When I installed our solar array, I signed a Net Metering Contract with our electric utility Central Main Power. Like most Net Metering arrangements, the utility credits me for every kilowatt hour that I export to the grid. This means that if I am paying the utility $.13 per kilowatt hour for the electricity that they deliver, if my solar array is producing more power than we use and we are exporting energy I am getting a credit of $.13/kWh. At the end of the month credits are applied to my bill, and if I exported more than we imported a credit is applied to subsequent bills until the credit is used up. If we were to have a credit after 12 consecutive months, then it would be cleared. So there is no incentive at this point to build a larger solar array than we need.
Since we purchased the Chevy Volt in May 2012, we have not had a net export for any given month because the Volt requires up to 13 kWh for a full charge (about as much energy as our house uses daily). We do not always use up a full charge - on some days when we only run short errands we may only use a partial charge (meaning less than 40 miles range in the summer). So our monthly electric bills have averaged between $15 and $18 from April to September 2013. Here is a chart showing our total daily import versus export:
While most of the variation in our daily net energy use has to do with cloud cover affecting the solar arrays output, there are also periods where our use of the Volt has a significant impact on the net energy for the day.
Below is a chart showing net energy for a typical solar day recently where we were exporting power while the sun was shining, but then we charged the Volt in the evening which used around 3700 W for 3 1/2 hours. So on this particular day we nearly broke even. If this had been a clear, sunny summer day - we would have had some slight export even after fully charging the Volt.
The bottom line is that we are driving our electric vehicle nearly for free within its electric range on average while also providing all of our domestic needs via the solar array.
Recently I was talking with a friend of mine whom I had assumed was somewhat knowledgeable about solar energy. I was surprised when he made the assertion that “solar panels only produce power in direct sunlight”. I corrected him and informed him that even on heavily overcast days, my solar array can produce over 3 kWh. But on a typical sunny day like yesterday, September 20 you can see the power curve shows power being generated before and after sun strikes my solar array which is facing West at 270°. Sunrise on this day was at 6:25 AM and it set at 6:38 PM, and my array was producing power for every minute of this time.
Total amount of power produced on this day was 22.9 kWH and the sun first hit the solar array at around 9:20 AM and Dropped behind the trees at around 5:30 PM (17:30). As you can see my solar array was producing hundreds of Watts before and after the sun was shining directly on the panels.
Compare this with the power curve from today which was heavily overcast with the sun barely making an appearance, yet the array produced 9.71kWH.
When I installed a tracking system for the two solar panels that I added to the south wall of my workshop back in June, I did not expect to notice them moving very often. Theoretically they would move incrementally every day to continuously adjust for seasonal solar elevation changes. However, what happens is that whenever clouds pass over the sun on a partially cloudy day the panels tilt down toward a brighter part of the sky in the south. Today I went out with my camera and took time lapse images every 10 seconds for around six minutes. The wind was gusting from 15 to 20 mph, and small dense clouds were moving across the sky fairly quickly. The GIF animation below condenses 6 minutes into 6 seconds and you can clearly see the panels tracking even the slightest cloud cover.
I cannot be certain, but I believe this is squeezing a few more Watts out of those 2 240 W panels that are fed through Enphase M215 inverters that can produce a peak of around 220 W when facing directly into the sun on an ideal day.
I have been keeping meticulous records since I installed our solar array in 2009, and have been keeping all this information in a spreadsheet. One of the things I have been doing is comparing the estimated power I should expect with the actual power generated. The National Renewable Energy Lamps tool called PVWatts allows you to input the specifics of your solar array including its location, number of panels, and orientation etc. so that it can give you an estimate of how much power you should expect. At the end of every utility billing cycle, I access the records from the web interface for the solar inverters on my system to determine exactly how much power was produced by the array for each month. The chart below shows the calculated estimate vs. actual production. Notice that the estimated production bumps up each year as I have added panels to the array. The scale on the left represents kilowatt hours per month.
Recently I discovered an unexpected benefit to the tracking system I designed for the 2 - 230 W solar panels I added to the south wall of my building. Since my office is on the second floor I notice whenever the linear motors activate to move the panels because I can hear the motors running. I had not expected to notice them moving since the whole idea is simply to adjust the tilt angle seasonally. But one day a storm blew in from the north and a big dark cloud came in over the building leaving the southern sky bright and sunny. My tracking system automatically tilted the solar panels down to point directly at the brighter part of the sky. And since then I have noticed the panels adjusting slightly whenever heavy cloud moves over. I had not expected to be squeezing every tiny watt out of these panels, and it is kind of fun to watch them adjusting to condition. in the photo below, you can see the tracking sensor in the red plastic dome. Inside there are two small light sensors on each side of a horizontal piece of metal painted flat black. When that metal shades one of the sensors it sends a signal to drive the motors to move it out of shade ensuring that the panel is always directly pointed at the brightest source of light.
Since we purchased the Chevy Volt, I have installed five more solar panels on my workshop building to achieve a total array specification of around 5.8 kW peak capacity. On a good sunny day our solar array can generate over 30 kWh, while the Volt requires a little over 13 kWh for a full charge. In theory that should leave us with a significant surplus in the summer and allow us to drive the Volt for free on those days. In reality we have yet to see a zero or negative monthly electric bill, partly because it has been an unusually overcast year to date (climate change?).
I had hoped that the additional panels would help to offset the charging power that the Volt uses via the 240 V charging station that draws 3700 Watts for 3.5 hours to fully charge the vehicle. So recently I decided to go through all of my solar power data and the records I keep of weekly charging power for the Volt and calculate what our electric bill would be if we did not have an electric vehicle. Here is a chart that I created showing our calculated monthly energy bills:
The calculated average monthly bill would be $10.57. This was my original goal for the solar power system - to provide almost all of the power for our home. So essentially I have accomplished that, now I have to think about where to put additional panels to offset the power needed to charge the Volt! I have been eyeing the east facing roof of the house, but my concern is that it is so high that I would be uncomfortable doing the installation with the help of neighbors and friends. I really dislike paying other people to do things I should be able to do myself. So for now this is a stalemate.
When I first mounted 2 additional panels to the south wall of my workshop it was March and I mounted them at a steep angle. Later in the season as the sun rose higher in the sky I extended the brackets to capture more energy in the summer months by tilting them further out from the wall. I realized that an optimal solution would be to adjust them frequently to optimize performance. So I decided to build my own tracking system that would adjust the tilt angle automatically. I started by designing a sun sensor that uses two photocells inside a little green plastic dome:
A flat black piece of metal separates them and when the sun shadows one of the cells it sends a signal to the control box that I designed that moves the linear actuators that I purchased. This image shows the control box mounted to the wall inside my window with the sensor mounted to the side of one of the panels so that it moves with the panel:
And here is a closer picture showing the control box that I built:
Here is my schematic if you are interested in making one for yourself.
I can set the slide switch to the left position which puts the system into automatic, or in the right (manual) position I can press the green button to tilt the panels out away from the wall, or the red button to tilt them back in. This may come in handy in the winter when I need to shed snow off the solar panels, or “furl” them in the case of high winds. The stroke of the linear actuators is not quite long enough to go all the way from winter to summer angles, but they have built in limit switches that shut off the motors at their maximum extensions. Since we are close to the summer solstice right now the actuators are fully extended, and over the next several months I will see the system adjust as the sun angle lowers.
I have been fooling around with this great web calculator produced by PVeducation.org that produces charts of available solar energy (insolation). So I have combined this with my recent interest in GIF animation to produce the chart below that shows the available solar energy in Watts/square meter in 10 day increments for the year at my location of 44° latitude North.
I made this animation by taking screenshots at ten-day intervals by adjusting the slider on the calculators webpage.
This clearly demonstrates how the amount of available sun hours per day varies significantly at my latitude. Of course, at the equator there would be very little variation at all, and at the poles you would see almost no sunlight for half of the year.
When I recently added 2 solar panels to the south wall of my workshop, it was early spring and I felt that a steep angle would be appropriate to catch the low winter sun so I mounted them at 70°. Recently, when I looked at my performance statistics on my Enlighten web interface I noticed that I was no longer seeing a peak power of over 220 Watts on each panel. So I did some research to determine how much more power I would get if I changed the panel to a 45° tilt angle. Since I live at about 44° latitude, this should make for an ideal average angle. I found a web calculator that allows me to see the relative per module power for each tilt angle, see below:
I have animated the two tilt angle curves so you can see the comparison. This made it clear that I would gain a significant amount of power in the summer by changing the tilt to 45°, so I cut two longer pieces of angle aluminum and have moved one panel up so that I can see the difference:
Here is the proof that tilting to the the panel up to 45 degrees improves performance, this shows peak Watts at solar noon:
Now to go up there an tilt the second panel!
After installing the two additional solar panels on the south wall of my workshop, I have been keeping an eye on the Enphase Enlighten web interface for my solar array to see how well they are performing. I quickly noticed that in the morning between 8:30 and 10:30 one of the panels was getting a lot less power than the other. When I walked around to eyeball things I realized there was a poplar tree that was shading one panel during those hours and compromising the power output dramatically:
The tree shown in the image below is leaning at a bad angle and poplar trees often are weak and will collapse under heavy snow loads or winds.
So I felt it was time to remove that tree before it came down and hit a power line or the building. I am mentoring a high school student named Ben on Friday afternoons, and used this opportunity to teach him how to use a chainsaw safely. Here he is finishing up cutting the trunk into firewood length pieces using my electric chainsaw. We split and stacked the wood together and got enough to heat my workshop for a few days.
The following morning I checked my Enlighten webpage to confirm that I am now getting more power on the one solar panel:
The thing about solar panels is that they often run into shading issues as trees grow. I recently had to move a solar panel on the house since it was easier to move the panel than cut the tree in question.
There is a tool available from the NREL (National Renewable Energy Labs) called PVWatts that allows you to estimate the energy that should be produced by a given sized solar array within the United States. It uses meteorological data for your location to accurately estimate the actual number of sun hours per day/week/month etc. along with specifics of your solar array such as tilt angle and azimuth etc. Here is a chart that I have been keeping showing estimated power vs. actual power generated for the exact same time periods. I get the power generated by my array from reports from my Enphase Envoy web monitoring system. The shape of the calculated curve changes year-to-year as I have been adding solar panels and adjusting the estimate based on the additional wattage available.
As you can see, it appears that actual energy produced is varying more and more from the calculated estimate. I am not sure whether this is natural variability, or if Climate Change is having an impact. As I talk to locals who have lived here in Maine for decades they are all noticing that the lakes are icing out earlier in the spring and freezing much later in the fall. But people don’t really notice a substantial change in cloud cover year-to-year because that is not remembered easily. It is clouds that are the enemy of solar power, and of course their absence is a bonus as can be seen in the fall of 2012.
When I decided to add two more solar panels to the south wall of my workshop, I began by adding the panels to the 3-D model I had already built using Google’s free SketchUp CAD program. I was able to explore the optimal location and tilt angle of these panels before I installed them since this would save me an enormous amount of trouble climbing around on my building trying to eyeball dimensions and locations. I did not bother to model the line of trees along the east side (on the right in the images) of the building, because they would have a relatively minimal impact on these two additional panels in the summer mornings. Since some of those trees are on my property, they may fall victim to my chainsaw if they are shading my panels too much.
I began my solar power installation in 2009 by installing 21 panels at a cost of $3.28 per watt. Over the years I have added panels to the point that I now have 31 panels installed for a total of about 5.8kW. Solar panel prices have dropped consistently since I installed my first panels and the most recent panels cost me $1.08/Watt. This means that if I had waited until now to install my system I would haveve saved many thousands of dollars and the break even return on investment would be much shorter. Depending on how one finances the system today one can see a breakeven point of 12 to 18 months, and in states with very high electrical costs a much shorter period than that. In solar parlance, this is referred to as “grid parity”, meaning that solar now is as affordable as utility power in many cases.
Here’s a list and chart of all of my solar panel purchases correlating $/Watt:
Click here to see live statistics from my solar array and learn more about the installation.
Since I have run out of room on the west facing roof slope of my workshop building, the only place I have left is the south facing wall above my solar collectors. The ideal angle to mount these panels would be around 40°, however I choose to go for around 25° for several reasons. Wind loading is a concern, As is shading the collectors below, but also these new 245W panels are somewhat overrated for the 215W rated Enphase micro-inverters, so a less than optimal angle will be more than sufficient.
There are no off-the-shelf solar power module racks made when you want to mount a panel at an angle, so I had to make my own from aluminum angle that I painted with black hammertone paint.
Here is a shot of the brackets mounted to the wall before we put the panels up.
I planned the installation very carefully so that all the brackets were in place and all I needed to do was borrow my good friend John and his ladder so that we could both haul each panel up and simply drop a bolt into the mounting holes at the top to secure it. Here is John putting a bolt in the bottom strut:
Thanks to my careful planning, both the panels went up in less than a half hour with only a few minor glitches where some bolt holes did not line up quite right. Here is a picture of the two panels mounted - taken today. Notice that the solar panels shade the collectors slightly, but this is a very little concern since I will be draining them within a month. As the summer sun angle goes lower in the fall and winter the collectors will be more exposed.
I was very pleased to see the power output peg at 225 W for over three hours today (Click the image to see live details):
It seems that the Enphase M215 inverters are conservatively rated at 215 Watts. I reached under the panel and put my hand on the inverter to see if it was running warm, and it was not in our 40° ambient weather. The inverters are suspended under the top center of each panel where they get plenty of air circulation for cooling.
Yesterday, I drove down to Massachusetts to pick up to more solar panels from the altE warehouse. This was a 240 mile round-trip that I decided to take in the Chevy Volt after taking very careful measurements to ensure that the solar panels would fit. The panels measure 66″ x 37.5″ And I had to fold down the rear seats and rest the panels on top of the front headrests. I am so pleased with the cargo capacity of this vehicle! Since I was using “range extender” mode for the entire trip I got 40 miles per gallon average. I decided to drive down to pick them up, because shipping solar panels is crazy expensive since they must be crated and shipped by truck freight. Here’s a picture of one of the panels removed from the car when I got home:
I also picked up 2 Enphase M215 inverters along with the special 240 V cable with connectors:
I would like to get these panels installed as soon as possible, but we have more snowstorms on the way, so it will be a while. Stay tuned …
Since we purchased the Chevy Volt last year, it has been clear that I will need to add more solar panels to make up for the extra energy required to charge it (around 13 kWh per day).
Last year I added three panels to the west facing array which now totals 5.1 kW. This leaves no more space on that roof. The only remaining surface on the building that is exposed to the sun is the south wall above the collectors, so I intend to put two more panels on each side of the window angled out at about 25°. Here is a Google SketchUp model showing the proposed location for these two new panels:
In keeping with my desire to install only American-made hardware, I have purchased two Solar World 245W panels, and two Enphase M215 inverters from altE. Rather than pay the very expensive truck freight to have this hardware delivered, I will be driving down to Marlborough, MA (a 2.5 hour Drive) to pick them up from the altE warehouse. After measuring very carefully I have determined that the 2 solar panels should barely fit inside the Chevy Volt which is fitting since the panels will be providing power to charge it.
I hope to mount them as soon as the snow clears off the roof, and the weather is amenable for outdoor work on ladders.
The Weather Underground site (wunderground.com) has a tool you can use to evaluate the solar potential for your location. While it does not account for the specifics of your roof, or potential shading issues, it does give you a perspective on the solar energy available in a relative context.
Click the image or here to go to the calculator page. The seasonal variations become quite clear, if you live closer to the equator there will be much less variance of course. Here in the northern latitudes the variance requires that we plan the size of our solar installations thoughtfully. Solar thermal system need to be designed so they do not over produce in the summer - or a heat dump mechanism needs to be considered. And solar power systems can be designed to overproduce in the summer so that a grid-tied system can “bank” energy credits under a Net Metering contract with the utility, and use those credits in the winter.
When people learn that I am committed to living sustainably, and power my home from solar, they often jump to asking me about what is involved in putting solar panels on their home. My immediate response is “First get your house in order”. By this I mean that there’s no point in putting renewable energy systems on your home if your home is an energy waster. What I prescribe an energy audit. An audit will cost anywhere from $300 to $600 (some utility companies or state agencies will do a simple audit for free). An audit will reveal all of the energy waste issues in your home. An ideal audit will give you a list of things to do to reduce energy waste and improve the efficiency of your home starting with the least expensive. Most of the suggestions will yield a return on investment within a year or two. Typical initial suggestions involve plugging the leaks in the outer envelope of the building. This is done by caulking cracks around electric outlets and pipes or wires that enter the building. The next step might include adding insulation in the attic or basement ceiling.
My friend Paul Kando running a blower door test (image courtesy of the Midcoast Green Collaborative)
Auditors will typically do a blower door test which involves installing a large fan in one of the exterior doors of your home and pressurizing the building. By measuring the difference between the inside and outside pressures auditors can determine the equivalent leakage area of the envelope of your home. For instance the calculated leakage area of my home is around 124 square inches, or just under 1 square foot of total leakage. According to the audit this creates the equivalent of 0.33 air exchanges (ACH) per hour. This is considered optimal by most standards. Less leakage reduces the number of air exchanges and can result in unhealthy air quality. Most homes however, leak too much which means that heating and cooling energy is being wasted. This is the greatest area of concern for improving building efficiency.
So any readers of my blog who are thinking about putting some expensive solar panels on the roof would be well advised to invest in an audit first. This can potentially reduce the energy operating cost of your home which will also reduce the size of the renewable energy system you may wish to install in the future. It’s a win-win deal.
Below is a net energy chart from MyEragy.com recorded on July 3, that shows the Volt being charged after driving about 25 miles starting from a full charge - so about 7kWh. Energy below the zero line is being exported to the grid and “banked”, above the line is being imported from the grid. This shows that our solar array now can provide all the power needed to charge our Volt in the middle of the summer. (The MyEragy site gets it’s data from my TED5000 energy monitor and displays it on the web on a user accessible page).
Today I drove down to the Alt-E warehouse in Marlborough Mass (3 hours in my Ford Escape Hybrid @ 29MPG) to pick up 3 Solar World 230 Watt panels.
Got back around 5:00pm and scrambled up on ladders to add some length to the mounting rails for these bigger panels. Then I bolted down the 3 Enphase M190 inverters and plugged them in (turned power off FIRST!).
My neighbor John helped me haul them up on the roof and bolt them down to the rails. Here is the new array (total 5190Watts) at dusk.
Too bad it will rain hard for the next 2 days…