Performance of concrete structures for nuclear reprocessing plant - review and research requirements

British Nuclear Fuels (BNFL) operates fuel cycle facilities in the UK, principally the Sellafield reprocessing complex. This includes concrete structures dating from the 1940's. Durability is most critical for structures storing reprocessed material and radioactive waste. To allow retrieval and decommissioning, the life of some facilities may extend to 2150. The concrete appears to have performed well in general, but has suffered some degradation from specific causes. To ensure continued performance of critical plants, life limiting features are identified and matched against the available methods of testing and examination. The difficulty of obtaining condition data limits prediction of structure life. Access can be difficult because of radioactivity, and there is a reluctance to take cores from sensitive liquid retaining walls. Based on this review, requirements are defined for research into improved examination techniques. The primary need is for remote, non-destructive methods. Better data is needed on the properties of concrete used in Sellafield structures, and the current condition of the concrete and associated materials. Design of most buildings was by the same design office, so that the specification to which concrete was produced was broadly consistent across site, at any particular time. Trends should therefore be visible when sufficient data is captured, and the need for intrusive investigations may be reduced as a result. The project to repair the Sealine pipe-bridge has recently gathered a lot of information on chloride content, carbonation, half cell potential readings, core strengths, etc, for the concrete concerned, but this represents only a small fraction of the concrete on site. The condition surveys produced over the years include some reports of tests, which may be used to predict trends over time. However, the conditions under which the tests were done are not known, and the results may be of limited use. Due to the radiological hazards involved, sampling from operational plants is difficult and expensive. It is therefore important that every opportunity is taken to use material which becomes available for testing, and to ensure that all results are included in a common database. For example, when concrete cores are removed to provide access for pipes or cables, they will be specified so that useful data can be obtained from them. When concrete structures are demolished, wholly or partly, the opportunity to test and examine them will be taken. Samples of irradiated water-bar will be especially valuable. In most cases, the lack of data results from radiological constraints on access, coupled with concerns that structures may be damaged by taking cores. The proposal is now to establish a site wide database to record the characteristics of concrete, reinforcing steel and other construction materials, based on original design records and test results. This database would be populated via a significant increase in characterisation work across all available structures, then be used in conjunction with an appropriate management framework for monitoring, life prediction and repair work. In addition to the general need for remote NDE to determine the extent of carbonation, and other relevant properties of concrete, there are a few technical areas (such as condition and details of prestressing ducts deep in concrete) where no suitable characterisation method has been identified. There is also a need to determine leakage rates through cracked concrete in order to justify the safety performance of structures where cracking is predicted. Water-bars represent a potential threat to service life. Considerable theoretical work and sample testing has been done but there is very limited data on the real performance of water-bars subjected to radiation, and the potential for significant variation in the material properties of the water-bars. Failure of some water-bars could be significant, and there is therefore a need to predict their performance. (author)