Heat Reuse & District Heating (Sustainability & Economics)

A gigawatt AI campus is, thermodynamically, a gigawatt heater. Essentially all of the electricity it draws — minus a rounding error that leaves as light down fiber — is converted to low-grade heat and then spent, at additional energy cost, being thrown away. Chapter 5.9 treats the engineering of capturing that heat; this chapter treats the decision that sits on top of it: is the waste stream a cost you minimize, or a product you sell? The answer is not free to choose at commissioning. It is set far upstream — by the temperature of the loop you plumb, by whether there is a heat offtaker within trenching distance of the site you picked, and by the regulatory regime of the country you build in. Get those three right and waste heat becomes a modest revenue line and a powerful social-license and compliance asset. Get them wrong and you have a thermodynamically embarrassing building that pays twice — once to make the heat, once to reject it — in a jurisdiction that is increasingly going to fine you for the privilege.

This is a sustainability-and-economics chapter, with the engineering deliberately left to its home in Chapter 5.9. We cover the heat-grade economics that govern everything, the district-heating integration models (and why Stockholm's tradable-contract approach is the one worth copying), the payback math and the EU regulatory drivers that are converting reuse from a nice-to-have into a mandate, and finally designing for reuse as a constraint: what it costs to build the option in versus bolt it on. Heat reuse is cheap if you design for it and expensive if you retrofit it, and the gap between those two worlds is decided before steel is cut.

The economics of waste-heat reuse collapse to one number: the temperature of the heat you can deliver. Heat is not fungible. A megawatt of 30 °C heat and a megawatt of 60 °C heat are radically different products, because a district-heating network or an industrial process needs its input at a usable supply temperature — typically 60–70 °C for a conventional European 3rd-generation network, lower for modern 4th-generation low-temperature grids. If your heat arrives below that, someone has to spend electricity in a heat pump to lift it, and that lift is the entire economics of the deal.

This is where the cooling decision in Chapter 5.4 reaches downstream into the sustainability ledger. An air-cooled or chilled-water hall delivers a return stream around 30–45 °C — too cold to use directly, so a heat-pump lift to 60–75 °C is mandatory, and the heat pump's electricity and capex eat most of the margin. A warm-water direct-to-chip loop, by contrast, can return facility water at 45–60 °C or higher; the ASHRAE TC 9.9 push toward a 30 °C-and-above coolant baseline exists precisely to widen free cooling and to make the return grade high enough that heat reuse needs little or no lift. The fork is stark: design your thermal plant cold and you have committed your future heat-reuse business to a permanent heat-pump tax; design it warm and you can hand an offtaker a stream they can use almost directly. The chip vendors are pushing inlet temperatures up and DLC is becoming the 2026 default — which, almost incidentally, is the single most important thing that has ever happened to data-center heat reuse, because it raises the grade of the byproduct.

Heat reuse only works where there is a sink — somewhere for the heat to go, all year, at a price. That is why the technology's center of gravity is northern Europe: cold climates create year-round heating demand, and decades of public investment built dense district-heating (DH) networks that can absorb it. Stockholm's DH network, begun in the 1950s, now spans roughly 3,000 km; that pre-existing pipe in the ground is the asset, and the data center is merely a new heat source plugged into it. In a heat-sink-poor market — much of the US Sun Belt, the Gulf, Australia — there is simply nowhere for the heat to go regardless of how good your grade is, which is why US reuse remains marginal despite far larger compute. The siting consequence is direct: heat reuse is a reason to value a Nordic or German or Dutch metro site, and it is nearly worthless in a desert. (The water-and-climate siting gate is in Chapter 3.7; this is its mirror image on the heat-rejection side.)

The decisive innovation is not the heat exchanger; it is the commercial structure. The naive model is bilateral: one data center negotiates a bespoke deal to dump heat into one utility, splitting capex and arguing over price annually. It does not scale and it does not bankably finance. Stockholm Exergi's Open District Heating rewrote this by making heat a standardized, tradable product: any heat producer can sell into the network under a published, temperature-indexed tariff — the price you receive rises with the grade you deliver — settled like a spot market rather than negotiated one-off. That turns waste heat from a charity donation into a commodity with a price signal, and it is why the model now connects 30-plus data centers across 16-plus providers. The strategic lesson for an operator anywhere is to insist on a tradable, indexed structure rather than a bilateral give-away, because the indexed contract is what makes the warm-loop investment pay back and what lets the offtake be financed.

The table sorts by who you are selling to and why. If a real DH market exists, the tradable-indexed model dominates because it is the only one that scales and finances. If you have exactly one offtaker — the Meta Odense facility feeds Fjernvarme Fyn under a bilateral structure, delivering on the order of 100,000 MWh/yr into the Odense grid — a long-term bilateral PPA is the pragmatic answer. If you are in Germany after mid-2026 you may have no choice: the EnEfG can compel you to offer the heat at cost price, which is not a business at all but a regulatory cost of doing business. And if you sited in a market with no network, your only reuse path is a captive on-site user (greenhouses, aquaculture, an adjacent industrial process, office heating) — which is why the heat-reuse question is, at bottom, a siting question.

Stripped of enthusiasm, the heat-reuse business case is modest and front-loaded with capex. The revenue is the heat price times the energy delivered, and DH heat is cheap energy — you are not selling electricity, you are selling lukewarm water. Against that you carry the cost of the heat-exchanger plant, the interconnection to the network (trenching and pipe is the silent budget-killer; running a hot-water main several kilometers to reach a network can dwarf the in-building cost), and, if your loop is cold, the heat pump and its lifetime electricity. The honest framing: at a warm loop, next to an existing network, reuse pays back in a handful of years and is genuinely accretive; at a cold loop, far from a sink, the payback stretches past the asset's useful life and the project only exists because a regulator or an ESG commitment requires it. That bimodal outcome — cheap-and-good vs expensive-and-mandatory — is the whole decision.

Which is exactly why Europe stopped leaving it to economics. The EU Energy Efficiency Directive, via Delegated Regulation (EU) 2024/1364, now requires data centers at or above 500 kW IT load to report ERF — the energy-reuse factor, reused energy divided by total facility energy — alongside PUE, WUE, and REF, into a public EU database. Reporting is not a mandate to reuse, but it is the instrument that makes the absence of reuse visible and comparable, and visibility is the precursor to standards. Germany's EnEfG went further and made it binding: data centers commissioned after 1 July 2026 must achieve an ERF of at least 10%, rising to 15% (post-July 2027) and 20% (post-July 2028), with the heat offered to a nearby network at cost price unless narrow exemptions apply. The metric stack behind all of this — PUE/WUE/ERF/REF and the standards that define them — is built out in Chapter 15.1; the broader disclosure regime in Chapter 15.7. The point here is the consequence: in a growing slice of the world, heat reuse is no longer scored on payback. It is a condition of being allowed to operate.

The recurring discipline of this guide applies cleanly here: some heat-reuse decisions are reversible and cheap to defer, and some are irreversible and must be hedged at scoping. Irreversible (decide once): the site relative to a heat sink (you cannot move the campus to the network); the facility-water loop temperature architecture (a hall plumbed for a cold chilled-water loop cannot cheaply become a warm-water hall mid-life); and the physical provisions for a future connection — pipe-rack space, knockouts, a heat-exchanger room, and reserved easement for the trench to the network. Reversible (defer): whether you actually energize the connection now or later; the specific offtaker contract terms; and the heat-pump sizing, which can follow demand. The strategic move is the same one the rest of the lifecycle teaches — convert an irreversible decision into a reversible one cheaply by provisioning the substrate. Reserving a heat-exchanger room and a pipe easement costs little at design; adding them to a finished, occupied, slab-poured hall is a major retrofit.

The emerging frame that ties this together is heat-as-a-product. The most sophisticated operators no longer treat waste heat as an externality to dispose of; they treat it as a co-product to be sold, and they design the thermal plant — loop temperature, capture topology, metering — so that the heat leaving the building is merchantable. That reframing is what justifies the warm loop, the indexed contract, and the reserved easement. It is also what turns a regulatory burden (the EnEfG ERF mandate) into a hedge: an operator who designed for heat-as-a-product is compliant by construction, while a competitor who optimized purely for PUE has to retrofit. In a world where the binding constraint is power and every megawatt is scrutinized, being able to show that a fraction of your rejected heat warms a city is not just an ESG line — it is social license, and increasingly, license to build at all. The grid-and-energy-systems framing of that social-license argument continues in Chapter 15.8, and the embodied-carbon side of the sustainability ledger in Chapter 15.6.