Sunday, August 04, 2019

Suburban 5-4-and-a-Door Solar: Two Years of Data

We have had solar on our house for just over two years now. As of the end of July we have 24 months of data collected every fifteen minutes, including power produced by the panels, power consumed by the house, power exported to the grid, and power imported from the grid. This is all collected in a massive set of Excel workbooks that only number fanatics would find interesting. Here's a summary for the rest of you.

This is the one my wife uses the most as it's the simplest one to present. How have your power bills changed since installing solar? The orange are years with solar, the blue are years before solar. The problem with this graph is that it doesn't account for higher temperatures, changes in usage, etc. It's just summing up the power bills. For example, September of 2018 was the hottest and most expensive electricity month we've ever had - with or without solar - which led to the increase in the 2019 bar. Still, there was definitely an improvement.

I like this graph better. For each calendar month, the data from the array was used to calculate the costs with and without solar. The blue lines are what the cost would have been with solar. The orange lines are what the bill would be without solar. The green line is the savings per month. Here you can see that horrendous September bill, but you can also see the savings that month.

The high bills in the summer of 2018 led me to re-evaluate our power plan and, based on results from simulations using the data from the array, I switched to the PEV rate plan in January 2019. The impact can be seen in the summer 2019 (June/July) numbers, and I believe that the overall savings should improve throughout August and September.

The question that I get asked about the most is return on investment (ROI). When will the savings form the array equal the amount put into it. This graph is where I am tracking those numbers. I had hoped for (and still hope for) about a 10 year return. Right now, it's on track for a 15 year return.

These numbers assume about a 2% growth in power rates per year going forward, which is probably conservative based on the latest information from Georgia Power. These numbers also do not take into account any improvement in savings from the recent move to PEV rates. However, for our region with the current rate plans (e.g. buying power at a higher rate than we sell it back), 15 years is where I am forecasting at the moment. Check back next year to see if that has changed.

I like these two graphs as they show very clearly the impact of changes in power savings, solar, and billing. The top graph is how much power (kWh) we used from the grid for a 12-month period. (For example, the July 2019 number is for power used from August 2018 to July 2019.) The bottom graph is how much we paid to the power company for those 12 months.

We were already doing things to save power back in 2015 (installing LED lights, using smarter cooling and heating) before we installed the solar array, which is why power use was dropping even then. The big drop in June 2017 was when solar was installed. We reached equilibrium a year later (12 months of solar numbers), then the big September hit and there was a jump. We hope to see a corresponding jump back to normal after this September when the bad numbers wash out of the rolling 12 month calculations.

What might be less obvious but is more interesting is the tail end of the graphs when we moved to PEV rates. Our use in May/June/July stayed high, but the amount spent went down sharply. That's because we're paying less for the power that we're using. You can see it better in the following graph from July, which shows power usage in each of the time categories and the cost of the power used in each category..
In July we used about 40% as much "On Peak" energy (kWh) with solar than we would have without solar. However, that 40% in power reduction was responsible for 70% of the cost savings for the month. Reducing power usage in the on peak hours with the PEV plan has a much greater impact than reducing overall power usage on a flat rate plan.

Another way to see the impact of PEV rating is to look at the raw $/kWh for each month. In this graph, the red dots are from before installing solar and the green dots are with solar, but on the flat rate plan. The blue dots are from months with the PEV rate in place. There was not a huge impact in the winter when power use is down. However, there is a clear impact in the summer, with June and July $/kWh significantly lower than on the flat rate.

These final two graphs show power produced and consumed (solar + grid) for every day of those two years of data. We scaled the array to provide about 30%-50% of our use, which is why the peaks in the graphs don't line up.

Mean solar production is 17.8kWh per day with a median of 19.6kWh and a maximum of 33.3kWh. Mean power use is 46.5kWh per day with a median of 48.6kWh and a maximum way up at 137.8kWh. Those 100kWh+ days are rare and we hope to avoid them going forward by moving home office locations, sending a child to college, etc.

We have a lot more information that is fun to plow through on occasion, but what's more exciting is watching the savings improve as we change the way we use power based on information. Most people don't really know more than the amount of their bill. Information is power (pun intended).

Notes for those who have read this far
  1. Power bill calculations and representations do not include any rebates as those would skew the $/kW, $/month, etc. results and would, after all, apply to both with and without solar numbers.
  2. Calculated costs per month don't always line up with actual bills since those bills don't always start and end with the calendar month. All fees, fuel costs, etc. are calculated into the calculations just as they are in the actual bills.
  3. Fee calculations are based on the formulas provided by Georgia Power and have been checked to be accurate within a few cents each month. Calculations of the dollars credited for power sent back to the grid are based on actual $/kWh paid each month as the formulas used by Georgia Power to calculate that rate are impossible to recreate.
  4. The PEV rate charges considerably less for power used overnight and slightly higher for power used during the day. In the summer, the PEV rate charges significantly more for power used between 2pm and 7pm on weekdays. Those happen to be the hours of highest solar production, so the PEV rate can save quite a lot for solar users.

No comments: