Taking the rotor diameter as 100metres (you quote a range of 80 to 110m diameter rotors) it is readily evident that to replace a single 1000 MW plant you will have to cover vast areas of landscape or seascape with wind farms.

If you arrange your 1500 windmills blade tip to blade tip 30 along one side and 50 along the other the you need 3000 metres by 5000 metres or 15 x 10E6 square metres. That means for your single 1000 MW plant you have consumed an area of 1.5 x 10E3 hectares. A hectare being 10,000 square metres. So 1500 hectares or 5.79 square miles are needed for each 1000 MW plant equivalent. It should be readily apparent that one will quickly use up all the available useful sites just to replace present demand. The marginal sites have worse capacity factors because the wind blows less frequently and less strongly at these sites. Logically then as the best sites are used up you will need even more area to replace the same 1000MW plant. While the use of land and sea areas for wind plants poses a nutural limit to wind generation capacity there is also the question of grid stability.While it may not be apparent to the casual user the grid is very carefully controlled for voltage, frequency and power factor. The control of large numbers of variable output power plants connected to the grid is only possible if there are in addition to the wind machines large, stable and highly controllable power plants as well. In other words if (as the proponents argue) we lived in a world where electricity is entirely produced by wind machines then the stability of the electricity grid itself is in severe doubt. This is already becoming an issue in Alberta, Canada where the wind generated electricity is becoming significant but still relatively small proportion of grid capacity. If grid instabilities are occurring now with small amounts of wind power then as the proportion increases those instabilities will become worse and more frequent power outages will occur. The final major problem I see with wind generated electricity is how to deal with the days when the wind does not blow or is not blowing strongly enough. If we could store the electricity then the picture is different but we cannot do that in large quantities. Surely we are not proposing to operate a highly industrialized society that would have to shut down on calm days are we?

I am not against any form of electricity generation but those who advocate one method to the exclusion of all others need to employ logic - not emotion and do their sums. With wind energy, whatever the politics, the math does not make sense. I envisage tidal energy as a much more likely large scale energy generator than wind but that is restricted to coastal sites of course. Great for New York City but not much good in the mid-west.

I would like to hear more from experts on grid stability. It is this issue - not birds - that will be the key problem to resolve if wind is to be anything more than a bit player in the electrical generation sector. Malcolm

And Malcolm: You may be surprised at what ingenuity can accomplish to enable a "free-fuel" energy source to compete with fossil fuel at $15 /MMbtu.

Thanks for the discussion. Always good material!

Malcolm

Many other possibilities are being investigated which I think make wind generation far more financially practical in the long term than tides/waves. All that's required is a price-effective bulk energy storage system. (flow batteries, compressed air, gravity?,.......)

My point was that it's fun to predict the future but risky to bet against it. In 1980 when 30 Kw wind turbines were the standard it was widely predicted that 300 Kw was the maximum potential size. Two days ago I read in ReFocus in two separate articles two prediction of upper maximum unit size of 10 MW and 30MW, and I'll happily take any bet they're both wrong.

In the case of wind, 8-10 wind generators are required to "reliably" produce the output of 1 of the generators, although the average power output of the combined sites will almost always exceed the output of any single generator. This additional "source of opportunity" power could be used in a number of ways, such as the production of hydrogen, as suggested above. However, such applications require either that grid capacity be expanded to accommodate the higher (and highly variable) power flows or that the hydrogen production be co-located the with the wind generators and capable of operating intermittently (a difficult requirement for most process applications).

I do not suggest that wind should not be a component of the future energy supply in the US. I do suggest, however, that there are several real issues which must be dealt with as the combined contribution of intermittent sources of power approach the conventional capacity reserve margin, as will likely happen first in California in the not-too-distant future as the result of a combination of low conventional capacity reserve margins and aggressive renewable portfolio standards.

Hysteria aside, the fundamental technical cause of the California power shortages was a persistent drought which caused the quantity of hydropower available to serve the State's needs to fall below the percentage of hydro capacity historically considered to be "reliable" as opposed to "source of opportunity". The two major failures of the western grid in 1996, on the other hand, were the result of efforts to take advantage of atypically abundant "source of opportunity" power available from BPA.

All of these issues can be dealt with, given both the required technology and the appropriate economics. It appears that, at the moment, both are lacking to some degree. The role of strategic energy planning is to determine where "there" is; and, how we get "there" from here. To date, the federal approach to energy strategy is neither strategic nor serious.

Mr. Duffin has convinced himself, and I believe is in the process of convincing many others, that the time to get both strategic and serious is upon us. I congratulate him for his efforts and applaud him for his insight. I look forward to his assessment of approaches to dealing with non-technical issues, such as "NIMBY" and "BANANA", which may also be very difficult to resolve.

Using hydrogen as a buffer as Len suggests is easy in principle....but difficult to do in practice. I thought at one time that was a solution but the volumes of hydrogen required to be stored require a whole new infrastructure to be constructed...a very expensive proposition. If that cost is borne by electricity consumers then the cost of wind energy becomes uneconomic - whether or not the fuel is free.

WInd energy data from wind farms here in Canada has not been anywhere close to achieving an 85% capacity factor. A wind farm on the Gaspe Peninsula here has been turning in disappointingly poor capacity factor data in the range of 15 - 25% yet according to meteorological data this is one of the windiest places in the country. Actual data like that does not bode well for your prediction of 85 %. Part of the reason is that the met data was far too optimistic about wind patterns. As I said it is dangerous policy to place the electrical generation capability of the continent on such data.

The problem of the proportion of generation from one type of plant or another is not confined to wind. One of the major problems with nuclear in fact is that it is essentially baseload and it is difficult to follow load patterns. Note that it is completely impossible to follow load patterns with wind generation. While many new designs of reactor can load follow none are yet built and we have no operating experience on which to draw. The point is (and maybe the subject of another of your excellent papers Murray) that it is the MIX of generation that is important much more so than the good or the bad of any particular type.

In any event going back to wind it seems to me inherently unwise to build plants with variable power ouput then build another plant to ensure that when it is not operating there is something to take its place. You might as well build just one plant with a predictable fuel supply.

There are many proponents of wind energy who do seriously contend that all electricity can and should be generated by wind. So while there are many thinking people out there like yourself Murray there are also many others who know little or nothing about electrical energy who are easily swayed by the "free fuel" argument. Some of those are our political decision makers and that is why it is not a red herring.

While you and I know that an energy mix is essential there are others who firmly believe that shutting down coal oil gas and nuclear plants and running our entire grid on wind and hydroelectric is feasible. Decisions are being made concerning electrical supply for political reasons not practical reasons that will affect the stability of the grid network in North America. We in the North East of the continent have recent and very unpleasant experience of what happens when the grid fails.

As the percentage of wind installed capacity increases so does the vulnerability and stabiltiy of the grid that delivers it. That is why I say again that wind energy cannot ever be more than a few percentage points of the installed capacity. As Edward points out we already know what happens when we base large amounts of our electrical energy production on natural phenomena over which we have no control. That is a very precarious footing for a modern industrial society.

Wind energy is simply too unreliable to be anything more than a small part of the market.

Appreciate everyones comments. Malcolm

Again I iterate, much of "what is known" is based on outdated basis data (real production-level capital costs of wind units, excess costs presently associated with being "grid compatible" such as inverters etc., not required if "the grid" is not the target, costs of trying to do big jobs with little tools such as long-haul transmission with current A/C grid technology)

Please don't park your liquid hydrogen fueled vehicle in the garage. The vapor "boil-off" is very difficult to detect and very easy to ignite once mixed with sufficient air.

Also, while liquid hydrogen is "very easily stored", it is not as easily produced in intermittently powered process trains.

As usual, those most critical of the analysis on the basis that it did not objectively consider all factors are the most guilty of this shortcoming. Malcolm's analysis of the area required to site wind plants seems to compare the foot print of wind turbine farms only to the footprint of fossil fuel plants. It completely ignores the vast areas of some of the most ecologically important areas on earth that are laid to waste each year to extract fossil fuel (especially coal), not tomention all of the land used in the supporting infrastructure. There are two additional flaws in Malcolm's critique: It assumes all the area under the wind turbines is "consumed;" while in fact, most of the land in a wind turbine plant is still available for other uses such as agriculture. Another important difference is if Murray's estimates prove wrong and wind turns out not to be a boon, the wind turbines may always be dismantled and that land is still suitable for other use. The current razing of the Appalachian Mountains to feed all of the coal power plants with their supposedly "small" foot prints will render these areas effectively ruined for any future use for the foreseeable future. I do not disagree with Malcolm's contention that we cannot simply do away with fossil fuel or other power plant technologies and that we need a diversity ofpower sources; however, that was not Murray's point either.

More efficient lighting, heating and air conditioning, a return to the 55 MPH speed limit, a regaining of our sanity on mpg requirements for new cars and trucks, including an elimination of the Hummer/SUV tax incentive, investment in light and heavy rail, a 99% reduction in the number of commercial flights, since about that many of them are for unimportant purposes (for now I will not digress into the usefulness of military flights, although I am a veteran), recognition of the folly of suburbia. Until most of us decide to become citizens rather than mere consumers, there is no hope for any of these measures.

Thanks to Joe Schiller for mentioning the stupendous, mindless and irreversible destruction caused by mountaintop removal for coal production. See the April 2005 issue of Harper's magazine for an article on this.

When Len Gould mentioned wind power for transport, I did not think of hydrogen production, but electric vehicles. I believe we have enough information now to conclude that the hydrogen economy is a fantasy. EV's are proven, and they would constitute an enormous electricity storage plant. Since electricity storage is crucial for increasing the efficiency of electrical power production and use, what are we waiting for? Mainly, for a decent understanding of the fact that the obstacle to progress is the auto and petroleum industries. The auto companies don't like the reliability of EV's, and the oil companies don't like the fuel.

On nuclear power, see http://www.antenna.nl/wise/621-22/621-22_en.pdf. We must rid ourselves of the false hope that nuclear power is a solution to our energy problems, including global warming.

On global warming, see the current series of articles in the New Yorker, beginning with April 25, 2005 issue. Those who believe that the current global warming is natural, or at least benign, have been hoodwinked by those who make these assertions for selfish reasons. From the second article in the series:

"It's not something dramatic now--that's why people don't really react. But if you can convey the message that it will be dramatic for our children and our children's children--the risk is too big not to care. The time is already five past midnight." (Konrad Steffen, University of Colorado)

eg. "According to their data, nuclear power production causes the emission of just 3 times fewer greenhouse gases than modern natural gas power stations" cited from Stormy VL etc., but those people (and incidentally all the more reliable and accurate ones, eg. ExternE who set the ration closer to 30 times lower) all ignore facts such as the volumes of CO2 co-produced at the well during natural gas extraction.

Reconcile this quote "Southeastern Asian countries such as Indonesia, Malaysia and Thailand have gas fields that are rich inand are therefore difficult to develop commercially. For instance, the Natuna gas field in Indonesia has huge reserves of more than 200 TCF. However, due to its highcontent (70%), theseparation cost is anticipated to have a major negative impact on the feasibility of the development." from JAPEX at http://www.japex.co.jp/en/technology/g_liquids.html. Seems you fossil fuel investors should be required to reconcile that sort of CO2 production as well. Also noteworthy are eg. offshore US floating production platforms needing to so special handling to deal with the CO2 co-produced in domestic fields, could find the reference if you needed it.

"Southeastern Asian countries such as Indonesia, Malaysia and Thailand have gas fields that are rich in CO2 and are therefore difficult to develop commercially. For instance, the Natuna gas field in Indonesia has huge reserves of more than 200 TCF. However, due to its high CO2 content (70%), the CO2 separation cost is anticipated to have a major negative impact on the feasibility of the development. "

Adding 70% / volume CO2 up front to natural gas production makes it worse than coal.

I too was surprised that Murray Duffin didn’t consider the effect of rotor speed on bird kill, but his conclusion that bird kill is a minor problem seems well supported. I hope we will be able to say the same about bat kill.

The EROEI section is also good, but discussions of tower head weight and the design of generator and drive train are missing.

The tower head weight (THW) of 4MW and larger turbines is approaching the ridiculous. According to the article “Wind Turbines: How Big Can They Get?” (REFOCUS magazine, March/April 2005), the THW of the REPower 5MW is about 400 tons. That unit is geared. For the gearless Enercon-112 4.5MW, the THW is thought to exceed 500 tons. In the gearless design, the elimination of the gearbox weight is apparently not enough to compensate for the increase in generator weight.

Judging by the article, about one-third of the THW is in the rotor: blades, blade hub and control mechanism, low-speed shaft and bearing. The rest, about 270 tons for the REPower 5MW, is in the nacelle: gearbox, generator, electronics package, auxiliary equipment, structural elements and enclosure.

The THW of the rotor probably can’t be reduced much, but there is a simple way to reduce the THW of the nacelle: put most of it at ground level. This can be achieved with a slow speed right-angle drive at the top, driving a vertical drive shaft that reaches to the ground level. There would be several benefits:

-reduction in size of crane and time needed to erect the turbine -reduction in size of crane and time needed to repair and maintain the turbine -increased safety and convenience of erection, maintenance and repair -lighter towers and foundations -long drive shaft acts as torsional spring, greatly reducing the shock that is the most important cause of gearbox failures

The gearbox can also be greatly improved. The conventional design for large turbines is two planetary stages on the low-speed end, coupled to a single parallel shaft stage for the high-speed end, consisting of just one parallel shaft.

The planetary design suffers from load imbalance of the planets, reverse bending of the planet teeth and, in wind turbines, very severe loading of the planet bearings. It will also be extremely expensive to produce the ring gears that would be needed for the planetary gearboxes of larger turbines. The single parallel shaft stage also has its shortcomings. Remember the huge gearbox retrofit program a few years ago on NEG-Micon turbines? High-speed bearing failures.

The solution to these problems is the multi-branch parallel shaft gearbox. It has none of the disadvantages of the conventional arrangement, and many advantages over it.

The power electronics now used in variable speed wind turbines can also be eliminated. The drive train can be designed to produce constant output speed from the variable input speed of the rotor. This would allow the use of a synchronous generator to produce constant frequency power in pure sin wave form, not the chopped-up, spiky kind that is produced by power electronics. The synchronous generator also inherently possesses the ability to control power factor and voltage, which is crucial for large wind farms. There is no need to use expensive, delicate power electronics in an attempt to approximate the performance of the synchronous generator.

I doubt that the gearless design will prove effective for large turbines. Instead, they will be geared, either for medium or high speed generators. With the improvements I have listed, they will become cheaper to manufacture, erect and maintain. They will be more reliable and efficient, and last longer. In other words, the EROEI will be greatly increased.

George: You've piqued my interest regarding a "drive train can be designed to produce constant output speed from the variable input speed of the rotor." Where could I learn more about this? Also re: electronics, have you seen ABB's latest design for offshore wind farms? They convert the individual generator's 3 phase 600V output to 1 phase 30KV AC for aggregation to a central location using cheap ( and much more efficient) thyristors at the nacelle. That then allows much cheaper 1 phase transformer to step up to 500 KV at the central DC transmission station, which still uses more costly IGBT's to produce smooth DC, but still a lot cheaper overall and doesn't care what RPM the turbine is running. I'm still guessing it's an error to bet against electronics in the long term, given trends.

IHC Holland is not advertising the VARIBLOCK for converting variable to constant speed in a wind turbine, but it could be used that way. It is a continuously variable transmission of a type that can handle the high power of a wind turbine with high reliability and long life. Most CVT's cannot pass this basic wind turbine test, as the director of the National Wind Technology Center in Colorado will tell you. They have tested a lot of them. I will describe a second test below.

There is an excellent Yahoo site for continuously variable transmissions, http://cvt.com.sapo.pt/why/why_cvt.htm You can join it to participate in the forum.

As the Yahoo CVT website explains, there are many kinds of CVT. The VARIBLOCK type can be built with any standard gearbox having in-line input and output shafts such as the planetary, or a multi-branch parallel shaft type. The gearcases of these gearboxes can be rotated about the shafts. That is how the speed control is achieved--by varying the speed of rotation of the gearcase so that the output speed is constant, no matter how the input speed varies (shaft registration effect).

The second test that a CVT must pass in a wind turbine is control of the output speed with enough accuracy for constant frequency power generation. I have developed such a control method, and a multi-branch parallel shaft gearbox that can be used with it in a CVT, or as a replacement for the planetary type in non-CVT applications of many kinds.

I was not aware of the ABB developments you mentioned. From your description, they are converting the wild AC from the wind turbines to HVDC power for transmission. At the other end, then, they must convert the HVDC to AC for the grid using power electronics. Since I am not a proponent of power electronics, I regard such systems with suspicion. Power electronics work, and they keep getting better, but I doubt that they will ever be able to produce power equal in quality to that of a synchronous generator.

This may present an insurmountable barrier to widespread use of power electronics. Variable speed wind turbines and other renewable energy sources such as solar power all rely presently on power electronics to produce constant frequency AC from a DC bus. But this AC power is not clean. Clean enough, maybe, for now. But what happens when the percentage of power on the grid from these renewable energy sources becomes much higher? The harmonics and voltage spikes produced by power electronics might be a problem then. Please see the discussion in the Energy Pulse article "Distributed Energy Resources: Why IEEE 1547 won't be the Last Word", by Roger Arnold, http://www.energypulse.net/centers/article/article_display.cfm?a_id=944