Raw Green: The 380 Exajoule Opportunity
In short
The grid runs at half capacity while investors chase new power
Rushad Nanavatty of RMI and Third Derivative told Raw Green's hosts that the biggest energy gains lie in heating, cooling and a half-used grid, which need more investor attention.
"If you think of the grid as a machine, it is a machine that is only operating at something like 40 or 50% of its capacity," Rushad Nanavatty said. Count generation, transmission and distribution together, and most power systems use about half of what they have at any given time. No factory owner would accept that from a piece of equipment. Yet new generation, sold on the load growth from data centres and AI, is where the attention goes. Rushad, managing director at RMI and Third Derivative, called much of that spending incredibly wasteful, and knew the view was contrarian.
The alternatives already exist. Appliances in homes, offices and factories can act as grid assets, grid-enhancing technologies get more out of existing lines, and storage avoids the need for new plants. Venture investors, in Rushad's view, pay too much attention to generation and too little to the demand side.
Francesco De Lieto and Emma Mee had built their live Raw Green episode around a bigger number. The world consumes about 600 exajoules of primary energy a year, and roughly 380 of them never become useful energy. Rushad recalled a line from RMI's co-founder Amory Lovins, that people want hot showers and cold beer and nobody wants lumps of coal. Energy is lost at every step in between. Burning fuel is the first loss, and renewables remove most of it. The other half of the job is efficiency, where the world is far behind. Renewable capacity has grown about 15% a year, close to the pace needed to triple it by 2030. The goal of doubling the rate of efficiency gains is lagging very significantly.
If you think of the grid as a machine, it is a machine that is only operating at something like 40 or 50% of its capacity.
Heating and cooling sit at the centre of it, taking between a half and two-thirds of all useful energy. Asked where to put money, Rushad chose industrial heat. About half of industrial emissions come from process heat, and about half of that heat is at low or medium temperatures that mature heat pumps can supply. The obstacle is the spark gap between gas and electricity prices, and the war in the Middle East has made it more favourable. Cooling came next. Air-conditioning demand from the Global South is likely to dwarf the load growth from data centres. Solid-state cooling could be three to five times more efficient than today's vapour-compression machines.
The difficulty is that efficiency does not look like venture. Its returns are very predictable, money in the bank, Rushad said, but it rarely offers the single deal that returns a whole fund. One of the hosts said most efficiency investments return 20% IRR or more, yet in a fundraise a pitch about smart sockets lacks the pull of one about nuclear fusion.
Francesco listed technologies that are proven and patented, some of them decades old, that still wait for wide adoption. Passive radiative cooling is one. It is a paint or film that reflects sunlight and radiates heat through the atmosphere into deep space. Developers aim it at data centres and livestock feedlots, which will pay, while the people who need it most live in poor, hot countries. When Rushad's team pitched a big global philanthropy on bringing it to them, the funder turned it down as not mature. The mature alternative is ordinary white paint, which works far less well.
Emma asked about bankability, calling it Francesco's favourite word. It comes up most around first-of-a-kind plants costing $100 million or more, which need infrastructure money. Waiting for venture investors to change their criteria would take a very long time, Rushad said. Projects instead have to be de-risked until they pass a financier's due-diligence checklist, which can run to 150 items. Rushad named four kinds of de-risking, technical, commercial, social and capital, and said many portfolio companies manage two, often three, and very rarely all four.
Start-ups spend their early years optimising a technology, while a project financier wants predictability. Venture LPs want excitement, Rushad said, and infrastructure investors want something really boring. The work is getting from one to the other, turning promising technology into "really, really boring but bankable infrastructure".
Recorded live at The Drop 2026 for Raw Green, hosted by Francesco De Lieto and Emma Mee, with Rushad Nanavatty.