ML20043C225
| ML20043C225 | |
| Person / Time | |
|---|---|
| Issue date: | 02/29/1988 |
| From: | Alexander B NRC |
| To: | Cunningham D NRC |
| Shared Package | |
| ML20042C963 | List:
|
| References | |
| FRN-53FR49886, RULE-PR-CHP1 NUDOCS 9006040254 | |
| Download: ML20043C225 (14) | |
Text
.
.4
,...y
p UNITED STATES ET A i
NUCLEAR REGULATORY COMMISSION
._ r maanmooTON,0A 33555
)( ) {.D I-7 h
[
e..++
4169111%
>l>
ff E 1 JdKdi A I I'
'I 1 J 5 s A ).1
)
L X XtA 4 A X T 1 W.x 1 1
4 :=_7' ~ ~
l Sk Az +
cum
' Y:
(
YhW A=.m -
f
=
h W kW= -,= = = ;,
Be t, A
_n Q,
2.
W.a. kJ d & #
w. '>t
,/,f 12 )
n K4 ;,d' l-w i
i h
n; i
.O h
E l
9006040254 891130
'PDR' PR
- CHP1 53FR49886, PDC l:
j%
qy CLARIFICATION ON RADIATION RISK CONTROL ODNCEPTS t
s"~
R. E. Alexander i
The Dose Limitation System and ALARA The radiation dose. limitation system employed by the NRC has two basic components -- regulatory limits and the ALARA concept.
These components work-together to maic. Lain individual and population (collective) radiation risks at an acceptable level.
For a given licensed activity 4. dose distribution exists in which few if any people are exposed at the limit.
ALARA (reasonable dose-reduction) methods are used to assure that.the dose distribution curve has a peak considerably below the limit.
The population risk is proportional to the collective (personrem) dose, which in turn is proportional to the area underneath the dose distribution curve.
All three of these are determined by the average individual dose.- Thus individual and collective risks may be controlled by establishing dose limits and by controlling the average individual dose using reasonable ALARA methods.
In the ICRP system, an occupational fatality criterion of one in 10 thousand per year, average, is used as an acceptable level of risk for all of the' risks I
to which workers are exposed.
This risk level is associated with an annual
' dose of 0.5 rem, averaged over a worker population.
However, the annual dose limit is 5 rems, associated with a risk of one in 1 thousand per year or, for example, 4% in the case of a 40 year career.
It is unnecessary for an individual worker to accept a lifetime radiation risk of this magnitude, but it is necessary to provide for operational flexibility under a dose limit somewhat above the average.
Otherwise operational inefficiencies may increase the collective dose.
Experience has shown that the ALARA concept in combination with a 5-rem per year limit will provide an annual average risk level below the h
1
1!$'.t I
+
a.-
.ICRP acceptable' risk level of one in 10 thousand.
For example, in 1986 the
.i average individual dose among workers in the U.S. nuclear power. industry receiving measurable doses was 0.41 rem, or 0.8 in 10 thousand.
It is evident that the ALARA concept is an essential component of the dose limitation system, but a question of great interest arises as to the conditions under which implementation of the concept should cease.
Under the ICRP
" optimization" concept, this point is reached when the monetary costs for protection, added to the monetary costs for health effects, pass through a
(
minimum cost.
A more commonly used method involves the use of a selected
-dollars per personrem criterion, e.g., $1,000 per personrem in Appendis. 1, 10 CFR Part 50, with a requirement to provide a protection measure if its cost, divided by the number of personrems it will save, is less than the criterion.
For example, if the purchase, installation, and operational cost of an effluent clean-up system is to be $20 million over the life of a plant, but the system
~
would save 40 thousand personrems, the dollars per personrem ratio would be
~
500, and the system would be considered cost effective.
In many other countries ratios of less that 1,000 dollars per-personrem are used since 1,000 is associated with a cost of about $5 million per theoretical life saved.
Applicability of the ALARA Concept I
-The ALARA concept is an integral and essential part of any risk control system in which a' dose limit is used; implementation of the concept determines the l'
shape of the dose distribution curve (and thus the collective dose) beneath a L
regulatory limit that few if any peor?e receive.
Dose limits for workers, or for members of the public from eny of the various sources of public exposure, 1
2
aitiL h;;
.are accompanied by a regulatory requirement or exhortation to maintain doses as
}
far below the limit as is reasonably achievable.. This approach to radiation risk control has been almost universal for several years.
The radiation protection community is now considering the role of the ALARA concept with respect to Below Regulatory Concern and below personal interest Itvels.
It-appears likely that the idea of regulatory implementation of the ALARA concept at levels that are already Below Regulatory Concern will be abandoned as incompatible-with the concept of " concern." However, since risks Below Regulatory Concern are not neeessarily trivial, voluntary implementation of the-
+
0 ALARA concept is to be expected, particularly in view of potential litigation.
It appears possible that even voluntary implementation of the concept will eventually cease at levels below personal interest as the public becomes better
[
informed. ALARA concept applicability is summarized in the table below.
1 ALARA CONCEPT APPLICABILITY l-Dose Range ALARA Concept L
Above the Limit Doses Not Permitted from the Limit to BRC Regulatory Implementation From BRC to Below Personal Interest Voluntary Implementation From Below Personal Interest to Zero Inapplicable
(
It 'is important to recognize (1) that the shape of the dose distribution curve
.[
beneath each limit is determined primarily by implementation of the ALARA concept, (2) that such implementation therefore controls and determines the
!~
magnitude of the collective (personrem) dose, and (3) that the collective dose is proportional to the average individual dose.
Thus the ICRP dose limitation L
system for workers relies on a 5 rem per year limit and an average individual 3
S$*. :
,, ' ' 4;
.gr, 7
~
a'
~ dose of 0.5 rem per year.
RES'is now studying the possibility of a BRC-level E
Lfor occupational; exposure; below this level, ALARA implementation would become voluntary.
An interagency committee organized by the EPA is studying 1
appropriate ways to apply all of these concepts in the protection of the public from the various sources of radiation to which they may be exposed. The status j
of.this work.is discussed in the following section.
t c
a.
Radiation Protection of the Public 1
l Current guidance to Federal agencies on the subject of protecting the public 1
i 1-from radiation was issued by. President Eisenhower in 1960.
An interagency L
committee under EPA leadership is now revising the guidance.
The~ committee is attempting to develop the new guidance within a logical framework-that covers
,i c
all sources of public exposure and incorporates the c'oncepts previously-j i
discussed in this paper..In summary:
j
-i (1) OfLthe many sources of.public exposure that are subject to governmental control, it is unlikely'that any group of people would be exposed
't i
repeatedly and simultaneously to more than four of them.
4 i
(2) Thus if an annual individual dose limit of L millirems is established in the guidance for all. sources combined, the upper bound for any individual source should not exceed 0.25 L.
(3) An individual operator would be deemed to be in compliance with L if doses
?
from sources under the operator's control were maintained below 0.25 L.
S 4
i
.b
s s
..s (4) Agencies would control the collective dose by requiring implementation of
^9,*
the ALARA' concept for each source, with the objective of maintaining the average individual dose at 0.1 L.
(5) l An across-the-board BRC. level would be established, most probably to
~
coincide with the ALARA objective as in the case of occupational radiation protection, i.e., the BRC level would be an average individual dose of 0.1-L, and implementation of the ALARA concept would be voluntary while this average is maintained.
(6) The guidance would also establish a below personal interest level; the 1 millirem per year level recently recommended by the NCRP will-undoubtedly receive strong' consideration.
It appears likely that the value for L will be 100 millirems per year, ai currently recommended by the ICRP.
If so, this proposed framework is fully compatible with current EPA standards for radiation protection of the public-as published in 40 CFR Part 190 and the Clean Air Act.
These standards are applicable to most NRC-licensed operations.
-Lowering the non occupational dose criterion from.500 to 100 millirems per
- year, Land continuing-to apply this limit to workers who are incidentally-exposed to radiation, would in some cases create a new and unnecessary requirement for training and monitoring.these workers. 'This problem would be extensive, for example, within the 000 where doses for these workers often cannot be maintained below 100 millirems per year.
Some NRC-licensed medical institutions are also concerned about this problem.
Therefore the Committee is 5
i 7',
14..
y t
y,.
i x
m conside' ring 500 millirems per year as the criterion for: incidentally exposed workers.
It should be noted that the doses for this group are readily.
controlled and that the group would not normally include children or unhealthy personnel; under these conditions the higher dose limit appears to be E
'justifled, i
~
i;
'Below Regulatory' Concern.
1 h
Many people'tielieve that levels of radiation risk exist sufficiently low that continued implementation of the ALARA concept by regulation cannot be
' j usti fied.- The term Below Regulatory Concern (BRC) is used in reference to 1
these levels.
At a BRC level, regulatory restrictions, including implementatio'n of the ALARA concept, are discontinued.
The risks associated p
with' BRC levels; are not-.necessarily trivial.
Therefore, governmental adoption 1
L
'of a BRC level. is not-intended to discourage further implementation of the j
E ALARA concept on a deregulated and voluntary basis.
For example, a BRC r-
.1 p,
' occupational risk level of one in 10 thousand per year average, the ICRP g
acceptable; risk criterion, could be considered by the NRC.
If adopted, the L,
agency would then leave implementation of the ALARA concept as a voluntary i
matter as long as the average annual risk for a given worker population would be less than one in 10 thousand.
Current NRC regulations include several
,s applications of BRC levels, although they are seldom referred to as such.
L
'These levels were. adopted by previous Commissions not because levels lower than
=these'are believed to be' risk free but because the degree of risk is too small L
ito ' justify governmental intervention.
6 i
i
M l
c.
.g m.
M y~
De Minimis 0
\\
The term "de minimis" is taken from a Latin phrase often translated as "the law a
is not concerned with trifles." The word " law" gives the phrase a "Below Regulatory Concern" connotation, but the word trifles" disqualifies the phrase for BRC purposes because BRC risks are not'necessarily negligible.
In common usage, the triviality connotation outweighs the legal connotation; and many 4
people associate the term de minimis with radiation risks that are below y
personaliinterest.
For example, in the United Kingdom:the NRPB has adopted a y
l~
de minimis-level of 5 millirems per year and describes any associated risk as-l too low to be considered in personal decision-making processes.
For the major revis' ion of 10 CFR Part 20 the NRC proposed a de minimis level of 1 millirem per year to be applied to collective dose calculations perforced in connection with NRC requirements.
More recently the NCRP has recommended 1 millirem per year as a " Negligible Individual Risk Level." The decision by the NCRP to bypass the term de minimis has been welcomed by many as the best way to avoid the intrinsic implication that it is appropriate for the -law to bn concerned at risk levels <immediately above those that are trivial,. leaving no range'in which N
l-
. people are allowed to. provide protection on an entirely voluntary _ basis.
It appears likely that the term de minimis will fall into disuse and that a term connotating the idea of "below personal interest" will take its place.
p 1'
The Data Base u
3 The prospect of BRC levels raises the question of whether such levels can be established safely.
The linear-nonthreshold hypothesis used in the development
'of radiation protection standards has led to the adoption of two familiar I
7 i,
i i.-
l J
M w
J.
- < 4:.
axioms:
(1) every radiation dose, however, small, causes a risk of cancer, and (2) no radiation dose should be permitted without justification.
If probabilities are not taken into account, these axioms are consistent with the hypothesis. With respect to BRC levels, the question of interest is whether the axioms are consistent with the data base.
Regulatory decisions regarding radiation protection depend to a great extent on radiobiological and epidemiological data.
The principal data base supporting the occupational dose limitation system comes from an epidemiology study of l-Japanese atomic bomb survivors.
Pertinent details of this study through 1982 are summarized in_the following table.
l LIFE SPAN SAMPLE STUDY (1950 - 1982) q l
Number of survivors included 91,231 L
Number receiving absorbed doses greater than 1 rad (to 600 rads) 54,058 l
Number receiving less than 1 rad (control group) 37,173 Total deaths from all causes 31,043 Total deaths from cancer 6,270 Cancer deaths in exposed group 3,832 Cancer deaths in control group 2,438 l
Cancer deaths in exposed group, not bomb related, inferred 3,514
-Cancer deaths in exposed group,-bomb related; inferred 318 Standard Mortality Ratio:
1.09 Statistical procedures were used to infer, from control group data, the number of cancer deaths that would have occurred among the exposed group if the bomb explosions had not occurred, viz,, 3,514.
Subtraction of this number from 8
~":...
i
'I n.y.:
3,832, the total cancer deaths recorded for the exposed group, provides an o
inferred number of cancer deaths caused by atomic bomb radiation.
The Standard Mortality Ratio (SMR) shown in the table is the ratio' of cancer deaths among the exposed group to those among tne control group. As indicated, the SMR for this study is 1.09.
To the extent that the control group experience does not match that of the exposed group had no exposure occurred, an error is' introduced.
Since such errors can be large, epidemiologists usually express concern only when the SMR is considerably greater _ than 1.09.
An SMR of 1.09 implies a 9% mortality increase if the estimate of deaths that would have occurred without the radiation is without error.
If this error exceeds 9%, the study results do not provide overall evidence as to whether atomic-bomb radiation-induced cancers occurred.
However, more conclusive evidence can be obtained by focusing attention on specific cancer sites receiving very large absorbed doses, and these data clearly support the need for controlling occupational radiation exposures.
In the 1980-BEIR Report information is provided as to the relationship between absorbed dose and the sample size required'in an epidemiology study to test a small absolute cancer excess, assuming that the excess is ac.tually proportional to the dose.
An approximate sample size of 50,000 exposed subjects was accepted by a majority of the BEIR-III Committee for doses of 10 rads or more.
From the information provided it appears that the following sample sizes might be acceptable for low gamma or x-radiation doses:
l 9
L 1
a
. $h,4h f
s.
- 99r
~
RELATIONSHIP BETWEEN DOSE AND REQUIRED EPIDEMIOLOGY SAMPLE SIZE FOR LOW-LET RADIATION Absorbed Dose to Number of People Required the Whole Body in the Exposed Group l
10 rads 50 thousand 1 rad 5 million-100 millirads 500 million 10 millirads 50 billion 1 millirad 5 trillion For reasons evident from this discussion, the BEIR-III Committee did not-provide radiation risk coefficients for populations in which absorbed doses of less.than 10 rads are received.
Their report does not state that there are no risks below 10 rads; the indication is that the data base is insufficient for i
useful risk estimates for lower doses.
This data base provides single-exposure.
information.
- owever, to assist government agencies in the establishment of radiation protection standards, the Committee provided similar coefficients for use with
- large irradiated populations subjected to lifetime dose rates of I rad / year or more.
This decision by the Committee is consistent with a statement published by the NCRP according to which protracted doses are believed to: be less effective in causing cancer by factors of 2 to 10.
Results of the atomic bomb survivor study for low absorbed doses are summarized below.
The data do not support the linear, non-threshold hypothesis; but as j
)
t.
l~
10 1
y;
~
yn* y t~
z, p 4, ; : t -
- .i b
j
.5 8
4 4
. LOW-DOSE RESULTS FROM ATOMIC BOMB SURVIV0R STUDY 1
v>
Leukemia Solid Tumor l
< s:
' Dose (rads)
Expected Observed Dose (rads)
Observed 0 -11 72 46 i
< 10 fewer than expected.
1-9 52 24 10 - 49 34-33
> 10 all excess
> 50 all excess e
previously indicated, the statistical power for doses of 10 rads or less is insufficient for definitive conclusions.
The absence of excess leukemia at P
doses less than 50 rads is significant, and the finding that all deaths inferred from solid tumors occurred among survivors receiving more than 10 rads I
.is.also considered to be.important.
1 t
h These data do not support the lir.aar-nonthreshold hypothesis where individual doses are very low.
They indicate that the probabilities of radiation-induced f
'+
tcancer at doses only a small fraction of natural background are too low to be of governmental concern.
They also indicate that the axioms mentioned earlier E
care not applicable where probabilities of harm are known to be very low.
Risk Coefficients and Their Use
- 1 The epidemiology study previously discussed has provided statistically sound evidence that the consequences of a collective dose, delivered to a large population, could be estimated as about 2.3 fatal cancers per 10,000 person 11
.__.l.,
.u.
m
I
.y j,
N rads.
The BEIR-III Committee, National Academy of Sciences, has provided similar coefficients for large populations exposed on a lifetime basis to 1 rad / year.
The use of such coefficients with lower doses and dose rates has a progressively weaker scientific basis as doses and dose rates are reduced, t
In the development of basic radiation protection standards such as dose limits, gaps in technology are normally compensated for by adopting conservative
^
assumptions.
This practice provides for progress without compromising safety; and for this purpose the use of hypotheses and models based on risk 1-coefficients for doses less than 10 rads and dose rates less than 1 rad / year.
I may be beneficial to a regulatory program.
Many of the unnecessary costs attributable to these assumptions can be reduced in time through research.
i 1.
1; Extrapolation of the use of these coefficients into the lower dose and dose i
L
. rate regions to provide information for other applicatiuns can lead to l
scientifically unsound decisions, depending on the magnitude of the doses involved.
As the doses become progressively-lower, hypotheses become conjecture, and conjecture is reduced to speculation. The severity of this problem is strongly dependent upon the importance of the decision to be-1 reached.
l s
This coefficient is being used in the reactor safety study, NUREG-1150.
12
[L...
f-.
n,. '-
43
- Document Name:
COMMISSION LETTER ENCLOSURES Requestor's ID:
HAYES Author's Name:
RAlexander/BH Document Comments:
'- ENCL. 3 and 2--- ENCL 6 3 in CM/ RELATIONSHIP OF DE MINIMIS i
e 1
.]
L q
L 7;
a
)
1 l.
y l-
_l l..
l
?
[
l l
}
l
)
.l l'
.j
..j l':
'l i
-. - - - - - - ----