An operational experiment in an apartment in the Romanian city of Bucharest showed that semitransparent photovoltaic windows can supply a water heater directly with direct current, using the water tank as thermal storage instead of an electrical battery. According to measurements published by the experiment’s owner, the system covered between 43.7% and 46.7% of the energy required for hot water by a household of four people during two consecutive monitoring periods.
The experiment, presented by Adrian Băisan of PhotoVoltaic Windows SRL, targets a practical problem facing apartment residents: the lack of private roof space for installing solar panels. In such cases, windows, balconies, and glass barriers become some of the few surfaces available for generating energy, but electricity production does not always coincide with its use inside the home.
What happened in practice in the apartment?
The system was installed on a south-facing glazed balcony and consists of 11 semitransparent photovoltaic windows. Its corrected nominal capacity is approximately 845 watts peak, after earlier project documents had indicated 683 watts peak based on an older data sheet dating from 2021. The manufacturer explained that the glass delivered belonged to newer, higher-capacity models. This correction does not change the measured energy results or the hot-water coverage percentage.
The windows operate outside the household electrical grid, as they are connected directly to the resistive heating element in a conventional water heater. When solar power is insufficient, the controller switches the heater to alternating grid current after sunset until it reaches the temperature set on the thermostat. With this configuration, the system does not require a solar inverter or an electrochemical battery.
During the period from 2023 to 2024, the windows delivered 446 kilowatt-hours to the water heater, covering 46.7% of the home’s hot-water energy demand. During the period from 2024 to 2025, the delivered energy rose to 462 kilowatt-hours, but the coverage rate reached 43.7%. Each period includes approximately 11 actual months of use because the household was away for about one month.
Why was water heating chosen instead of a battery?
The basic idea is to separate the time when energy is produced from the time when it is used. A water tank can receive 100, 300, or 500 watts over several hours and then retain the energy as heat for later use. This differs from a conventional solar system connected to instantaneous electrical loads, where the system might produce 500 watts while the home is consuming only 150 watts at that moment.
The material cites a reference example from the German Environment Agency, UBA, for an 800-watt balcony photovoltaic system installed vertically and facing south. Its annual output is estimated at approximately 532 kilowatt-hours, with direct consumption inside the apartment of about 240 kilowatt-hours, or roughly 45%, in the absence of a battery. Directing solar electricity to the water heater instead allows production to be absorbed over a longer period, potentially increasing the use of available energy without adding electrical storage.
Electrical matching is the decisive factor
The results do not mean that photovoltaic glass is inherently more efficient than conventional silicon panels. The more important point is the match between the array voltage and the heating load. When two conventional panels with a combined capacity of approximately 800 watts are connected directly to a 3-kilowatt heating element operating at 230 volts, the comparison presented in the material indicates that the transferred power may not exceed approximately 300 watts because of the voltage and current characteristics.
By contrast, the photovoltaic windows use a larger number of lower-power generating elements, allowing them to be arranged to reach a direct-current voltage better suited to the heating element while keeping the total capacity moderate. The system recorded a measured peak of 499 watts at the heater from a corrected nominal capacity of 845 watts peak.
An active MPPT converter may increase instantaneous power under some lighting conditions, but it will not necessarily convert that increase into an equivalent amount of useful annual energy; the heater stops when the set temperature is reached, and the converter also adds consumption, losses, cost, and complexity. This conclusion remains linked to the experiment’s configuration and is not a general rule eliminating the need for power electronics in every system.
Additional benefits and limitations to consider
The material states that producing hot water locally reduced the waiting time for centrally supplied hot water to reach the shower. Based on a measured flow rate of 6.7 liters per minute and a wait of approximately five minutes to reach about 43 degrees Celsius in late September, the case model estimated that approximately 62.1 cubic meters of water would be wasted annually while waiting by a household of four people. The source emphasizes that this is a case-specific estimate rather than a general figure, because pipe length, circulation systems, and operating conditions vary between buildings.
The photovoltaic glass also functions as a shading element. Of a total glass area of approximately 12.9 square meters, the semitransparent photovoltaic glass covers 7.92 square meters. The model estimates that replacing clear glass with it reduced solar heat gain through the façade by approximately 40%, with an estimated reduction in cooling consumption in this apartment of between 433 and 481 kilowatt-hours in summer, with a central estimate of approximately 457 kilowatt-hours. These are calculated figures, not actual measurements of air-conditioning savings.
Editorial perspective from certi.news: The importance of the experiment lies in redefining the use of solar energy in apartments: it is not always necessary to convert electricity into alternating current or store it in a battery; the existing water tank may be a flexible and useful load. However, the results alone do not establish economic viability or scalability, and the data come from an entity associated with the technology itself, while some of the cooling and water benefits are based on models and estimates. Therefore, comparisons with conventional panels, installation costs, and building safety and regulatory requirements require independent evaluation before the solution can be broadly adopted.