In order to understand the achievements made re­garding the general progression towards on-load re­fuelling at maximum reactor power, it is necessary first of all to examine briefly some of the history of AGR operation. At Hinkley Point B, the earliest attempts made to handle fuel stringers with the re­actor at power were restricted to the charging of new fuel only at the beginning of the fuel cycle, when some of the vacancy stringers were exchanged with feed fuel. During subsequent premature discharge of one of these vacancy replacement stringers which was thought to contain failed fuel (i. e., pin failure(s)), it was discovered that the graphite sleeves had failed on two of the fuel elements within the stringer. A period of exhaustive investigation began and it was eventually established that the presence of undetected cracks re­sulted in weaknesses within the graphite sleeves, which were responsible for their failure during loading at high reactor power when they had been subjected to a differential pressure of some 2.5 bar. Furthermore, it had previously been discovered that the normal cooling gas flows within the charge-path could, if the irradiated stringer became stuck at a crucial position during its removal from a reactor at power, result in flow stagnation, leading to consequent overheating of the tie-bar and standpipe seals.

Taking these and many other factors into consi­deration, on-load refuelling operations were suspended until all the various safety issues could be satisfac­torily resolved. In the meantime, refuelling had to proceed in large batches (perhaps 8-12 stringers at a time, or even more) with the reactor shut down. Apart from the economic ramifications, this resulted in unwanted thermal cycling of the fuel and reactor plant, difficulties with regard to reactor optimisation and an uneven loading of the working patterns of the fuel route. There was naturally a large incentive to re-establish an on-load refuelling regime as quickly as possible and its eventual return was only made possible following major modifications to the charge machine, in which a more effective and dependable cooling system and a prompt reactor trip system based on hoist load sensing, were provided. Together, these major system modifications provided guaranteed pro­tection against the likelihood of overheating and con­sequent tie-bar failure during on-load refuelling, as well as ensuring that even if a fuel stringer were accidentally dropped, there would be no significant release of fission products to the environment. In addition, special equipment was provided in the fuel route so that all new fuel elements could be eddy current and pressure tested on site before assembly into fuel stringers, thereby enabling any sleeve de­fects, not visible by eye, to be located prior to load­ing of the fuel to the reactor.

Current practice

On-load refuelling of AGRs now takes place in a totally different climate to that which existed in the early days. New operating techniques have been de­veloped for control of reactor power during the re­fuelling operations and new legislation and procedures (i. e., Operating Rules and related instructions) have therefore followed. Current practice is to refuel the reactor at about 30% power, which produces a small differential pressure of approximately 0.5 bar across the fuel element graphite sleeves. The batch refuel­ling concept has been retained, as shown in Fig 3.52, involving usually 6 to 8 exchanges at a time, which roughly represents the refuelling needs of each reactor following a month’s operation at maximum power. In practice, operational penalties result if the refuel­ling ‘backlog’ is allowed to exceed about half-a-dozen stringers since the initial reduction in power down to the 30% condition (approximately 450 MW(Th) or 150 MW(E)) then needs to be made more slowly so that the reactor does not ‘poison out’ during the en­suing xenon transient. As each exchange is completed, power is raised to an intermediate ‘parking’ condition, currently equivalent to about 500 MW(E) and set from considerations of temperature cycling upon boiler life­time. Under normal circumstances about three ex­changes can be completed in a day.

Much work is already in hand in support of ele­vating the power at which refuelling can safely take place. It is generally considered however that the po­tential of the current design of fuel element is some-

Fig. 3.52 Typical power variation during on-load
batch refuelling

Reactor power is initially reduced from maximum
down to about 450 MW(Th) (150 MW(E)), at which
refuelling can take place. On completion of the batch,
re-establishment of maximum power can take many
hours, depending primarily upon the resulting control
rod positions.

what limited in this respect, and that significant pro­gress can only be made by using a more robust design incorporating a single thick-graphite sleeve and other improved performance features.