence is that the Beyea Report as a whole does nnt portray.to the scientific community a consistent, believeable picture of the dosimetry of the accident and of our scientific and medical, understanding of radiogenic cancer. The various assumptions required do not seem to fit together and relate easily to plausible mechanisms of the radiation releases, the radiation doses, or of radiogenic cancer in exposed human populations.
- 2. Validity. 'The analysis of the radiation dosimetry in the-Beyea Report does not conform with recorded data where it is possible to observe them. The cancer risk estimates in the Beyea Report do not conform with observed risks of radiogenic and nonradiogenic cancer where it is possible to observe them.
In the Beyea Report, the properties, 1
of the methods used clearly violate fundamental principles and empirical observations.
- 3. Degree of Certainty. The doses estimated and the cancer risk estimatesNWBeyea Report are uncertain at best and unreliable and unsubstantiated assertions at worst.
They are biased in the most conservative direction only---to high doses and high mortality estimates---
without considering the preponderance of the scientific evidence. By compounding these uncertainties, preliminary asserttons, and examples
.Fabrikant 51 4
i of' data 'that Beyea admits to be uncorroborated and unsubstantiated,
-the Beyea' Report attempts: to conclude that th'e dosimetry has been.
inadequate and analys s misguided and incorrect, and that the implications
- for. delayed health effects would be much greater numbers of cancers-in the general population. On this basis, by simply declaring that much research must be done to reduce these uncertainties, Beyea proposes a series of research projects, most of which appear irrelevant,.
unnecessary, and frequently trivial.
In areas-where research has been needed, e.g., scurce krm investigations and the dosimetry of.the cleanup of TMI-2, much has already been done by competent scientists, has been published, and has become available, This makes the 1984 Beyea Report obsolete before the meetings planned to discuss it can take i
place, thereby invalidating it and discrediting its findings and conclusions.
t 1
l Jacob I. Fabrikant, M.D., Ph.D.
[
November 1, 1984 l-c I
1
-+
-c-
-r-,
a
,--n-n
Fabrikant 52 Jacob-I. Fabrikant, B.Sc. (McGill University), M.D., C.M. (McGill University), Ph.D. (University of London)-is presently: Professor of Radiology, Department of Radiology, University of California School of Medicine at San Francisco, and Department of Biophysics and Medical Physics, University of California at Berkeley; Member, the Board of Radiation Effects Research, National Academy of Sciences, National Research Council; Member, International Commission on
-Radiological Protection; Member, Safety Advisory Board, Three Mile Island-2; and is formerly Member, BEIR-I, BEIR-II and BEIR-III Committees, National Academy of Sciences, National Research Council; Director, Public Health and Safety, President's Commission-on the Accident at Three Mile Island.
He is certified in diagnostic radiology, therapeutic radiology and nuclear medicine (The Johns Hopkins Hospital and University).
1
,s e
e
1 h..hk% D 5 g$$ ^7hu ^
s N.kk.g 55.,h2 " # 'W].jk' :.1,W^.,..
~
!r
- f. i c
xm.>.usW - IM. eI?na~:*:*r, - -'c s-
- tu aM m e. _;._ w w - m a =a x.=u.-
- ' n n-
.c c w
N.
l APPENDIX A REPORT OF THE
~
PUBLIC HEALTH AND SAFETY TASK FORCE D,
p
.a
V-s.
r.
ON
/
HEALTH PHYSICS AND DOSIMETRY i.
?.
s i
BY
[
j c
HEALTH PHYSICS AND DOSIMETRY TASK GROUP g
9 h,_ '
John A. Auxier, Thomas F. Gesell (Task Group Leader)
School of Public Health f
Oak Ridge National Laboratory University of Texas
{
Oak Ridge, Tennessee
~f Houston, Texas e,.
Carol D. Berger Oak Ridge National Laboratory Alun R. Jones i*.
Oak Ridge, Tennessee Atomic Energy of Canada, Ltd.
b"f Chalk River, Ontario i
Charles M. Eisenhauer-Ibi.-
National Bureau of Standards Mary Ellen Masterson M2_
Gaithersburg,, Maryland Memorial Sloan-Kettering 7('
Cancer Center bh.fy New York, New York
~34W
$Y3 Tet s
- M
+
,i 2
9 u
October 1979 ri
..? V 'D '
D'
~~
Washington, D.C.
- tD49 fj 35
,' [ E N S
/
t
- ' L
f ~ 'i W m_. M sp, m, w, p.w y,, Cf'OM1C. M c s y g'l w g [c g i k. &.;.-
d&
p-E@wr: > YAf%
r..
n.~
t E M % 3.s u._L A -
%Ch2 " ' ~ '
" QEET
,r I.
SU:ClARY Th= primary task of this group was to determine the radiation doses that the worker population and the general public within a 50-mile rs'ius of Three Mile Island (TMI) received as a result of the incident th.: began on March 28, 1979. Estimations were made for dose to the shale body, lung, thyroid, skin, and extremities; details and calcula-g,j tic 31 techniques for the estimations are included in the body and 4,
appendices of this report.
I,.
L-The whole-body dose to the population was estimated through thermo-
[
luminescent dosimeter (TLD) measurements and through the use of computer
{
modeling of radioactive releases from the plant as they dispersed in the environment. Two different figures for the most likely collective
~
population dose within_a 50-mile radius of the plant between the dates of March 28, 1979, and April 15, 1979, were obtained. These numbers are
+
2,800 person-rems (by TLD measurenant) and 500 person-rems (by computer modeling).
Insufficient time has elapsed to analyze the possible areas of dif ference between these two techniques, but the task group has not ellainated either number as incorrect. For this report and for the use of other task groups, the stated current best value of collective dose s
is the more conservative one -- 2,800 person-rems. The fact that the N
cost probable collective dose lies below 2,800 person-rems cannot be i
ruled out.
f.
t This collective dose of 2,800 person-rems is applicable to those
{_
yho remained outdoors during the first few days of the accident.
There 7
ts some motection afforded by staying inside, as most people did, and i
th:refore the actual dose, incorporating a shelter factor, is estimated
[
to be 2,000 person-rcms.
1.
I s
The collective dose to TMI plant personnel from the day of the accident to the end of June 1979 is approximately 1,000 person-rems based on analysis of personnel dosimeter data. The maximum whole-body 7
d2se received by an individual was 4.2 rems.
f.
Based on the above and additional dose calculations from internal t,
deposition of radionucludes (determined by environmental and effluent
+
sa:Pling), average exposure levels to various organs and the whole body A
are summarized in Table 1 and in the body of this report.
Discussions Cf calculational, analytical, and other details are included in the b-various appendices.
3.:
1C The health physics and monitoring program was reviewed extensively.
%D might be expected, it has both important strengths and weaknesses.
2 task group found that considerable work in this area had been done by contractors, that the overall monitoring program was aimed at docu-
- tating routine releases as opposed to those due to accidents, and that
?
"Irmal maintenance of instruments and housekeeping were beh v the standards for a good health physics program.
J Q.
Ok 9m
'fz.q
.hiA e^ M - $*,
4 39
~t y-=.
M.c'27.
=? e.,
- m...
v,r23:y*is per,mrL.
+:
~ r. '
.. r..
Fabrika t.55; n.
-.. - ~
_... 3 :Cy.;. : =
w,c:s ~
r
.2
,s.c n~,_k W.w f9n.;gn'*w.4.,.%3 q~Tm=
- 3. x :
~
1 a-y~
--:3-
.:gn
- w r,-,; *:+ r-2 w =~
a:, ;- -
' APPENDIX B REPORT OF TIF.
PUBLIC HEALTH AND SAFETY TASK FORCE ON.
4' PUBLIC HEALTH AND SAFETY SUMIARY i
BY s.,
3....
gt.
Jacob I. Fabrikant
,y?
y?b.
- .+ ~
Head Public Health and Safety Task Force F.4 i,x
.l Pr -
+. q 1 t12 YS 4
- 4 -
+,;
~ %:sw Yk
,49..
s.
~
g p, -
. ax.
w., ~. -
.'e m '..
.myr -
October 1979 K'
JP f
Washington, D.C.
F. 4.., )
, --W&f Y(!av i
a ns.3
.6,N
, ei-A X
- +,
%',e2,+. i' e 1.
t s
9
Fabrikant 56 SUMMAP.Y OF THE HEALTH PHYSICS AND DOSIMETRY TASK GROUP REPORT l _
INTRODUCTION i
The general objectives of the Health Physics and Dosimetry Task Group inc.luded:
(1) to determine the radiation dose to the people living within the area of 50 miles around the Three Mile Island Nuclear q.p g Station during the period of March 28 to April 15, 1979; (2) to deter-t:Ei mine the radiation dose to the workers at the nuclear power plant during W[
the period of March 28 to June 30, 1979 -- the cutoff date necessitated by the deadline of the Commission's report; and (3) to evaluate federal, state, and utility company programs concerned with the protection of a[
j human populations and their environment from the possible hazards of ionizing radiation, and the efficacy of these radiation protection programs during the nuclear accident at TMI.
The task force identified the important events requiring analysis for the measurement of the radioactivity released into the environment,
-(%g for the assessment of the radiation doses to the public and to the workers, and the response of federal, state, and the utility company 4y programs for radiation protection. Among these are:
the identification m
of initial damage to the nuclear fuel; the release of radioactivity into Q) the atmosphere; the declaration of the site emergency and notification M
of the Pennsylvania State Bureau of Radiological Health; the notification
-t of the national radiological assistance program to draw on extensive i
resources to provide assistance during the emergency; the radiological indications of the uncontrolled escape of large amounts of radioactivity into the containment building; the declaration of the general emergency because of high radiation levels; the earliest releases of radioactivity into the environment resulting in raised levels of radiation in the areas where the general public lived; and the identification of the radioactive noble gases and iodine in the radiation releases.
RADIATION DOSE TO THE GENERAL POPULATION Normal Radiation Exposure j
Radioactivity occurs naturally in the environment and is constantly
}
being created in nature. Humans receive radiation exposure from this l
natural radioactivity, from cosmic rays from outer space, from the earth's crust, and also from those various human activities involving radiation and unrelated to nuclear power. Natural radioactivity occurs everywhere -- in air, in water, in soil, in foods, and in cur own bodies -- and is called " background" radiation. The radioactive elements (or radioisotopes) found in our external and internal environ-ment are extremely varied in the energies of their different radiations, and in the time of their decay -- that is, to undergo rpentaneous disinte-gration with the emission of radioactive particles or rays. The radiation dose absorbed in the cells and tissues of the body, whether from natural or manmade radiation, is frequently measured in rems; the rem is one form of physical radiation unit which takes into account the amount of j
radiant energy deposited In the body tissues and the type of radiation -
1
\\
6 i
- h.w G.;G igfq4 g v'E ~ F ~ # 'CS, m.W.c.4W p.W_. T V2 9 T E_
% p w.r, -
m e.,h " $. y 4,k W $
h-.*
. ?y W % Y,,--
. : -- ~
.-i m-3 q.
,,4
~-4 4
m.
.r.
b*
~'
- 4? x x
' d% k u_
__ a 2_ _ -
v--
~
, w.a?w$~N :^.
?
~
_ - m %'f s r t ;c. -;mm.mu m m: ~ --- mz ~ + ~ e_
&.. y,.n m
alpha, beta, or gamma radiation, or neutrons.
When the dose is measured over a time period, say rems per hour, this is called doserate. When the radiation dose level
- is low, as in the case of natural background,
' the radiation dose unit frequently used is the millirem (mrem), or one-thousandth of a rem.
Some familiarity with these quantities and radiation units is y
necessary for understanding the significance of normal or accidental 4
radioactive releases to the environment f rom nuclear power plants. Man is constantly exposed to naturally occurring radiation; each year, the
'y.-
cverage American is exposed to about 100-200 millirems of natural back-ground radiation depending on where that person lives. The variation
~
depends primarily on altitude and on the long-lived radionuclides in the carth's crust.
In Harrisburg, Pa., the average annual whole-body dose to the individual due to natural background radiation is estimated to be 116 millirems.
In general in Harrisburg, about 45 millirems per year of this whole-body dose come from cosmic radiation and 45 millirems per year from terrestial radiation. By comparison, each of these annual dese-rate values is about doubled in Denver, Colo., to about 75 milli-rems per year from cosmic radiation and 90 millirems per year terrestial radiation, respectively. The internal radiation annual dose-rate is relatively constant in all individuals (about 28 millirems per year) from naturally occurring radioisotopes in the body, primarily potassium-40.
About half of the radiation to which the general population is exposed annually comes from natural sources and the remainder from 3.
can-made sources. The average annual background radiation exposure to 4;.?
an individual is very low; comparisons between levels in Harrisburg, Pa.
1.5.f (average), Denver, Colo. (high), Las Vegas, Nev. (low), and the overall W
range in the United States, in millirems per year (mrem /yr), are given
.hf5.
in the following table:
M-Fj Er,Wk Harrisburg,
- Denver, Las Vegas,
- Range,
- i%
Radiation Source Pa.
Colo.
Nev.
U.S.
g Cosmic Radiation 42.0 74.9 49.6 40-160 0 _ %.
i;&-
Terrestial Radiation 45.6 89.7 19.9 0-120 q! tid!Y Internal Padiation 28.0 28.0 28.0 28 C.: :_,,
n..y p-Total (mrem /yr) 116 193 98 70-310 MM j
~
". :.a..
w.a. n. --
Tha remainder of man's radiation exposure, due to manmade radiation, is W y 45 i$
primarily (an additional 40 percent) due to medical and dental x-rays.
t ?bp6 Nuclear weapons testing and fallout, technologically enhanced natural 7 g+ g radiation (e.g., uranium tailings), consumer products (e.g., television 4
sats), and nuclear energy plants provide only a very small fraction y
~
(cbout 0.15 percent) of the total amount. The 1978 estimates of the
.w m annual collective dose (that is, the average yearly dose summed up for 1
a th2 entire population) of radiation exposures to the U.S. population --
?
.9 somewhat more than 200 million Americans -- based on data summarized by m%
gg,ST.
th2 Interagency Task Force on Ionizing Radiation (1979) -- are listed y_
gg below:
2 m.
7 xW cf
Fabrikan,t 58 Annual Collectiva Dosa Radictica S::ure?
(Parsen-reas par Year)
Natural background (e.g., cosmic and terrestial radiation) 20 million Medical and dental x-rays (e.g.,
F7 x-ray diagnosis) 17 million g
r;_
Nuclear weapons (e.g., manufacture
.}7 -
and testing) about 1.3 million
,(
Technology-enhanced (e.g., uranium qq tailings) 1 million 5i '
- T.7 Nuclear energy (e.g., nuclear power 4.,.{
plants) 0.06 million
=
",;7 7
.+
Consumer products (e.g., television
~E.
sets) 0.006 million Total about 39 million h;
Under normal conditions, the 2,163,000 persons living in the 50-mile area surrounding TMI would receive an annual collective dose of about g,,i 440,000 person-rems; about 240,000 person-rems would come from natural
~Fjf background radiation.
(In contrast, the collective dose to that population y
resulting from the radioactive releases during the TMI accident was
-E approximately 0.5 percent of the normal annual exposure rate, or about 1 A.-
percent of natural background radiation.)
[
Radiation Exposure During the TMI Accident
- 7 Nuclear radiation doses are measured with instruments or detectors
-l called thermoluminescent dosimeters (TIDs); TLD measurements formed the l
basis for estimating the total external gamma radiation doses (due almost exclusively to the radioactive noble gas xenon-133 and a few other short-lived radioactive gases in the radioactive cloud) to the population during the TMI accident.
The main TLD dosimetry instruments were located within a 15-mile dists. ace of the plant.
Individual doses within a few miles of the nuclear plant were relatively low; some 260 people living mostly on the east bank of the Susquehanna River possibly each received between 20 to 70 millirems. One person on a nearby island for 9-1/2 hours during the initial days of the accident received about 50 millircas.
All other persons living outside a one-mile radius and within 10 miles from the plant could have received an average dose of i
less than 20 millirems.
Almost all recorded excess exposure above l
background Icvels occurred within a 10-mile radius. There were no recordable radiation levels above natural background at a distance greater than 10 miles from the nuclear plant at any time during the accident.
l j
The total release of radioactivity into the atmosphere frcm the j
damaged nuclear power plant during the period of March 28 to April 15, 11 s-tm
+
=
db
&LQ:sy,yR&Qlik.b[
6.QLi;M Qf.L Wp, M n
u m.
m:
m
=&-
$$ $$h?-M *$YNh.Uf hk?$$$ L'.~bMYh;,M j$f 4$hMfM:ht
$d
@hMA%NQijdVMe; pg.
f,p;q,
%?5n...
1979, was calculated to be about 2.4 million curies,1/ primarily con-sisting of radioactive noble gases.2/ Approximately 10-15 curies of radioactive iodine were released into the environment.
This total release of radioactivity, known as the source term, was one way to determine the radiation doses to the entire population (collective dose) r and to t'.e individual in the population (average dose), taking into the collective dose was by use of the TLD radiation dose measurements.
. p",
cccount meteorological weather conditions and population distribution demographic data at the time of the accident. Another way to determine The collective dose to the population is a measure of the potential 3
health impact resulting from the total radiation dose received by the entire population; for the TMI site, a 50-mile radius and approximately 2,163,000 persons were included in the calculation. Since this value is obtained by summing the estimated radiation doses (measured in rems) y received by each person in the affected area, the collective dose unit is the person-rem. The collective dose above normal background levels y
to all persons within a 50-mile radius of TMI, based on the TLD radiation dosimetry, was estimated to be about 2,800 person-rems outdoors and unshielded.
Since most people spent most of their time indoors and
(
partially shielded by buildings, and assuming that the radiation dose i~
indoors was about three quarters of that outdoors, a more accurate y_
collective dose to this exposed population is estimated to be about t
2,000 person-rems above normal background levels.3/ The average dose to P.
any individual in the population living within 50 miles of the nuclear y,
reactor, therefore, is estimated to be about one millirem. The average
- ~9 dose to an individual living within 10 miles of the plant is estimated b
to be about 6.5 millirems.
~i; MW There are a number of ways to evaluate the magnitude of the radiation I Q releases and the exposures to the general population.
If the maximum i1 @; $,
dose to any member of the public exposed within just a few miles of the reactor site was no more than 70 millirems, this may be considered to be g*
t equivalent to about one-half of the normal exposure the average American 1 x%'
receives from natural background radiation each year; probably no more than 250 persons out of the entire population could have received this 6
dose, and most of them received less.
Another way of considering it is
/).
that this dose is equivalent to the difference between annual background p
radiation exposure in Harrisburg and Denver, Colo. An average dose of
,d 6.5 millirems is about 5 percent of the exposure from natural background p,
h.
radiation annually in Harrisburg, and equivalent to the difference of w {t:-[;gt living 2 weeks in Denver.
c; L'C.G The radioactivity released during the accident entered the air,
'. "WF w:ter, soil, and food, and could ultimately have become incorporated f
.M into the human body by breathing, swallowing, and absorbing it through i
Wh the skin. This could result in an internal radiation dose to the tissues ef the body. During the TMI accident, the identity and concentrations Mi Nd cf radionuclides present in the environment were determined by the utility company and by the various federal agencies.
Sampling analyses included milk, air, water, fruit and vegetable produce, soil, vegetation, 4
fish, river sediment, and silt. Any increase in internal radiation dose due to radioactivity released during the accident came primarily from am 9
yM 5M '
r
Fabrikant 60 ridierctiv2 x:n:n-133, iodin;-131, cnd casium-137. Extr=21y smll incrmsts in tha radionuclida c:ncantreticas of iodina-131 wara rcpartcd in cows' and goats' milk, and in water and air; of cesium-137 in fish, and of xenon-133 and krypton-85 in air.
The highest doses due to inges-y tion and inhalation of iodine-131 would occur in the thyroid gland, 9
since iodine concentrates in that gland. However, wholebody scanning of p
a large. number of the general public living near TMI during the accident detected no radioactive iodine in this population; no radioisotopes
[u related to the TMI accident were found.
sw
~
The internal radiation dose due to ingestion of cesium-137 was g
negligible. The internal dose from inhalation of xenon-133 and krypton-85, gm primarily due to radiation exposure of the lung tissue, was only a small 74 {h, fraction of that of the external dose. Overall, the internal doses due MNU[
to the radioisotopes released at TMI were negligible, and would have Q
been only 'a minute fraction of the average annual dose received due to naturally occurring, internally deposited radioisotopes in the body.
Y$
RADIATION DOSES TO THE WORKERS AT THREE MII.E ISI.AND
. ~ SN.-
w 7.W The radiation exposure to the nuclear plant workers during the M
accident at TMI came primarily from external radiation and some from 3;
internal radioactivity. Thermoluminescent dosimeters in badges were h
used to measure the external gamma and beta radiation doses.
Before the M
accident, the collective dose to about 1,000 workers at THI under normal Q
operating conditions varied from about 20-150 person-rems each month.
I--
About 5,000 workers were on-site at some time during the March 28-June 30, 1979, interval; the majority received no recordable radiation exposure. Most of these additional workers were brought to the Three
' 'J Mile Island plant during the accident and did not receive measurable exposures. About 1,000 workers received meacurable doses of radia-y.
tion -- that is, greater than 50 millirems during the accident. The 7
I collective dose for these 1,000 workers from the time of the accident
'I on March 28, 1919, through June 30, 1979, was about 1,000 person-rems.
I
{
The average whole-body dose to these 1,000 workers was about one rem during this 3-month period. Two hundred and seventy-nine workers received more than 0.5 rem, but less than 3 rems of whole-body gamma radiation exposure; three workers received about 4 reras (on March 28 or 29); and none received more than 5 rems, the annual limit permitted.
In l
addition to the three workers who received whole-body overexposures
{
during the accident - greater than a 3-rem whole-body dose per quar-ter -- two workers received overexposures to their hands of about 50 and 150 rems, respectively. The worker who received 150 rems to his fingers also received a vbole-body dose of about 4 rems. No overexposures were 7
recorded due to beta radiation.
Uhole-body counting of plant personnel 3
was inaccurate, and the procedures and the collective records provided
- 4. I little reliable information on internal body doses of the workers.
A
.I T few showed measutable levels of radioactive iodine-131 and cesium-137; it is probable that the radiation recorded by whole-body counting other than natural background was due to external contamination.
1 In spite of the high gamma radiation exposure rates of up to 1,000 R/hr 4/ neasured in the auxiliary building on March 28, the radiation i
f 10 e
.-s.'.nw. _
- ...~.x Q
a
- ~
- %l'
.lbn%
c ~-2.a l l V % L ', ;; ' * ;y,d::=uu%M R
.w; l '-
y L.
I $.$.m$:s~;w. '~ l.2.. ?, ' '.,k ~R ' % ~
- f.,i;i
,. _w&#.Q,.s l 581ii U d S. 4, m tkint 4,..4 %,,
-9
~
y*.
a v
~.-
r,
+....
I$ti^i hM l -tal. E;82R*~
.ar r w, "
s'
- " * M i*
' " ' ' - * * ~ }p 3
\\,i doses to the workers were quite low.
However, the collective dose to the workers of about 1,000 person-rens will increase as the decontamin-ation and recovery at the TMI plant proceeds. It is difficult to predict the eventual total collective dose, since that will depend on methods of decontaraination and recovery of the ~ containment building and the reactor vessel.
u i
D l
p l
f k,
r l
4.-
x'.
w
(~
i*
k.
t t.
L EC
. i' %.
v r
C..
, : -~
>.4 7
- s V.
- s ;Q.
W
'?'
?
- - r y- -
$Y
.e iE M.':4*
-1
,i{,.c '.
5 6
'% 3
'33555 rre m E-.V$f 11
.t-3.,...
/b 4
,,,A,
Fabrikant 62
- i 1
~
SUMMARY
OF THE RADIATION HEALTH EFFECTS TASK GROUP REPORT y
INTRODUCTION The highly publicized events during the early days of the accident included:
(1) the various releases of radioactive materials into the atmosphere and into the Susquehanna River; (2) the accumulation of hydrogen generated in the reactor pressure vessel; and (3) the risk of
- m major releases of large amounts of radioactive debris from the damaged nuclear core. These threatened the health and safety of the public and
_@6 the workers, and led to concern about possible acute and delayed health jy effects of exposure to ionizing radiation, s
]Y' 7 Some release of low levels of radioactivity normally occurs into n
the environment during the routice operation of a nuclear reactor power
~D plant.
The accident at TMI set off a series of events that raised the M
threat of risks of much higher levels of radiation exposure of the 3,
public due to uncontrolled releases of radioactivity.
Low-level ionizing
?
radiations (e.g., radiation doses of a few rems or less) are thought to
(
be able to contribute to three kinds of health effects. First, some of Q
the cells injured by radiation may occasionally transform into potential d--
cancer cells, and after a period of time there may be an increased risk
~E of cancer developing in the exposed individual. This health effect is called " carcinogenesis." Second, if the embryo or fetus is exposed 1.y during pregnancy, sufficient radiation damage of developing cells and j
1
. pp tissues may lead to developmental abnormalities in the newborn.
This 4
I*'
health effect is called " teratogenesis." Third, if radiation injures reproductive cells of the testis or ovary, the hereditary structure of the cells can be altered, and some of the injury can be expressed in the descendants of the exposed individual.
This health effect is called x.
" mutagenesis" or " genetic effect." There are other health effects of ionizing radiations, but these three important health effects -- car-cinogenic, teratogenic, and genetic -- stand out because it is possible that lov levels of radiation may increase the risk of these effects.
~
Much scientific information on these effects has been gained from animal experiments, and for carcinogenesis, from epidemiological studies of exposed human populations.
Scientists generally believe or assume i
that any cxposure to radiation carries some risk of carcinogenesis, or -- if reproductive cells are irradiated -- some risk of genetic effect, and that as the dose of radiation increases above low levels, the risk of these health effects increases in exposed human populations.
These latter observations have led to public confusion and fear about the possible health effects of low-level ionizing radiation from the radioactive releases during the nuclear accident at Three Mile Island.
Radiation scientists are generally in close agreement on the following broad and substantive issues of such health effects:
w i
Cancer arising in the various organs and tissues of the body o
is the principal late effect in individuals exposed to low or internediate levels of radiation.
The different organs and 12 4
r.$. *
- s ( %
..... ng;g:.p _.__
.Wg A
3_. g y g.g...
\\, m k V: ~
. F,.*
.L*'I :T, f
L.
3
,~
A.cW%%<rf.;[Jsw "P 7 ~
L O.
.~
C:
.qn w. ~ M~y
.; 2
.p g*p-
- 2
. h 4
4
2
_m abEjg~. 63 dp y
,3.,%- %.,. -.~
w ir;fynf@ %.
..a..
y y%..p;,m2syrgQ-iq.
_-t s
% ng. hpq p%
- g 7-g - ;. 4.. >
.c
~
a.
6
-~ n M g"h'w
.g b'N#4 g d '7.'rt 'f W /4. M %tM,hh % W N y y.
tis;u n vary in relative susceptibility to radiation-induced cv.c>c; the female br. east, the thyroid gland (especially in youcq children and females), and the blood-forming organs (in regard to leukemia) seem to be more susceptible than some other organs.
o The deleterious effects on growth and development of the j
cmbryo and fetus are related to the stage at which the radia-tion exposure occurs. A threshold level of radiation dose may
{
exist below which gross clinically evident developmental abnormalities will not be observed. However, these levels r
would vary greatly depending on the particular developmental p
abnormlity.
i The paucity of data from exposed human populations has made it o
g necessary to estimate the risks of genetically related ill-t health based mainly on laboratory mouse experiments.
Knowledge of funda.aental mechanisms of radiation injury at the genetic level permits greater assurance for relating scientific informa-tion from laboratory experiments to man, s
Ho ever, there is still very much scientists do not know about the
.tential health hazards of low-level radiation:
E,-
o We do not know what the radiation health effects, if any, are j
at dose rates as low as a few hundred millirems per year --
i higher than natural background radiation.
It is probable that i
if health effects do occur, they will be impossible to dis-
[
tinguish from similar effects owing to nonradiation related i o environmental or other factors.
[
{y ~'
o The epidemiological data on exposed human populations are uncertain regarding the dose-response relationships for various i p' radiation-induced cancers. Since this is especially the case I
for low radiation levels, where no unequivocal data exist, it g
i has been necessary to estimate human cancer risk at low radiation 7
levels primarily from observations at relatively high radiation levels on the basis of various assumptions. However, it is i.
not known whether the carcinogenic effectiveness observed at j
high radiation dose levels applies also at low levels.
s; S
- s...
o There are no reliable methods of estimating the repair of vf; -
injured cells and tissues of the body exposed to low radiation doses, nor is it possible to identify persons who may be
,Q1 Particularly susceptible to radiation injury (as, for example, y % g{
a genetically determined increase or decrease susceptibility
M to radiation injury).
g y'g k
ups -
All epidemiological surveys of irradiated human populations k g$
0 exposed in the past are incomplete with respect to ascer-2
' N Fj!
tainment of cancer incidence in terms of providing a basis for Q
-- 3 analysis and conclusions, since there is only limited infor-
.4 4 /
mation on the radiation doses in some of these studies, and T
limited and incomplete data on cancer incidence and/or variable NNA e
followup data, p
J A -Gi T ^ W@
13 f
Fabrikant 64 L
F o
W2 d9 n:t kn:w tha role of compating cnvir:nmentc1 cnd eth:r host factors -- biological, chemical, or physical factors --
existing at the time of exposure, or following exposure, which may affect and influence the carcinogenic, teratogenic, or genetic health effects of low-level radiation.
M RADIATION-INDUCED CANCER 2
y There are valid practical reasons for assuming proportionality in
_. m dose-effect relationships for the estimation of radiation-induced cancer 6$1 risk in the general population exposed in the vicinity of THI.
It
'lh should be recognized, however, that the assumption that the risk for pf C low-level gamma radiation (the predominant radiation exposure at TMI),
UL2 is proportional to observed risk at high levels may overestimate the th cancer risk; the actual risk would be much less.5f It is estimated that h<
the number of excess fatal cancers, if any, that might occur over the
$M remaining lifetime of the 2 million persons living within 50 miles of
$1 the nuclear power plant and exposed to an average whole-body dose of M
about one millirem is much less than one; a similar number is estimated
' JD for excess nonfatal cancers.
These numbers are estimated to be only a
~; -
very small fraction of the potential lifetime risk of radiation-induced i ~;
cancer which may arise in this population from natural background radia-l l
tion exposure, i
s; The estimated number of cancer cases from all causes normally occurring in this population of about 2 million people over its remain-ing lifetiine is 541,000 (325,000 fatal cancers and 216,000 nonfatal cancers).
The estimated excess number of fatal and nonfatal cancers associated with the increase in radiation exposure due to the accident is extremely low, and could be zero; it would not be possible to detect or to distinguish this excess either in the population or in the indi-vidual.
The number of excess cancers, if any, would be so small that it would not be possible to detect such an increase statistically in the I
more than half a million cancers that would occur in the population even if the TMI accident had not happened.
Furthermore, cancers caused by radiation are no different from any other cancers resulting from other causes; therefore, a particular cancer cannot be distinguiched as having been caused by radiation.
The lifetime cancer risk in individaals exposed to maximum doses of approxir.:ately 50 creras is about one or less chance in 100,000 for fatal and a like risk for nonfatal cancer, i.e., a total cancer risk of about two in 100,000, with zero not excluded.
The additional radiation-induced risk of skin, lung, or thyroid gland cancer j
due to beta radiation and internally deposited radioisotopes is estimated to be extremely small, and may be regarded as encompassed within the cancer risk values expressed above for whole-body radiation exposure.
We conclude, therefore, since the total amount of radioactivity released during the accident at TMI was so small, and the total popula-tion exposed so limited, that there may be no additional detectable cancers resulting from the radiation.
In other words, if there are any i
additional cancer cases, the number will be so small that it will not be possible to derenstrate this excess or to distinguish these cases ar.ong
.I the 541,000 persons (of the 2 million population) living within a 50-mile radius of TMI, who would for other reasons develop concer during the course of their lifetimes.
14
' M'r
~ ~ N W '"N
. ce =
N-
% 'Q.
~p e r:y "'
s.. m
_w-
'7
- 7,
. ], %.l.yg * ~
. e L. -
u
% " % D d;[ h *
. N ' S M
M-O N
.p s-
f {$
5jjkhA~[Qqv. 3yL*;5jaWAnL 65_3 Nyu.
t C5CEPTOFESTIMATIONOFRISKOFRADIATION-INDUCEDCAECER In all these calcutations of the risk of radiation-inheed cancer, several dif ferent methods have been applied for estimating the number of While cancer cases that may be caused by the radioactivity released.
different methods may lead to different estimates, all of them arrive at a very sml1 number -- less than one and possibly zero -- in 2 million For example, consider an estimate of "0.7 additional cancer people.
What does this mean?
deaths due to the released radioactivity."
l The number 0.7 is an estimate of an average, which is a mathematical "The average concept such as the one that appears in the statement:In the case of TMI, what it really American family has 2.3 children."
meant is that each of some 2 million individuals have a very small additional chance of dying of cancer, and when all_ of these very small In such a probabilities are added up, they add up to the number 0.7. situation a m af ter a French mathesatician) applies.
If the estimated average is 0.7, There is a roughly 50 then the actual probabilities work out as follows:
percent chance that there will be no additional cancer deaths, a 35 percent chance that one individual will die of cancer, a 12 percentcha that there will not be as many as five cancer deaths.
Similar probabilities can be calculated for the other estimates.
It is entirely possible All of them have in common the following fact:
that not a single extra cancer death will result from the radioactivity And for all the released during the accident at Three Mile Island.
estimates, it is practically certain that the additional number of cancer deaths will be less than 10.
We know from statistics on cancer deaths that in a population of this size, eventually some 325,000 people will die of cancer, for reasons Again, this having nothing to do with the nuclear power plant accident.
number is only an estimate, and the actual figure could be as much as Therefore, there is no conceivable statis-1,000 higher or 1,000 lower.
tical method known by which fewer than 10 additional deaths could evet be detected. A cancer caused by nuclear radiation is no different than a cancer from other causes. We conclude, therefore, that there may be no additional deaths due to this radiation, or if there are, they will be so few that it will never be possible to determine that even a single death occurred as a consequence of the accident at TMI.
GENETICALLY RELATED ILL-lEALTH There is persuasive scientific evidence which suggests that if an
[
average human population were exposed to one rem (1,000 millirems) of F>
irradiation during their reproductive life span when they can produce
_ %p children, we might expect to see about 5 to 75 cases of additional genetically related diseases (such as mental retardation or diabetes) in one million children born to the irradiated parents. Genetically related a
H ill-health is extremely common in humans under normal conditions; about d
10 percent of all live births are affected. Therefore, the increase due
- EU
' ^%"
.M 7
15 mC,
~,
b.,
=
Fabrikant 66
)b tafi,000 tillircs cf rediatica wruld r:prIs;nt c vary small numb:r cf cases of genetically related ill-health in addition to the 107,000 cases (an increase of only about 1 one-thousandth of one percent) of genetic disorders expected to develop in that newborn population.
t Since there are no direct data from human epidemiological studies, the basis for this estimate comes mainly from laboratory experiments in which the reproductive cells of the testes and ovaries of mice are irradiated. That such experiments in inice have applicability to man is 3
suggested by the following:
1 5-l'l R
1.
The hereditary material of life, or genetic material, of all Dy organisms is chemically similar.
-1 sM 2.
The reproductive cells of the testes and ovaries of mice are
[
similar to those in humans and are expected to be pertinent j
ff..
for assessment of genetic ill-health due to irradiation, ji I1 3.
Radiation, as well as a great many other toxic agents, can y
produce similar kinds of changes in the hereditary material in
.k both the mouse and humans, both within the genes and chromo-
~
These changes, or mutations, in the genes of the
/
somes.
parents can, under certain circumstances, be transmitted to the offspring and thus result in inherited or genetically related diseases -- abnormal anatomical, physiological, or f
behavioral health conditions.
1 3'
4.
Many of the inherited diseases appear to have analogues in inherited diseases in mice.
Genetic mutations resulting in genetically related ill-health probably do not only come from exposures to radiation or chemicals.
Most of the newly arising genetic mutations in humans result from unkncwn or as yet unidentified events, called " spontaneous mutations," wit hin the reproductive cells that can lead to " mistakes" in genes when they s
are being formed and reproduced for newly formed reproductive cells.
Natural background radiation in our envirorunent appears to account for only a very small fraction of mutations resulting in genetic disease.
k'e know very little about the precise contribution of chemicals in our i
environment to genetic ill-health. Radiation and other toxic agents will j
increase the probability of a genetic mutation occurring, but they will not produce any dif ferent kinds of genetic diseases than occur from other causes of mutations.
During the accident at Three Mile Island, the collective dose to the reproductive cells of the testes and the ovaries of the 2 millica persons living within 50 miles of the plant was about 2,000 person-rems, with an average individual dose of one millirem.
In this population, assuming a 30-year generation time, we would expect about 3,000 cases of genetically related ill-health among the approximately 28,000 live children born each year; these are unrelated to the radiation from the unclear power plant accident. From an additional dose of one raillireo g
above natural background radiation, we would expect about 0.0001 to
~
about 0.002 additional radiation-induced cases of genetically related
'. 6
, y
=A J,. M #.
r.g,,
,4,.
-4 4 _
o
" mm a,.
.: " r~~my ~ m T=!
e
-~
m
~
- A. w m n.-
n n.- r 3
- g-
~** y* - ;..' '.,__
, {,,,
s r
4 l
- - w F..,;
-Q n.(9 a.y,. - n ~:
..%-.r.
. w. s-
.,s.
.p
~
q%,,e %ge =fN
. 7d W-Q >3..~ef. y%Nps,v Q27.$w,e;:931 @Mptkphp7. ;,gg.y
_ y _ ~ G y.. An- & m, s:
- s. u.,. - a.-
..a mhg&p.,.,y,3
_ ;;n.x. - mwg a
n..
ygg 5
3;_-
~:>
N
..1 ;.n1th.
This 0.002 case is an " average" nucber and is miniscule,
- . _ r. ting less than 1 in 10 million live births.
Furthernore, this
- r: re.; ult ultimately in a total of no more than about one additional
- ih o't genetically related ill-health in a million liveborn children brin; all generations in the future. This number of " additional cases" is so snalt that it can never be detected or distinguished, if it does occur, among the cases of genetically related ill-health in each genera-tion during all future human existence. We conclude, therefore, it is probible that there will be no detectable cases of genetically related ill-heilth resulting from the radiation exposure to the general popula-tien followin; the accident,at Three Mi_le Island.
- qElp?MFyTAL ABN0ptALITIES Approximately 2,160,000 people live within a 50-mile radius of ihr-e Mile Island; it is estimated that in this population, based on ei
- i1 statistics data, about 28,000 childmn will be born in 1979.
In this ne,. born population, about 300 children would normally be expected to ba born with developmental abnormalities in the absence of any added
- i M ition exposure as a result of the accident at TMI.
The estimated v.erige individual radiation dose to tiG fetus of pregnant women exposed daring the accident (perhaps only onehalf of the one millirem) was below an/ threshold dose level known to cause detectable casas of develop-
- -at.il abnormality in the human embryo or fetus, or in laboratory animal exp-riments.
In addition, the estimated dose may be too high, since
- niny pregnint women left the area in the vicinity of the nuclear plant.
And finally, if the maximum dose received by the workers were received by a pregnant woman working at the plant during the accident, the dose level to the fetus still would not exceed a threshold to cause any detectable developmental abnormality. We can conclude, therefore, that no case of developmental abnormality may be expected to occur in a new-born child as a result of radiation exposure of a pregnant woman from the acci-dent at Three Mile Island.
5 1~.
T
[v'
?.:v z;, 2
.A w
+;n V
2 h,.;
e
- *tr
..x 493'
~. :.
17 P
s APPEN[XC Fabrikant 68',
Coe.nents ti l~crril Eisenbud on "A Review of Dose Assessments at Three Milt Island and Recocnendations for Future Research,"
by Jan Beyea, dated August 15, 1984 On page 15, the statement is uade that "There is evidence in the literature that the original TLD's lef t significant angular gaps through which bursts of radioactivity might have passed entirely undetected or They don' t only partially detected." This is an uncritical statement.
document the evidence except by reference to the Thocas report
( AIF/IIESP-023) from which they have' taken their Figure 1 on pa6e 16.
The figure they present is for stability Class F (moderately stable) which was not typical of conditions that existed during most of the accident. This part of their argument requires critical review by some.
body more f aailiar with the post-accident ueteoroloby. I would think that Pickard, Lowe, and Garrick uould have the information right at their fingertips.
Para. 3.2, Doses from Radiciodine: The aucunt of iodine released is very important and has been thoroughly investicated.
The Beyea report trica to cast doubt on the validity of the estimates but it succccds only by innuendo and not witn hard facts.
The subject is discussed extensively in Appendix C, which starts out by saying there are three "najor puzzles associated with the bchavior of radiolodino at Three Pdle Island."
The first puzzle prcsonted is that 11 million curies of the cere's radiciodine inventory is unaccounted for.
But if even a cuall fraction of the radiciodir.o escaped, it would be easily detectable by a variety of ccans.
Only 20,000 curies escaped during the Uindscale accident in Octoocr 18, 1984
Fabrikant 69 s
C cm:nts cn Bcyca. Report !!. Eissnbud 1957, and except for a minor amount of cesium, no other radionuclide was deposited downwind of the Windscale reactor.
As a result of that.
accident, there were large downwind areas in which the canna ~ radiation levels due to I-131 deposition were in excess of 150 pr, which is about 20 tbses norual.
The milk in those areas contained radioiodine in con-centrations greater than 500,000 pCi/ liter.
In the post-accident gamma surveys around TMI, a 20% increase could have been casily detected.
Assuming the relationship between the source strength and deposition were comparable at Windscale and TIE (though only to a first approxination), a 20% increase in the gaama background at THI would have been attributable to a release of 200 curies.
If you think it uorthwhile, this approximation can be refined by taking microueteorological factors into consideration.
Beyea (page C-39) uses Windscale in a similar way and esticates that the Tt!I emission was actu-ally 4 Cil All things seen to point in the richt direction: I-131 in crass, human ceasurenents, etc. lead one to conclude the I-131 cc1 case was niniscule.
One would not expect to see an elevation in the cauaa background unlees. the caission uas core than 10 tinos hicher than ertiaated.
It is not " puzzling" that nest of the radiciodine inventcry is unaccounted for: it recained within the reactor building and has long since decayed.
Uo learned froa the Uindscale accident that when radiciodine is released in quantities significant to health, it can be readily detected not only by the increase in a=bient catua raciation, but also by high concentrations of radioiodine in grass sud cow' c uilk, as uell as bucan thyroids.
Octotcr '16,1964
Fabrikant 70,
i te-M. Eic2nbud Comments on Beyea Report.
1 Incidentally, one cow ue were monitoring in Pennsylvania af ter a Chinese weapons lest in 1980. delivered milk containin61000 pCi/ liter!
~
In the early 1960's, the milk from some of cajor eastern milksheds fre-
.~,
quently contained radiolocine in concentrations greater than 100 The methods for radioiodine pCi/ liter during periods of cany days.
detection are very sensitive and, when present, it is one of the easiest radionuclides to" find in the environment.
The lengthy discussion in Appendix C of the various pathways by which radiciodine cay have escaped i,s of no importance insof ar as public Had the I-131 escaped in significant quantities, health is concerned.
it would have been detected in the environment.
On pcde C-27, they propose a search for residual I-129 in the reac-One of the problems is that the reactor operated for such tor building.
In a short period of tiac that there was very little build-up of I-129 cethods would peruit I doubt that the uaupling and analytice1 any case, A basic proolca is that it uould be developing useful inforcation.
necessary to obtain a small nurbar by subtracting tuo larce am;bors (the I-129 estiuated to have been present originally, and the 129 esticateu This is alucys a risky pro-to be present at the tice of acasurement).
cedure where one or both of the large numbers are subject to uncer-tainty.
The cocond " puzzle" identified in the report is that airborne radioactivity, inferred froa uilk ceasurecents, is tuch higher than the It is propcsed aucunt inferred froa othcr environnental..catvecutnts.
Ui.ct's the point of all that there be research to reconcile the data.
to iuprove our I-131 models, and perhaps soue of this? Surely ue want October 18, 1954
L<
Fabrikant 71 s
y
- Crcu;nts on Bryc3 lleport th Eisenbud the intercation that could be obtaired will be useful from this point of view, althoubh J doubt it.
If the milk concentrations were lower than m
- expected, then people roccived lower doses.
They Lake reference to the 2
peak concentration o 900 pCi/u on April 15.
They then use the Windscale information and conclude that if the radiciodine at THI behaved the way the Windscale iodine behaved, the T!!I release would be 4 Ci, " number which is not wildly inconsistent with the officJak THI esti-mate of 15 C1." As a matter of fact, all things considered, it is excel-lent acrecuent and shows that the THI estinate' of 15 Ci was arrived at in a conservative canner.
The third " major puzzio" is that it is not clear uhat percentage of the radiciodine was organic.
I don' t see what difference this would make.
The iodine rc,tention system uight be less efficient for =cthy-lated I-131, but once it entered the environment it should behave the saac.
I!y recollection is that the various biological uptake facters are no different for crganic iodine than they are 'for inorcanic.
This,_hcu-cver, should be checked by so.2 cone.
The discursion of radio.odine then sces on to conuent on measure-uents that utro nado of voles.
Sone nay find this interesting radioecolog, but I don' t see how the inforcation enn affect the esti-nates of the dcsos roccived by people.
This is also true of neasure-cents unde in rabbits, Coats and sheep.
The review of the ccw's milk studies, beninning, on page C-51, is mainly concerned with the question of u,lacchcr the ccus absorbed radiciodine frca the air or frou crass. (I don' t think it uutters as long as the.uilk cencentrations assure that the dose to children's October 18,.1934
Fabrikant 72 Couu:nto B:ycc Raport H. Eisanbud thyroids, cycn with the assuuption of the highes*. radioiodine concentra-tions in milk, were less than about 10 crem.
Tnis is broucht out very well in the report of the Ad floc Population Dose Assessment group s9at-tist, 1979).
I have even more trouble understanding Beyea's reasoning concerning radiocesium.
Considering the f act that so little radiciodine was enit-ted, I don' t understand why.anyone would suspect that cesium-137 would be a problem.
The xenons have short half-lives and blow out of the area in a matter of hours af ter release.
I-131 has a 8-day half-lif e, so that measurements might be possible for many days af ter the accident.
Cesium-137, with its 30-year half-life, remains near the surf ace of soils for long periods of time and can be measured easily.
Most of the back round described by Beyea as being due to residual fallout from 6
nuclear weapons tests has been in the soil since 1962! On the other hand, as pointed out by Boyca, Cs-134 uith a 2.1-year half-life is asso-ciated with the cesium-137 and can be used to diff erentiate f allout froa reactors and weapons because Cs-137 is not present in weapons f allout.
2 The amount of ecciuu-137 reported (100 uCi/u ) is consistent with what would be expected to be present froa ucapons fallout and the absence of cesium-134 can be taken as a definite indication tnat there was no con-tanination by reactor caterial.
On page 23, Beyea cays "In the absence of confinaation of this presuaption (which could have been checked by testing for the ratio of cosiu:-134 to cosiu=-137), it is not scientifically valid to conclude
'l that no radiocesiua frou the accident uns present." The DOE Envircnavn-tal !!casureuents Labcratory is highly skilled in ceciua neasureacnto and October 18, 1984
T Fabrikant 73 '
Cor:nnts ci D: yea Rsport Ii, Eisenbud found no cesium-134 und reported their findings to the EPA.
I have not had an opportunity to check this in the EPA report, but the inforuation comes from Harold Beck at EtL, who made the measuremen's.
Beyea should t
have known that these measurements were made.
On pace 57 the report " assumes" that 25% of the measured cesium 2
contauination '(25 nci/m ) could have originated from the accident.
This is not possible becauso cesium-134 was not datected.
Boyca's general conclusion, given in the second paragraph on page 25, is that "For all these reasons, it appears that the official estimates for whole-body and thyroid population doses should not be regarded as final at this time.
Such a statement is not meant to imply that, in f act, the official dose estimates have been proven wrong, but only to judge that much arcater uncertainty than heretofore acknowledged should have been assi ned to the doses delivered to the population and, b
as a result, to the estinated health offects projected from the deses."
I believe the GPU position should bc: 1) that the doce estinates were made by scue of the best teams in the country, operating independently, and that they agreed within a roaeonabic factor; and 2) that the uncer-tainty in the dose estiuates is well within the uncertaintics accepted by public health authorities in risk assessuent when low levcis of risk are involved.
The highest credible actinatos place the individual popu-lation doses at less than would be received by the population due to natural sources of radiation in one year.
The report cutgests that the Public 1;calth Fund should support a cocprehensivc'research program to iuprcyc the dosiuetry.
In support of i
their rococ.nendations, the report states on page 29 "It has aircady j
October _ 18, 1904 j
-i.
Fabrikant 74 C:tm2nto ri B:yeS Raport II. Eisenbud '
becoce cicar from this preliminary study of the dosimetry that in order to minimize radioiodine in milk, not only should cows be kept indoors af ter a release of radioactivity and kept from grazinc, but they should
' be shif ted to feed that has been stored indoors or brought from distant locations..." Here the author displays his ignorance of the subject.
The Federal Radiation Council discussed countermeasures against I-131 in its reports in tile early 1960's, and identified all of the options men-
~
tiened by Beyea.
The FRC rc'cocuendations were at that time incorporated into state energency plans to deal with contamination of the milksheds by I-131. This subject is also cove' red in the 1977 report by llCRP,
" Protection of the Thyroid Gland in the Event of Releases of Radioiodine. "
In Section 6 th,e report considers the health impacts of the Three lille Island accident and states "The conversion of population dose to health iupacts for low-level radiation is conventionally accomplished by cpplying dose response estinatos roccarch cnd published by the ilational Acadccy of Sciences." The report then goes on to Give the ranse of risk coefficients used by the Acadcuy.
Dcyca fails to point out that the Academy was careful to note that the risk coefficients are derived from high doses at high dose rates, and that there is some question about their applicability to exposures less than about 1 rad.
As a matter of fact, the BEIR III report states (pabe 3) that "The Cocaittee does not knew whether doce rates of Gauca or X rays of about 100 uR/yr are detri-cental. to ::an." Decause of this positicn, the Comaittec uould not cake risk estinates for single exposures to less than 10 rads, or to continu-l ous lifetiae exposure to 1 R/yr.
l October 18,198f4
h Fabrikant 75
~
8-H., Eieanbud Co xnnts en Beys 3 Raport A fundamental problem with the report is that it attaches equal v
weir,ht to thf, Takeshi and Kepford estimates as it does to the core
~7 s
thorough studies of others listed in the report's Table 3.
- This, o
despite the fact that the Takeshi and Kopford reports wcre critically revicwed.
I am sure you will ask Pickard-Lowe to deal with the 12,000 prem estimate Beyea derived from the Uoodard report.
N i
l
'f; u
October 18, 1984
., -