The Wrong Way to Sell Wind and Solar

A reader sent me this article on renewables by Tom Randall at Bloomberg.  I would like to spend more time thinking about it, but here are a few thoughts. [Ed:  sorry, totally forgot the link. duh.]

First, I would be thrilled if things like wind and solar can actually become cheaper, without government subsidies, than current fossil fuels.  I have high hopes for solar and am skeptical about wind, but leave that aside.

Second, I think he is selling renewables the wrong way, and is in fact trumpeting something as a good thing that really is not so good.  His argument is that the decline in capacity factors for natural gas and coal plants is a sign of the success of renwables.  The whole situation is complex, and a real analysis would require looking at the entire power system as a whole (which neither of us are doing).  But my worry is that all the author has done is to demonstrate a unaccounted-for cost of renewables, that is the reduction in efficiency of coal and natural gas plants without actually being able to replace them.

Here is his key chart.  It purports to show the total US capacity factor of each energy mode, with capacity factor defined as the total electricity output of the plant divided by what the electricity output could be if the plant ran full-out 24/7/365.

capacity factors

First, there is a problem with this chart in terms of its data selection -- one has to be careful looking at intra-year variations in capacity factor because they vary a lot seasonality, both due to weather and changes in relative fuel prices.  Also, one has to be hugely suspicious when someone is claiming a long term trend but only shows 18 months of data.   The EIA can provide some of the data for a few years ahead of his table.  You can see it is pretty volatile.


I won't dwell on the matter of data selection, because it is not the main point I want to make, but the author's chart looks suspiciously like cherry-picking endpoints.

The point I do want to make is that reducing the capacity utilization, and thus efficiency, is a COST not a benefit as he makes it out.  Things would be different if renewables replaced a lot of fossil fuel capacity at the peak utilization of the day (the total capacity of a power system has to be sized to the peak daily demand).  But the peak demand in most Western countries occurs late in the day, long after solar has stopped producing.  Germany, which relies the most on solar, has studied this and found their peak electricity demand is around 6PM, a time where solar provides essentially nothing.   Wind is a slightly different problem, because of its hour to hour unpredictability, but suffice it to say that it can't be counted on in advance on any particular day to provide power at the peak.

This means that one STILL has to have the exact same fossil fuel plant capacity as one did without renewables.  Yes, it runs less during the day and burns less fuel, but it still must be built and exist and be staffed and in many cases it still must be burning some fuel (even if producing zero electricity) to be hot and ready to go.

The author is arguing for a virtuous circle where reductions in capacity factors of fossil fuel plants from renewables increases the total cost per KwH of electricity from fossil fuels (because the capital cost is amortized over fewer kilowatts).  This is technically true, but it is not the way power companies have to look at it.  Power companies have got to build capacity to the peak.  With current technologies, that means fossil fuel capacity has to be built to the peak irregardless of their capacity factor.  If these plants have to be built anyway to cover for renewables when they disappear during the day, then the capital costs are irrelevant at the margin.   And the marginal cost of operations and producing power from these plants, since they have to continue to exist, is around $30-$40 a MwH, waaaay under renewables still.

In essence, the author is saying:  hurray for renwables!  We still have to have all the old fossil fuel plants but they run less efficiently now AND we have paid billions of dollars to duplicate their function with wind and solar plants.  We get to pay twice for every unit of electricity capacity.

Environmentalists are big on arguing that negative externalities need to be priced and added to the cost of things that generate them -- thus the logic for a carbon tax.  But doesn't that mean we should tax wind and solar, rather than subsidize them, to charge them for the inefficiently-run fossil fuel plants we have to keep around to fill in when renewables inevitably fail us at the peak time of the day?

By the way, speaking of subsidies, the author with a totally straight face argues that renewables are now cheaper than fossil fuels with this chart:

solar costs


He also says, "Wind power, including U.S. subsidies, became the cheapest electricity in the U.S. for the first time last year."

I hate to break it to the author, but a Ferrari would be cheaper than a Ford Taurus if the government subsidized it enough -- that means nothing economically other than the fact that the government is authoritarian enough to make it happen.  All his chart shows is that solar is more expensive than coal and gas in every state.

And what the hell are those units on the left?  Does Bloomberg not know how to annotate charts?  Since 6 cents per Kw/hr is a reasonable electricity cost, my guess is that this is dollars per Mw/hr, but it is irritating to have to guess.


  1. J Bryan Kramer:

    OK I worked in a large power plant for 30+ years. The reason for the seasonal dips is that power plants typically schedule maintenance outages for the fall and/or summer. Those are periods when the demand is the lowest. The plants in a region coordinate their shutdowns so that too many plants aren't down at the same time. It gets to be a fairly complex affair with a lot of other factors involved.

    The big problem with periodically available power is that in order to maintain a base power load there have to be fossil or nuke powered plants sitting idle. But these plants have to have a full crew of operators and maintenance people sitting idle too and being paid for it. The crews are needed to bring those plants from 'hot standby' to full operations as quickly as possible when the wind dies or the skies cloud up.

    That cost should be added into the costs of the none base load capable plants. The technicians operating the base load plants do not come cheap either, they are highly paid for their skills.

    I don't see how this problem will ever be solved in a cost efficient way. Batteries are a joke, the materials (lead, lithium ect) to make those batteries just don't exist. And every other scheme I've seen is even less likely. Furthermore all these storage schemes ignore the realities of thermodynamics, every time you change the form of energy to another form you lose energy. Changed electricity to pressurized gas, you'd lose a lot to friction, pumping losses, other mechanical losses and heat. You want to change it back? The same in reverse occurs. You be very lucky to get back half the power that you started with.

  2. The Outsider:

    There's another thing to consider. Generation shifts to wind and solar when they are available even when they are uneconomic because most states require utilities to produce a certain percentage of their energy from "renewables." In that scenario, the cost compared to coal or gas is irrelevant.

  3. herdgadfly:

    With government-financed wind and solar monstrosities taking over our landscape due to wrong-headed politicians wasting money and dictating conversion to renewables, plants burning carbon-based fuels are declining because of Obama's war on coal and because of legislative taxation and regulation of the new answer to cheap utilities - plentiful natural gas.

    Science doesn't support this carbon pollution idiocy and economics for renewables will remain untenable for as long as the government interferes - which will now be forever. What recovery the US has shown since 2008 is directly attributable to shale gas and oil - and that too is our only shining star.

  4. Brad Warbiany:

    Offbeat question... After reading a piece by [I think] Megan McArdle that you may have pointed us readers to, I wonder if the peak electricity demand in America is at the same time of day as it is in Germany.

    After all, Europe is not generally as hot nor as dependent on air-conditioning as the US. I could see US peak, as a result, being much more in the 1 PM to 4 PM range (the hottest part of the day), while our air conditioners are less active as the sun starts to fade into the early evening.

    Any thoughts on that one?

  5. Daniel Nylen:

    Capital costs and total lifetime/cycle costs are important as that is what the consumer pays absent subsidies. It is actually very hard to get the total costs and real capital costs outside of subsidies, so there are studies where the authors try to take out effects and adjust--and reality then goes out the window. We have a pretty good idea of actual costs for coal, oil, and gas, but not for nuclear and renewables. How long does a wind turbine actually last and what is it s maintenance cycle/cost? Same with the new solar cells versus output?

  6. J Bryan Kramer:

    Peak power in the US occurs when people are getting ready for work say around 6-8 AM and the second larger peak occurs when they get home, turn the a/c on higher, do laundry, run other appliances and so on. That's from 4 until 8 PM generally. I can't see how the Europeans would differ much. They don't use as much a/c because where they live they are at higher and colder latitudes. Living in Florida I'm always amused when I hear them complain about a 'heat wave' in the upper 80's low 90's.

    I cannot think of a less appropriate place for solar than central Europe. It's cloudy frequently and they are at higher less efficient latitudes. If you go due east from NYC you'll hit Madrid Spain. And most of the US is south of NYC.

  7. joe:

    Kramer - I cannot think of a less appropriate place for solar than central Europe. It's cloudy frequently and they are at higher less efficient latitudes. If you go due east from NYC you'll hit Madrid Spain. And most of the US is south of NYC.

    You mention that most of Europe is further north than most of the USA -

    On totally different subject obviously overlooked by liberals in promoting Obamacare due to the shorter live expectancy in the US, As a general rule, life expectancy increases world wide as you move further away from the equator up to around the 60th pararal. Obviously there are numerous other reasons unrelated to the quality of health care, but latitude does account for some portion of the difference.

  8. J Bryan Kramer:

    A lot of the shorter life expectancy in the US vs Europe is due to a medical fact. In the US premature babies of any age are given medical care to try to save them. In Europe from what I hear, they just let the baby die unless it is able to live w/o life support. Also if you subtract the big inner cities from the US number along with the premie affect you'll see that the US numbers are as good and better than most.


  9. Lawrence Karch:

    Denmark is a good example of a country that has grown its renewable capacity significantly (almost totally wind), but has failed to decommission any of its thermal plants. The inescapable conclusion is that while the percentage of energy delivered by wind has increased, the pre-existing thermal plants cannot be decommissioned for reasons of reliability. Oh yes, the price of electricity in Denmark is something north of $0.30 per KWh, among the highest in Europe.

  10. Not Sure:

    To proponents of green energy, higher costs are a feature, not a bug. Because, of course, you shouldn't be using as much energy as you do, and higher prices will encourage you to use less. In fact, you probably shouldn't even be alive since there are far too many people already.

  11. Lawrence Karch:

    " probably shouldn't even be alive since there are far too many people
    Now you're talking. Several noted climatologists have commented that an earth with five billion people would be a lot more manageable temperature wise than a ten billion person earth. Apparently this concept has yet to make it into the Pope's plan for climate control

  12. Shane:

    Thanks Coyote for the analysis. I wished that renewables could make traction but it seems that there are a lot of hurdles to overcome. But hey it makes great news when a likable technology seems to make headway. Maybe we eventually we will see some truly cool tech hit the energy scene :)

  13. Pinebluff:

    Regardless not irregardless

  14. xtmar:

    I don't see how this problem will ever be solved in a cost efficient way.

    Obviously I don't have the experience that you have, but it seems like aero-derivative natural gas turbines are a pretty good solution here, as they have a proven record of reliability in the aviation industry, as well as in other similar functions like powering offshore drilling rigs, plus their startup time is on the order of minutes, rather than hours or days. I'm not sure what their end to end efficiency is compared to a well designed dedicated coal or natural gas plant with more heat reclamation, superheaters, etc, but I would imagine the modularity, simplicity of use, lower capital costs, and peakability would compensate for the lower peak efficiency, or at least that seems like a promising avenue.

  15. xtmar:

    Why are you more optimistic on solar than wind? If anything, I would think that wind is a better bet, because it can produce 24/7 if there is wind, while solar can probably produce at most 12 or 13 hours a day. Obviously you need a fair amount of geographic diversity to account for windless days, but that doesn't seem insurmountable, especially if you do more offshore production, where winds are generally higher.

    That being said, nuclear seems like the best option for carbon-less energy production, even though that's not renewable in the common sense of the term.

  16. J Bryan Kramer:

    Sure we had 3 of them onsite and used them for quick loads. If one of the big units had an unscheduled outage for example (it was broken in other words) or if another utility had a quick call for power (for which they were paying a premium price). But they are very inefficient compared to full sized power plants which are optimized to the max. And inefficient means pricy power.

  17. Trevor:

    The biggest kick in the teeth for renewables has come from fracking. Natural gas producers in the Marcellus and Utica formations had done such a tremendous job of proving up natural gas reserves that they are sitting on decades worth of drilling inventory. Natural gas prices will continue to be stuck in the 2.00-4.00 MMBtu for some time to come. Additionally, even in oil-dominate shale plays like the Bakken and Eagle Ford, wells heavy on oil production will begin to produce increasing amounts of associated natural gas as the wells age. Renewables are going to need a heavy dose of Moore's Law before they can even approach the cost structure of natural gas.

  18. mesocyclone:

    The big problem with solar and wind is intermittency. With their current low penetration, that cost is minimal, but if we get up to 15-20% penetration, there will be a big problem with grid stability. The current grid stores some energy in the angular momentum of the gigantic generators, and the magic of three phase synchros means that these will automatically (and I mean open loop, no controls) compensate for sudden shifts - which has to be one in less than 5 milliseconds. Solar does not have this immediate storage. For shortages longer than a few milliseconds, turbines need to be rotating and fueled, even if they are generating no power, in order to take up the slack. This is quite expensive. For the CO2-phobes, these are also emitting CO2, with an effective CO2 rate per kilowatt hour of infinity, since they are not generating power. Of course, we have some amount rotating anyway (those generators with their angular momentum are not all running at full power), to handle other variations, but alternative sources will require a lot more.

    If we had economical storage (such as pumped hydro), we could easily handle intermittent generation. We don't, and there is no technology on the horizon to provide it. The Dutch, who insist on solar/wind, are digging a reservoir a kilometer underground to use for pumped hydro. Needless to say, this is not economical.

    If we really have to reduce CO2 emissions (and some day, we will if we stay on fossil fuels), nuclear seems like the way to go. Today, it is too expensive, but it doesn't have to be. The expense is a combination of ancient technology, regulatory hurdles (which among other things make modern technology very costly to license) and general obstinate obstruction from environmentalists and NIMBYs.

  19. CapitalistRoader:

    Take fatal injuries out of the calculation and the US is #1 in life expectancy. The US is a very big and racially diverse country, unlike the EU countries or Japan. We drive a lot to get around our big country so we get killed in more car accidents. And our racial diversity results in more violent interactions—frequently resulting in death—than the racially homogeneous Japan and EU states.

  20. ano333:

    I agree that nuclear power may be promising, but I see it more as backup to a renewables + battery system to get us through a spate of cloudy (or windless) days. Battery technology seems to be constantly improving.

  21. Dan Wendlick:

    The big thing with Solar is that it appears that there are still considerable gains to be made in the technology of photovoltaic cells (I think solar thermal is a dead-end technology). There are some promising developments around processes that allow the active layers to be essentially painted onto a glass substrate instead of the ion or vapor deposition of today.
    the real question is whether people will tolerate the environmental disruption that would result from the real estate needed to house the cells.

  22. ano333:

    I would be careful about taking correlation to be causation with the latitude analysis...

  23. ano333:

    According to IEEE (using a NREL study), 2008 PV tech would need 0.6% of the land in the US to power the whole nation. This can only improve...

  24. J Bryan Kramer:

    A nuke plant is the worst possible unit for that application, they do not make good peaking units. Nuke plants are complex to start up and take more time to do so than fossil fuel plants. The make excellent base load units tho which is what they are designed for. It would be pure stupidity to take a nuke unit off line in order to run one of these 'green power' units. Nukes produces no CO2 and no fly ash. The nuke waste issue has been blown far out of proportion by the same sjw mob.

  25. joe:

    Ano33 - I would be careful about taking correlation to be causation with the latitude analysis.

    just an observation - which has some reasonable correlation with life expectancy, (not sure if the correlation holds true going south of the equator, but does hold reasonably true going north of the equator). My broader point is that there are numerous factors that effect, life expectancy as opposed the lower US life expectancy due to "poor and/or limited" access to health insurance/health care in the US

  26. xtmar:

    Sure, but improved solar output doesn't really address the energy storage/time diversity issue, which it seems like wind is more amenable to.

  27. Dan Wendlick:

    That doesn't sound like a lot, until you realize that on average, every standard township would need to devote 138 acres to solar collection.

  28. sean2829:

    I think people are thinking of Moore's law backwards when it comes to energy. Moore's law reduced chip size which resulted in more computational power but less energy consumption per computation performed. The battery and power supplies did not change by orders of magnitude, they just got smaller because of lower demand. If there is an application of Moore's law regarding energy, I think you stand a much better chance on the consumption side such as with LED lighting, better insulation, designs that passively heat and cool, more efficient cooking and cleaning devices, high efficiency vehicles. Reduce energy consumption by a factor of 5 and intermittent energy coupled with storage devices begin to make more sense.

  29. disqus_DCPbdeGXic:

    Having developed the integrated resource plan for an electric utility, I can say that Coyote nailed it spot on. The general public and media doesn't understand the need to build capacity to meet peak demand, nor do they care to take the time to understand. Anything that shows renewable energy in a positive light serves their explicit and implicit biases and will be presented to the public with the typical spin shown here. I'm not really sure what the solution is here and with increasing cost pressures and an unreliable power grid I would really hate to be running an electric utility these days.

  30. Daublin:

    The land usage is especially interesting if you compare it to the concerns about landfills using up too much real estate. Landfills are much smaller than that!

    To get an idea of the order of magnitude, here are the numbers using the above IEEE article, and the top Google hit I could find on landfills that gives a straight answer.

    solar power for nation: 5,760 square miles (0.1%)
    landfills: 250 square miles/century