The evolution of the various decay-data libraries for nuclear applications has resulted in programmes of intensive testing to re-assure users that the recommended data sets are reliable and comprehensive. During the course of these benchmark exercises, errors and inadequacies have inevitably been discovered. Many can be dealt with rapidly, but some problems have proved more difficult to overcome without a significant amount of additional work. Although the assessments given below are not comprehensive, a number of important discrepancies are noted, along with known efforts to improve the contents of the decay-data libraries. Sometimes the analyses required to highlight and identify the culprit radionuclide(s) can be as taxing as subsequent attempts to correct the problematic data. Furthermore, attempts to improve the quality of the recommended decay data in this manner may be seriously undermined by the increasing lack of expertise available to undertake such work.

After much consultation and debate in the mid-1990s, a number of important radionuclides were judged to be inadequately characterised for various applications (ranging from estimates of radiotoxicity and input data for reprocessing flowsheets, to nuclides that undergo delayed-neutron emission). A set of 27 fission products (plus short-lived daughters and related metastable/ground states) was identified as important for various thermal reactor studies, including decay-heat calculations (Table 15). The majority of these radionuclides are short-lived fission products with half-lives significantly less than 3 hours. Considerable improvements were made in their recommended decay data as a consequence of including a number of relatively recent measurements in the evaluations, although the complexity of many of the decay schemes may still pose mean energy problems (Nichols, 1998; Nichols et al, 1999a).

Various decay parameters and continuum spectra have also been theoretically derived for 35 neuton-rich nuclides deemed to be important at short cooling times in decay-heat calculations (Table 16). While each of these fission products contributes significantly to the decay heat of irradiated fuel (>0.01 of the fractional cumulative yield), they have proved extremely difficult (if not impossible) to prepare, isolate and study experimentally, and are not included in the JEF-2.2 decay-data library. The US ENDF/B-VI decay-data library contains files of theoretical data for 33 of these nuclides, and they were considered as possible candidates for incorporation into the JEFF library, with supportive data from Takahashi et al (1973) and Audi et al (1997).

The precise content of any data library for decay-heat calculations is based on subjective judgements, although the desire for completeness can overcome many constraints. While the nuclides contained within Appendix B provide a reasonably comprehensive set of fission products and actinides for this type of study, changes can always be made to improve and expand the data base (Storrer, 1994). The addition of further isomeric states is a particular good example of this form of improvement; recent experimental studies have furnished evaluators with evidence for further short-lived radionuclides of this type that could be incorporated into the decay-data (and fission yield) libraries:

117mPd. 107mAg 119mAg ^^Ag 122mAg. 123mCd 125mCd* 120nIn 122nIn 126mIn 130m,130nIn 131m,131nIn. 114mSn 128mSn* 129mSb* 132mp 132mXe* 144mCs* 146mLa* 148mP^ 153mSm* 155mGd

While the postulated decay schemes for many of these radionuclides are simple (e. g., 107mAg, 113mPd and 128mSn), others are relatively complex (e. g., 112mRh, 122nIn, 125mCd and 146mLa). Their decay data can be found in ENSDF, and could be assessed and transferred to nuclear applications libraries.

Additional short-lived daughter and related metastable/ground state radionuclides are in parentheses, and were also evaluated.

*Beta-decay mode only.

An interesting piece of detective work has been carried out by Yoshida et al (1997 and 1999) to offer some explanation for the discrepancy between experiments and calculations of the gamma component of decay heat over cooling times between 300 and 3000 sec (Fig. 31, instantaneous fission burst for 239Pu). Experimental data measured by Dickens et al (1980 and 1981), Akiyama and An (1983), and Nguyen et al (1997) were combined and compared with decay-heat calculations using the JNDC-FP-V2 library (adoption of either fast-neutron or thermal-neutron fission yields had little impact on the resulting curves). Yoshida et al introduced a gamma — ray emission that is effectively missing from the decay-data files, and was postulated to belong to an ill-defined fission product (half-life of ~1000 sec). An additional P — decay chain was artificially added to the JNDC data base (as two radionuclides) on the basis of this proposed omission.

An energy release of 1.5 MeV per decay was assumed to occur (e. g., 104Mo (half-life of 60 sec) ^ 104Tc (half-life of 1092 sec) represents a suitable candidate for such an emission). The results of these artificial calculations are shown in Fig. 32 for, , U and Pu in comparison with the measurements of Akiyama and An (1983): the discrepancies between 300 and 3000 sec disappeared for all four sets of data. Fig. 33 shows the impact on the beta-energy component for the fission burst of 239Pu; there is some decrease against the original calculation, but this change falls well within the error bars of the measurements. Re-assessments have been made of beta-strength functions to identify possible candidates, including 102Tc, 104Tc and 105Tc. Overall, the proposal of Yoshida et al would appear to be a reasonable suggestion.

The incorrect assignments of ground and metastable states may also play an important role in explaining some of the discrepancy in the gamma component between 300 to 3000 sec. Fig. 34 compares experimental data with decay-heat calculations for 239Pu; improvements can be achieved by adjusting the decay data for Rh. The ground and metastable states of Rh are assigned half-lives of 17 sec and 6 min respectively in JNDC-V2 (labelled ‘JNDC-V2 original’ in Fig. 34), while these assignments are reversed in JEF-2.2 (labelled ‘Rh108-JEF2.2’ in Fig. 34). Adoption of the JEF-2.2 data increases the half-life of 108gRh by a factor of ~ 20, leading to a considerable rise in decay heat around 1000 sec. Another feature is worthy of note: a lack of gamma-ray transition energy in the 102Tc decay-data file of JEF-2.2 gives rise to a collapse in the gamma energy component (labelled ‘Tc102- JEF2.2’ in Fig. 34), implying that this radionuclide represents a suitable candidate for the missing gamma energy (as noted earlier). These studies demonstrate that adjustments to the decay-data of only one nuclide can dramatically change the decay — heat predictions, and that beta branching ratios to the ground and metastable states need to be measured with good accuracy in order to define the situation correctly.