Natural radioactivity of residues from groundwater treatment facilities in Finland

The accumulation of naturally occurring radionuclides in solid residues was investigated from groundwater treatment facilities (GTFs) in Finland. Natural radionuclides U-238, Ra-226, Pb-210 and Ra-228 were found in various precipitates, sludges and filters at concentrations exceeding the general clearance level of 1 kBq kg−1 used for solid materials in the European directive 2013/59/Euratom. The accumulation of natural radionuclides in different solid residues was observed even when the activity concentrations in the untreated groundwater were relatively low, and when there was no measurable change in the concentrations between raw and treated groundwater within analytical uncertainties. Based on mass and activity balance considerations this is thought to be due to the large volumes of treated water per year. The exposure of workers to natural radiation from solid residues in the regular use of a groundwater facility was found not to be likely to exceed 0.3 mSv a−1 if the activity concentrations are <10 kBq kg−1 for U-238, Ra-226, Pb-210 and Ra-228. The worker exposure from solid residues is therefore likely to remain below the reference level of 1 mSv a−1, and indoor radon is more of a concern for the radiation protection of workers at GTFs. However, the natural radionuclide content in the different solid residues from groundwater treatment needs to be characterised properly to be able to ensure safety in the final use of the residues with respect to the potential exposure of the public.


Introduction and background
Groundwater in Finland frequently contains radon and its short-lived daughters [1,2], but also uranium and other uranium series radionuclides can be present [2][3][4][5][6].Relatively high levels of radon and uranium have been reported especially in the groundwater from the bedrock [4,7].The source of the uranium-and thorium-series radionuclides in groundwater is the bedrock and soil [8].Geologically, Finland occupies a part of the Fennoscandian shield, and the bedrock contains abundant granitic rocks which have elevated concentrations of uranium in some parts of Finland [9].The uranium content of the overlying mineral soils [10] is correlated with the concentrations in the bedrock.In the European context, the uranium concentration in the soil in Finland is not anomalously high [11], however, and this alone would not explain the groundwater composition.In Finland, groundwater is typically soft bicarbonate water with a mildly acidic to neutral pH [7], and the bicarbonate content in the water is the main factor leading to the solubility of uranium [2,8].In oxidising conditions uranium forms soluble uranyl carbonate complexes which are stable over a comparable range (6)(7)(8) of pH values [8,12] as is typical for groundwater in Finland [7].The radium concentrations in groundwater in Finland are low compared to for example Spain, Sweden, Germany and Estonia [13][14][15][16].
In Finland approximately 65% of the drinking water originates from groundwater [17,18] which is slightly above the European Union (EU) average of 50% [19].The groundwater used for drinking water can be from the soil or bedrock, and it can be either natural or artificial.In Finland the activity concentrations of natural radionuclides in surface waters are very low, while those in groundwater from soil layers are slightly higher, although the concentrations are generally below the regulations for radioactivity in drinking water [20].The highest activity concentrations can be found in groundwater from the bedrock [21][22][23].Most groundwater treatment facilities (GTFs) in Finland exploit groundwater from the soil, and the facilities are in areas where the soil is permeable to water and the formation of groundwater is abundant.In contrast, drilled wells in bedrock are more common as small private household wells rather than the main water source for large water treatment facilities.
Groundwater can be distributed as drinking water without treatment if the quality is sufficient but often the water is treated before distribution.In Finland the raw groundwater is commonly treated to decrease the corrosivity of the water or to remove iron or manganese by filtration [23,24].When the water is treated by a filtration method the natural radionuclides may accumulate to filters, precipitates, or sludges [25].The elevated activity concentrations of natural radionuclides in filters, sludges and other residues may cause exposure to natural radiation, and this should be considered for the radiation protection of workers and the public.
In Finland, the directive 2013/59/Euratom [26] has been implemented with the Radiation Act 859/2018.Production of drinking water in a GTF is listed as an industrial sector which may cause exposure to natural radiation due to the accumulation of naturally occurring radionuclides.The reference levels used for natural radiation in Finland are 1 mSv a −1 for workers and 0.1 mSv a −1 for the public (excluding indoor radon, building materials and background).The reference level for radon concentration at workplaces is 300 Bq m −3 , and for radon exposure of workers 500 000 (Bq h) (m 3 a) −1 .The general exemption and clearance levels of 1 kBq kg −1 for U-238, Th-232 and their progeny, and 10 kBq kg −1 for K-40 apply to solid residues from GTF's (excluding use in building materials).Exposure assessment for workers and the public is required if the activity concentrations of natural radionuclides exceed exemption and clearance levels.
Compared to the reference levels, the main cause of worker exposure to natural radiation at GTF's is indoor radon [27,28].It is also known that granulated activated carbon (GAC) filters can accumulate short lived radon progeny during the operation of a GTF at a level that can be of concern for the radiation exposure of workers [29].There is a fair amount of information about drinking water, indoor radon and GAC filters at GTF's in Finland, but there is little data about the long-lived natural radionuclides from the solid residues formed at GTF's.Internationally, there are studies from e.g.Germany, Spain and Estonia [15,16,25,30,31], but these are from areas where the geology and groundwater chemistry can be very different from Finland.
In this study the accumulation of natural radionuclides in the solid residues of GTFs in Finland was investigated.The radioactivity of different filter, sludge, precipitate, backwash-affected soil, and water samples was analysed from Finnish GTF's to characterise the materials and better understand the processes that could produce residues with radiation risks.Additionally, one aim of the study was to gather information to give better guidance to the GTF operators to help them make more targeted investigations and exposure assessments as required by the legislation [26] for radiation protection.

Materials and methods
Since the focus of this study was on the radionuclide content of solid residues such as filters, sludges and precipitates at GTF's, the investigated sites were purposefully selected to include different types of filtration steps in the treatment processes.Basic information of the studied facilities is given in table 1.Many of the facilities were located on granitoid bedrock areas with granites and granodiorites as the main rock types.The studied GTF's used groundwater filtration with limestone, sand, sand-anthracite and GAC filters.Aeration and UV-disinfection were used by nearly all of the facilities, and alkalisation, ozonation and chloramine were used by some of the facilities.
Sampling and measurements were carried out during 2020.The sampling sites were located with the personnel at each facility.Samples of raw water and treated water (N = 46), backwash water (N = 4), filter materials (N = 13), and backwash precipitates and sludges (N = 7) were collected.Additionally, one precipitate from an aeration tower and one precipitate form inside a raw water pipe were collected.Soil samples of backwash discharge-affected soil (N = 5) and control soil samples (N = 5) were taken if the facility had previously discharged backwash water into the soil.Data for GAC filter samples from which Pb-210 was reported in [29] are present here for the other radionuclides as mean values for each facility.
Gross alpha, gross beta, and radionuclide (U-238, Ra-226, Rn-222) activity concentrations in water samples were determined by accredited methods of liquid scintillation counting (LSC) and inductively coupled plasma mass spectrometry (ICP-MS) at the laboratory of Radiation and Nuclear Safety Authority (STUK), Finland.For solid samples, activity concentrations of Pa-234m, Ra-226, Pb-210, U-235, Ra-228, and Th-228 were determined using gamma-ray spectrometry at the laboratory of STUK, and U-238 was Only the radionuclides found in GAC-filter materials are discussed for these facilities.
calculated from the U-235 measurement results using the natural uranium isotope ratio.Additional details of the sampling, sample treatment and analytical methods are provided in the electronic supplement.

Water samples
The activity concentrations of gross alpha and gross beta in both raw groundwater and treated groundwater were at the most 0.5 Bq l −1 in all the 12 facilities from which water samples were taken (table 2).Most of the gross alpha is from uranium, as U-238 concentrations (0.01-0.24Bq l −1 ) are higher than Ra-226 (<MDC to 0.1 Bq l −1 ) (table 2, figure 1).The higher activity concentration of uranium compared to radium is a typical feature in the groundwater in Finland [4].Overall, the concentrations in the water samples are lower than what is found in groundwater which is specifically treated for radionuclides [14,16].The natural radionuclides in backwash water were also analysed at four facilities and the activity concentrations were similar to raw groundwater within the measurement uncertainties (table 2).The activity concentrations in water samples show mostly no change before and after the treatment processes within the measurement uncertainties, except for Rn-222 (table 2, figure 1).For Rn-222, the activity concentrations in raw groundwater were mainly distributed around 50-100 Bq l −1 , while in treated groundwater the concentrations were lower from 1 to 50 Bq l −1 (table 2).This is showing the expected loss of gaseous radon during water treatment (figure 1(d)).All of the measured Rn-222 concentrations are below the parametric value of 300 Bq l −1 for radon in drinking water in Finland.According to the EU directive 2013/51/Euratom the parametric value for radon is to be set by the member states somewhere between 100 Bq l −1 and 1000 Bq l −1 , and remedial action is always justified at concentrations exceeding 1000 Bq l −1 [20].

Sludges and precipitates
Activity concentrations of natural radionuclides from U-238 and Th-232 decay chains in the sludges and precipitates from GTF's are shown in table 3 and figure [15,16,25], although in the study from Germany there are many more samples and the few extreme highest values are significantly higher than in this study.Nearly all of the sludge and precipitate samples have higher Ra-226 concentrations compared to U-238 (table 3, figure 2), even though in the raw groundwater those concentrations are the opposite way round (table 2, figure 1).It is known that groundwater systems in general are split into radium-type and uranium-type systems depending on the chemical conditions and age of groundwater [8].Interestingly, the sludges and precipitates of this study had a higher frequency of radium-enriched (Ra-226 > U-238) residues even though the groundwater is of the uranium-type (U-238 > Ra-226).Radium in groundwater occurs mainly in ionic form [8,32], and radium is known to be efficiently removed from water through coprecipitation and adsorption into manganese oxides [31,32].Finnish groundwater often contains manganese that needs to removed from the drinking water in addition to iron (table 1), so it is a possibility that the formation of manganese oxides during aeration and filtration is involved in the capture of radium to residues, but it was not possible to determine a detailed water chemistry or mineralogy of the residues as a part of this study.Most of the samples of sludges and precipitates (7 out of 9) had at least one radionuclide concentration above the clearance level of 1 kBq kg −1 .This result indicates that if backwash sludges are formed as a solid waste fraction at a GTF, their radionuclide content needs to be characterised even if the activity concentrations in the raw groundwater are low.The potential amounts of solid residues formed are estimated with mass balance calculations in section 3.4.The analysed samples included a precipitate from inside a raw water pipe.It had activity concentrations close to the clearance level of 1 kBq kg −1 , U-238: 900 Bq kg −1 , Ra-226: 800 Bq kg −1 , Th-228: 700 Bq kg −1 and Ra-228: 900 Bq kg −1 .This implies that also the residues formed in groundwater pipelines may accumulate natural radionuclides at a significant level.This should be considered in the waste treatment and working protocols when old water pipelines are removed.
The highest activity concentration in a solid residue in this study (6400 Bq kg −1 of U-238) was found in a precipitate which had formed in an aeration tower (table 3).This sample had mainly accumulated uranium and not the other radionuclides (figure 2).Visually this rusty-looking sample appeared to consist of mainly iron oxyhydroxides, and it is known that uranium may be efficiently incorporated to iron oxides through sorption and coprecipitation [32][33][34].A detailed mineralogical study of the residues was beyond the scope of this investigation, however.

Filters and backwash-affected soil
Sand and limestone filters had lower activity concentrations of natural radionuclides compared to the sand-anthracite and GAC filters (table 4, figure 2), and precipitates and sludges (table 3, figure 2).In filter materials, the highest activity concentrations were found in sand-anthracite filters with up to 5200 Bq kg −1 Ra-226 and 1900 Bq kg −1 Ra-228 (table 4, figure 2).The sand filters contain only up to 600 Bq kg −1 Ra-226 and even less is found in the limestone filters.The Ra-226 and Ra-228 concentrations in the filters and in the sludges are correlated (figure 2), which is a feature typical for groundwater systems at least in radium-type groundwater systems with higher concentrations of radium [8].However, the ratio of Ra-226 and Ra-228 is higher than one in the filters and sludges, but close to one in the soil, GAC-filters, aeration precipitates and pipeline residues (figure 2).Ra-228 and Th-228 show disequilibrium with Ra-228 > Th-228 (mean ratio 1.6 in filters), which is understandable due to the higher mobility of radium compared to thorium in groundwater systems [8].Because the samples were taken from operating facilities, this disequilibrium is maintained by the dynamic system where radionuclides have different concentrations in the groundwater, capture rates into solids can be variable, solids are periodically removed by backwash, and occasionally new filter material is added.Even U-238 and Pa-234m show disequilibrium in some samples.The radiation safety of GAC filters has been investigated previously due to the gamma radiation emitted by radon progeny and including the accumulation Pb-210 in the GAC filters [29].Here, we report also the other natural radionuclides in the same GAC filter samples (table 4).The GAC filters had variable accumulations of U-238 (170-460 Bq kg −1 ) and Pb-210 (210-1000 Bq kg −1 ), but there was significantly less radium especially compared to the anthracite filters (table 4, figure 2).
By visual observation it seemed that sand and limestone filters had a layer of precipitated material on top, which is mostly washed from the filter during the backwash.Sand, limestone and sand-anthracite filters are only very rarely changed, and filter waste is not frequently formed from these types of filters.When natural radionuclides are accumulated into the filter material such as in the sand-anthracite filters, the waste amounts can be substantial only when the filter is changed.GAC-filters are changed when the removal  2) from selected groundwater treatment facilities in Finland (table 1).Measurement uncertainties are shown with a coverage factor of 2. Table 3.Activity concentrations (Bq kg −1 dry weight) of naturally occurring radionuclides in precipitates and sludges from groundwater treatment facilities in Finland.The uncertainties are shown with a coverage factor of 2. U-238 was calculated based on the measured activity of U-235 and the natural ratio of uranium isotopes.efficiency drops, or the water quality does not comply anymore with regulations.For GAC-filters, the time of use before the change effects the amount of accumulated Pb-210 [29].

Uranium series
The filters are periodically backwashed to prevent clogging and maintaining the functionality.Sand filters are backwashed more often than lime bed filters.Sand filter clogs more frequently because the function is to remove precipitated particles.The lime bed filter is mainly used for alkalisation and clogging is not typical.The backwash water from the filters may be discharged from the facility via a settling tank or sedimentation tank to a discharge area, directly to a discharge area, or into sewage system.The radioactivity of the analysed top 8 cm soil layer in the discharge area had slightly elevated activity concentrations of radium-isotopes in comparison to the unaffected control area (table 5, figure 2).The highest Ra-226 activity concentrations in the discharge affected soil were 250-300 Bq kg −1 which was five times the activity concentration of the control areas (45-66 Bq kg −1 ).The activity concentrations of natural radionuclides in backwash water were not significantly elevated compared to the raw or treated groundwater (table 2), however, indicating the additional activity in the soil is mostly in the form of solid precipitates transported by the backwash (table 3).Compared to the backwash precipitates and sludges (table 3) the Ra-226 concentration in the discharge-affected soil is roughly diluted by a factor of ten, giving an indication of the amount of solid residues that could be inserted into the soil by the backwash discharge.The observed level of Ra-226 (300 Bq kg −1 ) in the soil in a small contained area is not a radiation protection issue for the public directly.However, the affected soil would not be suitable for example as a building material without further assessment [26].The pumping of backwash precipitates directly into the soil is not necessarily the preferred method and therefore it is important to look at options for backwash practices such as sludge or settling ponds, or backwashing into the sewage system, when investigating the natural radioactivity at GTF's.

Mass and activity balance of solid residues
The gross alpha, U-238 or Ra-226 in raw and treated groundwater show no systematic change in concentration due to the water treatment (table 2).Considering that the analytical uncertainty (k = 2) of U-238 (ICP-MS) and Ra-226 (LSC) in the water at low concentrations can be 20%-50% (table 2), it is possible that for example 10%-20% of these nuclides could be captured in the filters and sludges without it showing up in the measurements from water samples.In support of this, it has been reported previously that for anthracite filters the removal of uranium occurs at this 10%-20% rate [23].An added complication is that the water composition could be variable in time.The temporal variations in the water composition were outside the scope of this study but in previous studies from water treatment facilities the compositions of Table 6.An example of a mass and activity balance calculation of solid residues from groundwater treatment, to estimate the potential mass of solid residues that can form from a given groundwater volume, activity concentration in groundwater, fraction of captured activity in residues, and activity concentration in solid residues.

Groundwater Solid residues Fraction of activity captured in residues
Volume (Mm radon and uranium were seen to be constant over time in repeated sampling [23].For the purposes of estimating the potential masses of solid residues that could be formed at a GTF, a simple mass and activity balance is calculated (table 6).The activity per year in raw groundwater is the product of groundwater volume per year and the activity concentration in raw groundwater.The activity per year captured in solids is determined by multiplying the activity per year in raw groundwater by the fraction of activity captured in residues.Finally, the mass of solid residues is the activity per year captured in solids divided by the activity concentration in the solids.The calculation in table 6 shows that if the change between raw and treated groundwater activity concentrations are within analytical uncertainties (captured fraction of activity 0.1-0.2), the total mass of formed solid residues with activity concentrations exceeding 1 kBq kg −1 is likely to be of the order of few tonnes to few tens of tonnes per year at one GTF with treated water capacity of 1 Mm 3 a −1 and a moderate activity concentration of 0.5 Bq l −1 .For larger volumes of treated water, or for higher activity concentrations in the raw water the potential mass of residues can be larger at hundreds of tonnes per year.The amount can be crudely estimated from table 6 directly (or by repeating the calculation with case specific data) as the mass of the residue grows linearly with the water volume and water activity concentration, if the captured fraction stays constant.This mass balance method is a simple order-of-magnitude estimate that does not consider separately the mass of the filter which is important especially for anthracite filters were radionuclides are adsorbed into the filter material.However, as discussed in section 3.3, the precipitates in sand and limestone filters are mostly loose particulates that are removed from the filter during the backwash process.For these types of solid residues the mass balance calculation can give estimates of the limits for the masses of residues that can form at a particular GTF.
In Europe, the treated water volume of 1 Mm 3 a −1 would be used by a population group of approximately 23 000, based on an average water consumption of 120 l per person per day in Europe [19].A population of this size produces sewage sludge at a rate of approximately 500 tonnes per year (20 kg a −1 dry weight per capita [35]).Continuing the mass balance calculation of table 6 further, if all the backwash precipitates and sludges at a GTF were inserted into the sewage during the backwash of filters, it would lead to an average dilution by a factor of 5-500, if the fraction of activity captured from groundwater into solids was 0.1-0.2 and the initial concentration of the radionuclide in the untreated groundwater is <0.5 Bq l −1 .In that case the solid residues forming at a GTF would not be likely to cause radiation exposure issues at the sewage treatment facilities or in the further recycling or disposal of the sewage sludge, because the activity concentrations would be less than the clearance level of 1 kBq kg −1 in the mixed final sewage sludge.

Exposure to natural radiation
Filters that are in use under a water layer or inside containers do not cause internal radiation exposure to workers at GTF's through inhalation or ingestion during normal operation.In GAC-filters the short-lived radon daughters may lead to elevated external dose rates close to the filter [29], but the occupancy of GTF workers very close to filter materials is low for GAC-filters [29], and also for backwash sludges [15].The filters and associated backwash sludges may cause exposure or workers mainly when handled for removal.However, the limestone, sand and sand-anthracite filters at GTF's in Finland are used mainly without Table 7. Estimate of radiation exposure from solid residues for a worker at a groundwater treatment facility in Finland.Activity concentrations of 6 kBq kg −1 for the uranium-series and 2 kBq kg −1 for the thorium-series are used for the sludge composition.Dose coefficients for workers are taken from ICRP publication 137 [36].GAC = granular activated carbon.changing, and the amount of filter is kept sufficient by backwashing and adding new filter material as needed.To get an idea of the radiation risk to workers, the natural radiation exposure for a worker at a GTF was estimated (table 7) based on the maximum activity concentrations of materials found in this study.In a simple theoretical assessment for a hypothetical worker at a GTF, the worker was exposed to sludges (maximum concentrations from tables 3 and 4 for each decay series), and external radiation from radon progeny (GAC filters [29]).The working time 1500 h a −1 was divided to 100 h a −1 in the vicinity of GAC-filters, 100 h a −1 handling sludges with radium and uranium, and 1300 h a −1 for other work.The working time of 100 h a −1 with sludges is conservative, as in a more detailed study from Germany 20 h a −1 was used for facility workers, and 100 h a −1 was used as a worst case example of a conservative case [15].Similarly, a more detailed study of GAC-filters stated that water plant maintenance personnel do not work long periods of time near the GAC-filters, and an exposure time of 8 h a −1 was used [29].The estimated external exposure was calculated to be 0.25 mSv, internal exposure by inhalation 0.05 mSv and ingestion 0.02 mSv, using conservative assumptions (table 7).The total calculated effective dose is ∼0.3 mSv a −1 , which is significantly less than the reference value of 1 mSv a −1 for workers.The inhalation dose is an overestimate, because a dust concentration of 1 mg m −3 was used and in reality the backwash sludges are handled mainly when wet.The ingestion rates used are also conservative (table 7).The external exposure was the most significant contributor for the total dose, and this is dependent on the dose rate and working distance from the GAC filters [29].The conclusion is that if the solid residues at GTF's contain <10 kBq kg −1 of U-238, Ra-226, Pb-210 and Ra-228, it is unlikely that the exposure of workers could be higher than the reference level of 1 mSv a −1 .In support of this, in an exposure assessment at water treatment facilities in Germany (500 samples from 400 companies), where the solid materials had higher maximum activity concentrations than in this study, the exposure of different workers remained also under the reference value of 1 mSv a −1 with a median exposure of 0.02 mSv a −1 , worst case maximum of 0.65 mSv a −1 , and 99% of the cases <0.5 mSv a −1 [15].A special case could be workers that specifically work full-time in the service and replacement of GAC-filters in many facilities throughout the year, but it is unclear whether such workers exist in Finland.For workers at GTF's, the exposure from indoor radon is a much more prominent issue compared to the solid residues.In Finland, the indoor radon concentrations at GTF's have so far been found to be larger than the reference level 300 Bq m −3 in 39% of the measurements, and in 13% of the cases the concentration has been >1500 Bq m −3 [28].So far, the highest measured indoor radon concentration at a GTF in Finland has been 31 600 Bq m −3 [28], which is more than a hundred times the reference level for indoor radon at workplaces in Finland.The public is not significantly exposed to the solid residues from GTF's if the waste materials are disposed of properly for example at appropriate landfill sites, or the backwash including precipitates goes through sewage treatment facilities.However, if filters, sludges or other solid residues from GTF's would be used for example in agriculture or in building materials [25], there could be potential issues with the exposure of the public.Therefore, the characterisation of natural radioactivity of solid residues from GTF's is essential for proper reuse, recycling, utilisation or disposal of solid residues, even if the activity concentrations in the untreated groundwater are low.

Conclusions
Measurements of natural radioactivity of samples from GTFs in Finland were made to characterise the solid residues and their possible exposure risks for workers and the public.Naturally occurring radionuclides were found to be accumulated in sand-anthracite and granular activated carbon filters, backwash sludges and aeration precipitates at concentrations exceeding the general clearance level of 1 kBq kg −1 for radionuclides in the uranium-and thorium-series.The occurrence of such solid residues could not be easily predicted based on the measured radioactivity levels in the raw and treated groundwater.Even at relatively low activity concentrations in the raw groundwater, or with no measurable loss of activity between the raw and treated groundwater, accumulation of natural radionuclides to over the clearance level was observed in filters, sludges and precipitates.This is thought to be due to the large amounts of water processed per year at GTFs and this type of accumulation is shown to be possible through simple mass and activity balance calculations while also considering the measurement uncertainties of the water samples.The same mass balance approach can be used at any groundwater treatment facility to estimate the potential amounts of solid residues that could be of concern for radiation protection.Unlike the case of indoor radon, the radiation exposure from solid residues at GTFs in Finland is not likely to be a significant issue from the point of view of exposure of workers.However, when planning for the reuse, recycling, utilisation, or disposal of solid residues from GTFs, the characterisation of the natural radioactivity in the materials is essential to ensure proper risk management and radiation protection of the public in the final use of the materials.

2 .
The activity concentration of U-238 varied from 20 to 6400 Bq kg −1 , Ra-226 from 130 to 4300 Bq kg −1 , Pb-210 from 15 to 1200 Bq kg −1 , and Ra-228 from 30 to 1200 Bq kg −1 .The distributions are similar to what has been observed in groundwater treatment from Germany and Spain

Figure 1 .
Figure 1.Relationships of gross alpha (A), U-238 (B), Ra-226 (C), and Rn-222 (D) activity concentrations in raw and treated groundwater (table2) from selected groundwater treatment facilities in Finland (table1).Measurement uncertainties are shown with a coverage factor of 2.

Table 1 .
General information of the studied groundwater treatment facilities (GTFs) in Finland.GAC =

Table 2 .
Activity concentrations (Bq l −1 ) of naturally occurring radionuclides in water samples from groundwater treatment facilities in Finland.Facilities marked with 'a' and 'b' mean separate raw groundwater extraction sources which are treated at the same facility.Backwash frequency is marked in brackets.The measurement uncertainties are shown with a coverage factor of 2. nd = not determined.

Table 4 .
[29]vity concentrations (Bq kg −1 dry weight) of naturally occurring radionuclides in filter samples at groundwater treatment facilities.Uncertainties are shown with a coverage factor of 2. U-238 was calculated based on the measured activity of U-235 and the natural ratio of uranium isotopes.The Pb-210 results for granular activated carbon (GAC) are from[29].

Table 5 .
Activity concentrations (Bq kg −1 dry weight) of naturally occurring radionuclides in soil samples from groundwater treatment facilities in Finland.From each facility, soil areas affected by backwash water discharge, and areas unaffected by the discharge were analysed.The uncertainties are shown with a coverage factor of 2. U-238 was calculated based on the measured activity of U-235 and the natural ratio of uranium isotopes.