How bad regulations make infrastructure expensive

It's more complicated than you think

Whether it is a high-speed railway from London to Manchester Birmingham, a road tunnel under the Thames, or a nuclear power station in the South West: when Britain builds, it almost invariably does it far more expensively than our international peers.

Britain Remade has reviewed over 300 infrastructure projects from around the world. Only the Americans (who are a third richer) manage to build infrastructure more expensively than us. Even then there are categories where we are in a league of our own: Hinkley Point C nuclear power station is on course to be the most expensive ever built.

Some achievement when the dubious honour of ‘world’s most expensive nuclear power plant’ belonged to a project (Vogtle 3 & 4 in the US) where almost everything that could go wrong went wrong. Regulators insisted on major design changes extremely late in the day, important structural work had to be redone at great cost, and Vogtle’s designer and initial project manager Westinghouse even filed for bankruptcy during construction.

Caption: This account is not wholly inaccurate.

Britain Remade’s focus has primarily – though by no means entirely – been on the impact of bad rules and regulations. We have written extensively about bat tunnels and fish discos, about £300 million pound planning applications containing 30,000 page environmental impact assessments, and failed legal challenges delaying work.

Yet, at least at face value, there’s a challenge to our work. Planning regulations may produce absurdities on a regular basis, but can they really explain why HS2 is nine times more expensive than the Tours-Bordeaux TGV line? Or why Hinkley Point C is more than twice as expensive as Finland’s Olkiluoto 3 power plant despite using the same reactor design?

Few would argue a £120m bat tunnel is a good use of money – even the most vociferous of bat advocates would concede the money could be spent to help bats more effectively elsewhere. Still, £150m is a small fraction of the overall cost of HS2’s £62bn and rising budget. Likewise, the decision to bury large parts of the railway to protect views in the Chilterns will have clearly added costs. Yet, it is unlikely the decision added much more than a few billion to HS2’s bill. It may be that HS2’s price tag could be the sum of hundreds of bat tunnel-esque decisions, yet as it stands we haven’t uncovered enough to total a tenth of HS2’s eye-watering costs.

Likewise, when we look at nuclear – it isn’t hard to find examples where Britain’s regulator has insisted on expensive design changes for extremely small reductions in radiation exposure. Whether it is Hinkley Point C’s separate hardwired control and instrumentation system, the bulky HEPA filters that the Office for Nuclear Regulation (ONR) insisted be installed on every vent and duct of the ABWR (to prevent one banana’s worth of radiation exposure), or the requirement that Westinghouse use a completely different method for managing spent-fuel, it is clear nuclear plants are more expensive than they ought to be due to unnecessary safety measures.

Yet when I’ve spoken to industry experts they’ve suggested these specific changes, while infuriating, only added a few billion directly to the cost of UK projects. (Neither the ABWR or Westinghouse’s AP-1000 were actually built.)

So, is the focus on bad or, more precisely, disproportionate safety and planning regulations misguided? Should we instead focus on other problems – such as supply chains or a lack of skilled engineers in government able to scrutinise spending decisions?

A new study from Zachary Liscow, Cailin Slattery, and Will Nober suggest the latter can have a huge impact on the success of a project. Looking at US motorway projects, they find that when talented highway engineers retire, costs increase. In fact, switching from a decent engineer to a good engineer can cut a project’s costs by 14%.

Clearly, regulation isn’t the only thing that matters, but it is more important than the totting up above might suggest.

When doing the research for Britain Remade’s Policy Playbook for Cheaper Nuclear, I spoke to Tony Roulstone. He established and teaches the Nuclear Energy Masters programme at Cambridge, but before he moved to academia he was a senior director at Rolls-Royce working on aerospace and nuclear engineering. In short, he knows a thing or two about how to deliver complex engineering projects, like nuclear submarines, on budget.

Our conversation clarified that there were two ways regulation could increase costs. There were of course the direct costs of installing bulky filters on every duct, hiring ecologists to run lengthy surveys, and constructing a bat tunnel. In some cases, these direct costs can be very large. The cost of planning and permitting for the Lower Thames Crossing – a tunnel between Kent and Essex – is twice as much as what it cost Norway to build the world’s longest road tunnel.

Direct costs are only part of the story. The real issue is that regulation can act as an impediment on the methods and strategies that are essential to reducing costs like the fleet approach that has enabled South Korea to build nuclear plants at a sixth of the cost of us.

There are big advantages to modular construction with fewer complex civil works. Modular construction typically involves upfront investment in production lines that needs to be spread across projects. When projects are delayed or cancelled, production lines sitting idle can effectively wipe out the advantage of this approach.

Legal and General attempted to apply the modular approach to housing. They built a factory and employed almost 500 people. The aim was to build homes that were not only cheaper, but greener too. In theory, the upfront investment in a factory could be spread over hundreds of projects across the UK. The problem was every single project they tried to build saw lengthy planning delays (and in one case, outright refusal). The factory closed soon after.

It is near-impossible to quantify these indirect costs of regulation, yet they are real and are why regulation can have a much larger impact on costs than it appears at first glance.

In the case of nuclear, we identified six key ways that regulation blocks proven infrastructure cost-reducers. (Below from A Policy Playbook for Cheaper Nuclear)

1. Limiting gains from learning across projects: There is strong evidence that First-of-a-Kind (FOAK) nuclear projects tend to be substantially more expensive than the third, fourth, or even tenth plant in a fleet. Construction teams become more productive and are less likely to make preventable mistakes as experience with a design grows. Welding at Hinkley Point C’s second reactor building is being done at four times the pace as welding at the first reactor building. This is estimated to generate a 30% saving. If frequent design changes and project-specific environmental mitigations impede standardisations, it is likely the cost savings associated with a fleet approach are less likely to be found.

2. Making it harder to innovate in reactor design and construction: In most markets, innovation is a key driver of cost-reductions. For example, solar panel factories have changed substantially over the last 15 years and have become more productive as a result. In nuclear, changes must be justified with detailed safety cases and can trigger planning delays. Hinkley Point C was required to gain additional planning permission to change its approach for storing nuclear waste based on operational experience.

3. Deterring investment in supply chains: One key way the fleet approach to nuclear saves money is by creating a strong incentive to invest in the nuclear supply chain. Nuclear construction involves not just specialist skills like nuclear welding but also specialist equipment. For instance, the world’s largest crane, ‘Big Carl’, lifted a 245‑tonne dome onto Hinkley Point C’s first reactor building. Costs can be saved when expensive investments in workforce training and machinery can be spread across multiple projects. If uncertain planning processes, legal challenges, and the risk of design changes lead to stop-start construction then these investments will either be more expensive, or not be made at all, such as by using foreign welders instead of training new British welders.

4. Creating barriers to entry to disruptive competition: Competition forces companies to stay lean and to innovate. In markets where only a few producers can profitably operate (e.g. industries with high-upfront costs such as nuclear), market leaders keep prices low and operate efficiently to guard against innovative competitors who have the potential to undercut them offering an entirely new product. Expensive licensing and planning processes can create a significant additional hurdle for new innovative competitors who bear large costs before a single order is made.

5. Increasing the risk associated with financing fleets: Building fleets, as Britain did in the 1960s, France did in the 1980s, and South Korea does now, means that large upfront costs can be spread over multiple projects and leading to substantial ‘learning-by-doing’ between projects. However, the fleet model, in order to work properly, requires that nearly every decision, especially planning and financing, are ‘one and done’ at the beginning. No developer or operator can make a fleet model investable if they are subject to enormous risks that there will be planning delays or required design changes.

6. Preventing the use of modular construction through increased design complexity: Additional environmental mitigations and safety features such as a hard-wired backup control and instrumentation system increase the complexity of plant designs. As a result, projects are less able to use modular construction (e.g. built in a factory assembled on site) and more reliant on civil works with greater construction risk.

At Britain Remade, we are sometimes asked by policymakers for a cost-breakdown: What share of the blame for high infrastructure costs can we attribute to planning, to safety regulation, to supply chains, and to factors like geology? I often struggle to give them the answer they want. The problem, with the exception of the last one, is that they are all inter-related.

Take the paper from Liscow, Slattery and Nober, they have strong evidence that employing better engineers is a reliable way to cut costs. Yet, it seems fanciful to believe that the British state could not only employ a corps of highway engineers, but pay them enough to keep the best from going private. In the years where roadbuilding programmes are halted by legal challenges – as they often are – the highly-paid engineers sitting around doing nothing will be the first on the Chancellor’s chopping block.

Fewer unnecessary bat tunnels, lengthy environmental impact assessments, and time-wasting legal challenges will not cut the cost of building new nuclear power stations to South Korean levels on their own, but fixing them is essential. There is so much more to do to bring down costs – from investing in fleets to building up a pool of skilled engineers, builders, and welders – but each necessary step is made so much harder by long, unpredictable planning processes and safety regulations that seem to demand constant change in the pursuit of tiny improvements in safety.

To read Britain Remade’s plan to reform regulations so they don’t impede proven cost-reducers, click here.