The table below shows the house’s predicted and actual electrical usage. As part of the design, we measured appliance energy consumption using a Watts Up? Pro energy meter, used Energy Star specifications for the appliances we expected to install, and estimated lighting consumption based on the number of fixtures and our expected usage. We estimated consumption for the well pump, solar hot water circulation pump, and the ERV based on component specifications and national consumption averages.
After eliminating backup space and water heating, the remaining actual electrical load was close to our predictions. However, some specific estimates were off by quite a bit. We underestimated SHW circulation pump consumption by 183 kWh per year. We overestimated the well pump consumption, having based it on a water consumption of 120 gallons per day—our actual average is about 36 gallons per day. We also overestimated the electric clothes dryer’s consumption because we are using a “solar dryer” (outdoor clothesline) much more than anticipated. Our total electrical consumption, excluding the EV, averages 338 kWh per month. The EV has raised this average to 521 kWh per month, which is still much less than the average U.S. home’s 920 kWh per month.
The energy our EV would consume was the most critical, yet most difficult, amount to predict. It would be the household’s largest energy consumer. We did not know which vehicle we would purchase or what our driving patterns would be, and there was precious little manufacturer data on recharging consumption. Our prediction was based on published Chevy Volt charger specs and an assumption that the EV would require a full charge every night. Charging consumption from empty to full is higher for our Nissan Leaf compared to the Volt specifications, but it is consuming less than predicted, since full discharges are rare.
During its first 61 days with us, the Leaf consumed 366 kWh for an average of 6 kWh per day. We are eager to see how this will play out over the course of a year. Of course, with only two months of EV data under our belts, it is difficult at this point to make a long-term consumption prediction.
Solar Hot Water. Even though the system met all of our hot water demand, the Taco pump is undersized. The low flow rate causes the collector to overheat, which in turn shuts down the system daily, except during cloudy weather. The other sign that the pump is undersized is the 55°F temperature difference between the collector inlet and outlet. Ideally, there would be only a 20°F difference. The installed pump should be able to provide a 6 gallon per minute (gpm) flow rate (2 gpm per collector) with the 10-foot head from the drainback tank to the top of the collector, but only provides 2.5 gpm. I suspect that the 12 or so 90° elbows are introducing excessive pipe friction, which is creating the 15-foot effective head—and low flow.
However, provided that the system continues to meet 100% of our demand, and we do not need the SHW system for backup space heating, it’s unlikely we will replace the pump, since the pump that meets the 15-foot head requirement uses twice as much energy as the current pump.
Solar Electricity. For 10 days in October and five in January, we lost 10 modules’ worth of PV production (about 11.0 and 8.8 kWh per day, respectively). The fault was finally traced to a failed 240 VAC rooftop disconnect. After switching to an air-conditioner-type heavy-duty disconnect, the problem was eliminated.
There was another inverter-related problem we were aware of during the system’s design. At the time of installation, the Enphase M210 microinverter was not available, so we used the M190s that Enphase specified for our Sharp 224-watt modules. These inverters (rated at 190 watts) have a maximum output power of 199 watts, which means that, under ideal conditions, module output would be clipped at 199 watts rather than 224 watts. Our high elevation and low temperatures may be contributing to more clipping than expected—it happens three to four days per week.
Finally, a minor issue is snowfall that sticks to the modules longer than anticipated. Even with their 39° tilt, the modules don’t shed snow for at least two days if snowfall is greater than 2 inches—and they are too high to be swept. One advantage of microinverters, however, is that as the snow melts unevenly, individual modules start producing electricity rather than an entire string remaining shut down until all of the snow melts.
Space Heating. In hindsight, our 10 kBtu per hour Mitsubishi minisplit heat pump was probably overkill as backup—although staying warm was a key design area where we did not want to come up short. Any future backup heating needs (they were zero for this first year) could have been handled by a small 1,500 W to 2,000 W electric heating element in the supply trunk of the ERV, and cost much less.