Figures - PV System
My Grid-Tied Solar Electric System: What, Why, How, and How Much
THIS IS THE STORY of my 2 kilowatt rooftop solar photovoltaic (PV) system ("solar panels"). It's the story of why I went solar, what things need to be considered, the cost, the rebates and tax credits, what it's like to own solar panels, the system's performance, and the answer to the bottom-line question: "looking back, would I do it again?" Measurements of my solar electricity production and household electricity use are shown, and readers may want to check out the informative essay about how electricity is measured called Power and Energy (Watts and kWh) in a Nutshell.
IT'S VERY IMPORTANT to say up front that solar panels are not just for people who want to save the Earth. They also make solid economic sense. I no longer pay for electricity (for life), and I'm protected from future increases in the price of electricity.
IT ALL STARTED ONE DAY in 2007 when I opened my Xcel Energy bill, and read in their newsletter for the umpteenth time about rebates for installing a grid-tied solar PV system on my roof. But this time, for some unknown reason, it struck me that this is exactly what I've been waiting for, and if not me, who, and if not now, when? So I read everything on the Xcel Energy SolarRewards website, did some online research, thought about the economics of it, and met with a solar installer. It's now five years later, and I love my PV system. It occurred to me that if it took so long for someone who's already on board with renewable energy and energy independence, then hardly anyone who's not already on board would give it a second thought. The idea of this webpage is to describe my experience and some lessons learned, in order to inform and hopefully make readers more comfortable through knowledge with the idea of producing their own electricity.
A SOLAR PHOTOVOLTAIC (PV) SYSTEM converts sunlight into electricity, and shouldn't be confused with a "solar thermal" system that converts sunlight into heat for water and space heating. Unfortunately, both types of solar energy systems use the term "solar panels", sometimes leading to confusion. This page is about my solar electric system, although I do hope to install a couple of solar thermal panels one day.
A STANDARD SOLAR PANEL consists of many individual solar cells that output electricity when the sun shines on them, and a "solar array" is a number of connected solar panels. The output of a solar panel is direct current (DC) electricity like the output of a battery, and an "inverter" (picture) converts the DC electricity to alternating current (AC) electricity that is used by most households and by the national power grid. Stand-alone (or "off-grid") PV systems use the DC output of solar panels to charge a bank of batteries, which stores the electricity as chemical energy for use at night and on cloudy days. However, this page is about my "grid-tied" PV system, where the AC output of the inverter is connected to the power grid at my house's main electrical service panel ("breaker box"). The beauty of a grid-tied PV system is that during the day when my panels generally produce more power than my house is using, the excess power goes onto the grid for everyone else to use, and my electrical meter (a "net meter") runs backward and gives me credit for that power. Then at night or whenever I'm producing less power than my house is demanding, I draw power from the grid like a regular customer. All of this happens automatically and eliminates the need to store electricity in batteries. THERE IS ZERO DIFFERENCE in how my house's electrical system works, and there's no way to even tell whether or not I have solar panels or where my electricity is coming from at any given moment. The downside is that whenever there is a power outage on the grid, my power goes down just like everyone else's. For uninterrupted power, there is a third type of solar electric system called "grid-tied PV with battery backup", meaning that you have a small battery bank to get you through outages (at the cost of battery expense and maintenance). I haven't paid for electricity since March 2008, but I still do pay Xcel Energy the base charge of about $7.50/month, which I see as a good deal for having the grid act as my zero-maintenance battery bank.
I CAN'T RESTRAIN MYSELF from devoting one paragraph to the physics behind solar electricity, because it's so cool. No moving parts, just shine light on solar cells and you get electricity. It's based on something called the "photoelectric effect", which was explained by Albert Einstein in 1905 and for which he received the Nobel Prize in Physics in 1921. The photoelectric effect refers to the observation that light shining on a metal will cause electrons to be ejected from the surface, indicating that light carries energy and acts like a tiny particle (called a "photon"). Very importantly to physics, this experiment also showed that a photon is a "quantum" of light, sort of like a penny is the quantum of money...it's the fundamental indivisible unit and you can't have fractions. In solar cells, rather than shining light on a metal, a solar cell uses a specially-prepared semi-conductor material like silicon, which has electrical properties intermediate between a good conductor like a metal and a good insulator like rubber or glass. The electrons that are liberated by incoming photons of light are captured and accumulate on one side of the solar cell. Electricity is simply the movement of free electrons (those not bound to atoms), and like any electrical circuit, electrons repel each other and will naturally flow through wires from the side of the solar cell where there are many electrons to the side where there are few. With a grid-tied PV system, rather than putting an electrical motor or a lightbulb in the circuit to make use of that electrical energy, we essentially put the whole grid into the circuit.
I HAVE A RELATIVELY SMALL 2 KILOWATT (2000 Watt) PV SYSTEM that consists of ten Sharp 208 Watt panels mounted parallel to the roof on a Unirac mounting system with about 6" of ventilation underneath. The array is wired as one string of 10 panels in series, with a rated output at standard illumination and temperature of 2080 W (or 7.3 Amps at 285 Volts DC). A Fronius IG2000 inverter converts the DC power to AC power at 240 Volts AC, which is fed to a breaker in the main service panel, where the house's electrical system is connected to the grid. The official connection of my PV system to the grid occurred on Dec. 19, 2007 at 10am, when an Xcel electrician replaced my electrical meter with a "net meter" that allows electricity to flow in both directions. As of this writing, the system has run perfectly and transparently. It is essentially zero maintenance, and there are zero differences in how my house's electrical system works. The one thing that I'll call "optional maintenance" is to push snow off the panels after a storm, which I usually do just because it drives me nuts if I'm at home and it's sunny and I'm producing no power, plus I have a handy snow-removal setup (picture) that makes it an easy task.
THE AMOUNT OF POWER produced by a solar array under clear skies at any given moment depends on the angle between a line perpendicular to the panels and a line to the sun. Production is greatest when the panels point directly at the sun, and production decreases as the sun moves further above, below, or off to the side of the panel's perpendicular. The sun-panel angle changes continuously during the day and shifts with the seasons as the sun's path through the sky changes. The path of the sun depends only on your latitude and can't be changed of course, but you do choose the orientation of the panels. The total annual production of a solar array is determined by two angles that characterize the installation: the tilt angle (the slope of the panels from horizontal), and the orientation angle (the compass direction that the panels face, where True South is 180° and East is 90°). The tilt angle for my panels is 18° (the slope of a 4/12 roof), and the orientation angle is 123° (or 33° south of East, roughly East-Southeast). The ideal installation that maximizes annual PV production would be oriented Due South and tilted at an angle equal to one's latitude (40° in my case). However, even with my system's fairly large departure from ideal, it still produces 90% of the energy it would produce at the ideal angles. I found this acceptable in order to have a nicer-looking low-profile array, but others might prefer to mount at the ideal angles and get that other 10%, possibly as a ground-mounted installation.
INSTALLING A ROOFTOP SOLAR PV SYSTEM is a financially sound move in areas where utilities and/or states offer Rebates that offset part of the cost, or in areas with high electricity rates. In 2007, the total rebate from Xcel Energy of Colorado for my system was $4.50 per rated DC Watt (or $9360 for my 2080 Watt system), plus there was a Federal tax credit of 30% of the after-rebate cost up to a maximum of $2000. The rebates or other incentives vary a lot by State and by Energy Provider, and they change with time. Beginning in 2010 the Federal tax credit is 30% of the after-rebate cost with no maximum, and the Xcel rebate is much lower, and the price of solar panels has dropped substantially (especially in 2011). For these reasons, it's best not to attach much meaning to my numbers below, especially outside the Xcel service area in Colorado. That said, here's how the economics worked out for my 2 kW system in 2007, and how it would work out for my same system in 2010 and 2012 (they were kind enough at Real Goods Solar to re-run the numbers for me):
THE REBATE INCENTIVES IN COLORADO resulted from the passage of Amendment 37 in 2004, which requires large power providers to obtain 20% of their electricity from renewable sources (solar, wind, geothermal) by 2020, with at least 4% from solar, and at least half of that from residential systems. In 2010 Colorado's renewable energy requirement was raised to 30% by 2020. Xcel Energy and Black Hills Energy of Colorado offer a 2-part rebate as an incentive to install a grid-tied residential or commercial PV system, where Part 1 depends on the rated system size in Watts, and part 2 depends on how much energy the system is predicted to produce (which depends on the tilt and orientation angles). The rebates in Colorado have been decreasing with time, but this is partially offset by the decreasing price of solar and the increase in the Federal tax credit. In exchange for the rebates, I assigned my "renewable energy credits" (RECs) to Xcel to use for meeting their renewable energy requirement (some states handle RECs differently, with a private "REC market"). Part of the agreement is that I (or the next owner of my house) must leave the PV system connected and maintain it for the contract period (20 years), or else I must return a pro-rated portion of the rebate. Here are links that describe the current solar rebate programs for Xcel Energy and Black Hills Energy of Colorado. An excellent compilation of renewable energy incentives organized by state is at dsireusa.org, which also lists all sorts of conservation and other energy-related incentives. The current Federal tax credit can be found on Form 5695, "residential energy credits", under "forms" at irs.gov.
I DON'T LIKE TO ESTIMATE A PAYBACK PERIOD for my PV system because it's too uncertain, but I'll throw out some numbers anyway. My system produces about 3000 kWh/year (see next section). At 10¢/kWh that's $300/year in electricity savings, which pays off the $6000 cost in 20 years. That's less than the estimated 30-40 year life of the panels, but on the other hand the inverter will probably need to be replaced in 15-20 years. The tricky thing is that nobody knows what the price of electricity will be in 5 or 10 or 20 years, but I'd wager it will be more than 10¢/kWh, which would reduce that payback time. Also, having solar panels will be great when we move to time-of-day (or "demand") pricing for electricity, because that pricing will be highest during peak demand in the middle of the day when I am producing excess power and selling it to the grid, whereas electricity prices will be lowest at night when I need to buy power from the grid. Buy low, sell high. Another thing often overlooked is that the resale value of my house immediately increased when I put the panels on the roof, because surely "free electricity for life" is a great selling point. I also place a high value on being protected from future increases in the price of electricity, and I like having lower monthly living expenses. All of this taken together makes a solid economic case for the investment in solar panels. On top of that, I also value that solar-produced electricity displaces coal-produced electricity, so coal is not mined and burned and the resulting carbon dioxide and other pollution is not put into the atmosphere on my behalf.
REGARDING FINANCING, I don't have much information, but there may be financing options through utility incentive programs or state or local programs. Some solar companies have arrangements where you essentially rent the PV system from them and pay with the electricity savings, possibly paying no up front cost at all. Boulder County Colorado has an innovative loan program that was approved by voters and has no cost to taxpayers. A loan can be obtained for a wide variety of renewable energy and energy efficiency measures, whose repayment is made by a special assessment on the property taxes over 15 years, so the repayment is attached to the home and not the borrower. It's called the Climate Smart Loan Program, but the last time I checked this program was on hold pending resolution of some issues at the Federal level.
THERE ARE MANY WAYS to judge the performance of a PV system. One way is to compare the amount of electricity it actually produces to the predicted electricity production for the system using NREL's PVwatts solar production calculator (see Main Planning Considerations or Useful Links). The PVwatts calculation for my system is based on these inputs: 2.08 kW system size, location Boulder Colorado, tilt angle 18°, and orientation (or azimuth) angle 123°.
MY SYSTEM BEAT the PVwatts estimate by 5.6% the first year and about 2% the second and third years. This small bonus isn't too surprising because the output of solar panels usually exceeds the rated performance for the first year or two, then slowly decreases with time. Also, there is annual variability in cloudiness, plus I often use my handy snow-removal setup to clear off snow after storms, and I'd guess that PVwatts has a built-in "snow loss" factor.
ANOTHER WAY TO LOOK AT PERFORMANCE is in terms of the impact on my electricity bill (green curve in figure). On average over the year I produce 64 kWh/month more than I consume. In the summer I produce 100-150 kWh/month more than I consume, but in the winter months I need to buy back 50-100 kWh/month from Xcel. The figure also shows that replacing my older refrigerator in 2004 reduced my total electricity use by 87 kWh/month, which paid for the new fridge in electricity savings in 6 years, and saved my needing to buy another 3.7 solar panels.
CLEAR-SKY PV PRODUCTION at any given moment depends mainly on the sun-panel geometry, namely the angle between a line to the sun and a line perpendicular to the panels. This angle changes continuously through the day as the sun follows its path across the sky, and the sun's path shifts position slowly during the year because the Earth's axis of rotation is inclined 23.5° from the plane of the Earth's orbit around the sun. A summary of my PV production throughout the day and year (blue line in figure) illustrates how PV production varies with the sun-earth-panel geometry. The figure also shows the seasonal effect on my production caused by shadows from trees to the southeast (winter morning shadow) and northeast (summer morning shadow). Another figure shows that PV production decreases as the panel temperature increases (so ventilation underneath is important).
THIS SECTION IS A TRANSITION from the sections above that describe my PV system, to the sections below that are practical considerations for anyone thinking about installing a PV system. Earlier I suggested that economic considerations alone are sufficient to justify buying a solar PV system, but there are additional "bonus factors" worthy of mentioning that may or may not matter to a given individual, but they don't have to matter because they're bonus factors. This is my indulgent section where I get to opine in a blog-like way about Philosophy and the Big Picture, and then tie it back to solar panels. To summarize, it's about National Security, Economic Leadership, Climate Change, and Children.
THE FUTURE OF THE COUNTRY depends on energy choices made today. We import over half of our oil from these countries, and people across the political spectrum, as well as the U.S. military, agree that it's a threat to our national security to be vulnerable to disruption of the oil supply, not to mention the enormous expense and political destabilization that results from protecting countries with oil and the oil lanes. Regarding coal, it was great for powering the industrial revolution and electrifying America, but now we know there are long-term global environmental consequences from the massive burning of coal. It's been 200 years and now we know how to power civilization in a smarter way than making fire to boil water to turn a generator. Considering economics, I heard a story on NPR (National Public Radio) that Japan holds 40% of the patents recently filed for things that fall into the broad category of "green technology", and the U.S. is far behind with 12%. It reminds me of how Detroit dropped the ball a few decades ago and Japan got a jump on a whole sector of the auto industry, and that shortsightedness probably contributed to the 2009 disasters and bailouts in Detroit. China is quietly moving much faster than the U.S. on renewable energy. We Americans, for very good and well-deserved historical reasons, like to think of ourselves as leaders in inventing things and then prospering by selling them to the rest of the world, but we risk dropping the ball again when it comes to recognizing that the future on a crowded planet requires using energy and resources more efficiently and cleanly, and at present we're about a decade behind Europe and China. What does any of this have to do with putting solar panels on the roof, you ask? Well, I admit it's a loose connection, but in all new successful technologies there are "first adopters" who are ahead of the mainstream, and they are important because they fund the expansion of companies and development of more efficient and cheaper solar panels (or CFLs or heat pumps or whatever). So, the first adopters are the ones who jumpstart the technologies of the future, the technologies that we should be on top of in order to have jobs and flourish in the next generation. Putting solar panels on the roof contributes to that, and it contributes more to do it earlier than most other people. Nonetheless, I still stand by the notion that solar is a good move on personal economic grounds alone.
THE FUTURE OF CIVILIZATION depends on energy choices made today. The country and the world need to move toward a long-term sustainable way for massive numbers of humans to live in comfort on our one and only planet without ruining it or using it up. It's our moral responsibility to future generations and our own children to leave them a tolerable planet with a decent future. Hopefully we can all agree on that. Powering civilization by burning massive quantities of coal and oil puts massive quantities of carbon dioxide (CO2) into the atmosphere, which absorbs and traps heat in the atmosphere, warming the planet (see "Global Warming in a Nutshell"). That aspect of Global Warming (or Climate Change) is pure fact based on long-term measurements of CO2 and temperature, and it is well-understood in terms of basic chemistry and physics. Some predicted consequences of continued global warming are warmer and more acidic oceans, altered ocean circulation and rainfall patterns, more energetic and more frequent hurricanes, more severe droughts and floods and shifting agricultural viability, melting ice caps and sea level rise that will create millions of climate refugees, and so on. These aren't even the worst-case scenarios. Some of these things we observe happening today (see the NASA Evidence Page or the Center for Climate and Energy Solutions (C2ES) Impacts Page). Something like 99% of climate scientists, some of the best critical thinkers on the planet who study this for a living and are therefore experts, are convinced by an overwhelming amount of data and understanding of the physics that these consequences will occur if we don't change our energy sources, yet there is an irrational amount of skepticism out there. Why? The coal and oil lobbies are very powerful and well funded, and they know how to buy support in Congress and how to fund advertising aimed at deceiving the average person who doesn't have a good understanding of climate science. Another source of irrational skepticism is that the skeptic hasn't made the effort to actually investigate the science that makes the conclusions of climate change research so clear to those who do understand it. That makes it easy for well-funded powers to deceive people and turn what should be a moral, economic, and national security question with long-term consequences into a political and ideological question with short-term consequences. When it comes to the question of who to believe, I like this simple analogy. If I want to know how to fix my plumbing, I will ask a plumber, not a climate scientist. But if I want to know if all the hoopla about Global Warming is real and should we do something about it, it seems like common sense to ask a climate scientist, not the Coal Lobby or the American Petroleum Institute or many (but not all) politicians. Our energy choices today will determine the climate of the planet in coming decades and the quality of life for the next generation, and putting solar panels on the roof contributes to that in a positive way. Nonetheless, I still stand by the notion that solar is a good move on personal economic grounds alone.
FIRST INVESTIGATE REBATE PROGRAMS in your area. One source of information is this list of incentives organized by state, or one of the Colorado programs listed under useful links. My investigation began with the Xcel Energy newsletter, which gave the website and phone number for their program. I followed all the links on the website and learned about the program. Although it is possible to install a system yourself, it has to be approved by the local building inspector and by the power company, so check to see if there are restrictions on do-it-yourself installations. I bought a fully-installed system from a solar company, although I chose the components and designed the panel layout on the roof to minimize shading. Some advantages of a package deal are: they handled the building inspection and coordination with the power company, they took care of the rebate (which I assigned to them and they applied to the cost up front), they gave me a discount on parts, answered my many questions, and then did the installation in one day. Another possibility is to purchase all the parts and then separately hire an installer (including an electrician who is solar-qualified). Rebate programs may have a list of solar suppliers and installers, and of course google is an endless source of information.
TWO ANGLES MUST BE MEASURED to plan the size of a PV system and calculate the rebate. The tilt angle is the slope of the panels from horizontal, which is also the slope of the roof if the panels are mounted parallel to it. The orientation (or azimuth) angle is the horizontal direction the panels will face, as a compass direction. Any solar sales person invited to your house should be able to measure these angles, but you can measure the angles beforehand once you know the installation site (see next section).
TWO IMPORTANT PLANNING NUMBERS can be calculated from those angles using NREL'S "PVwatts" solar production calculator. Installers and rebate programs will probably also use PVwatts estimates of PV production for your location and panel angles, and then compare it to the maximum production you would get if your installation had the ideal orientation angle (Due South, 180°) and ideal tilt angle (equal to your latitude). From the PVwatts webpage, select "version 1 calculator" (twice), then select a nearby location from the map. First calculate the maximum possible PV production at your location by entering the ideal tilt and orientation (or azimuth) angles (your latitude and 180°), and enter a DC-rated system size of 1 kW (to get a "per kilowatt" result), then select 'Calculate'. Save the total annual PV production in kilowatt-hours from the "AC Energy (kWh)" column. Then go back one page and repeat the calculation using the measured tilt and orientation angles for your installation. The result is the total annual energy production in kWh for a 1 kW system at your location for the specified tilt and orientation angles. Once you determine your desired total annual energy production (see System Sizing section), it will be easy to calculate the needed system size from this "per kilowatt" production number. The PVwatts calculation for my system (2.08 kW system size, location Boulder Colorado (40°N latitude), tilt angle 18°, and orientation angle 123°) yields a total annual energy production of 2753 kWh. The maximum possible production for my system at the ideal tilt and orientation angles is 3034 kWh, so my installation produces 2753/3034 = 0.907 = 90.7% of the maximum possible production at my location, which was good enough to receive the full rebate (less than 90% and the rebate is decreased proportionally). Interestingly, at my location PVwatts estimates a slightly greater PV production for an orientation angle of 170° than for the "ideal" angle of 180°. The reason is the tendency in this area to cloud up in the afternoon, so over the course of the year you come out slightly ahead by facing the panels a little toward the morning sun. If you're less than clear about the difference between a kW and a kWh, read Power and Energy (Watts and kWh) in a Nutshell.
FOR MAXIMUM POWER PRODUCTION the panels should face True South (bearing 180°) with a tilt angle equal to your latitude. However, my installation is pretty far from ideal and still produces 90% of the maximum possible energy, so low-profile installations parallel to the roof are a reasonable option with minimal compromise to production for many situations. Another option is to mount the panels with a more favorable orientation and tilt that is not parallel to the roof. Another good option is a ground-mounted system if you have the space and good sun exposure, with the bottom edge of the panels above the level of snow and vegetation, and ideally mounted at the optimal tilt and orientation angles. There are also more productive but more expensive mounting options where the tilt and/or orientation of the panels can be changed, either continuously throughout the day (a Tracking system), or periodically during the year to optimize for either summer (high sun) or winter (low sun). Good ventilation beneath roof-mounted panels is important because they get really hot (I've measured up to 170°F), and the production efficiency decreases with increasing temperature.
AN INVERTER CONVERTS THE DC ELECTRICITY produced by the solar panels to AC electricity, which is fed to a circuit breaker in the main service panel where the house's electrical system is connected to the grid. The inverter should be mounted near the service panel if possible, mainly so the required DC and AC Disconnects are nearby. It's good to minimize distances between the panels, the inverter, and the main service panel in order to minimize "line loss", or electrical energy lost to heating of the wire. Line loss increases in proportion to the length of the wire, but it decreases with increasing wire size, so longer distances should be compensated for with bigger (and more expensive) wire. There is a new inverter option that wasn't available to me in 2007, which is to replace one or a few big wall-mounted inverters with a "micro-inverter" on the back of each panel that outputs grid-ready AC power directly. Besides eliminating big inverters, this allows each panel to be independent of the others, which makes the array much less susceptible to shading or a weak panel than is my single string of 10 panels connected in series. There are other advantages to micro-inverters, and I would look into them further if I was planning a PV system today.
SHADOWS ARE A MAJOR CONSIDERATION for siting and designing a PV array. Hard shadows from trees, buildings, and roof protrusions are bad for production and should be avoided to the greatest extent possible. A cloud shadow is not an issue because it affects all panels equally, so the power output just decreases in proportion to the decrease in illumination. But with partial shading by hard shadows, even a small amount of shadow can substantially reduce the output, way more than in proportion to the area of panel it covers. Similar to the way that the flow rate of water from a pinched hose is controlled by the pinch point rather than the diameter of the hose, a hard shadow "pinches" the flow of electricity through a whole string of solar panels that are connected in series. With the micro-inverter approach mentioned above, only the panels that are affected by hard shadows experience reduced output. For planning purposes, it's useful to observe the path of shadows throughout the day, but it's difficult to imagine how the path of a shadow changes through the year. For example, it might not be apparent during winter that a tree to the northeast will cast an early morning shadow on the panels in the summer, or it might not be apparent during the summer that a tree to the southeast will cast a shadow on the panels in winter when the sun is lower. Some installers use a domed gadget called a Solar Pathfinder (or some such thing) to get a rough idea of how much shading there will be. Some things can be changed (cut down tree or roof protrusion), and some things can't (neighbors' trees, buildings), so some shading may be unavoidable. A general guideline is that it's most important to be shadow-free during the peak power producing hours of 9am-3pm (or 10am-2pm in winter). Add an hour to those times during Daylight Savings Time, which the sun doesn't recognize. My two tree shadows take a bite out of production at different times of year, which I knew about up front because I wrote a computer program to calculate shadows based on equations for the path of the sun through the sky. The main lesson is to be sure your installer seems knowledgeable about shadows and how to deal with them. Some sites are just bad for solar.
PLANNING THE SIZE OF A PV SYSTEM (in kW) depends on what percentage of your total annual household electricity consumption (in kWh) you use as a target. I planned my system to provide 100% of my annual electricity use, and it came out to be a 2 kW system (which is small). For various reasons, my electricity use is a lot lower than the average American, so the average household would need either a bigger system or a lower target percentage. My 2 kW system was based on an average monthly electricity consumption of 225 kWh (or 2700 kWh/year). The U.S. average monthly electricity consumption of 675 kWh (8100 kWh/year) is triple my electricity consumption and would require a 6 kW system to supply 100% (or a 2 kW system would supply 33%), assuming my same location and panel tilt and orientation angles. So, the first step is to determine or estimate your total annual electricity consumption in kWh. If you have old electricity bills you can just add up the kWh for a year, or you might be able to get that information from your electricity company, possibly through their solar incentive program. You could also make a guess based on the kWh from your next electricity bill, and possibly modify it depending on how your electricity use varies through the year. It makes a lot of sense to reduce the needed size of a PV system by reducing your electricity consumption, for example by replacing an older refrigerator or making other energy efficiency improvements suggested in the next section.
THE RATED PV SYSTEM SIZE (in kW) is the size needed to produce your target annual electricity production in kWh, a number we'll call "Ptarget" for "annual kWh Production target". Back in the Main Planning Considerations section you calculated how much annual PV production in kWh you would get from a 1 kW PV system for your location and panel angles. We'll call that number "P1" for "annual kWh Production from a 1 kW system". The number of 1 kW systems needed to produce Ptarget kWh annually is the ratio Ptarget/P1. This is the same thing as the needed system size in kW. For my installation location and panel angles, PVwatts shows that a 1 kW system would produce 1324 kWh of electricity annually (P1), and my target annual PV production was 2700 kWh (Ptarget), so my needed system size was 2700/1324 = 2.04 kW. Knowing the system size allows you to estimate costs and plan a panel layout, which solar companies will figure out too, so you don't have to do it yourself unless you want to. Panels come in different sizes, shapes, and rated power output, and these characteristics determine how many panels are needed and how much area they will cover. A common rated output from a solar panel these days is about 200 W, so it takes 5 such panels per 1 kW of rated system size. If you're less than clear about the difference between a kW and a kWh, see Power and Energy in a Nutshell.
SOLAR PANELS ARE EXPENSIVE. It's much cheaper to decrease electricity use by improving the energy efficiency of a house than it is to buy enough solar panels to power inefficient devices. Inefficient devices use more electricity to do the same job because of poor or cheap design, basically by turning electricity into waste heat. There are many ways to use less electricity by using it more efficiently, some of which are obvious and some not so obvious. I'll mention a few things that have been particularly effective in reducing my energy consumption and saving money while at the same time making my house more comfortable. There is much information and many ideas out on the internet, but I especially recommend the 1-page "skinny" sheets on various energy efficiency and renewable energy topics at the Center for Resource Conservation. The Electropedia is an excellent source of information on the generation, storage, and economics of electricity, and on energy sources (fossil fuels, biofuels, solar, wind, nuclear), renewable energy systems, energy efficiency, and more. One gadget I use frequently to measure how much electricity something uses is a power and energy meter called a Watts Up? (standard model). Anyone in an Xcel Energy service area can borrow one of these meters from a participating library.
THE ENERGY EFFICIENCY OF REFRIGERATORS was first regulated in 1993, then standards were raised in July 2001, and a review of standards is in progress in 2010 (reference). The plot of my electricity use since 1997 (right panel) shows that replacing my 1980s model refrigerator with a more efficient model decreased my electricity consumption by a little over 1000 kWh/year. At 10¢/kWh, that's $100 in electricity savings per year, so my new $600 refrigerator paid for itself in electricity savings in 6 years. Now I have a newer, nicer refrigerator FOR FREE, and I'll continue to save $100/year in electricity. But wait, there's more. That extra 1000 kWh/year used by the old fridge comes from inefficiency related to poor (cheap) design, and that energy didn't help cool food but was just converted to waste heat and dumped into my kitchen. The last thing my house needs in the summer is an interior heat source. For reference, a small hair dryer or portable electric heater (the kind with wires that glow red) uses about 1 kW of power, so the 1000 kWh/year of extra heat put into my kitchen by the old fridge is equivalent to running a hair dryer in the kitchen for 1000 hours/year (2.7 hours each day). What a waste and discomfort! People with air conditioning need an efficient fridge even more than others, because YET MORE electricity must be used to remove that extra heat from the house. When a solar PV system is considered, it's EVEN MORE important to have an efficient refrigerator. If I still had the old fridge then my PV system would have needed to produce an extra 1000 kWh/year, and my design annual electricity consumption (previous section) would have been 3700 kWh/year rather than 2700 kWh/year (37% more), so I would have needed an extra 3.7 solar panels (13.7 instead of 10), and my installation would have cost roughly 37% more, just to make an extra 1000 kWh/year of electricity for the sole purpose of being pointlessly converted into waste heat and dumped into my kitchen. In summary, it makes ultra good sense to replace old refrigerators, especially pre-1993 models and probably also pre-2002 models (but they'll have a longer payback period since they're more efficient than pre-1993). I wish, especially now that I have solar panels, that I would have looked into a more expensive but super-efficient refrigerator like a Sunfrost, or at least I'd search among the most efficient models at the Energy Star refrigerator database. Very old refrigerators, like the white 1930s to 1960s models, use an extraordinary amount of electricity, and replacing one would pay for itself quickly. Some utilities have an Appliance Recycling Program that will pay you to take an old, working, energy-hog refrigerator off your hands. A Watts Up? is particularly well suited for measuring refrigerator energy use in kWh/month, which must be averaged for at least 24-48 hours to get an accurate measurement. When choosing a refrigerator, consider both the purchase price and the cost to operate it, namely the total life cost of electricity. Spending more up front often pays off big in the long run. Also, be aware that ice-makers and through-the-front water and ice dispensers reduce the energy efficiency, and freezer-on-top is more efficient than side-by-side, all else being equal.
ATTIC INSULATION AND AIR-SEALING has been an improvement in both winter and summer, lowering my utility bill for both natural gas and electricity while at the same time making my house more comfortable. One day I'll make a webpage about this project because it was so successful, and because I learned a few tricks that made the job easier. Originally I had 6" of old and poorly laid insulation between the ceiling joists, and my first step was to add a second layer of 6" batts perpendicular to the first layer, and also a first and second layer above the previously-uninsulated attached garage. I also improved the soffit ventilation, which was partially blocked by the first layer being stuffed over the top of the wall, and I did some air sealing. Next I loose-laid a layer of radiant barrier over the top of the new insulation, but later I stapled the radiant barrier to the roof trusses as shown in the photo. Radiant barrier comes in rolls up to 4' wide and consists of a sheet of plastic with a reflective aluminum coating on both sides (like aluminum foil), and perforated with tiny holes so it doesn't trap water vapor in the insulation (see how it works). My attic temperature often reaches 130-140°F in the summer, and radiant barrier reflects 95% of radiant heat energy, so it is very effective at keeping that extreme heat from "shining" on the top of the insulation. Regular insulation inhibits conductive and convective heat flow, and radiant barrier completes the picture by inhibiting radiant heat transfer. Finally, I took advantage of rebates for attic insulation and air sealing in Boulder County to have cellulose insulation blown in to a depth of 15" (approximately R-50 insulation value). Although I benefit from lower gas bills in winter, the most dramatic improvement has been in the summer, when the temperature inside my house before this project could reach the low 90s in the midst of a hot spell. After the attic project (and the new refrigerator), it now gets up to the low 80s indoors during the hottest periods, and most days it's in the upper 70s. So now my house is more comfortable in summer and my energy bills are lower year-round, all without the large expense of A/C. I got rid of my evaporative cooler (swamp cooler) and now rely on small fans for the hottest days and a window fan for moving cool air through the house at night. The effect of attic insulation and air sealing on a house with air conditioning would be to run the A/C less, or move to a smaller A/C unit, or move to a more efficient swamp cooler (if the climate is dry enough). The cooler indoor temperature during the summer also reduces the electricity consumed by the refrigerator, which must work harder when the kitchen is warmer. A refrigerator's electricity use increases by 2.5% for every 1°F increase in temperature - so don't put a refrigerator in a hot garage!. Also, an efficient A/C unit is doubly important for a house with solar panels, because it's much cheaper to buy a more efficient unit than to install more solar panels simply to power inefficiency that does no cooling.
THERE COULD HARDLY BE ANYTHING LESS EFFICIENT than making light by heating up a wire until it glows (incandescent lights). Only a few percent of the electrical energy is converted to light, and the rest is converted to heat. Usually you don't want that heat (summer), but even if you did (winter), using electricity is the least efficient and most expensive way to make heat. Compact fluorescent lights (CFLs) are the corkscrew bulbs that work the same way as the larger office or garage fluorescent tubes, namely by passing electricity through a gas to excite all the atoms, then the excited atoms drop to a less excited state by giving off light. This process is 4 times (400%) more efficient than incandescent lighting. A CFL uses only 15 Watts of power to produce the same amount of light as a 60 Watt incandescent, and CFLs supposedly have a lifetime 10 times longer than incandescents. They cost more up front (about $2 each, sometimes cheaper), but they pay for themselves in electricity savings many times over. See "Power and Energy (Watts and kWh) in a Nutshell" for the CFL payback calculation (and other interesting things), but here is the bottom line. The payback time for a 15 W CFL (equivalent to a 60 W incandescent) is 7.5 months if used 2 hours/day, or 3 months if used 5 hours/day, or 0.6 month (18 days) if used continuously (assuming electricity costs 10¢/kWh). Since CFLs are more efficient and put out less waste heat, they are doubly cost effective for a house with A/C. They are yet more cost effective for the owner of a PV system, who can avoid buying an extra solar panel just to provide electricity that is converted into waste heat in incandescent lights. CFLs don't perform identical to incandescents and may take a little getting used to by some people. When they're cold and first turned on, they are less than full brightness and take a few seconds to warm up, but I don't find this to be a problem.
YES, ABSOLUTELY. I've had zero problems, the panels have withstood vicious winds and modest hail, they've gotten better than predicted electricity production, and if I wasn't interested in them I'd hardly know the panels were there. The one maintenance task is snow removal, which I do by choice rather than necessity, although in some climates it may be more of a necessity for flat or low panel mounting angles. With the utility rebates and Federal tax credit the system will eventually pay for itself in electricity savings, even without considering higher electricity prices or time-of-day pricing in the future. I also like being electricity-independent and having lower monthly living expenses. They're also great fun to own and study for a data geek like me. And if that weren't enough, I value knowing that my share of the electricity is produced cleanly and contributes toward getting us on an energy path where future generations aren't saddled with the consequences of our short-sighted, old-fashioned, and frankly, stupid way of powering civilization with dirty coal now that we know better. In conclusion, I have no regrets about installing a PV system, except I wish I would have added an extra panel or two in the beginning.