PEE is aiming to establish which of the six programs is the most accurate. The greater the accuracy the lower the cost to wind developers of deviating from their scheduled production. Companies taking part in the research effort include wind plant operators, forecasting specialists, and the national electricity market operator, Operador del Mercado Eléctrico (OMEL). OMEL will help PEE establish the commercial implications for wind plants using different forecasting models.
PEE has applied to the EU's Intelligent Energy initiative to allow it to extend its forecasting research to France, Germany, the UK and the Netherlands, each involving the respective national market operator. The proposed starting date of the EUR 300,000 Intelligent Energy exercise is January 2005.
According to PEE, its research will mark the first time that a range of forecasting models will be subject to systematic and direct comparison. PEE's Alberto Ceña says that until now wind forecasting has mainly been used by grid operators to help them balance the variable input from wind plant with generation from other sources. The new research will reveal the economic value of wind forecasting for wind plant owners.
PEE's initiative compliments a EUR 4.3 million European forecasting exercise called Anemos, involving 29 companies and institutions and covering six wind plants, two in Spain. Anemos mainly involves representatives from the global meteorological industry and aims at developing new and improved forecasting models for wind plant, says Ignacio Martí of Spain's Centro Nacional de Energías Renovables (CENER). "PEE aims at immediate benchmarking; Anemos at building know-how and software over four years," adds Carlos García from Spain's energy efficiency agency.
"The [PEE] program will give us a five-year leap in forecasting knowledge," says Ignacio Lainez of Dutch utility Nuon, whose Spanish wind affiliate Desarrollos Eólicos is participating. If wind plant operators were to individually try out various models it would take five years to compile and process meaningful data equivalent to what PEE hopes to come up with in a year's time, he adds.
"It remains to be seen, but less sophisticated models might be enough to operate profitably, in many cases, given the nature of the electricity market," says OMEL's Pedro Basagoiti. Although power producers in Spain must lodge a 24 hour schedule, 36 hours ahead of production, they also have six openings on the secondary market on the day of production where adjustments to schedules can be made up to an hour ahead of production. "It's much easier to forecast six hours ahead than 24 hours ahead," says Basagoiti. PEE's midway report in December should give a preliminary indication of which models might be sufficient for wind plant operators to avoid costs incurred for not delivering power as promised on the wholesale market.