Scientists from Germany and Tanzania have found a way to significantly reduce the energy consumption of refrigeration equipment by optimizing its load management. This result was achieved during a study conducted at a residential facility in Kenya with the use of smart devices and simulation of a hybrid energy system combining solar panels, batteries and a power grid connection.
Today, about 20% of global electricity is used for cooling technologies, from household and commercial refrigerators to freezers. In developing countries, especially in hot climates, the number of refrigerators could double by 2030, which means that authorities need to develop effective demand response measures in advance.
To solve this problem, the scientists from Germany’s Paderborn University and Tanzania’s Nelson Mandela African Institution of Science and Technology have conducted a study at a residential facility in the Kenyan settlement of Lwak. The facility, which consists of three buildings connected to the power grid, was equipped with a 40 kW backup diesel generator, which provided power in the event of power outages.
As is known, a key feature of refrigeration equipment is thermal inertia: after being turned off, refrigerators can maintain the required temperature for several more hours. This allows the user to temporarily turn off the refrigerator or shift its operating cycles without risking the quality of food storage. It was this feature that formed the basis of the proposed load management method in Kenya.
Eight refrigerators and freezers were installed at the facility, each equipped with an AM2302 temperature sensor that was connected to a Raspberry Pi 3 mini-computer for continuous monitoring. Power management and data collection were conducted with Edimax SP-2101W smart sockets, which made it possible to switch devices on and off remotely, as well as to monitor their energy consumption. This approach allowed the scientists to accurately assess equipment operation and set the optimal operating mode for each device. In order to calculate these modes, they used the particle swarm optimization algorithm, a method that simulates the behavior of a group in which many particles (possible solutions) simultaneously explore different scenarios and focus on both their own successful attempts and the best solution overall. The algorithm adjusted strategies on a stepwise basis until it found the most efficient load distribution over time.
Since the temperature inside the refrigerators remained within safe limits for 3 to 7 hours after the switch-off, an individual operating cycle was set for each device based on its type and thermal inertia level.
The results showed that a noticeable change could be achieved even by controlling only eight devices: the energy consumption of the refrigerators fell by 18%, total daily consumption dropped by 8%, and the evening peak load decreased by 23% and shifted to a later time. As a result, annual electricity costs went down by about $300.
A hybrid power system was also simulated: it included solar panels with a capacity of 12.45 kW, a 37.6 kWh battery and SMA inverters. The implementation of control measures made it possible to increase the share of energy coming from solar panels and also reduce the consumption of grid electricity by 25%.
The researchers are planning to apply the newly developed algorithms in hotels, stores and restaurants, where refrigeration equipment is used especially intensively, and to supplement the control systems with water heaters, pumps and air conditioners.
