Variability not such a big problem -- Alberta revelations

Managing fluctuations in wind power generation in the Canadian province of Alberta is likely to be far less costly than estimated a year ago by ABB, a multi-national electrical equipment supplier. In a study for the Alberta Electric System Operator (AESO), ABB claimed that to accommodate the expected variability of output from 1200 MW of wind on the province's power grid, regulating reserves would need to increase by 120-481 MW and contingency reserves by 375-805 MW. The added cost to the system, it said, could be between C$91,000 and C$370,000 a day for regulating reserves and C$45,000 for contingency reserves.

The findings were hotly disputed by the wind industry, which pressed the AESO for a new study. Twelve wind developers took the unprecedented step of handing over confidential monitoring data to help demonstrate that ABB's estimates were out of synch with the growing body of experience in other markets. "We absolutely had to do this and we had to do it not in a half measure kind of approach," says Vision Quest Windelectric's Kevin Van Koughnett. "We had to go and absolutely nail this and put it out of everyone's mind."

Analysing data

The result is a new study, commissioned and just released by the AESO, demonstrating that the relative variability of wind generation in the province will decrease as production facilities are spread over a wider area. Conducted by Alberta's Phoenix Engineering, it looked at wind data for 2004 from existing wind farms and from monitoring towers scattered through the southern part of the province, creating three future development scenarios with 895 MW, 1445 MW and 1994 MW of wind. At the time of the study period, Alberta had 254 MW of wind generation.

The analysis found that that the magnitude of power fluctuations caused by wind speed variations was seldom extreme, estimating that 97.5% of the ten-minute fluctuations for the three scenarios were less than 40 MW, 56 MW and 70 MW, respectively. As a percentage of total capacity, the fluctuations decreased as more wind was added, from 7.4% for the existing wind farms, 4.5% when installed capacity jumped to 895 MW, 3.9% at 1445 MW of wind generation and 3.5% at 1994 MW.

"Our contention always was that when you had a geographic dispersion or spreading out of wind farms it would wash out a lot of the variability, and that is very evident from this study," says Van Koughnett.

Fluctuations

The study also estimated power fluctuations in a one-minute timeframe and found that 97.5% were less than 9 MW, 12 MW and 14 MW for the three scenarios. The maximum fluctuations were 116 MW, 149 MW and 222 MW over the one-minute interval. Over ten minutes, fluctuations were as high as 202 MW, 208 MW and 259 MW.

The AESO will now use the study's findings to examine the effects of wind's variability on the operation of the province's electricity system. Using load data from 2004, the AESO will redispatch the entire system as if it had the wind capacity contemplated under each scenario available to it, allowing it to assess the level of required reserves. "We really don't know what is going to happen on this next part, but we're feeling pretty confident that the impact should be reasonable," says Jason Edworthy, also of Vision Quest.