4. Summary of the case studies

The Swedish ICRP Project (SwIP)
Report March 2002
4. Summary of the case studies
4.1. Introduction
Already in the terms of reference for the project case studies were mentioned as an important
basis for the analysis in this report. Six areas for such studies, all linked to Swedish
applications, were identified at an early stage of the work. The cases include doses from
normal operation of nuclear power plants in Sweden (environmental doses and occupational
doses); post-Chernobyl decisions taken by Swedish authorities; aspects on radiation
protection criteria applied to geological disposal of radioactive waste (spent nuclear fuel);
residential radon; and consumer products (smoke detectors and tritium-light watches).
There are four sites for nuclear power generation in Sweden, the Ringhals nuclear power plant
with four units (1 BWR and 3 PWRs) being the largest with a total output of 3570 MW. The
four reactors at Ringhals started commercial operation over the period 1975–1983. The
Ringhals nuclear power plant has been used as an example in the first two studies that cover
the normal operation of a nuclear power plant.
Also medical use of ionising radiation was discussed as an area where case studies or more
detailed information on actual doses could be of interest. See Section 10.
The case studies are summarised in this section and the full text of each of the case studies
(including references when relevant) are available at www.analysgruppen.org as a separate
report under the heading ‘Swedish ICRP Project: six case studies’. All but one—the study of
occupational doses in a Swedish nuclear plant— are in English. (Printed versions may also be
obtained from KSU AB, P.O. Box 1039, SE - 611 29 NYKÖPING)
Note that conclusions by the authors as presented below, may not be complete from their own
presentations and may have been chosen by us from different parts of their reports.
4.2. Environmental doses from the normal operation of nuclear power plants in Sweden
The case
In the report of the case study, the dose limits—past, present and expected in the future by the
authorities—for releases into the environment from Swedish nuclear power plants are
presented by the author, as well as an analysis of how the proposed ICRP system can be
applied to the regulation of releases from nuclear power plants under normal operation. The
author of the case study has also presented some situations with varying levels of radioactive
releases into the environment from the Ringhals nuclear power plant.
Important features and figures
The maximum radiation doses assessed from radioactive releases into the environment during
periods (up to 1997) with many fuel leaks (such as pin-holes in the fuel canning) in old BWRs
have been in the range of 40–60 µSv/year, calculated by means of old conservative methods.
The major part of these releases has consisted of external radiation from noble gases. New
and revised (although still conservative) assessment models have been developed since then.
The maximum dose calculated using the new methods decreased to 20–30 µSv/year.
Section 4 Page 1 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Releases in the early 2000s usually yielded doses in the range of 1–10 µSv/year. The adoption
of revised models will reduce the doses to 0.5–2 µSv/year, with even lower doses for some
Based on observations from many cycles of fuel leakage in Swedish reactors, the practice of
stopping operations and replacing heavily leaking fuel (especially in BWRs) has been
developed. The main reason for this is to maintain low contamination levels in the plants, and
thus, to achieve low occupational doses. This also ensures that doses to the environment are
In addition, the introduction of improved release delay systems and filters has contributed
significantly to the decreased airborne releases and radiation doses to ‘critical’ groups living
in the neighbourhood of the plant. These measures fulfil the requirements for optimisation of
doses to critical groups using the collective dose concept.
With some exceptions, modern nuclear fuel design and modern methods of operation usually
yield few fuel leaks. In Sweden nowadays it is typical to have less than one leaking fuel rod
per reactor per year on the average.
Table 4.1. Radioactive releases from Swedish nuclear power reactors compared with ICRPs
Bands of Concern (ref. ICRP 2001)
Dose, microSv/year
Band of DescripRelease 2010 Reference Actual
concern tion
min max mean
Band 4 Normal
1000 10000 3000
Band 3 Low
100 1000 300
Band 2 Trivial
Band 1 Negligible < 10 <100 <30
* Applies to single release paths and/or species
** Historical maximum doses recalculated with new models
*** Current doses recalculated with new models
It can be seen in Table 4.1 that the existing regulatory release limit falls between Bands 3 and
2. The target value for 2010, recently decided on by the Swedish Parliament, falls between
Bands 2 and 1. The highest doses from historical releases fall in Band 2, while the current
doses are within Band 1.
The reference and target levels are intended to predict the operation of a plant or a system in
order to achieve reduction of releases and to check if the operation complies with the BAT
concept. These levels may also be used to reduce the dose contribution from single nuclides
and/or release paths.
Conclusions arrived at by the author of the case study
If we compare doses from current radioactive release levels at Swedish power reactors with
the new ICRP proposal, we can see that there should be no need for any further protective
action. In fact, the doses are in the band where they should be ‘excluded from the ICRP
system of protection’.
There are, however, some international treaties under discussion that might force radioactive
releases from Swedish reactors to be reduced even more over the next 20 years. These
Section 4 Page 2 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
reductions are more far-reaching than the ICRP proposal. In the past years, there has also been
some pressure from the media to reduce releases even further.
Comments by the project group
It is interesting to note that it was possible to develop new and refined methods for calculating
environmental doses to critical groups, which has meant a reduction of the dose figures by a
factor of 2–5. A high degree of ‘conservatism’ may be found in other similar applications.
This is an example of how to approach more realistic conditions.
The case study describes the situation under normal operation of a Swedish nuclear power
station. The releases to the environment are low compared to the limits stated by the
authorities. It might be of interest also to study how the proposed ICRP system would
function in a single year with casually increased radioactive releases close to accepted limits
but still within the framework of normal operation of a nuclear power plant.
4.3. Occupational doses and radiation protection procedures in normal operation at
nuclear power plants in Sweden
The case
The actual occupational radiation doses at the Ringhals nuclear power plant during the period
1997–2001 have been analysed in terms of the proposed ICRP system. In 1997 the doses were
higher than normal because of an extraordinarily long maintenance period, which included
large-scale jobs being performed close to the reactor. During the period 1998–2001 the
situation was normal with few high-dose jobs. The study includes occupational radiation
doses both to persons employed full time at the Ringhals plant and to contractors hired for
special jobs.
Important features and figures
The ALARA principle has been applied in all internal radiological protection procedures used
at Ringhals. In the 1990s a new culture developed through systematic education so that the
ALARA principle is used not only by the management and by the radiation protection
personnel but by all persons working within the controlled areas of the plant.
In the day-to-day work, individual doses are usually the focus of concern, but collective doses
are of great concern when jobs are planned to take place during the maintenance periods
(normally once a year for between 4 and 8 weeks, when the reactor is closed down) and
longer periods (up to around 12 months) for replacing major components and/or for
performing activities involving modernisation. The rule in planning is that protection
procedures must always be taken when the cost per avoided manSv is less than 0.5 MEuro
(valid as of 2001). Procedures with higher costs per manSv can be motivated after an
individual assessment has been made.
The principal legal limit for occupational radiation doses requires that no one should be
allowed to receive a total dose higher than 100 mSv during a five-year period. According to
internal procedures at Ringhals, the goal calls for the average dose to any individual to be
lower than 20 mSv per year. Over the odd year, an individual might receive more than 20
mSv, but only after an investigation has been performed in the planning stages. If the dose
Section 4 Page 3 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
per year to an individual is regularly in the neighbourhood of 20 mSv, a special investigation
of the working conditions will be carried out.
According to internal radiation protection rules, 100 mSv is the absolute upper limit for an
individual dose during any rescue work that might be needed in the event of an acute
emergency arising.
The doses to individuals are routinely registered by means of thermo-luminiscence detectors
(TLD) and doses below 0.1 mSv/4 weeks are not measured or reported. In practice, however,
there is also a redundant system capable of measuring electronic down to 0.001 mSv per shift.
But doses at that level are not usually reported to the central dose register.
A conservative assessment shows that the dosimetry system might miss registering 2.5 % of
the total collective dose at the plant.
Some rather extensive maintenance projects were carried out in 1997 at Ringhals and during
this year, which was extreme, 172 persons received a dose higher than 20 mSv. Over the four
years that followed, 1998–2001, doses returned to normal and in all only two persons got
doses above that limit.
The distribution of individual doses within the proposed bands is shown in the following two
diagrams. The first diagram demonstrates the situation at the Ringhals nuclear power plant
over four rather normal years, 1998–2001 and the second shows the dose distribution during a
rather extreme year, 1997.
Bands of concern about individual effective dose in a year 1998 - 2001, number
of persons. Total for Ringhals.
Number of persons
100 - 1000
> 100 x normal
10 - 100
> 10 x normal
1 - 10
0,1 - 1
> 0,1 x normal
0,01 - 0,1
> 0,01 x normal
0,001 - 0,01
< 0,01 x normal
Fig. 4.1 The distribution of personal radiation doses, in mSv, at the Ringhals Nuclear Power
Plant during some rather normal years.
Section 4 Page 4 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Bands of concern about individual effective dose in a year 1997, number of
persons. Total for Ringhals.
Number of persons
100 - 1000
> 100 x normal
10 - 100
> 10 x normal
1 - 10
0,1 - 1
> 0,1 x normal
0,01 - 0,1
> 0,01 x normal
0,001 - 0,01
< 0,01 x normal
Fig. 4.2 The distribution of personal radiation doses, in mSv, at the Ringhals Nuclear Power
Plant a year with doses higher than normal.
Conclusions arrived at by the author of the case study
For occupational doses, bands 1 and 2 are not relevant to radiation protection management at
nuclear power stations.
To a great extent, the radiation protection system at Swedish nuclear power plants focuses on
individual doses but collective doses (workforce doses) are used in the optimisation of
protective action. The concept of workforce doses is a powerful tool for planning any major
maintenance and repair work at nuclear power plants.
With some minor changes, the existing procedures can be adapted to the new proposed ICRP
system. This system will probably make the application of the ALARA-principle more
understandable for everyone working at nuclear power plants. However, extra efforts to
spread information will be needed to make any advantages visible for the individual workers.
The proposed ICRP system’s focus on individual doses might force the nuclear plant owner
to take more direct responsibility for the total doses received for a year or longer periods by
hired-in maintenance workers. As things now stand at Swedish nuclear power plants, this is
only the responsibility of the contractor that employs the hired workers.
Comments by the project group
It is obvious that the use of the concept of workforce doses in the planning of maintenance
projects at nuclear installations must be encouraged in any ICRP RP system.
The case study shows that in a normal year at a nuclear power plant in Sweden few people
receive doses as high as in band 5 (protective action recommended: Reduce the dose) but in
years with extensive maintenance work several hundred persons are in the dose level of band
5. This is not a problem for people who are regularly employed at a nuclear power plant
because the average dose over several years for almost all individuals falls into Band 4 or
Section 4 Page 5 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
lower. On the other hand, the situation might be different for employees of contractors hired
for special jobs. The total dose received over several years might be rather high for specialists
moving from one high-dose job to another, perhaps in different countries. The situation has
not been covered in the actual case study but the problem ought to be analysed in more detail
preparatory to the introduction of a revised ICRP system.
4.4. Post – Chernobyl decisions taken by Swedish authorities
This case study is written as an official document from the Swedish Radiation Protection
Authority, SSI. The starting-point for the study is formal interviews, which were made
especially for this report, by an external consultant, with Gunnar Bengtsson (then acting
Director General of the SSI), Lennart Albanus (then working at the National Food
Administration), and Evelyn Sokolowski (the initiator and first leader of the KSU Analysis
Group, which was started up just after the Chernobyl accident), together with a selection of
references that describe the decisions and recommendations made in 1986-1987.
The case
The accident at the nuclear power station in Chernobyl occurred on the night of 25/26th April
1986, but became known only to the rest of the world on the morning of 28th April 1986,
when personnel passing through the monitoring detectors at the Swedish nuclear power
station in Forsmark, about 150 kilometres north of Stockholm, were found to be contaminated
with radioactive material. Field measurements showed high values of iodine and caesium on
the ground and a number of restrictions were introduced. Some restrictions for marketing fish
from small lakes and meat from reindeer are still in force.
The report concentrates on the question of the extent to which a different ICRP policy might
have affected the decisions taken by central authorities during the first years after the
Important features and figures
When it was realised that radioactive material had drifted in over Sweden, the authorities took
a number of actions. One of these measures, which was vital to assessing the need for
protective responses, was to take extensive measurements to determine the presence of
radioactive contamination throughout the country. Gradually, a picture of the extent and
intensity of the fallout took shape. By 29 April, the SSI already had a rough picture of the
fallout pattern over most of Sweden. The highest concentrations were found along the coast
north of Stockholm. The highest surface activities measured were about 100 kBq/m2 Cs-137.
During the period from immediately after the accident becoming known until about a year
after it, the public authorities made a number of decisions and published many
recommendations which, either directly or indirectly, affected the public. Some of these
recommendations are mentioned in the following examples.
Example 1. Evacuation, remaining indoors, iodine tablets. The authorities decided not to
recommend evacuation, intake of iodine tablets or staying indoors for people living in the
affected areas.
Example 2. Limiting values for foodstuffs. Some immediate decisions were taken on the use of
contaminated foodstuffs. A couple of weeks after the accident, the National Food
Administration introduced new (higher) guide values for radioactive iodine and caesium in
Section 4 Page 6 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
foodstuffs of 300 Bq/kg Cs-137 and 2 kBq/kg I-131. As the expert body involved in the
decision, the SSI stated that these values were intended to limit the average radiation dose,
over a 50-year period, to a maximum of 1 mSv/year (even for those receiving the highest
radiation dose), with a maximum of 5 mSv/year for any single year.
Example 3. Conflicting messages. Particularly during the period shortly after the accident, the
SSI issued a number of recommendations that were regarded as difficult to understand and,
not infrequently, as conflicting. On the one hand, they recommended that care be taken, in
particular as regards some aspects of behaviour etc., while on the other hand it seemed as if
they were saying that there was no need to worry if one did not follow the recommendation.
The risk of harm was 'non-existent'. The following are two examples of such conflicting
On 2 May, the SSI recommended that rainwater should not be used in households, and that
nettles, parsley etc. growing outdoors should not be eaten. They also recommended that green
vegetables should be rinsed. The SSI was aware that the intake of radioactive substances via
carriers such as nettles and parsley was small. Even if the concentration of radioactive
substances in these plants was high, the ingested quantities/activity levels were small, i.e. the
total intake would be limited and so also would be the radiation dose. In other words, the risk
to an individual was insignificant. It was felt that it would not constitute much of a sacrifice,
whether in economic or other terms, for an individual to refrain from ingesting these products
for a while. In other words, in exchange for a modest sacrifice by the individual, it was
possible to reduce the collective dose and thus reduce the consequences of the precipitation.
Another example, with a different kind of conflicting message, was the recommendation to
keep cattle indoors, or at any rate not to turn them out to spring pasture. This was interpreted
by many as meaning that the objective was to protect the cattle from radiation, while at the
same time saying that there were no risks to humans. In fact, the objective was to ensure that
milk would not be contaminated, thus protecting humans through lower concentrations of
iodine and caesium in the milk.
The concept of collective dose was used before, during and for a longer period after the
accident, in order to estimate future harm caused by exposure to radiation: for example,
expressed as the number of expected deaths from cancer. During the first few years after the
accident, the SSI published a paper estimating that there would be 300 deaths in Sweden from
the Chernobyl accident over a period of 50 years. In recent years, there has been a clear trend
in the SSI against providing any risk estimates that quantify the number of expected deaths.
According to the report, one of the problems with estimates of this type is that they also
include very small radiation doses to a large number of persons (which, in the light of the
present view of radiation protection, is correct when using LNT), where the individual risk is
negligible in practice, while at the same time it is not expected that it will be possible in
Sweden to identify any cases that can be explained by the radiation from Chernobyl.
The SSI states in the report that the main question remains: will there not always be questions
concerning the effects of an accident, particularly in terms of expected harm to persons? How
are we to answer this question, particularly in those cases where no observable effects can be
expected? The answer to this question needs to be further discussed.
Conclusions by the authors of the case study
Starting from experience of the effects of the Chernobyl accident, and on the basis of two
partly different fundamental radiological protection philosophies, the case study discusses
Section 4 Page 7 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
some of the decisions that need to be made, and recommendations that need to be published,
after a nuclear power accident. Among the conclusions that can be drawn from the discussion
are the following:
The fundamental risk picture, i.e. the view of the dose/effect relationship when low
radiation doses are involved, is the same in the present and in the proposed protection
A more individual-centred approach would probably have avoided conflicting messages
of the 'nettles/parsley' type. At the same time it is probable, based on the experience from
1986, that the SSI would probably not issue such recommendations in a similar situation.
According to the proposal for new ICRP recommendations that have constituted the point
of departure for this report, the concept of collective dose will remain. At present,
sufficient material is not available to assess how collective dose will be used in the field of
radiological protection. However, regardless of underlying policy, the concept of
collective dose should not be used as a consequential measure of future effects, i.e. cases
of cancer, which may develop some time in the future.
Any action taken by the authorities will result in people becoming affected or involved,
which produces a very complex situation in which radiological protection has to contend
with other interests, such as political commitment, or the interests of individuals in
forestry, agriculture, the dairy industry, the manufacture of food, etc. It is therefore
particularly important to be able to distinguish the elements of any special decision that
can be directly related to scientific knowledge, and those which depend on other factors,
primarily economic and social.
It is doubtful whether, in a situation similar to that which occurred after Chernobyl,
relatively limited changes in new recommendations (as compared with the present
recommendations), would actually result in any greater changes in the way in which
public authorities would act, or in how the mass media and the public would see radiation
Summarising, on the basis of this relatively limited review, it is doubtful whether a new ICRP
radiological protection system would have had more than a marginal effect on the way in
which the authorities reacted in 1986/87.
Comments by the project group
Some of the decisions by the authorities were based on the individual risk concept and others
on the collective risk concept. Mass media and the general public had difficulty understanding
that two different concepts were used and thought that all the recommendations were based on
individual risks. The authorities were seemingly not successful in communicating the
theoretical concept of collective risk. Therefore many persons wrongly believed that the risk
for cancer was rather high, which resulted in groundless fear in the affected areas.
It is probably impossible for a national authority to communicate two different types of risk
concepts with any efficiency in an acute situation as the one in Sweden after the Chernobyl
accident. One way out of the problem would be to base all the comments and
recommendations on the risk concept that is intuitively easy for the general public to
understand: the personal risk. Another way could be to inform the public specifically about
the concept of collective risk. However, this must be done not only in acute situations but
continuously and not only in connection with radiation risks.
Section 4 Page 8 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
4.5. Radiation protection criteria applied to geological disposal of Radioactive waste
(spent nuclear fuel)
The case
The Swedish Nuclear Fuel and Waste Management Co. (SKB), plans to construct a repository
for spent nuclear fuel somewhere in Sweden, to accommodate the high-level radioactive
waste from the Swedish nuclear energy program. According to the concept, called KBS-3, the
repository will be excavated in bedrock at a depth of about 500 m. No reprocessing will be
used and the spent fuel assemblies will be placed in canisters of steel and copper and
surrounded by a layer of clay in holes bored in the rock. The process of deciding on a specific
site for the repository is currently in progress.
The Swedish Radiation Protection Authority (SSI) and the Swedish Nuclear Inspectorate
(SKI), both licensing regulatory authorities, have sanctioned the KBS-3 concept as such.
From the perspective of radiological protection, the main problem is the long-lived radionuclides in the spent fuel. The future development—which is, to a large extent, unknown—of
our society, of our biosphere and geosphere will all influence repository performance over the
long term. Assessments of exposure at a time far in the future thus mean extrapolating to
situations, which are by-and-large unpredictable, or for which the accuracy of quantitative
assessments may be fundamentally difficult to evaluate. Full knowledge of a repository
system’s long-term qualities is simply not available.
Important features and figures
Several scenarios have been studied based on assumed loss or decreased effectiveness of different barrier functions. With the exception of the most extreme cases—involving relatively
rapid transport in the geosphere by colloids, or transport under oxidised conditions that not
only allow relatively short-lived radio-nuclides to be released, but also diminish the retarding
129 36
influence of sorption processes— the nuclides I, Cl, Se, and Ni dominate as agents of
exposure over the time period of main concern, i.e. within 10 000 years after closure of the
repository. Over longer periods several other radio-nuclides are expected to contribute to
exposure as well, particularly Cs and Ra.
Section 4 Page 9 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
In most cases, calculated
individual doses in populations
that receive the highest exposure
is about 10 Sv per year or
considerably lower; whereas
under unfavourable and extreme
conditions, such as those
involving colloidal transport or
oxidisation, doses may be two
orders of magnitude higher.
The span of individual dose rates
(Sv/y) covered by such scenarios
is illustrated in Figure 4.3, which
also indicates the level for the
ICRP dose constraint according
to present regulations, the
‘trivial’ dose in the suggested
new approach, as well as the
dose provided by natural
background exposure.
Dose rate
10 -2
10 -3 Natural background in Sweden
10 -4
10 -5
10 -6
10 -7
10 -8
ICRP: dose constraint
New approach: Trivial dose
Present range
for maximum
dose rate in
specific scenarios
in KBS-3
10 -9
10 -10
Nuclides of particular interest
36 Cl, 59 Ni, 79 Se, 129 I, 135 Cs, 226 Ra
Fig. 4.3 Dose comparison
Constrained optimisation is the central approach to evaluating the radiological acceptability of
a waste disposal system; it presupposes putting constraints rather than limits on doses or risks.
By changing the focus from limitation to optimisation, the need to apply, in practice,
radiological protection systems to the disposal of long-lived solid wastes can be met.
The key criterion is the individual source-related constraint. The ICRP has recommended an
upper numerical value for the dose constraint of 0.3 mSv over a year for application in normal
exposure situations (ICRP-60). This corresponds to a risk constraint of 10 .
Later recommendations deal more specifically with radioactive waste. In the ICRP’s view,
future doses should be taken into account in the protection of both populations and individuals
from radioactive waste storage, although not necessarily by applying the same system of rules
that is applied today, to current doses. The Commission recognises the problems of estimating
collective doses over long periods of time in the future. In principle, the collective dose
involves the aspects of spatial and long-term temporal distribution. However, at present the
ICRP is vague about how to perform collective dose assessments; moreover, they do not have
specific recommendations on appropriate time periods for such assessments.
According to ICRP-81 reasonable efforts should be made at the repository development stage
to reduce the probability of human intrusion or to limit its consequences in circumstances
where human intrusion could lead to doses that are sufficiently high so that intervention based
on current criteria would almost always be justified
Section 4 Page 10 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
The proposal for a new system of protection requires that exposures subject to control are first
justified and then restricted by individual-based Protective Action Levels. Moreover, it
proceeds from the idea that sources are controllable. This means that the need for protective
action is influenced by the individual dose, but not by the number of exposed individuals.
Furthermore, there remains an additional requirement to do all that can be done to make
exposures As Low As Reasonably Practicable (ALARP).
The SSI is more explicit than the ICRP about how to apply collective dose assessments in the
future. Furthermore, the SSI applies relatively rigorous conditions for dose to members of a
critical group in the future, as a precaution for avoiding higher doses to future generations
than are accepted for people today.
Conclusions by the author of the case study
Despite notable differences in the basic principles, there appears to be practically no major
change in how the present ICRP recommendations and the suggested new protection system
take into account the protection of a critical group when the wastes are finally disposed of.
Yet, the lowest dose level that prompts action to reduce exposure in case of human intrusion,
as expected in the present ICRP recommendations, exceeds that of the proposed new system
by one order of magnitude.
The essential difference in these principles concerns the optimisation process and the basis for
optimisation (global collective dose on the one hand, doses to individuals on the other).
By comparison, the criteria set by the SSI mean stronger emphasis on the precautionary
principle, and, in particular, on considerations of inequities between present and future
generations. Contrary to the ICRP’s rather vague recommendation regarding the application
of collective dose, the SSI provides specific conditions for different time periods.
In Table 4.2 there is an attempt to present a qualitative summary of how certain radiation
protection aspects have been dealt with in the proposed new protection system and by the SSI,
as compared with how these aspects are considered in the present ICRP recommendations.
Section 4 Page 11 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Table 4.2 Comparison of three different RP systems
Comparisons —in relation to the principles in ICRP-60—of the
considerations in the Proposed New Protection System and in the
current conditions expressed by the SSI with regard to inequity or
precautions against future exposure and human intrusion
Inequity or
The proposed
new protection
The national
radiation protec+
tion authority
Explanation of the matrix symbols:
+ means strengthened; = means similar;
Inequity or
means not applicable
Comments by the project group
The project group agrees with the comments by the author of the case study. The author is in
fact a member of the project group.
4.6. Residential radon
The case
The focus is on two different aspects of radon exposure (222Rn) indoors in Sweden: (a)
inhalation of radon and its progeny and (b) intake of radon-rich water. Radiation doses of
naturally occurring gamma radiation from building materials are not included in this study. In
Sweden, both of the studied phenomena are regulated. The reasons for selecting them are:
• that inhalation of radon gives high exposures compared with most other radiation
• that radon is an example of how natural radiation sources can be regulated
• that exposure to radon from drinking water might correlate with residential radon
exposure from the ground
• that communicating the risk involving radon is quite different from the communication
of many other risk situations
In addition, it has been possible to estimate the risk directly (from epidemiological studies) for
inhalation of radon. For ingestion (drinking water) it has only been possible to estimate the
risk by conventional calculation of the dose.
Section 4 Page 12 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Important features and figures
Indoor radon
Radon gas occurs naturally in the air outdoors and in all buildings. Radon is therefore
unavoidable, but in some buildings radon levels may be very high and could be reduced by
technical means. Radon disintegrates into radon progeny, the short-lived (with an overall halflife of about 30 minutes), and the first long-lived (with about 22 years half-life) in the chain of
decay. It is the short-lived progeny that give, via alpha radiation, the highest doses to the
respiratory passages and the lungs are transported via the lymphatic system and the blood to
other organs in the body.
Sweden was the first country to establish a system for regulating exposure of the general
public to radon in dwellings. This legislation was enacted in 1980, earlier than the
international organisations were to give their recommendations for reducing radon exposure.
This has led to a somewhat different approach compared with other countries, where
recommendations and regulations were, as a rule, given later.
For newly built dwellings, the limit in Sweden of 1980 was mandatory: 70 Bqm-3 of radon
progeny (corresponding to 140 Bqm-3 of radon gas). For existing dwellings the limit was also
mandatory: 400 Bqm−3 of radon progeny (corresponding to 800 Bqm-3 of radon gas).
However, the author points out that local authorities (primarily responsible for supervising
compliance with the rules) usually did not demand that levels of radiation activity exceeding
these figures be mitigated, although sanctions are known to have been imposed, for example,
in the case of children in one-family houses.
The limit for existing dwellings was decreased to 200 Bqm−3 in 1990. The limits were
changed from radon progeny to radon gas in 1994: for newly built dwellings to 200 Bqm−3,
and for existing dwellings to 400 Bqm−3.
The results of the regulation system for residential radon in Sweden have been followed up by
radon measurements in a representative sample of the housing stock in 1976 and 1988. The
radon concentrations in dwellings show an almost log-normal distribution with an overall
domestic average of about 100 Bqm-3, and a median of 53 Bqm-3, in the 1988 study (see Table
4.3., the column to the extreme right). The highest concentrations are found mostly in
detached houses, which display an average of 141 Bqm−3. For dwellings in multi-family
houses the average is 75 Bqm−3. Although average concentrations decreased to about half in
dwellings built after 1980, partly due also to other causes than the regulations, no significant
decrease could be seen as regards the whole housing stock between the two observation
periods. This was referred to the fact that newly built dwellings are only a small part of the
total stock. The aim of the Swedish system for regulating radon is to establish an average of
50 Bqm-3 in the future, although it will take very long time to achieve this objective.
The risk estimate of lung cancer from exposure to residential radon and its progeny is based
directly on epidemiological studies, unlike studies of exposure effects from other radioactive
The latest SSI estimate (in 2001) is that 500 lung cancer cases per year in the Swedish
population are the result of radon exposure in dwellings. Of these cases, 90 % are estimated to
be related to smoking in combination with the radon exposure.
Section 4 Page 13 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
If the radon concentration in all dwellings with more than 400 Bqm-3 were decreased,
potentially 150 lives per year could be saved. Active measures in all dwellings in the interval
200 – 400 Bqm-3 could save about 50 lives per year based on the latest Swedish risk estimate.
The proposals for new ICRP recommendations are mainly based on a concept that focuses on
individuals. This approach has been used for radon exposure earlier, both in ICRP-65 and in
the Swedish regulations. The main basis for the author’s comparison of Bands of concern and
Protective Action Levels according to ICRP’s proposal with present conditions and
regulations in Sweden has been summarised in Table 4.3.
Table 4.3. Swedish radon data seen in the context of the ICRP proposal (ICRP 2001).
The first four columns are based on tables 1 and 2 in the ICRP proposal, complemented with Swedish data.
The concentrations are calculated using the risk estimate and dose conversion convention established in
ICRP-65. Within brackets the corresponding concentrations have been calculated on the basis of the
Swedish risk estimate of 2001.
Band of
Band 6
Band 5
Band 4
Level for
Relocate or
shelter in
Reduce the
Level of
Swedish Rnconcentrations
in dwellings
mSv y-1
Bq m-3
Bq m-3
Bq m-3
Band 3
Band 2
Band 1
from the
system of
Max. measured
Mean 100
Median 53
Outdoor conc.
Mean 10
Outdoor conc.
Outdoor conc.
a) Within parenthesis, multi-family houses
Section 4 Page 14 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
As seen in Table 4.3, the limits are found in the same Bands - despite the difference in the
dose conversion convention.
The Protective Action Levels in Band 4 are proposed as a warning to ‘reduce the dose’, as
regards the individual. The average radon concentration is higher than the lower end of the
range. The Swedish limits and the present ICRP recommendations (ICRP−65) are within the
range. According to the author, it would not be realistic to have a limit at the lower end of the
interval, when half the number of Swedish dwellings exhibit higher levels. The table also
shows that radon concentrations corresponding to dose levels in Bands 5 to 6 have been
measured in the Swedish housing stock. The ‘typical protective actions’ in these bands are
radical, especially for Band 6. The Swedish regulation system only advises that ventilation be
temporarily increased by airing until corrective measures can be taken.
The outdoor concentrations are in Bands 1 - 3. They are usually low but can—in extreme
cases over short-term periods—go up to about 100 Bqm−3 (Band 3).
Finally, one may also take note of the following questions: The proposal for new ICRP
recommendations is based on effective doses. For radon exposure via inhalation, the problem
will be the same with the calculated effective dose as it is for the present ICRP
recommendations, which are based on the dose conversion convention. According to the
Swedish regulatory system the risk estimate is based directly on the results of epidemiological
studies of radon exposure and lung cancer without the use of any dosimetric calculations. So
the question is, which comparison (with a level of dose that calls for protective action with
regard to the individual) should be used to analyse the effect of the ICRP proposals: the
effective dose or the risk-based dose conversion convention? Further, the magnitude of the
risk estimate for lung cancer differs between ICRP−65, in 1993, and the SSI in 2001, by a
factor of 2−3. Thus we may ask which risk estimate should be used as a basis for calculating
the dose conversion convention?
Radon in drinking-water
The highest risk caused by radon in tap water is its exhalation, almost 90 % of the dose, from
water into the air. On intake, the stomach absorbs radon. The greater part of the intake leaves
the body by exhalation within an hour.
From 1997 on, the Swedish limits for radon in drinking water have been 100 Bq L-1 and 1 000
Bq L-1, where the latter figure indicates what is unfit for consumption. These limits are
mandatory for public water plants and recommended for private wells. According to a survey
in 1988, all larger water plants were well within the limits imposed later, the average being 17
Bq L-1. Many private wells are rock-drilled, of which a majority show activities > 100 Bq L-1
and a fair number > 500 Bq L-1. Private wells using groundwater also show great variations,
the average being about 200 Bq L-1. Towards the end of the 1990s, 4% of these had
concentrations above 1 000 Bq L-1, corresponding to Bands 3-4 in the ICRP proposal.
Insufficiencies in interpreting the communication of risk
Tenants and owners of dwellings should be aware of the risks taken by living with high levels
of radon exposure. In Sweden there have been major efforts of information directed towards
the general public, politicians, estate agents, etc., by various means and media including
training consultants to perform measurements and how to take proper measures for reducing
the air concentrations.
Section 4 Page 15 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Studies of risk perception and risk communication have been carried out. Some findings (by
L.Sjöberg in 1989) were that the radon risk, both individual (for oneself) and general (for
others), was assessed as lower than the average risk for all hazards; that the likelihood of harm
occurring was assessed to be lower than the average risk for all hazards; but also that the
possibility of protecting oneself against risk was assessed as lower than the average risk for all
As expressed in the case study, the author’s personal impression is that most Swedes know
something about radon but that there may be different reasons for risk denial as well as for a
house owner not wanting to know about the actual radon conditions in case the price of the
house might plummet. In other cases, some people may have overestimated the risk and taken
more costly measures to mitigate it than the family could afford.
Conclusions by the author of the case study
The fact that radon exposure can be seen as ‘natural’ may be grounds for ignoring the risk it
poses. In many people’s opinion, anything ‘natural’ is good. However, we use radioactive
building materials and water from rock-drilled wells, and keep radon indoors by bad
ventilation - all of which are man-made developments.
Not only does the risk seem to be underestimated by the general public; our potential for
decreasing risk through our own actions is underestimated as well. Therefore, the wisest
course of action would be to increase public awareness of the health hazard, while at the same
time improving awareness of measures that might mitigate the danger. Informing people
about risk is unlikely to worry them, because most people are prone to retreat comfortably
into denial. Changes that have been made in the regulation system for radon in dwellings have
given rise to widespread information campaigns by the authorities. Such changes have
sometimes resulted in financial problems for individuals as a result of their property values
being reduced.
Another aspect on communication is the fact that the whole population might be exposed to
high radon levels in their homes and/or at their workplaces. This implies great difficulty in
formulating and distributing information, which is not the case with most information about
other radiation sources.
As to a new radiological protection system in accordance with the ICRP proposal, it would
not change the present concept whose major aim is to protect individuals, nor would a new
system change the control organisation (in Sweden) or its functions.
Comments by the project group
Radon gas is a ‘natural’ phenomenon but doses from radon in dwellings can certainly not be
characterised as unavoidable in the usual sense of this word. Doses to the lungs can be greatly
reduced by constructing houses in a way that minimises the danger and by arranging efficient
ventilation systems in both old and new dwellings.
It is interesting to note that according to the model used for calculating radon risks, radon
cases are in fact being transferred to lower Bands of concern (in the ICRP proposal) as the
average consumption of tobacco decreases.
Section 4 Page 16 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
The main concern in terms of radon exposure is that in practice there are still so many
dwellings in Sweden that cause high doses to the lungs. The case study indicates that
communicating health risks in this context and getting a response shown as measures taken to
reduce indoor exposures, has met with some difficulties. In spite of this we agree with the
author that it should be possible to increase public awareness of the individual cancer risks
from radon by means of effective information programmes. Likewise, we also think that,
above all, it is important to provide information on practical means of reducing existing
indoor radon air concentrations, as has also long been a strategy of the Swedish authorities.
For more discussion on the main principles connected to cancer risk assessments from radon
exposures and other comments, see Section 6.3.
4.7. Consumer products
The case
Two consumer products containing radioactive sources have been presented in the case study:
smoke detectors for domestic use and watches containing gaseous tritium-light devices
(GTLS). The first product is in wide use in Sweden but in the second case the authorities have
decided that the use of wristwatches containing GTLS could not be justified, and they have
thus been banned from the Swedish market.
Important features and facts
Ionising chamber smoke detectors (ICSD) for domestic use are constructed with a sealed
americium-241 source, 35 kBq (at the most). The detector is operated by a battery. The first
regulations on the import, approval, use and disposal of smoke detectors were issued by the
authorities more than 20 years ago. The fire safety and security authorities as well as the
Swedish Consumer Agency were called in to examine the benefit of the product and these
authorities decided that the use of ICSDs was justified from the point of view of home safety.
The Swedish regulations state that the import, manufacture and wholesale trade of ICSDs
require a licence, although retail sales and home use do not. Before being licensed, each type
of ICSD must be type-approved by a notified body or by the Radiological Protection
Authority, SSI.
Today there are some millions of ICSDs in Swedish homes, and in a few years there will be
several million more of them in use because the installation of smoke detectors is now
compulsory in every newly built apartment and house to ensure fire safety.
Radiation doses to the users are estimated at 0.07 microSv/year or a total collective dose of
about 1 manSv/year (9 million people in around 3.4 million households).
The SSI has stated that used ICSDs may be disposed of as domestic waste. A conservative
estimate shows that if all the ICSD waste were incinerated, doses to the workers involved and
to people in the immediate vicinity would be much less than 1 microSv/year.
There is now a growing demand in Sweden that all wastes be separated to protect the
environment. Several large communities have already decided that smoke detectors must be
disposed of separately as hazardous waste because they contain a radioactive source. This
development might cause the people who separate the wastes to be exposed to radioactivity.
Section 4 Page 17 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
Still another problem is the waste treatment of electronic equipment. ICSDs are defined by the
Swedish Environmental Protection Agency as electronic equipment because they contain a
printed circuit board. Thus, the first step in the waste treatment of these products must be the
retrieval of electronic components.
Dismantling the detectors and separating out the americium sources can be questioned from
the perspective of occupational protection because the direct intake of released americium
from a damaged housing could result in a considerable individual dose to the person handling
the device.
In the late seventies, the SSI decided that the use of wristwatches containing GTLS was not
justified. The decision was prompted by the fact that there are battery-run watches with
battery-run lights on the market as well. Furthermore, the watch had neither a lifesaving nor
any other safety function. The decision was appealed by the importing company, but when the
case went to court, the court upheld the agency’s decision.
A wristwatch may contain a gaseous tritium (H-3) light that illuminates the clock face. The
activity is of the order of 10 GBq. The doses are trivial; normal use will give only a skin dose
of the order of 1 mSv/year (equivalent dose), whereas continuous leakage in a room yields a
dose of the order of 1 microSv/year. Even if the glass housing broke, it would be impossible
to inhale more than a small fraction of the activity, which would equal a maximum effective
dose of 0.1 mSv on that single occasion.
Conclusions by the author of the case study
The new ICRP proposal still demands that the use of such devices be justified, and typeapproved ICSDs will continue to be considered justified by the public and the authorities as
long as no alternative in function and cost appears. Radiation doses to users will be
negligible—appearing in the lowest band of individual doses in the ICRP proposal—and the
use of smoke detectors will still be exempted from licensing.
Nevertheless, the proposal does not solve some problems that might appear in the near future.
The real stakeholders, the general public, might begin to demand much more expensive ways
of disposing of ICSDs, which would increase the cost dramatically so that the detectors would
not be as popular as they are now.
In the long run the new proposal by the ICRP will probably lead to more open discussion and
more information being spread concerning risks to society. In all likelihood, this will lead to
public acceptance of very low doses.
The ICRP’s publications do not contain much information about how decisions on
justification are made. Perhaps this might be something to consider in the future work on the
new proposal.
Comments by the project group
The new problems arising from the demand for separation of different types of electronic
waste is interesting from the point of view of principle. In this specific case, the problem has
arisen when going from dilution to concentration of the hazardous component, whereas it may
be the opposite in many other cases. In the future, society will probably have to analyse and
compare, in particular low, risks from chemical and radioactive components to get optimal
Section 4 Page 18 (19)
The Swedish ICRP Project (SwIP)
Report March 2002
waste treatment systems. This is an interesting area on which the ICRP would do well to
begin focusing its interest.
With regard to the proposed ICRP system of protection, we believe that justification would
still be required (cf. Section 9.2) in a case like the wristwatches containing gaseous tritium
giving low radiation doses to the individual. However, one may speculate as to the general
signals given in the proposal and whether or not these watches would have been accepted if
the decision were based on the suggested ICRP system.
Section 4 Page 19 (19)