A free-market energy blog
Random header image... Refresh for more!

Wind Integration Realities: Case Studies of the Netherlands and of Colorado, Texas (Part I: Introduction)

There is no convincing proof that utility-scale wind plants reduce fossil fuel consumption or CO2 emissions. Although there are are a number of reports claiming gains can be made that will combat climate change, free us from fossil fuel “addiction,” provide energy independence and needed 21st century industrial development, such reports are not substantiated by definitive and comprehensive analyses.

To determine the actual effects will require long-term time series, at intervals significantly less than one hour, of wind production and fuel consumption due to fast ramping of fossil fuel plants to compensate for wind’s volatility in an electricity system where wind represents approximately at least 1-2% of production.

As opposed to wind proponents’ claims, studies based on actual experience with wind integration are emerging  that demonstrate the fossil fuel and CO2 emissions gains are not valid. The two reviewed here are examples but are limited by the lack of availability of complete information on operational performance, especially of wind plants. Fortunately, enough information can be gleaned that provides a strong indication of what those who have studied this objectively have long suspected.

Why is more complete information about wind performance and integration not available? Is it because wind proponents, including some policy makers and wind industries, do not want the realities disclosed, or, in the case of many environmentalist organizations, because they would interrupt established agendas? Or is it that these groups believe it unnecessary because they do not understand the realities of utility-scale wind power?

Two New Studies: le Pair/de Groot (Netherlands) and Bentek (Colorado, Texas)

The two studies reviewed were released this year and show increases in fossil fuel or CO2 emissions with the introduction of wind plants. The first is based on the Netherlands experience by C. le Pair and K. de Groot, and the second for Colorado and Texas by Bentek Energy. Their findings will be compared to each other, as well as to results from my fossil fuel and CO2 emissions calculator. The analytical approaches taken by le Pair and de Groot, Bentek and the calculator are different, but the results are very similar. This is therefore a very revealing and instructive exercise.

Le Pair and de Groot take a very analytical path and apply the formulas they derived to published information on the Netherlands system, for which some actual information on fossil fuel inputs for electricity production is available. Bentek uses detailed information on increases in coal-plant cycling since the introduction of wind plants, along with the impact of wind “events” on reported emissions. Because the Public Service Company of Colorado (PSCO) does not publish hourly wind production, Bentek is restricted to a few such events, from which they draw general conclusions for Colorado.

To validate the Colorado findings, Bentek uses the same analysis approach for Texas with information from the Electricity Reliability Council of Texas (ERCOT), which reports wind production at 15 minute intervals. This not only provides validation of the PSCO analysis, but also conveniently adds experience from a third jurisdiction. The calculator is a general model of the interaction between an amount of wind generation in an electricity system and the fossil fuel plants (coal and gas) involved in balancing wind’s volatility.

Summary of Results

This is the first in a four part series that analyzes and compares the findings of these studies with each other and to the calcualtor. Briefly, the results are:

The Netherlands

Le Pair and de Groot show that when the entire fossil fuel fleet efficiency is reduced by about 2% due to the presence of wind, the fossil fuel consumption saving is zero. This is the calculated efficiency reduction in the fossil fuel fleet for the Netherlands for a wind penetration of about 3% based on the published fossil fuel input and electricity production information. Their conclusions include the following:

The use of wind energy for electricity generation in combination with the requirement for fossil fuel powered stations to compensate for wind fluctuations can easily lead to loss of the expected saving in fuel use and CO2 emission. In addition, the conventional stations will be subject to accelerated wear and tear.

It is recommended to get an accurate and quantitative insight into these extra effects before society sets out to apply wind energy on a large scale. All producers must be required to publish data on the efficiency effects and fuel use when wind energy is added on.

Colorado and Texas

The study by Bentek Energy, aptly named “How Less Became More: Wind, Power and Unintended Consequences in the Colorado Energy Market,” is a ground-breaking analysis of the effects of the introduction of wind power into electricity systems. The study is based on actual results for the PSCO system in Colorado and ERCOT in Texas and their overarching conclusion is that there are unintended consequences to the implementation of Renewable Portfolio Standards (RPS). One of the key findings is:

Contrary to their stated goals, implementation of RPS in Colorado and Texas appear to be adding to the air pollution problem, especially in areas where older plants are cycled more frequently.

Calculator

The fossil fuel and CO2 emissions calculator was applied to each of the jurisdictions studied and shows similar results. In each case an explanation of the calculator input parameters is provided.

Comparison of Results

The congruence of results from these three different approaches is a convincing confirmation of the questionable value of new alternative energy sources, especially wind, in an electricity system. RPS programs, and similar initiatives to encourage new renewables, should be withheld until such time as objective and comprehensive evaluations can be made in a completely transparent manner about the real benefits.

Part II will provide more details on the Netherlands study and Parts III and IV for the Bentek study on Colorado and Texas respectively.

14 comments

1 Charles { 05.22.10 at 2:24 am }

The main issue of course, which is mentioned in the body of this article, is that most of the information regarding wind farm performance is being kept from the public.

In most of the countries who have been stupid enough to go down this path, the general public opinion is that wind is a clean, renewable (almost infinite) source of energy. However, it is rare for individuals to look logically at the big picture and see the yawning holes in the theory.

The most irritating outcome though are the Deep Green adherents, trying to push us under the bus driven by all their traditional greatest enemies (Big Oil, Big Electricity, Big Banks, Big Law, and Big Government), who are usually the only ones to benefit from this exercise.

What are(n’t) they thinking?

2 Jon Boone { 05.22.10 at 6:18 am }

Since the Bentek report is impressed to sell a natural gas/wind tandem, presenting data that shows, accurately, that wind integration primarily by coal generators increases CO2 emissions and coal consumption–but fails to produce data about how wind volatility affects natural gas performance–I await Kent Hawkins’ evaluation, particularly since Hawkins (and Peter Lang) show that wind/natural gas could, at best, achieve very marginal results in offsetting CO2 and conserving natural gas consumption.

Le Pair and de Groot’s analysis seems provisional, not only because it has yet to be tested by reality (because the researchers don’t have access to the necessary performance data) but also because their basic methodology is not well understood, whereas Hawkins’ calculator approach is transparent. I hope that Hawkins’ has been able to clarify and enhance the Dutch methodology approach, making it one that others can apply throughout the world.

I look forward to the unfolding of this article.

3 Peter Lang { 05.29.10 at 6:23 pm }

Kent,

Sorry, no detailed comments yet. I’ll get to it.

I mentioned to you elsewhere that Australian data should be excellent for extending this work of calibrating the calculator. Australian generation data is recorded at 5 minute intervals and is publicly available. The Australian National Grid is the largest in the world by aerial extent (so I am told) . We have 18 main wind farms spanning some 12ookm in areal extent. Importantly, one region of the grid is connected by just two small interconnectors to the main part of the grid. This is the South Australia region and it contains about 50% of the installed wind capacity. So it would be an ideal small, isolated region for further studies. We also have the Western Australian grid which is totally isolated from the NEM grid. I am not sure how accessible the data is from this grid. I am not sure how to get the fuel use data from the generators. They would need to cooperate.

Below are some links to data and some sites that are downloading it and making it available in various ways.

http://www.landscapeguardians.org.au/data/aemo/

Here are charts from this data

http://windfarmperformance.info/?date=2010-05-17

Here is the AEMO site. http://www.aemo.com.au/data/csv.html#nsgendata

You can down load all generation data from all registered generators connected to the the National Electricity Market grid. This grid does not extend to Western Australia or Northern Territory.

The site is an absolute mess and I cannot provide any help on how to find your way around it. I’d need to do a PhD on the site to begin to understand its layout.

Here is another site downloading and charting the wind data (they have only just got started. They would appreciate any comments from you.

http://www.oz-energy-analysis.org/index.html

This explains the Australian National Grid:

http://www.aer.gov.au/content/item.phtml?itemId=732297&nodeId=797fa2c37535f919f67fa34dc4970e13&fn=Chapter%201%20%20Electricity%20generation.pdf

This gives the information about all the main fossil fuel generators connected to the NEM.

http://www.aemo.com.au/planning/419-0035.pdf

I don’t know how we could get 5-minute or 30-minute fuel use data from the generators.

4 Jeff C { 07.01.10 at 10:59 pm }

You imply that all of the studies in favor of wind are invalid due to the nature of data collection (no long enough, or frequent enough, etc). Then you go on to talk about two studies against wind?!! You just said studies are inaccurate!

Also, the claim that wind power increases fossil fuel emissions and CO2 is misleading. Take a look at the extreme – the following link is regarding an island that is 40% powered by wind and the rest is bio fuels that come from the crops grown on the island. According to your information, the wind power on the island should be producing tons of fossil emissions and CO2, which clearly isn’t the case. You should really get some critical thinking skills. http://inspiredeconomist.com/2009/01/31/want-a-sustainable-community-try-an-island/

5 PB { 07.15.10 at 5:34 am }

Jeff C- Clearly your island is not going to produce fossil fuel emissions if there is no fossil fuel generation. But that misses the point; back in the real world, the real world which provides your island with all the goods and services it can’t provide for itself, electricity is made from fossil fuels. The rest of the world can’t live off 60% biofuels- just do the calculation on the land take required and you’ll see why.

6 KHawkins { 07.15.10 at 2:43 pm }

PB has dealt with your comment about living on an island. I am surprised at your other comment.

You are incorrect in saying that I “imply” that “…all of the studies in favor of wind are invalid due to the nature of data collection (no [sic] long enough, or frequent enough, etc).” What I said in this post is that there is no convincing proof that utility-scale wind plants reduce fossil fuel consumption or CO2 emissions. Your comment about available data applies to all studies, both for and against wind power. The question is: Why do the wind proponents, some of whom likely have access to the necessary information, not base their analyses on such?

In the absence of such information, all analyses are thus incomplete, although as I have said elsewhere all studies make a contribution – the question is how much?. For a general description of the shortcomings of typical wind proponent analyses see my post at http://www.masterresource.org/2009/11/wind-integration-incremental-emissions-from-back-up-generation-cycling-part-i-a-framework-and-calculator/ .

Not having the “inside information”, so to speak, studies by wind opponents are left to glean what they can from available information, and as I said these studies are emerging. The Bentek and Netherlands studies are based on actual experience, data for which is slowly becoming available, and they are a refreshing departure from most other studies. The available information is still very much limited – you will note in the subsequent post, Part III, I did say that there were “notable limitations” in the Bentek study. Further, the Bentek study authors call for more comprehensive analyses for confirmation, as do I and the authors of the Netherlands study.

On the other hand, for an analysis of typical wind proponent studies see http://www.masterresource.org/2009/12/wind-integration-incremental-emissions-from-back-up-generation-cycling-part-iii-response-to-comments/ , http://www.masterresource.org/2010/04/case-study-on-methods-of-industrial-scale-wind-power-analysis-part-i/#more-8609 and http://www.masterresource.org/2010/04/case-study-on-methods-of-industrial-scale-wind-powerii/

If you can find a wind proponent study that takes the sound and comprehensive approach that myself and others call for, and you can demonstrate that it does do this, I would appreciate knowing about it.

7 Das schmutzige Geheimnis der Wind- und Sonnenkraftnutzung | { 02.24.11 at 3:57 am }

[...] Kent Hawkins: Wind Integration Realities – Case Studies of the Netherlands and of Colorado, Te… [...]

8 Does Wind Power Actually INCREASE CO2 Emissions? — True North Reports { 03.11.11 at 10:16 pm }

[...] is the conclusion that Kent Hawkins reached after a four part study on “Wind Integration Realities“.  Mr. Hawkins  holds electrical engineering degrees from Royal Military College of Canada [...]

9 Wind Power? — True North Reports { 04.25.11 at 7:51 am }
10 Taxpayers Get Hosed on Duke Energy’s Wind Farm Buying Spree : Greener Energy Sales { 08.04.11 at 9:20 pm }

[...] not save energy, or decrease pollutants and carbon dioxide emissions.  Because wind (and the sun) energize intermittently (on calm or cloudy days), they require back-up power from coal or natural gas generators, and when [...]

11 Taxpayers Get Hosed on Duke Energy’s Wind Farm Buying Spree : Get Greener Energy { 08.05.11 at 4:33 am }

[...] not save energy, or decrease pollutants and carbon dioxide emissions.  Because wind (and the sun) energize intermittently (on calm or cloudy days), they require back-up power from coal or natural gas generators, and when [...]

12 Subsidizing CO2 Emissions via Windpower: The Ultimate Irony | Save The Eagles International { 04.06.12 at 1:14 pm }

[...] recent four-part Wind Integration Realities reviewed two new studies, based on actual experience, that show fossil fuel consumption and CO2 [...]

13 Michael Goggin, AWEA { 09.24.13 at 11:14 am }

A new study conclusively answers this question, and it turns out that wind energy’s emissions savings are not undermined by additional cycling at fossil-fired power plants:

http://www.nrel.gov/electricity/transmission/western_wind.html

14 Kent Hawkins { 09.25.13 at 3:51 pm }

Michael,

Nice try.

Because of their highly variable and random production over short time periods (minutes), wind and solar PV cannot be realistically or properly analysed using statistical methods to evaluate real time events. This statistical approach hides the real time impacts, and electricity systems must balance load and generation in real time (again minutes).

To illustrate, I have done some analysis of the BPA wind performance, at 5 minute intervals (which still masks some greater volatility) and 14% wind penetration in energy terms, which shows that, statistically speaking, the standard deviation of the increase in ramping of load minus wind versus load alone is 26 MW. However, the real time effects show over 8 instances per day (one every three hours on average) of over more than three times this level (three standard deviations), with one event per day (on average again) in excessive of 200 MW which is over 8 standard deviations. Note that the ‘averaging’ over time of these events does not mean that their occurrence cannot be more concentrated in real time. There are more ‘black swans’ in wind production than statistical approaches are designed to handle, if you want to avoid severe consequences. For real world experience, talk to the financial industry about this.

Statistical analysis is based on the assumption that the distribution of 99.7% of all events is contained within three standard deviations. Therefore the BPA analysis shows that three standard deviations should exclude only 7 ‘black swans’ or extraordinary events. Over the period measured there were 52 or over 7 times this number. This shows that it is questionable and risky to apply statistical analysis here in connection with our electricity systems on which so much of our well-being depends.

For the purposes of planning reserve requirements based on the use of reliable, dispatchable generation plants in the electricity system (and wind and solar PV are neither), statistical methods have some value.

Having said all this, statistical analysis of wind and solar PV performance may have some value if limited to providing some indication of the relative effect of their use (1) between jurisdictions, and (2) within a jurisdiction between time periods and wind and solar PV penetrations, for example. But, and this is important, it is not valid for the determination of absolute values, especially actual reserves required, stress on wind balancing plants and, importantly, emissions.

I should introduce you to the ‘Statistics Professor’ who some time ago in a comment tried to convince me that statistical methods are applicable in the analysis of real time wind performance, claiming that the ‘square root of the sum of the squares’ formula properly captures the real time effect of combining two random series – in this case the net of wind and load and load alone. Perhaps you already know him, being of like mind on this issue.

In short, you have neither proved your case, nor disproved mine.

It looks like this issue is not going to go away, so for a more complete response I will have to look at completing my BPA analysis when time permits.

Leave a Comment