1.1.1. Fission products

Schmittroth (1976) studied the impact of the uncertainties in fission-product yields, half-lives, decay energies and the assignment of isomeric states. Thermal-neutron fission of 235U was considered in detail, and this assessment indicated that decay heat can be calculated to an accuracy of 7% or better for cooling times > 10 sec. The major sources of uncertainty at cooling times < 1000 sec arise from ill-defined decay energies and fission-product charge distributions. This work was extended further by Schmittroth and Schenter (1977) who undertook a sensitivity analysis of the calculated decay heat associated with the thermal fission of 235U and the fast fission of U and Pu. Both burst and 10 sec exposures were considered (Figs. 2 and 3).

Fig. 2. Total decay-heat uncertainties for thermal fission of 235U (Schmittroth and Schenter, 1977)

Fig. 3. Total decay-heat uncertainties for fast fission of 239Pu (Schmittroth and Schenter, 1977)

Schmittroth and Schenter were able to attribute the main sources of uncertainty in decay-heat summation calculations to existing uncertainties in the fission-product yields and decay energies. Uncertainties in fission-product half-lives were judged to be relatively unimportant for most cooling times in decay-heat calculations. Overall, decay energies were found to be the major source of decay-heat uncertainties, especially for short cooling times (< 100 sec). These studies underline the sensitivity of decay heat to the mean energies of decay; various efforts have been made to improve these data by measurement and theoretical modelling, as outlined in Section 5.

A similar sensitivity analysis has been made of the uncertainties in decay heat when using the nuclear data contained within the JEF-2.2 library (Storrer, 1994). The calculated decay heat is dominated by radionuclides with well-defined decay schemes for cooling times > 3 x 106 sec, while the largest contributions come from poorly-defined nuclides for cooling times < 3 x 105 sec. Finally, at cooling times less than ~10 sec, fission products based completely on theoretical data contribute approximately 25% to the resulting decay heat.

Developments of the gross theory of beta decay form the main source of decay data for poorly-defined radionuclides in the JNDC-FP and US ENDF/B-VI libraries (see Sections 5.3 and 5.5). Oyamatsu et al (1997) have undertaken extensive studies of the suitability of these data libraries in decay-heat calculations. Their sensitivity analyses were extremely detailed, and highlighted a series of specific inadequacies.

For example, Fig. 4 shows the variation in uncertainty of the total P+y decay heat as a function of cooling time, following the thermal-neutron fission of 235U:

(a) at short cooling times, the uncertainty in decay heat is dominated by uncertainties in specific independent yields and decay constants, although there is also a increasingly significant contribution from uncertainties in the decay energies up to 1000 sec cooling time;

(b) particular peaks in the uncertainty profile contain significant contributions from the uncertainties of specific parameters (see Table 4) — for example, peak 4 contains significant contributions from uncertainties in the decay energies of 93Sr and 102Tc, and peak 7 is dominated by uncertainties in the independent fission yields of 97,97mY.

These analyses are extremely informative for a wide range of fissioning nuclides, and a further example is given in Fig. 5 and Table 5, for the fast fission of 238Pu. Particular nuclear parameters appear regularly in the assessments (e. g., independent yields of 97Sr and 97mY in peak 1, decay constant for 101Zr in peaks 1 and 2, decay energies of 103Mo and 103Tc in peak 3, and cumulative yields for 102,102mNb in peak 4). The main contributors to the decay-heat uncertainties are highlighted in the tables with respect to each numbered peak, providing clear indications of the specific needs for improved fission-product data.