Large Volume Calorimetry : A Technique to Quantify Tritium

1
Large Volume Calorimetry A Technique to
Quantify Tritium

• Jean-Charles Hubinois, Alain Godot, Sonia de
  Boyer, Guénaëlle Coindet, Armelle Collardey

Commissariat à lEnergie Atomique, Centre
dEtudes de ValducDTMN/SAD/Laboratoire de Chimie
Analytique21120 Is-Sur-Tille, France
Introduction Characterization of nuclear
samples can be achieved in terms of impurities by
classical chemical analytical techniques, such as
emission, absorption or mass spectrometries, but
the determination of nuclear element inventory
implies the use of specific techniques. Determinat
ion of plutonium can be carried out by
coulometry, isotopic dilution mass spectrometry,
UV-visible, alpha or gamma spectrometry.
Techniques that are available to quantify tritium
are for instance liquid or solid scintillation,
mass spectrometry, helium-3 method or
calorimetry. All these techniques differentiate
by accuracy, limit of detection, analysis time,
dependence on matrix but also by the fact that
some are destructive when others are non
destructive. Among non destructive and non
intrusive methods, calorimetry is considered to
be the most accurate, even for large volume
samples. On account of these characteristics,
determination has been achieved for more than 35
years by calorimetry at CEA Valduc. However, need
to quantify lower amounts of tritium leaded CEA
to invest in new apparatus. CEA entered into
collaboration with Setaram Instrumentation in
order to develop and manufacture high sensitive
and accurate large volume calorimeter.
Experimental results
Calibration lines obtained with an electric
heater sample
Aspect of registered signal, until stady state
(60 L)
Predictive determination of the resultby means
of numerical calculation on a 2 W tritiated
sample
Reproducibility (calculated on 10 different
measurements)
Zooms on stady state signals (for blank, 1 W, 5 W)
Conclusions

• New large volume calorimeters are integrated in
  glove box and are shown to be capable to quantify
  tritium in 60 and 90 liters volume
• Sensitivity is measured to be in the range of 10
  nW/mL
• Uncertainties on measurements are calculated to
  be less than /- 1 (60 L) and than /- 1,5
  (90 L) on the whole range
• Accuracy is measured to be better than 0,3 on
  a 1 W certified standard
• Limit of detection is calculated to be 3 mg of
  tritium for 60 L calorimeter, and 15 mg for 90 L
• Predictive determination is possible within less
  than 1 deviation from the steady state baseline
  (for a 2 W sample), after only 5 hours