New thinking now needed for new-build nuclear

Public interest in Britain’s plans for new nuclear power plants will be re-ignited by events at Fukushima and details like evacuation plans, which before seemed distant, will now seem very real for communities near the proposed sites.

There will be major implications for the specific technologies proposed. It was five years after the Three Mile Island (TMI) accident before it was possible to get into the reactor and find, to everyone’s surprise, that a significant proportion of the fuel had melted. At Fukushima, there are four reactors affected, each one probably much more seriously damaged than TMI, so the post mortem on the site will be lengthy and hazardous.

In the past 35 years, there have been four major events that have had a profound influence on nuclear design. All these events would have been regarded by nuclear designers as incredible before they happened and all took at least a decade between the event and new designs being available that took full account of the lessons from these accidents. In 1975, at the Browns Ferry site in the USA, an electrician’s lighted candle was dropped into a cable tray causing a fire that disabled the control systems for all three reactors at the site.

Although there were no significant immediate health and safety consequences and the damage was relatively straightforward to repair, this led to a requirement worldwide for independent systems for each unit at new sites.

The 1978 TMI accident, when the core-cooling system was thought by the operators to have failed (it actually had not), the core was uncooled for several hours leading to a partial meltdown. This led to a requirement for ‘defence in depth’ so that if a safety system failed, there was an independent system to replace it.

Chernobyl had a range of implications despite the huge differences between that technology and those used in the West. In particular, ‘passive safety’, in other words, if everything else failed, the reactor would naturally tend to go into a safe condition, became an important requirement for new designs.

9/11, despite it not involving a nuclear reactor, led to a requirement in the West that new nuclear plants should be able to withstand an impact from a full-size passenger aircraft. Chernobyl and TMI took a particularly long time to work through because of the difficulty of access to the wreckage at the sites and the difficulty of piecing together what had happened.

Existing plants

A particularly difficult issue when a major event with safety implications occurs is how to deal with existing plants especially where it is not economically feasible to modify them to meet the new requirements. In 2007 in Japan, an earthquake near the Kashiwazaki Kariwa site (only 6.6 on the Richter scale), where seven reactors are sited, caused structural damage at some of the units and all seven were temporarily shut down.

Four years later, three units are still out of service. Two reactors at another site not directly affected by the quake were permanently closed because it was judged that their ability to withstand earthquakes was not adequate. Unit 1 at the Fukushima site was completed in 1970 and was one of the oldest operating plants in the world. Plans worldwide to extend the lives of existing plants from the expected 40 years to 60 years will come under renewed scrutiny and the extent of updates required is likely to be increased.

For the UK, earthquakes are probably not going to raise many issues but inundation of the sites for example, from tidal surges and sea-level increase will be of concern. Other issues that will be difficult to resolve are sequences that knock out all electrical supplies to the plant and spent fuel ponds that are outside the relative safety of the reactor containment. None of the UK plants have spent fuel stores within the containment and many contain much more fuel than they were originally designed to take.

The use of plutonium in Mixed Oxide (MOX) fuel, even before Fukushima a highly dubious proposition, is likely to come under renewed scrutiny, especially if the condition of reactor 3 at Fukushima is significantly  complicated by the use of MOX at that unit.

As with Chernobyl, the UK nuclear industry will argue that all except one, the Sizewell B PWR, of our operating reactors are of unique British designs. They use gas rather than liquid as coolant and this makes ‘loss of coolant accidents’ less problematic.

Sizewell B was ordered in 1987 so should have most of the lessons of Browns Ferry and TMI incorporated but is much more like the Fukushima than the other plants.

Two out of eleven of Britain’s old ‘Magnox’ stations (both comprise twin reactors), Oldbury (the oldest operating plant in the world) and Wylfa, are still in service. One reactor at Oldbury is scheduled to be closed in June 2011 and the other a year later, while both reactors at Wylfa are due to have closed by the end of 2012.

The fact that these old plants are still operating seems to have much to do with the needs of their owner, the government’s Nuclear Decommissioning Authority, which is desperate for the income they generate to fund its main job of cleaning up the existing nuclear sites.

Seven out of eight of the other operating reactors are Advanced Gas-cooled Reactors (AGRs) and five of these were designed and ordered in the 1960s (the other two being ordered around 1980). Two (Hunterston B and Hinkley B) were completed in ten years in 1976, while the others encountered huge problems in construction and were only completed in the mid-80s but all five are, in many respects, very old plants. The AGRs had a design life of 25 years but the UK safety authorities extended their lives, in the case of Hinkley and Hunterston for a third period of five years to 40 years. Whether the line can hold that AGRs are too different to Fukushima for there to be much to learn remains to be seen.

New build

However, the big question is whether Fukushima will derail the UK’s attempts to launch large-scale nuclear ordering. The Generic Design Assessment (GDA) was launched in Britain in 2007 and was aimed at providing full generic safety approval for at least two new competing designs of reactor. This would mean that a company wanting to build one of these designs in the UK would know that all major safety issues had been resolved leaving only site-specific issues to be sorted out.

Two designs are being examined in the UK, the European Pressurised Water Reactor (EPR) supplied by the French company Areva and the AP1000 supplied by the Westinghouse company, now owned by Toshiba. No plants of these designs are in operation anywhere in the world and they have not received full generic design approval anywhere. Two EPRs are under construction in Europe, but as has been widely discussed elsewhere, both are way over budget and very late.

The UK safety authority, the Health and Safety Executive (HSE), was given a requirement to complete its review in July 2011 and it seems determined to keep to that target. In practice, the HSE acknowledges that all it will be able to give then is conditional approval with some significant issues still to be resolved, perhaps requiring a year or more.

The conditional approval likely to have been given in July 2011 would have been essentially worthless because it was not clear when, if at all, final approval would be given and what additional requirements would be needed to gain it.

Until approval is given and the design finalised, no firm order can be placed and no site work allowed. A parallel process of generic review is also underway in the USA, but, before Fukushima, it was not expected to be complete for the AP1000 before end 2011 and for the EPR till mid-2013.

The Fukushima accident seems likely to set these generic design reviews back indefinitely. How can a regulator credibly claim that a new design takes account of all the lessons from Fukushima when it is still not known exactly what happened, much less what is needed to ensure it can’t happen with the new design? It is hard to say at this stage whether EPR or AP1000 will come out better from this process. AP1000 relies much more than EPR on ‘passive’ safety design features. EPR relies on mechanical systems which, it may be judged after Fukushima, are too prone to failure even when there are multiple back-ups.

On the other hand, EPR, unlike the AP1000, has a ‘core-catcher’ that is designed to ensure that if the core burns through the bottom of the reactor vessel, it remains isolated from the surrounding environment. If it turns out that there was significant fuel melting and the reactor vessel was at risk, designs including a core-catcher, a feature that adds significantly to the construction cost, may be mandatory. The most likely outcome is that both designs will require significant modifications adding to a construction cost that was beginning to look prohibitively high even before.

Insurance and liability

The clean-up bill for Fukushima will be monumental. At TMI, it took about 15 years and US$1bn for the clean-up including removing the damaged fuel and decontaminating the site. No start has been made on actually removing and decommissioning the damaged reactor, but current estimates are that a further US$1bn will be needed to complete that. At Chernobyl, all that has been done is to try to contain the radioactivity with a massive ‘sarcophagus’. How the site can be decontaminated and the wrecked reactor removed is anyone’s guess.

Fukushima seems likely to be somewhere between TMI and Chernobyl in terms of clean-up. Owners of plant are fully responsible for the cost of cleaning up the site and the impact on insurance premiums is likely to be large. However, their third party liability is limited by International Conventions (Paris/Brussels Conventions). The coalition government had proposed to increase the limit seven-fold from £140m to £1bn before Fukushima.

This increase might be large but it has already been dwarfed by the costs imposed by Chernobyl, and Fukushima may well reinforce this point if there is significant off-site contamination.

Likely outcomes

Already, EDF has said there is no need for delay but this looks like whistling in the dark. EDF grossly overpaid for Britain’s existing nuclear plants, paying about €15bn in 2009 for the 7 AGRs and Sizewell B, plants that had bankrupted their previous owner in 2002. This caused EDF’s level of debt to spiral and caused them to sell off any assets it could claim were not ‘core’, for example its UK electricity networks and its German energy business — bringing its debts down by €20bn to €34.4bn. If a few of the AGRs are forced to be retired early or even if they just have to be closed for modifications, the sum paid for British Energy will look even more misguided.

If UK plants are closed early, the government and EDF might try to use the gap in our electricity supplies they create to argue for a more ambitious and more rapid nuclear programme. This is the oldest and most threadbare tactic in the nuclear industry’s playbook — dreadful things will happen if we don’t launch a new nuclear programme immediately. However, the reality is that the five oldest AGRs would probably all have been off-line by 2020 anyway, well before any new nuclear capacity could come on-line, so that argument will not hold water.

Chris Huhne has announced that the HSE will carry out a review of the implications of the Fukushima accident for the UK to report within six months but the government is deluded if it believes the issues raised for the UK by Fukushima can be resolved in only six months. Even such determined nuclear proponents as China have ordered a pause for breath.

The logical response is to assume that new nuclear construction cannot start much before 2020 and new capacity cannot be on-line till after 2025. That means we must accelerate options, such as energy efficiency and renewables that can be deployed quickly, cheaply and reliably and without increasing our greenhouse gas emissions. However, logic has been in short supply when it comes to government nuclear power policy in the UK since Blair claimed it was ‘back with a vengeance’ five years ago.

Steve Thomas is Professor of Energy Policy at the University of Greenwich.