Sellafield's Evaporator D Project - A Lesson for La Hague?

Martin Forwood

If nuclear power has been labelled an extraordinarily expensive way of boiling a kettle, then Evaporator D--known to the workforce as the 'Big Kettle'--must be breaking all records. Initially costed at £90 million in 2007 (then US$185 million) and due to come into operation in 2010-2011, the cost has increased eight-fold to £740M (now US$920M) by September 2015. With a still 'challenging' operational date penciled in as 2017-2018, and with updated figures yet to be disclosed, the sky is clearly the limit for Evaporator D.

Sellafield's Evaporator D story might bear some serious lessons for the upcoming evaporator replacement planned for the French La Hague reprocessing plant (see Evaporator corrosion at La Hague threatens future reprocessing - Time for a strategy rethink?). La Hague operator AREVA NC's replacement strategy looks strikingly similar to the largely failed approach implemented at Sellafield (see also Nuclear Intelligence Weekly, 2 December 2016).

The tortured hit and miss progress of the new Sellafield evaporator, designed to reduce by evaporation the volume of the self-heating High Level Waste (HLW) liquors (raffinate) produced by spent fuel reprocessing, reveals a catalogue of project mismanagement and eye-watering cost hikes that show little sign of abating. Originally promoted by British Nuclear Fuels plc (BNFL) and subsequently by the Nuclear Decommissioning Authority (NDA) as being urgently needed to support continued reprocessing operations in the B205 (metal fuel) and THORP (oxide fuel) plant, Evaporator D is currently being squeezed into the HLW complex, where it will join its three fellow but semi-crippled evaporators (A, B & C) whose track record, outlined below, provides the clearest indication of what lies behind the urgency for Evaporator D.

Warning signs of the frailty of the existing evaporators resulting from the corrosive combination of hot liquids and nitric acid in the raffinate processed through them, had long been noted by BNFL. Designed to serve B205's Magnox reprocessing, evaporators A and B were commissioned in 1970 and 1984 respectively, each with four internal cooling coils and an external cooling jacket. By 2000, some coils were already being taken out of service because of corrosion-induced pin holes and the thinning of the coil walls beyond a minimum specified thickness. Evaporator C, commissioned in 1990 with six cooling coils and designated to serve both B205 and THORP, has also fallen prey to the same corrosive damage. As of today--and making a mockery of Sellafield's historic boast about the layers of redundancy provided by the number of coils--Evaporator A's four cooling coils have all been 'retired' leaving it wholly dependent on its external cooling jacket. Evaporator B has just one coil in service and half of Evaporator C's has also been retired. That all three ageing evaporators, with their depleted complement of coils, must remain in service or on standby until Evaporator D is operational further questions the industry's claims of 'defense in depth'.

Leaving THORP and its UK and overseas contracts and expectant customers at the mercy of just one evaporator (C) is viewed retrospectively as being a major planning blunder by Sellafield, for with Evaporators A and B frequently forced off-line for inspection and repair, the use of Evaporator C had to be shared between both reprocessing plants with the inevitable impact on reprocessing throughput and schedules. With priority given to B205's raffinate by the safety authorities for hazard reduction purposes, THORP has been left high and dry on several occasions over the years, facing either a moratorium imposed by the regulator on further reprocessing or a limitation on its throughput until either A or B return to service.

Whilst no cost figures for Evaporators A and B are known, BNFL's £70M cost of Evaporator C would make the £90M cost of Evaporator D projected in 2007 not appear unreasonable. But despite claiming not to recognize that £90m estimate, the NDA was nevertheless happy to confirm a price tag of £100M in 2008, since when the cost of Evaporator D has escalated annually - with the biggest hike to over £600M ($740M at 2012 rate) reported in a National Audit Office (NAO) report in late 2012.

NAO was highly critical of the NDA's project management and that of its contractor Nuclear Management Partners (NMP) who had acted for the NDA as Sellafield's Parent Body Organisation (PBO) from 2008 to 2016 when it was stripped of its contract. Also in 2012, a report by the UK's Public Accounts Committee was similarly critical of NMP deficiencies over Evaporator D, with the Committee's Chair describing NMP's inaction as having lost taxpayers' money 'while your shareholders have, no doubt, made a good buck out of this'.

At its formation in 2008 the NMP consortium was comprised of France's AREVA, UK's AMEC and URS from the United States and there can be little doubt that with its historic experience at the La Hague reprocessing complex, AREVA was way in front of its partners in terms of evaporator experience. Though the seeds of Evaporator D were sown several years before NMP's arrival at Sellafield, the consortium's involvement with the old Evaporators A, B and C is well documented. For example, one NMP report (Performance and Progress at Sellafield) acknowledges its involvement with a reference to the 'detailed inspections and assessments of the residual evaporator lifetimes, through the expertise brought by NMP and in particular parent company AREVA'. A 2013 report by auditors KPMG refers specifically to NMP's handling of Evaporator D and noted that the project was suffering from weak contract and design management and that attempts to accelerate progress had led to out-of-sequence construction work that had to be re-worked. Matters relating to supply chain and seismic design decisions were highlighted by KPMG as being issues for NMP to manage.

Whilst this behind-the-scenes bungling by AREVA and its partners--overseen by an apparently equally incompetent NDA--may largely explain the cost hikes and delays to Evaporator D, its novel fabrication method has also played its part. For recognizing the difficulties of constructing the evaporator, and the new building to contain it, in the confines of an HLW complex adjacent to operational facilities and on a restricted footprint, Sellafield opted for the Evaporator to be built off-site in modular form by Interserve at Ellesmere Port, Cheshire. The modules would be delivered directly to Sellafield beach by barge. At Sellafield, the building 'shell' to house Evaporator D is comprised of 11 primary cells into which the prefabricated modules would be shoe-horned, the largest of which would measure 12.5 x 7.5 x 27 meters tall and weigh 500 metric tons.

Novel and 'innovative' as the option may have been, it soon fell prey to two fundamental issues: the overly complex seismic design leading to increased material costs and construction complications, and supply chain quality issues ranging from basic materials not meeting specification to manufacturing flaws. Design changes, including the conversion of Evaporator D from a liquid to a solids evaporator imposed further delays. Thus, deliveries of the modules to Sellafield beach were completed in late 2013, the 27-meter Evaporator D itself being the last to be offloaded from the barge, with the project cost then standing at £650M.

The subsequent escalation to £740M, at September 2015, has largely been attributed by the NDA to the delivery to Sellafield of 'incomplete' modules, which has resulted in 'a knock-on effect of scope transferred to site which has created significant resource demands on the site contractor and increased issue with congestion in the facility which impacted site productivity and efficiency levels in comparison to those estimated' (Programmes and Major Projects Report: Sellafield, March 2015, Appendix 13, page 26). The exact nature and extent of the 'incomplete' modules has not been divulged but is described as leading to an increase in overall weld numbers due to access issues requiring pipe spools to be cut, and the difficulties of undertaking the work to connect together the component parts of the evaporator in the confined workspace of the 11-cell structure added to one corner of the HLW complex.

With updated cost figures for 2016 yet to be published, plus the additional costs that are certain to arise in 2017 and 2018 as the struggle continues to physically 'tie in' Evaporator D and move it from inactive to active commissioning, the final cost of the project is likely to approach the £1 billion mark.

Faced with this exorbitant cost for what has been described by the local media simply as 'a piece of kit' to support reprocessing, taxpayers will take small comfort from the Sellafield announcement in 2012 that, following design changes to Evaporator D, plans for a £600M Evaporator E had been cancelled.

Putting aside the cost aspect of the troubled project, the greatest irony of this latest of a long list of Sellafield sagas is that despite the early hullabaloo about its pivotal support role for reprocessing at Sellafield, Evaporator D can be of service to THORP reprocessing (due to finish in 2018) for no more than one year at best. At worst, it will be of service only for THORP's post-2018 clean-out, the remnant days of B205 reprocessing, which is due to end around 2020, and the planned site decommissioning work. Faced with this prospect and the embarrassing reality that, in the event of further delays, its much-vaunted Evaporator D could indeed miss the THORP reprocessing boat for which it was primarily designed, the NDA and Sellafield Ltd damage-limitation teams have recently swung into top gear by stressing the evaporator's future role in decommissioning through its ability to deal with the larger waste particles expected to be encountered during the coming years of clean-up work.

When Evaporator D is eventually fully operational, it will take over the processing of the raffinate waste streams from both B205 and THORP. At which point Evaporator C will be placed on standby for both reprocessing plant and A and B will be taken out of service and moved to post-operational cleanout.