TRITIUM RADIOECOLOGY AND DOSIMETRY TODAY AND TOMORROW

1
TRITIUM RADIOECOLOGY AND DOSIMETRY - TODAY AND
TOMORROW D. Galeriu, P. Davis, W. Raskob, A.
Melintescu IFIN-HH Romania AECL Canada
IKET Germany
8th International Conference on Tritium Science
and Technology September 16-21, 2007 Rochester,
New York
2
Environmental Tritium Processes
3
Resistance Approaches to Deposition and Exchange

• Aerodynamic resistance Ra depends on turbulence
  and wind speed
• Boundary layer resistance Rb depends on
  turbulence, wind speed and surface properties
• Total surface resistance Rc can can be split up
  into canopy and ground related resistance
• Canopy resistance depends on surface properties,
  temperature, photosynthetically active radiation,
  humidity, water content in soil
• For HT deposition, ground resistance depends on
  the rates of diffusion and oxidation in soil, and
  is much lower than the canopy resistance

Deposition velocity1/(RaRbRc) This is also an
exchange velocity at air to plant (soil) interface
4
HT is oxidized to HTO, mostly in soil by bacteria
containing the hydrogenase enzyme
After OKula and East 2000
Oxidation in the atmosphere is a very slow
process with a half-life of gt 5 years. Most
significant oxidation occurs at the
atmosphere-soil interface
5
HTO vs. HT Predicted Dose
37 TBq/y chronic DCART ?/Q1?10-6 s/m3
370 TBq accidental UFOTRI distance 1200m
The only significant dose impact occurs following
oxidation of HT to HTO and subsequent HTO
exposure. For similar source terms, an HT release
has a 1-10 dose impact compared with HTO,
depending strongly on weather and soil conditions.
6
Models for Routine Releases
Since meteorology is time varying and ROUTINE
RELEASES ARE TIME VARYING!, specific activity
concepts do not strictly apply as full
equilibrium is not established. The model must be
adapted to partial equilibrium and the data
averaged appropriately.
Weekly HTO release rates from the CANDU 6 reactor
at Cernavoda, Romania (2002)
7
Specific Activity Models for Tritium (EMRAS
approach)

• Transfer from Air to Plants

HTO concentration in plant water (Bq L-1)
RH relative humidity Cair tritium
concentration in air (Bq m-3) Ha absolute
humidity (L m-3) Csw HTO concentration in
soil water (Bq L-1), and ? ratio of the H2O
vapour pressure to that of HTO ( 0.909).
Relative humidity is a weighting factor that
reflects the partial equilibrium of the plant HTO
with air and soil water HTO.
8
The equilibrium plant water equation can be
derived starting from a complex dynamic model
involving the transpiration flux, many transfer
resistances, the exchange velocity and the
following assumptions

• the mass of plant water is constant and
  consequently the transpiration flux is linked
  with the exchange velocity and water vapour
  density deficit (realistic)
• the concentration at the soil surface is equal
  to the concentration in the root zone
  (reasonable)
• the soil surface, leaf and reference temperature
  are the same and it is possible to define a
  unique saturation humidity and relative humidity
  (affordable)

The equilibrium plant water equation gives robust
and reliable results for the average
concentration averaged over a few weeks prior to
harvest.
9
Soil HTO

• The soil water concentration is assumed to be
  proportional to the concentration in air
  moisture, with proportionality constant CRs
• CRs varies between sites and a generic value can
  be recommended only tentatively.
• If the HTO concentration in precipitation (and
  irrigation water) are known or well predicted,
  better assessments can be done, including
  individual contributions from wet and dry
  deposition

10
Transfer to Animals

• The model considers transfers from HTO and OBT in
  feed to HTO and OBT in animal products
• All 4 transfer coefficients can be defined using
  data on animal nutrition and hydrogen metabolism
  and were tested with available experimental data
• The transfer coefficients show large variability
  from animal to animal and productivity
• The variability can be reduced by using
  concentration ratios in place of transfer
  coefficients
• CRHTO from HTO in diet to total tritium in
  produce
• CROBT from OBT in diet to total tritium in
  produce

11
Models for Accidental (Dynamic) Releases

• The prediction of time-dependent concentrations
  in plants must take into
• account
• Uptake and loss of HTO in plant water (dependent
  upon the canopy resistance)
• crop growth
• HTO to OBT conversion via photosynthesis during
  the day. Night OBT production is also important.
• distribution of dry matter to plant parts
• respiration
• the water cycle
• the plant development stage at the time of
  exposure

12
Processes driving tritium dynamics in plants
13
Canopy resistance can be linked directly to
canopy photosynthesis rate
Important parameters include air temperature,
plant development stage, photosynthetically
active radiation, leaf area index and water
deficit.
The aerodynamic and boundary layer resistances
depend on weather and crop height. These
resistances have low values at midday (50-100
s/m) and higher values (300 s/m) at night.
During the day, the canopy resistance is
comparable to the sum of the aerodynamic and
boundary layer resistances. At night, the canopy
resistance controls HTO uptake, although all
stomata do not close in some plants.
14
Effect of soil grid size on HTO concentration in
soil and plants

• The upper soil (0-5 cm) concentration varies by
  more than 3 orders of magnitude as the grid size
  increases from 0.1 cm to 1.6 cm.
• For a plant with a 30-cm rooting depth, the HTO
  concentration in the plant water varies by a
  factor of 4 as the grid size increases from 0.1
  cm to 1.6 cm.
• As a compromise between computational time and
  accuracy, a grid size of 0.5 to 1 cm seems to be
  acceptable. This was also the depth where
  deposited HTO was found after a short term
  exposure of bare soil in an experiment conducted
  at FZK.

15
Dynamic modeling of tritium in plants requires
knowledge of plant growth dynamics
Partition fraction of newly produced dry matter
to roots, leaves, stems and edible grain as a
function of development stage (0emergence 1
flowering 2 full maturity) Data are for maize
cultivar F320 from South Romania
Above ground dry matter dynamics for sunflowers.
Note the difference between an average cultivar
from the central EC and a local cultivar from
Romania
16
Dynamic animal model

• A recent metabolic model for adult mammals is
  based on the following
• assumptions
• The most important body organic compartments are
  the viscera (including heart), muscle, adipose
  tissue, blood (plasma and RBC) and remainder
  (including brain). Mass and composition are
  known.
• Tritium in body water equilibrates rapidly and a
  single body water compartment suffices.
• The loss rate from the organic compartment is
  similar for intakes of HTO, OBT or OBC and can be
  assessed directly from the energy turnover rate
  (net maintenance) of organs . The organ specific
  metabolic rate (SMR) in basal and active states
  must be known.
• SMR has been obtained experimentally for a few
  mammals and a zero order approximation is
  generally used depending on mature mass.
• There is metabolic conversion of HTO to OBT and
  the equilibrium value does not vary across
  mammals for the ratio of OBT derived from HTO or
  from intake OBT.
• The energy (heat) and accompanying matter lost in
  transforming the metabolisable input in net
  requirements is considered as a single, fast
  process.

Under these hypotheses, the model gives reliable
predictions with no calibration. However, there
are difficulties in expanding the model to
growing mammals.
17
Flowchart of the Metabolic Model
HTO intake
Red blood cells
OBT intake
This figure shows the complexity of the processes
in animals. The metabolic model can be used to
build simpler models with more robust parameters.
Remainder (skin, skeleton, brain)
18
Model tests (no calibration)
Sheep fed tritiated glucose acetate
Cow fed hay contaminated with OBT for 28 days.
Results are shown for the present model and
for TRIF and UFOTRI
19
Uncertainties in dynamic modeling
Soybean scenario The models all overpredict in
the first few hours after exposure. The large
differences in predictions at longer times have
little significance because the concentrations
are so low. The mispredictions carry through to
the predicted OBT concentrations.
Hypothetical scenario, ingestion dose Large
variability exists in the predicted doses as well
as in the contribution of the various food items
to the dose. This degree of uncertainty makes it
difficult to manage the accident and to set
interdiction levels on food trade.
20
Diurnal and seasonal effects on ingestion dose
after an accidental tritium release (calculated
with RODTRIT)
Dose is given in arbitrary units (a.u). The
plants and animals were exposed to a constant HTO
air concentration for one hour.
21
The mammal model can be extended to humans,taking
into account the longer maturation period, the
larger brain (which consumes much of the
maintenance energy) and different diet.
22
Human Dose Coefficients (in 1011 Sv/Bq)
ICRP as presently recommended H RBE1 ignores
the non-uniform distribution of OBT in the body
(as in ICRP) shows higher retention. E
(RBE1) RBE1 considers the non-uniform
distribution of OBT (high concentration in
adipose tissue with low radio-sensitivity). E(RBE
gt1) as above but allowing a range of RBE. New
model is probabilistic
The dose coefficients may vary with gender. The
RBE at low doses remains to be clarified.
Potentially the RBE should be increased for the
fetus.
23
Tritium in the Aquatic Environment

• HTO in the body water of aquatic fauna and flora
  is in equilibrium with the HTO concentration in
  the surrounding water.
• For routine releases, the OBT concentration is
  related to the HTO concentration through a
  discrimination factor.
• For accidental releases, OBT is treated
  dynamically with separate equations for primary
  producers and consumers. For phytoplankton, we
  deduced and tested an original model. For
  consumers, OBT is formed in a single compartment
  from HTO and OBT from food. The OBT loss rate is
  taken from experimental data or from a metabolic
  model for fish. The model has been tested for
  plankton, mollusk and small fish. Experimental
  data are needed on OBT formation and loss in
  larger fish species.
• Predictive power for detailed dynamics is
  tributary to use of single OBT loss rate.
  Improvements must await additional experimental
  data.
• For HTO releases of similar source strength, an
  aquatic release is 10-100 times less harmful than
  an atmospheric release, depending on the source
  of drinking and irrigation water.
• Many experimental data show higher OBT
  concentrations than expected, implying the
  release of dissolved organic tritium directly to
  the water body.

24
Dissolved organic tritium (DOT) release to
rivers or seawater
High OBT concentrations in mollusks and fish have
been observed in some rivers and in the Severn
estuary (UK), while HTO concentrations in the
water were low. The most probable explanation is
linked with the ability of plankton, mollusks and
crustaceans to selectively uptake dissolved
nutrients from water. These nutrients, if marked
with tritium from industrial sources, will
increase the OBT concentration in the organic
matter of aquatic fauna above the levels expected
on the basis of the HTO concentration in the
water. The uptake efficiency of DOT varies
depending on the chemical form and on the species
of plankton or mollusk.
The release rate of DOT to the Severn estuary
decreased 10-fold. OBT concentrations in aquatic
organisms also dropped by a factor of 10 with an
environmental halftime of 1 year.
25
TOWARD CONCLUSIONS

• The 1990 Aiken list was amended in 1997 by Raskob
  and Barry. Sensitivities and hence importance
  in this list vary with both inputs and end
  points. Site- and task-specific analyses must be
  done to identify the most important processes in
  a given application.
• Areas Requiring Further Work
• plant uptake of HTO at night
• rates of OBT formation in plants, particularly
  at night
• dispersion in soil
• reemission from soil and plants
• rates of OBT formation and loss
• in animals
• translocation to fruits and roots
• tritium behavior in winter
• HTO concentrations in the environment following
  an HT release.

26
With present knowledge, it can be argued that the
expected dose to members of the public from
routine tritium releases is unlikely to be higher
than 30 µSv/a for todays nuclear facilities.
Accidental releases of HT or aquatic HTO releases
have much lower radiological impact than an
accidental atmospheric release of HTO. EU
guidance on response to accidental releases is as
follows
The next generation of models for accidental HTO
releases must be improved to decrease
uncertainties and to cope with tighter regulatory
requirements.
27
Requirements for the Next Generation of Dynamic
HTO Models

• Reliable atmospheric transport and dispersion
  codes (particle models) with good representation
  of reemission and inclusion of turbulence,
  topography etc.
• Changing environmental conditions must be taken
  into account
• Several sub-models are needed to describe the
  behaviour of tritium in soil and crops
• The crop sub-model is most important and here the
  plant physiological parameters must be considered
• Conversion processes from HT to HTO and further
  to OBT have to be modelled
• Sub-models have to be based on physical
  approaches knowledge from other disciplines
  should be used to derive general dependencies
  based on data for other substances than tritium
• Site-specific information on land use, soil types
  and crop genotypes should be applied, together
  with realistic habits for the maximally exposed
  individual.