NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Evaluation of EPA Guidelines for Exposure to Naturally Occurring Radioactive Materials. Evaluation of Guidelines for Exposures to Technologically Enhanced Naturally Occurring Radioactive Materials. Washington (DC): National Academies Press (US); 1999.

Cover of Evaluation of Guidelines for Exposures to Technologically Enhanced Naturally Occurring Radioactive Materials

Evaluation of Guidelines for Exposures to Technologically Enhanced Naturally Occurring Radioactive Materials.

Show details

8Indoor-Radon Guidelines and Recommendations

National and international agencies operating under different directives have been responsible for addressing the health risk associated with indoor radon and for addressing its regulation. This chapter provides a review and comparison of national and international guidelines and recommendations regarding radon in dwellings, schools, and workplaces. It also examines the differences in the scientific information and in risk-management polices used for developing the guidelines.

In 1970, through the enactment of several statutes, the US Environmental Protection Agency (EPA) became responsible for the establishment of environmental protection standards for both radiologic hazards and chemical agents. Although there is not now a federal regulation for indoor radon, EPA has issued guidance under the Indoor Radon Abatement Act about the risk, measurement and remediation of radon in homes and schools (EPA and DHHS 1994; 1986). EPA's published guidelines and recommendations for indoor radon are different from those of other bodies that develop guidance for radiation exposure of the public.


Concentrations of radon gas in air are normally given in units of picocuries per liter (pCi/L) or becquerels per cubic meter (Bq/m3); and 1 pCi/L is equal to 37 Bq/m3. Concentrations of radon decay products (RDPs) normally are expressed in working levels (WL); 1 WL is defined as any combination of the short-lived RDPs in 1 L of air that results in the ultimate release of 1.3 × 105 MeV (2.1 × 10−5J/m3) of alpha energy—about the amount of energy emitted by the short-lived decay products in equilibrium with 100 pCi of radon. In general, equilibrium does not occur in houses, because ventilation removes some of the radon and its decay products; also, it takes time for the entering radon to produce its decay products. Because RDPs have a static charge, they plate out on walls, furniture, and other solid objects; this reduces the equilibrium ratio (ER)—the concentration of radon progeny in air divided by the concentration that would exist if the progeny were in equilibrium with the radon gas. The ER ranges between 0.3 and 0.7; an ER of 0.5 is commonly assumed as an average. At less than equilibrium, 1 WL is equal to the product of ER and the radon concentration in picocuries per liter divided by 100. A house with 150 Bq/m3 (4 pCi/L) is likely to have 4 × 10−7 J/m3 (0.02 WL). A working level month (WLM) is a measure of time-integrated exposure and is the product of time in working months, which is taken to be 170 hours, and working levels (WL). Thus, 1 WLM is equal to the product of average WL and hours of exposure divided by 170. Under full occupancy conditions (8,760 h/y), residence in a house at 150 Bq/m3 results in about 1.0 WLM per year of exposure. In SI units, 1 WLM is approximately 3.5 mJh/m3.


Inhalation is the principal route of radon exposure of humans. The dose contribution from the inhaled radon gas itself is small under normal conditions of exposure. Exposure to radon is due mainly to the inhalation of its short-lived decay products (polonium-214, polonium-218, lead-214, and bismuth-214), which deposit nonhomogeneously in the human respiratory tract and irradiate the bronchial epithelium. Two progeny, 214 Po and 218Po, deliver the most important alpha-radiation dose to the lung (NCRP 984c). About 90% of RDPs can attach initially to airborne particles (ICRP 1993b); tobacco smoke provides additional attachment sites for RDPs. The unattached fraction (10%) has a higher rate of deposition and is more efficient in delivering a dose to the critical cells (basal and secretory cells) of the lung (UNSCEAR 1993); about two-thirds of the total dose in homes from radon comes from the unattached fraction (National Research Council 1988). The assumed health effect end point of high exposure to indoor radon is radiation-induced lung cancer.

This chapter focuses on the lung-cancer risk associated with inhalation exposures to RDPs. Although risks to other tissues posed by radon can occur through ingestion of water with high radon concentrations, they are much smaller than those associated with inhalation exposure to RDPs. All other exposure pathways distribute smaller amounts of radon and progeny over a much larger tissue mass with correspondingly lower doses and risks.


The existence of high mortality among miners in central Europe was recognized before 1600, and the main cause of death was identified as lung cancer in the late 19th century. It was suggested in 1924 by Ludewig and Lorenser that the cancers could be attributed to radon exposure (ICRP 1993b). EPA classifies radon as a known human carcinogen on the basis of data from epidemiologic studies of underground miners. That classification is supported by a consensus of national and international organizations (IARC 1988; National Research Council 1988; ICRP 1987b; NCRP 1984c). Further information on the deleterious health effects associated with exposure to radon has been provided by experimental studies of animals (National Research Council 1988).

The main source of quantitative information on the risks posed by radon exposure is the epidemiologic studies of miners which uses data on thousands of occupationally related lung cancers among many diverse groups of miners. The epidemiologic evidence of the induction of lung cancer after inhalation of radon comes from several cohort and case-control studies of underground miners, particularly uranium miners. The evidence has been reviewed and summarized in other reports (ICRP 1993b; National Research Council 1988; UNSCEAR 1988; UNSCEAR 1986). Most of the data are consistent with the assumption of a proportional relationship between risk and cumulative exposure (linear, no-threshold response model). The exception to linearity occurs at very high exposures (over 2,000 WLM), where the response per unit exposure decreases; this exception is attributed partly to the reduced life expectancy of the miners at such high exposures (ICRP 1987b).

The epidemiologic findings have to be extrapolated to provide risk estimates for long periods of exposure and for populations other than those studied. For estimating lifetime risk from data covering shorter periods, projection models are used. Different types of risk-projection models have been proposed to estimate the possible lifetime risk of lung cancer posed by inhaled radon progeny in homes on the basis of the results of the epidemiologic studies of miners (National Research Council 1988; ICRP 1987b; NCRP 1984c).

The National Council on Radiation Protection and Measurements (NCRP) model is an attributable-risk projection model based on information obtained from several groups of underground miners in the United States, Canada, and central Europe (NCRP 1984c). The model expresses lung-cancer risk uniformly with time after exposure, with the restriction that tumors do not occur either before a 5-y latent interval or before the age of 40. The model uses an initial, age-averaged risk coefficient as derived from data on miners and assumes a decrease in the initial potential excess rate with time after exposure according to an exponential function. That functional structure provided an age dependence so that the model would fit the observations of lung-cancer frequency among radon-exposed miners. The NCRP risk reduction with time since exposure is supported by the followup studies of underground miners (National Research Council 1988).

The International Commission on Radiological Protection (ICRP) model is a constant relative-risk projection model based on the lung-cancer incidence data from the uranium miner cohort studies (United States, Canada, and Czechoslovakia) and on information from the atomic-bomb survivors (ICRP 1987b). The model assumes that the excess risk of lung cancer in miners associated with a given radon exposure is constant with age and over time after the end of exposures. ICRP made three modifications to the radon relative risk coefficients from the miner data to reflect presumed differences in residential indoor exposures. First, because of potential cocarcinogenic influences that might be present in the mines but not indoors (such as exposures to diesel fumes, dust, and other forms of radiation), ICRP assumed that the risk coefficients for residential indoor exposures would be 80% of those for mine exposures. Second, because of potential differences in breathing rate and the unattached fraction between residential and mine exposure conditions, ICRP assumed that the observed dose of alpha radiation per unit of cumulative radon exposure for the general population is only 80% of that for miners. Third, on the basis of findings from studies of the atomic-bomb survivors, ICRP assumed a risk coefficient for exposure of people under 20 y old that was 3 times the risk coefficient for people 20 or older; this had the effect of increasing the overall lifetime risk by about 40%. The latest recommendations by ICRP (1993b) retain the multiplicative risk-projection model as in previous publications (ICRP 1987b).

The BEIR IV committee model (National Research Council 1988) is a relative-risk projection model based on reanalyses of cohort studies of underground miners (US and Canadian uranium miners and Swedish iron miners). The model assumes that the rate of excess lung cancer due to radon exposures increases with age-specific baseline lung-cancer mortality. The BEIR IV modified relative-risk model is somewhat different from the ICRP 50 model (ICRP 1987b) in the added assumptions about the effects of time since exposure and attained age. It incorporates the BEIR IV finding that excess relative risk in the miners decreased with time since exposure and attained age. Direct evidence on the sensitivity of children to radon is sparse. The BEIR IV committee did not find an effect of age at first exposure after controlling for other correlates with age (National Research Council 1988). That is consistent with the publication of the BEIR V report, which found no evidence of dependence of lung-cancer risk on age at exposure for external radiation (National Research Council 1990). The effect of any higher relative risk in the period soon after exposure of children would probably be offset by the decrease in excess relative risk with time.

Although each of the above three models incorporates risk coefficients derived from the studies of miners, the biologic assumptions underlying the models differ. The different features of these risk-projection models are summarized in table 8.1 (National Research Council 1991). The NCRP model (NCRP 1984c) assumes additivity of the risks posed by radon progeny and the background risk of lung cancer and a time-dependent decline in risk after exposure. In contrast, the ICRP model (ICRP 1987b) assumes that the background rate is multiplied by the additional risk associated with radon progeny. The model developed by the National Research Council (National Research Council 1988) is also multiplicative, but it incorporates a time-dependent decline in risk after exposure. With regard to the relationship between exposure in the mining environment and exposure in the home environment, the three models make different assumptions. The BEIR IV model makes no adjustment, whereas the ICRP model reduces the risk by 20% for adults in the general population, and the NCRP model increases the risk by 40% for the residential exposures, because of a higher calculated unattached fraction. In addition, the ICRP model increases risk for exposures before the age of 20 y, and the NCRP model assumes that risk commences at the age of 40 y. In the BEIR IV model, risk varies with attained age. With regard to smoking, the NCRP model is additive, whereas the other two models are multiplicative (National Research Council 1991).

Table 8.1. Comparison of principal risk-projection models for radon and lung-cancer.

Table 8.1

Comparison of principal risk-projection models for radon and lung-cancer.

The estimate of risk based on chronic occupational exposure to radon in the BEIR IV report (National Research Council 1988), given as a lifetime fatality coefficient, is 3.5 × 10 −4 per WLM for a US population. The corresponding ICRP value is 3 × 10−4 per WLM based on a “reference” population with somewhat lower baseline cancer mortality. EPA's estimates of lung-cancer risk posed by radon exposure at 150 Bq/m3 (4 pCi/L) —1.6 × 10−3 for never-smokers and 3 × 10−2 for smokers—are based on the report of the National Research Council (1988) and an adjustment recommended by the National Research Council (1991). EPA has made two adjustments to the BEIR IV model in estimating radon risks. In the first, age-specific baseline lung-cancer mortality was adjusted by eliminating projected deaths due to an average background radon exposure of 0.24 WLM per year, reducing the lifetime risk estimates by about 10%. The second was based on differences in dose to the bronchial epithelium per unit of radon-progeny exposure in mines and homes due to a number of physical and biologic factors that are expected to differ in the two environments. Among the factors considered in the 1991 National Research Council report are age, sex, aerosol size distribution, unattached fraction of radon progeny, breathing rate and route (oral vs. nasal), pattern and efficiency of deposition of radon progeny, solubility of radon progeny in mucus, and growth of aerosols in the respiratory tract. This comparison of exposure-dose relations in the mining and home environments indicated that the dose per unit of exposure to radon progeny is about 30% lower in the home environment. Therefore, in calculating the risks associated with residential exposures, EPA multiplied the risk coefficient in the BEIR IV model by a factor of 0.7 (EPA 1992c).

The assumption underlying the EPA estimates of radon risk by smoking category is that radon risk varies in proportion to smoking risk (radon and smoking act multiplicatively in causing lung cancer). The data source used by EPA for the prevalence of and relative risks associated with smoking was the surgeon general's report (DHHS 1989). The calculated lung-cancer death rates in each smoking category in the general population were compared with the average lung-cancer death rate for the general population to obtain radon risk multipliers. For example, the number of lung-cancer deaths per 100,000 current smokers in the general population (males and females combined) is 10,329. That is 2.33 times the 4,433 lung-cancer deaths expected in the general population, averaged over all smoking categories (EPA 1992c). From the presumed multiplicative interaction between radon and smoking, the radon risk among current smokers also would be about 2.33 times the radon risk for the general population. The risk multipliers were used in conjunction with a standard life-table analysis based on 1980 vital statistics and the EPA-adjusted BEIR IV relative-risk model to estimate the lung-cancer risks. As discussed previously, the risk coefficients used in the BEIR IV risk model were adjusted by a factor of 0.7 to correct for an estimated lower bronchial radon dose per WLM in homes than in mines (National Research Council 1991). The lung-cancer baseline risk was also adjusted for an annual background radon exposure of 0.24 WLM. Table 8.2 shows the risks for never-smokers and current smokers, with the risks for the general population, for selected radon exposures (EPA 1992c). The lung-cancer risk to current smokers associated with exposure to radon progeny is substantially greater than the radon risk to never-smokers.

Table 8.2. Estmated lifetime lung-cancer risk for never-smokers, current smokers, and the general population .

Table 8.2

Estmated lifetime lung-cancer risk for never-smokers, current smokers, and the general population .

The analysis of the effects of smoking on Rn risk is subject to uncertainty about the nature of the interaction (multiplicative or submultiplicative), the variation in the relative risk associated with smoking by age and gender, the changes in age-specific relative risks for smokers as smoking habits and types of cigarettes change over time, and the influence of environmental and passive cigarette smoke on Rn risk (EPA 1992c).

More quantitative information on lung-cancer risks posed by exposure to radon progeny was provided by a joint analysis of data from 11 studies of underground miners (Lubin and others 1994). The authors examined over 2,700 lung-cancer deaths that occurred among 68,000 miners. The analyses confirm the linear relationship between cumulative exposure to radon progeny and lung-cancer risk and a decrease in excess relative risk (ERR) per WLM with attained age, with time since exposure, and with time after cessation of exposure. Among miners first exposed to radon progeny under the ages of 10-20 y, the ERR per WLM was not related to age at first exposure. The report also noted that for equal total exposure, exposures of long duration and low rate (typical of exposures in homes) were more harmful than exposures of short duration and high rate. Among cohorts with tobacco-use information, the slope of the radon exposure-response function for never-smokers was 3 times that for smokers, indicating a much greater risk for never-smokers relative to their background risk of lung cancer from all causes. Assuming that the miner-based findings apply to residential radon exposure, the study estimated that about 9% of all lung-cancer deaths among residents of single-family dwellings in the United States could be attributable to indoor radon exposure. The estimates are similar to estimates based on the BEIR IV risk model. On the basis of the relative differences in ERR per WLM for smokers and never-smokers, it was estimated that indoor radon-progeny exposure could be responsible for 10-12% of the lung-cancer deaths among smokers and 28-31% of the lung-cancer deaths among never-smokers. For the roughly 15,000 estimated lung-cancer deaths in the United States in 1993 that might be attributable to indoor radon-progeny exposure, those percentages translate to about 10,000 lung-cancer deaths among smokers and 5,000 among never-smokers (Lubin and others 1994).

Several studies have aimed at detecting the correlation between the incidence of lung cancer and exposure to radon in dwellings. The results are mixed. A study of radon and lung cancer in women, with 480 lung-cancer cases and 442 controls, has reported a statistically significant trend with increasing residential radon concentration after adjusting for smoking and age (Schoenberg and others 1990). However, another study showed no statistically significant association between radon exposure in homes and lung-cancer risk (Blot and others 1990). A study of indoor radon and lung cancer in Swedish women, with 210 lung-cancer cases and an equal number of controls, reported increasing trends of lung-cancer risk with radon exposures exceeding 150 Bq/m3 (Pershagen and others 1992).

Recently, Lubin and Boice (1997) provided additional information on the risk of lung cancer associated with indoor radon. They conducted a metaanalysis of eight residential case-control studies that included at least 200 case subjects each and that use long-term indoor radon measurements. The analysis included a total of 4,263 lung-cancer cases and 6,612 control subjects. From the published results of each study, relative-risk (RR) estimates for various categories of radon concentration were obtained, and weighted linear-regression analyses were performed. The combined trend in RR was significantly different from zero, and an estimated RR of 1.14 (95% CI, 1.0-1.3) at 150 Bq/m3 was found. The exposure-response trend was similar to model-based extrapolation found. The exposure-response trend was similar to model-based extrapolation from miners and to RRs computed directly from miners with low cumulative exposures.

A summary of lifetime risk estimates of lung-cancer mortality associated with chronic exposure to radon progeny estimated by various organizations is provided in table 8.3.

Table 8.3. Comparison of lifetime-risk coefficient estimates associated with chronic exposure to radon progeny.

Table 8.3

Comparison of lifetime-risk coefficient estimates associated with chronic exposure to radon progeny.


National Council on Radiation Protection and Measurements (NCRP)

One of the earliest recommendations for domestic radon exposure in the United States was developed by NCRP on the basis of available data on lung-cancer risk (NCRP 1984c). The NCRP recommendation states that an excess risk of death from lung cancer of 2% (a doubling of the average background risk of lung cancer) or more over a lifetime for individuals exposed to enhanced levels of radon decay products should be avoided. The NCRP recommendation was based on evaluation of the lung-cancer risk and the avoidance of an unacceptable exposure and thus risk. The recommendation may be considered as an “upper bound based on maximum tolerable risk.” The excess risk of 2% corresponds to an annual exposure of 2 WLM (equivalent to 8-10 pCi/L if 0.4 or 0.5 is used for the equilibrium ratio) and is the recommended NCRP remedial action level for radon exposure. It is about 10 times the average background exposure of 0.2 WLM assumed for the US population. NCRP recommended that exposure above the remedial action level be reduced using appropriate actions. It also stated that exposures just below the remedial action level might not be acceptable to some individuals, who could of course try to reduce their exposures further.

The assumption by NCRP of an excess risk of death from lung cancer of 2% posed by a lifetime exposure of 2 WLM was based on the available underground-miner epidemiologic data at the time and on the attributable-risk projection model (NCRP 1984b). The NCRP estimate is less than the risk estimated by more-recent projection models, such as those of BEIR IV (National Research Council 1988) and ICRP 50 (1987b). The most recent recommendations of NCRP (1993a) retain the same action level for indoor radon as previously recommended by NCRP (2 WLM) on the basis of an excess lifetime risk of no more than 10 times the risk associated with the average annual background levels found in homes and consideration of the feasibility of remediation.

US Environmental Protection Agency

The discovery of extremely high levels of indoor radon in the northeastern United States in 1984 created a major public-health concern (Oge 1992). As a result, EPA has initiated activities to increase the public's awareness of radon risk and understanding of the options for reducing exposures. EPA used the data and knowledge available at the time to begin immediately reducing the risks posed by elevated radon levels. EPA adopted a nonregulatory approach to the problem and recommended a primary action level of 150 Bq/m3 (4 pCi/L) as the point above which mitigation is always advised to reduce radon in existing buildings (EPA and DHHS 1986).

The 1986 EPA and DHHS recommendation stems from the guideline originally developed for homes built on uranium mill tailings in Grand Junction, CO (Harley 1996). The guideline, published in 1976, was developed by scientists in the Colorado Department of Health, the Surgeon General, Public Health Service, and the Atomic Energy Commission. The guideline was as follows:

  • 0.05 WL in excess of background, remedial action is indicated (this is equivalent to -10-12 pCi/L, depending on whether 0.4 or 0.5 is chosen for the equilibrium ratio of the radon decay products).
  • From 0.01 to 0.05 WL in excess of background, remedial action may be suggested.
  • At less than 0.01 WL in excess of background, no action is indicated.

In Colorado, the Department of Energy (DOE) was attempting to remediate homes built on tailings to the guideline of “no action necessary” levels of less than 0.01 WL above background. The background level of indoor radon decay products is about 0.01 WL, so DOE was attempting to remediate to a gross value (contamination plus background) of 0.02 WL. The EPA radon action level of 4 pCi/L (equivalent to 0.02 WL at 50% equilibrium of the radon decay products) is consistent with the DOE undertakings to remediate to the “no action necessary” level of the original guidelines (Harley 1996).

In its revised guidance, EPA has maintained the same 4-pCi/L action level for indoor radon (EPA and DHHS 1994). EPA sought to balance several factors, including the findings of its technical analysis of risk, cost-effectiveness, and the practical limitation of radon testing accuracy and mitigation technology. It examined five action levels (2, 3, 4, 8, and 20 pCi/L) for the guideline. Higher action levels did not reduce the population risk posed by exposure to radon nearly as much as did lower action levels. The agency has focused its attention on action levels of 150 Bq/m3 (4 pCi/L) or lower. An action level of 4 pCi/L was determined to be incrementally cost-effective (EPA 1992c). For example, the average cost per life saved by using this action level is about $700,000—well within the range of the costs per life saved by other government programs and regulations, such as highway safety, air-transportation safety, and occupational safety. Furthermore, EPA believes that the 150-Bq/m3 (4-pCi/L) action level is technologically achievable in the vast majority of homes. The current guideline also recommends that mitigation be considered for indoor radon concentrations in the range of 75-150 Bq/m3 (2-4 pCi/L), provided that concentrations can be reduced to less than 75-150 Bq/m3. EPA has emphasized several priorities regarding indoor radon: target the highest-risk areas first, promote radon-resistant new construction, support testing and mitigation in connection with real-estate transactions, use information and motivation programs to promote public awareness, and develop a coordinated research plan for radon-related issues.

As previously discussed, EPA has relied primarily on relative-risk projection models to estimate radon risks to the public (National Research Council 1988). EPA's Science Advisory Board (SAB) did not recommend use of the NCRP Report 78 model (NCRP 1984b), which is an absolute-risk model. Absolute-risk models have been described as less appropriate for the estimation of lifetime radon risk because they do not assume the temporal correlation with the baseline lung-cancer rate as indicated by available data (ICRP 1987b). Moreover, the NCRP model presumes that the effects of radon and cigarette smoking are additive, contrary to epidemiologic evidence of a near-multiplicative relationship (EPA 1992c). The SAB also recommended, on the basis of two pieces of recent information, that the agency use only the BEIR IV model and discontinue use of the ICRP 50 model. The first was evidence from epidemiologic studies of a decrease in lung-cancer risk with time since exposure, which had been incorporated into the BEIR IV model but not the ICRP 50 model. The second was the publication of the BEIR V report (National Research Council 1990), which found no evidence of dependence of lung cancer on age at exposure for external radiation.

For exposure at the 4 pCi/L level, the EPA's estimated lifetime risks of fatal lung cancer are 1.6 × 10 −3 for never-smokers and 3 × 10−2 for smokers (EPA and DHHS 1994). EPA's risk estimates at 4 pCi/L are in line with the recommendations of ICRP (1991; 1987b) and BEIR IV (National Research Council 1988). Using the adjusted BEIR IV risk estimates with results from the National Residential Radon Survey (EPA 1992b), EPA calculated about 14,000 lung-cancer deaths per year caused by indoor radon exposure of the US population, with an uncertainty range of 7,000-30,000.

The following sources of uncertainty in the estimate of lifetime risk were addressed by EPA: statistical variability in the miner data, projection of risk beyond the period of epidemiologic followup (projection of risk over time), age dependence of risk, extrapolation from mines to homes, the influence of mine exposures other than to radon, the exposure-rate effect, extrapolation to females, and the relationship between radon risk and smoking risk (EPA 1992c).

International Commission on Radiological Protection

ICRP is one of the principal sources of guidance for radiation-protection policies for most international bodies. Its recommendations for protection of the public from radiation distinguish two circumstances of exposure: one in which human activities introduce new sources or modes of exposure and thus increase the overall exposure, and one in which they decrease the exposure to existing sources. The first it calls practices and the second, interventions (ICRP 1991). Reducing exposure due to pre-existing natural sources, such as indoor radon in existing structures, is clearly an intervention. The system of intervention is based on two principles. The first, justification of intervention, requires that an intervention itself do more good than harm; the reduction in the radiation detriment should be enough to justify the harm, including the cost of intervention. Under the second principle, optimization of intervention, the form, scale, and duration of the intervention are chosen to obtain the maximum net benefit.

ICRP (1991) recognized the complex problems involved in controlling exposure to indoor radon, which is by far the largest source of average human exposure to natural background radiation. It recommended that the choice of an action level for radon should depend not only on the magnitude of risk to individuals, but also on the likely scale of action required and its economic implications for communities and individual homeowners. ICRP also stated that the particular national action level chosen could be one that defines a sizable, but not unmanageable, number of houses in need of remedial work. Intervention is to be applied to reduce the risk to those most highly exposed and not primarily to reduce the collective dose to the population.

Because of its widespread occurrence and relatively high concentrations, radon has been treated separately by ICRP. ICRP recommends that countries carry out radon surveys to identify radon-prone areas, defined as those in which more than 1% of homes contain radon at more than 10 times the national average. That is a way to identify a manageable number of homes with the greatest risk to occupants (ICRP 1993b).

The principle of justification is used to set action levels at which intervention would almost always be justified to reduce radon exposure. ICRP considered that simple countermeasures would be virtually certain to be justified to avert an effective dose of 10 mSv in a year (the upper bound of the maximum tolerable dose to individuals).

The principle of optimization will lead to a lower action level than 10 mSv, but not lower by as much as a factor of 10, because that would often correspond to the national average radon exposure. Using an effective dose per unit exposure of 4 mSv per WLM (ICRP 1993b) for exposure in homes, ICRP recommended a radon concentration of 600 Bq/m3 as that at which action is almost certain to be justified, and it expected optimization to suggest an action level no lower than 200 Bq/m3. It recommends that national authorities set an action level of about that and advise residents whose homes are at higher levels that they should initiate remedial measures. It recommends that new homes be designed to avoid the problem of high radon. ICRP no longer sets another, lower target level for new homes as it did previously (ICRP 1984).

ICRP and EPA use similar risk-management approaches to address control of the hazard posed by indoor radon (Overy and Richardson 1995). They agree that risks posed by domestic radon should not be treated in the same manner as other risks, because the social and economic effects are greater and more complex.

International Atomic Energy Agency

The International Atomic Energy Agency (IAEA) basic safety standards for protection against ionizing radiation have been recently published (IAEA 1996a). The guidelines specific to indoor radon are based on the ICRP recommendations (ICRP 1993b; 1991). The main IAEA recommendation is that the optimized action level related to chronic exposure involving radon in dwellings should fall within a yearly average 222Rn concentration of 200-600 Bq/m3 in air. For the optimized action levels, account should be taken of the benefits and costs assessed in the remedial action plan.

Commission of the European Communities

The current indoor-radon control policy recommendations of the Commission of the European Communities (CEC) were published in the Official Journal of the European Communities (EC) in 1990 (EC 1990). The recommendations are not legally binding, but constitute, within the EC, the reference framework for the initiation of policies at the national level. The main CEC recommendations, adopted mostly from ICRP (1984), are:

  • To set a reference level of 400 Bq/m3, above which consideration should be given to reducing radon concentrations in existing homes, and a design level of 200 Bq/m3 for all new dwellings.
  • To use annual average radon concentrations as a basis for radiologic protection decisions and to develop criteria for identifying regions, sites, and building characteristics likely to be associated with high indoor radon concentrations.
  • To have national authorities provide information to the public on radon exposure, risk, and available remedial measures.


Various countries' and organizations' current recommendations for action levels for existing houses and for upper limits (bounds) in new buildings are summarized in table 8.4. Additional comments and information are also provided in the tabular summary.

Table 8.4. Summary of national and international recommendations for indoor radon in dwellings.

Table 8.4

Summary of national and international recommendations for indoor radon in dwellings.

Values for existing dwellings are mostly of an advisory nature. In Sweden and Switzerland, the levels are legally enforced, and both apply “recommended” action levels, lower than the regulatory limits, above which remediation is advised. Although most guidances are not enforced standards for limiting indoor radon exposures of the public, they are widely used as de facto standards in the real-estate and insurance industries. For example, in the United States, lending institutions often require radon concentrations (based on short-term measurements) less than the EPA action level of 150 Bq/m3 (4 pCi/L) as a condition for financing home purchases.

Of the 15 member states of the European Union, only Austria and Finland have adopted the values proposed in EC (1990). Belgium, Germany, Ireland, Luxembourg, Sweden, and the United Kingdom have adopted somewhat different values. In the Netherlands, an approach based on limiting individual risk has been adopted. Peak radon concentrations in the Netherlands are relatively low in comparison with those of other countries. The 20-Bq/m3 reference level in the Netherlands reflects the low indoor radon concentrations generally found and provides lifetime risks of 10−4 or less.

Australia, Austria, Germany, Ireland, Norway, Switzerland, and the United Kingdom have set their action levels on the basis of recommendations of ICRP (1993b) and other considerations regarding cost, risk, and feasibility. In Canada, cost-benefit analysis was used as a basis for a radon reference level of 800 Bq/m3. A similar approach was initially adopted in Sweden, which has recently reduced its reference level after the publication of ICRP 65 (1993b). Similarly, the reference level in Luxembourg was reduced from 250 to 150 Bq/m3 in 1992. The value of 150 Bq/m3 recommended by the US EPA is based as much on technologic limitations and cost-benefit analyses as on health risks (EPA 1992c).

Although national and international guidance for radon in dwellings varies, most values have similar scientific and technical bases and are within only a factor of about 2 from each other. Most differences are related to policies and risk-management decisions by the various bodies that develop radon guidance.


Workplaces are defined as ordinary places of work, such as offices, schools, stores, theaters, libraries, and hospitals (Clarke 1995). For the purposes of this discussion, workplaces do not include nuclear fuel-cycle facilities. Radon is present in all workplaces. In some cases, such as uranium mines, exposure to radon is already subject to occupational control. Because of the lower occupancy rate and associated lower accumulated exposure, radon in ordinary workplaces is widely ignored or of secondary concern compared with radon in dwellings. Some countries, however, have placed strong emphasis on radon measurement and remediation in schools (for example, Sweden, the United Kingdom, and the United States).

Exposure to radon in schools is often viewed separately from that in other workplaces, to emphasize protection of young people and because large numbers of people are potentially exposed. Although occupancy rates of schools are lower than those of dwellings, some countries have adopted the same action levels for both. In Switzerland, the reference level of 400 Bq/m3 for schools is lower than the legally enforced 1,000 Bq/m3 limit for dwellings but is equal to the level above which remediation of homes is recommended. In the United Kingdom, Finland, Switzerland, and Norway, the values for schools are legally enforced. The US EPA advisory reference level of 150 Bq/m3 for schools is the same as for dwellings.

There is considerable diversity in the approach to radon in workplaces, and the range of action levels is wider than that for dwellings (table 8.5). Levels range from a target value of 20 Bq/m3 in the Netherlands to a statutory limit of 3,000 Bq/m3 in Switzerland.

Table 8.5. National and international reference (action) levels for radon in workplaces excluding those linked to the nuclear fuel cycle .

Table 8.5

National and international reference (action) levels for radon in workplaces excluding those linked to the nuclear fuel cycle .

ICRP (1993b) recommends that intervention levels for exposure to radon in homes be carried over to workplaces for exposure to radon. The level at which intervention in the workplace is almost certainly justified is the same as in homes—10 mSv in a year. Because of the different occupancy factor—2,000 h at work and 7,000 h at home each year—and an effective dose per unit exposure of 5 mSv per WLM (ICRP 1993b), one arrives at a radon concentration in the workplace of about 1,500 Bq/m3 as the level at which action is almost certainly justified. With optimization, the suggested range within which an action level should be set is 500-1,500 Bq/m3. The IAEA remedial action level for chronic exposure involving radon in workplaces is a yearly average 222Rn concentration of 1,000 Bq/m3 (IAEA 1996a). This guideline appears to be based on the average range of 500-1,500 Bq/m3 recommended by ICRP (1993b).


  • The first recommendations for domestic radon exposure in the United States were developed by NCRP in 1984. The NCRP recommendations were based on evaluation of the lung-cancer risk and the avoidance of an unacceptable risk. A personal avoidance of lifetime lung-cancer risk of 2% was proposed, that is, avoidance of a continuous exposure to 2 WLM (equal to radon concentration of about 10 pCi/L). This lifetime risk was compared with a normal lung-cancer risk in smokers of about 10% and in nonsmokers of about 1%.
  • On the basis of different risk projection models, the BEIR IV Committee and other major radiation protection organizations (ICRP) have estimated that higher lung-cancer risks are associated with indoor radon exposure since the NCRP publication of 1984. Although all three models (NCRP, BEIR IV, and ICRP) incorporate risk coefficients derived from the studies of miners, the biologic assumptions underlying the models differ.
  • EPA's current estimates of lung-cancer risk associated with indoor radon exposures are based on the BEIR IV report and later adjustments. EPA risk estimates related to domestic radon exposure are 1.6 × 10−3 for never-smokers and 3 × 10−2 for smokers at 1 WLM (equal to 4 pCi/L, which is recommended as the remedial action level). EPA guidelines are generally comparable with the recommendations of ICRP.
  • The above approaches for estimating lung-cancer risks from the miner data require numerous adjustments for estimating comparable risks in homes, and those approaches assume that cancer incidence observed in miners at high radon levels can be extrapolated linearly to zero exposure. Other sources of uncertainty include statistical variability in the miner data, the age dependence of risk, extrapolation to females, the relationship between radon risk and smoking risk, and the impact of risk extrapolation of the different levels and types of particles in uranium mines and in homes.
Copyright © 1999 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK230646


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.0M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...