FROM PASSIVE CUSTOMERS TO VALUABLE ASSETS

Demand Response Field Tests in District Heating

The earliest known DR field test dates back to 1982 in Stockholm, Sweden, where heat delivery to eight residential and office buildings was remotely reduced to secure supply for customers furthest from the heating plant. DR events were designed to fit the building time constants and allowed a maximum indoor temperature drop of 3 °C, with the longest event lasting 24 hours. [2]

In 2002, Kärkkäinen et al. conducted DR tests in Finland and Germany. Finnish buildings with water-based heating showed a 25–30% peak cut and 20–25% heat load reduction during 2–3-hour events. In Germany, an air-heated office building achieved a 4.1% peak reduction. The study highlighted that air-based systems respond faster than radiator systems, requiring tailored DR strategies. [3]

Kensby et al. (2010/11) confirmed similar findings in Gothenburg, Sweden, where five multi-story residential buildings with radiator heating were tested. The results emphasized the importance of HVAC characteristics in DR planning. [4]

Sweetnam et al. (2015/16) tested DR in 28 English homes without utility involvement, aiming to reduce peak demand by 15% through demand-shaping technology. While peak demand dropped, overall demand increased by 3%. Occupant surveys showed willingness to participate in DR for small financial incentives. [5]

In Denmark, Christensen et al. (2013–2015) tested DR in 72 single-family homes using night-setback strategies. Energy savings averaged 7%, with significant variation. Notably, residents continued using DR strategies beyond the initial heating season, indicating long-term acceptance. [6,7]

Mishra et al. (2017/18) demonstrated 11 DR strategies in a refurbished campus building in Finland. Events were controlled via the building management system (BMS), adjusting inlet water temperature based on price signals. While occupant satisfaction remained stable, some rooms became non-compliant during DR events, suggesting future strategies should be decentralised and room-specific. [8]

Ala-Kotila et al. (2018) tested DR in eight student-occupied residential blocks in Finland. The goal was to reduce power peaks without affecting tap water or indoor temperature. DR events reduced peak loads by 14–15% and total energy demand by 11%, resulting in €26,000 in cost savings . These buildings already use heat optimisation services, implying greater potential in traditional buildings. [9]

Christensen et al. (2018/19) conducted a DR test in a low-energy multi-storey building in Copenhagen with underfloor heating and mechanical ventilation. The strategy aimed to shift morning peak loads using price signals sent to apartment controllers. Peak load was reduced by 85% compared to similar apartments. However, residents objected to DR in bathrooms, highlighting the need for user input in DR design. [10]

Hagejärd et al. (2019) tested DR in eight multi-residential buildings in Malmö, Sweden. Five control schemes reduced heat supply by 25–100% for 0.5–3 hours, and one increased supply by 25% for 1 hour. The study focused on occupant satisfaction and found that successful DR must consider building insulation, ventilation, user temperature perception, individual control, and communication of load shifts. [11]

Christensen et al. (2021) tested DR in three low-energy houses in Aalborg, Denmark. The strategy involved shifting morning space heating using local radiator valve control. Resident feedback showed that without clear explanations and perceived benefits, DR strategies were not accepted. In homes with night-setback or morning ventilation, pre-heating benefits were lost, and dissatisfaction increased due to lower indoor temperatures. [12]

Van Oevelen et al. introduced the “STORM controller,” a DR strategy leveraging building thermal capacity to meet system objectives. Three control strategies were tested: peak load reduction, load shifting based on electricity prices, and enhancing thermal exchange in heating/cooling grids. In Sweden’s Rottne DH network (2018/19), peak heat production was reduced by 3.1%, and 96% of the heat load was shifted via building charging. In Heerlen, Netherlands, the controller increased system capacity by 37–49% and reduced peaks by 7.5–34%. [13]

Further testing in Brescia, Italy (2021/22), targeted daily peak load reduction at a mixing substation serving 35 homes. By adjusting supply water temperature and using building inertia, peak loads were reduced by 30–40%, depending on outdoor temperature. [14]

Guelpa et al. conducted a DR test in Turin, Italy, over three days with 10°C outdoor temperatures. Only 32 of 104 buildings had shiftable loads, yet peak reductions exceeded 5% daily, with DR events limited to 20 minutes to maintain indoor comfort. [15].

These demonstrations show that DR can be initiated by various stakeholders (i.e., utilities, technology providers, developers, and researchers), in different DH generations and building typologies