Energy Transition Today

The episode delivers a reasonable density of useful data points - rack power jumping from 5-10 kW to 100-150 kW for AI, Germany's 2026 legal mandate for waste heat offtakes, UK's 14,000 heat networks covering only 3% of heat demand - but these are spread across substantial filler, repetition, and generic framing about grid constraints and PPAs that any sector-follower already knows.

“just essentially the racks...even about five years ago their typical capacities was about 5 to 10 kilowatts. And since then that's now increased to between about 100 and 150 for AI driven facilities”

“in Germany it's now a legal requirement. So under the Energy Efficiency act which was recently passed, uh, from, from this year, from 2026, all the all new build data centers have to be able to secure an offtake for at least a proportion of their waste heat”

The claim that the UK has not a single operational data-centre-to-city-heat-network connection, and the contrast with Germany and Paris's legal requirements, is genuinely informative and underreported; however, the bulk of the episode - grid constraints, PPA additionality, AI driving demand - is standard industry commentary that circulates widely.

“as far as I'm aware we don't have a single data center that's providing waste heat to a city scale heat network”

“I don't think we're seeing a classic demand bubble. Uh, I think it's just that the market is restrained because the grids are constrained”

Lorcan McCalloden is a genuine practitioner - Development Director at Engie for city-scale district energy networks for 10 years and Project Director at One Energy where he personally negotiated waste heat terms with a live data centre - giving him directly relevant first-hand experience; he is not a C-suite hyperscaler executive or widely recognised authority, but he has clearly done the work.

“in my last role as development director, um, I was responsible for the growth and the expansion of some of our flagship cityscale district energy networks”

“I was able to identify uh, an energy center location close to an existing data center which was operational. And then I went on to negotiate um, terms so that we could secure a future waste heat supply”

The episode is well-stocked with named projects, companies, and figures: Pure/AVK's 110 MW off-grid Dublin facility, Deep Green's 5.6 MW Bradford approval, Hemico's £36M Green Heat Network Fund grant and potential £600M investment, underground pipe costs of £2,000-5,000/metre, and the sub-10 MVA threshold to avoid reinforcement studies - these are concrete and verifiable rather than hand-waved.

“Deep Green has just secured plan and approval for a 5.6 megawatt data center to be located just next door to the Bradford Energy Network”

“they secured 36 million of government grant funding in the form of the Green Heat Network Fund...could be up to a 600 million pound investment by Hemico”

The host identifies a genuine structural tension (chicken-and-egg between data centre location and heat network viability) and lands one sharp pushback on the 'couple of years' timeline claim, but most questions are leading or overly open, and several interesting assertions - like the additionality standard for PPAs or the NESO zombie-application reforms - are left without meaningful follow-up or challenge.

“That's certainly not possible. Come on, couple of years. You can't build a solar farm in a couple of years.”

“don't you run into the chicken and egg problem? Because they want uh, these data centers to be in less populous areas for obvious reasons”