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


1.1 Statement of the Problem

The Agency for Toxic Substances and Disease Registry (ATSDR) is responsible forconducting public health assessments of communities adjacent to hazardous waste sites. Aspart of this public health assessment process, ATSDR directly queries area residents andcommunity groups to determine if they have any specific health concerns that may be relatedto the hazardous waste site. During the health assessment at the Department of Energy'sLawrence Livermore National Laboratory (LLNL), community members expressed specificconcerns related to the environmental monitoring and dose evaluation of tritium.

Their concerns are that existing tritium monitoring procedures are primarily for tritium in theform of water (HTO) and do not measure the tritium in an organic form ( i.e., as organically-bound tritium or OBT) that may be present in the environment and in foods. Radiation dosesfrom this form of tritium may therefore be missed. There is also the concern that thebiological effect of radiation doses from OBT may be underestimated. Specifically, thesecommunity members have requested that ATSDR conduct or recommend to other regulatoryagencies the direct sampling of OBT in the LLNL environment.

The community concerns about OBT evaluation arise from the known longer retention timeof tritium as OBT in the body compared with that as tritiated water (HTO) (Hill and Johnson,1993; Diabate and Strack, 1993) and the current uncertainty in the relative effectiveness ofa given radiation dose from tritium in the organic form compared with the same value ofdose from tritium in the water form (Straume and Carsten, 1993). The concern is that currentrisk factors underestimate the risk from OBT. These factors, combined with the paucity ofdirect measurements of OBT in the environments surrounding DOE facilities that areundergoing ATSDR public health assessment, require ATSDR to address how environmentaltritium exposures are evaluated.

The seemingly simple community request for direct OBT monitoring at LLNL also presentsseveral areas of uncertainty. Specifically, ATSDR believes the following questions must be answered before environmental OBT measurements will provide useful information forevaluating tritium doses:

  1. Do the potential total tritium exposures, including OBT, present doses ofpublic health concern that require the collection and analysis of additionalenvironmental OBT data?

  2. If such data are warranted, what environmental media should be sampled,how would such sampling be integrated with ongoing tritium monitoring, andwhat methods should be used to measure what are likely to be very lowconcentrations of environmental OBT?

ATSDR convened a panel of experts in the fields of tritium analysis and dosimetry toevaluate site specific tritium monitoring and evaluation programs and determine whetheradequate data and dose evaluation models currently exist for assessing the public healthimplications of tritium exposure and uptake at those sites. The DOE sites chosen for theseevaluations are the LLNL because of the community concern about tritium exposures and theSavannah River Site (SRS) because it was the primary DOE tritium production site withcommensurately large environmental releases and monitoring efforts to document tritiumtransport and fate in the environment.

ATSDR also decided to include a review of the SRS site in the overall problem. AlthoughATSDR is not currently developing the SRS public health assessment, this review of tritiummonitoring and evaluation would be relevant to the planned SRS assessments. Also, thesignificantly greater magnitude of tritium releases at SRS magnifies the potential for anyhealth effects from such releases and should make environmental OBT easier to detect. Additionally, past studies of the environmental fate and transport of tritium at SRS mayprovide useful information for assessing the importance of OBT in environmentscontaminated with tritium.

This report is the collective effort of the five panel members. Following this statement ofthe overall problem is a brief explanation of how the expert panel was selected and how ithas defined and addressed the overall problem. The next section of the report presentsreviews of the important issues and uncertainties related to tritium analysis and dosimetryincluding the chemical forms of tritium, OBT in the environment and foodstuffs, and doseconversion coefficients for HTO and OBT. After this introductory review of tritium issues,specific information on the monitoring programs, estimated releases, environmentalconcentrations, and estimated doses at SRS and LLNL are presented. The calculations oftritium doses at each site include estimates based on currently accepted dosimetry modelsand explicitly include doses from OBT.

The subsequent sections discuss the uncertainties in the estimation of risk from tritium in itsvarious forms and lead to an estimation of the risk to public health from current levels oftritium in the respective local environments, including that from OBT. The final sectionpresents the conclusions and recommendations related to tritium dose and health riskassessment at LLNL and SRS. The overall goal of this report is to identify the specific stepsof the tritium assessment process, the data and models needed, and the uncertaintiesassociated with each step. It is hoped that this will help to identify whether improvementsin tritium monitoring are needed, evaluate whether procedures for assessing the doses andrisks from environmental tritium assessment are appropriate, and will help ensure that the site-specific environmental tritium programs are protective of public health.

1.2 Expert Panel Process

A brief ATSDR review of tritium analytical procedures and published values ofenvironmental tritium at LLNL indicated that direct sampling of OBT using currentmonitoring and dosimetry procedures would very likely be inconclusive with respect to bothOBT activities and the resulting doses. Consequently ATSDR determined that a technicalreview of the site-specific tritium monitoring programs and potential magnitude of OBTactivities and dose evaluation procedures was the most appropriate response to the LLNLcommunity concerns.

Upon determination that a technical review of tritium monitoring and dosimetry by an expert panel was an appropriate response to the community concerns, ATSDR developed thefollowing topics to focus the panel review:

  1. Review of LLNL and SRS tritium sampling programs for measuring environmental tritium with specific emphasis on quantification of organically bound vs. inorganic tritium and potential human uptake;
  2. Assessment of ATSDR procedures for evaluating the health effects of tritium exposure with emphasis on utilizing the most appropriate value of relative biological effectiveness;
  3. Evaluation of probability-based calculations vs. conservative point values in assessing dose and health effects from tritium exposure.

Using the subject area of these topics, panel members were selected based on theirqualifications and experience in the following specific areas of expertise:

  1. Radiochemistry of Tritium
  2. Sampling and analysis of tritium or hydrogen in environmental media
  3. Human uptake of environmental radiological contaminants
  4. Partitioning of organic hydrogen compounds in humans and animal models
  5. Dosimetry of organically- bound tritium, health effects of tritium (or other low energy beta emitters)
  6. Uncertainty analysis

Site community members were provided an opportunity to review and comment on the topicareas and specific areas of expertise listed above prior to the selection of the final expertpanel. The panel members, SRS and LLNL, ATSDR, and community representatives meton October 31 and November 1, 2000 to discuss the problem and outline the proposed report.This document represents the collected reviews and conclusions of the panel members to theoverall problem of environmental tritium health assessment with special emphasis on theuncertainty of OBT measurement and dosimetry. Additionally, this report includesrecommendations for ATSDR to pursue in evaluating potential environmental tritiumexposures.

1.3 Background

Most tritium (T) released from LLNL and SRS is generally in the elemental form (tritiatedhydrogen gas - HT or T2 - referred to here as HT) or the oxide form (tritiated water - HTO,DTO or T2O - referred to here as HTO). Some tritium may also be released in variousorganic chemical forms, mostly as tritiated methane gas (CH4-xTx), but other more complextritiated organic compounds might be present. In the organic compounds tritium may bebound to oxygen, nitrogen, sulphur or carbon atoms. When bound to any of the first threethe tritium is exchangeable with hydrogen in water to varying extent, depending on themolecular structure. The tritium bound to carbon is non-exchangeable (and is sometimesexplicitly called non-exchangeable tritium, NET) and is only freed by decomposition of theorganic compound. In this report, the NET is the form of tritium that we refer to asorganically bound tritium, or OBT.

In the environment, the chemical form of the released tritium may change. Tritium releasedas HT may be converted to HTO with subsequent transformation to OBT, some released asHTO may become bound as OBT in biota, and some released as OBT may be taken up bybiota or converted to HTO. In water, the OBT may be included in any organic compoundcontaining tritium such as microorganisms, pump oil, cleaning agents etc. People maytherefore be exposed directly to the tritium in its released form and in food and water informs to which it has been converted in the environment. Figure 1 illustrates the varioussource and exposure chemical forms.

Tritium can enter the body in any of these chemical forms. Elemental tritium can be inhaledand HTO can be inhaled, ingested, or absorbed through the skin. OBT in gaseous form (e.g.,CH3T) can be inhaled and other OBT can be ingested as a constituent of food. Some of theinhaled HT may be converted to HTO or OBT in the body and some of the tritium taken inas HTO and as OBT may be converted to the other form in the body. Hence, the dose froman exposure to tritium in any particular form will be a composite of doses from tritium in thebody in a variety of chemical forms (see Figure 1).

To assess the doses from tritium released from various facilities, estimates of the doses delivered in the body by tritium in its various forms are needed (DHT, DHTO, and DOBT in Figure 1). The dose from inhaled HT is several orders of magnitude less than that from the same amount of tritium inhaled or ingested as HTO. Generally, for conservatism in assessments, any released tritium is often assumed to be in the form of HTO and often only the doses from intake of tritium as HTO are considered. For more realistic assessments though, estimates of the relative emissions in the HTO and OBT forms are needed as well as the extent to which OBT might be present in food. Total doses from tritium releases may be underestimated if they are based on measurements of tritium released only in the HT and HTO forms or on measurements of HTO in the environment (which do not include the OBT components), or if dosimetric models consider tritium only as HTO.

The questions faced in this review, therefore, concern the extent to which existingmeasurements provide sufficient information for the estimation of the total doses from allforms of tritium that might have been (1) released, (2) distributed and possibly changedchemically in the environment and ingested with food and drinking water, and (3) formedby conversion in the body.

The following section (Section 2) summarizes current understanding of the behavior of thevarious forms of tritium in the environment and in humans, and the dosimetry of the variouschemical forms of tritium. The emphasis is on the formation, behavior and dosimetry ofOBT. This section provides background for the following two sections (Sections 3 and 4)in which we review the published information on the emissions and environmentalmonitoring of tritium released from the Savannah River Site and the Lawrence LivermoreNational Laboratory, and estimate doses to the public from HTO and OBT.

Chemical forms of Tritium responsible for exposures and doses
Figure 1: Chemical forms of tritium responsible for exposures and doses. The shading indicates the tritium components that are generally estimated or measured. The dose effectiveness of HT is much less than HTO or OBT and the current dosimetric assumption is to consider that all inorganic tritium is present as HTO. This report specifically addresses the potential magnitude of the dose from OBT (Dobt).

The estimations of radiation doses follows conventional models and methods and from theseone can assess a measure of radiological impact on health through the so-called "nominalrisk coefficients". These risk coefficients relate the risks of a deleterious impact on healthto particular types of radiation dose. The appropriate risk coefficients for use in assessmentsof dose from tritium in various forms are discussed in Section 5. There are, though,uncertainties associated with such a measure that need to be discussed in evaluating anypossible impact on public health from radiation doses in general and tritium in particular. Such uncertainties arise in the judged validity of the nominal risk coefficient, theappropriateness of the weighting accorded the radiation from tritium in comparison withother radiations, and the possibility of any effects that are peculiar to tritium because of itsincorporation in key biological molecules. These are discussed also in Section 5.

The public health implications of the doses resulting from releases of tritium at SRS andLLNL are discussed in Section 6 and the overall conclusions and recommendations arepresented in Section 7.

Throughout the text the term "dose" generally means effective dose, or committed effectivedose if the dose is that from tritium (ICRP 1992). The effective dose is the dosimetricquantity that reflects the weighting assigned to the absorbed dose to take into account thephysical characteristics of the radiation involved and its relative effectiveness in causingdeleterious biological effects. The committed effective dose is the dose that is eventuallyreceived from a radionuclide - in this case, tritium - until it is eliminated from the body afterintake. An exception is presented in the discussion of risk in Section 5. The effective doseas currently defined is sometimes referred to as the effective dose equivalent or EDE. These quantities are discussed in more detail in Section 5.


2.1 OBT formation in the environment

Information on the physical and biological behavior of tritium in the environment wasreviewed by NCRP in 1979 (NCRP 1979a). This report provides a useful compilation ofcurrently available information pertaining to elemental (HT) and oxide (HTO) forms oftritium. The current understanding of the processes affecting tritium movement in the localenvironment and, in particular on the processes that result in the formation of OBT fromHTO and from HT released to the environment have been described by Davis et al. (1997). The following paragraphs summarize the main points from that report.

HTO dispersed in the environment can be incorporated into organic compounds to formOBT. This may occur by exchange of tritium with labile hydrogen atoms attached to oxygen,nitrogen or sulphur compounds to produce exchangeable OBT, which is in equilibrium withHTO and behaves in the same manner. Tritium can also enter into stable bonds with carboncompounds through various metabolic processes, primarily photosynthesis. This results inthe formation of non-exchangeable OBT, which has a much longer retention time in plantsand animals than HTO. Plants with a high organic content may have a large fraction of theirtotal tritium content in the form of OBT. OBT is formed only in the green parts of plants,but can be translocated to edible fruits. OBT concentrations are reduced by slow conversionback to HTO and, if the exposure to tritium is not a continuing one, by plant growth anddecay.

HT dispersed into the atmosphere can diffuse into the soil and can be converted to HTO byan enzyme-mediated reaction, the rate of which depends on the porosity, water content andmicrobial activity of the soil (Dunstall et al. 1985, Taeschner et al. 1988). The converted HTis subsequently transported as HTO.

Tritium released to fresh water systems is transported by the receiving waters with itsconcentration determined by dilution, dispersion and evaporation. Many aquatic organismsare totally immersed in water and all have high water exchange rates. Uptake of HTO istherefore very quick and concentrations in tissue become equal to water concentrationswithin minutes or hours (Blaylock et al. 1986). Aquatic plants form OBT and HTO throughphotosynthesis but production rates and concentrations are lower for OBT than for HTO. Fish and invertebrates also convert HTO to OBT and can incorporate OBT taken up throughingestion. Tritium concentrations in organisms ingesting OBT are higher than those resultingfrom intake of the same amount of tritium as HTO.

Dynamic models of environmental tritium transfer have recently been tested in aninternational comparison (SPRI 1996). The conclusion from the study was that although theconceptual model of tritium behavior in the environment was generally agreed, thesignificance of some transport processes was still a matter of debate and those processes thatare accepted as important are often modeled using different approaches. This can causesubstantial differences in model predictions. For example, little is known about uptake ofHTO and formation of OBT in the absence of light. Other processes that contributedsubstantially to variation in results included deposition of HTO from air to plants and soil,OBT production in dairy and beef cattle and HTO transport through the soil.

Existing studies indicate that OBT can be expected in foods produced in environmentscontaminated with tritium that has been released as HTO or HT. The processes that resultin the formation of OBT in the environment are broadly understood but there can be largeuncertainties in estimates of the levels of OBT based on models, certainly for brief exposuresto HTO or HT. The results of measurements that have been made in various studieselsewhere can provide some indication of the proportion of tritium that might be present in foodstuffs in the form of OBT close to the SRS and LLNL facilities. These results are discussed in the next section.

2.2 OBT in foodstuffs contaminated with tritium

Tritium occurs in foodstuffs in three forms: (1) as free water, (2) bound to oxygen, nitrogen,and sulfur atoms, and (3) bound to carbon atoms. As noted in Section 1.3, the second formis exchangeable with hydrogen in water to varying extent, depending on the molecularstructure. The third form, tritium bound to carbon, is non-exchangeable and is only freed bydecomposition of the organic compound. Brown (1988) points out that the exchangeabletritium will be in equilibrium with the aqueous component of the system and should beconsidered part of the aqueous component. It will tend to equilibrate with local atmosphericmoisture. The third form, which is OBT in this report, will initially have the same specificactivity as the tritium in the water since net isotopic fractionation in the biochemicalreactions involved is slight. However the difference in metabolic pathways and turnovertimes of HTO and OBT in foods will result in the ratio differing from unity.

The concern is whether circumstances can arise in which the OBT/HTO ratio is so high thatestimates of doses from ingested foods that consider only the tritium in the form of HTOsignificantly underestimate the actual dose. Given the quantitative uncertainties associatedwith the behavior of tritium in its various chemical forms in the environment, directmeasurements of tritium in foods may be needed for confidence in the assessment of thelevels of OBT.

Tritium is also taken into animals, which may be consumed by humans, via inhalation,ingestion of food and water, and percutaneously (Robertson 1973, van den Hoek et al. 1979). Losses occur through respiration, sweating, excretion and (in the case of cows) milking. HTO diffuses freely and rapidly across all cell membranes and equilibrates with body fluidswithin minutes. A small fraction of HTO taken into the body is converted to non-exchangeable OBT. Some of the OBT ingested with food is broken down to HTO duringdigestion and assimilation. OBT can be localized in the body in a relatively small numberof cells and at relatively high concentrations. The constant synthesis of organic compoundsin dairy cows results in the continuous formation of OBT, which is excreted in milk, urineand feces at a rate only slightly slower than that of HTO. OBT is retained longer in bloodand hence in meat.

Brown (1988) provides a comprehensive list of 71 reports and papers that bear on OBT. Particularly important were those reporting the series of measurements by Koenig et al.(1987) and by Hisamatsu et al. (1987) who directly measured OBT and HTO in the rangesof food items. Brown's conclusion was that ratio of specific activities of OBT/HTO infoodstuffs was generally greater than unity but the mean ratio in a diet would be unlikely toexceed 2. Much greater ratios were observed in some studies but these studies were eithercarried out when HTO and HT levels in the environment were fluctuating widely as a resultof thermonuclear weapons tests in the atmosphere (e.g., Bogen et al. 1979) or there seemedto be problems in the analyses. In the former, the persistence of OBT in the environmentfrom the preceding decade when environmental levels were generally high led to theobservation of high ratios in the next decade. It was also apparent that meat would not attaina specific activity ratio for the two forms greater than that of the food fed to the animals,based on the results of experimental feeding.

More recent surveys of OBT/HTO specific activity ratios in foods have strengthened theseconclusions. The most extensive has been that of Brown (1995). He measured 77 food itemsobtained from three areas with different average concentrations of tritium in atmosphericmoisture and precipitation and obtained from supermarkets in one of the areas. Averageatmospheric concentrations ranged from <0.1 Bq/m3 to 40 Bq/m3 (< 3 pCi/m3 to 1000pCi/m3) and precipitation from 5 Bq/L to 3000 Bq/L (0.1 nCi/L to 100 nCi/L)(2). (As will beseen, the low end of these ranges encompasses the ambient concentrations observed in theLivermore Valley.) The organic material and the water in all samples were separatelymeasured by 3He mass spectrometry and the water samples were also measured by LS counting.

Table 2.1 shows the average values obtained in the various foodstuffs sampled. The overallaverage value for the ratio of the specific activities of OBT to HTO (OBT/HTO) was 1.25+ 0.75. This was similar to values reported by others (Koenig et al. 1987, Hisamatsu et al.1987). The range of values was 0.4 to 4.3; the ratios for vegetables tended to be higher than1.5, while those for meats tended to be lower than 0.8. The variability of the ratio precludedspecific values being assigned to particular foods or locations. The high value of 2.8 (for eggs) was attributed to a particularly low value of HTO concentration.

Further measurements of OBT and HTO, reported by Kim, et al. (2000), have specificactivity ratios in various foods including, rice (0.88; range = 0.58-1.44), chinese cabbage(1.34; range = 0.58-2.61), radish, 1.01 (range = 0.52-3.03), and green onions 1.02 (range =0.45- 2.00). We conclude that a reasonable default value for the ratio of the OBT/HTOspecific activities, in foodstuffs would be somewhere between 1 and 2 (say 1.2) with a rangefrom 0.5 to 3. Note that the actual ratio of OBT/HTO specific activities in any given foodwill generally be less than this ratio because of the high moisture content in foods (grains arean exception) and the lower proportion of hydrogen in organic materials (except fat) than in water.

Item Average Values: Bq/L water equivalent
Area 1 Area 2
Area 2
(local produce)
Area 3
OBT 8.2 11.4 157.6 3,245
HTO 4.9 9.9 97.9 2,136
OBT/HTO 1.67 1.34 1.66 1.70
OBT   6.1 172.7 3,839
HTO   6.7 151.0 2,870
OBT/HTO   0.99 1.16 1.34
OBT 5.2 8.1 18.0  
HTO 4.3 15.0 37.8  
OBT/HTO 1.19 0.52 0.56  
OBT 9.2   67.6  
HTO 2.2   45.3  
OBT/HTO 4.23   1.49  
OBT 9.0   25.3  
HTO 3.4   63.9  
OBT/HTO 2.80   0.40  

Table 2.1 Results of a study in Canada by Brown, 1995, which measured OBTand HTO in foods produced in three areas that had different levels of environmentalcontamination by tritium and in foods from supermarkets in one of those areas. Thewater equivalent = [dry weight] x [water equivalent factor]. The water equivalentfactor of dry material was taken as {(% protein x 0.07) + (% fat x 0.12) + (%carbohydrate x 0.062)}/100 x {18/2}.

2.3 Relative daily intakes of HTO and OBT in foodstuffs from an environment contaminated with tritium.

Most measurements of tritium in environmental media and in foods have been of tritium inthe form of water. As will be seen later, this is the case for the environments around SRSand LLNL. Hence, although there may be a few samples in which OBT has been measured,it will be helpful for this discussion to estimate what the intake of tritium in the form of OBTin foods might be if there are only estimates of the concentration of tritium, in the form ofwater, in some foodstuffs and in some drinking water supplies.

The last section discussed the values that have been observed for the ratio of the OBT/HTOspecific activities in food grown in, or exposed to a tritium-contaminated environment. Default values of this ratio and default food ingestion rates are the links to estimating theintake of OBT in food when there are only estimates of the concentration of tritium in water.

If the measured concentration of tritium in a food is C, in Bq/kg wet weight and the dailyintake of the food is I, in kg/d, then the daily intake of measured tritium with that food is Ix C Bq/d. If we assume that only HTO has been measured in this food, then theconcentration of tritium as HTO in the water will be C/m Bq/kg where m is the fraction ofthe food that is water.

In any food, the proportions of hydrogen in proteins, fats and carbohydrates differ from thatin water, so, for any food, a "water equivalent factor", f, can be estimated for the organiccomponent. This is the mass of water that would have the same mass of hydrogen as theorganic material of the food.

In the previous section, typical ratios for the ratio of specific activities as OBT and as HTOwere suggested. If such a value is designated R, then the concentration of tritium in the massof water equivalent to the organic material will be R x (C/m) Bq/kg. The concentration inthe organic component itself will be f x R x (C/m) Bq/kg.

The intake of tritium as OBT would then be: (1-m) x f x R x I x (C/m) Bq/d.

In summary, if A(HTO) is the activity intake rate of tritium taken in with an intake of I kg/d of food and A(OBT) is the activity intake rate as OBT in the same food, then:

A(HTO) = I x C Bq/d(1)
A(OBT) = (1-m) x f x R. x I x (C/m) Bq/d.(2)

Brown (1995) provides values of m and f for a variety of foods, the latter estimated from thefractional hydrogen content of protein (7%), fat (12%) and carbohydrate (6.2%) and thecomposition of the respective food items.

Values of the daily intakes of food [I] and of the daily intake of hydrogen may be estimatedfor individuals of various ages from data given in ICRP (1975) and a more recent report fromHealth Canada (1994). For example, the intakes of food and of drinking water (includingdrinking water-based beverages) by an adult are given as 1.6 kg/d and 1.5 kg/d respectively. Also, the daily intakes of hydrogen in food and fluids are given as 0.350 and 0.245 kg/d foradult males and adult females respectively, an average of ~ 0.3 kg/d. Hence, since this intakerate of drinking water will account for 0.17 kg/d of hydrogen intake, in the 1.6 kg/d of foodintake there is 0.14 kg/d of hydrogen or about 8.3% of the food intake. We can also make thesimplifying approximation that the hydrogen intake in food is proportional with the foodintake of individuals at various ages. By interpolation from the two above references, thevariations of food and hydrogen intake with age for the age groups, corresponding to thoseselected by the ICRP for dosimetric purposes (ICRP 1996), are as shown in Table 2.2.

The above formulae and the values in Table 2.2 provide sufficient information to estimatean individual's intake of tritium as HTO and as OBT in food, based on some sampling ofmoisture in the environment and foods. If a more detailed analysis is warranted, the more detailed diet in Health Canada (1995) or locally specific food consumption could be used.

Intake (kg/d)Age
3 months1 year5 years10 years15 yearsAdult

Table 2.2 Estimated daily intakes of food and of hydrogen. Values were obtained by interpolation from ICRP (1975) and Health Canada (1994).
* Water intake is drinking water plus made-up beverages.

2.4 Doses from intakes of tritium as HTO and OBT

The behavior of tritium inhaled, ingested or taken in percutaneously as HTO has been wellstudied in humans and the dosimetry is straightforward. Most of the tritium is retained in thebody with a half-time in the range 5 to 15 days with the average being about 10 days (Pinsonand Langham 1957, Butler and Leroy 1965, Osborne 1966). Hamby (1999) analyzed theresults of reported biokinetic studies with HTO in humans and concluded that the distributionof estimates of dose per unit activity intake of tritium as HTO, based on a singlecompartment model, was characterized by a geometric standard deviation of 1.4.

A small fraction of the tritium that enters the body as HTO becomes bound in organiccompounds through biochemical and metabolic processes and is retained for much longer. Observed retention curves can be fitted with one or more exponential components with half-times of tens of days to several hundred days. A single half-time of 40 days is often chosento represent this longer component. The important parameter, however, is the fraction (about 30%) of organically-bound hydrogen in the body that can be replaced by tritium fromHTO; this puts a bound of about 10% on the relative dose from tritium as OBT that is formedfrom HTO (Osborne 1972). A recent study by Trivedi et al. (1997), exploiting data frompeople who had been exposed to HTO, definitively verified this long-standing generalconclusion that the contribution to dose from tritium as OBT was less than 10% of that fromHTO when the intake was tritium in HTO.

The situation is more complicated when some of the tritium that is taken in is already in theform of OBT. A variety of studies have provided data from which estimates of the dosesfrom tritium in the OBT form can be compared with those from tritium in the form of HTOwhen tritium is taken in both forms. Table 2.3 illustrates some of the estimates from studies with animals.

Source of Tritium% increase indose from OBT Reference
tritium contaminated environment
(air, water, food)
40 to 50Evans 1969
in all normal diet100Commerford 1984
in fresh plants16 to 35Myers and Johnson 1991
in dried plants40 to 100Pietrzak-Flis et al. 1978
in dried milk50 to 120Kirchmann et al. 1977
as tritiated proteins infoods400Commerford et al. 1983

Table 2.3 Percentage increase in dose from tritium as OBT above that from tritium in the form of water (HTO), estimated from studies with mammalian animals exposed to tritium for long times.

Evans (1969) has shown that, in deer exposed to a tritium environment over the long-term, organically bound tritium in various organs was appreciable. Evans calculated thatthe upper-limit to the increase in dose for humans exposed to tritium as a result of theorganic binding was about 40% to 50% over that assumed by the then current methodsthat considered only the dose from water.

This is similar to the conclusion of Thompson and Ballou (1956) that organically-boundtritium in rats exposed to HTO for half of a year was equal to about 20% -30% of theactivity in body water. Laskey et al. (1973) determined that if only tritiated water wereconsumed, the specific activity of OBT in rats would be about 25% of the water-specificactivity. Koranda and Martin (1973) found that the total tritium in kangaroo rats exposedchronically to environmental tritium was, on average about 50% greater than the tritium in body water. These studies all tended to show that the organically bound tritium in thebodies of animals exposed long-term to environments contaminated with tritium wouldreach 20% to 50% of the activity present as HTO.

These types of study have been reviewed extensively by Myers and Johnson (1991), Hilland Johnson (1993), Diabate and Strack (1993) and, most recently, by Richardson et al.(1998). It is apparent that long-term exposures to HTO-contaminated environments inwhich tritium will be present in foodstuffs as both HTO and OBT may result in dosesfrom tritium in the form of OBT that are a substantial fraction of the dose from thetritiated water itself. The relative magnitudes of the doses will depend on the specificactivities of the tritium in the food items, daily intakes, oxidation rates, gut absorptionrates, and retention and transfer rates in the tissue components. For specific tritiatedorganics (such as proteins in Table 2.3), the dose associated with OBT can be severaltimes that associated with HTO. Ingestion of tritiated nucleic acid precursors (e.g.,thymidine) can result in doses from OBT 5 - 9 times greater than that from tritium takenin as HTO (Balonov et al. 1984).

There have been only a few studies directly with humans who have been exposed totritium as OBT in diet. Richardson et al. (1998) in their review of studies by Bogen et al.(1979), Belloni et al. (1983) and Hisamatsu et al. (1989) showed that they indicated thatincreases in dose could be 12% -106% over that from tritium as HTO, though a morelikely range (based on an indication that one study may have had problems withcontamination) was 12% - 32%.

A variety of dosimetric models have been developed to reflect the observed behavior oftritium taken in as HTO and as OBT. The simplest (Crawford-Brown 1984) is a variantof one for HTO intake that had an HTO compartment linked to a second OBTcompartment. Crawford-Brown added a direct input of OBT into the OBT compartmentfrom which there is bi-exponential clearance that varies with age (see Figure 2.1). Fiftypercent of the intake of OBT is assumed to be incorporated as OBT; the other 50% isinitially oxidized to HTO. A recent review by Richardson et al. (1998) of the literatureon the metabolism and dosimetry of OBT-containing foods concluded that the 50%incorporation of ingested OBT was at the high end of the values observed, which weregenerally in the range of 9% to 45%.

Model of the behavior of Tritium ingested as HTO and OBT
Figure 2.1 Model of the behavior of tritium ingested as HTO and OBT (Crawford-Brown, 1984)

Other models have additional compartments. Examples are NCRP (1979a), which has threecompartments in total, including two for OBT; Killough (1981), five total with two OBT andtwo bone compartments; Etnier et al., (1984), four total with three tissue solidscompartments that handle carbohydrates and fats differently; Belloni et al. (1985), three totalwith direct input into two OBT compartments; and Saito (1992), three total with direct inputinto only one of two OBT compartments. There are various values for the transferparameters between compartments in the various models. None of the models isphysiologically based so they are all limited in their validity beyond the data to which theyhave been fitted. The simplest (Crawford-Brown, 1984), although having only one OBTcompartment, serves to indicate the key dosimetric consequences when intakes of both HTOand OBT occur.

The values chosen by Crawford-Brown for the parameters, K1 - K4 were based on biologicalhalf-times of 1,000, 33, 100, and 10 days, respectively, for adults. The radioactive decayconstant, l, has little effect since the half-life of tritium is so much greater than the biologicalhalf-times. The values of fS (0.90) and fL (0.10) indicate the amount of tritium released fromorganically bound sites with relatively short and long half-times, respectively. Agedependency is assumed for individuals younger than 21 years, after which all adults areassumed to possess the same retention characteristics.

The quantity that is important for dosimetry is the integrated activity of tritium in the body- the becquerel-days. The pattern of intake in time is not important for determining the totaldose; only the dose rate. The effective dose is a simple multiple(3),(4) of this quantity(1.25 pSvBq-1d-1 or 4.63 remCi-1d-1). The integrated activity can be estimated for both formsof tritium when the intake is either as HTO or as OBT or is a mixture of the two. Figure 2.3shows the variation of the integrated activity for various intake mixtures against age ofindividual.

For an intake of 1 Bq tritium as HTO, the integrated activity is 14.4 Bqd, corresponding toa dose of 18 pSv (67 rem for a 1 Ci intake). With this model, the integrated activity (and,hence, dose) for an intake of tritium as OBT is 2.7 times greater for an adult than the valuefor an intake as HTO. For an infant, the dose is 2.5 times greater. The actual dose that isdelivered by HTO after an intake of OBT is the same as that delivered after an intake of thesame activity of HTO (since, in this model, all the tritium goes through this HTOcompartment). The dose delivered by tritium in the form of OBT after the same intake ofOBT is 1.7 times that from tritium in the form of HTO. This distinction is important for anyconsideration of the possible difference in effectiveness of the tritium in these two forms (seeSection 5). Table 2.4 provides a summary of these relative doses for the different chemical forms of tritium.

Graph of integrated activity versus age
Figure 2.3. Variation of the integrated activity with age of individual from 1 Bq tritium taken in as a mixture of HTO and OBT with the percentages as OBT as shown on the lines. The model is that of Crawford-Brown (1984). The ratios of dose from 100% OBT intake to dose from 100% HTO intake are shown for infants and for adults.

From these estimates of dose from given intakes of HTO and OBT, we can estimate thedosimetric consequences of ingestion of food from tritium-contaminated environments. InSection 2.3, the daily intakes, A, of tritium in the two forms, HTO and OBT, if the food has C Bq/kg wet weight and intake rate I kg/d were given by

A(HTO) = I x C Bq/d(1)
A(OBT) = (1-m).f.R.I.(C/m) Bq/d.(2)

Form of tritium and component considered

Relative dose from intake as:
Dose from tritium as HTO in the body 1 (reference) 1
Dose from tritium as OBT in the body 0.035 1.7
Dose from total tritium in the body 1.035 2.7

Table 2.4. Contributions of tritium in the forms of HTO and OBT to integrated activity (and, hence, dose) after intakes of unit activity of tritium as HTO or as OBT. The contributions are expressed relative to that from tritium as HTO after an HTO intake. The estimates are based on the Crawford-Brown compartment model (Crawford-Brown, 1984)

If we take the dose per unit intake of tritium as HTO to be EHTO pSv/Bq (18 pSv/Bq for anadult in the model above), and the dose per unit intake of tritium as OBT to be EOBT pSv/Bq,then the dose, D, from a day's intake will be given by the sum of the doses, dHTO and dOBT,from the tritium intake in the form of HTO and OBT respectively;

D = dHTO + dOBT pSv
D = I x C x.EHTO + (1-m).f.R.I.(C/m). x EOBT pSv
    = I x C x { EHTO + ((1-m)/m) x f x R x EOBT} pSv

By way of example, for a food such as milk, the moisture content can be taken as 0.9, thewater equivalent factor as 0.75 (Brown, 1995). If unit rate of ingestion (I = 1 kg/d) and unitconcentration in the milk (C = 1 Bq/kg) are considered, then, with R =1.2, and values forEHTO and EOBT as given by the Crawford-Brown model (18 and 47 pSv/Bq respectively):

D = 18 + 4.7 pSv
     23 pSv from a daily intake of 1 kg of milk at 1 Bq/kg.
           (85 mrem for intake of 1 kg of milk at 1 mCi/kg).

Hence, in this example, when OBT is taken into account, the dose is increased by 26% abovethat resulting from the intake of just the tritium in the food as water. Note, however, that,in the model explored here, 37% of the dose from OBT that is taken in is actually deliveredby tritium as HTO ( 1 / 2.7 in above Table). Therefore, in the example, of the 4.7 pSv fromthe OBT intake, 1.7 pSv would be from HTO in the body; 3 pSv from OBT in the body. With rounded values, the dose delivered by HTO in this example (DHTO in Figure 1) wouldbe 18 + 1.7 20 pSv and the dose delivered by OBT (DOBT in Figure 1) would be 3 pSv (74 mrem and 11 mrem for intake of 1 mCi in milk).

Clearly, for different foods that have different moisture contents and proportions of fat,carbohydrates and protein, the relative doses from the two forms of tritium will be different. For example, with the tritium-related parameters as above (e.g., R = 1.2), for meat (with m= 0.65, f = 0.8) the dose would be increased by 135%, and for vegetables (with m = 0.9, f =0.55) the dose would be increased by 19%. Note, though, as pointed out in Section 2.2, theR-values for meat and vegetables are typically less than 0.8 and in the range 1.5-2respectively. With these values of R, the OBT-associated increase in dose from tritiumwould be 90% and 30% respectively.

In the model, the ratio of the values of the integrated activities does not change much withage. It follows that these relative increases in dose when OBT is taken into account will notvary much either.

The above has been considering the dose from OBT in a macroscopic way. That is,distributions of OBT within cells and tissues have not been considered; the energy depositionhas been averaged over the soft tissues of the body. Since risks from radiation are couchedin the same terms with doses being averaged over tissues the macroscopic approach isgenerally justified. However, for tritium, because of the short range of the tritium beta intissue (maximum 6 mm, mean 0.69 mm) the distribution within tissues, and possibly withincells, may need to be considered.

For example, a study involving the collection and analysis of 52 deer from the SavannahRiver Plant site showed that the distribution of OBT in the bodies was not uniform (Evans1969). Concentrations of OBT varied by factors of almost two between muscle and fat, andthere also appeared to be inter-animal variations of the same order of magnitude. Anotherstudy, involving mice maintained on tritiated drinking water for up to 600 days, found thatthe average tritium content (and thus dose rate) of epididymis and ovary were only about halfof that observed for the blood plasma (Carsten 1979; Carsten and Cronkite 1976).

A general correction factor for estimated doses delivered by tritium in the form of OBT toaccount for variation in labeling between tissues does not seem warranted. However, thedistribution of OBT within cells themselves deserves further comment. Experimentally, itis difficult to distinguish whether a given level of biological effect from OBT, relative to thatfrom some other radiation, is a result of a concentration of the tritium in a cell or cellnucleus, rather than being uniformly distributed, or is a result of an inherent greatereffectiveness of the tritium decaying from a particular organic molecule. If the dose isunderestimated, the effectiveness of the tritium radiation appears to be higher - and viceversa. Because of this complication, this effect, which is peculiar to tritium and otherbiologically active radionuclides with short-range radiations, is included in the more generaldiscussion of risk in Section 5.3.

We conclude, therefore, that for foods obtained from environments that have beencontaminated with tritium, the dose from OBT that is ingested in the food may increase thedose attributed to tritium by not more than about a factor of two, and in most cases by afactor much less than this.

The current internationally recommended values for the doses from tritium taken in as HTOor OBT are based on modeling assumptions that are similar, though not identical, to thosediscussed above. The recommended model (ICRP 1994, ICRP 1996b) for HTO has twocomponents of retention in the whole body with fractions and halftimes for adult clearanceof 0.97 (10 d) and 0.03 (40 d). The recommended model for OBT has two components ofretention in the whole body with fractions and halftimes for adult clearance of 0.50 (10 d)and 0.50 (40 d). The effective doses for ingestion of HTO and OBT by infants (3 monthsold), 1, 5, 10, and 15 years and adults are given in Table 2.5.

These coefficients are taken as being identical for all body organs and tissues. The dose conversion factors are highest in infants, which are between 2.9 and 3.6 times larger than for adults when tritium is taken in as HTO and as OBT respectively. The ratio of the coefficients does not change appreciably with age (see the fourth column). For tritium in the elemental form (HT) and as tritiated methane, the recommended (ICRP 1996a) are four and two orders of magnitude less than for tritium as HTO.

Age Dose Conversion Coefficient pSv/Bq (rem/Ci) Ratio of coefficients
HTO intake OBT intake
3 months 64 (240) 120 (440) 1.9
1 year 48 (180) 120 (440) 2.5
5 years 31 (120) 73 (270) 2.4
10 years 23 (80) 57 (210) 2.5
15 years 18 (67) 42 (160) 2.3
Adult 18 (67) 42 (160) 2.3

Table 2.5 Dose Conversion Coefficients for tritium taken in as HTO and OBT (ICRP 1994, 1996b).

As in the Crawford-Brown model considered above, the ICRP dose conversion coefficientgiven here for OBT lacks a firm physiological base. Nevertheless, we have concluded thatthese nominal values that are in international use are reasonable values to use in thisassessment. The modeling review above provides insight into the variation in the form oftritium from which the doses are delivered.

2 Metric, or SI, units are used in the report with the older units in parenthesis (e.g., Ci, rem). "Soft" rounding is used commensurate with the precision of the values of the quantities.

3 Taking the soft tissue mass as 63 kg; average energy imparted per tritium disintegration as 5.69 keV; radiation weighting factor as 1; and a physical conversion factor of 1.38 x 10-11 kg.Gy.s.kev-1.d-1.

4 The SI prefixes are used in this report to help readability:

k 103;M 106;G 109;T 1012;P 1015;E 1018;
m 10-3;µ 10-6;n 10-9;p 10-12.

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. #