In my last post on Gorona del Viento (GdV) I brought comments to a close with the portentous words: “new information that greatly complicates the picture has come to my attention and we are not going to get much farther until this issue is resolved.”
Well, it turns out that this was a bit of an overreaction on my part. The new information in fact confirms what we were thinking all along – that GdV is indeed pumping water uphill and allowing it to flow back down again without passing through the hydro turbines. Here’s a brief review of the evidence:
We begin with the Red Eléctrica de España (REE) “hydro” values. These tell us how much energy has been consumed pumping water up from the Lower to the Upper Reservoir, with negative numbers showing net pumping and positive numbers net hydro generation. Here we will look at the period between February 20 and March 10, 2016. According to the REE data some 720MWh were consumed pumping water up from the Lower to the Upper reservoir over this period, enough to lift roughly 250,000 cu m of water up the hill (Figure 1):
Figure 1: Cumulative pumping from lower to upper reservoir, February 20 to March 10, 2016
So we would expect to see a large increase in the level of the Upper Reservoir over this period, correct?
But we don’t. Here are two images of water levels in the Upper Reservoir taken by Rainer – our man on El Hierro – at 0837 on February 20 and 0902 on March 10. The water levels are essentially the same. There is no sign of the 250,000 cubic meters that according to the REE data were pumped into the Upper Reservoir over this period:
Figure 2: Changes in water level, GdV upper reservoir, February 20 to March 10, 2016. Image credits Rainer.
So where did all the water go?
Well, if it didn’t go up it must have gone back down, and according to the photos Rainer took of the Lower Reservoir on February 20 and March 10 that’s exactly where it went. (Note that the water level increases between February and March in the lower reservoir. Presumably this is because of water added from the Tamaduste desalination plant, which reportedly produces about 1,200 cu m/day, or about 25,000 cu m in 20 days, This fits the increase in water level quite closely.)
Figure 3: Water entering GdV lower reservoir, February 20 and March 10 2016. Image credits Rainer
But there’s a problem here too. Rainer’s images clearly show water flowing down, but according to the REE hydro values water was being pumped up the hill at the time. How to reconcile these contradictions? Here’s how I see it:
1. Water pumped up the hill from the Lower Reservoir does not make it to the Upper Reservoir. As soon as it gets to the Caseta de Válvulas (Valve House, ringed in yellow on the graphic below) it’s sent back down again. (According to the Spanish text the Valve House is where “pumping accumulates surplus energy and functions as a generator”, although I’m not sure exactly what that means).
Figure 3: GdV schematic showing “Valve House”. Image credit El Mundo
2. The water flowing back downhill is metered only if it generates power in the hydro turbines, but most of it doesn’t, so it doesn’t get counted.
3. We therefore have a semi-continuous flow of water going up the hill to the Valve House and flowing back down again, wasting a lot of power but mostly without generating any.
Why do this? Because experience apparently shows that this is the best – and maybe the only – way of matching GdV generation to El Hierro demand while maintaining grid stability. Pumping is being used to “waste” surplus energy the grid can’t accept.
Here are the demand and generation data for February 20 to March 10 for reference:
Figure 4: Balancing demand by switching wind between the grid and pumping, February 20 to March 10, 2016
Some concluding observations:
The Upper Reservoir has evidently fallen into disuse, not that it was ever used much to begin with. The Figure 5 images show no appreciable variations in water level since 2013. One speculates that there might be an unresolved water dispute between GdV and the Island’s farmers behind this, but for all the use that’s presently being made of it the Upper Reservoir might as well not be there. Another problem is that if you look carefully you will see that the three images that cover the north end of the reservoir (2, 3 and 5) all show the drain barely covered by water. So how much usable water is there in storage in the Upper Reservoir? Would anyone care to make an estimate? (The upper reservoir contains 385,000 cu m when full).
Figure 5: Date-order images of upper reservoir water levels since March 2013.
A final question is how much water is stored in the Lower Reservoir. This is an even more important point because if pumping is needed to ensure grid stability then it’s a good idea to keep the Lower Reservoir full. But Rainer’s most recent photo, taken on March 17, shows the reservoir only about a third full, i.e. containing roughly 50,000 cu m. Limited water availability may in fact explain why GdV sometimes curtails turbine output at relatively low levels, but more work would be needed to confirm this.
Figure 6: Water level in lower reservoir, March 17 2016, image credit Rainer
Finally, a big danke to Rainer for the images. Keep them coming, Rainer.