Harnessing waste heat for community heating is not a new concept. Exhaust heat networks that capture heat from factories, power plants, and wastewater treatment plants have existed in the past and the waste heat generated has previously been used to provide hot water and warm buildings. However, with the advent of disruptive digital technologies, data centres have started gaining more prominence and tech companies have realised that the heat they generate in their data centres can be utilised through district heating networks for multiple purposes like water heating and farming.
This blog explores how data centres can supply waste heat for community heating. In doing so, it explains how tech companies (through their data centres) contribute to waste heat collection, the regulations that are impacting date centre site selection, the use of disruptive technologies like AI and IoT for heat management across data centres, and some of the obstacles that data centre operators face.
Want to read about community heating in detail? Simply click here: Community Heating: A slow growing market with 85 million devices by 2034 fueled by climate resilience and energy-saving requirements.
There are currently 8,000 data centres globally, with 33% located in the US, 16% in Europe, and 10% in China. These data centres use a lot of electricity and generate a huge amount of heat, since their power-hungry servers work around the clock. Since the use of disruptive digital solutions like artificial intelligence (AI) is rapidly increasing, existing data centres will work continuously for data storage and to enhance efficiency. It is clear that more data centres will be required to store and manage the ever-growing demand for data that organisations rely on for their day-to-day operations.
As per the International Energy Agency (IEA) Electricity 2024 report, the electricity demand in data centres mainly arises from two processes: computing (accounting for 40% of electricity demand) and cooling requirements to achieve stable processing efficiency (taking up another 40%). The remaining 20% comes from other associated IT equipment. Data centre operators work around the clock to ensure that equipment does not get overheated, and some operators also focus on reusing the heat instead of wasting it. Dispelling the heat from data centres is not only bad for the circular economy but also contributes to climate change. Re-using it for community heating is an attractive alternative. Community heating simply involves generating heat in a centralised location and then distributing it across residences, businesses, and industries in a local area.
Global technology and internet service companies are trying to adhere to sustainability commitments and many of them are re-using the heat they generate to power district heating systems. Tech firms like Microsoft have collaborated with Fortum to reuse data centre heat for district heating in Finland. Meta has been redistributing surplus heat generated from their data centres in Denmark for local district heating solutions. At the Paris Olympics 2024, Equinix data centres played a supporting role in warming the Olympic swimming pool. A few data centres located in West London are being used to heat up to 10,000 homes, and similar schemes are relatively common in other parts of the world.
Governments around the world have established guidelines which mandate data centres to reuse the waste heat they generate. Some of the prominent examples in this context would be that of the European Union, which has implemented laws that mandate the need to capture and utilise excess heat generated by data centres and require feasibility studies to grant heat reuse in building permits. Some European countries such as Denmark have introduced laws requiring data centres to recover excess heat and have also imposed taxes on wasted heat. In Germany, the Energy Efficiency Act mandates that data centres must reuse a certain percentage of their waste heat, which increases over time. Across various locations in Europe, governments now require new data centres to conduct feasibility studies to use the excess heat generated to benefit local communities before granting building permits and power agreements.
Therefore, companies now prefer opening and operating their data centres only at suitable locations that are not too far from the users for exhaust heat energy. In fact, if there aren't enough users to consume all the heat being recycled, business organisations may not consider it as a viable site option.
The increasing usage of disruptive digital technologies like AI and IoT applications do contribute to a large amount of data generation and increase the requirement and need for data centres all the more. Yet, the use of these connectivity technologies also allows data centres to operate flexibly by responding to the actual heat demand of consumers. AI is used to enhance data centres’ operations as they analyse large amounts of data generated from IoT sensing devices and other sources across all domains and derive real-time insights for facility and IT optimisation opportunities for overall risk management and optimisation. AI is also useful in site selection for community heating projects as it helps identify sites with optimal energy resources and predict energy demand and supply capacity. Advanced models can help adjust cooling or power distribution automatically, based on real-time data centre operations, thereby preventing unnecessary energy wastage and risk.
Data centre operators usually face immense pressure from governments for site selection due to varying norms and regulations. However, there are a few obstacles that they face while building data centres. Firstly, it must be noted that when a data centre is initially built, it has minimal occupancy and heat exports could be low. This can create uncertainty in terms of heat generation for community heating stakeholders at least during inception. Additionally, community heating projects cannot assume that heat produced by a data centre will increase over time. Servers usually have a lifecycle of three to five years and technology improvements generally reduce the overall heat they produce. Another factor is the varying seasons, with demand for heating lower during the summer months, which is just when a data centre needs the most cooling. The off-season for heat demand should therefore be considered carefully, so that data centres can maintain viable cooling options in hot weather.
The demand for data centres is going to increase continuously and reusing exhaust heat is one way to build more efficient and sustainable infrastructure for data centres. Potential benefits of connecting data centres to heat networks include lower carbon footprint and reduced cost of heating for households. For data centre operators, connection to community heating systems can reduce the amount of electricity and water required for cooling IT infrastructure, and it might even become a source of revenue. However, successful deployment of these data centres requires choices to be made in terms of site planning and designing, as retrofitting could be difficult.