Aagesta-BR3 Decommissioning Cost. Comparison and Benchmarking Analysis

This report presents the results of decommissioning cost analyses focusing on discrete working packages within the decommissioning program of the BR3 reactor in Mol, Belgium and comparison of them with cost estimate data for the Aagesta research reactor in Sweden. The specific BR3 work packages analysed were: Primary coolant piping decontamination; Primary coolant piping dismantling; Vulcain reactor internals dismantling; Westinghouse reactor internals dismantling; Reactor vessel dismantling. The main conclusions to be drawn from the analyses are that: The fixed costs related to decontamination and dismantling activities generally are a very important part of the overall resources needed to execute the work, with the Reactor Pressure Vessel (RPV) seemingly being significantly more demanding than other major components. Cutting activities tend to need something like 150 to 200 labour hours per m2 of reactor equipment dismantled. Fixed investment costs to set up the equipment needed to cut up major vessels or internals appear to be in the range of MSEK 4 to 8. Consumables costs vary according to the nature of the equipment being dismantled. The thicker the metal being cut, the higher the attrition rate for things such as cutting blades. The range of consumables costs at BR3 have been in the range of MSEK 0.1 to 0.2/m2 dismantled. The extent of detailed information available in the 1996 Aagesta estimate is not sufficient to enable a full comparison with the BR3 decommissioning results. A global first comparison has been attempted by summing the resources expended on the BR3 work packages described in this report with the combined dismantling data presented in the 1996 Aagesta cost estimate report. Very broadly the cost of decontamination plus dismantling of the main process equipment at Aagesta appears to be in the order of MSEK 70, of which MSEK 4 is labour on preparatory/planning work, MSEK 40 is labour on actual decontamination and dismantling and MSEK 25 is equipment. The BR3 work packages described in this report add up to something like 83,000 labour hours plus about MSEK 13 of investments and consumables costs. At Swedish average team labour rates 83,000 hours would equate to about MSEK 52. Adding the investment cost of MSEK 13 gives a total of about MSEK 65. This of course is quite close to the Aagesta figure but it would be wrong to draw immediate, firm conclusions based on these data. Such a comparison should take into account, inter alia: The number and relative sizes of the equipment decontaminated and dismantled at Aagesta and BR3. The assumed productivity in the Aagesta estimate compared to the actual BR3 figures. The physical scale of the Aagesta reactor is somewhat larger than the BR3 reactor, so all other things being equal, one might expect the Aagesta decommissioning cost estimate to be higher than for BR3. Aagesta has better access overall, which should help to constrain costs. The productivity ratio for workers at BR3 on average was high - generally 80 per cent or more, so this is unlikely to be exceeded at Aagesta and might not be equalled, which would tend to push the Aagesta cost up relative to the BR3 situation. There is an additional question of the possible extra work performed at BR3 due to the R and D nature of the project. The BR3 data analysed has tried to strip away any such 'extra' work but nevertheless there may be some residual effect on the final numbers. Analysis and comparison of individual work packages has raised several conclusions, as follows: The constructed cost for Aagesta using BR3 benchmark data is encouragingly close to the Aagesta estimate value but it is not clear that the way of deriving the Aagesta estimate for decontamination was entirely rigorous. The reliability of the Aagesta estimate on these grounds therefore might reasonably be questioned. A significant discrepancy between the BR3 and Aagesta cases appears to exist in respect of the volumes of waste arising from the decontamination activity. A factor of 15 different in ion exchange resin volume is puzzling and no explanation has been found, other than that the volumes were 'estimated' rather than calculated in accordance with a clearly defined method statement. The work analysed for comparisons included preparation work before the actual dismantling of the primary pipes and the auxiliary circuits plus the actual cutting of the primary pipes into small pieces of 0.8 m long to fit in the chemical reactor of the BR3 decontamination process. The estimated grand total resources required was 4,734 hours, for a unit requirement of 740 hours/MT. This is very close to the Westinghouse Test reactor benchmark figure for this activity. The total resource requirements for primary pipework dismantling were dominated by the preparatory activities rather than the cutting activity itself. A comparison with the Aagesta cost estimate would be possible only if a more detailed breakdown of projected manhour information could be provided for Aagesta. The basis for the preparatory work and actual removal of the Aagesta RPV is unclear and may not have been particularly rigorous. Comparison with BR3 benchmarking data suggests that the Aagesta estimate for the RPV could be significantly low. However, actual experience of steam generator (SG) removal at Aagesta provides evidence of very efficient execution of similar work, which might contradict the BR3 experience. The available data does not support reaching a detailed conclusion. A clear possibility is that the nature of the two jobs (SG and RPV) is radically different, either due to size, radiological conditions, physical access etc., or a combination of all factors, with RPV work being more demanding. If this is correct, the Aagesta SG experience may not be particularly relevant, whilst the BR3 experience would indicate the need for further scrutiny of the Aagesta RPV estimate