- Lennart Baumgärtner

- Aug 19
- 4 min read
Executive summary: Britain’s water‑supply “megaprojects” are arriving: 30 schemes with ~£50bn of lifetime spend, most to be let via DPC or SIPR structures. The choice of delivery model shifts who bears risk and how cash flows are regulated—VAPRI shows 220-320 bps impact on required equity returns.
The UK’s water regulator, Ofwat, has proposed a pipeline of 30 “megaprojects” worth roughly £50bn in lifetime costs. 27 thereof will be delivered by SVP investors under competitive tender, with 11 due to start construction this regulatory period and most spend arriving after 2030. The menu includes nine reservoirs alongside transfers, recycling and desalination such as London water-recycling plants, Teddington and Beckton, and the Fens Reservoir. How these projects are regulated determines who underwrites what risks and what customers ultimately pay.
There are three current routes to build Britain’s water megaprojects: in-house under the existing price control; Direct Procurement for Customers (DPC) SPVs for discrete schemes over £200m; and Specified Infrastructure Projects Regulations (SIPR) SPVs for complex outliers. For DPCs, bidders compete on the all-in charge to design, finance, build, and run the asset. Revenue is paid by the commissioning water company and passed through to bills. SIPR is also a competitive route, but bidders compete on the return they require, which effectively sets the price; the winner gets its own licence and operates like a stand-alone, single-asset utility. Thames Tideway is the most prominent SIPR example, a success Ofwat aims to replicate. The labels may sound bureaucratic but have real impact on risk allocation and required investor returns.
Figure 1: Comparison between DPC and SIPR
Dimension | DPC — Direct Procurement for Customers | SIPR — Specified Infrastructure Projects Regulations |
|---|---|---|
Brief overview | The water company runs a competition; the winner builds, finances, and operates the asset. The water company pays an agreed annual charge and recovers it from annual bills | For very large/complex schemes, a stand-alone project company with its own licence delivers the asset and is paid by customers under that licence |
When Ofwat uses it | Default for discrete projects >£200m; Ofwat expects most major projects to use DPC | Used selectively for larger, more complex jobs where direct licensing’s benefits outweigh higher set-up/oversight costs |
Notable projects | None yet | Thames Tideway Tunnel |
How many projects are planned | Ofwat’s PR24 Major Projects portfolio totals 30 schemes (~£50bn whole-life). Of these, 27 have DPC as the proposed route | Within the same 30, 3 are proposed for SIPR |
Who pays for the project | The water company pays the winning contractor’s agreed annual charge and passes it on to customers through bills | A licensed project company operates like a mini-utility, and is paid directly under its licence from customer bills |
Construction risk | Equity investor & contractor based on shared bid; extremes may be socialised at regulators/politicians discretion | Contractor bears small overruns, equity bears routine risk, overruns excess of fixed threshold are shared with customers via the licence |
Counterparty risk | Paid by the commissioning water company — cashflows depend on its creditworthiness | Project company is paid under its licence — not reliant on a water company’s balance sheet |
Big civil projects are prone to cost blowouts. Oxford’s sewage-works upgrade, for example, is reported by the Financial Times to have jumped from about £40m to roughly £435m over just four years. Under DPC, much of that risk sits with investors. Under a SIPR, part of a shock can be capped and shared across contractors and customers.
Then the counterparty risk. Under DPC, payments to the project depend on the water company’s ability to pay. In a sector where both Thames and Southern Water bonds have been pushed into junk territory, that risk is material, independent of Ofwat’s investment grade requirement. Under a SIPR which is a mini utility, revenues are collected from the end-customers directly and do not rely on the balance-sheet of a counterparty.
Risk allocation changes the price of capital—and, ultimately, customer bills. For Thames Water’s upcoming Beckton Water Recycling plant, Vallorii’s Price of Risk (VAPRI) model suggests that compared to a DPC, SIPR trims the required Cost of Equity by ~220-320 bps (roughly 9% vs 12%) by shrinking two risk premia. Under DPC, investors add ~250 bps premium for the risk of construction overruns, and ~150 bps for the risk that Thames Water defaults on its payments. Under SIPR, the licence can cap and share extreme overruns—cutting ~100 bps from that construction premium—and revenues are licence-backed, effectively eliminating the ~150 bps counterparty premium.
Figure 2: SIPR CoE analysis

Financing Britain’s water needs is a complex trade-off, not a slogan. DPC harnesses competition but leaves more of the tail risks with equity; SIPR can steady cashflows and cap extreme overruns, but adds regulatory heft and the risk of socialising too much. Ofwat’s task is to share only the rare, costly shocks—and make that sharing explicit—while resisting blanket guarantees that bloat bills. VAPRI helps turn these choices into numbers, showing where a little underwriting meaningfully trims the cost of capital and where it merely shifts costs. Get the balance right and long-term money will fund reservoirs and recycling plants at a fair price; get it wrong and the country will pay a premium.
