Skip directly to search Skip directly to A to Z list Skip directly to site content


Tritium Releases and Potential Offsite Exposures


March 11, 2002


SRS is a large Department of Energy site, currently operated by the Westinghouse Savannah River Company with 12,000 employees. The site, on the South Carolina side of the Savannah River, is approximately circular with a 16-km radius (see Figure 3.1). Reactors, separations plants, and ancillary facilities for producing and purifying radionuclides - primarily tritium (3H) and plutonium (239Pu) for the Defense Department- began operation in October 1954. Of five reactors, one was shut down in 1964, a second one was shut down in 1966 and restarted in 1984, and all reactors were shut down by 1988. One reactor was briefly restarted and shut down at the end of 1991. Reactor power levels were relatively high from 1957 to 1986. The separations plants and ancillary facilities are still operating.

Generation of tritium, its movement within SRS, and its transfer through the environment to persons off site are complex processes that are described here only briefly. A detailed source of information is Murphy et al. (1993). Tritium control technology at SRS and other facilities was described by Rhinehammer and Lamberger (1973). Much of the tritium information has been reviewed by an independent group (Till et al. 1999). Tritium releases, concentrations in environmental media, and radiation dose model calculations are reported annually (Arnett and Mamatey 2000).

Savannah River Site Map
Figure 3.1. Savannah River Site Map

3.1 Sources of tritium releases at SRS

Tritium was produced in the reactors by neutron bombardment of lithium in lithium-aluminum targets. These targets were dissolved in the tritium separations plants and the tritium product was then purified and collected as the gas. Some of this tritium was lost from production through release to air, release in wastewater, and retention in solids. The airborne releases are discharged from the SRS separations plants and reactors into air through 60 m stacks that are clustered near the center of the site.

Tritium was also formed adventitiously in reactors from deuterium in heavy water (deuterium oxide, D2O) that was used as coolant and neutron moderator, and as a fission product in fuel. The formation from deuterium is through the (n, g) reaction supplemented to a small extent by an (n, p) reaction on the tritium decay product, 3He (Holford and Osborne 1979). Heavy water that escapes from reactor systems carried the tritium with it as DTO. Some tritium accompanied the fraction of heavy water that is vaporized into the reactor blanket gas. Deuterium and tritium gases formed by radiation-induced dissociation of coolant water were passed with the blanket gas through a recombiner system with oxygen to form water. Some gas with DTO vapor occasionally was vented for discharge to the stack.

Some heavy water with tritium drained into sumps from leaks. Some clung to surfaces of the lithium targets, uranium targets, and uranium fuel as they were removed periodically from the reactor and is collected in drip trays and with wash water. Any remaining DTO was carried to the storage facilities and separations plants. The Rework Facility through which heavy water is recycled for purification discharged some DTO with tritium. Laboratory and test plant facilities also may discharge tritium to air and wastewater.

Tritium formed by neutron-induced ternary fission in uranium can leak from failed fuel with other fission products into the reactor coolant. Fission products were also carried in discharged air and wastewater from fuel storage. Most fission products are transferred to the uranium separations plants in the targets and fuel that are dissolved there to extract plutonium. Tritium accompanies fission products that may become airborne or are stored as high-level and low-level radioactive waste. Spent fuel from other facilities also is stored and processed at these SRS separations plants.

Wastewater may be discharged directly to streams that flow through the site and into the Savannah River or may be treated by evaporation and solidification in concrete, with subsequent discharge of residual treated water. Until a few years ago, much of the tritium-bearing wastewater was discharged into seepage basins so that a fraction of the tritium could decay as the water moves slowly through the shallow aquifer before seeping into the local streams. This tritium in groundwater continues to enter the streams now that the seepage basins are closed. The basins, vegetation, and on-site surface water bodies are also minor sources of airborne tritium by evaporation and evapotranspiration.

Liquid high-level radioactive wastes that contain tritium are stored in buried high-level waste tanks. These wastes, which are reduced in volume by evaporating water (including tritium), are currently being processed to form low-level radioactive liquid waste and high-level radioactive solids. This low-level liquid waste is solidified as concrete and the high-level radioactive solids are contained in steel cylinders. Both solid wastes are stored on site. Solid radioactive wastes stored on site at the Solid Waste Disposal Facility in near-surface repositories are sources of tritium that migrates into groundwater and air.

3.2 Monitoring of releases of tritium

3.2.1 Air effluents from separations plants

Since 1985, airborne releases from the tritium separations plants have been measured with dual Kanne ionization chambers. One chamber records the total beta-gamma activity, while the other chamber records beta-gamma activity from an air stream that has been stripped of water vapor, including tritiated water (HTO). Because the ionization in the chambers is mostly from tritium, the first chamber is considered to record total tritium, and the second, total tritium minus HTO, which SRS generally reports as tritiated hydrogen gas (HT and T2).

Only a single Kanne chamber was used at the tritium separations plants before 1985, which did not allow any distinction to be made between HTO and non-HTO. Gaseous or volatile fission and activation products that accompany the airborne tritium are also measured in the Kanne chambers so that the total tritium activity may be overstated.

The tritium released by the separations plants that process enriched uranium fuel and depleted uranium targets was not monitored. Its activity was calculated on the basis of ternary fission yield, estimated evaporation rates, and irradiation history. Tritium constitutes only a small fraction of the discharged radionuclides, which are mostly 85Kr. Of the total fission-produced tritium, 10% was assumed to be discharged as HTO to air, and the remainder discharged with waste water (Whitney 2001).

3.2.2 Air effluents from reactors

Airborne releases from the reactors consist of tritium plus radioactive noble gases, mostly the 41Ar activation product and several Kr and Xe fission products. The Kanne chamber record represents the upper limit of total tritium, but cannot be used for monitoring tritium. Instead, four monitoring systems specifically for tritium were installed in succession between 1954 and the present; the first three were each in turn replaced, apparently because of imperfections (Lee 1998).

The first reactor stack tritium monitoring system, used in 1954 through 1957, collected water vapor on silica gel during brief sampling periods. The collected water was recovered from the silica gel by distillation and measured by liquid scintillation (LS) counting. From 1957 to about 1972, the second system collected water vapor continuously by condensation for subsequent LS counting. From about 1972 to about 1986 (installation and replacement dates varied among reactors), the third system, designated "Stack Tritium Monitor" (STM), measured radioactivity with dual ionization chambers. The difference between total beta-gamma activity in the first chamber and total activity minus HTO in the second chamber, which received a dried air stream, was reported as HTO activity.

The current Berthold Tritium Monitor (BTM) system is a two-channel proportional counter that records low energy beta particles attributed to tritium in the first channel and higher-energy beta particles from other radionuclides in the second channel. A correction factor based on the second-channel count is applied to subtract from the first-channel count rate the low energy beta particles from the other radionuclides. Records indicate that the STM and BTM were calibrated periodically with check sources, and occasionally with tritium gas standards. Because the second and third systems continued to be used as the subsequent system was installed, data for comparisons were collected. A few of these records are still available, although they are from periods after reactor shutdown (Lee 1998).

3.2.3 Liquid effluents

Liquid effluents are monitored from all facilities by collecting water samples and measuring tritium with LS counters. Continuous samplers are used at points of discharge. Other sampling is performed in on-site streams and wells, in the Savannah River, and at the two public water supply intakes in the lower Savannah River. Tritium releases to groundwater are estimated from measurements performed at groundwater monitoring wells.

3.2.4 Environmental media and biota

Tritium in the environment, unlike most other radionuclides discharged at SRS, has been at readily detectable levels. Tritium concentrations have been measured and reported annually since 1971 for nearby rain, water vapor, milk, foodstuffs, wildlife, surface water, groundwater, and fish in the Savannah River (Arnett and Mamatey 2000). Tritium in some of these media also is measured by the environmental protection agencies of South Carolina (Brownlow et al. 1999) and Georgia (GDNR 2000) on their respective sides of the river. The US EPA monitors tritium in a surface water monitoring station in the Savannah River downstream from SRS at Allendale (USEPA 1994).

3.3 Estimated releases and environmental concentrations

3.3.1 Tritium in general

Airborne tritium continues to be released from the operating tritium and fuel separations facilities, and from reactor buildings in which tritiated heavy water is currently stored. Figure 3.2 shows the reported total airborne releases of tritium since 1954. The values are the sum of the total tritium from the tritium separations plants and all airborne tritium measurements and calculations at other SRS facilities. Figure 3.2 also shows the estimated HTO component in the total release from 1985 on, which is when this information became available.

Total Airborne Releases from Savannah River Site, 1954-1998
Figure 3.2. Total airborne releases of tritium from 1954 through 1998 (represented by gray shaded bars) and directly measured HTO from 1985 through 1998 (represented by black shaded bars) from SRS. The data for 1954-1991 are from Murphy et al. 1993; those for 1992-1999 are from Arnett & Mamatey 2000.

In addition to the routine releases of airborne tritium from the separations plants, the reported totals include hundreds of inadvertent releases (Murphy et al. 1993). Each of the 11 largest incidents released between 0.2 PBq and 20 PBq (5 kCi and 500 kCi) of tritium. The plumes of major releases were mapped by monitoring the environmental path. The amounts released could be calculated from such mapping; in some cases, monitoring at the point of release or a material balance was used to calculate amounts released.

Table 3.1 shows the total reported releases to air since 1954. The calculated annual amounts from processing enriched uranium fuel and depleted uranium are included in the Separations Plants totals (Murphy 2000).

Reported discharges, PBq (kCi)
Air Water
1954-1991 1992-1999 1954-1991 1992-1999
Separations Plants 650 (17,500) 22 (590) 0.3 (9) 2 (59)
Reactors 260 (7,000) 6 (160) 32 (870) 0.7 (20)
Other Facilities 3.5 (94) 0.11 (3) 5.6 (150) 0.11 (3)
Outdoors (Basins/Solid Waste) 15 (410) 0.05 (1.4) 17 (470) 0

Table 3.1 Reported Total Tritium Discharges from SRS. The data for 1954-1991 are from Murphy et al. 1993; the data for 1992-1999 are from Arnett & Mamatey (2000). Discharges to basins had been discontinued by 1992.

A review of available records and recalculation of the airborne tritium releases for the years 1955 to 1991 has confirmed the annual totals shown in Figure 3.2 and Table 3.1. The review, performed as part of the SRS Environmental Dose Reconstruction Project (Till et al. 1999), found individual deviations from the annual values shown on Figure 3.2 but the total value between 1955 and 1991 was only a few percent higher than shown.

Information is not available in the reports by Murphy et al. (1993) and Till et al. (1999) on monitoring at other SRS facilities for airborne tritium releases, but Table 3.1 indicates that relatively little tritium was released at facilities other than the separations plants and the reactors. Tritium releases to air from outdoor sources at SRS were estimated from evaporation rates and tritium concentrations in water.

Annual direct releases and indirect seepage of tritium to streams are shown for 1995 on in Figures 3.3 and 3.4. The direct tritium discharges have been measured. The other tritium releases are monitored and checked by modeling tritiated groundwater flow. Values of total aquatic discharges, calculated from a model for groundwater flow in the case of indirect discharges, agree with measured values of tritium concentrations in water multiplied by estimated flow rates in the on-site streams and also in the Savannah River downstream from SRS (Arnett and Mamatey 2000).

The total tritium discharge reported by SRS to the end of 1998 was 940 PBq (25.5 MCi) to air and 60 PBq (1.6 MCi) to water. Airborne discharges were more than 40 PBq (more than 1 MCi) for each year from 1957 to 1964, but by 1998 had decreased to 3 PBq (83 kCi). Liquid releases peaked at 2 - 4 PBq (50 - 100 kCi) per year from 1961 to 1976, and had decreased to about 0.4 PBq (10 kCi) in 1994 through 1998. Of the radionuclides discharged at SRS, tritium and 85Kr (during earlier years) constitute the largest activities.

Releases to Streams, Savannah River Site, 1954-1998
Figure 3.3 Releases to Streams from SRS. The data for 1954-1991 are from Murphy et al. 1993; those for 1992-1999 are from Arnett & Mamatey (2000).

The results of tritium analyses of air, water supplies, and foods are reported annually (Arnett and Mamatey, 2000). Air samples have been collected at the site perimeter; water samples, at downstream water supplies; food and milk samples, generally within 16 km of the site; and fish, from the SRS reach of the Savannah River. The minimum detectable concentrations were reported to be approximately 2 Bq/m3 (50 pCi/m3) in air, 15 Bq/kg (0.4 pCi/g) in liquids, and 1.5 Bq/kg (0.04 pCi/g) in solids. Tritium concentrations were generally higher in previous years when tritium releases were higher.

In 1999, the maximum concentration observed in off-site air was 6 Bq/m3 (150 pCi/m3). For other media and food the observed concentrations ranged from undetectable to the maximum values reported in Table 3.2. Also shown in the Table are the arithmetic means of all measurements in each category. Because of the low concentrations, the measurement uncertainties are relatively large. For some foods, the concentrations in the more distant control area (10 - 25 km distant) are higher than those closer to the SRS.

Seepage to Streams from Savannah River Site, 1954-1998
Figure 3.4 Seepages to Streams from the SRS. The data for 1954-1991 are from Murphy et al. 1993; those for 1992-1999 are from Arnett & Mamatey 2000.

Exposure Route Concentration;
Bq/kg wet weight (pCi/g wet weight)
Mean of all samples measured Maximum observed
Drinking Water 37 (1.0 ) 60 (1.6)
Greens 5.2 (0.14) 8 (0.22)
Fruit 7.7 (0.21) 44 (1.2)
Beef 4.3 (0.12) 5.3 (0.14)
Milk 4.5 (0.12 ) 25 ( 0.67)
Fish 12.5 (0.34) 90 (2.4)

Table 3.2 Tritium concentrations in selected media measured in 1999 (Arnett and Mamatey, 2000).

The tritium concentrations measured in these environmental samples have been compared with concentrations predicted with the aid of transport models from release rates or concentrations in other environmental samples. Annual average concentrations of airborne water vapor have been reported to be within a factor of two of the values predicted from tritium release rates (Simpkins and Hamby 1997). Agreement by this margin was considered reasonable, given the approximations and averaging in the model. The ratio of tritium concentrations in moisture in vegetation relative to moisture in air averaged 0.54 ± 0.10 from 1982 to 1990 but ranged widely (Hamby and Bauer 1994). Tritium concentrations in fish tissue samples appeared to be consistent with the observed concentrations in river water (Arnett and Mamatey 2000).

3.3.2 Releases and environmental concentrations of OBT

The activity of tritium as OBT in releases, environmental media and foods is not addressed in SRS reports of airborne and liquid releases. Therefore, one can only attempt to infer whether (1) the recorded tritium releases include OBT or (2) OBT was discharged in amounts above and beyond the recorded annual tritium releases. The extent to which OBT has been included in the reported release values is less important than is the extent to which OBT is released as unmeasured tritium. This is because the modeling of tritium releases as HTO is likely to be reasonably conservative for OBT. In addition, OBT in the environment necessarily has to be considered following releases as HTO and HT. It is important to establish whether there could have been activities of tritium as OBT released that were substantially larger than reported tritium releases.

Any OBT in airborne releases from tritium processing in the separations plants would be included in reported tritium releases. The earlier ionization chamber measurements in the stacks measured all forms of airborne tritium. After the dual chambers were installed, any OBT was included in the record as part of the HT component, i.e., the non-HTO-category.

A special study of gaseous tritium forms in various tritium processing activities defined OBT as the fraction oxidized to HTO and measured after HTO and HT had been removed. Less than 1% was OBT in six cases out of seven. Only for the process associated with by-product helium purification was the gaseous form mostly OBT (Milham and Boni 1976). Several sources of carbon, such as graphite crucibles, pump oil, and carbon dioxide in the process system, were suggested as the constituents of the organic molecules. Tritium released from this process constitutes only a small fraction of the total airborne tritium release.

At the reactors, only the record from the current monitor, the BTM, includes OBT as part of total airborne tritium. The three earlier tritium monitoring systems did not measure OBT. Hence, tritium released to air as OBT before the BTM was installed (1985 to 1988) would not be included in the reported releases. The SRS Dose Reconstruction Project (Till et al. 1999) cites two documents (Longtin et al. 1973; Jacober et al. 1973) for the statement that "nearly all the tritium released from reactors" was HTO. These documents provide excellent descriptions of tritium control at SRS but give no data to support the statement. Another document (Miller and Patterson 1956) is also cited (Till et al. 1999) as stating that "tritium losses from the reactor area were estimated to be 100% oxide), but the reference could not be reviewed. Milham and Boni (1976) infer, from the large mass of heavy water in each reactor, that the released tritium should be in the form of HTO.

Airborne tritium releases from the Heavy Water Rework facility were monitored with silica gel columns and dehumidifiers that would not have measured OBT (Murphy 2000). Hence any tritium discharged as OBT would have been unreported. The calculated total tritium release at the uranium processing plants would have included OBT. In the calculations of tritium releases to air from basins and solid waste disposal sites, the model was that of HTO vaporization and OBT was not considered.

For monitoring tritium in liquid effluents and streams at SRS, water samples are counted directly (without distilling them). Hence total tritium, including OBT, is measured by LS counting. Savannah River water, by contrast, is distilled before counting so that only HTO is measured in the distillate (WSRC 2000). We interpret the consistency of the annual tritium discharges reported on the basis of Savannah River samples with those estimated on the basis of concentrations in the sampled on-site streams and the streams' flow rates to indicate that OBT in Savannah River water is a small fraction of HTO.

The SRS environmental surveillance program does not consider OBT in food and air samples. The airborne tritium sampler collects water vapor on silica gel. Tritium in foodstuffs, vegetation, and fish is measured with an LS counter in water extracted by distilling or freeze-drying. Special environmental studies provide information on OBT in the SRS environment but not on OBT in releases from SRS. Any observed OBT may be present as a result of conversion of HTO to OBT or HT to OBT (via HTO) in soil, plants, and animals (see Section 2.1), or the OBT may be from other sources.

Special studies of OBT at SRS have been reported in deer, trees, vegetation, soil, and fish (Murphy et al. 1993). Normally, the OBT was obtained from a dried sample by ashing the sample and collecting the water vapor that was produced. Tritium was measured with an LS counter and reported relative to the amount of water recovered from the sample during ashing, or, in one case, relative to the dried weight of the sample.

Tritium analyses of seven organs in 52 deer collected on site at SRS in 1966 yielded specific activity ratios of OBT/HTO with averages near 1.0 in six organs and of 0.6 in fat (Evans 1969). The OBT concentrations in the cellulose of tree rings for the periods 1950 - 1970 (Sanders 1976) and 1954 -1992 (Kalin et al. 1995) were measured to observe the changing pattern of environmental tritium levels with time. The former studied loblolly pine trees at various distances on site from the tritium production facility; the latter analyzed a sweetgum tree with roots at a groundwater seep line. The tritium concentrations in the tree rings appear to have responded to tritium in both groundwater and air. Tritium in pine litter and the soil beneath had OBT/HTO specific activity ratios generally between 1 and 4; the uptake pattern suggested atmospheric HTO as source of tritium in pine needles; conversion to OBT in pine needles; and transfer of this OBT to soil (Sweet and Murphy 1984; Murphy 2000). During 1991, in five locations in Four Mile Creek, the OBT/HTO specific activity ratios were between 12 and 0.9 in grass and leaves, and between 2 and 0.6 in the fish. On average, both HTO and OBT concentrations in the bass were slightly below those in the water, while grass and leaves by the stream had the lowest concentrations (Eaton and Murphy 1992).

In summary, recent monitoring has shown that OBT is a minor component of the total tritium release. It is not known if earlier discharges from the reactors contained significant quantities of OBT in addition to the reported total releases of airborne HTO. It is also unknown if OBT continues to be discharged from various minor sources at SRS. Tritium in liquid effluents has been measured as total tritium and will have included any OBT.

Further efforts may provide some of the information that is now missing. Additional document searches may yield at least an upper limit on the fraction of OBT - or of total tritium including OBT - that was released from reactors as airborne effluent. Comparison of effluent monitoring by BTM with data from earlier tritium air effluent monitoring systems that are still operating in parallel with the BTM could provide such information for reactors. Such comparison measurements appear to be available only for periods after the reactors were shut down. Direct measurements of the OBT fraction in effluent air and water at least could determine the current fraction of tritium discharged as OBT at this time. Routine monitoring of OBT in milk, foods, and air could provide a measure of the OBT intake by persons off site that is currently ignored by monitoring only for HTO.

3.4 Estimated radiation doses

3.4.1 Doses from intakes of tritiated water

Tritium currently is a major contributor to the dose of the maximally exposed persons off site. Estimated doses reflect the considerable dilution of both airborne and waterborne releases because of the large distances between the stacks and the SRS boundary and of the dilution by Savannah River water.

The estimated annual radiation doses since 1954 to the maximally exposed person off site from airborne tritium as HTO are shown in Figure 3.5 (Arnett and Mamatey 2000). They reached an effective dose equivalent of 18.7 mSv (1.87 mrem ) in 1958 and decreased to less than 0.2 mSv (0.02 mrem ) in 1999. The annual radiation doses from tritium from waterborne effluents reaching public water supplies reached a maximum of 3.3 mSv (0.33 mrem) in 1965 (at Port Wentworth); the dose was 0.6 mSv (0.06 mrem) in 1999. The estimated annual waterborne radiation doses since 1954 to the maximally exposed person off site (just downstream of SRS) are shown in Figure 3.6.

Annual Dose to Maximally Exposed Individual from Airborne Releases from Savannah River Site, 1954-1998
Figure 3.5. Annual Dose to the maximally exposed individual from airborne tritium released from the Savannah River Site 1954-1999. The data for 1954-1991 are from Murphy et al. (1993); those for 1992-1999 are from Arnett & Mamatey (2000). [microsievert, sic]

The estimated radiation doses from airborne releases of HTO to persons off site have been calculated for the combined effect of inhalation, absorption through the skin, and ingestion of food and water.

The dispersion of airborne tritium from all SRS sources is calculated for various distances in 16 directions with a dispersion model that is based on atmospheric conditions compiled over a five year period. The tritium is assumed to pass into the ecosystem as part of the water cycle. Radiation doses to persons are then calculated from tritium concentrations in intake media, intake rates, dose conversion factors and other generic or site-specific parameters. The highest annual dose value beyond the fence line is assigned to the maximally exposed person. Doses due to inadvertent tritium releases are usually calculated separately and are included in the dose from routine releases during that year.

Annual Dose to Maximally Exposed Individual from Waterborne Release from the Savannah River Site, 1954-1998
Figure 3.6. Annual Dose the maximally exposed individual from waterborne tritium released from the Savannah River Site 1954-1999. The data for 1954-1991 are from Murphy et al. 1993; those for 1992-1999 are from Arnett & Mamatey 2000 (microsievert; sic).

There have been a series of evaluations of the validity of, and uncertainties in, the values of the more important parameters applied in the models for calculating these radiation doses. Three such evaluations - comparison of tritiated moisture at ground level calculated from airborne releases with measured concentrations, comparison of measured tritium concentrations in airborne moisture and in vegetation, and comparison of tritium in fish tissue water and in river water - have been noted above (Section 3.3.1).

A fourth evaluation has been conducted to see if annual doses estimated on the basis of the same year's annual average of the atmospheric dispersion parameters would be substantially different from those estimated on the basis of the 5-year running average of the dispersion parameters (Kock and Hamby 1998). The parameters involved were the wind direction, wind speed, and atmospheric stability class for each of 16 directional sectors. The resulting radiation doses to the maximally exposed person were between 0.8 and 1.2 of the reported annual doses based on the 1987 - 1991 five-year composite set of parameters.

The influence of the statistical distributions of selected model parameters on the estimates of dose to the maximally exposed off-site adult from tritiated water in ground-level air has also been evaluated by Hamby (1992). Distributions (such as normal, log-normal, or Gaussian) were selected for each transfer, intake, and dose parameter. The median of the distribution of estimated annual dose was only 6% less than the reported point estimates. The dose distribution was lognormal, with 5th and 95th percentiles that were approximately 4.4 times lower and higher, respectively, than the median (Hamby 1993; Hamby 2000.) In this study, the major contributors to the dose distribution were the distributions assigned to dosimetric parameters.

Radiation doses from tritium in surface streams and in groundwater have been calculated for maximally exposed persons off site at three locations. Two locations are Beaufort-Jasper in South Carolina and Port Wentworth in Georgia, where public supplies provide drinking water. The third location is the Savannah River just below SRS, where fish may be caught for consumption and water intake is possible although no water supply as such exists. The estimated annual doses were consistently of similar magnitude (Murphy et al. 1993). Doses for the third location, which generally are slightly higher than those at the other two, were shown in Figure 3.6 above.

Doses to individuals at different ages have also been estimated (Kock and Hamby 1998). Doses to infants were found to be higher than reported for the conventional maximally exposed adult at SRS. The main tritium exposure pathway for infants was considered to be milk consumption, which was 54% of the estimated dose. For adults, the two main pathways were estimated to be vegetable consumption (44%) and inhalation plus skin absorption (41%).

Doses to maximally exposed individuals have been estimated by SRS staff from the concentrations of tritium measured routinely in air, drinking water, fish and terrestrial foods environmental media as a check on the values of doses calculated from releases. They report the dose estimated for the maximally exposed persons from drinking water and from eating fish but do not report dose estimates based on the low concentrations measured in air and terrestrial foods.

3.4.2 Doses from intakes of organically bound tritium

Since there is no direct monitoring of OBT in environmental media and foods that may have been contaminated by tritium from SRS, estimates of doses to individual members of the public from OBT intakes can only be made following the relations outlined in Section 2.4. These dose estimates are based on computer-modeled concentrations.

If we assume that these components constitute the total intake of food that has been contaminated by tritium in the local environment and that they are representative of their food groups, then the doses from the measured concentrations of tritium in Table 3.2 may be estimated from the relations developed in Section 2.4 and tabulated consumption rates (I) for the food types represented by these samples (Health Canada 1994). Table 3.3 provides these estimates. The values for moisture content (m), water equivalent factor (f), and the ratio of specific activities (R) are as suggested in Section 2.4.

The tabulated foods total 0.9 kg/d. Although these food types are likely to be those subject to contamination by tritium in the local environment, an alternative estimation of the OBT dose is to take the mean concentration of tritium in the moisture in these foods and estimate the doses from HTO and OBT for a total daily food intake of 1.6 kg (see Table 2.2)

The unweighted average concentration of tritium in the food moisture in the measured samples listed in Table 3.2 is approximately 9 Bq/kg (0.24 nCi/kg), corresponding to a tritium specific activity of 81 Bq/kg hydrogen (2.2 nCi/kg). If we assume that all the tritium in the moisture in a reference daily food intake of 1.6 kg has this specific activity and that the specific activity of the tritium in the organic material is 1.2 times this - i.e., 98 Bq/kg (2.6 nCi/kg) then, with a total hydrogen daily intake of 0.14 kg and 85% moisture in the food, the annual dose from food would be 0.1 mSv (10 mrem), with 36% coming from the OBT intake.

Diet component Assumed parameters Mean values of annual dose* Total annual dose from maximum values
From HTO intake From OBT intake Total
m f R I
Drinking water 1 1 1 1.5 360 0 360 590
Vegetables 0.9 0.55 1.5 0.25 9 2 11 16
Fruit 0.9 0.55 1.2 0.19 9 2 11 64
Meat 0.65 0.8 0.8 0.14 4 4 8 9
Dairy products 0.9 0.75 1.2 0.28 9 2 11 58
Fish 0.65 0.8 1 0.04 4 4 8 53
Total from food 35 14 49 200
Total from food and water 393 14 407 790

Table 3.3. Estimated doses from the measured concentrations of tritium in Table 3.2 based on the relations developed in Section 2.4 and tabulated consumption rates (I) (Health Canada 1994) for the various food types represented by the measured samples. The values for moisture content (m), water equivalent factor (f), and the ratio of specific activities (R) are as suggested in Section 2.4. The right-hand column shows the estimated doses if the highest observed values are selected for each item.
* For clarity, doses are expressed only in SI units in this Table. 10 nSv = 1 mrem.

Given the overestimation in the last paragraph for the level of contamination actually observed, we conclude that the annual dose from food will have been less than 0.1 mSv with less than half of this coming from OBT ingested in the food. The annual dose from drinking water with the tritium levels reported would be about 0.4 mSv. Even if the extreme assumptions are made by taking the highest observed concentrations in measured items, the annual dose would have been less than 1 mSv (0.1 mrem).

Next Section     Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #