Combining energy harvesting sources for increased system availability and using an input switching algorithm to maximise battery lifetime

Sammanfattning: Energy harvesting is commonly used for powering low power electronics, for example Low Power Wide Area Networks (LPWAN). Many energy sources are natural (e.g. solar or wind), thus the harvested power can depend on weather, time of day, and time of year. By combining different energy harvesting techniques the power delivery becomes more reliable, as the different harvesters complement each other. This increases the availability of the wireless system. The thesis investigates which energy harvesting techniques (solar, wind, or rain) should be combined in a multiple-input energy harvesting system to give the most reliable power delivery. An input switching algorithm is developed for the energy harvesting system to select input source depending on available power. By not charging nor discharging the battery fully, the algorithm extends the battery lifetime. The evaluation assesses if such a multiple-input system with the algorithm can eliminate the need of an energy storage element. A simulation was constructed in Simulink to test combinations of the different harvesting sources. Weather data from 2008 to 2021 provided by the Swedish Meteorological and Hydrological Institute (SMHI) was used to accurately simulate Swedish weather. All combinations of solar, wind and rain harvesting were tested, resulting in seven combinations. A hardware system using solar and wind harvesting along with the input switching algorithm was constructed. The simulation results show that solar cells and wind turbines have the highest power density, while the density of rain harvesting is very low. When comparing the simulation with the constructed hardware system, the harvested solar power is similar while the harvested wind power is much smaller in the real system. Most times the real wind power is so low that it is measured as zero, while the simulated wind power is not. The few times the real wind turbine is generating power, the simulated wind power is often several times greater. A combination of solar and wind power results in an availability of 99%, and due to the high power-density of solar and wind harvesters, the system size is kept small. The input-switching algorithm decreases the amount of time the battery is completely discharged, but a small energy storage is still needed for sufficient availability. Compared to using a single harvester, a smaller energy storage can be used when combining harvesting sources. Using the algorithm, the battery state of charge is limited to a maximum of 80% which extends the battery lifetime.