The below studies ARE based on 1/4-hour and 1-hour data published by grid operators.

http://theenergycollective.com/willem-post/57905/wind-power-and-co2-emissions
http://theenergycollective.com/willem-post/64492/wind-energy-reduces-co2-emissions-few-percent

On the reserves question the wind generators are getting a true bargain at $1.29/kw-month. PJM’s capacity auctions for this year established prices of roughly $3.25/kw-month. Most other pools were higher, some more than twice the PJM figure.

Finally, price collapses in spot markets have been common in NW Europe when the wind roars on the North Sea. With the exception of Norway with its enormous hydro capacity, there is simply no way to buffer the changes in wind output in Germany. However, prices to consumers do not reflect such fluctuations and industrial customers or distribution companies cannot sign contracts for firm delivery of wind energy.

It is not clear that wind integration has supplied any useful electrical energy to Germany in the past several years. We will have a post on this very subject shortly.

Here are two previous posts on MasterResource that illustrate the level of complexity involved – http://www.masterresource.org/2010/01/how-many-households-can-a-large-wind-project-serve-lessons-from-texas-and-the-uk-part-1-of-2/ , and http://www.masterresource.org/2010/01/how-many-households-can-a-large-wind-project-serve-lessons-from-texas-and-the-uk-part-2-of-2/ .

The comment about improved forecasting is also questionable. It is difficult enough to forecast day-ahead weather on an hourly basis, let alone specific local wind conditions. Even if forecasting was perfect, this would not remove the need for other generation to balance the random and often large variations on a sub-hour basis. Even a major wind proponent, Gross, acknowledges this and maintains that wind balancing needs are separate and addition to normal reserves. Gross does suggest that this does not introduce significant concerns but admits that the studies to properly evaluate this need to be conducted.

I will not add to the comments of Jon Boone, Don Hertzmark and Tom Tanton, who have pointed out some of the realities involved, for example, important errors in understanding the costs.

The claim that ”BPA does not back wind megawatt for megawatt. For the more than 2,800 megawatts of wind connected to our transmission system, we have set aside about 850 megawatts of clean hydro capacity to produce electricity if scheduled wind power doesn’t materialize” requires comment.

First of all this, comment demonstrates the misunderstanding that back-up is required only when wind does not materialize. It fails to show any appreciation of what happens when wind does materialize, and has to be balanced on a very short-term basis (minutes) on an ongoing basis. In this case what the BPA representatives are saying is that this is like balancing short-term variances in load with a capacity of only 30% of the variances. The often relied upon reply that geographic dispersion improves wind reliability has been shown to be not valid.

On the subject of wind balancing requirements, there is sufficient experience in Germany that wind balancing capacity in the order of 90-95% on a statistical basis over a year is required (E.ON Netz). Of course this statistical averaging is for system wide planning considerations. What must be taken into account in real time (the world we live in) include (1) grid limitations, and (2) wind variation over most if not all of the total wind capacity. Further, deNet (a consortium of 100 German research institutions and services providers) acknowledges this as well, based on the dena grid study.

We need full, transparent information on the behavior of the generation means within a specific grid structure as a result of the presence of wind, especially as the wind penetrations increase.