Solar economically viable: back up figures

Picture: The Unity House: a working example of the kind of grid-tied solar installation covered in the following article.

One of the commentators on a piece I wrote that happened to get posted on Andy Revkin's NYT blog asked for some back-up figures as to the viability of solar PV.

The claim I made in the original piece was that solar power is cost-effective now if purchased using lease or long-term finance. The original comment referred to both solar thermal hot water and solar PV. The commentator specifically asked for references for the solar PV claim, and that is what I give here, but I could also work up some numbers for solar hot water if required.

This is a conditional claim, as I mentioned in the piece, but because the piece was written to be "punchy" and so actually get read, some may have missed the conditions:

An ordinary solar PV installation is currently cheaper than your electricity bill in several states, if you use longer financing and a lower interest rate, or if you lease. Cost effectiveness varies by state and location. In some states you need the federal rebates, in some states you need state and federal, in others you can swing it without, and in some it doesn't pay at all. (A database listing available rebates and other subsidies is available here.)

In all cases, you need a flat or south-facing roof exposed to full sun.

Lease versus purchase? For outright purchase to be cost-effective, you need to have a supplier of capital (either yourself or a financial institution) that is willing to see solar PV as part of the house, much as we currently see the furnace or hot water tank, and calculate opportunity costs or finance accordingly.

Since this requires uncommon vision for banks right now, at least on a first mortgage, and since few of us will have the dollars up front, for most of us ordinary folks, whose capital is tied up in house and retirement savings, this would require a home equity loan.

For the lease option, you need a company such as Sungevity or New England Solar.

The cost structure is much the same in either case.

So, in the case of purchase, for instance, if you live in Maine and currently draw power from the grid to the tune of 10,000 kilowatt-hours (KWH) per year (a little more than the New England average), you'll pay around $150/month to the power company, give or take twenty dollars depending on the source of that power.

If instead of paying this large bill, you place a 3 KW per hour rated solar power system on your roof, opting to pay a smaller power bill plus a solar system purchase bill instead, you'll make about 5,000KW of your needed 10,000KW/year with that 3KW of solar modules.

(3KWH/hour x 4.5 hours/day average sunlight x 365 days per year = 4,928 KWH/year)

Half is about right. Currently, without further development of the grid, and with Maine's current net-metering law, although you don't legally have to do this, I would size the system so it runs the parts of your house that keep running during the middle of the day, so fridges, freezers, hot water tanks, and the like. This is usually about half the total. This particular choice puts the least stress on the grid, and is not likely to cause the negative externality of a major grid demand-supply imbalance unless an awful lot of us do the same thing in one swell foop.

To the grid the initial connection of the panels, once the sun shines on them, looks more or less as if you disconnected a few appliances. Later, once the grid gets used to solar and distributed generation in general, and has figured out how to "dispatch" it to where it's needed, or if we ever get a "feed-in" tariff law, you can add more. Oversizing the original inverter would facilitate adding more modules later.

The current installed cost of that 3KW-rated solar PV system is about $25,000 in Maine. I got my figure from my regular communications with local installers.

The commentator, who lives in Oklahoma, asked for specific references. This is a bit like asking for a reference for the price of a new furnace installation in Oklahoma City. I don't know any installers there, and these kinds of pricing questions are not normally the concern of referenced scientific works. Instead, I would refer him to the price of ordinary modules and suitable inverters in a catalog such as Northern Tool, or use Google shopping, or any other outlet whose prices are the same nationally, or ask him to fill out the Sungevity online quote form.

A lot of installers will buy solar equipment from online vendors for the simple reason that you can do your ordering at night or early morning, when you're not out installing. I also buy from the same kinds of outlets, albeit for my educational and research work, and for our own two modest properties, which is why I'm familiar with the pricing, brands, and outlets.

To provide some objective back up for the figure of $25,000, I went to these sources and priced the equipment needed. Click on the links above and below to see the prices.

From these ordinarily priced sources, I see that I can get 110 watt panels for $410 and change, 135 watt panels for around $450, and so on. I need 28 units of 110 watt panels to make my 3KW system, so if we buy them from Northern Tool and include some dollars for shipping and taxes, the total is about $13,500 (at $480 each), a conservative estimate. The inverter is about $3,000. There's some wire needed, a couple of new kilowatt-hour meters, and a big old circuit breaker, about $1,500 all-in. The rest is design and labor.

It's actually only a couple days work for two guys to install, but installers tend to charge top dollar, as far as contractors go. This will change as our local community colleges are cranking up to qualify many more installers for solar and wind power devices using ARRA and other recent federal investments.

Still, $25,000 is a lot of money just to get out from under half your electricity bill. But you get 30% back on your taxes in the first year or years through the governments current rebate programs for renewable energy, so the cost is then $17,500. Financed on a mortgage or home equity loan at 20 years at current low interest rates, the price of that much money is between $95 and $120 per month.

The mean point of that, say $105, compares unfavorably with the $75 you pay for half your electricity bill, but as you can see it's very close. I also picked an unfavorable state, Maine, with relatively high power rates but relatively few average hours of sunshine daily. In California, Arizona, or Florida or any sunny state marked on this map, the average hours of daily sunlight are more like 7 or 8, so there is a considerable increase in power produced, enough to close the gap.

Or you can install a smaller system.

It turns out that Maine has additional state-level rebates currently that also close the gap, but I wanted to keep the numbers simple so people can begin to see the kind of thinking needed here.

Folks can and should quibble with my numbers (if you have nothing better to do) but as with many economic and ecological modeling problems, the insight here is not so much to do with precision as it is to do with finding new and different ways of looking at the system's dynamics. In this case, that means understanding solar PV as long-term capital, perfectly appropriate when you consider how long the panels last, 25 years or more.

Actually we don't know how long PV panels last. There are several systems I know of that were put up in the late 1970s or early 1980s and the panels are still working fine, although the inverters and batteries have all been replaced. In some cases there is a small drop in output, around 15%, but only after long time periods. Grid-tie systems don't need batteries, and inverters have improved a good deal, so these concerns are far less in the kind of system we're talking about here.

The take-home message:

The commercial viability of solar depends as much or more on the kind of finance used, any subsidies, and the hours of sunlight per day, as it does on the price per installed watt.

On that last, though: It's coming down. There are two ways this is happening. The first is market reorganization as speculators take advantage of these price points. Sungevity is a case in point. Their business is less solar design and more arbitrage. What they are selling is less solar installations and more market knowledge and various kinds of guarantees to homeowners and to their own investors. They are also aggregators of a kind, taking advantage of various economies of scale. Lease financiers such as Sungevity already face more favorable prices for PV installation, so they can bring the capital price down. I'm not privy to their exact price structure, but I can guess. They buy modules wholesale from the factory, which saves them a large percentage. They have one engineer for dozens of projects, saving design costs. They don't customize each set-up, at least not so much, so their installers can knock them out quickly, much as the Dish or Direct TV guys get your leased satellite dish up in an hour or two. And instead of being a bank and having to satisfy a loan committee and loan regulations, they're using private equity money, which they can access faster.

This kind of money does, however, require a higher return, so I would expect that the homeowner who is prepared for the nuisance of ordering up the installation and filing for the rebates herself can do as well or possibly better if she can find cheap capital.

The rebates also reorganize the market. Installers emphasize them in advertising and audits. I've gotten one of these rebates every year for the past five years, although so far only for insulation, not solar. Eventually I'll get around to solar, but only after I've gotten all my other energy needs down using insulation, renewable heat, and the like. That's the right way to think about it because the return is greater on insulation, up to a point, than on solar. Again, this is ordinary economic thinking. I'm not a wealthy man. Opportunity costs for the use of my available capital prevent me from doing this right now, but once I've paid down the home equity loan I'm currently using to pay for wood stoves, Energy Star appliances, windows, and insulation, I can use the same for solar.

Additionally, the actual price of solar PV modules and associated inverters has dropped by roughly 40% in the last three years. This has been reported nationally, but I also have personal experience to draw on. The last time I priced a Kyocera panel it was $699 for 135 watts, now it's $450.

Actually, I just refitted the small solar cabin we own, an off-grid summer/hunting/recreational camp, with a new inverter and a charger, for about a third of the original price. The solar module was still fine after 8 years' use. They're hard to break.

If this trend continues, as we expect it to do because new factory capacity is being built in the US, China, Japan and elsewhere, and new solar production technology is much cheaper to run than older technology, even Maine would be cost effective without the state-level rebates, in one to two years.

It is also a real possibility that the federal rebate would be unnecessary within a few more years.

So, my personal conclusion is that the solar age is here, and in some ways inevitable, except that we tend to see solar as short term, not long term capital. I expect we'll figure that out as people begin to save money with solar. I'm looking forward to buying a system in about five years here, by which time, with all the new production capacity, I would guess the price of modules should be about 70% of what it is now. Of course, I plan to install it myself, so I'll pay less. I also plan to buy the panels at wholesale price, and because I run only energy-efficient appliances and don't use electrical power for hot water, I don't need as much as 3KW. I can get by with half that.

The real problem, as the commentator mentioned, is the fact that we currently have sufficient grid-based electrical production capital to make our power demand, in the form of ordinary power stations, so the solar capacity is redundant. The other problem is that the solar power comes only during the day when the sun is shining.

But this last turns out to be a good thing with solar. (It's more problematic with wind power.)

Maine, and New England in general, like most northern regions, has a daily electrical power demand peak that is roughly twice its nighttime "base load." Adding solar power to the grid in this fashion, one roof at a time, means that the power companies will have to change the way they dispatch power. Each time an inverter is connected to a breaker panel, when the sun shines, the ebb and flow of power in the grid is slightly altered during the day. For the time being and until there is a lot more solar capacity installed, any large amounts of solar power that arrive during this daytime peak mean only that the grid operators have to slow down some stations, and these are generally peaking stations, not base load stations.

A peak load station is one that can be powered up and down quickly. These include natural gas stations and some kinds of hydropower stations. A base load station takes time to power up or down, and instead likes to run continuously. Coal and nuclear power are typically used for base load.

Currently, grid operators probably can't see or measure any solar-powered drop in peak demand anywhere except perhaps in California. This will change eventually. But for a few more years yet of adding additional capacity, they'll be able to cope. The problems will arise when the new solar capacity begins to approach the capacity of peak load power stations in any particular region.

Until then, any solar power capacity additions will have much the same effect on the grid as a football game ending, except that this football game ends each day, as the sun climbs in the sky. The balance between demand and supply is momentarily upset, and operators reduce supply by powering down certain kinds of peak load power stations, particularly natural gas-powered ones. It wouldn't be hard for weather forecaster and geoscientists to build computer models to predict this power production timing, which would help the power companies cope. We're currently learning to do this with large scale wind power.

You never get rid of all the base-load and peaking power stations because the sun doesn't shine 24/7. There will still be night-time, and still be cloudy days. In this sense, solar, and wind are best seen as a replacement for fuel in the grid, not a replacement for power stations. You still need much the same amount of base load and peak load power station capital, but you can get rid of quite a bit of fuel use.

When making cost comparisons across grid sectors, the cost structure of the solar sector looks to an economist more like the cost structure of fuel purchasing. Solar power happens to have a stable price, and one that is coming down, while oil and gas have recently been unstable and can spike upward, so this is an attractive feature of solar for grid and power company managers.

But you can get rid of some power stations if you reconfigure the grid.

I happen to think that some of the new "fourth generation" nuclear power systems such as the "Hyperion" reactor, or the new "Bloom Box" fuel cells, might also be interesting new capital options for power and grid operators because they offer the new option of distributed base load, which we can use together with solar to harden electrical generation against natural disaster at the local level, create "microgrids" and reduce the extent of outages. But there are other interesting new ways to think about base load capital too. One is electric vehicle storage.

You can also get rid of more stations, and even some base load stations, if we begin to drive battery electric cars, which we charge up during the day and draw down at night. You have to charge them up at work using solar power on the roof of your workplace, or placed around the parking lot. You then drive home and plug your car into your house's computer-run power center, which makes sure only to draw down enough power that you can still get back to work the next day.

Whatever our commentators may think of all this, the bottom line is, there's a lot more electrical generation technology around now, some of it renewable, others, like the Bloom Boxes, still need fossil fuels (natural gas, in the case of the Bloom Box), but are better for the environment.

This is US-owned tech, and we should be thinking about how we use it to get our climate emissions down, meet or exceed our international commitments, and reduce the cost of energy to consumers.

The idea of using leases or long term finance for solar is just one example of new ways to think about solar power, power distribution, and energy demand, and associated ways to reduce climate emissions.

There are others: In Maine, one local renewable energy company is proposing a solar power station to reduce the cost of what would otherwise be a very expensive power line upgrade, another example. In California, the new thin-film PV produced by Nanosolar is going primarily to power stations on the roofs of big box stores.

The greatest objections to all of this interest in solar will come from people tied to old ways of producing power, either economically, politically, or culturally. They will object that this is all pie-in-the-sky, utopian, not "real" energy thinking. Apparently "real" energy has to produce pollution to get taken seriously. This is a cultural tendency I've observed among, for instance, older petroleum geologists and engineers. The cleanliness and efficiency of, say, a Nanosolar plant is foreign to them. Although some "paleotechnology" is involved in solar power production, it's minimal compared to coal and oil, and the life cycle production per unit pollution is much greater.

In any case these are not rational objections and will not withstand even another 20% or 30% drop in the price of solar PV modules.

Such folk of course also hate the 30% federal subsidy for solar, and object to that too. But even were the subsidy to be removed, the price would still likely catch up after only a few more years. And remember, the hidden costs or "externalities" of coal and oil are substantial, and include oil spills off our shores, dead miners killed in accidents, the alteration of Appalachian skylines by mountaintop removal mining, and, last but not least, climate change. The cost of such things are major subsidies from the general welfare to destructive forms of power production and should also be taken into account. In a sense this is partly what solar rebates do, correct the positive externality that is supplied to the general public when a homeowner or business turns their roof into a small power station for their own and the general good.

Another option would be a carbon price or carbon tax to correct the negative externality of power production pollution and accidents. I quite liked Maine (Republican) Senator Susan Collin's recent proposal for a carbon tax-and-dividend system because a dividend would reduce any regressive effects of a carbon tax.