Efficiency of Domestic Hot Water Heaters through Real-Time Price-Driven Model Predictive Control – a Reality Check

Domestic hot water heaters (DHWHs) provide hot water for daily activities such as bathing, cleaning, and cooking and therefore considered essential appliances in residential and commercial buildings. They consist of a storage tank that heats water using electricity, gas, or solar energy and maintains it at a desired temperature until needed. However, DHWHs are considered inefficient with significant energy consumption due to fixed heating schedules and continuous standby power use which makes them prime candidates for energy optimization specially for smart grids and renewable energy integration. Although simulations and lab-based studies have demonstrated the potential of model predictive control (MPC), real-world applications often face discrepancies due to unpredictable user behavior and system inefficiencies. To overcome these challenges, in a new study published in Frontiers in Energy Research Journal and conducted by Prof. (FH) Dr. Peter Kepplinger, Gerhard Huber and Prof. (FH) Dr. Markus Preißinger from the illwerke vkw Endowed Professorship for Energy Efficiency, Energy Research Centre at Vorarlberg University of Applied Sciences in Austria, the researchers performed a long-term field test involving 16 DHWHs in Austrian homes to evaluate the performance of a real-time price-driven MPC. Their goal was to determine how effectively MPC could optimize hot water heaters’ operation in practical settings compared to traditional control methods like hysteresis-based or night-only switching modes. Each heater was retrofitted with hardware that enabled data acquisition, including temperature sensors, power meters, and real-time electricity price inputs, all controlled through a Raspberry Pi system. The MPC system adjusted heater operation based on these real-time price signals, predicted user demand, and estimated system states, with the aim to minimize energy consumption and cost while maintaining user comfort.

The authors found that the MPC system outperformed traditional control methods specially in homes with moderate water usage which resulted in significant cost savings and higher thermal efficiency because it shifted heating to times when electricity prices were lower. In contrast, traditional hysteresis control and night-only switching modes either maintained fixed schedules or reacted to immediate demand which often led to inefficient energy use. They also reported that the night-only mode had lower electricity costs but frequently caused cold tap events during high daytime usage whereas hysteresis control failed to adapt to fluctuating energy prices. Therefore, the MPC’s dynamic approach in their studies allowed for greater flexibility in heating times, effectively balancing cost and user comfort. The authors also ran simulations to compare the real-world field test results with idealized scenarios. They designed the simulations to test the performance of the MPC under conditions where user demand and system states were known perfectly to identify any errors in demand prediction and system state estimation. The team’s findings showed that field test resulted in 30% cost savings compared to hysteresis control but with simulations with perfect knowledge of the system state and demand prediction demonstrated that savings could reach up to 50%. According to the authors, that means, although MPC has strong potential, its effectiveness is constrained by inaccuracies in demand prediction and state estimation. The researchers observed the largest improvements was when the system’s internal state was better understood which suggests that further development in state estimation algorithms could yield even greater energy savings. It is noteworthy to mention that the authors also found that the degree of utilization or how heavily a heater is used had an influence on the MPC system effectiveness with users that had low to moderate hot water demand, the MPC consistently delivered better results in terms of both cost and energy efficiency. On the other hand, homes with high water demand, the advantages of MPC diminished because the system was forced to heat water continuously to meet usage which left little flexibility in optimizing based on electricity prices. This observation indicated to the authors that MPC is particularly effective for homes where water use fluctuates rather than being constant throughout the day.

In conclusion, Dr. Peter Kepplinger and colleagues successfully demonstrated a retrofittable MPC system in a field test, which now give confidence for a clear path for the broader application of the technology in residential energy management. Moreover, their findings showcased how MPC has the potential to significantly reduce electricity costs. Especially when demand fluctuates, this can result in both environmental and economic benefits. This will be of high importance to energy providers, policymakers, and technology developers who want to implement more responsive and adaptable energy systems in line with the growing reliance on renewable energy sources. We believe that smart energy systems especially in residential settings can greatly benefit from integrating MPC to optimize energy consumption especially DHWHs which have excellent flexibility for load shifting.