Learning from Others' Mistakes

Here is the summary of an October 2009 report from the Rheinisch-Westfälisches Institut für Wirtschaftsforschung titled: "Economic impacts from the promotion of renewable energies: The German experience"


Although renewable energies have a potentially beneficial role to play as part of Germany’s energy portfolio, the commonly advanced argument that renewables confer a double dividend or “win-win solution” in the form of environmental stewardship and economic prosperity is disingenuous. In this article, we argue that Germany’s principal mechanism of supporting renewable technologies through feed-in tariffs, in fact, imposes high costs without any of the alleged positive impacts on emissions reductions, employment, energy security, or technological innovation.

First, as a consequence of the prevailing coexistence of the Renewable Energy Sources Act (EEG) and the EU Emissions Trading Scheme (ETS), the increased use of renewable energy technologies triggered by the EEG does not imply any additional emission reductions beyond those already achieved by ETS alone. This is in line with Morthorst, who analyzes the promotion of renewable energy usage by alternative instruments using a three-country model. This study’s results suggest that renewable support schemes are questionable climate policy instruments in the presence of the ETS.

Second, numerous empirical studies have consistently shown the net employment balance to be zero or even negative in the long run, a consequence of the high opportunity cost of supporting renewable energy technologies. Indeed, it is most likely that whatever jobs are created by renewable energy promotion would vanish as soon as government support is terminated, leaving only Germany’s export sector to benefit from the possible continuation of renewables support in other countries such as the US. Third, rather than promoting energy security, the need for backup power from fossil fuels means that renewables increase Germany’s dependence on gas imports, most of which come from Russia. And finally, the system of feed-in tariffs stifles competition among renewable energy producers and creates perverse incentives to lock into existing technologies. Economic impacts from the promotion of renewable energies

Hence, although Germany’s promotion of renewable energies is commonly portrayed in the media as setting a “shining example in providing a harvest for the world” (The Guardian 2007), we would instead regard the country’s experience as a cautionary tale of massively expensive environmental and energy policy that is devoid of economic and environmental benefits. As other European governments emulate Germany by ramping up their promotion of renewables, policy makers should scrutinize the logic of supporting energy sources that cannot compete on the market in the absence of government assistance. Such scrutiny is also warranted in
the US, where there are currently nearly 400 federal and state programs in place that provide financial incentives for renewable energy.

History clearly shows that governments have an abysmal record of selecting economically productive projects through such programs. Nevertheless, government intervention can serve to support renewable energy technologies through other mechanisms that harness market incentives or correct for market failures. The European Trading Scheme, under which emissions certificates are traded, is one obvious example. Another is funding for research and development (R&D), which may compensate for underinvestment from the private sector owing to positive externalities. In the early stages of development of non-competitive technologies, for example, it appears to be more cost-effective to invest in R&D to achieve competitiveness, rather than to promote their large-scale production.

In its country report on Germany’s energy policy, the International Energy Agency recommends considering ‘‘policies other than the very high feed-in tariffs to promote solar photovoltaics.’’ This recommendation is based on the grounds that ‘‘the government should always keep cost-effectiveness as a critical component when deciding between policies and measures.’’ Consequently, the IEA proposes policy instruments favouring research and development. Lesser and Su concur with this viewpoint: ‘‘Technologies that are theoretically promising, but unlikely to be competitive for many years, may be best addressed under other policies, such as publicly funded R&D.’’ This reasoning is particularly relevant for solar cells, whose technological efficiency is widely known to be modest and, hence, should be first increased substantially via R&D.

Instead of a policy instrument that aims at pushing technological improvements, however, Germany’s support scheme of renewable energy technologies resembles traditional active labour market programs, which have been demonstrated in the literature to be counterproductive. It bears particular noting that the long shadows of this economic support will last for another two decades even if the EEG were to be abolished immediately.


You're an ER doctor and three patients enter the waiting room at once. One is having a heart attack, one has a broken arm, and one has the flu. Whom do you treat first?

If you think it's the heart attack patient, I'm with you.

It's time we face the fact that resources are limited and we must triage our energy problems in a logical order.

People talk about competing priorities in addressing the national energy situation; clean air, global warming, and dependence on imported oil. It's useful to look squarely at this issue first, and there appears to be an elephant in the room.

If one compares the total known reserves of oil in the U.S. to our rate of use, we have about 20 billion barrels of known reserves (source EIA), and we use about 20 million barrels a day. So, we have about 1000 days of oil reserves (2.7 years) if we were forced to supply all our daily need with only our current proven reserves. The same calculation for natural gas yields about 10 years of known reserves (BP Review of World Energy 2009) but we have a whopping 234 years of coal!

We have lots of coal and not a lot of oil or gas. Now consider what we do with coal, oil, and gas. In very general terms, we use oil to run our transportation system, and we use gas and coal to make electricity. We are critically low on the fuel needed to operate our planes, trains, and automobiles, but we have lots of fuel to make electricity.

So what problem is more important to work on first, electricity or transportation? I vote for the transportation/oil issue. I trust you agree.

So why are we spending a lot of money in subsidies to promote the use of expensive solar power to make electricity? I'm not sure. It seems to me that electricity is not the problem and if it were, solar is not the answer (See post titled "Grid Tied Photovoltaics") We have enough coal, nuclear energy, and hydro to make electricity for centuries to come. Our electrical problem is like the patient with the flu. Our pending transportation trouble is like the guy having the heart attack. I'm concerned about the guy with the flu and the guy with the broken arm, but let's treat the problems in a logical order. Transportation first, then electricity. Am I not concerned about CO2 from burning coal? Yes, but we have a bigger near term problem that left unaddressed will cause us much more pain and suffering. If our oil situation becomes desperate, we will collectively cast off any concerns for the environment in favor of simple survival. So let's steer clear of those rocks.

As our oil supply depletes, oil prices will rise, potentially in a disruptive way. Policies that anticipate the pending oil/transportation paradigm shift would be wise. But subsidies to make electricity with PV are treating the wrong patient with the wrong medicine.

Cost of Various Electricity Options

Click on chart to enlarge

This analysis has a few important caveats. First, a CO2 penalty was imposed on the cost of the coal and gas technologies. The penalty was applied as a 3-percentage point increase in the cost of capital. The report stated this was equivalent to a $15/ton cost for CO2. Why not just add the CO2 penalty directly to the fuel cost? That would be easy to grasp and analyze. Why increase the cost of capital?

Second, the fact that PV and wind energy is worth less due to its intermittency is NOT captured in the cost data.

In summary, solar PV is shown to cost about 3-4 times more than traditional generation methods, but without the CO2 penalty on coal and with a correction for the intermittency of PV energy, the real price of PV is about 10 times more than traditional alternatives and several times higher than other renewable alternatives.

A more intuitive and fair comparison, would include externalities clearly and separately. The structure of the analysis, makes it difficult to apply the lower value of intermittent power. Perhaps a simple $/kWhr would be a better approach, if clarity on the cost of renewables is an objective.

PV Vs. CFL Bulbs

Someone is given $10,000 to spend on either a grid-tied PV system or on compact fluorescent lights. Their goal is to save energy, reduce CO2 output, and delay the building of more power plants. Which option is better, PV or CFL?

Here are a few assumptions. The CFL is replacing a 75W incandescent light which runs, on average, 4 hours a day. Also assume that the CFL draws 15 watts, 1/5 the power of the incandescent bulb for the same lumens. Finally, assume that you install the PV system in sunny southern Arizona with a net production output of 1,650 kWh/y for each kW of installed rated power.
With the $10,000 one can buy a 2kW grid-tied PV system ($5/w), which will produce about 3,300 kWh/y of usable power. If the PV energy is offsetting energy from a coal plant (like in Tucson), about 3.3 tons of CO2 will be saved a year. But it will not help reduce peak load much since the utility can't count on its availability with the level of certainty they demand. (see post "Grid Tied PV and Peak Demand").

If one chooses the CFL, then the $10,000 will buy about 5000 lamps. Each lamp saves 60W for 4 hours a day (75W - 15W). 60W x 4 hr x 365 days x 5000 lamps = 438 MWh/y. This reduced CO2 by about 438 tons a year. But also very important, efficiency gains reduce peak load requirements and can delay or avoid the construction of a power plant.

So the CFL options saves 132X more energy, reduced 132X more CO2 per dollar, and easies the burden on the utility grid and generation loads.

So why don't utilities give away free CFL bulbs to anyone that asks rather than subsidizing grid-tied PV?

The point is that there are many things which can and should be done in the areas of energy efficiency before one spends any significant money promoting grid-tied PV. But it is not happening and the reason is a mystery.

Grid Tied Photovoltaics Economics

The cost of a typical grid-tied PV system is about $5/watt installed (for a large system). Such a system located in sunny southern Arizona will actually generate power for about 1650 net sun-hours a year. A 1kW system will produce about 1,650 kWhr/year. This system will cost about $5,000 to install. So what is the value of the electricity it produces?

If 100% of the power could be used real-time by the end user, then one might say it’s worth the real time retail rate (See post titled “The town of Coalville” to appreciate the pitfalls of such an assumption). In my town I pay about 11 cents/kWhr. So making 1650 kWhr * $0.11 = $182. This has a simple payback of 27 years. But what if I can't use the power during the day when it is being produced and I have to sell it back to the utility?

In Arizona and in many states there are 'Net Metering' laws which require the utility to buy power back from the end user at full retail rates. This is unfair to the utility and is a subsidy for the PV user (which we all pay in the form of higher utility rates and fees). It is extremely important to have a firm grasp on the concept of "no free lunch." It is very easy to confuse real economic returns with artificial returns in a heavily subsidized environment.

Forcing the utility to buy back power at retail rates is unfair because they are required to provide power 24/7, install and maintain generation equipment and transmission lines and provide service and billing. The actual value to the utility of this intermittent power (PV) is much closer to the fuel avoidance cost.
For reasons explained in the Coalville post I think the value of homeowner produced PV is close to the fuel avoidance cost which is about 2.7 cents/kWh in my area. So the real payback of PV is about 111 years.

But what if PV costs fell substantially? Well it has, especially in the last 18 months. The $5/watt figure is based on the new current lower prices (Mar 2010). But what if they fall more? How cheap do they need to be to make systems break even without subsidies. If PV panels were available free, the cost to install the panels, including the inverter, wire, breakers, enclosures, conduit, and panel mounting structure would be about $2.50/watt. So the free 1 kW panels would cost $2500 to install and connect to the grid, and will produce about $45 worth of power a year, for a 56 year simple payback.

Grid Tied PV and Peak Demand

Does installing grid-tied PV really reduce the power company's peak generating or transmission requirements?

I often hear that PV can be used to delay or even avoid building a new power station.

In my town we have our highest annual electrical loads in the mid to late summer. We also have some great afternoon thunderstorms. We call them monsoons. In the last 3 of 5 years, at the moment of peak annual load, our town was covered by clouds. You may think this is impossible. If there are clouds, then the electrical loads (air conditioning) should be lower. I admit this sounds a bit bazaar.

But the real world is often stranger than fiction. In our case, we were having a typical hot sunny day. Then in a 10 minute period of time, our town which is about 10 miles by 10 miles square was overshadowed by one of those great monsoon rain clouds. The interesting part is that the electric demand of the town continued to rise for 15 more minutes. It seems that there is a delay between the time the building get shadowed and when the ACs respond to the lower load. Or maybe everyone just started turning on their lights. It does get dark when these clouds roll in.

So this actually happened 3 of the last 5 years! For the power company to depend on a power source like PV, their reliability guidelines only allow 1 such event every 10 years. So this means they must build in the generation capacity for the case that the PV is cloud covered when the annual peak demand is occurring. It happens too often for them to ignore it. The only relief the PV offers is that even under dark cloud cover it can still produce 10% of its rated output.

One last point. Your town might have much different weather patterns than mine and distributed PV might have a better chance to be illuminate at the moment of annual peak power demand 9 of 10 years or whatever the utility demands for reliable service. The drawback is that peak demand occurs at 4:30 to 5:00 pm where I live and it might be the same for you. This is several hours past solar noon. PV output is down 30-50% by then. You can face the panel more southwest to try to synchronize the peak PV output with peak energy demand, but you then lose in total energy output and burn more fossil fuels annually. But all this can be a fun optimization problem.

So in my town, PV only marginally helps reduce peak generation requirements for the utility and if we grow and need to add more generation capacity, adding PV will not delay this much.

Chasing your Tail

In college, I made an important mistake and learned a valuable lesson. I was working on a solar thermal cooling system and made a cost/benefit analysis. I determined this technology would break even if the price of oil exceeded $50 per barrel. Adjusting for the CPI, this is $121/barrel today. Well, oil exceeded that recently, but this technology did not come to market. My error was that as oil becomes more expensive, so do solar technologies. There's a lot of energy embedded in the cost of things. When energy prices rise, so do the cost of these items. There is a multiplier in each product, such that a 1% increase in energy prices results in a fraction of 1% increase in cost. Some things are energy intensive (like food or aluminum) and some things are not. Although, at the moment, I have trouble naming anything that is not.

If something has an energy cost factor of 25%, it means that a 1% increase in energy prices causes a 0.25% increase in price. Imagine an alternative energy technology that can produce power for 20 cents/kWh at today's energy cost of 10 cents/kWh. Also assume the technology has an embedded energy cost factor of 25%. So when energy prices rise to 20 cents/kWh, the cost of the energy produced by the alternative supply also increases to 22.5 cents/kWh. So it's not quite at break even yet.

In general, the price of energy might need to rise more than one might think before alternative technologies break even, because these technologies have an energy content that increases in cost. If the energy content is too high, then you can chase your tail, so to speak, and not reach break even for a very long.

If you know any good papers about the impact of energy prices on food prices, please comment. Thanks.