黑洞社区

Last week energy secretary Ed Miliband promised a 鈥済olden age鈥� for nuclear power as the government confirmed it is pushing ahead with the 拢14.2bn Sizewell C project in Suffolk. It has done this despite the fact that investor interest in building nuclear in the UK has nosedived in the wake of EDF鈥檚 2016 deal to build Hinkley Point C in Somerset.

That scheme is now massively over budget on construction 鈥� EDF has said it will cost almost 拢47bn (in current prices) compared to the 拢10bn the firm said it would cost in 2010.

While the 拢92.50 鈥渟trike price鈥� deal with the government 鈥� a guarantee to buy energy at a certain price 鈥� has protected bill payers to an extent from these cost rises, it also ensured a big risk premium was priced into the job from the start, making nuclear power expensive, with the saying a risk-sharing approach could have been cheaper. The deal 鈥渓ocked consumers into a risky and expensive project with uncertain strategic and economic benefits,鈥� the NAO said.

Either way, as EDF and other nuclear promoters have struggled financially, no one has rushed to repeat the exercise in the decade since. Until now.

So, what has changed in order to make Sizewell C 鈥� a project EDF once promised would be operational by 2025 鈥� viable again?

The difference, as Miliband confirmed to the House of Commons last week, is that Sizewell will be financed under a totally different model to Hinkley Point. Known as the regulated asset base, or RAB, model, it is a tried and tested system for raising private finance from bill payers for large programmes of infrastructure investment.

The RAB model is that used since the privatisation of UK utilities to draw investment into monopoly providers鈥� infrastructure, while continuing to ensure value for money for their consumers. Crucially, it is also the model under which the 拢4.6bn Thames Tideway project has been successfully delivered over the past decade. 

The Thames Tideway Tunnel is a 25km 鈥渟uper sewer鈥� built along the tidal section of the river between Acton and Abbey Mills. It is designed to capture almost all the raw sewage and rainwater which would otherwise overflow into the river during heavy rain. Construction began in 2016 and it finally became operational at the start of this year. 

While the exact form of RAB to be used is not yet clear, the government has  that using this model could save bill payers up to 拢80bn on new nuclear compared to a Hinkley Point C-style system. So, what is the model, how can it make projects cheaper, and what are the prospects of it being used on other major schemes?

The RAB utility model

The RAB model has been responsible for ensuring billions of pounds worth of investment in privatised utilities over the past four decades. While critics of the service provided by water and power companies would certainly question just how effectively it has operated in that period, no one can question whether or not it has successfully levered in private cash.

For example, water industry trade body Water UK estimates that UK firms invested 拢236bn of privately raised cash in water networks between 1989-90 and 2023-24, all generated under the RAB model.

Beth West, director at Navigate Advisory and an adviser on 黑洞社区鈥檚 Funding the Future initiative, worked on Thames Tideway briefly prior to the construction phase. She says: 鈥淚t鈥檚 a really common model, used by utilities all the time, such as for power stations. It totally makes sense when you鈥檝e got a customer paying through a bill for a service.鈥�

In basic terms, the RAB model works in utilities where you have a monopoly company that looks to raise money against its existing assets, its income and its operational track record to invest in its infrastructure. A regulator is established in order to ensure that costs and profits are controlled, given the lack of competitive tension, while providing investors with assurance over financial stability.

The idea is that companies borrow money, with investors paid a return out of a surcharge on customers鈥� bills. The regulator vets spending to ensure that only qualifying investment is allowed to be added to the 鈥淩AB鈥� and thereby clawed back through customer bills.

The key conditions required for a RAB model to work are, therefore, a base of customers (typically, utility bill payers), a regulator seen as credible by the market, a monopoly, and a market willing to fund the company on the basis of regulated returns.

Project RAB

The innovation that happened around Thames Tideway was to apply this logic to a single major project, rather than to a company running a large programme of works. Tideway took advantage of bespoke legislation (namely the catchily-entitled Water Industry (Specified Infrastructure Projects) (English Undertakers) Regulations 2013) created to allow such a venture to happen.

What this allowed was for a special purpose vehicle to be established to build Tideway, that also became a regulated water company able to borrow under a RAB structure, with money paid back through Thames Water bills. Matthew Duncan, chief financial officer at Tideway, says this regulated structure, which capped costs and returns within known bounds, automatically provided a much greater level of confidence to potential shareholders and investors.

A key part of the investor attraction to the RAB model 鈥� compared to traditional project finance 鈥� is that investors can start getting a return on their money from the day construction starts. The regulator allows bill payers to start getting charged from day one, as opposed to at project handover, which in the case of Tideway would have meant a delay of 10 years.

This means that institutional investors are much more likely to be interested, and Tideway鈥檚 Duncan says 拢600m has already been generated from bills through the RAB model during the construction phase, keeping borrowing costs down.

Craig Elder, partner at law firm Browne Jacobson, says: 鈥淥ne key advantage of RAB is that borrowing costs can be reduced by passing construction costs onto consumers through bills before the project is completed 鈥� reducing the need for borrowing and interest payments through the riskier construction phase.鈥�

Tideway in the end secured 拢1.3bn from its initial investors, and has drawn in a further 拢3.6bn from lenders and the bond markets, Duncan says.

Amar Qureshi, joint CEO of infrastructure consultant Agilia, which is now advising the government on Sizewell, formerly led the Thames Water team involved in creating the structure for Tideway. He says the structure, given the regulatory mechanisms and incentives, allows investors to treat projects as 鈥渕ore akin to investment in a newly established regulated utility company, as opposed to an investment in an infrastructure project exposed to big gnarly construction risks鈥�. Crucially, this helps reduce the cost of borrowing money.

The structure allowed Tideway to reduce borrowing costs to such an extent that initial estimates in 2011 that the project would see 拢80 added annually to Thames Water customers鈥� bills has proven way over the top 鈥� in the end customers will pay just 拢25 per year (in 2016-17 prices).

The reduction in the cost of capital was 鈥渁bsolutely鈥� behind this saving, says Duncan. 鈥淚t made such a big difference.鈥�

Government backing

However, it is by no means the RAB model alone that gave investors such security. Tideway also benefited from a significant government support package. This took the form of a range of guarantees where the government agreed to step in 鈥� scenarios described by Agilia鈥檚 Qureshi as 鈥渉igh impact but low probability鈥�.

Most significantly perhaps, the government said it would cover costs if the construction price rose by more than 30% (or 拢960m) above the target cost of 拢3.2bn (at 2014-15 prices).

鈥淎s we developed the model for Tideway, it became clear that, on its own, it wouldn鈥檛 quite drive the pricing benefits that we wanted without a government support package,鈥� says Duncan. 鈥淏ecause with the RAB model you have quite a heavily regulated return, investors want to know how those construction risks are going to be managed 鈥� particularly those high impact low probability events. The support package gave that confidence.鈥�

鈥淥ne key advantage of RAB is that borrowing costs can be reduced by passing construction costs onto consumers through bills before the project is completed鈥� 

Craig Elder, partner at law firm Browne Jacobson

Dr Alex Budzier, a former McKinsey consultant, now fellow in management practice at the Sa茂d Business School, University of Oxford, says: 鈥淭his model is different to traditional procurement, in that essentially you鈥檝e got the capital provided by the private sector, and the risk taken by the public sector. The RAB with that support is a powerful combination.鈥�

This combination resulted in an investment-grade rating for the Tideway business which kept the cost of borrowing money low. The Tideway deal was nevertheless widely criticised for opening the taxpayer up to potentially limitless liabilities.

In the event, build costs on Tideway have risen, in part because of a 拢200m hit from closing and then reorganising sites during the covid pandemic. The final outturn price is now expected to be 拢4.55bn at current prices.

However, with inflation adjusted back to 2014-15 prices, this puts the cost at around 拢3.85bn, says Duncan, well within the point that would trigger the support package. So, the taxpayer will not have paid anything, but yet customers will have felt the benefit of the guarantees in lower bills.

Amar Qureshi thinks the significance of the RAB structure should not, however, be underplayed, despite an understandable focus on the government support package. He says the model allowed a more realistic approach to managing construction costs, in which all parties could respond to the true cost of carrying out the works, with external independent regulation ensuring this was done fairly.

In comparison, a traditional project finance approach would seek to pass the majority of construction risk onto a contractor, with the response being simply to price in a big contingency, paid whether those costs arose or not. Or to underprice the job, relying on pushing up the price post award.

He says: 鈥淎s a licensed special purpose vehicle, you鈥檙e remunerated via the RAB for effective expenditure and cost management. You aren鈥檛 expected to take cliff-edge risks, so your base returns aren鈥檛 dependent on whether something you can鈥檛 control or reasonably anticipate arises during the construction of the project.鈥�

Sizewell and beyond

The government 鈥� actually the previous Conservative government 鈥� has already legislated to allow the RAB model to be used in the nuclear sector. The Nuclear Energy (Financing) Act 2022 sets out the statutory provisions necessary for setting up a new regulator to oversee the construction of Sizewell C 鈥� and any other power stations that come forward.

Miliband told the House of Commons last week that the RAB system was 鈥渢he right system [for building Sizewell C] and will cut the cost compared with Hinkley Point C鈥�.

Certainly, some in the sector agree with him. Budzier says the system would work well for very big projects such as Sizewell C, which is large enough to generate the logic for its own quango to regulate it, and where 鈥渢he risks are quite large, but we do understand them鈥�.

Qureshi says it is 鈥渁bsolutely鈥� the case that an 鈥渁ppropriately structured RAB model with cover for high impact low probability risks鈥� could 鈥渟ecure low cost finance to help deliver much needed infrastructure such as Sizewell C, reducing the cost impact on bills for customers and consumers.鈥� An analysis by the Department for Business, Energy and Industrial Strategy published in 2021 found that a RAB model could save between 拢30bn and 拢80bn on the cost of new nuclear stations for bill payers, compared to using the system that funded Hinkley.

But, in terms of major projects, it is of course not only Sizewell C that could take advantage of the RAB. The water regulator Ofwat is exploring using it to fund the proposed 30 million cubic metre reservoir outside Abingdon, known as the South East Strategic Reservoir Option (SESRO), with some of the nine further reservoirs the government is considering before 2050 also looking over the model.

鈥淚 do see options for using this model in other sectors such as nuclear, but there鈥檚 got to be a strong understanding of the revenue streams it鈥檚 going to bring in鈥� 

Matthew Duncan, chief financial officer at Tideway

An assessment of SESRO by Agilia in 2022, undertaken for Thames Water, found the project 鈥渕ay offer the potential to generate greater value for money鈥� through the particular form of RAB that was used on Thames Tideway.

In addition, National Highways has publicly stated that RAB is one of three funding models it is considering for the 拢10bn Lower Thames Crossing project 鈥� with most believing it is the favoured option.

The right model

RAB does, of course, have its downsides. And it is not suited for every situation 鈥� social infrastructure construction, for example, does not have the customer revenue stream against which investors can invest.

And the requirement for a regulator, too, imposes a large cost overhead. The structure also relies on faith that the regulator will do its job 鈥� a faith being sorely tested in the wider water sector right now, given the situation in which Thames Water鈥檚 wider business finds itself.

West says: 鈥淚t [RAB] needs to have a strong regulator because you鈥檝e got a monopoly provider. Their job is to ensure that the company is making the right investments 鈥� if they鈥檙e not strong enough, it鈥檚 a problem.鈥�

Both the structuring of contracts and the offer of government guarantees, all taking risk off the private sector, also open up questions about what happens when the worst does happen. Browne Jacobson鈥檚 Elder says: 鈥淭he most obvious downside is the transfer of risk away from the contractor.

鈥淭here has been criticism that allocation of excess construction costs to taxpayers, for example, constitutes a 鈥渂lank cheque鈥� for the developer.鈥�

Hence Tideway鈥檚 Duncan says that, while RAB can work for nuclear, there are a number of necessary conditions that should be adopted, including certainty over revenue streams and an understanding of the construction risk.

He says: 鈥淚 do see options for using this model in other sectors such as nuclear, but there鈥檚 got to be a strong understanding of the revenue streams it鈥檚 going to bring in 鈥� some ability for investors to get visibility. That鈥檚 a really big driver of whether it will work.

鈥淎nd there鈥檚 got to be some understanding of construction risk, and how manageable that will be.鈥�

West says that, while RAB makes 鈥渢otal sense鈥� for reservoirs, nuclear remains 鈥渃hallenging鈥�. She adds: 鈥淚t鈥檚 just because of cost overruns and the scale. It鈥檚 not to say it鈥檚 not possible, but it鈥檚 just they鈥檙e big, they always overrun, and they take a really long time.鈥�

The RAB model has undoubtedly proved effective at drawing in cheap funding for Tideway, keeping costs down for London bill payers. 鈥淚t鈥檚 about having the right model for the right circumstance,鈥� West says.

Whether it becomes a force in the financing of mega-projects may be dependent on what projects end up getting the go-ahead.