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 SavannahRiver Company with 12,000 employees. The site, on the South Carolina side of the SavannahRiver, is approximately circular with a 16-km radius (see Figure 3.1). Reactors, separationsplants, 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 andrestarted in 1984, and all reactors were shut down by 1988. One reactor was briefly restartedand shut down at the end of 1991. Reactor power levels were relatively high from 1957 to1986. The separations plants and ancillary facilities are still operating.

Generation of tritium, its movement within SRS, and its transfer through the environmentto persons off site are complex processes that are described here only briefly. A detailedsource of information is Murphy et al. (1993). Tritium control technology at SRS and otherfacilities was described by Rhinehammer and Lamberger (1973). Much of the tritiuminformation has been reviewed by an independent group (Till et al. 1999). Tritium releases,concentrations in environmental media, and radiation dose model calculations are reportedannually (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 thetritium product was then purified and collected as the gas. Some of this tritium was lost fromproduction through release to air, release in wastewater, and retention in solids. The airbornereleases are discharged from the SRS separations plants and reactors into air through 60 mstacks 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 fissionproduct in fuel. The formation from deuterium is through the (n, g) reaction supplementedto a small extent by an (n, p) reaction on the tritium decay product, 3He (Holford andOsborne 1979). Heavy water that escapes from reactor systems carried the tritium with it asDTO. Some tritium accompanied the fraction of heavy water that is vaporized into thereactor blanket gas. Deuterium and tritium gases formed by radiation-induced dissociationof coolant water were passed with the blanket gas through a recombiner system with oxygento 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 ofthe lithium targets, uranium targets, and uranium fuel as they were removed periodicallyfrom the reactor and is collected in drip trays and with wash water. Any remaining DTO wascarried to the storage facilities and separations plants. The Rework Facility through whichheavy water is recycled for purification discharged some DTO with tritium. Laboratory andtest plant facilities also may discharge tritium to air and wastewater.

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

Wastewater may be discharged directly to streams that flow through the site and into theSavannah River or may be treated by evaporation and solidification in concrete, withsubsequent 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 coulddecay as the water moves slowly through the shallow aquifer before seeping into the localstreams. This tritium in groundwater continues to enter the streams now that the seepagebasins are closed. The basins, vegetation, and on-site surface water bodies are also minorsources of airborne tritium by evaporation and evapotranspiration.

Liquid high-level radioactive wastes that contain tritium are stored in buried high-level wastetanks. 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-levelradioactive solids. This low-level liquid waste is solidified as concrete and the high-levelradioactive solids are contained in steel cylinders. Both solid wastes are stored on site. Solidradioactive wastes stored on site at the Solid Waste Disposal Facility in near-surfacerepositories 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 withdual Kanne ionization chambers. One chamber records the total beta-gamma activity, whilethe other chamber records beta-gamma activity from an air stream that has been stripped ofwater vapor, including tritiated water (HTO). Because the ionization in the chambers ismostly 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, whichdid not allow any distinction to be made between HTO and non-HTO. Gaseous or volatilefission and activation products that accompany the airborne tritium are also measured in theKanne chambers so that the total tritium activity may be overstated.

The tritium released by the separations plants that process enriched uranium fuel anddepleted uranium targets was not monitored. Its activity was calculated on the basis ofternary fission yield, estimated evaporation rates, and irradiation history. Tritium constitutesonly a small fraction of the discharged radionuclides, which are mostly 85Kr. Of the totalfission-produced tritium, 10% was assumed to be discharged as HTO to air, and theremainder 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, mostlythe 41Ar activation product and several Kr and Xe fission products. The Kanne chamberrecord 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 between1954 and the present; the first three were each in turn replaced, apparently because ofimperfections (Lee 1998).

The first reactor stack tritium monitoring system, used in 1954 through 1957, collected watervapor on silica gel during brief sampling periods. The collected water was recovered fromthe silica gel by distillation and measured by liquid scintillation (LS) counting. From 1957to about 1972, the second system collected water vapor continuously by condensation forsubsequent LS counting. From about 1972 to about 1986 (installation and replacement datesvaried 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 counterthat 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 factorbased on the second-channel count is applied to subtract from the first-channel count rate thelow energy beta particles from the other radionuclides. Records indicate that the STM andBTM were calibrated periodically with check sources, and occasionally with tritium gasstandards. Because the second and third systems continued to be used as the subsequentsystem was installed, data for comparisons were collected. A few of these records are stillavailable, 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 atreadily detectable levels. Tritium concentrations have been measured and reported annuallysince 1971 for nearby rain, water vapor, milk, foodstuffs, wildlife, surface water,groundwater, and fish in the Savannah River (Arnett and Mamatey 2000). Tritium in someof 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. TheUS EPA monitors tritium in a surface water monitoring station in the Savannah Riverdownstream 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 separationsfacilities, 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 arethe sum of the total tritium from the tritium separations plants and all airborne tritiummeasurements and calculations at other SRS facilities. Figure 3.2 also shows the estimatedHTO component in the total release from 1985 on, which is when this information becameavailable.

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 reportedtotals include hundreds of inadvertent releases (Murphy et al. 1993). Each of the 11 largestincidents released between 0.2 PBq and 20 PBq (5 kCi and 500 kCi) of tritium. The plumesof major releases were mapped by monitoring the environmental path. The amounts releasedcould be calculated from such mapping; in some cases, monitoring at the point of release ora material balance was used to calculate amounts released.

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

Sources 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 years1955 to 1991 has confirmed the annual totals shown in Figure 3.2 and Table 3.1. Thereview, 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 totalvalue 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) onmonitoring at other SRS facilities for airborne tritium releases, but Table 3.1 indicates thatrelatively little tritium was released at facilities other than the separations plants and thereactors. Tritium releases to air from outdoor sources at SRS were estimated fromevaporation rates and tritium concentrations in water.

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

The total tritium discharge reported by SRS to the end of 1998 was 940 PBq (25.5 MCi) toair and 60 PBq (1.6 MCi) to water. Airborne discharges were more than 40 PBq (more than1 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 haddecreased to about 0.4 PBq (10 kCi) in 1994 through 1998. Of the radionuclides dischargedat 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 (Arnettand 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; andfish, from the SRS reach of the Savannah River. The minimum detectable concentrationswere reported to be approximately 2 Bq/m3 (50 pCi/m3) in air, 15 Bq/kg (0.4 pCi/g) inliquids, and 1.5 Bq/kg (0.04 pCi/g) in solids. Tritium concentrations were generally higherin previous years when tritium releases were higher.

In 1999, the maximum concentration observed in off-site air was 6 Bq/m3 (150 pCi/m3). Forother media and food the observed concentrations ranged from undetectable to the maximumvalues reported in Table 3.2. Also shown in the Table are the arithmetic means of allmeasurements in each category. Because of the low concentrations, the measurementuncertainties are relatively large. For some foods, the concentrations in the more distantcontrol 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 comparedwith concentrations predicted with the aid of transport models from release rates orconcentrations in other environmental samples. Annual average concentrations of airbornewater vapor have been reported to be within a factor of two of the values predicted fromtritium release rates (Simpkins and Hamby 1997). Agreement by this margin was consideredreasonable, given the approximations and averaging in the model. The ratio of tritiumconcentrations in moisture in vegetation relative to moisture in air averaged 0.54 0.10 from1982 to 1990 but ranged widely (Hamby and Bauer 1994). Tritium concentrations in fishtissue 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 addressedin SRS reports of airborne and liquid releases. Therefore, one can only attempt to inferwhether (1) the recorded tritium releases include OBT or (2) OBT was discharged in amountsabove and beyond the recorded annual tritium releases. The extent to which OBT has beenincluded in the reported release values is less important than is the extent to which OBT isreleased as unmeasured tritium. This is because the modeling of tritium releases as HTO islikely to be reasonably conservative for OBT. In addition, OBT in the environmentnecessarily has to be considered following releases as HTO and HT. It is important toestablish whether there could have been activities of tritium as OBT released that weresubstantially larger than reported tritium releases.

Any OBT in airborne releases from tritium processing in the separations plants would beincluded in reported tritium releases. The earlier ionization chamber measurements in thestacks measured all forms of airborne tritium. After the dual chambers were installed, anyOBT 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 OBTas the fraction oxidized to HTO and measured after HTO and HT had been removed. Lessthan 1% was OBT in six cases out of seven. Only for the process associated with by-producthelium purification was the gaseous form mostly OBT (Milham and Boni 1976). Severalsources of carbon, such as graphite crucibles, pump oil, and carbon dioxide in the processsystem, were suggested as the constituents of the organic molecules. Tritium released fromthis 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 oftotal 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) wouldnot 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 excellentdescriptions of tritium control at SRS but give no data to support the statement. Anotherdocument (Miller and Patterson 1956) is also cited (Till et al. 1999) as stating that "tritiumlosses from the reactor area were estimated to be 100% oxide), but the reference could notbe reviewed. Milham and Boni (1976) infer, from the large mass of heavy water in eachreactor, that the released tritium should be in the form of HTO.

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

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

The SRS environmental surveillance program does not consider OBT in food and airsamples. The airborne tritium sampler collects water vapor on silica gel. Tritium infoodstuffs, vegetation, and fish is measured with an LS counter in water extracted bydistilling or freeze-drying. Special environmental studies provide information on OBT in theSRS environment but not on OBT in releases from SRS. Any observed OBT may be presentas a result of conversion of HTO to OBT or HT to OBT (via HTO) in soil, plants, andanimals (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 thesample 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 duringashing, 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 specificactivity ratios of OBT/HTO with averages near 1.0 in six organs and of 0.6 in fat (Evans1969). 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 changingpattern of environmental tritium levels with time. The former studied loblolly pine trees atvarious distances on site from the tritium production facility; the latter analyzed a sweetgumtree with roots at a groundwater seep line. The tritium concentrations in the tree rings appearto have responded to tritium in both groundwater and air. Tritium in pine litter and the soilbeneath had OBT/HTO specific activity ratios generally between 1 and 4; the uptake patternsuggested atmospheric HTO as source of tritium in pine needles; conversion to OBT in pineneedles; and transfer of this OBT to soil (Sweet and Murphy 1984; Murphy 2000). During1991, in five locations in Four Mile Creek, the OBT/HTO specific activity ratios werebetween 12 and 0.9 in grass and leaves, and between 2 and 0.6 in the fish. On average, bothHTO and OBT concentrations in the bass were slightly below those in the water, while grassand 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 tritiumrelease. It is not known if earlier discharges from the reactors contained significant quantitiesof OBT in addition to the reported total releases of airborne HTO. It is also unknown if OBTcontinues to be discharged from various minor sources at SRS. Tritium in liquid effluentshas been measured as total tritium and will have included any OBT.

Further efforts may provide some of the information that is now missing. Additionaldocument searches may yield at least an upper limit on the fraction of OBT - or of totaltritium including OBT - that was released from reactors as airborne effluent. Comparisonof effluent monitoring by BTM with data from earlier tritium air effluent monitoring systemsthat 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 reactorswere shut down. Direct measurements of the OBT fraction in effluent air and water at leastcould determine the current fraction of tritium discharged as OBT at this time. Routinemonitoring of OBT in milk, foods, and air could provide a measure of the OBT intake bypersons 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 offsite. Estimated doses reflect the considerable dilution of both airborne and waterbornereleases because of the large distances between the stacks and the SRS boundary and of thedilution by Savannah River water.

The estimated annual radiation doses since 1954 to the maximally exposed person off sitefrom airborne tritium as HTO are shown in Figure 3.5 (Arnett and Mamatey 2000). Theyreached an effective dose equivalent of 18.7 mSv (1.87 mrem ) in 1958 and decreased to lessthan 0.2 mSv (0.02 mrem ) in 1999. The annual radiation doses from tritium fromwaterborne effluents reaching public water supplies reached a maximum of 3.3 mSv (0.33mrem) in 1965 (at Port Wentworth); the dose was 0.6 mSv (0.06 mrem) in 1999. Theestimated 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 beencalculated for the combined effect of inhalation, absorption through the skin, and ingestionof food and water.

The dispersion of airborne tritium from all SRS sources is calculated for various distancesin 16 directions with a dispersion model that is based on atmospheric conditions compiledover a five year period. The tritium is assumed to pass into the ecosystem as part of thewater cycle. Radiation doses to persons are then calculated from tritium concentrations inintake media, intake rates, dose conversion factors and other generic or site-specificparameters. The highest annual dose value beyond the fence line is assigned to themaximally exposed person. Doses due to inadvertent tritium releases are usually calculatedseparately 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 thesame year's annual average of the atmospheric dispersion parameters would be substantiallydifferent from those estimated on the basis of the 5-year running average of the dispersionparameters (Kock and Hamby 1998). The parameters involved were the wind direction,wind speed, and atmospheric stability class for each of 16 directional sectors. The resultingradiation doses to the maximally exposed person were between 0.8 and 1.2 of the reportedannual doses based on the 1987 - 1991 five-year composite set of parameters.

The influence of the statistical distributions of selected model parameters on the estimatesof dose to the maximally exposed off-site adult from tritiated water in ground-level air hasalso been evaluated by Hamby (1992). Distributions (such as normal, log-normal, orGaussian) were selected for each transfer, intake, and dose parameter. The median of thedistribution 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 approximately4.4 times lower and higher, respectively, than the median (Hamby 1993; Hamby 2000.) Inthis study, the major contributors to the dose distribution were the distributions assigned todosimetric parameters.

Radiation doses from tritium in surface streams and in groundwater have been calculated formaximally exposed persons off site at three locations. Two locations are Beaufort-Jasper inSouth Carolina and Port Wentworth in Georgia, where public supplies provide drinkingwater. The third location is the Savannah River just below SRS, where fish may be caughtfor consumption and water intake is possible although no water supply as such exists. Theestimated annual doses were consistently of similar magnitude (Murphy et al. 1993). Dosesfor the third location, which generally are slightly higher than those at the other two, wereshown 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 maximallyexposed adult at SRS. The main tritium exposure pathway for infants was considered to bemilk consumption, which was 54% of the estimated dose. For adults, the two main pathwayswere estimated to be vegetable consumption (44%) and inhalation plus skin absorption(41%).

Doses to maximally exposed individuals have been estimated by SRS staff from theconcentrations of tritium measured routinely in air, drinking water, fish and terrestrial foodsenvironmental media as a check on the values of doses calculated from releases. They reportthe dose estimated for the maximally exposed persons from drinking water and from eatingfish but do not report dose estimates based on the low concentrations measured in air andterrestrial 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 havebeen contaminated by tritium from SRS, estimates of doses to individual members of thepublic 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 beencontaminated by tritium in the local environment and that they are representative of theirfood groups, then the doses from the measured concentrations of tritium in Table 3.2 maybe 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 providesthese estimates. The values for moisture content (m), water equivalent factor (f), and theratio 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 subjectto contamination by tritium in the local environment, an alternative estimation of the OBTdose is to take the mean concentration of tritium in the moisture in these foods and estimatethe 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 actuallyobserved, we conclude that the annual dose from food will have been less than 0.1 mSv withless than half of this coming from OBT ingested in the food. The annual dose from drinkingwater with the tritium levels reported would be about 0.4 mSv. Even if the extremeassumptions are made by taking the highest observed concentrations in measured items, theannual 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. #