…Misleading Costs for Wind and Solar?…
Recently the media has reported that wind and solar were competitive with coal and natural gas for generating electricity.
The Wall Street Journal, for example, published an article with a headline, Economic impact of wind farms is changing the political dynamics of renewable energy.
These media reports could lead people to believe that wind and solar were competitive with coal-fired and natural gas power plants, which is not the case. Electricity generated by coal-fired power and natural gas combined cycle power plants remain the lowest cost methods for generating electricity, especially when the unreliability of wind and solar are taken into consideration.
In trying to determine the source of the media claims, two sources became apparent.
- Contract purchase agreements
- Studies performed by financial firms such as Lazard
In the first instance, the lower contract prices were the result of subsidies. The lower prices did not accurately reflect the true costs for wind and solar: The subsidies resulted in low prices and low LCOEs.
In the second, some assumptions in the studies performed by financial groups resulted in low LCOEs (Levelized Cost of Energy).
Furthermore, equating LCOEs of wind and solar with those of coal and natural gas power plants is fallacious. Beyond a certain point, it’s impossible to replace coal and natural gas with wind and solar on a one for one basis, interchanging them as though they were LEGO pieces.
A review of the Lazard study established why the study produced very low LCOEs for wind and solar: LCOEs that were atypical of previously determined LCOEs.
Lazard study is for New Construction
It’s important to point out that the LCOEs determined by Lazard were for new power plants, something not mentioned in their report.
Existing coal-fired and natural gas power plants have LCOEs of around three cents per kWh because their construction and financing costs have already been amortized.
It makes no economic sense to deliberately replace existing coal-fired and natural gas power plants with wind and solar units even if the LCOEs of new wind and solar power plants are below six cents per kWh.
Lazard Assumptions
Lazard held financial costs, such as the cost of debt and equity, constant when making calculations for each type of facility.
This was an effort to ensure that calculations between facility types were fair. However, there were at least two instances where this assumption was misleading.
- Natural gas combined cycle (NGCC) plants were assumed to have a life of twenty years, which is half the life that should have been used. Financial costs should have been amortized over 40, not 20 years.
- The investment cost for a coal-fired power plant was assumed to be $3,000 per KW. This is higher than actual historical costs for supercritical plants and slightly higher than for ultra-supercritical plants. This imposed a financial penalty for coal-fired power plants.
There were two important assumptions in the Lazard study that were either questionable or that slanted conclusions unfairly to the benefit of wind and solar. These are addressed in a and b. A third factor was omitted from the study and is addressed in c.
a) Capacity Factor
Capacity factor (CF) is defined as the amount of electricity produced over a year by an installation, compared with the amount that could theoretically be produced based on the facility’s nameplate rating.
The Lazard study refers to “resource availability”, and it is unclear whether the CFs used in the study are true CFs or ersatz CFs based on some undefined resource availability calculation.
Because this is unclear, both possibilities are addressed for wind.
Alt 1: Traditionally Defined CFs
The capacity factor (CF) for wind used in the Lazard study was significantly higher than experience from existing installations. The study used 55% in one instance and 38% in another.
Actual CFs, as reported by DOE in its 2015 Wind Technology Report, averaged 32.8%, between 2011 and 2015; 31.8% between 2006 and 2010; and 30.3% between 2000 and 2005.
New, taller units with longer blades will probably have higher CFs, but not anywhere near 55%.
Wind installations in high wind areas, such as Montana where CFs could be higher, require long and expensive transmission lines, the costs of which are not included in the Lazard or many other studies.
The use of higher CFs and lower capital costs in the Lazard study, skewed the LCOEs for wind, making them unreasonably low.
Alt 2: Ersatz CFs
The Lazard study may have used a specially designed “capacity factor as a proxy for resource availability”.
Why this would be done is unclear since actual wind resources have been carefully mapped across the United States for heights of 30 meters, 80 meters and 100 meters above ground level.
The best winds for generating electricity are predominantly in the upper plains states such as Montana, and across the front range of the Rocky Mountains.
The regional factors used in the Lazard study do not appear to align with the wind maps available from NREL, though these regional factors were apparently used to represent wind availability across the country.
The Lazard study did not explain how these ersatz capacity factors were determined, so there is no way to determine their appropriateness or accuracy.
For this reason, the LCOEs developed by Lazard using ersatz CFs for wind are suspect, and not comparable to traditionally determined LCOEs.
b) Solar
The Lazard study seems to have used a specially designed “capacity factor as a proxy for resource availability” when determining LCOEs for solar.
Presumably “resource availability” refers, in some manner, to insolation levels.
“Resource availability” was apparently used to establish, what can best be described as ersatz capacity factors for solar installations.
Insolation levels are readily available for all areas of the world, so it begs the question of why Lazard chose to create a “resource availability” factor for solar.
Insolation levels for the Southwestern United States are twice those for the Midwestern United States, yet the LCOEs arrived at for solar by the Lazard study did not reflect these substantial differences.
For this reason, the solar LCOEs developed by Lazard are suspect, and not comparable to traditionally determined LCOEs.
Again, The Lazard study did not explain how these ersatz capacity factors were determined, so there is no way to determine their appropriateness or accuracy.
The report did confirm that rooftop PV solar is uncompetitive. As demonstrated in Nothing to Fear, PV rooftop solar is uneconomic in every state except possibly Hawaii.
c) Reliability
Both wind and solar are intermittent, and in some respects unreliable.
Beyond small amounts, it’s impossible to replace coal and natural gas power plants with wind and solar on a one for one basis. As mentioned earlier, these are not interchangeable LEGO pieces.
The CAISO Duck curve demonstrates this very clearly.
CAISO Duck Curve with Effect of Extended Renewables
For example, wind and solar must also include expensive storage if the evening ramp-up is to be minimized. Coal and natural gas power plants must be retained to provide power at night and for when the sun stops shining or the wind stops blowing.
These limitations become increasingly worse as greater amounts of wind and solar are placed on the grid.
At the very least, LCOEs for wind and solar are misleading because wind and solar require the use of costly storage. More about the CAISO Duck curve is found in Nothing to Fear.
Conclusion
If an undefined “resource availability” is used to calculate LCOEs, the resulting LCOEs can’t be compared with a traditionally derived levelized cost of electricity (LCOE): It’s like comparing cashews with apples.
Pawning these LCOEs off as equivalent to traditionally calculated LCOEs is misleading at best, and at worst, could be considered deceptive.
In addition, wind and solar are unreliable, and LCOEs do not reflect the extra costs associated with having to compensate for their intermittency and unreliability.
The Lazard report and virtually all media articles attempting to compare LCOEs between wind and solar and coal and natural gas are fallacious and meaningless.
Wind and solar cannot replace coal and natural gas on a one for one basis … They are not interchangeable LEGO pieces.
Coal-fired and natural gas combined cycle power plants continue to be the least costly methods for generating electricity, notwithstanding the latest Lazard study.
…
(253)
Donn,
Thanks for this post. I have been suspicious of how LCOE is calculated for solar and wind.
Questions – why do you think they do this?
Lois
Unfortunately, I believe they have an agenda to promote wind and solar.
I also believe that it’s rooted in their belief that CO2 is the cause of global warming.
It’s frustrating when the media parrots this misinformation, especially when the WSJ does too.
I have written to the WSJ editors and asked that they stop saying that wind and solar are less costly than electricity generated using coal or natural gas.
We’ll have to wait and see how they respond.
How about a petition sent to WSJ. I will sign.
Thanks. I’m not planning on a petition drive to the WSJ. A letter from you and others would help to establish that their news reporting is biased and that their reporters should use greater care when establishing facts. You can use my article as the basis for any letter.
My opinion on the questionable use of capacity factors for wind and solar is that it is a deliberate attempt to mislead. Many seasonal generators, mid-range generators, and peakers have low capacity factors. By using the term capacity factor, wind and solar appear to be as good or better than fossil technologies. The term used should really be availability factor because without wind or sunshine, respectively, these generators are not available.
Good point. Wind and solar are severely limited by when the sun shines or the wind blows. Capacity Factor is a term that has been used for a long time, and, you are correct, it shows that some fossil fuel installations have low CFs. This could lead the non-technical person to jump to the conclusion wind and solar are better, which they obviuosly are not.
http://www.eia.gov has capacity factors for “generators not primarily using fossil fuels” https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_6_07_b and on a separate page for “generators primarily using fossil fuels” https://www.eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_6_07_a. There are also graphs of monthly capacity factors at https://www.eia.gov/todayinenergy/detail.php?id=14611
Thanks.
Good links.
The EIA data always needs to be double checked. It’s also necessary to read the fine print that accompanies the various tables.
For example, an EIA table claims that the LCOE for coal-fired power plants is 10 cents per kWh, without noting that the calculation included a $15 charge for CO2 emissions. This was noted in the text, but most people don’t read the text.
Today’s coal-fired power plants are being affected by having to follow load which is reducing their overall capacity factor. I’ll have to review the data on NGCC plants to see how they compare with other data.
An important oversight is that solar/wind backup systems such as gas peakers, simply enter “backdown mode” — a secretive term in Calif. that hides the fact they remain on & burning gas while their output is unneeded and replaced by the variable wind/solar sources for some hours.
This has the effect of increasing gas dependence 7 emissions in Calif., where the allowed gas leakage more than wipes out all Calif. benefit from all Calif. wind/solar installations.
This is one item we’ve introduced in our legal testimony to CPUC re PG&E’s desire to close Diablo Canyon nuclear plant. Sections 2.2 & 2.6 of our sworn testimony may be of interest…
http://www.cgnp.org/CGNP-OpeningBrief-A1608006_05-26-17.pdf
http://cgnp.org/CGNP-Reply-Brief-A1608006.pdf
http://www.cgnp.org/CGNP_Direct_Testimony_01-27-17.pdf
http://www.cgnp.org/CGNP_Rebuttal_Testimony_03-17-17.pdf
—
Dr. A. Cannara
650 400 3071
Thanks. Excellent points. Thanks for the links.
Worldwide, nuclear power is the lowest cost source (except in South Korea and USA). Diablo Canyon’s price (lowest in California) is 5¢/kWh; 3.7¢/kWh is capital amortization, 0.5¢ is fuel, 0.8¢ is operations. Palo Verde’s price (lowest in Arizona) is 4.2¢/kWh. Construction costs for new nuclear power plants in the US are artificially magnified by delays, lawsuits, dysfunctional NRC regulations…, all of which comes from ignorance and irrational hysteria that has been intentionally generated by (ignorant) loud-mouth opponents.
In the US, we have 80,000 tonnes of spent nuclear fuel and 700,000 tonnes of depleted uranium. Spent fuel consists of two parts:
5% is fission products, which are less radiotoxic than uranium ore before 300 years. This is a trivial problem, especially given that the production rate is only one tonne per GWe-year. If the expense to separate caesium and strontium is bearable, the amount is reduced to 92 kg/GWe-yr (these are the only fission products needing long-term custody).
95% is unused fuel, which is more radiotoxic than uranium ore for 300,000 years. Contemplating isolating unused fuel from the biosphere is madness. The only way to destroy this is in a fast-neutron reactor such as IFR. An all-nuclear-electric American energy economy could consume this in 450 years. Wind, solar and other unworkable green favorites just slow this down.
In the US, we have 1,200 tonnes of fissionable material, for which there are only two uses: weapons, and electricity. Which do you prefer?
According to the Paul Scherrer Institut and the United Nations Scientific Committee for the Effects of Atomic Radiation, nuclear is by far the safest-ever way to make electricity: 43 deaths in its entire six-decade worldwide history — all at Chernobyl, a nutty design that nobody will repeat.
It’s tragic that extremists have created an environment among the American public that largely precludes the use of nuclear power.
Nuclear power is safe, and getting safer. The Chernobyl disaster was no accident. It was the result of incompetence, where a low load test was conducted when everyone knew the reactor was unstable at low load.
It’s possible to treat spent nuclear fuel to reduce the amount of dangerously radioactive material to a small percentage of the total. France was treating their spent fuel, but I don’t know for certain that they still are.
It’s truly unfortunate that nuclear power in the United States is dying.
So lets take the latest big investment news in the US regarding renewables.
A 2 000 MW windfarm would be about 670 turbines with a standard 3MW turbine. That would give an total investment of ~6,7 MUSD / Turbine. Since a 3MW turbine is ~2,5 MUSD for the turbine including installation and comissioning (excluding foundation) the figure tells me that 4,5 BUSD should represent the total investment everyting included in the total lifespan of the turbines (service/maintenance, grid, insurance, taxes, roads, electrical consumption, land rental, decomissioning, etc). That is unless there is a huge difference in landlease, grid costs and taxes which besied the turbines and S/M are the major costs.
So, production on a standard 3 GW is today ~8 GWh yearly for a descent wind site with both descent windspeeds as well as good windquality (wind-shear, turbulence etc) That would make around 5 360 GWh/year for the total farm (losses included). So total production would be:
For 15 years – 80,4 TWh, LCOE 56 USD/MWh
For 20 years – 107,2 TWh, LCOE 42 USD/MWh
For 25 years – 134 TWh, LCOE 34 USD/MWh.
30 and 80 meter for wind ? Even 100 is considered a low hubheight today.
Thanks for the comment.
A few quick comments:
First, you don’t mention the capacity factor.
Second, why did you exclude the foundation, since it is an important expense?
Finally, without getting into reliability issues, most locations for wind turbines aren’t as good as you are assuming. Capacity factor for most locations is below 30%. The best wind sites also require the construction of transmission lines to bring the electricity to where it is needed.
I think people are interested in those items also.
Pingback: The Great Pretenders: Intermittent Wind & Solar Simply Can’t Compete with Coal & Gas & Nuclear – Climate- Science.press
Pingback: The Great Pretenders: | ajmarciniak
The first failing with LCOE is that it places zero value on dispatchability and transmission/distribution efficiency. Because of this LCOE is a useless metric.
No electricity system can operate without dispatchable electricity.
I agree that wind and solar are not suited for use on the electric grid.
I also agree that LCOE should not be used for wind and solar for several reasons, including those you mention.
LCOE is an economically useless term. It’s appalling the EIA gives it any credence at all.
It adds in, somehow, external ‘costs’ of something, to what would otherwise be Generally Accepted Accounting Procedures.
You can’t add apples to oranges.
If they had any sort of integrity, with regard to ‘externalities,’ they’d also add in external BENEFITS.
If one could add apples to oranges, why not add in bananas?
Pingback: Distorting the Levelized Cost of Electricity – Power For USA