© EDP Sciences 2017

1 Introduction

Many epidemiologic studies prove that indoor radon exposure is responsible for a substantial number of lung cancers in the general population (Darby et al., 2005). After cigarette smoking, radon is the most frequent cause of lung cancer. Regarding radiation protection, the most important isotope of radon is ²²²Rn, and in Austria, the inhalation of indoor radon contributes with about 54% to the annual effective dose of the population (Maringer et al., 2012). Following the ICRP recommendations, current dose conversion factors for the inhalation of radon daughters, which are derived from epidemiologic studies, will be revised. It is likely that the result will be an increase in effective dose per unit exposure of around a factor of two (ICRP, 2009). Dosimetric evaluations of the absorbed dose to basal cells of the bronchial epithelium provide a conversion factor 2.5 times higher than the epidemiologic approach (UNSCEAR, 2008).

Since 2013, a revised Basic Safety Standards Directive (Council Directive 2013/59/Euratom, 2013) lays down basic safety standards for protection against the dangers arising from exposure to ionising radiation. This Directive shall be brought into national law of the member states by 2018. The updated Directive covers the whole range of radiation sources and categories of exposure including occupational, medical, public and environmental exposure. One special focus is on the protection of workers in workplaces with indoor ²²²Rn and in activities processing Naturally Occurring Radioactive Materials (NORM).

Industrial activities can lead to an enhanced exposure to NORM in products, by-products, residues and wastes. There are specific working areas, in which elevated ²²²Rn activity concentrations can occur. Waterworks are part of these areas (BMLFUW, 2008). ²²²Rn can escape from the ground and is dissolved in ground water. In waterworks, it can be outgassing in large amounts due to open water surfaces and high water flow rates.

In order to perform traceable, accurate, standardised measurements, improved measurement practices and procedures are required for NORM industries. The standards for ²²²Rn determination of indoor air are mainly designed for dwellings, in which different conditions regarding the influence on the ²²²Rn activity concentration are predominant. A detailed literature study is the first step to develop improved measurement practices for specific working areas like waterworks. The main objectives of this review are to compare and discuss different studies focussed on ²²²Rn in indoor air of waterworks from the aspect of the presented results as well as the applied methodology of the ²²²Rn measurements. Special emphasis is given to the annual effective dose of the workers. Besides, strategies for waterworks testing are compiled from literature.

2 Overview – radon in waterworks

Generally, the ²²²Rn activity concentrations in water tank facilities and other parts of waterworks, such as the water treatment or fountains, can be much higher than in common domestic buildings (Nagels et al., 2009). The main source of ²²²Rn emanation within a waterworks is the water itself. Though potable water can be a source of ²²²Rn in indoor air of dwellings (Nazaroff et al., 1987), it is much more important in waterworks. ²²²Rn can outgas from the water into indoor air of waterworks and can cause a significant contribution to the annual effective dose of the waterworks staff (Tab. 1).

Many factors have an impact on the ²²²Rn activity concentration in the water, such as the pathways of the aquifers. The mineralogical composition of the aquifer and the permeability of soil affect the ²²²Rn activity concentration in water (Loomis, 1987; Trautmannsheimer et al., 2003). Rock types differ in the uranium concentration and therefore in the radium concentration, which influences the ²²²Rn activity concentration in soil gas and ground water. Relevant parameters for the exhalation rate of ²²²Rn in water into the indoor air of waterworks are the water flow rate and the type of water input (Hingmann et al., 2001). Besides the exhalation rate, the ventilation heavily influences the ²²²Rn activity concentration in indoor air of waterworks. For that reasons an assessment of ²²²Rn in air on basis of ²²²Rn in water is not reasonable (Stietka et al., 2014).

The facilities of a waterworks differ in terms of size and structure. High ²²²Rn activity concentrations tend to occur at those parts of a waterworks with high water flow rates and a small air volumes, e.g. the tapping of a spring. The farther a facility is located from the spring the more ²²²Rn could already have been outgassed or decayed. Like in dwellings, the ²²²Rn activity concentrations in waterworks are subject to diurnal and seasonal variations, also tending to higher concentrations in the cold season (Stietka et al., 2013).

Overall, there are rare publications in the field of radon in waterworks, therefore some statements, which are applicable to radon in waterworks, are related to radon in indoor air in general.

2.1 Measurement aspects

Generally, for the monitoring of radon, both active and passive techniques exist. Passive methods require a preparation and an evaluation process, whereas the results are instantaneously available with active devices. In waterworks, the latter ones are mainly used for short-term measurements serving as an overview and for detailed investigations of individual sites. According to the WHO (2009), short-term refers to measurements that average radon concentrations over days or weeks, while long-term measurements generally cover a season or more (several months to one year). Due to temporal variations of the ²²²Rn activity concentration in waterworks, short-term measurements are not suitable as basis for dose assessments (Stietka et al., 2013). To provide a realistic exposure scenario, long-term measurements at various measurement points are necessary for the most waterworks. For such simultaneous long-term measurements at numerous sites, passive devices offer advantages because of their moderate costs.

For long-term indoor radon measurements in general, solid-state nuclear track detectors or SSNTDs are a reliable, sufficient and inexpensive passive measurement technique (Ahn and Lee, 2005; Banjanac et al., 2006). The alpha particles leave nuclear tracks at the detection films of the SSTNDs, which can be evaluated after etching, usually with NaOH. The most popular of these detector types are based on CR-39 polycarbonate films, because of its good sensitivity, stability against various environmental factors, and high degree of optical clarity. CR-39 films inside a closed detector cup are not as sensitive to humidity and dust effects as other methods (Nikezic and Yu, 2000). With closed detector cups, ²²²Rn diffuses through the encapsulation into the measuring chamber, whereas the ²²²Rn decay products, dust and moisture are blocked. This is an important factor for ²²²Rn measurements in waterworks. On the other hand, open cupped detectors, frequently on basis of LR115 nitrate cellulose films, are sensitive to the alpha particles of ²²²Rn and ²²⁰Rn as well as their progenies. ²²⁰Rn (the so-called thoron) with a half-life of 56 s can rarely occur in waterworks, see chapter 2.2. Special double detectors with a closed and an open cup allow for a ²²²Rn/²²⁰Rn discrimination (Tokonami et al., 2005).

In contrast to passive methods, active devices require a battery pack or power supply, which is not available in every part of a waterworks. The most popular active techniques applied for monitoring are ionization chambers and solid-state silicon detectors. The latter ones have small-dimension diffusion chambers. Therefore, long integration times are essential to reach a statistically stable value, whereas ionization chambers having generally larger measurement chambers can fulfil the requirements for quick overview measurements in waterworks. There also exist active devices for the determination of ²²⁰Rn, ²²⁰Rn progenies as well as ²²²Rn progenies.

Considering their operating humidity, some devices or measurement methods are not suitable for the application in waterworks in general.

2.2 Dose aspects

Though the inhalation of short-lived ²²²Rn progeny and not ²²²Rn itself is dose relevant, determining the ²²²Rn activity concentration is a common, more convenient practice. For dose assessment on basis of the ²²²Rn activity concentration, the consideration of the equilibrium factor is essential. The equilibrium factor characterizes the disequilibrium between the mixture of the short-lived progeny and their parent nuclide, radon, in air in terms of potential alpha energy (ICRP, 1993). UNSCEAR (1988) and the ICRP (1993) have adopted a typical worldwide equilibrium factor of 0.4 for ²²²Rn in indoor air. In situations where the equilibrium factor differs significantly from the usually assumed value of 0.4, a direct determination of the ²²²Rn decay products can be performed (WHO, 2009). Within surveys in Upper Austria, Bavaria, Hesse and Baden-Wuerttemberg mean values of 0.52, 0.5, 0.37 and 0.55 for the equilibrium factor of indoor air of waterworks were found (AGES, 2006). Out of Table 1 and the values above, it can be seen that there is no clear indication to use another value for waterworks than the generally recommended indoor value of 0.4 (UNSCEAR, 1988; ICRP, 1993). There is only a very limited amount of data published on the equilibrium factor determined in the indoor air of waterworks.

Besides the equilibrium factor, other influence factors on the dose conversion such as the distribution of aerosol diameters and the unattached fraction (Yu and Nikezic, 2012) may differ from those in dwellings. No publication was found, which addresses these parameters for waterworks.

The short-lived ²²⁰Rn can occur in some parts of waterworks. Dosimetric evaluations in UNSCEAR (2008) give dose coefficients in the range of 6–15 nSv (Bq h m⁻³)⁻¹ for ²²²Rn and 40 nSv (Bq h m⁻³)⁻¹ for ²²⁰Rn (for equilibrium equivalent concentrations). Within a Bavarian survey of Trautmannsheimer (2002), ²²⁰Rn, exhaling from the water, was found within a water chamber of a waterworks. Due to its short half-life, the activity concentration of ²²⁰Rn was much lower in the other parts of the same building. In this study, it was found out that the ²²⁰Rn caused dose is no higher than 10% of the dose of the ²²²Rn decay products. In Stietka et al. (2014), ²²⁰Rn was detected in 3 out of 47 measurements. In an Upper Austrian study (AGES, 2006), where ²²⁰Rn was investigated in five waterworks, no ²²⁰Rn was found at all.

Table 1 summarizes the results of the literature survey and gives an overview on studies with the focus on the determination of radon in waterworks. For each study only the most comprehensive or relevant publication was mentioned in Table 1, so the shown results rely on different datasets. Also very specific studies or small ones, based on measurements in a small number of participating waterworks (e.g. Aakerblom et al., 2002), were excluded because of lack of representativeness. As far as possible, the results were presented in their original form to avoid a bias of the outcomes. Besides the values for the ²²²Rn activity concentration in indoor air of waterworks, other results and relevant information are presented, which are the used methodology for the determination of ²²²Rn in air, the activity concentrations of ²²²Rn in water, the equilibrium factor F applied for dose calculation, and the annual effective dose of the staff. The maximum values reported in these studies were more than 50000 Bq m⁻³ (Marinko et al., 2004) for ²²²Rn in air, 1200 Bq I⁻¹ (Trautmannsheimer et al., 2003) for ²²²Rn in water and 56 mSv y⁻¹ (Ringer et al., 2008) for the annual effective dose of the waterworks staff.

3 Strategies for waterworks testing

3.1 Methodology

As it can be seen in Table 1, the results of the ²²²Rn measurements of indoor air in waterworks mostly serve as basis for dose assessment. The calculation of the effective dose can be done either on basis of the personal exposure (e.g. Trautmannsheimer et al., 2003; Nagels et al., 2009) or on basis of the local activity concentrations and the residence times (e.g. Fisher et al., 1996; Vaupotic, 2002; Marinko et al., 2004; Chmielewska et al., 2014; Stietka et al., 2014). Regarding possible, subsequent dose reduction measures, a determination of the local activity concentration is needed. In case of future changes in the residence times at specific locations, this method is also favourable. The effective dose can simply be recalculated out of the representative activity concentrations and no new measurements are required.

When collecting the data about the residence times of the affected employees, special attention should be paid to the exposure during cleaning works or repair works at facilities with open water surfaces. During these activities, a different ²²²Rn concentration can occur (Hingmann et al., 2001).

At facilities with a water treatment, there can be large diurnal variations depending on its operation mode, e.g. ²²²Rn can outgas in large amounts during the backwash process of a filter (Fisher et al., 1996). In some of these cases, it is expedient to consider these diurnal variations of the activity concentration with separate measurements. These should be conducted with an active measurement device with a sufficient time resolution (Hingmann et al., 2001). Except for some above mentioned selected measurement points in the studies of Fisher et al. (1996) and Hingmann et al. (2001) all annual effective doses reported in Table 1 are based on passive measurements without time resolution. Both studies show, that the average ²²²Rn concentrations may significantly differ from those during the residence times of the staff. In these cases the average ²²²Rn concentrations do not allow to determine the correct annual effective doses.

For water purification buildings, the radon concentrations are much higher in case of the presence of the process controller than the average concentration (Trautmannsheimer et al., 2003). Considering this problem and the insufficient availability of the exact time spent in the water supply facilities, Trautmannsheimer et al. (2003) sees personal detectors as the most reliable method to evaluate the radon exposure of the staff. In his study, he conducted ∼3000 ²²²Rn long-term measurements. Usually the lack of sufficient equipment does not allow the general use of active ²²²Rn measurement devices with time resolution.

On the other hand, Hingmann et al. (2001) reported problems with the handling and regular wearing of the personal detectors. In addition, personal ²²²Rn detectors do not allow identifying the locations with high activity concentrations to set dose reduction measures. Nagels et al. (2009) pointed out, that effective doses for workers of external companies could be calculated on basis of the local ²²²Rn concentrations.

3.2 Equipment and determination of the radon activity concentration

During visits at waterworks, it could be helpful to start with active measurements (as conducted by Fisher et al., 1996; Hingmann et al., 2001; Vaupotic, 2002; Ringer et al., 2008; Nagels et al., 2009; Stietka et al., 2014). Such short-term measurements − preferably conducted with a high-efficiency active device with a display operated in flow mode (e.g. an ionization chamber) − can give a quick overview about the situation. Works in the course of the measurements can reduce the present activity concentration by ventilation effects or low water flow rates. Keeping the ventilation and operation conditions in the usual state during the measurements can prevent this and ensures representative results. This is especially important for short-term measurements.

Out of the active short-term measurements, one can decide to shorten the measurement time of the passive long-term measurements or to split it up to prevent overexposure, in case of high ²²²Rn activity concentrations. Overexposure causes high track densities on the detector film, which can make the evaluation impossible. Further, the results of the active devices can help to speed up for prompt remediation measures. In order to obtain more reliable values for rapid technical remediation measures, it can be helpful to conduct passive short-term measurements.

For ²²²Rn measurements in indoor air in general, passive SSNTDs have a well-documented history of application (Tommasino, 2001). For the determination of the annual effective dose caused by ²²²Rn in waterworks, the literature listed in Table 1 shows, that passive long-term measurements with SSNTDs are a reasonable choice. The closer the measurement time is to one year, the more representative it will be of annual average radon levels.

As can be seen in Table 1, the determination of the local ²²²Rn activity concentration is the mainly used method, whereas in the studies of Nagels et al. (2009) and Trautmannsheimer et al. (2003), the dose calculations were based on the personal ²²²Rn exposures.

Normally, installing one detector in every room where persons stay is adequate for the determination of the local exposure. Within rooms with large air spaces, the radon activity concentration can vary considerably. In this case, several detectors within one room provide a more realistic exposure scenario.

For ²²²Rn in indoor air in general, an optimal height for positioning the detector is the breathing zone (EPA, 1993). A position should be selected, where the detector will not be disturbed during the measurement period or disturbs the ongoing works of the waterworks staff. For waterworks, especially in the field of drinking water production, hygiene is very important issue, which has to be considered e.g. for positioning the measurement devices. In addition, measurements should not be conducted near drafts caused heating or vents (EPA, 1993).

For the proof of effectivity of the set dose reduction measures, it is necessary to repeat the measurements after a remediation process (e.g. Ringer, 2008). In case of high annual effective doses, the repeat of measurement after remediation is enshrined in the national act or ordinance, e.g. BMLFUW (2008).

Additionally to the ²²²Rn measurement devices, a dose rate meter is frequently used as a supplementary instrument in waterworks (Ringer et al., 2008). It can help to identify water filters enriched with ²²⁶Ra, which could be a source of ²²²Rn emanation. Besides, the filter materials as well as the residues from the backwash process can contain several other, natural radionuclides (Fonollosa et al., 2015).

3.3 Evaluation of measurement results

If the measurements are intended to provide a representative annual-average radon concentration, there are some general considerations. These should be given to measurement variability and the predictive value of the results, given the uncertainties that arise from spatial, temporal and instrument variations (WHO, 2009). Especially for waterworks, temporal variations depending on the operation mode of the water treatment have been documented in Fisher et al. (1996) and seasonal variations in Stietka et al. (2013).

Concerning seasonal variations for radon in indoor air in general, depending on the country, there are different circumstances and approaches to handle this problem. In some countries, measurements made in various seasons are adjusted to estimate an annual-average ²²²Rn concentration based on “typical” seasonal variations (Baysson et al., 2003). For measurements of a few months, the WHO (2010) suggests, that a seasonal adjustment factor, if available, may be applied to obtain an estimate of the annual average value. In Austria, these factors have been defined especially for waterworks, which include a correction for the ²²²Rn activity concentration, depending on the season, as well as for its uncertainty depending on the measurement duration (Stietka et al., 2014).

To evaluate the measured ²²²Rn exposures with regard to the effective dose of the waterworks staff, usually an equilibrium factor of 0.4 is applied, as can be seen in Table 1. The applied equilibrium factor of 1.0 in Fisher et al. (1996) was declared as conservative basis for the calculation of the working-level month.

4 Conclusion

This literature review on studies directly associated with ²²²Rn measurements in waterworks, covering more than 800 waterworks in nine regions, shows that ²²²Rn can exhale in large amounts through open water surfaces in this working area and cause a significant dose contribution for the staff. For the ²²²Rn activity concentration in indoor air of waterworks, values higher than 50000 Bq m⁻³ were documented in the literature (Marinko et al., 2004) and annual effective doses as high as 56 mSv y⁻¹ were reported (Ringer et al., 2008).

In this study, measurement strategies for ²²²Rn in waterworks were collected and analysed. The results of ²²²Rn measurements in indoor air of waterworks mostly serve as basis for dose assessment. The assessment of the workers' effective dose can be done either on basis of the personal exposure or on basis of the local activity concentrations and the residence times. The determination of the local exposure allows for the identification of spots with high activity concentrations to set dose reduction measures.

Personal detectors do not require the survey of the residence times at specific locations and therefore avoid problems caused by variations of the ²²²Rn activity concentration. Especially at facilities with a water treatment, large diurnal variations depending on its operation mode were reported. Also during cleaning or repair works, different ²²²Rn concentrations can occur. The average ²²²Rn concentrations may considerably differ from those during the time spent of the staff and do not allow to determine the correct annual effective doses.

For the calculation of the annual effect dose, there is no clear indication in literature to use another equilibrium factor for waterworks than the generally recommended indoor value of 0.4. Overall, there are rare publications in the field of radon in waterworks. More investigation is needed, especially regarding ²²⁰Rn, and influence factors on the dose conversion, e.g. the equilibrium factor, the distribution of aerosol diameters and the unattached fraction.

Acknowledgements

This work was done within the Joint Research Project IND57 MetroNORM in the framework of the European Metrology Research Programme (EMRP). MetroNORM is aimed to develop improved methods for laboratory and in situ measurements and to design standardised procedures, as well as perform testing in industrial environment. The EMRP is jointly funded by the EMRP participating countries within EURAMET and the European Union.

References

All Tables