[en] An extensive programme of work has been carried out to evaluate two candidate inhibited cleaning solutions for possible implementation on plant, which would be the first chemical clean of an AGR boiler. The two candidate cleaning solutions considered were a Stannine-inhibited citric acid/formic acid mixture (GOM106) and inhibited hydrofluoric acid. Citric acid-based cleaning processes are widely used within the UK Power Industry. The GOM106 solution, comprising a mixture of 3% citric acid, 0.5% formic acid and 0.05% Stannine LTP inhibitor, buffered with ammonia to pH 3.5, was developed specifically for the AGR boilers during the 1970's. Although a considerable amount of materials testing work was carried out by British Energy's predecessor companies to produce a recommended cleaning procedure there were some remaining concerns with the use of GOM106, from these earlier studies, for example, an increased risk of pitting attack associated with the removal of thick 9Cr oxide deposits and a risk of unacceptable damage in critical locations such as the upper transition joints and other weld locations. Hence, additional testing was still required to validate the solution for use on plant. Inhibited hydrofluoric acid (HFA) was also evaluated as an alternative reagent to GOM106. HFA has been used extensively for cleaning mild and low'alloy steel boiler tubes in fossil-fired plant in the UK and elsewhere in Europe and is known to remove oxide quickly. Waste treatment is also easier than for the GOM106 process and some protection against damage to the boiler tube materials is provided by complexing of fluoride with ferric ion. Validation of the potential reagents and inhibitors was achieved by assessing the rate and effectiveness of oxide removal from specimens of helical boiler tubing and welds, together with establishing the extent of any metal loss or localised damage. The initial materials testing resulted in the inhibited ammoniated citric / formic acid reagent being chosen. This paper describes the results of the latter two activities: additional materials testing necessary to support a full scale plant clean and the trial clean on a model boiler rig replicating a single boiler tube together with the pre- and post-clean thermohydraulic behaviour. (authors)