Lynne Kiesling
I am enjoying the conversation in the comments on my prior post about wind subsidies. I forgot to include in that post a link to John Whitehead’s post on wind farms at Environmental Economics, building off of a recent John Tierney column in the New York Times.
Tierney points out that, as some commenters on my prior wind subsidy post observed, fossil fuels and nuclear energy receive subsidies too. He cites evidence that renewables receive twice the subsidy that they do per unit of energy.
The biggest problem with windfarms (as Germany and the Netherlands found out) is that you still need just as many traditional (coal, nuclear or natural gas) plants as before in order to deal with wind fluctuations. The problem is especially acute with coal plants – you can’t ramp up their power output instantly unless you already have a full head of steam.
Once you get into the subsidy game you end up subsidising both.
Although wind turbines are not dispatchable, wind variability or intermittency is far from the unsolvable obstacle the previous commenter suggests. I want to be sure simplistic misconceptions aren?t reinforced here.
First, wind turbines are likely to be geographically dispersed with varying orientations and surrounding topographies. The aggregation of many wind turbines therefore significantly reduces the volatility of output.
We must also remember that the electric power network is designed to accommodate fluctuating loads. Demand isn?t very dispatchable (yet) either! While new grid management strategies are necessary in areas where wind exceeds 5-10% market share, I would suggest this obstacle is being sufficiently addressed.
Many of today?s wind turbines have sophisticated electronic controls that actively contribute to grid reliability by contributing reactive power, controlling voltage, riding through outages, and communicating with grid operators.
Weather forecasting models are also currently being used to assist proper load following and grid management. These tools make options such as operational reserve control and demand response control more effective.
Studies from Xcel Energy, PacifiCorp, BPA, We Energies, and the Utility Wind
Interest Group (UWIG) report ?the need for additional generation to compensate for wind variations is substantially less than one-for-one and is often closer to zero.?
For more info see the International Energy Agency?s report on wind variability at http://www.uwig.org/IEA_Report_on_variability.pdf
I would be interested to see further evidence of the problematic experiences in Germany/Netherlands to which Nordic refers.
“First, wind turbines are likely to be geographically dispersed with varying orientations and surrounding topographies. The aggregation of many wind turbines therefore significantly reduces the volatility of output.”
The industry suggests that the next generation of wind turbines will be 5mW machines with 35% availability. Twenty of these machines, properly distributed as suggested above, would provide: 100mW at 35% availability; or, 5mW at 99.99% availability and greater outputs at progressively lower availability within the range from 5-100 mW. Proper distribution increases the availability of any fraction of the combined capacity of the machines, but not the availability of the total capacity.
Yes. But my point is volatility of AVERAGE output (against which the grid operator plans) is significantly reduced.
I’m not saying 100% of total “nameplate” generating capacity is available more often. I’m saying the magnitude of fluctuations in AVERAGE output (the fraction of total output) is less.
Read the EIA report to see data and figures showing this smoothing effect.
Derek, I can tell you that in my experience we need significant backup reserves to our wind projects. Hydro is our best backup, in that in that its very easy to ramp up and ramp down, but of course it’s also our lowest cost resource with means we sacrifice some very low cost power to backup the high cost wind resource (another cost of wind power). A major limiting factor to expanding our wind projects is sufficient backup generation. We can use our gas turbines… but they are of high heat-rate and constant ramping up and down beats the hell out of the machines. I don?t see how wind can be much more than a small portion of the countries? energy portfolio unless we devise efficient ways to store electricity.
And yes Derek, it’s true power grids can handle some degree of variability… but our wind facilities can go from well over 100MW to close to zero MW within the hour… all day long. That is an expensive amount of variability to manage?. especially if your wind resource is outside of your system (in which case your going to have to pay all kinds of charges to the transmission provider for going over or under your expected generation).
Even with these limitations however, there?s still a lot of room for wind power to grow beyond it?s current use in the US.
It would appear that the limit to “source of opportunity” wind power fraction is the conventional capacity reserve margin in the affected market, unless storage is available.
It would appear that the limit to “source of opportunity” wind power fraction is the conventional capacity reserve margin in the affected market, unless storage is available.
Ed Reid posted:
“The industry suggests that the next generation of wind turbines will be 5mW machines with 35% availability. Twenty of these machines, properly distributed as suggested above, would provide: 100mW at 35% availability; or, 5mW at 99.99% availability and greater outputs at progressively lower availability within the range from 5-100 mW. Proper distribution increases the availability of any fraction of the combined capacity of the machines, but not the availability of the total capacity”
This is a misunderstanding of the concept of “capacity factor” as it applies to wind turbines. It is NOT “availability”. Turbines typically are available at near 100%.
A nuclear plant, for example, is essentially a “binary” device. It is either producing 100% of its capacity or zero. A reactor running at say 50% is a sick reactor.
What this means is that a reactor with a 90% capacity factor is cold and dead – unavailable – for more than a month per year, often requiring a thousand megawatts of backup power.
A wind turbine is completely different. Here is an analogy: Consider a car engine. It is “nameplate” rated at 100 horsepower. The actual typical demand on the engine may be only 10 horsepower. Only for passing or big accelerations does it reach near 100 HP.
Running at a 10% capacity factor DOES NOT imply that it is unavailable 90% of the time.
A wind turbine is the same. Its “nameplate” rating is sized for the occasional gusts of high winds that carry lots of valuable energy. Ordinarily, the winds generate at 30-40% of maximum capacity. The turbine, if downsized to produce 100% at average wind speeds would then waste the power of the gusts.
The key point is that a wind farm operating at 35% capacity factor may, in principle, be available 100% of the time producing a steady 35% of rated capacity. Or it may fluctuate widely. That is not an inherent feature of wind turbines, but of individual sites. Off-shore farms may be very consistant, while desert sites may be quite variable.
Do not misapply the concept of capacity factor.