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 for conducting public health assessments of communities adjacent to hazardous waste sites. As part of this public health assessment process, ATSDR directly queries area residents and community groups to determine if they have any specific health concerns that may be related to the hazardous waste site. During the health assessment at the Department of Energy's Lawrence Livermore National Laboratory (LLNL), community members expressed specific concerns related to the environmental monitoring and dose evaluation of tritium.

Their concerns are that existing tritium monitoring procedures are primarily for tritium in the form 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 doses from this form of tritium may therefore be missed. There is also the concern that the biological effect of radiation doses from OBT may be underestimated. Specifically, these community members have requested that ATSDR conduct or recommend to other regulatory agencies the direct sampling of OBT in the LLNL environment.

The community concerns about OBT evaluation arise from the known longer retention time of 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 of a given radiation dose from tritium in the organic form compared with the same value of dose from tritium in the water form (Straume and Carsten, 1993). The concern is that current risk factors underestimate the risk from OBT. These factors, combined with the paucity of direct measurements of OBT in the environments surrounding DOE facilities that are undergoing ATSDR public health assessment, require ATSDR to address how environmental tritium exposures are evaluated.

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

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

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

ATSDR convened a panel of experts in the fields of tritium analysis and dosimetry to evaluate site specific tritium monitoring and evaluation programs and determine whether adequate data and dose evaluation models currently exist for assessing the public health implications of tritium exposure and uptake at those sites. The DOE sites chosen for these evaluations are the LLNL because of the community concern about tritium exposures and the Savannah River Site (SRS) because it was the primary DOE tritium production site with commensurately large environmental releases and monitoring efforts to document tritium transport and fate in the environment.

ATSDR also decided to include a review of the SRS site in the overall problem. Although ATSDR is not currently developing the SRS public health assessment, this review of tritium monitoring and evaluation would be relevant to the planned SRS assessments. Also, the significantly greater magnitude of tritium releases at SRS magnifies the potential for any health 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 may provide useful information for assessing the importance of OBT in environments contaminated with tritium.

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

The subsequent sections discuss the uncertainties in the estimation of risk from tritium in its various forms and lead to an estimation of the risk to public health from current levels of tritium in the respective local environments, including that from OBT. The final section presents the conclusions and recommendations related to tritium dose and health risk assessment at LLNL and SRS. The overall goal of this report is to identify the specific steps of the tritium assessment process, the data and models needed, and the uncertainties associated with each step. It is hoped that this will help to identify whether improvements in tritium monitoring are needed, evaluate whether procedures for assessing the doses and risks 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 of environmental tritium at LLNL indicated that direct sampling of OBT using current monitoring and dosimetry procedures would very likely be inconclusive with respect to both OBT activities and the resulting doses. Consequently ATSDR determined that a technical review of the site-specific tritium monitoring programs and potential magnitude of OBT activities and dose evaluation procedures was the most appropriate response to the LLNL community 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 the following 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 their qualifications 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 topic areas and specific areas of expertise listed above prior to the selection of the final expert panel. The panel members, SRS and LLNL, ATSDR, and community representatives met on 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 the overall problem of environmental tritium health assessment with special emphasis on the uncertainty of OBT measurement and dosimetry. Additionally, this report includes recommendations for ATSDR to pursue in evaluating potential environmental tritium exposures.

1.3 Background

Most tritium (T) released from LLNL and SRS is generally in the elemental form (tritiated hydrogen 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 various organic chemical forms, mostly as tritiated methane gas (CH4-xTx), but other more complex tritiated organic compounds might be present. In the organic compounds tritium may be bound to oxygen, nitrogen, sulphur or carbon atoms. When bound to any of the first three the tritium is exchangeable with hydrogen in water to varying extent, depending on the molecular structure. The tritium bound to carbon is non-exchangeable (and is sometimes explicitly called non-exchangeable tritium, NET) and is only freed by decomposition of the organic compound. In this report, the NET is the form of tritium that we refer to as organically bound tritium, or OBT.

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

Tritium can enter the body in any of these chemical forms. Elemental tritium can be inhaled and 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 the inhaled HT may be converted to HTO or OBT in the body and some of the tritium taken in as HTO and as OBT may be converted to the other form in the body. Hence, the dose from an exposure to tritium in any particular form will be a composite of doses from tritium in the body 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 existing measurements provide sufficient information for the estimation of the total doses from all forms of tritium that might have been (1) released, (2) distributed and possibly changed chemically in the environment and ingested with food and drinking water, and (3) formed by conversion in the body.

The following section (Section 2) summarizes current understanding of the behavior of the various forms of tritium in the environment and in humans, and the dosimetry of the various chemical forms of tritium. The emphasis is on the formation, behavior and dosimetry of OBT. This section provides background for the following two sections (Sections 3 and 4) in which we review the published information on the emissions and environmental monitoring of tritium released from the Savannah River Site and the Lawrence Livermore National 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 these one can assess a measure of radiological impact on health through the so-called "nominal risk coefficients". These risk coefficients relate the risks of a deleterious impact on health to particular types of radiation dose. The appropriate risk coefficients for use in assessments of 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 any possible 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, the appropriateness of the weighting accorded the radiation from tritium in comparison with other radiations, and the possibility of any effects that are peculiar to tritium because of its incorporation 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 and LLNL are discussed in Section 6 and the overall conclusions and recommendations are presented in Section 7.

Throughout the text the term "dose" generally means effective dose, or committed effective dose if the dose is that from tritium (ICRP 1992). The effective dose is the dosimetric quantity that reflects the weighting assigned to the absorbed dose to take into account the physical characteristics of the radiation involved and its relative effectiveness in causing deleterious biological effects. The committed effective dose is the dose that is eventually received from a radionuclide - in this case, tritium - until it is eliminated from the body after intake. An exception is presented in the discussion of risk in Section 5. The effective dose as 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 was reviewed by NCRP in 1979 (NCRP 1979a). This report provides a useful compilation of currently available information pertaining to elemental (HT) and oxide (HTO) forms of tritium. The current understanding of the processes affecting tritium movement in the local environment and, in particular on the processes that result in the formation of OBT from HTO 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 form OBT. 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 with HTO and behaves in the same manner. Tritium can also enter into stable bonds with carbon compounds through various metabolic processes, primarily photosynthesis. This results in the formation of non-exchangeable OBT, which has a much longer retention time in plants and animals than HTO. Plants with a high organic content may have a large fraction of their total 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 conversion back to HTO and, if the exposure to tritium is not a continuing one, by plant growth and decay.

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

Tritium released to fresh water systems is transported by the receiving waters with its concentration determined by dilution, dispersion and evaporation. Many aquatic organisms are totally immersed in water and all have high water exchange rates. Uptake of HTO is therefore very quick and concentrations in tissue become equal to water concentrations within minutes or hours (Blaylock et al. 1986). Aquatic plants form OBT and HTO through photosynthesis 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 through ingestion. Tritium concentrations in organisms ingesting OBT are higher than those resulting from intake of the same amount of tritium as HTO.

Dynamic models of environmental tritium transfer have recently been tested in an international comparison (SPRI 1996). The conclusion from the study was that although the conceptual model of tritium behavior in the environment was generally agreed, the significance of some transport processes was still a matter of debate and those processes that are accepted as important are often modeled using different approaches. This can cause substantial differences in model predictions. For example, little is known about uptake of HTO and formation of OBT in the absence of light. Other processes that contributed substantially 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 environments contaminated with tritium that has been released as HTO or HT. The processes that result in the formation of OBT in the environment are broadly understood but there can be large uncertainties in estimates of the levels of OBT based on models, certainly for brief exposures to HTO or HT. The results of measurements that have been made in various studies elsewhere 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 form is exchangeable with hydrogen in water to varying extent, depending on the molecular structure. The third form, tritium bound to carbon, is non-exchangeable and is only freed by decomposition of the organic compound. Brown (1988) points out that the exchangeable tritium will be in equilibrium with the aqueous component of the system and should be considered part of the aqueous component. It will tend to equilibrate with local atmospheric moisture. The third form, which is OBT in this report, will initially have the same specific activity as the tritium in the water since net isotopic fractionation in the biochemical reactions involved is slight. However the difference in metabolic pathways and turnover times 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 that estimates of doses from ingested foods that consider only the tritium in the form of HTO significantly underestimate the actual dose. Given the quantitative uncertainties associated with the behavior of tritium in its various chemical forms in the environment, direct measurements of tritium in foods may be needed for confidence in the assessment of the levels 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 fluids within 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 during digestion and assimilation. OBT can be localized in the body in a relatively small number of cells and at relatively high concentrations. The constant synthesis of organic compounds in dairy cows results in the continuous formation of OBT, which is excreted in milk, urine and feces at a rate only slightly slower than that of HTO. OBT is retained longer in blood and 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 ranges of food items. Brown's conclusion was that ratio of specific activities of OBT/HTO in foodstuffs was generally greater than unity but the mean ratio in a diet would be unlikely to exceed 2. Much greater ratios were observed in some studies but these studies were either carried out when HTO and HT levels in the environment were fluctuating widely as a result of thermonuclear weapons tests in the atmosphere (e.g., Bogen et al. 1979) or there seemed to be problems in the analyses. In the former, the persistence of OBT in the environment from the preceding decade when environmental levels were generally high led to the observation of high ratios in the next decade. It was also apparent that meat would not attain a 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 these conclusions. The most extensive has been that of Brown (1995). He measured 77 food items obtained from three areas with different average concentrations of tritium in atmospheric moisture and precipitation and obtained from supermarkets in one of the areas. Average atmospheric concentrations ranged from <0.1 Bq/m3 to 40 Bq/m3 (< 3 pCi/m3 to 1000 pCi/m3) and precipitation from 5 Bq/L to 3000 Bq/L (0.1 nCi/L to 100 nCi/L)(2). (As will be seen, the low end of these ranges encompasses the ambient concentrations observed in the Livermore Valley.) The organic material and the water in all samples were separately measured 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 overall average 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 than 1.5, while those for meats tended to be lower than 0.8. The variability of the ratio precluded specific 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 specific activity 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/HTO specific activities, in foodstuffs would be somewhere between 1 and 2 (say 1.2) with a range from 0.5 to 3. Note that the actual ratio of OBT/HTO specific activities in any given food will generally be less than this ratio because of the high moisture content in foods (grains are an exception) and the lower proportion of hydrogen in organic materials (except fat) than in water.

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 OBT and HTO in foods produced in three areas that had different levels of environmental contamination by tritium and in foods from supermarkets in one of those areas. The water equivalent = [dry weight] x [water equivalent factor]. The water equivalent factor 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 in the form of water. As will be seen later, this is the case for the environments around SRS and 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 OBT in foods might be if there are only estimates of the concentration of tritium, in the form of water, 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/HTO specific 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 the intake 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 daily intake of the food is I, in kg/d, then the daily intake of measured tritium with that food is I x C Bq/d. If we assume that only HTO has been measured in this food, then the concentration of tritium as HTO in the water will be C/m Bq/kg where m is the fraction of the food that is water.

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

In the previous section, typical ratios for the ratio of specific activities as OBT and as HTO were suggested. If such a value is designated R, then the concentration of tritium in the mass of water equivalent to the organic material will be R x (C/m) Bq/kg. The concentration in the 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 the fractional hydrogen content of protein (7%), fat (12%) and carbohydrate (6.2%) and the composition of the respective food items.

Values of the daily intakes of food [I] and of the daily intake of hydrogen may be estimated for individuals of various ages from data given in ICRP (1975) and a more recent report from Health Canada (1994). For example, the intakes of food and of drinking water (including drinking 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 for adult males and adult females respectively, an average of ~ 0.3 kg/d. Hence, since this intake rate of drinking water will account for 0.17 kg/d of hydrogen intake, in the 1.6 kg/d of food intake there is 0.14 kg/d of hydrogen or about 8.3% of the food intake. We can also make the simplifying approximation that the hydrogen intake in food is proportional with the food intake of individuals at various ages. By interpolation from the two above references, the variations of food and hydrogen intake with age for the age groups, corresponding to those selected 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 estimate an individual's intake of tritium as HTO and as OBT in food, based on some sampling of moisture 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 months 1 year 5 years 10 years 15 years Adult
Food Total 0.8 1 1.6 1.8 1.9 1.6
Hydrogen 0.07 0.08 0.13 0.15 0.16 0.14
Water* Total 0.75 0.8 0.85 1 1.3 1.5
Hydrogen 0.08 0.09 0.1 0.11 0.15 0.17

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 well studied in humans and the dosimetry is straightforward. Most of the tritium is retained in the body with a half-time in the range 5 to 15 days with the average being about 10 days (Pinson and Langham 1957, Butler and Leroy 1965, Osborne 1966). Hamby (1999) analyzed the results of reported biokinetic studies with HTO in humans and concluded that the distribution of estimates of dose per unit activity intake of tritium as HTO, based on a single compartment 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 organic compounds 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 chosen to 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 from HTO; this puts a bound of about 10% on the relative dose from tritium as OBT that is formed from HTO (Osborne 1972). A recent study by Trivedi et al. (1997), exploiting data from people who had been exposed to HTO, definitively verified this long-standing general conclusion that the contribution to dose from tritium as OBT was less than 10% of that from HTO when the intake was tritium in HTO.

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

Source of Tritium % increase in dose from OBT Reference
tritium contaminated environment
(air, water, food)
40 to 50 Evans 1969
in all normal diet 100 Commerford 1984
in fresh plants 16 to 35 Myers and Johnson 1991
in dried plants 40 to 100 Pietrzak-Flis et al. 1978
in dried milk 50 to 120 Kirchmann et al. 1977
as tritiated proteins in foods 400 Commerford 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 that the upper-limit to the increase in dose for humans exposed to tritium as a result of the organic binding was about 40% to 50% over that assumed by the then current methods that considered only the dose from water.

This is similar to the conclusion of Thompson and Ballou (1956) that organically-bound tritium in rats exposed to HTO for half of a year was equal to about 20% -30% of the activity in body water. Laskey et al. (1973) determined that if only tritiated water were consumed, the specific activity of OBT in rats would be about 25% of the water-specific activity. Koranda and Martin (1973) found that the total tritium in kangaroo rats exposed chronically 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 the bodies of animals exposed long-term to environments contaminated with tritium would reach 20% to 50% of the activity present as HTO.

These types of study have been reviewed extensively by Myers and Johnson (1991), Hill and 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 in which tritium will be present in foodstuffs as both HTO and OBT may result in doses from tritium in the form of OBT that are a substantial fraction of the dose from the tritiated water itself. The relative magnitudes of the doses will depend on the specific activities of the tritium in the food items, daily intakes, oxidation rates, gut absorption rates, and retention and transfer rates in the tissue components. For specific tritiated organics (such as proteins in Table 2.3), the dose associated with OBT can be several times 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 taken in as HTO (Balonov et al. 1984).

There have been only a few studies directly with humans who have been exposed to tritium 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 that increases in dose could be 12% -106% over that from tritium as HTO, though a more likely range (based on an indication that one study may have had problems with contamination) was 12% - 32%.

A variety of dosimetric models have been developed to reflect the observed behavior of tritium taken in as HTO and as OBT. The simplest (Crawford-Brown 1984) is a variant of one for HTO intake that had an HTO compartment linked to a second OBT compartment. Crawford-Brown added a direct input of OBT into the OBT compartment from which there is bi-exponential clearance that varies with age (see Figure 2.1). Fifty percent of the intake of OBT is assumed to be incorporated as OBT; the other 50% is initially oxidized to HTO. A recent review by Richardson et al. (1998) of the literature on 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 were generally 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 three compartments in total, including two for OBT; Killough (1981), five total with two OBT and two bone compartments; Etnier et al., (1984), four total with three tissue solids compartments that handle carbohydrates and fats differently; Belloni et al. (1985), three total with direct input into two OBT compartments; and Saito (1992), three total with direct input into only one of two OBT compartments. There are various values for the transfer parameters between compartments in the various models. None of the models is physiologically based so they are all limited in their validity beyond the data to which they have been fitted. The simplest (Crawford-Brown, 1984), although having only one OBT compartment, serves to indicate the key dosimetric consequences when intakes of both HTO and OBT occur.

The values chosen by Crawford-Brown for the parameters, K1 - K4 were based on biological half-times of 1,000, 33, 100, and 10 days, respectively, for adults. The radioactive decay constant, l, has little effect since the half-life of tritium is so much greater than the biological half-times. The values of fS (0.90) and fL (0.10) indicate the amount of tritium released from organically bound sites with relatively short and long half-times, respectively. Age dependency is assumed for individuals younger than 21 years, after which all adults are assumed 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 total dose; 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 forms of tritium when the intake is either as HTO or as OBT or is a mixture of the two. Figure 2.3 shows the variation of the integrated activity for various intake mixtures against age of individual.

For an intake of 1 Bq tritium as HTO, the integrated activity is 14.4 Bqd, corresponding to a 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 value for an intake as HTO. For an infant, the dose is 2.5 times greater. The actual dose that is delivered by HTO after an intake of OBT is the same as that delivered after an intake of the same activity of HTO (since, in this model, all the tritium goes through this HTO compartment). The dose delivered by tritium in the form of OBT after the same intake of OBT is 1.7 times that from tritium in the form of HTO. This distinction is important for any consideration of the possible difference in effectiveness of the tritium in these two forms (see Section 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 the dosimetric consequences of ingestion of food from tritium-contaminated environments. In Section 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 an adult 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, the water equivalent factor as 0.75 (Brown, 1995). If unit rate of ingestion (I = 1 kg/d) and unit concentration in the milk (C = 1 Bq/kg) are considered, then, with R =1.2, and values for EHTO 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% above that 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 delivered by tritium as HTO ( 1 / 2.7 in above Table). Therefore, in the example, of the 4.7 pSv from the 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) would be 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, the R-values for meat and vegetables are typically less than 0.8 and in the range 1.5-2 respectively. With these values of R, the OBT-associated increase in dose from tritium would be 90% and 30% respectively.

In the model, the ratio of the values of the integrated activities does not change much with age. It follows that these relative increases in dose when OBT is taken into account will not vary 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 deposition has been averaged over the soft tissues of the body. Since risks from radiation are couched in the same terms with doses being averaged over tissues the macroscopic approach is generally justified. However, for tritium, because of the short range of the tritium beta in tissue (maximum 6 mm, mean 0.69 mm) the distribution within tissues, and possibly within cells, may need to be considered.

For example, a study involving the collection and analysis of 52 deer from the Savannah River Plant site showed that the distribution of OBT in the bodies was not uniform (Evans 1969). Concentrations of OBT varied by factors of almost two between muscle and fat, and there also appeared to be inter-animal variations of the same order of magnitude. Another study, involving mice maintained on tritiated drinking water for up to 600 days, found that the average tritium content (and thus dose rate) of epididymis and ovary were only about half of 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 to account for variation in labeling between tissues does not seem warranted. However, the distribution of OBT within cells themselves deserves further comment. Experimentally, it is difficult to distinguish whether a given level of biological effect from OBT, relative to that from some other radiation, is a result of a concentration of the tritium in a cell or cell nucleus, rather than being uniformly distributed, or is a result of an inherent greater effectiveness of the tritium decaying from a particular organic molecule. If the dose is underestimated, the effectiveness of the tritium radiation appears to be higher - and vice versa. Because of this complication, this effect, which is peculiar to tritium and other biologically active radionuclides with short-range radiations, is included in the more general discussion of risk in Section 5.3.

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

The current internationally recommended values for the doses from tritium taken in as HTO or OBT are based on modeling assumptions that are similar, though not identical, to those discussed above. The recommended model (ICRP 1994, ICRP 1996b) for HTO has two components of retention in the whole body with fractions and halftimes for adult clearance of 0.97 (10 d) and 0.03 (40 d). The recommended model for OBT has two components of retention 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 months old), 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.

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 coefficient given here for OBT lacks a firm physiological base. Nevertheless, we have concluded that these nominal values that are in international use are reasonable values to use in this assessment. The modeling review above provides insight into the variation in the form of tritium 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. #