Reading a Wind Developer’s spin is sometimes amusing and always infuriating. So it was with a laugh and a grimace that I read the following chart and comment in BluEarth’s April 2013 newsletter. Even they are surprised that wind plants produce electricity. I was annoyed that they apparently take their readers to be equally simple.
“Of course it is not always windy, but it is quite amazing when one looks at operating wind power facilities in Ontario how consistently wind plants produce power throughout the day and night. In response to some questions posed to us about whether it was windier in the night time versus in the day time, we completed an analysis of available wind data.
We examined the publicly available hourly power production data for all operating wind farms in Ontario from January 1st, 2011 through to December 31, 2012. This involved over 200,000 data points providing a good sample to look at. We then mapped the production data by season and time of day. The results are presented in the table below.
What we find is that overall, the production of wind power in Ontario is quite evenly distributed across the day and night. There is a small dip in production on summer mornings. What is also clear, and readily known, is there is seasonal variability in wind production, with less wind in the summer months compared to the winter months.” (emphasis added)
At first I thought the numbers in the final column referred to the percentage of annual generation each season represented, except the numbers add up to 101%! Another example of how efficient Wind Generating Stations are? New Math?
“If you torture the data long enough, it will say anything you want.” say the statisticians. These claims of ‘consistent’ wind power production ‘evenly distributed’ across day and night is a case in point.
The use of “average” production is problematic as averages are, by definition, meant to smooth out the variations in a series of numbers. But averages can be deceptive and of limited value. Try telling a Police Officer who pulls you over for doing 130Km/Hr on the Hwy. that he shouldn’t have stopped you as you have been only averaging 90Km/Hr on your trip. The series of numbers 3, 19, 37, 111, 10, 0 averages 30, however you wouldn’t call the numbers consistent or evenly distributed. The use of the average smooths out the fluctuations, in this case hiding them. Given a random event like wind, the more data you have, the more you smooth out the variations. Using 2 years of Wind production helps to smooth the fluctuations further.
It is always wise to start with a look at the raw data, unmolested by averages and other statistical manipulations.
I happened to be analyzing the same data when this newsletter came to my attention and here, in Charts 1 through 4 are the actual production values for the 2011 seasons, also taken from the IESO’s website
That is not what I call consistent or evenly distributed generation across the day and night. Chart 5, the 2011 Total Annual per Hour Wind Generation, shows how averaging a larger data set smooths the fluctuations yet further. Note however that the characteristic pattern of Wind Generation is exhibited with the greatest production between 8PM and 7AM, overnight in other words, and least production during the day between 7AM and 8PM:
The actual data represented in these charts paints a different picture, one of constantly fluctuating production. The same pattern can be seen in all seasons and for the year as a whole. On average power output drops approximately 15% between midnight and 10AM and doesn’t recover till 9PM. As Chart 6, April 2011 Wind Generation vs. Ontario Demand shows, wind power generation is poorly matched to demand, tending to randomly decrease as demand increases and increase as demand decreases.
Even the summing of hourly production figures required to produce these seasonal charts tends to smooth the fluctuations in Wind power generation. In Ontario the IESO (Independent Electricity System Operator), which runs our electrical grid, only makes hourly wind performance data available. Other countries, such as Ireland, make Wind production data available for shorter time frames. The variability of Wind power is even more evident when looked at over these short time frames. Chart 7, below, is an example taken at random which shows a drop of 500MW over 5 hours, an 83% drop in output and an increase of 760MW over 6.5 hours, a 475% increase in output.
The wind industry claims that wind forecasting enables the grid to compensate for these swings in output. However, to date, comparing the IESO’s forecasts to actual performance doesn’t justify this assertion at all. The wind is still unpredictable and constantly variable. What this means is that conventional generation such as Nuclear, Hydro, Gas and Coal must be kept running to supply electricity when it is needed and as Wind fluctuates.
(Note: Above, Wind scaled up by 8x to chart)
No one denies that Wind Turbines can generate electricity, that is not the issue. The issue is that they cannot generate electricity on demand, reliably, constantly and predictably and they usually generate the most electricity when we need it least and due to the need for backup they don’t reduce our fossil fuel consumption or our CO2 emissions.
This results in 80% of the electricity generated being unusable and sold on the export market below cost, sometimes we even pay others to take it from us. This problem will only get worse as Ontario adds more Wind to its grid. What is more, Ontario presently exports about 2,000 MWh each and every hour of the day. We have a surplus of generation capacity and don’t need the extra generation.
One has but to look at the mess that the German grid is in to see how large amounts of wind power can destabilize an electrical system and cost the ratepayer many times more than they would otherwise pay for electricity produced by conventional means.
Then there is the problem of energy sprawl…