During October the hybrid wind-hydro Gorona del Viento (GdV) plant achieved 19.8% renewables generation, higher than the 13.5% achieved in September 2015 but barely a third of the 58.2% achieved in August 2016. The cause was an abrupt mid-month degeneration in wind conditions (renewables generation averaged 32.4% in the first half of October but only 7.7% in the second, and for a total of 199 hours during the month GdV generated no electricity at all). Percent renewables generation since full operations began at GdV in June 2015 now stands at 38.8%, down from 40.1% at the beginning of the month. This represents only 8.9% of El Hierro’s total energy demand if we accept that electricity supplies only 23% of El Hierro’s total energy. Data on GdV plant layout, operation and capacities are given in the September 2015 review. Previous posts on GdV are accessible through the El Hierro portal
Figure 1 plots daily average percent renewables generation since project startup on June 27, 2015 (data from Red Eléctrica de España (REE). The abrupt decrease in renewables generation in October 2016 is apparent:
Figure 1: Daily mean percent renewables generation since June 27, 2015. The monthly divisions are approximate
The following Table updates the monthly grid statistics. Note that a column showing GdV generation as a percentage of El Hierro’s total energy demand has been added. The numbers in this column assume that electricity generation fills 23% of El Hierro’s total energy demand in each month, which is in line with the 20% estimate published in a pie-slice chart by Climatique for all the Canary Islands. The 23% estimate is, however, not verified, so the numbers in the column are italicized to signify uncertainty:
Figure 2 plots the REE 10-minute generation data for October. Surplus wind is still being wasted by pumping water uphill and hydro is still being fed to the grid in small quantities. There were no periods of 100% renewables generation during the month:
Figure 2: 10-minute REE grid data for El Hierro, October 2016
Figure 3 compares 3-hour wind speed readings at El Hierro airport with GdV’s total wind generation during October (10-minute data). In this case wind generation is perhaps not as high as the airport wind speeds suggest they should be in the second half of the month, and the most likely reason for this is changes in wind direction. Wind directions were dominantly northerly in the first half of the month but ranged over all points of the compass in the second, with frequent rapid changes in wind speed. Periods of higher wind speeds and higher wind generation during the second half of the month also tended to correlate with periods of northerly, and in some cases southerly winds, indicating that the GdV wind turbines, which are aligned east-west on a ridgetop, perform more efficiently when the wind comes from the north or south.
Figure 3: Gross wind generation versus wind speeds at El Hierro airport, October 2016
There is little to discuss here other than to speculate on what the wind might do in November. Accordingly I have performed a brief analysis of wind curtailment, the results of which are discussed below.
Wind curtailment at GdV occurs in two stages. The first is curtailment above the thresholds that have been adopted to maintain grid stability, which are generally in the 5-7.5MW range but sometimes lower. It is not possible to estimate how much wind power was curtailed above these thresholds, but in windy months like July 2015 it would have been substantial:
Figure 4: GdV wind generation, July 2015. Wind is curtailed to an unknown extent above all the flat-topped intervals
The second curtailment stage involves wasting surplus wind power by using it to pump water uphill and allowing the water to flow down again without generating any electricity in the hydro plant, meaning that the hydro pumping circuit is functioning as a dynamic resistor. Here the wind losses can be estimated from the REE grid data, and the results by month are summarized in the Table below. The wind power used to produce hydro power is calculated assuming that the hydro system is 60% efficient:
Average pumping curtailment has been 30.4% since project startup. It’s easy to see how the curtailment added by generation thresholds such as those shown in Figure 4 could have increased total curtailment to 50% or more in some months, but unfortunately it’s not possible to come up with hard numbers.
Finally, Figure 5 plots percent pumping curtailment against total monthly wind generation since project startup. Curtailment increases with generation, as we would expect, and at a fairly high level of correlation (the R squared value of the trend line is 0.77).
Figure 5: Percent pumping curtailment versus total wind generation by month since June 27, 2015