ABSTRACT Grid-connected solar photovoltaic (PV) systems are gaining popularity as consumers look to embrace renewable energy technology and lower energy bills. Traditional grid-tied solar PV can meet electrical loads in a home when solar radiation levels enable adequate generation of electricity. The amount of solar generation available to a home varies directly with the amount of solar radiation at any point in time. This approach has helped many reduce electricity consumption, but the impact on demand reduction has not been significant. This is largely due to the fact that residential demand happens between 5:00 pm to midnight during weekdays; a time when solar PV generation is not always available.
We hereby propose our “Innovative Solar Demand Response” tool, which utilizes Green Button Data to size a Solar PV and a battery system based on average peak energy demand of a home during different hours of the day. The idea is that the battery would be charged by solar PV during daytime (when solar radiation is available) and a charge/discharge controller would be responsible for fluctuating battery system output to offset demand as it rises above the minimum average demand of the building throughout the year.
According to U.S. Department of Energy, an average home uses about 11,000 kWh annually. The house we chose for our test run, which is in Berkeley, CA, has an annual energy consumption of 7,644 kWh. The peak demand for this house occurred between 6:00 pm and midnight and has a maximum average hourly peak demand of 1.1 kW. We executed our software on the Green Button Data to find that this house would need a 1.5 kW solar PV system with two 12-V, 450 Ah batteries. This would allow the system to shift the maximum average hourly peak down to 0.4 kW consistently throughout the year. According to Go Solar California, approximately 114,450 solar installations have already occurred in California. If these existing systems performed as described in this case study, over 80,000 kW of peak demand could be reduced each day.
The proposed solution could save around one-third of a residential customer’s energy bill and has the potential to save even more as time-of-the-day pricing is more widely distributed. Future development could also integrate weather forecasts to allow more effective control of the battery power usage. Inverse building models allow electrical consumption to be accurately predicted at various outdoor air temperatures. Incorporating these features into a solar PV control system would achieve significant demand reductions and financial gains for a residential customer.