UK decommissioning firms share lessons learned

27 June 2017

The two companies responsible for the clean-up of the UK's nuclear legacy are willing to share lessons learned with the global decommissioning industry, Martyn Jenkins, managing director of Enkom Consulting, told a recent conference. Both Sellafield Ltd and Magnox Ltd have "moved from operational to project delivery" and have identified five main challenges, Jenkins said.

Jenkins chaired a panel discussion, A UK perspective on how to address the rising costs of liabilities, at the Nuclear Decommissioning & Waste Management Conference Europe held in Manchester, England on 25 May.

"Increasing costs tends to cause a problem with confidence and that's an issue we all face," said Jenkins, who was joined on the panel by John Millington, head of estimating at Sellafield Ltd, and Jeremy Richards, Project Management Improvement at Magnox Ltd.

"We've talked a bit about the problem of sharing information, whether that be commercial issues, or completeness of information, but we'd like to ask you what you think. Don't expect us to tell you everything, but we'll run through some of the challenges we currently see," Jenkins said. He listed these challenges as: culture; project; financial; design/engineering; and capability.

Major change

At the same time as moving towards performance delivery, both Sellafield Ltd and Magnox Ltd are undergoing structural and ownership change.

In April last year the Nuclear Decommissioning Authority (NDA) become the owner of Sellafield Ltd, the site licence company (SLC) responsible for managing and operating Sellafield in Cumbria. The new arrangements replaced the parent body organisation model and therefore ownership of Sellafield Ltd by Nuclear Management Partners. Sellafield Ltd published its first corporate strategy since the change on 20 April.

Magnox Ltd, owned by Cavendish Fluor Partnership (CFP) Ltd, is the management and operations contractor responsible for safely managing 12 nuclear sites and one hydroelectric plant in the UK. In March, the NDA announced it had decided to terminate its contract with Cavendish in 2019, which is nine years earlier than planned.

"Magnox has been decommissioning probably for longer than Sellafield, but both have extensive experience and a lot of project knowledge. Sellafield's current focus is on hazard reduction, operational activities and constructing new facilities; so not quite yet at the full-blown decommissioning phase," Jenkins said.

"Contrast that with Magnox, which has been decommissioning for a number of years. Magnox is into a situation now of waste retrieval, where buildings are coming down. They're looking for a means to retrieve and encapsulate the materials, whether they be hazardous, radioactive or just conventional on the decommissioning sites. But what we have between these organisations - one which is a very congested and complex site and the other being spread out over a large area - is mature plans."

Sellafield Ltd is enjoying "exciting times", Millington told delegates.

"We've just released our first corporate strategy now that we are a wholly owned subsidiary of the NDA. There's been a lot of preparation for transformation, we've got a new board formed and a new executive team preparing for the end of reprocessing round about 2020, but clearly our key strategies remain safe and secure stewardship of the site, progress and of course return on investment.

"On top of all that Sellafield remains probably the most hazardous and complex nuclear facility in Europe. The site is near the west coast of Cumbria and covers an area of about six square kilometres. We've got about 200 facilities housing nuclear waste, ranging from low-level right to extreme high-hazard. There's been a lot of talk at this conference about costs, but some of the engineering behind the numbers, some of the technology and science that’s going on, we're doing some incredible things."

Richards, previously head of project controls and estimating at Babcock Nuclear (now Cavendish Nuclear), highlighted the fact Magnox comprises 12 nuclear licensed sites that are geographically dispersed all around the UK. Richards was seconded to Magnox from Cavendish Nuclear in September 2014.

He said: "I came in as part of the parent body organisation in 2014 and there were two SLCs in existence at the time - RSRL and Magnox. We've brought them into one entity. We inherited both the decommissioning and generating emphases because when we took over we were still in the process of generating at one site, in Wylfa. That subsequently ceased generation some time ago and we are now in the middle of defueling. It's the last site to be defueled in the fleet. But essentially, in terms of the hazard, we are into waste management now." 


On the financial challenge, Jenkins said: "Annualised funding is a big issue for us; it means that we have to plan the work around an annual position."

He added: "The NDA has in its portfolio what we would consider to be mature plans, but it's government funded. So, unlike the German or Swedish scenarios where the funds are ring-fenced or defined, and although the mature plans are in place, they are effectively funded with taxpayers' money. So, what Magnox and Sellafield are really tasked with is demonstrating value for money."

Another challenge concerns design and engineering work. Jenkins said: "We might argue that we've got a very mature plan, but sometimes we can't define the work at the very outset. Equally, we can't define the end scope. We don't know how we'll leave it, whether it's brownfield, greenfield or somewhere in between. That creates a problem for us in terms of design and engineering."

The 'project' challenge concerns "unforeseen work and interfaces between projects", Jenkins said. This concerns the risks that can occur, "whether they are conventional risks, such as asbestos, or planning or scheduling risks when a project interfaces with other areas".

The 'culture' or 'change of mindset' challenge, Jenkins said, is "certainly an issue" when turning an operational team into a project delivery team. "Do you replace that team? Do you retrain them or do you bring in specialists to do that work? How do you get an organisation that has been operating for a number of years to a position where it can start decommissioning sites? The skillsets are very different."

Program-based approach

The 'capability' challenge concerns the availability of skilled labour and productivity levels, Jenkins said, and Richards added that Magnox Ltd has the advantage of having two parent bodies.

"We can draw on these two parent bodies in terms of expertise, from Fluor in the United States and Cavendish at different sites in the UK, such as Dounreay. One of the things we recognised, when we came in and we took over the SLCs, is that it's all about hearts and minds," Richards said.

"We inherited 12 sites that were quite disparate in their thinking, 12 fiefdoms effectively, that we had to roll into a generic culture that is still evolving. One of the ways that we've chosen to do that, rather than operate on a site-by-site basis and reinvent the wheel, is to put in a series of programs. We've programmised the whole approach, so that once we do an activity, once we knock a building down, once we retrieve waste, once we devise a new product, we roll that approach out consistently against the fleet.

"We try and actively promote that lead-and-learn concept. It has taken some time to embed, but it is starting to pay some rewards and capture all this learning to go forward."

Jenkins asked one of the delegates - Joseph Boucau, global director of decommissioning & dismantling and waste management business development at Westinghouse – about his experience of a "programmised approach".

Boucau said: "Westinghouse is thinking in a broad manner. We are thinking multi-site and also about the cultural mindset, the local aspects, needing to work within the local context in terms of the customer and the safety authority, and being totally aligned with local expectations. With design and engineering you need to be very clear with your customer about the scope of work and what the boundaries are. In terms of capabilities, you need to have the right skills and, to dismantle a heavy component, you need people who are capable of understanding the challenge and the risks involved from having learned lessons from other projects. It's important to be in production mode; if you can reproduce what you've done before, you can reduce the burden financially and this means less cost to the customer."

Asked about sharing data, Boucau said: "We are continuously improving. Innovation is important, but we are decommissioning here and not going to the moon. In some discussions people are thinking these are sophisticated things; yes, they are sophisticated but we need to control sophistication to a certain level and make it pragmatic and still with the mindset of improving from lessons learned."

Richards said a "big issue" for Magnox Ltd - and the wider estate - is that "the full extent of the waste volume involved is not fully known".

"It's understanding what's behind the closed doors, so to speak. The liabilities continuously grow because as you approach a new coal face, you can identify new issues that you haven't encountered before. We've also had regulatory compliance to adhere to - recognising that we have rules and regulations to adhere to and frameworks to work to, but whilst some of those can enable, some can slow progress of true decommissioning," he said.

An example he gave of "things that should have been picked up and provisioned for" was asbestos found at the Bradwell site. "One thing we started in Magnox in 2015 is, we went in to revisit the liabilities and identified asbestos that was not in the plan. We built in everything we could see but, essentially, it's one of those moving feasts that once you start scratching the surface you often find what's beneath and it's sometimes chasing things to the nth degree. The longer you take to define or change the end state has an ultimate impact on the decommissioning strategy, the durations, extends the schedule etc. and it actually adds to the liability cap.

"So, it's about understanding what we've got. I think now we're in a stronger position now than we ever were," he said.

"Other examples of liability increases include changes in policy - we've got the Scottish highly-active waste policy for example that was a big input into the liability last year," he added.


Boucau stressed the importance of operational experience before decommissioning starts.

"There's the need to conduct a lot of tests and training to make sure the first-of-a-kind project does not become the last-of-a-kind because that would be a disaster. We should use as much as possible from practice that exists perhaps somewhere else, including in the operational life of a plant," he said.

He referred to the Fort St Vrain high temperature gas cooled reactor, located close to Denver, USA, for which Westinghouse received the contract to perform its dismantling in 1990.

"Westinghouse dismantled this graphite reactor underwater. Why did we do that? To minimise the dose risk," he said. "We applied a concept we used to have for refuelling PWRs or BWRs; so, a flooded pool that meant personnel were well shielded. This is the kind of thinking people should have; it may be out-of-the-box thinking, but something that has been done in a different context can have value." 

Richards referred to the pioneering use of divers to clean former cooling ponds at the Dungeness nuclear power plant in England. This was a first-of-a-kind project in the UK, but not in the USA. He said it was difficult initially to get the divers' consent, but the project had been so successful that Magnox Ltd is planning to use divers at the Sizewell plant as well.

Millington highlighted two areas Sellafield Ltd has started to focus on recently.

"We've done quite a bit of analysis on the longevity of the projects. One area in particular I think we need to take a much more critical look at is the project initiation and that front-end engineering phase. There is a trend we've seen where the construction durations of some of our main projects tended to be relatively static and predictable, but what we've seen is considerable variation in that from the preliminary concept stage through to the detail. So that's an area to certainly focus on," he said.

Boucau's 'unique one-of-a-kind' comment resonated with practice at Sellafield, Millington said. "If we collect that data, we need the granularity to make sure some of the mistakes we made are measurable and can be fed back."

Budget setting

The other area Sellafield is trying to focus on is "early stage budget range setting", he said. 

"If you think of the longevity and complexity of some of these projects, it's critical that the first range is set realistically. All the work we're trying to do is to support and help us in those study estimates, the early concept estimates where you have little or no information. There's a drive for that predictability. Government and stakeholders are looking for almost single numbers. So that's an area we're really focusing on at the moment and we need some help to do that. It would be really powerful and useful if we were all looking at similar types of metrics and shared information."

Jenkins asked whether Sellafield had "reached the headline peak yet".

Millington said: "We're getting there, but clearly, there's a long way to go. I think it's recognition that what we're doing is first-of-a-kind, it's unique, and sometimes we discover things along the journey. It's that level of realism that budgets need to be based on."

Simon Carroll, a senior analyst with the Swedish Radiation Safety Authority, said a lot of cost estimates on decommissioning "are not that well informed". The OECD-Nuclear Energy Agency (NEA) - together with the International Atomic Energy Agency (IAEA) - plans to publish soon new guidance on improving cost estimates of nuclear decommissioning projects. The guidance will be within a report produced by the OECD-NEA Decommissioning Cost Estimation Group, which Carroll chairs.

Carroll said: "There are core elements of cost estimates that are not well informed by solid knowledge. They are assumptions, they are basic judgements that are made on the limited knowledge that we have in some areas, but we need to acknowledge where that knowledge is limited.

"To the extent that it is limited we are working with assumptions and it's only by going back once we have the experience and examining what we can learn to develop that platform to be confident about our assumptions. As Joseph [Boucau] was saying, the replicability of projects gives a reason to be confident. We're confident not just because we like the number, we're confident because we have something that that number is grounded on. And the only way to do that is to validate the assumptions that we had by comparing what has actually happened. And the only way to do that is to get the data."

Martin Stevenson, executive director of operations at Jacobs Engineering, said an understanding of psychology is useful. "A lot of we technologists and engineers dive into the detail to try and understand things and, as we get closer to thinking we understand, there's an optimism bias that makes us believe we're getting close to the truth. That's the way it's been for generations and I don't just mean decommissioning and nuclear.

"The notion is, 'Well people tell me it costs this much and my experience is it always costs that much'. If you look at the UK government's Green Book approach, the sorts of contingencies that people put on projects are vastly different than the ones generally thought of in an engineering and project delivery context. There's the matter of how you look at things and not losing the wood for the trees."

Lessons learned

Richards described approaches to standardisation and simplification that Magnox Ltd is using.

"The CFP way of standardisation is, through the repeatability, or trying to get something done in a pragmatic way numerous times to reduce that first-of-a-kind experience.

"My experience with Magnox was it had 12 sites - ten sites previously – and ten different solutions. Now clearly that's not efficient, that's not the way to do things. But equally you restrict innovation if you standardise things because if you keep on doing things the same way, you never get any new thinking, like the diving at Dungeness, for instance.

"Agricultural approaches to decommissioning have been used in the UK for years; 'agricultural' being the machinery that's applied to it. It makes no sense to try and become overly complicated designing retrieval systems which when they break down you need to design another retrieval system to bring the original one out. However, we understand that some of the areas on the sites are difficult to access and some of the materials in there are unknown. Whether it's a vault in Berkeley or a shaft in Dounreay, there are lot of examples that are first-of-a-kind, but I think simplification is clearly something that we can all strive for. That can be a process rather than just the design and construction of something because the processes in many instances can slow us down," he said.

Jenkins added that it is also important to "plan at the appropriate level". Referring to Enkom's recent exercise with NuGeneration, to assess the UK nuclear power plant developer's decommissioning and waste management plan, he said: "Now that plant when it's built won't be decommissioned for 100 years, so why build a decommissioning plan now and at the level of detail that is unreasonable, in too much detail? Using that as the far end of the scale and using perhaps Magnox as the near end, planning and costing the work needs to be appropriate to when the work's going to be executed. That determines the level of effort put into the accuracy of the estimates and the application of a contingency," he added.

Millington said: "We have to look at the value of putting detail to something that's going to happen in 50 years' time. We try to use common and consistent breakdowns of cost and work, so depending on where you are with a project you've got those various levels and at least one higher level breakdown you can monitor those to. Then, as that evolves, it becomes populated with more detail.

"In terms of simplification in the design process, we can look at commercial off-the-shelf solutions and ask whether that would fit the purpose, rather than an over-engineered nuclear solution. There are a lot of initiatives going on for us at the moment," he said.

Richards added: "It's interesting that you can look at a schedule with 200 activities and another site might have 1000 activities and by virtue of having 1000 activities you can find that the cost escalates. There are reasons for this, but you sometimes find that the more granularity you have, the more the estimate and the schedule will grow. It's really just about trying to get that appropriate level of information."

Jenkins added that Enkom had worked with Magnox on a "simple cost breakdown structure" in line with the NDA's approach to "incentivise Magnox to capture the information and reuse it".

"But that might prompt a conversation about sharing of data. In the UK, you have government funding and a lot of freedom of information. So, how can we share? One way might be through standardised coding, meaning if we can all talk about the same things and use the same sort of structure, then that's going to help us hugely, though what level of detail it has may be open to debate," he said.

"We could also share normalised outturn costs. There aren't many reactors that have been decommissioned, but what there has been is a lot of projects undertaken, so what you need to do from our experience is look to those projects. It takes 15 years on average to pull a reactor down, but you have to wait a long time to get that information and feed it back into the cycle. So maybe we need to look at parts of the project and not the entire project."

He noted that information is "commercial property that everybody, if they invest in it, would like to hang on to", but there's a "huge benefit" to sharing information.

"How we do it is a matter of face. It could be through various organisations - the Nuclear Industry Association, the International Atomic Energy Agency, among others - but we'd all like to get the costs down. It's a strange fact that we probably spend more money pulling things down than we do building them in the first place," he said.

Carroll noted that the California Public Utilities Commission (CPUC), when deciding on the decommissioning of San Onofre units 2 and 3, had required that the utility report its decommissioning costs in relation to the estimate that they had approved. This meant that any major changes to the costing had to be reported and clarified to the CPUC, followed by regular reporting according to the same structure as the cost estimate.

"That was a regulatory decision. Obviously, where government entities are involved, they can put impositions on utilities to report costs in certain ways, but there must be a win-win in here that people will want to share data in a good way in order to improve their project delivery, rather than impose an artificial decision," he said.

Richards said methodologies for costing had been introduced at Magnox Ltd. "The way it's structured now is definitely better than what we inherited. For example, we've got one consistent lifetime plan instead of the 12 inconsistent plans we inherited. We can give one report from a portfolio now that we can monitor against our funding and drill right down to whatever level of activity we want in the plan."

Millington added: "Since I've been at Sellafield, I've had two really good and sensible discussions with the NDA about realistic expectations, but cost estimating is the hardest game in the world."

Researched and written
by World Nuclear News