On 12/29/06, Eugene Coyle <[EMAIL PROTECTED]> wrote:
I did read the NY Times article by Mathew Wald as having a strong negative perspective. It is true that wind doesn't blow all the time, hence other capacity is needed to serve load when wind isn't blowing. But other types of plants have down time. Nuclear units are unavailable for months at a time for refueling -- that capacity has to be replaced when the plants are unavailable. And the stress on having stand-by units just to cover wind is wrong as well. An electrical load fluctuates diurnally as well as seasonally and utilites have dealt with that for a century. This is not to say that integrating wind into a system is without problems. But the industry is learning new tricks, and incorporating new technology to redce the problems. All in all I think Wald's story emphasized the negative too much.Gene Coyle
The point for this argument: If wind supplies up to 20% of power is pretty widely acknowleged that variability does not matter. The spinning and operating reserves you have anyway will cover what is needed for wind, possibly with trivial extra expenses for management and tracking wind data. A point I've just learned about, that is controversial but which I'm prepared to argue: If you connect distant wind farms to one another with long distance transmission lines (using HVDC to go really long distances and bridge grids with different frequencies) storage requirements become low enough that the limitation dissapears you can let wind power dominate the grid. Let me quote from something I'm going to publish in a few days: ============================================== In March of 2006, Windtech International magazine published an article on Vehicle to Grid interconnection[2], which included a chart based on unpublished wind data from a study covering eight sites[3]. The V2G article asked what percent of the time interconnected (hypothetical) wind farms could guarantee 20% or more of maximum rated capacity if located at these sites. Without storage, the answer was 80% of the time. But most of the periods where such a promise could not be met lasted three hours or fewer. Three hours of storage compared to that 20% is 36 minutes compared to the theoretical maximum generators could produce. (We compare to this maximum because it is the way wind capital costs are measured.)With that amount of storage eight wind farms could meet a 20% firm capacity 90% of the time. However, 20% of theoretical maximum represents around two thirds of average wind production. If we have to throw the remainder away, we increase costs per kWh by more than a third. Because three hours storage are already specified, and some extra power would be needed when produced, we don't lose the full third. But to tap all or most we would need to store additional hours. Ten hours compared to that 20% commitment would seem a pessimistic guess - and represents only two hours compared to maximum rated production. It allows full capture of unused excess, and time shifting of power to when it was needed. As a happy side effect, it would increase reliability to 95%. So, wind alone could produce 95% of the power of any nation with extensive distributed wind capacity. And note that this conclusion is based on data from only eight sites. Wind from one hundred or one thousand dispersed locations will be much more reliable. Both the U.S. and the U.K. get about 4% of their electricity from conventional hydropower, which can work very well in "shaping" other power sources. This leaves 1% to obtain from somewhere else, such as biomass. Note that even burning conventional carbon-containing natural gas to provide 1% of power would leave the grid better than 99% decarbonized. That leaves one problem : operating reserves. The problem with the scenario I've laid out is that the final 1% will occur in large blocks - more than hydro shaping can make up for. That means that on occasion backup will have to supply all or most power, when not enough wind is blowing, and stored power is exhausted. This does not represent significant fuel consumption, but it represents a huge amount of capital investment to cover around 88 hours of power per year. Standby diesel plants in the multi-megawatt range costing $200-$300 per KW would be the cheapest solution if we were starting from scratch. Lifespans can be as low as 1,500 hours; but even that would be more than 15 years with this degree of usage. In quantities this small we may be able to produce biodiesel without destroying rainforests; if not diesel turbines can run on 70%-80% natural gas. At any rate capital costs should be around .7 cents per kWh spread over all electricity consumption. Annual fuel and maintenance costs for these standby plants should be trivial in comparison. [Lastly as to type of storage - while pumped storage is cheapest, I favor vanadium flow batteries - made by Candian headquartered VRB, because they avoid the horrible ecological consequences of flooding huge areas for pumped storage purpose. With only two hours of storage compared to nameplate capacity, they bring the cost of a 95% wind grid up to 10 cents a kWh, more than many pay, but by no means the maximum price paid for electricity in the U.S. I believe there are parts of Florida that pay more than that. And I dimly remember Manhattan prices being in that range or above. Doug, how does 10 cents a kWh compare to what you and Liza pay?] ........................................................ Notes [2] Willet Kempton and Amardeep Dhanju, "Electric Vehicles with V2G Storage for Large-Scale Wind Power,". Windtech International Mar 2006, (accessed 27/Dec/2004) <http://www.udel.edu/V2G/docs/KemptonDhanju06-V2G-Wind.pdf>. Figure 2. [3]Cristina L. Archer and Mark Z. Jacobson, "Spatial and Temporal Distributions of U.S. Winds and Wind Power at 80 m Derived from Measurements,". JOURNAL OF GEOPHYSICAL RESEARCH 108, no. D9 16/May 2003, (accessed 27/Dec/2006) <http://www.stanford.edu/group/efmh/winds/2002JD002076.pdf>.Previously unpublished data in the V2G article had been compiled for this study.
