Making Clean Tech Bankable: Industrial Heat

RMI's modelling puts the payback on a renewable heat project at about 15 years. The industrial companies expected to host them are used to paying off infrastructure in two or three. One participant at the industrial-heat Ripple, who works in thermal storage, did the arithmetic. A 15-year payback is an internal rate of return of 2.5 or 3%, and with 5% on offer from US Treasury bills, nobody needs to take any risk at all. To be financed, paybacks have to come down to somewhere between two and seven years.

That gap sat under everything George Harris of RMI and Daniel Galis of the Renewable Thermal Collaborative staged. Their format was a role play. Daniel played a developer whose thermal battery turns cheap solar and wind power into on-demand process heat, the hot water, oil and steam behind almost everything physical, most of it now made with natural gas. The developer's first project, a five-gigawatt-hour battery, was funded entirely by venture capital, and in character Daniel said venture is no long-term option for assets this large. George played a sceptical bank.

Whose balance sheet

A participant asked why the industrial companies should not pay themselves. George said heat competes inside them for capital with productivity projects and AI. Daniel, in role, said a food and beverage company wants to make food and drink. It does not want to trade electricity or learn what grid-balancing markets are, so it would rather someone else owned the battery and sold it the heat. That moves the loan from a multinational's balance sheet to a small developer's.

As the bank, George liked heat-as-a-service contracts for their recurring revenue and their physical collateral. The worry was resale, and whether anyone would buy a bespoke heat battery if a project failed. Modular units that can be taken apart and sold again could be underwritten more cheaply. And the bank was happy with the credit of a large food company, far less so with a small developer's.

technical economic analysis eats the story for breakfast

Story against spreadsheet

A late-stage infrastructure investor in a suit said this was a communication problem more than a bankability problem. A pitch that begins with bespoke designs and two years of engineering loses them. A unit that comes off a truck, has been built a hundred times and is supplied to creditworthy customers is different. "It's not their credit I want. I'll bank the system," they said. Founders should stop using the word first at all, because the deal has to feel risk-free. The capital between venture and the big infrastructure funds is not missing, only scarce, and it goes to theses a non-engineer can follow in two minutes that do not hang on a policy change. The ten giant infrastructure funds will not touch the sector yet. Build 100 units, and they will buy the company.

An energy-services provider disagreed about what comes first. In their view "technical economic analysis eats the story for breakfast", and if the price per megawatt-hour does not beat what a customer pays now, the discussion is over. Many of their customers are choosing an electric boiler next to the gas boiler, with low capital cost and no technology risk. To win, heat has to cost no more than gas plus CO₂. The thermal-storage participant said storage can do that, but only where electricity and network charges are low. They had seen storage come in well below gas, and the EU's heat auction makes Europe a better market than the US for now.

Revenue that may not last

Daniel then switched sides and played the bank. Most heat-battery revenue now comes from balancing and capacity markets, which are shallow and may be crowded out by grid batteries within two years. A single customer can scale down the one site a project serves. And most projects are 50 to 100 megawatt-hours, too big for venture capital and too small for many banks and infrastructure funds.

Contracts do not fit either. Projects sold so far price heat at a 5% to 10% discount to gas, which leaves the exposure to gas prices with the equity. Heat as a service needs contracts of 10 to 15 years, while industrial companies sign gas contracts for two or three at most. An investor in biomass gasification said customers had signed offtakes of 15 to 20 years whenever gas prices spiked, three years ago and again recently.

Other sectors offered tools. In the early days of solar, UniCredit offered leasing repaid from the cash flow an asset generates. US residential solar contracts were pooled into single financings to spread customer risk, and one participant said debt should start at 20% or 30% of a project, not 90%. In European battery storage the same lenders now appear in every deal, and a participant warned that start-ups had spent too long perfecting the terms of a first facility instead of building something that works and refinancing it.

One of the hosts closed on something the role-play bank had never offered. Some banks, the host said, are accepting losses on their first couple of transactions so that a bankable pipeline can follow.

This Ripple was hosted by George Harris (RMI) and Daniel Galis (Renewable Thermal Collaborative) at The Drop 2026 on 16 September.