How The ‘Generic Design Assessment’ Has Failed
The Generic Design Assessment (GDA) is a process which was set up in order to examine the designs for new power stations and iron out any flaws in them before the power stations are constructed. Allowing the regulators to get involved with designers at the earliest stage was supposed to ensure an open and transparent process resulting in several competing designs with all significant design issues fully resolved.
The GDA was carried out by the Health and Safety Executive between 2007 and 2012. It was was undertaken in order to prevent the construction problems that have plagued the two reactors currently being built in Europe from being repeated in the UK. The two European Pressurized Reactors (EPR’s) are being built by Areva, a company which, like EDF, is largely owned by the French State. One of the reactors is at Olkiluoto in Finland, the other is at Flamanville in France. The original date for completion of the Olkiluoto reactor was May 2009, but in February 2013 it was announced that the reactor would not be finished until 2016. the cost has escalated from an original estimate of €3 billion to a whopping €8.5 billion. This has led to legal tussles over who is to blame for the delays as Areva has a fixed price contract to build Olkiluoto for €3 billion. According to some estimates, the Olkiluoto reactor will be the 5th or 6th most expensive structure in the world, even more expensive than the Large Hadron Collider in Switzerland. Similar cost overruns and delays have dogged the EDF owned Flamanville reactor which is now projected to cost about 8 billion euros, almost three times the original price.
In November 2010 Stephen Thomas, a Professor of Energy Studies from the University of Greenwich authored a report1 which stated that The EPR design is “in crisis” to such a severe extent that it is likely to be an economic failure. He said:
“The two sites in Europe where EPR is under construction, Olkiluoto and Flamanville, have gone dramatically wrong from the start of construction. It might have been argued that the problems at Olkiluoto were due to the lack of experience of the utility and the inexperience of Areva NP in carrying out the architect engineering. However, the fact that EDF, the most experienced nuclear utility in the world seems to be doing no better at Flamanville suggests the main problems are more related to the build-ability of the design itself than to specific issues at Olkiluoto.”
He was also concerned about the cost-
“As early as 1995 and again in 1997, there were concerns about the cost of the EPR .., it seems unlikely that EPR will be affordable except where huge public subsidies are offered and/or there is a strong likelihood of full cost recovery from consumers, no matter what the cost is.”
The fundamental cause of the escalating costs and delays is due to problems with the design which have come to light during the course of the construction.
In October 2010 The Royal Academy of Engineering published a report2 called ‘Engineering the Future Nuclear Lessons Learned’ in which they analysed factors to deliver reactors on time and to budget. The primary lesson from Olkiluoto and Flamanville is that the design must be finalized before construction
“New stations should be based on the application of proven technology and established design. This must be complemented by a high level of design completion in advance of construction”
“…design development does involve risk of delay and price escalation and should be resisted. There must be a rigorous, efficient and auditable design change process in place, indeed this is a requirement of the nuclear site licence, and a culture established that recognises that even seemly small changes can have unexpected implications and therefore require formal review.”
The government insists that we urgently need new nuclear power in order to replace the coal, gas and ageing nuclear plants that are scheduled to be closed within the next few years. In order to make the claim that nuclear power will plug the energy gap credible, the time frame that was imposed on the GDA process was strict. EDF proposed to begin construction of the first of the power stations in January 2013. In order for this to be feasible, it was planned that an Interim Design Acceptance Confirmation (I-DAC) would be granted in 2011, and a Final Design Acceptance Compliance (F-DAC) would be granted by December 2012
From the outset of the GDA process, reports indicate that the HSE was understaffed and that the information requested from EDF on the details of the design was not provided on time and was often of poor quality, which made it very difficult for the HSE to reach its targets. In a review of the GDA process in May 20103 , the review board discussed the issue of what to do in the event of being unable to meet the deadline:
“One further key issue… concerns ‘the end of GDA’, and particularly the outcomes of that process and the transition from Phase 1 (GDA) into the site-specific licensing work of Phase 2. In this respect, we were surprised and concerned at the level of uncertainty that now appears to have grown within ND (Nuclear Directorate) about these issues and at the potential damage that such uncertainty is likely to create to both the momentum of the work on the new build programme and to the credibility of GDA in particular.”
The Board went on to say:
“The final Step 4 of GDA has been scheduled to finish in June 2011 but, at the time of our review, we noted debate within ND as to whether or not, if a DAC (Design Acceptance Confirmation) were to be issued with ‘issues’, the GDA timetable should be extended until such time as these had been satisfactorily resolved. Doing so would, of course, turn what had originally been envisaged as a defined and time-limited process into a more open-ended one, with all the attendant risks of such open-endedness. The alternative approach that we learned had also been under consideration in ND would be to stick to the original finish time for GDA (in June 2011) and instead deal with any unresolved and outstanding generic issues as part of Phase 2 – the licensing process.”
Ironing out all the generic design problems before issuing the site licences (which allow construction to begin) was the very reason for setting up the GDA. Here the review body admits that if the generic issues are not solved, the licences could be granted regardless.
In order to stick to the schedule and issue the final design confirmation in December 2012, the HSE regulators introduced a process of ‘Assessment Findings’. Assessment Findings were unresolved concerns with the design which, though they were deemed to be important to safety, were not considered critical to the decision to start construction. Between December 2011 and November 2012, 240 Assessment Findings were raised. These were in addition to the 484 Assessment findings raised during the three year period up to the issue of the Interim-Design Acceptance Compliance (I-DAC) in December 2011
It has been strongly argued that a great number of the unresolved design issues that have been shunted off into Assessment Findings(AF’s) are in fact major elements of the design. A report4 examining the GDA process compiled by John Large a respected nuclear engineer concluded that:
“In effect, this AF deferral approach has lacked transparency at the time of the F-DAC grant and, moreover, in terms of nuclear safety the final performance of the plant (functionality, risk, effectiveness of protection, etc) will not be finally settled until well into the construction and, quite possibly, commissioning phases of the first EPRs scheduled for Hinkley Point. The existence of such uncertainties together with the quite obvious incompleteness of the plant design and development, particularly in the generic safety critical areas of Fault Studies and Control & Instrumentation must have, surely, rendered the GDA process itself incomplete and inconclusive”
1 The EPR in Crisis, Steve Thomas, 2010, (www.nirs.org/reactorwatch/newreactors/eprcrisis31110.pdf)
2 Engineering the Future. Nuclear Lessons Learned, The Royal Academy of Engineering, 2010, (www.imeche.org/docs/default-source/public-affairs/Nuclear_Lessons_Learned.pdf)
3 Third Report of the GDA Process Review Board, HSE 2010, (www.hse.gov.uk/newreactors/reports/third-report-gda-prb.pdf)
4 Final Report On The ONR Generic Design Assessment, Large and Associates 2013, (www.largeassociates.com/LA%20reports%20&%20papers/3206%20GDA/R3206-I3-06-06-13.pdf)