Going solar, part 2: Economics and hardware

Costing a solar system is complicated; there’s just no getting away from that. So I’m not going to present a detailed cost breakdown; instead, I’ll outline a few of the complexities.

First of all, solar panels vary in efficiency. Higher efficiency panels generate more power per unit area, but they also cost more. So do you buy more panels, buy better panels, or aim to use less electricity? Similarly, the inverter which turns DC electricity from the panels into AC has a certain capacity and efficiency, and there are lots of companies making them to all kinds of specifications.

Panels gradually degrade over the course of their lifespan. The output from monocrystalline silicon solar panels drops by about 0.4% per year. Manufacturers normally warranty that after 25 years their panels will produce not less than 80% of their original output. So, you need to allow for that in your calculations.

Meanwhile, new panels go down in price every year, and go up in efficiency every year. So not only do you need to work out if solar makes financial sense this year, you also need to consider whether it might make enough extra financial sense next year that you would be better off waiting a year.

solarindustry
SEIA

Don’t forget that electricity also tends to get more expensive each year. Although Austin hasn’t seen the kinds of rate increases the rest of Texas has, nevertheless it’s a pretty safe bet that prices will go up in future years.

In Austin, electricity rates also go up the more you use. So if you cut your usage by 50%, that won’t necessarily cut your bill by 50%. Furthermore, summer and winter rates are different.

This seems like a reasonable point at which to mention grid parity. This is an important concept to solar power; it’s the point at which solar power becomes as cheap as the cost of conventionally-produced electricity purchased from the utility grid, in dollars per watt. Ten states have already reached grid parity; the only reason Texas isn’t one of them is our relatively cheap electricity costs. You can find maps online showing projected grid parity dates for various states, as a very rough guideline for whether you are likely to find solar affordable.

Austin’s “Value Of Solar” buyback program currently gives a buyback price which is almost grid-parity level. However, a feature to be aware of is that if you generate more electricity than you use over the course of an entire year — that is, if your overall bill would be negative — you can’t carry over credits to the next year or make an overall profit for the year. Instead, the excess energy goes to the city to use, for free. This means that the ideal situation is to have your estimated generated electricity over the course of an average year be exactly the same as your electricity usage over the course of an average year, i.e. “net zero”. So if you’re thinking you might want to set up a solar power system any time soon, my first suggestion is to start recording all the consumption figures from your electricity bills in a spreadsheet, month by month.

Then there are the installation costs. They can vary a lot, and can be significantly more than the hardware costs. In our case I expected them to be significant, because our two story house with high ceilings meant that someone was going to have to hoist a dozen large solar panels up about 10 meters, probably in the middle of a hot Texas afternoon. I didn’t know who that someone was going to be yet, but I was damn sure it wasn’t going to be me.


Greens MPs

Speaking of hot summers, the expected capacity of your solar system will also depend on your geographical location. The reason Texas is hotter than Minnesota is that more solar energy reaches ground level per unit area, because of the curvature of the earth. That’s probably just a theory according to Texas schoolbooks, though.

Orlando-Ferguson-flat-earth-map edit.jpg
Orlando Ferguson-flat earth map, licensed under public domain via Wikimedia Commons

Finally there are possible effects on property taxes (and often exemptions you can apply for), rebates from the utility company, costs of permits, tax savings from federal tax rebates next April, and all the other bureaucratic pluses and minuses to consider.

I started off by ignoring most of the details, and doing some Internet searches for example costs of installed solar systems that were about the right size to match our average monthly electricity consumption. Those varied a lot from state to state, but the numbers did seem to suggest that a system would probably pay for itself in under 10 years, and not be too outrageously expensive to install.

Encouraged by this, I went to some online tools which attempt to estimate the cost of a solar system more precisely. I answered their questions, and looked at the figures they gave me. Those numbers varied less, and still looked to be in a sane price range. So, convinced that I wouldn’t be wasting anyone’s time, I decided to contact a company that does residential solar power installations. Faced with dozens of possible solar power companies, none of which I had heard of, I didn’t really have much idea who to call. However, one of my ex-managers at IBM had left to work in the solar power industry a few years back, so I contacted him and passed on my spreadsheet of a year’s electricity consumption statistics.

Solar installation companies have special purpose software to assist in designing and costing systems, which does day-by-day simulation of usage taking all the varying electricity prices and average weather patterns into account. They will do a site survey to ensure that the location is suitable, measure the angle of the sun and the number of hours of daylight, and so on. They should also help with the process of applying for any utility rebates or tax incentives.

The solar company came back to me with a proposed system that would hopefully get us to “net zero”, given Austin’s current electricity tariffs and our current typical usage. The projected break-even point would be after 7 years — that is, by year 8 we should have saved enough on electricity to pay for our out-of-pocket costs of the system. The out-of-pocket costs would be comfortably under $10,000, which seemed reasonable for a home improvement project.

The key word in the previous paragraph is “should”. No matter how much research you do, you should remember that residential solar power is in its early stages. Maybe next year someone will come up with a radically new solar cell design that’s twice as efficient, or a working nuclear fusion plant. Perhaps panel lifespan will be lower than expected, or the utilities will decide they don’t want solar customers any more.

I was an early adopter for hybrid cars. It didn’t make compelling financial sense at the time, but I wanted to help popularize the technology and vote with my wallet. I’m viewing solar power the same way. I’m pretty sure I won’t lose out overall, but I’m also comfortable with the idea that it may not be a massive money saver.

There were quite a few scare stories about maintenance costs of hybrid cars when we bought our Prius. I’m pleased to say that they have turned out to be totally false. In nearly 90,000 miles over 10 years, the only unscheduled failure has been one coolant pump. A solar system has a lot fewer moving parts than a car, and is subject to a lot less mechanical stress, so I have a sneaking suspicion that the scare stories about maintenance costs for photovoltaic solar systems are similarly baseless. Yes, we’ll need to make sure we’re more diligent about getting the tree trimmed each spring so it doesn’t shade the panels, but I don’t believe the manufacturers would be offering 20 year warranties on all the solar system hardware if they expected it to fail in that time period.

I’ve seen articles about the importance of cleaning solar panels. I asked about this, and was told that since our panels are at a fairly steep angle and high up, chances are the rain would do a pretty good job of keeping them clean. If you need to install panels at a flatter angle or near ground level, you might need to hose them down each spring. Still, not a major cost for a home-sized system.

Solar panels
Powerhouse Museum via Compfight

There are some design choices for solar systems that are a matter of opinion with no clear economic case one way or the other. To understand those tradeoffs, it’s necessary to understand how a grid-tied home solar photovoltaic system works.

The basic arrangement is that each panel consists of a number of individual solar cells wired up together. Cells are wired in series to multiply their output voltage up to some reasonable level, and in parallel to multiply their current output. Each panel thus ends up with a power rating; say, 230W at 31V.

The panels are then wired together to make the complete system. At this point, the same techniques of series and parallel wiring are used to raise the voltage and power output further. In the case of my system, the final voltage is around 200V.

There’s a snag, though. That’s DC voltage. Before it can be fed into the house AC circuits, it needs to be converted to AC. This is done by a device called a power inverter. The inverter steps the voltage down and converts from DC to AC, then feeds the result into the metering systems.


Mike Carter

There’s another way to do it, though. Some solar systems now have a small inverter (known as a microinverter) on each panel, rather than one big inverter. One advantage of this technique is that you can then ask each inverter to tell you the state of its solar panel. With one big inverter, you can get a “Christmas lights” problem, where one of the panels fails but there’s no easy way to find out which one. The downside of microinverters, of course, is that having one on each panel is more expensive than one big inverter. This is particularly an issue if you want your microinverters to have a warranted lifespan as long as your solar panels, which you probably do unless you fancy replacing them all at some point.

Ultimately, the benefits of micro-inverters are bigger the more panels you have which need monitoring. For a dozen panels, the difficulty of tracking down a fault (which should be very rare anyway) probably isn’t an issue. If you’re a commercial installation with 50 panels, on the other hand, microinverters are probably the way to go. The solar company I worked with ended up designing our system with a single inverter that connects to two independent strands. This hybrid approach seems reasonable for a small-scale installation.

Another design choice is polycrystalline vs monocrystalline panels. (There are other solar cell technologies, but those are the two leaders at this time.) Monocrystalline cells are the dark, almost black ones that have a smooth, uniform appearance. Polycrystalline cells are a lighter blue color. Generally monocrystalline cells are more expensive, but more efficient and have a longer lifespan and better heat tolerance. Polycrystalline cells are getting better, though, and ultimately it’s a tradeoff of cost versus lifespan and area required. One non-difference is that both kinds of panel can stand up to a hail bombardment, which is a significant concern in Texas.


Naturewise

For the record, our system has monocrystalline solar panels. The original design was going to use BenQ cells, but after measuring the roof and looking at the data the recommendation was to increase system capacity slightly and switch to Hanwha Q-Cells (from their Q-Peak range). The inverter is a Sunny Boy 4000 from SMA. It has a peak efficiency of 97% and is made in the USA; SMA itself is a German company. Q-Cells are also German, their parent company Hanwha is Korean.

The terrible cost of free doughnuts

During the Second World War, the Red Cross provided free doughnuts to US troops fighting in Europe. Then they started charging 2¢ for a doughnut and coffee. 70 years later, veterans still haven’t gotten over that betrayal. NPR Planet Money explain why: basically, it comes down to the special nature of “free”, and the fact that the change altered the perceived relationship between the Red Cross and the veterans, from “friendly charity” to “money-making business”.

It seems to me that this issue of categorical change vs price change is part of the reason why I am so vehemently against the TSA’s policy of charging for security screening that hasn’t been deliberately slowed down. Getting security screened as fast as possible was a free service, and it’s not part of what I’m paying for when I book a plane flight. Being charged extra to not get deliberately and unnecessarily delayed at the airport is a categorical change, right? Meanwhile, while I don’t like baggage check charges, they don’t fill me with the kind of furious rage that the TSA does.

There’s a lesson for Internet companies too. Facebook’s attempt to charge for sending a message to someone you don’t know? Their attempt to charge for having your update be seen by all your friends? Categorical changes, very unpopular.

The future of Bitcoin

[One of a series of articles.]

So far I’ve talked about Bitcoin myths and how Bitcoin is in a speculative bubble. In this article I’m going to discuss Bitcoin’s future.

My view regarding the long term future of Bitcoin is fairly simple: I maintain that it doesn’t have one. And not just because there are pre-existing superior digital cash systems. No, the fact that better digital cash systems were invented in the 90s points at the fact that there are specific reasons why they haven’t taken off, and that it’s not to do with technology.

Put simply, the adoption of digital cash is limited because of economic problems, not technical ones.

The first problem is that digital cash isn’t in the consumer’s interest. It’s more complicated to deal with than a credit card, and doesn’t give you any of the associated benefits (cash back, reward points, extended warranties, ability to get the card company to make a seller take back a faulty item, and so on). In addition, in the case of Bitcoin, you’re out of luck if your money gets stolen — which is currently becoming increasingly likely. DigiCash was a lot more resistant to theft and fraud, but it still didn’t seem a compelling value proposition to consumers, even though it was launched in the middle of the .com boom.

The second problem is that digital cash isn’t in the store’s interest. They love being able to track purchasers by credit card number or check account number. Many stores go to the extra step of having loyalty cards, so they can track you even if you pay with cash. So the anonymity of digital cash is a minus to retailers. There’s also a technological issue — US stores don’t even have NFC deployed widely yet, as I’ve discovered by trying to use my phone to pay for stuff digitally. (Here in Austin, Texas, H-E-B has NFC terminals, but they deliberately turn the digital payment functionality off!)

Problem number three: Digital cash isn’t in the bank’s interest, again because of the anonymity. Your credit card account is a rich data source the bank mines in order to sell you add-on services or other bank products. Banks also make a lot of money from the transaction fees that people would be able to sidestep if there was a global digital currency.

Finally, digital cash isn’t in any government’s interest, because it makes it a hell of a lot harder to control the economy, control money flow, catch drug dealers, and audit people’s tax returns. No government is ever going to voluntarily recognize secure anonymous digital cash.

I haven’t seen any proposals that would even begin to deal with (or sidestep) all the economic disincentives I’ve just listed, for Bitcoin or any other cryptocurrency. I suspect that’s because the people who are into Bitcoin are computer programmers, and computer programmers tend to blindly see everything as a technological problem, and dismiss economics as a pseudo-science that has nothing to say about the real world.

Notice that the situations where you see Bitcoin actually being used for commerce are exactly those where it has economic benefits — situations where everyone involved has a compelling desire to want everything to remain anonymous. You know, sale of illegal drugs, procurement of assassinations, that kind of thing.

So all in all — and speaking as someone who met David Chaum and had brochures about DigiCash and loved the idea and has therefore seen it all before — I rate the chances of Bitcoin going mainstream as zero. It’s a fad and a speculative bubble. Of course, you can still make some money off of it, but don’t stay in it for the long haul expecting to gain.

Next: Other dissenting voices…

Bitcoin and quantitative easing

[Part of a series of articles]

It’s frequently stated that Bitcoin is finite, that there can be exactly 21 million Bitcoins, and that’s it. Bitcoin fans like to point this out to distinguish it from normal currency, where the central bank can just print more money.

The claim that Bitcoin is finite is technically correct, which as we all know is the best kind of correct. However, it’s seriously misleading, to the point of not really being true at all. To understand why, you need to have been following the Bitcoin saga for a while.

In March 2013, Bitcoin forked. It turned out that versions 0.7 and 0.8 of the software behaved slightly differently, and after a particular block of data the ‘block chain’ (the registry of all bitcoin transactions) split into two: the 0.7 block chain, and the 0.8 block chain.

This was resolved by the community deciding that everyone who had mined or traded bitcoins on the 0.8 chain was out of luck — nobody was going to accept their coins as valid. So not only did Bitcoin turn out to be less finite than people had hoped, it also turned out that people could have perfectly valid Bitcoins taken away from them by fiat.

Of course, taking people’s Bitcoins wasn’t the only option. The Bitcoin protocols could have been extended to include a block chain indicator, which would specify whether each coin was a 0.8 chain coin, or a 0.7 chain coin. But that would have suddenly dramatically increased the number of Bitcoins by over 50%. The Bitcoin community values the idea of there being only 21 million of them, so collectively it decided that a better option was to effectively seize and invalidate the new coins of those who had adopted the new version of the software.

Since Bitcoin is just software, the block chain indicator idea could be extended further. You could start a new block chain each year, just like most national currencies have a year of issue on their coins and banknotes. Each Bitcoin would be labeled with the year of issue, so the software would know which block chain to transact with.

And effectively, this has happened. Except instead of a year of issue or some other numeric version fork of the block chain, people have come up with a new name for their Bitcoin fork, and usually made a few tweaks to the algorithms. Anyone can do it. Already there are dozens of them — Litecoin, PPcoin, Megacoin, Betacoin, AnonCoin, PhenixCoin, and so on. There’s even a satirical Dogecoin. And the Dogecoin community has decided to allow inflation.

Dogecoin

Just as Bitcoin 0.7 only remained limited to 21 million coins as long as everyone agreed not to use the 0.8 block chain, so Bitcoin in general is only immune from currency expansion as long as everyone agrees not to use any of the forked alternative crypto currencies. So far, it doesn’t look like that will be the case — everyone’s hoping that by picking a new fork, they can get in on the ground floor of a bubble just like the Bitcoin one.

Next: The future of Bitcoin…

Perverse incentives

It’s commonly believed by right-wingers that US public radio has a left-wing bias. If you want to hear an example of a program that disproves the assertion, I suggest This American Life episode 490: Trends With Benefits. It has turned out to be somewhat controversial, though TAL are standing by the facts they reported, unlike with the Mike Daisey piece.

As most people know, the US economy has a deficit problem. Part of the reason for the problem is crazy defense spending, yes; but there are two other chunks of the federal government’s budget that are about the same size as the military budget: Social Security and Medicare.

The This American Life show gives some insight into why the Social Security budget is so huge, starting from one simple statistic: That in Hale County Alabama, a quarter of the working-age adults are receiving disability payments, because they have been declared permanently unable to work.

What counts as a permanent disability? Well, the list includes:

  • High blood pressure
  • Diabetes
  • Chronic back pain
  • Sleep apnea

Now, I’ve known people with diabetes who inject themselves with insulin and go to work every day. I have chronic back pain, but I’ve only missed a couple of days of work in the last few years. My father-in-law had high blood pressure for the last couple of decades of his working life. None of these things should be permanent disabilities that render you unable to work.

At this point, the traditional right-wing media narrative is “Look at all these scrounging parasites”. But as is generally the case, it’s a bit more complicated than that.

If you live in Hale County Alabama, and you aren’t highly qualified — which is the case for the disability recipients — then there really aren’t any jobs you can do. Even if you had the time and money to get the necessary education, there aren’t any white collar jobs in rural Alabama. The only jobs on offer involve lifting heavy objects, standing up all day, or other physical tasks. So if you have high blood pressure, well, you probably really aren’t fit to work any of those jobs. Or at least, no company will employ you, in case you drop dead.

So the problem is bad. But then things get farcical. It turns out that when someone applies for disability and is turned down, they can file for a hearing. At that hearing, they can get a lawyer to present the case for why they can’t work. On the other side, the government has … Nobody. There’s nobody whose job it is to explain why the petitioner is actually able to work, or point out jobs they could be doing. So most of the time, the lawyer will get them approved for disability. At that point, the petitioner is liable for back-payment of disability from the time they first applied. The lawyer gets a portion of that back-payment. In fact, the federal government pays the lawyer directly. So if you do a Google search for (say) ‘disability’, you’ll find lots of lawyers ready and eager to take even a pretty marginal case, on a contingency benefit.

But that’s not the worst part. See, once you get disability payments, you no longer qualify for unemployment benefit. And whereas unemployment benefits are paid by the individual states, the Social Security disability system is a Federal program. So get this: States actively hire companies to go through their lists of the unemployed, spot any who might be persuaded to qualify for disability, and encourage and assist them with the process.

But even that’s not the craziest part. It turns out that children can qualify for disability payments too, for things like learning disabilities. Dyslexic? You might be able to get your parents a nice Social Security check each month. Of course, in time your parents will come to rely on that check, and they’ll only get the check as long as you continue to do badly in school. Once you reach adulthood, well, you can either stay on disability — or you can try and get a job which might not work out, and in the mean time lose your steady disability benefits.

Because the disabled don’t get unemployment benefits, government excludes them from the unemployment figures. Which might seem reasonable enough, but unfortunately right now there are more people being added to the disability benefits list each month than there are new jobs being created. So the actual percentage of people who could be doing some sort of job, but are not doing so, could be way bigger than the official unemployment figure of 7.7%. And we’re still sliding backwards.

So, what’s the solution? Well, there’s one radical theory that has been proposed for decades, under titles like “Universal Living Wage”, “Subsistence Allowance” or “Basic Income Guarantee”. It’s pretty simple to understand: You put together a minimal subsistence payment, enough to cover basic food, shelter and other bare essentials — and you give it to absolutely everyone. You filed a tax return? Here’s your subsistence payment.

This immediately eliminates a ton of bureaucracy. No application process, no means testing, no medical tests, no appeals process, no judges needed, very little chance of fraud. It also eliminates the problem of potentially losing the steady income you need to live by risking taking a job. Hence, hopefully nobody will ever feel like it isn’t worth taking on a job. Even a part-time, temporary or seasonal job will boost your income without the risk of reducing your subsistence payment.

Obviously, there are naysayers. The most common objection is that if nobody is absolutely forced to work in order to afford a roof over their heads and food to eat, why, surely nobody will work at all? Personally, I find that argument incredibly patronizing and offensive. Nobody who has earned more than a couple of million dollars needs to work, but there are plenty of millionaires who do carry on working. And we’re not talking about a luxury subsistence here; if you want a mobile phone, TV, PlayStation or car then you’re going to have to go out and get a job to pay for it, like the rest of us. If you want your own space, rather than a run-down government-run shared apartment space, you’re going to have to work hard.

Still, many doubt that such a scheme could ever work. Except it turns out that one country tried it. From 1974 to 1979, Dauphin Manitoba ran an experimental basic income program known as “Mincome”.

Mysteriously, the Canadian government locked away all the data that was collected, and prevented it from being analyzed. The astute reader can probably guess why. When academics finally managed to get access to the documents in 2009 and analyze the results, it turned out that the whole scheme worked pretty well. Rather than everyone sitting around idle on welfare, the only people who worked less were mothers and teenagers. Graduation rates for teenagers, not surprisingly, went up. Hospital visits dropped. There were fewer work-related injuries.

So, how about it, America? The current system is clearly a disaster, and an inefficient one at that.