STORM · Project

Smart Controller That Cuts District Heating Costs Using Self-Learning AI

energyPilotedTRL 7

Imagine your city's heating network is like a massive plumbing system delivering hot water to thousands of buildings — but nobody's really coordinating when to pump more or less. STORM built a smart controller that learns on its own how the network behaves and automatically balances supply and demand, shaves off expensive peak loads, and squeezes more value from waste heat and renewables. Think of it like a thermostat for an entire city district, except it teaches itself instead of needing engineers to program every rule. They proved it works on two real networks — a cutting-edge low-temperature grid in the Netherlands and a conventional heating grid in Sweden.

By the numbers

Real operational grids used for demonstration (Netherlands and Sweden)

Control strategies included (peak shaving, market interaction, cell balancing)

Consortium partners across 4 countries

Industry partners in the consortium

Total project deliverables produced

The business problem

What needed solving

District heating and cooling networks waste energy because they lack smart coordination between producers, the grid, and consumers. Peak demand spikes drive up costs, waste heat and renewables go underused, and traditional controllers require expensive custom engineering for every network configuration. Operators need a plug-and-play solution that learns and adapts on its own.

The solution

What was built

A self-learning district heating and cooling network controller that works as an add-on to existing SCADA systems, with 3 built-in control strategies (peak shaving, market interaction, cell balancing). The team also delivered a controller compatibility report covering communication protocols and system integration, and developed business models and replication plans for cross-country deployment.

Audience

Who needs this

District heating and cooling network operators looking to reduce peak costsEnergy utilities integrating more waste heat and renewables into existing gridsEnergy service companies (ESCOs) managing multi-building heating portfoliosMunicipal energy planners expanding or modernizing district energy systemsSCADA and building management system vendors seeking smart control add-ons

Business applications

Who can put this to work

District heating and cooling utilities

enterprise

Target: District heating network operators

If you are a district heating operator dealing with rising fuel costs and unpredictable demand peaks — this project developed a self-learning controller that plugs into your existing SCADA system as an add-on. It was demonstrated on 2 real grids across the Netherlands and Sweden, applying 3 control strategies including peak shaving and market interaction to increase use of waste heat and renewables without replacing your current infrastructure.

Building energy management

mid-size

Target: Energy service companies (ESCOs) managing large building portfolios

If you are an energy service company struggling to optimize heating and cooling across a portfolio of commercial buildings — STORM's controller uses self-learning techniques that adapt to different network configurations without custom engineering. The add-on design means it works with many existing network controllers, reducing integration costs while enabling smarter energy balancing across your assets.

Renewable energy integration

any

Target: Waste heat and renewable energy producers feeding into district networks

If you are a renewable energy or waste heat producer having difficulty getting your supply absorbed by district networks — STORM's cell balancing strategy coordinates producers, transporters, and consumers to maximize uptake of your energy. The project also developed business models to distribute the additional value created among all market players in the network.

Frequently asked

Quick answers

What would it cost to deploy this controller on our network?

The project data does not include specific pricing or licensing costs. However, the controller was designed as an add-on to existing DHC network controllers and SCADA systems, which suggests lower integration costs compared to full system replacements. Contact the coordinator for commercial terms.

Can this work at industrial scale on real networks?

Yes — the controller was demonstrated on 2 existing operational grids: a low-temperature DHC network in the Netherlands and a medium-temperature district heating grid in Sweden. These were real networks, not lab simulations, confirming industrial-scale viability.

What is the IP situation — can we license this technology?

The project was funded under an RIA (Research and Innovation Action) scheme, coordinated by VITO (Vlaamse Instelling voor Technologisch Onderzoek) in Belgium. IP ownership and licensing terms would need to be discussed directly with the consortium. Based on available project data, replication plans for other countries were developed as part of the project.

Will this work with our existing control systems?

The controller was explicitly designed as an add-on to many existing DHC network controllers and SCADA systems. The self-learning approach means it adapts to different configurations and generations of DHC networks without requiring custom model-based programming for each installation.

How quickly could we see results after deployment?

Based on available project data, the self-learning approach eliminates lengthy model calibration phases typical of model-based controllers. The project ran demonstrations over a 4-year period (2015-2019) and produced a controller compatibility report covering communication protocols and system integration requirements.

Does this comply with different national energy regulations?

The project specifically developed a replication plan addressing different market organizations and legal conditions across European countries beyond the Netherlands and Sweden demonstration sites. The consortium included partners from 4 countries (BE, FR, NL, SE) to ensure broad applicability.

Consortium

Who built it

The STORM consortium is well-balanced for commercialization with 7 partners across 4 countries (Belgium, France, Netherlands, Sweden). With 4 industry partners and 4 SMEs making up 57% of the consortium, this is clearly an industry-driven project rather than a pure academic exercise. The coordinator VITO is a major Flemish technology research organization in Belgium, providing credibility and long-term institutional backing. The mix of countries also mirrors key European district heating markets, particularly Sweden and the Netherlands where the demonstrations took place.

VLAAMSE INSTELLING VOOR TECHNOLOGISCH ONDERZOEK N.V.Coordinator · BE
STICHTING ZUYD HOGESCHOOLparticipant · NL
EUROHEAT & POWERparticipant · BE
SIGMA ORIONIS SAparticipant · FR
NODAIS ABparticipant · SE
MIJNWATER WARMTE INFRA BVparticipant · NL

How to reach the team

VITO (Vlaamse Instelling voor Technologisch Onderzoek), Belgium — a major Flemish research and technology organization. Search for STORM DHC project lead at VITO for direct contact.

Order a report on this consortium See VLAAMSE INSTELLING VOOR TECHNOLOGISCH ONDERZOEK N.V. profile →

Next steps

Talk to the team behind this work.

Want to know if this self-learning controller fits your district heating network? SciTransfer can connect you directly with the STORM team and provide a tailored feasibility brief.

Order a report on this topic More in Energy & Climate