DecisionData.org presents information collected independently from official provider websites. We regularly update the site in an effort to keep this information up-to-date and accurate at all times. The offers that appear on this site are from companies from which DecisionData.org receives compensation.
Installing solar yourself isn't easy, especially if you intend to put in more than a basic backup. We strongly recommend letting a professional installer handle it, but here's the info you'll need to understand intimately before getting started.
- How large does my system need to be?
- What do I need to know to install it?
- Should I go fully off the grid?
DIY System Size
Systems are generally sized in kilowatts, or kW. A 5kW solar system, for example, is capable of producing 5 kilowatts under normal sunny conditions. It will produce more on the sunniest days and obviously much less when the weather is darker.
Step 1: Peak Power
You need to determine is whether you want to completely replace power (e.g. if you're in an “off-grid” location) or to reduce your power bill and have a backup for power outages. A fully off-grid system (including generation plus storage) will need to completely meet the peak power you need to use, where a power-saving system will only need to meet the typical load.
Calculating peak power is your first step regardless. You almost never want a system that generates more power than you could ever use. While you might be able to sell some power back to your electric company, it's generally not a high enough rate to warrant buying a larger system.
To calculate peak power, go here. Enter all the highest-power things you might use at any given time and add them up. Don't put everything in your house in unless everything will be on at the same time. You're trying to calculate your maximum usage, and chances are you're not running every appliance at the same time. Add at least 30% to make sure you compensate for system loss and additional power availability.
System Power: Basic
Take your power bill and divide it into the number of days in the month, then divide it into the average number of daylight hours. If you're using 1,000 kWh in a 30-day month, you'll need to produce an average of 33 kWh per day. Divide that by 6 if you're in the red area, 5.5 if you're in the orange area, 5 if you're in the yellow area, and 4 if you're in the light tan area.
The result is the size of the system you'd need to buy to cover your average use. You can go with a smaller system, and you probably should if you live in a state like Minnesota where power is cheap and solar potential is low.
That's a quick and dirty version, but you should really read below if you're serious about DIY solar.
System Power: More Advanced
For most people (not going off the grid) the next step is to decide what percentage of your peak power you want to cover. The goal is to determine the optimal return on your solar investment. This guide should at least give you a better idea of system size and how to determine savings. First, decide what time period you want to look at. People generally use either 20 or 30 years. For this example, we'll use 20 years.
Before getting into the math let's illustrate why we don't just run with peak power. Imagine your power consumption looked like this:
Maybe you use a lot of energy in the morning getting ready, then even more when you fire up the furnace or AC for your return home. Your peak usage is about 10 kW, but does that mean you should get a 10-kW system? Probably not. You'd probably save nearly as much money with a smaller system that cost far less up front.
Contact your electric company and see if you can get daily graphs like the one above for a typical day. They probably won't have it. You can then have to either map out your hourly usage or get a product like Neurio to help look at your daily usage.
Now plug in a few different system sizes and check how many kilowatt hours (kWh) are covered by the system and how many are not. Assume a 5-kW system can only generate 5 kW in an hour (e.g. 5 kWh) and multiple by the price you pay per kilowatt hour. A table for your state may be found here. Your calculations will look like this:
The formula for a 2-kW estimate in Hawaii ($.30 per kWh) is =IF(B2<=2,B2*0.3,2*0.3): meaning if the kWh used is less than the system capacity, multiply the kWh by the price of power per kWh. If the number used is greater than the system capacity, multiply by system capacity instead. Obviously, we're not going to count hours where the sun isn't strong enough to produce solar power. The point here is that even in Hawaii you can't save more than $.60 per hour with a 2-kW system.
Note that despite being capable of producing 250% more power (and likely costing about twice as much), the 5-kW system in this example only saves 57% more on electricity than the 2-kW system. The most cost-efficient system will be the one that barely covers your lowest usage. Try to think of solar as a long-term investment, though. Our goal is not to buy the most cost-effective solution: it's to save the most money over 20 years.
Once you have an average sunny day's savings, multiply by 365.25 (.25 for leap year) to see yearly savings. Those are your savings if you live and multiply by the percentage of sunny days where you live. If, for example, 85% of days are sunny enough to produce 2 kW you'll want to assume 15% lower savings. Actual sun can be as high as 90% of the year in a city like Yuma, AZ to as low as 40% in Anchorage. This table should give you some idea, but you may also have to Google your area.
To see 20-year return, multiply by 20 and subtract the system cost. Then assume an average of 10% efficiency lost over 20 years. Most solar systems will run at 80-85% efficiency at the end of 20 years, so we're just using a mean efficiency loss to make it more realistic.
Technically you need hourly calculations for sunlight vs. your usage in those same hours for every hour for an entire year. While we may someday build a calculator using NOAA data to do this. For now, this is a more precise method than most solar calculators will give you.
In our silly example, the calculations show that a 5-kW system pays for itself several times over. Even though the savings with a 5-kW system don't look too impressive vs. the 2-kW system, it makes up for itself over time.
This is largely due to the excessively-expensive $.30 per kWh in Hawaii. At a more-common $.12 per kWh the 2-kW system yield greater 20-year savings. Again, be aware that most solar panel calculators are highly optimistic. Nuanced calculations are harder, and the sites with calculators are generally trying to sell solar.
This guide is not intended to teach you how to wire or fully install your panels. This is just an overview so you can decide whether you actually want to do DIY solar.
Panels and Parts
To get started you'll need panels, some way to mount/attach the panels, an inverter, a grid tie, and a PV disconnect. If you want to store energy, which is recommended, you will also want a battery and a charge controller.
Most solar panels you'll find are either monocrystalline or polycrystalline silicon. Either will work fine for most people. Thin-film solar cells are still around. They're cheaper, less durable, less efficient, and require a lot more space. There are lots of technologies available, but don't get bogged down in the details. Look for the PV rating. A 250-watt-rated panel will produce the same amount of power under test conditions regardless of what it's made of.
Top pick: Renogy 6-piece set
Amazon Rating: 4.0 (as of 2/22/2017)
Average price per watt: $.33 (varies)
Check price on Amazon
Make sure that your roof can support the weight of the panels with mounting equipment. Most modern houses will be fine with the added weight, but the weight on a low-sloping roof with two layers of shingles plus snow adds up quickly.
You'll also want to make sure you can mount the panels facing true south (not magnetic south). Keep in mind that some HOAs prohibit things like solar panels on your roof. Do whatever you can to reduce shade. Even a little shade from a tree can have a massive impact on the system, sometimes wearing them down prematurely.
Hopefully you have a big south-facing roof with one set of shingles that can easily support the weight. Putting panels on the east or west side will dramatically reduce their efficiency. You could buy a system on a track that follows the sun. While these systems are more efficient, they probably won't last 20 years. In almost every case it's best to mount panels fixed facing south.
DIY solar installers often fail to consider the optimal angle of tilt. The sun hits the earth dead-on on the equator. If installing panels right on the equator a no-tilt situation might work fine. Since we're talking about the US, almost all panels will be more efficient facing south at a raised angle. Get the angle too high and the sun won't strike them in the early morning or evening. Put the angle too low and you'll have shadows on the bottom of each panel every day.
While some inefficiency is unavoidable due to the earth's shifting axis, make sure to determine your latitude. You can use this calculator or simply check on Google maps. With your latitude, you can determine an ideal fixed angle.
Many people do adjust the tilt of their systems manually each season or even each month. You'll get 20-30% more out of your system if you adjust regularly, especially if you live further north. Even in Las Vegas, for example, the ideal angle varies from 30 degrees in the winter to 78 degrees in the summer. If this appeals to you, make sure you get mounting hardware that makes angles easy to adjust.
Cutting the Electric Cord: Renewable (Solar, Wind, etc.) Only
Going truly off-grid isn't a good idea for most people.
First, if you don't intend to use the grid as a backup you have two less-than-ideal options. One, resign yourself to not having electricity sometimes. “Roughing it” might sound appealing to some, but will likely grow tiresome after several months. Option two, buy expensive energy storage. Lithium ion batteries such as the Tesla Powerwall are the most common choice as they are basically plug-and-use options, but a 14kW Powerwall costs $5,500, and one won't even be enough to get most people through a cloudy day or two. Other options (flywheels, water pumps, and weight-based storage) are similarly expensive and much harder to install.
Second, you'll need a much more expensive system if you intend to fully cut your house off the grid, and you may have to make some lifestyle changes. For example, a standard clothes dryer can take between 1k-4k watts to run. Families tend to use between 1 and 3 kW for standard living. That means running your clothes dryer during the day might be too much for a 5-kW system to handle. Running any two major appliances at the same time — something most of us don't even think about — may be impossible or require more (expensive) power storage.
Third, there are some electrical uses that you can just forget about. Basically, any sort of heating will push all but the most expensive systems over their limits. A standard high-efficiency furnace can use up to 18 kW, which is far too much for most off-grid systems. Water heaters use about 4kW, so you'll have to either heat water some old-fashioned way or deal with cold showers. Even AC systems use 3-5kW, so regular use on a 6kW system will mean either no extra power or no energy going into backup power. Also remember some appliances will wear out prematurely if run on too little power.
The easiest way to solve the heating issue is to get a non-electric (gas) heater, dryer, oven, etc. The practicality of gas will vary by location, but you'll need a gas connection to make life convenient.
Finally, keep in mind that most electric companies will charge you a flat fee if your home is still connected to their grid, and they're not going to remove the connection for free.
It's certainly possible to go completely “off grid,” but it's not convenient or cost effective for most people.
In any case, before deciding whether to install your own system make sure to find your city and learn more about solar in your area.