October 29, 2009
Wind Turbine Transmission Technology
Recently a commenter asserted (with no evidence - correction: some evidence but not quantitative) that wind power has big reliability problems due to blades and the mechanical assemblies suffering high failure rates. No quantitative data was provided. This got me wondering if there is any publically available data on mean time between failures for assorted wind farms. Anyone know of such sources? Or are the operating companies tight-lipped about their problems? An article in Technology Review about the development of continuously variable transmissions for wind turbines mentions a number of hours that wind turbines are expected to operate: 80,000 hours.
The question is whether the CVT is tough enough. Viryd parent company Fallbrook Technologies has already commercialized its technology as a smooth-shifting alternative to gears and derailleurs in high-end bicycles and is working on larger vehicle applications. Wind power, however, is a particularly demanding application, according to Jason Cotrell, a senior engineer at the Department of Energy's National Wind Technology Center in Golden, CO. "Wind turbines are subject to very high torque for 80,000 hours of operation, so it's a very challenging environment," Cotrell says. "CVTs tend to be complex, and we haven't yet verified that they're suitably robust."
Well, there are 8760 hours per year. A wind farm that operated every hour of the year would run up 80,000 hours in about 9 years and 2 months. But if a wind farm only spins, say, 30 percent of the time then 80,000 hours of operation would occur over a period of 30 years.
I would be very surprised if a wind turbine's transmission could run for 80,000 hours without service. The spinning, vibration, temperature variations, humidity, and other stresses must take their toll. So do wind turbine transmissions last 80,000 hours before total replacement? How much maintenance gets done before total replacement? Do the transmissions or blades tend to fail first? Anyone know?
While I'm asking: What's more expensive, the blades, the transmission, or the generator?
"With no evidence"????
I posted this:
"Despite reasonable adherence to these accepted design practices, wind turbine gearboxes have yet to achieve their design life goals of twenty
years, with most systems requiring significant repair or overhaul well before the intended life is reached
[3,4,5]. Since gearboxes are one of the most expensive components of the wind turbine system, the higher
than-expected failure rates are adding to the cost of wind energy."
Annual cost of maintaining a wind farm is roughly 1-2% its cost to build.
I am by no means an expert, but I've been lead to belive that the blades (modern ones) do not require any significant / expensive maintainance. The gearbox on the other hand seems to bear the brunt of the load and is where most R&D dollars are spent on reliability.
BTW: typo in the post -- 80,000 years vs 80,000 hours in the second mention.
There is a gottcha in most claims for reliability. And that is the distribution of times/cycles/miles to failure. If the distribution is exponential then a claim of 80,000 hours Mean Time to Failure means that 63% of the units will have failed with less than 80,000 hours of service. However, for most mechanical systems times to failure are described by a Weibull distribution. Absent the details about the parameters of the distribution the claim is at best ambiguous.
There are too many dubious claims being made for wind power by academics and researchers who have no field experience nor actual work credentials in the field. Ivory tower pukes are the least reliable source for information.
Wind power sucks from start to finish. Unreliable, expensive, prone to failure on many fronts from gearbox to propeller to bearings. You want documentation? Go where the wind blows and you'll find it. Wishful thinking is going to cost the US trillions more than it needed to spend. Call it a junk tax on stupidity.
I am looking for quantitative evidence: MTBF, cost of repairs. Something that would give us a quantitative sense that bearings and gears failures are substantially increasing the costs of wind power. You can always find problems in any industry. But what do they cost? Preferably in fractions of a penny per kwh.
"Theoretically, offshore wind should be a low risk investment, in that fixed costs represent a high proportion of overall costs. This provides a level of certainty which, combined with guaranteed tariffs, makes it particularly attractive during times of volatility. But offshore installation is roughly 50% more expensive than for onshore, and O&M costs are roughly twice as much."
"This is underlined by an analysis of maintenance records, which shows that while service teams for offshore wind farms are supposed to make two scheduled maintenance visits every year, unscheduled visits to many installations are made 20 times a year."
"The classic example of this is the disaster at the Horns Rev wind farm in 2005, following which Vestas is reported to have removed and repaired 80 of its V90 models, designed for offshore use, owing to the effect of salty water and air on the generators and gearboxes, which became corrupt after only two years. A similar procedure has been reported this year, with Vestas' 30 turbines requiring a change of rotor bearings, at an estimated cost of €30m."
Some of the blade makers have had cracking problems. I know someone with a close enough connection to a wind turbine maker to know this. I never come across it in my reading. Not sure to what extent it is still a problem. Last I heard the problem had become small enough that they were worried more about other stuff.
In a free market high proportion of fixed costs are a negative, not a positive.
The Horns Rev wind farm was the first off shore wind farm in the world. For a world first it was problemless.
A summary of Wind Farm failures. Details in pdf.
"By far the biggest number of incidents found were due to blade failure. "Blade failure" can arise from a number of possible sources, and results in either whole blades or pieces of blade being thrown from the turbine."
Offshore wind is more expensive for 4 reasons:
1) The tower structure is more expensive to build.
2) The equipment must handle the much harsher marine environment. As your article demonstrates, the industry is still struggling to achieve high reliability offshore.
3) It is more expensive and dangerous to go out to the wind towers to do maintenance.
4) Getting the electric power to shore costs more money.
However, your article paints a much more rosy view of onshore wind farm reliability:
Failures are also harder to repair because they tend to happen in stormy conditions, and are often not dealt with when they happen, but on an aggregated basis at intervals. That means it can be as long as three months before a turbine failure is repaired. The contrast with onshore reliability is dramatic, and availability levels of 97% are regularly achieved.
Let's be honest. It doesn't matter whether wind farms work or not. They are simply a convenient (though expensive) large scale project for governments and inter-governmental agencies, so that they can claim to be doing something to solve a problem which may or may not truly exist. As long as politicians can throw money into the project (much of which will end up in the hands of friends and campaign contributors) they can claim to be doing something to save the world. And they add to the GDP, whether they ever provide reliable energy or not. In terms of GDP, it is better if they do not work, since some other expensive large scale project will then be required. More chances to skim off the top for friends of the powerful.
Later on, in the cleaning up stage of the aftermath, fingers can be pointed and perhaps a few persons will go to prison. But most of those involved are far too high up on the food chain to be indicted, or even publicly accused. Those foolish enough to have been cheerleaders for the debacle will try to forget their former stupidity, perhaps taking cold comfort in the fact that they were not alone.
I am reluctant to call the Danish insane. ;>
Political groups at northern latitudes who oppose fossil fuels usage and who also oppose nuclear power and who also do not have much hydro power (and curiously the Swedes have nukes and lots of hydro) obviously can't use much solar power. The winters are too dark and their electric demand winter peaks (Maine, Canadian Maritimes, Scandinavia, and most or all of northern Europe has peak electric demand in winter).
They aren't left with a lot of options. Northern Europe is densely populated. Plus, the people like scenic mountain views uncluttered by wind farms (and so do I fwiw). They end up with offshore wind and geothermal. But fears that geothermal will cause earthquakes puts them back at just offshore wind.
They've put themselves in a box.
Angela Merkel wants to delay the phase-out of existing German nuclear power plants. She'd even like to build new ones. Maybe she'll pull this off because, yes, offshore wind is problematic.
On the other hand, maybe the Danish wind firm Vestas will solve the offshore wind problems. Though even if they do one of the reasons Denmark can get such a high percentage of its electric power from wind is that it can buy electric power from neighbors when the wind doesn't blow.
I get a kick out of the whole thing. I know a guy in Germany who is paying the equivalent of 35 cents/kwh. I told him what the people in the Dakotas, Wyoming, and other cheap places pay and he put his head down on the table. Poor guy. He knows the German government's policies are stupid.
I wonder if the 33-40 GW output is nameplate or average real output. Also, how much of that 33-40 GW total power comes from the £100bn to be spent offshore? The article is not clear. But they are saying at least £3bn/GW for offshore wind. Probably much higher.
But suppose you had £100bn to spend. That's about $163 billion at the moment. This post by David Bradish of the Nuclear Energy Institute has some good numbers on nuclear power costs. A nuclear power plant might cost $8 billion to build and that's probably about 1.5 GW. So that £100bn could probably pay for 30 GW of nuclear generation capacity that would operate about 90% of the time.
How much nameplate wind power could be built offshore for the same amount of money? What would be its average output? Depends on the winds and also on reliability of the equipment.
I meant to add in the link above that wind turbine o+m is about twice what it is for coal and nuclear, what with capital costs for wind already 10 times per mw that of it's next closest energy source, whats another huge cost to government loving jerks, what matters most are not results, but soviet style central planners screwing things up.
A few thoughts:
Analysis of European renewable energy policy is interesting, but not very relevant to the US: the US has far better wind & solar resources.
European electricity was twice as expensive as US electricity long before renewables were added.
You can't compare US & European prices with a simple currency conversion: that exaggerates European prices. The PPP adjustment is roughly 30% lower.
European wind reduces the premium for power in the hourly markets, and arguably pays for itself.
European advocates for renewable energy generally regard the early premium for wind & solar as an investment in building economies of scale - we should thank them for this gift, which we see paying off in PV right now.
European antipathy for nuclear, IMHO, comes from it's association with weapons, and has some justification.
Wind O&M (onshore) is much lower than nuclear, and that needs to be included in lifecycle comparisons.
Wind capital costs in the US are about $7.50 per average watt (including conecting transmission), and has O&M of about $.005. $8B for 1.5 nameplate GW, and 90% capacity factor gives about $6 per average watt. When you include nuclear's $.02 O&M, wind is roughly at the same cost per KWH.
Wind can be built much more quickly than nuclear. The $8B is overnight costs - the additional 6-8 years to build nuclear must add at least a 10% premium in interest costs and project cancellation risks, not to mention the opportunity costs from CO2 emissions that are contributing to unpredictable tipping points/positive feedbacks in climate change.
A few thoughts:"
Thought 1: "we have better wind and solar" US wind is rated at its best locales at less than 10% reliability, so how much better is the solar? And substantiate it with facts please, not lefty blog BS only san franciscans fall for.
2: Kinda like california, the AGW lefty foolishness is costing both of them.
3: No wonder this bothers you, that makes the US costs 30% higher, if you have any real evidence they aren't, let's see it. Shenyang group from china is supplying all the turbines to that huge new West Texas wind farm, that would suggest to you that manufacturing costs here are higher? so if you're going to cherry pick, prove that isn't the same with all other.
4: Provide something other than your thoughts, Denmark, the biggest of all European windturbinesuckers has the highest electricity prices of all of Europe.
5: European advocates for renewable energy? Political groups aren't reliable sources until they reflect your biases where they become an authority at which point you then assume everyone has the same biases.
6: Thats the stuff of hollywood, however the opposite has happened, Soviet weapons nuclear material was sold to the US following their collapse and was used for power consumption.
7 thru 9: prove it, these are O+M for nuke and coal
Parker, I'm working on a poem that goes ..nickg is as weird as poet and we all know it
Nick, This seems high:
Wind capital costs in the US are about $7.50 per average watt (including conecting transmission)
Is that per average watt of nameplate? If so, don't you have to more than triple it for average generated?
Isn't the most expensive part of wind power that the wind sometimes doesn't blow?
My family toured America by RV this year. We encountered a number of wind farms, but only one time were they actually spinning.
When I say average watt, I mean nameplate of $2.25/W ($2 for windfarm, $.25 for connecting transmission) divided by capacity factor of 30%.