The Regenerative Strategist

Welcome and thank you for reading The Regenerative Strategist. A weekly update on harnessing the power of decarbonization, digitization, and enhanced data collection, delivering buildings of the future powered by renewable energy & automation.

Want to stay informed?

Executive summary

Data centers are not just power-hungry computing hubs; they are waste-heat factories. Traditionally, tapping that heat required the facility to be adjacent to a district energy network. Now, emerging heat highways aim to move usable heat long distances, turning distance into just another engineering parameter. A new peer-reviewed model (Feb 2026) demonstrates a 100 MW data center can economically supply heat up to ~24 km away. Europe is already laying groundwork: Germany’s new energy-efficiency law (July 2026) mandates rising shares of “reused energy” from data centers, and the Linz District Heating Forum (Mar 17–18, 2026) is explicitly focused on heat highways.

This edition explores the full landscape: from the history of district heat reuse to the physics of thermal transport. We explain the heat-highway concept (multi-stage heat pumps, hot water pipelines, thermal storage) and the critical constraints of latency and losses (minutes-per-kilometer delays, storage sizing). We detail recent case studies (e.g. Stockholm 30 MW plant, Helsinki 2026 startup) and policy moves. Finally, we map stakeholder incentives and sketch best practices for cities, utilities, and tech operators. The takeaway: in an AI-driven world, “heat” must flow as freely as data. ⚡♨️

Introduction: From Co-Generation to Heat Highways

District heating (DH) is a well-known legacy. Cities for a century have bought heat from local power plants or waste incinerators. Data centers changed the picture: they generate vast quantities of clean heat, yet often remained heat-isolated. Until recently, only adjacent reuse was practical (e.g., a campus built next to a heating network).

Heat highways flip that narrative. With modern insulation and boosting technology, heat can travel far. Thermal transmission networks are being designed to carry warm water (50–70 °C) many kilometers, then reheat it at the end point. This changes waste heat from a byproduct into a potential urban energy source on par with fuel imports.

Why now? Policy and technology. The European Commission’s data center rating scheme (2024) and delegated acts define how to count “heat reused”. Crucially, EU law (Dir. EU 2024/1364) will require data centers to disclose heat export metrics, effectively tracking compliance. Germany, for example, will demand 10% of a new data center’s heat be reused in 2026, rising to 20% by 2028. In short: regulators are pushing data centers to become thermal suppliers, not just loads.

At the same time, companies face rising scrutiny of their carbon claims. Reusing waste heat is an easy way to burnish sustainability metrics if done credibly. Thus, all pieces are aligned: the era of heat highways is dawning.

1. How Heat Highways Work

Heat highways are engineered networks that move low-temperature heat long distances, using specialized equipment to overcome the losses and temperature drops. The February 2026 Energy study provides the first detailed blueprint:

• Low-Grade Transport Loop: Waste heat (~50–70 °C) is collected and pumped into an insulated pipeline. Thicker insulation and a larger diameter pipe reduce losses (typically 15–25% over 20 km). This loop can include intermediate buffer tanks.
• Multi-Stage Heat Pumps: Near the receiving city, a compression heat pump boosts the heat to around 90–120 °C suitable for the local district heating return. The model used two compression stages: from 50→80 °C, then 80→120 °C. The overall COP was ~3.5, meaning each unit of grid electricity yields 3.5 units of heat in the city.
• Thermal Storage: Both ends often have hot-water tanks (the study sized them at 5.4% of annual demand). These buffers absorb the mismatch between 24/7 data center output and fluctuating heat demand, mitigating latency.
• Controlled Flow: Advanced controls vary pump flow based on demand forecasts and grid conditions, ensuring timely delivery. Think of it as traffic control on a highway of heat.

The result: in the modeled scenario, even with the extra complexity, the economics were compelling. A 100 MW facility could pay back the entire infrastructure (pipes + pumps + storage) in about 2.6 years by selling its “free” heat. Put simply, a city could ditch a gas boiler or oil furnace by buying virtually costless heat, easily justifying the capital.

2. Thermal Latency and Loss Mitigation

Heat highways introduce thermal latency and losses that must be managed:

• Latency: Heat moves with fluid. In a typical district pipe, expect roughly 20–25 minutes of travel per km. Over 20 km, that’s up to 5–8 hours round-trip delay! In practice, smart operation and storage smooth this: supply tanks run continuously, and demand tanks deliver on schedule.
• Losses: Even with good insulation, heat leaks into the ground. Models show ~15–25% of heat can be lost over 20+ km. The solution is not to eliminate loss (impossible) but to account for it. By sending more volume at a lower grade, the absolute loss can be lower, and then nearly all remaining heat is upgraded at the endpoint.
• Boosting Near Demand: Crucially, these systems do not try to deliver 90 °C heat 20 km away. Instead, they deliver 50–70°C flow and boost it right at the hook-up. This avoids wasting pumping energy on high-temperature fluids in the long haul.
• Storage Sizing: To combat latency, the study uses tanks sized for minutes of demand. For example, a 100 MW source might need on the order of 30–60 m³ of hot water storage per MW of demand. That lets the city heat demand ride through the travel delay. (In the example model, about 5% of annual heat load was in storage.)
• Flow Control: Advanced planning is used. If a cold snap is forecast, the system pre-charges storage and runs pumps at peak power beforehand. During low demand, it slows. Essentially, the heat highway is run like a smart grid, with heat “traffic signals.”

Technical trade-offs:

This architecture implies several choices:

• Temperature vs Distance: Higher pipeline temp (80°C+) could reduce pumping volume but requires even more insulation. Lower temp (50°C) allows moderate insulation but necessitates more booster lift.
• Pump COP vs Insulation Cost: A better COP (more pump stages) can justify thinner insulation or longer pipes but adds capital expense and maintenance.
• Storage Size vs Latency: Larger tanks cut the delay effect (heat is always ready), but cost space and money. Each system optimizes differently based on site and need.
• Water vs Ethylene Glycol: Some designs use glycol blends to prevent freezing, impacting heat capacity and pump work.
• Overall system COP: The total COP (heat_out / electric_in) may run ~3–4 for a multi-stage pump. If renewables supply the power, net system emissions can be near zero even before accounting displaced fossil heat.

In sum, engineers now see heat highways as balancing a complex multidimensional trade space. But the Feb2026 study proves the sweet spot exists: design curves show a clear economic range around 10–25 km for 100 MW, given modern tech.

3. Case Studies – 2025–2026 Projects

Real-world projects are already testing heat highways:

• Stockholm (atNorth, 30 MW): AtNorth is building a 30 MW data center (online 2026) tied into Stockholm’s district heat. They use a reversible heat-pump loop: servers reject heat into water which is then sent to the utility. This plant can both supply heat to homes and use DH return to cool the data center, yielding a highly efficient, near-net-zero loop.
• Kajaani, Finland (Borealis & Loiste): Borealis Data Center will, by May 1, 2026, deliver its 4.8 MW waste heat into Kajaani’s DH. They laid a small pipeline (<500 m), using heat pumps to raise 55°C water to 90°C. This displaces ~5.5 GWh of oil heat per year. Notably, corporate owners partly funded the district network extension, illustrating public–private cost-sharing.
• Berlin, Germany (ENGIE): ENGIE has retrofitted the NTT DATA Center’s waste heat (formerly discarded) into Berlin’s network. Starting late 2026, up to 3.6 MW of heat will be lifted via compressors into city heat. Combined with 300 m³ of hot storage, this pilot is converting an existing site to comply with upcoming 2026 reuse rules.
• Oslo Region (Asp & Lyse): Asp Data Center (Fornebu) is partnering with Lyse to integrate cooling and heating networks. A unique twist: up to 3 MW of summer cooling will feed Lyse’s district cooling grid, whose absorption chillers then dump heat to the municipal grid. Exact status is unspecified but expected in 2026.
• Copenhagen, Denmark (atNorth, Ballerup): Another atNorth campus in the works will channel heat to local DH (8,000+ homes) starting ~2028. Using direct liquid cooling (higher outlet temp), they can push heat efficiently just 1 km to the nearest network. This shows smaller scale but fully integrated reuse.

These examples highlight enabling technologies in use (heat pumps, TES, controls) and underscore that heat reuse projects now span 3–30+ MW and multiple countries. They also hint at variety: some are short links (Berlin, Copenhagen), others plan for tens of km (the model implies possible expansion).

4. Urban Planning and Economics

Heat highways are as much urban policy as engineering. Key implications:

• Thermal Rights-of-Way: Cities may need to allocate land for pipelines, just like for subways or utilities. For example, an urban masterplan might set aside corridors in industrial zones or even street easements for future heat lines. This proactive planning avoids costly renegotiation later.
• Ownership & Tariffs: Who “owns” the heat network? Possibilities include the utility owning pipes (data center pays a tariff), or the data center building a private line and selling heat wholesale to the utility. Regulators will likely require fair pricing so that heat buyers (often households) are not overcharged for the new infrastructure. Transparent pricing models (e.g. fixed fee + volume charge) will emerge, akin to water or broadband utilities.
• Permitting and Siting: Permitting can delay projects by years if a new heat highway requires crossing roads, water bodies, or sensitive land. Permitting reforms (like fast-track for strategic decarbonization) could help. Additionally, cities could integrate “data center with heat export” into zoning — for instance, allowing higher capacity only if reuse is part of the plan.
• Environmental Justice: While heat highways cut regional emissions, local impacts must be managed. Building pipelines and plants can disrupt neighborhoods. Policies should ensure community benefits (reduced local air pollution, potential district cooling in summer) and avoid unfair costs on low-income areas. Public engagement, environmental reviews, and benefits-sharing agreements are prudent.
• Economic Incentives: Some governments may offer subsidies or loans for waste-heat projects, recognizing the public good. Already, the EU’s Green Deal investment programs include district heating modernization funds. Aligning tax credits or carbon pricing to favor heat reuse would accelerate adoption. Conversely, ending incentives for exclusive data center development (e.g. the Georgia example of cutting tax breaks for new servers) can balance scales and encourage reuse commitments.

Stakeholder map:

• Developers/Operators: Seek low-cost energy & PR. Benefit from selling “renewable heating” but risk cost of infrastructure and uncertain ROI.
• Utilities/Heat Companies: Gain cheap heat source, stabilizing load. Risk if data center departs or export stops.
• Municipalities: Win local jobs and lower emissions. Must handle civic infrastructure costs and manage community input.
• Consumers (Households): Get cleaner, possibly cheaper heat. Concerned about rates and project delays.
• Regulators/Investors: Want verifiable carbon reductions. Can drive standards (like EU regs) and vet projects carefully.

Building consensus around these incentives will be critical. The coming years (2026–2030) will test models: will we see industrial-scale thermal utilities, or only isolated pilots?

For heat highways to work at scale, robust governance is needed:

• Measurement and Reporting: EU delegated Reg 2024/1364 standardizes how to account for waste heat: it sets the metering point at the data center’s heat exchanger and defines how to declare “heat reused” versus losses. Strictly following these rules ensures no double-counting and builds trust. Outside the EU, analogous standards may emerge (or be imported via GHG protocols).
• Verification: Third-party audits or automated telemetry should confirm heat deliveries. Smart meters on the network and blockchain accounting are being discussed to certify each MWh of reclaimed heat as genuinely extra (not just shifting existing loads).
• Rebound and Leakage: One risk is that “cheaper heat” from data centers encourages more heating in buildings (rebound). Planners must account for this: for instance, tying building renovations or thermostat controls to any heat subsidy. Leakage in GH term also means accounting properly in local CO₂ inventories.
• Stranded Assets: If a city invests heavily in a heat pipeline only for the data center to switch off or be repurposed, there’s risk of underutilized infrastructure. Mitigation: design networks to be multi-user (take heat from industrial parks or waste incinerators too) and ensure contracts include long-term take-or-pay clauses by the heat supplier.
• Regulatory Lag: Tech evolves fast, but rulemaking lags. Observers note that regulation already struggles with “getting net-zero wrong.” Heat highways need nimble frameworks — e.g. pilot permits or rapid updates to codes (like NFPA fire codes adapting to liquid-cooled racks).

Conclusion: Toward a Thermal Internet

Heat highways represent a shift in urban energy thinking. They promise to turn one of AI’s byproducts — thermal waste — into a civic asset. The necessary technologies (heat pumps, pipes, controls) exist now; the next steps are systems design and governance. For policymakers and planners, the charge is clear: integrate thermal planning with digital infrastructure planning. For engineers, the challenge is making these networks reliable and cost-effective.

If successful, we may look back and say: just as the internet connected data, the thermal internet connected every degree of heat. In that world, data centers become power plants and heating plants — architecting a more sustainable, resilient cityscape. ⚡♨️🌆