Another war tragedy -- the solar opportunity in Iraq


While I've written before about the trouble in making solar competitive with grid power, this is not true when the grid is being blown up by geurilla fighters on a regular basis. Over the past couple of years, Bechtel has been paid over 2 billion dollars, mostly to try to rebuild the Iraq electrical infrastructure. Perhaps it's not their fault that power is only on in Bagdadh for 2 hours a day after these billions have been spent -- but their might have been a better way.

Imagine if that billion had been directed at building a solar power system, with a lower-power grid for night power. A billion would have provided major stimulus to the solar industry, of course, and helped the companies that are working at making PV cost-effective. But it also would have generated a power infrastructure that was much harder to destroy in a civil war. Yes, they might take down sections of the grid, but these would only have been there for night and brownout power. Without them, people would still have had more power. And not just during the day. Mini "neighbourhood grid" systems could allow small areas to have backup diesel generators. Not quite as efficient as the big generators but much more difficult to take down. The "value" targets would still see their local panels and generators under attack, but that's the way of it.

It seems odd to think of this in a country with so much oil. But doing this would have also had a major effect on greenhouse gas emissions. Putting solar into Iraq would have made the US responsible for major emission cuts. Cutting emissions there so we don't have to cut them here.

Something to think about next time your country goes and destroys a foreign country's power grid and then works to rebuild it. (Of course, ideally that's never.)


This has been the wrong time. For the past 18 months and next 18 months (or more) the PV industry is capacity limited due to reaching the capacity limit for silicon production. It used to be that it could live off the excess capacity from semiconductor manufacturing. Semiconductor growth was enough to keep up with PV needs. Now PV needs its own dedicated silicon production, which is under construction, and factories take years to build.

PV has also been growing just fine. It's been a 20-30% CAGR for 20 years now. That may be small compared to semiconductors, but by industrial standards it is a superb growth rate. The demand has exceeded supply for over a year and prices have been increasing significantly. PV production is backlogged with orders out two years on some lines. Increasing demand is not useful for a while. I'm aware of the thick film technologies, but those are still very limited production and not suitable for many PV purposes. They also face the factory capacity and construction time limits at the moment.

When more factories are on line and production is catching up with orders, then it may be time to consider subsidies.

The most effective subsidies have been those using the European approach. There is no subsidy for construction at all. (Bechtel would not get a penny.) There is a subsidy for power generated. This eliminates the massive fraud and corruption that has been found in the construction subsidies. If you subsidize construction, you get corruption, hugely excessive construction costs, bad siting, inadequate maintenance, and abandoned operations. (See the US history with solar hot water subsidies for a domestic example.) If you subsidize based on actual power generated, it is much harder to have significant loss to corruption. The builder is motivated to build reliably and cost effectively, because they only get their profits from long term power generation.

The steady success for wind generation in the US and Europe is a result of subsidizing generation and not construction.

As for Iraq, there are many more complex factors at work. For example, a few bullets can easily destroy a PV panel. PV panels are no more likely to survive in Iraq than grid power. The bulk of Iraqi power in the post war years was from the myriad individual home generators. These just need gasoline/diesel. Now that the entire economy is in ruins it is also hard to get fuel. I doubt that the PV systems would have survived as long.

I had always heard the reason most of the wind turbines here in the bay area have stopped is they got subsidies to build them but not to maintan them, so when they fail, there is no financial justification to repair. I am glad if that's changed.

You can destroy any power infrastructure, but destroying distributed infrastructure is a lot harder, and the users of it live near it and protect it, unlike power lines out in the middle of rural areas. If solar capacity could not be made available with billions to spend, other distributed infrastructure would have made more sense, even if it was distributed diesel or natural gas generators and mini-grids. They are less efficient, but make some of that back with the lower transmission distance. And we're getting better at making this sort of stuff.

I'd overlooked that example of subsidy corruption.

The present subsidies in Europe tend to be of the form of 10-yr commitment to purchase power with a few-cent premium per kwh. For some sources, the guarantee to purchase is more important than the premium. A good wind site is already cost competitive if they can avoid the costs of marketing and selling to the grid. Even in the US you find a few turbines going in without needing subsidies, e.g., Hull.

The US subsidy comes and goes depending on the annual whims of Congress, but for a long time now wind subsidies have been 1-2 cents per kwh delivered. It has made a big change for wind producers. Probably the first change was the turbine manufacturers, who started designing for production efficiency and maintenance cost instead of construction cost. The subsidy change happened first in Europe, so you see the European manufacturers dominating wind.

Many of the old California turbines were never worth building. They were in windy locations, but not properly sited. When the subsidy is for construction why waste money on local siting studies. Some are being replaced. The new turbines are being sited based on the detailed wind data from the old turbines. They found that some were never worth building and will not be replaced. Those will be torn down for scrap.

Another microscale that is moving into the northern areas is co-gen home heating. It does not make good financial sense, but for $20,000 you can put in a high efficiency home heating system that is also an electricity generator. Another 10 years and the turbine designs may be buildable at a reasonable cost. At present, these sell to enthusiasts who don't mind the cost and to those few unusual customers where the finances work out.

I should have included the cogen reference. It is (of course) Honda that is making these units.

Especially if the waste heat is used to heat the house and water. Do they make one that also drives the AC compressor directly, since I presume that's a fair bit more efficient than generating electricity and then using an electric air conditioner?

Of course, if you need a methane infrastructure to power local generators, then geurillas can attack the gas pipes to shut you down. With solar, and even with diesel they have to attack in a lot more places. Places like Iraq need power with no single points of attack. The main point of my blog post, however, was to wonder if that special need might have helped finance a lot of interesting power technologies that are not yet economical otherwise.

I'm not sure about AC. In automobiles the MIT Auto research center found electrical to have superior efficiency. But they have the added issue of managing optimal engine speed and load versus optimal compressor, even when part of a hybrid car.

Honda chose natural gas to take advantage of existing heating infrastructures and avoid lots of pollution control issues. They could have used heating oil (which is diesel with different additives). But then they would have had lots of pollution control features added to the cost.

What is happening on the commercial and industrial scale are:
1. All sorts of interesting heat recovery combined with AC. The absorption boiler air conditioner is well adapted to use with low temperature waste heat. They are common in commercial locations when waste heat is already available.
2. Combination hot water/ AC systems are sold. These are electrical heat pumps, much like those that commonly are use for whole house air conditioning. But instead of having fans and radiator outside to dispose of heat, they heat the hot water tank. The equipment sizes, costs, and operations do not make sense for household use. These are found at commercial sites that need both hot water and AC.

The special need that has funded a lot of research is the military and field operations in general. They need highly mobile power sources, and are much less sensitive to cost. There are lots of interesting mobile and semi-mobile power applications. The biggest areas at the moment are power efficiency (to reduce the demand), batteries (to improve power/weight, storage/weight, and reduce volume), and mobile power sources. Rugged solar panels are becoming common. They are a bit heavy for back pack, but horseback sizes can do a lot.

The most common operational model is a battery charger at camp to recharge the numerous little batteries in radios, GPS, lights, etc. Then the soldiers just need to carry the batteries in the field. The whole thing needs to be easily packable whenever the camp is relocated.

For covert operations you cannot use noisy things like engines as a power source and you cannot use noisy things like helicopters to move camp. All the supplies need to be carried by people and animals (horse, mule, etc.).

I've got some experience with both on and off grid generation, and found that it's immensely easier to work with any grid at all - even having 5 or 6 small generators linked made the system much more robust (some generators will stall when a big load starts, motors often draw 2x or 3x rated load on startup and windmills, microhydro and IC generators do not like that at all). If you can link to the grid that problem goes away, and even having several small units linked worked well in one trial we did (using a petrol powered generator to fill in on top of a pair of grid-interactive inverters running off solar+wind). I suspect bigger windmills have enough rotational inertia to obviate most of that problem, but the sub-kilowatt units we had would just stop.

So yeah, mini-grids would be a good technical solution.

I wonder if in Iraq that would just amount to building targets - my understanding is that a fair bit of the civil war consists of brief raids not much more than drive-by shootings, and PV/wind on the roof would be an obvious target. Although thinking about it, I've never tried shooting a windgen but there's not much to them so they might be quite hard to disable without a ridiculous rate of fire. Hmm. Worth trying :)

It is a concern, but people protect their neighbourhoods more than they protect random power conduits. However, if they started shooting solar you could start protecting it, at some cost of power yield, but putting it in back yards rather than roofs, or surrounded by a metal or concrete wall so there isn't a line of sight to the panel, and putting a mesh screen over it (which blocks some of the light but also stops them from lobbing rocks.) If they are going to lob bombs, not a lot you can do about that.

But clearly external wires aren't a workable solution, even underground -- which are harder to take out but more expensive to install or fix. You still have them, still have a grid, but it doesn't carry as much power and if they take it out, you just lose some, not all of your power.

Some forms of solar panels, like the flexible unisolar one on my roof, can probably take a bullet without being destroyed unless you hit the right place.

Neighbourhood grids can run in back yards where they are harder to sabotage. You can attack decentralized infrastructure but it's harder. A place like Iraq would be great for wireless mesh networks, mini-grids and the like. Indeed, who needs single points of failure anywhere? It's more efficient, but we have better tech nowadays.

Add new comment