Environment & Energy
In reply to the discussion: 'Faster Than We Thought': An Epitaph for Planet Earth [View all]kristopher
(29,798 posts)In my initial post demonstrating the problem with EIA/IEA methods, I compared the projection of the WEO to the actual performance of wind and solar.
I added nothing, I speculated on nothing.
The only thing not controlled for was the content of what was being compared. For the WEO it was "other renewables"; a category poorly defined that appears to include not only solar PV and wind but also solar thermal, geothermal, and whatever else they saw on the horizon.
The achieved numbers reflect solar and wind only, are measured, and span 20 years with 20 individual evenly spaced data points.
Therefore the exercise I presented erred on the side of being conservative in regard to the point being made; which was that the forecasts from the EIA and the IEA are, in relation to the performance of renewable energy, invalid.
**********************
Here is your description of what you'd done:
Post 12
Extending the curves out another 20 years to 2040, renewables make a respectable showing, adding 1250 TWh. The problem is that fossil fuels add over 20,000 TWh - 16 times as much. And in that case hydro doesn't help - fossil fuels still add 9 times as much electricity as hydro and renewables combined.
Going by this table, fossil fuels go from generating 70.5% of the world's electricity today, to 81% in 2020, to 85% in 2040.[/font]
To create your slopes you used the WEO data from a summary table, which means you covered 49 years with 4 points of data - 2 real and 2 hypothetical. (That's 49 years with 4 data points vs 20 years with 20 data points)
The gap between the 1st/2nd point was 24 years, between the 2nd/3rd was 15 years;and between the 3rd/4th was 10 years.
This means you were trying to establish a complex set of trend lines only two real data points.
With this you make the point (I suppose you saw it as a point anyway) that the relative share of renewables is not adequate to be meaningful.
Your errors are clear.
Your sample size was small and the period you extrapolated to was large.
You used 4 points - 2 real and 2 hypothetical - over a 69 year period. Essentially giving you only a straight line between those first 2 measured points from which to derive the slope of the next 45 years from the remaining data (that you knew to be compromised at least in the area of renewables).
**************************
You obviously realized there were problems so you then did this in post 15:
So for the following graph I used the 2009 actuals in place of the estimated 2010 values. I also boosted the 2020 estimate for non-hydro renewables from 239 TWh to 600 TWh.[/font]
What that amounts to is that you:
a) removed hypothetical 3rd data points to use the actual numbers and;
b) addressed your error in using known false renewable data through inflation of the final renewable data point by 250%.
While a) is commendable it still leaves you with extremely poor granularity for your slopes and, more importantly there is a problem with your assumption about the amount of renewable generation in 2020.
We don't need to wing it, we can make the fit more accurate by recognizing the firmly established trend that is pretty created by the performance of solar and wind between 1992 and 20012.
The numbers below are year, capacity and amount generated for renewables from the WEO table:
1971 - 0
1995 - cap 13 gen 49
2010 - cap 43 gen 154
2020 - cap 79 gen 239
You changed it to:
1971 - 0
1995 - cap 13 gen 49
2009 - cap ?? gen ??
2020 - cap 79 gen 600
The slope established by the 20 data points predicts that instead achieving 79GW of installed renewable capacity by 2020 we are on track to hit 1200GW of solar and wind by 2020.
That suggests the WEO projection for 2020 is <6.5% of what we might expect given the limits of out modeling here.
In other words you have flattened your renewable curve by using a number the data does not support. You recognize that 239TWh is ruled out, but 600TWh is hardly better.
As the installed capacity numbers show, accurately plotting the trend line over the last 7 years gives us a lot of confidence that the 600TWh 2020 projection is woefully inadequate.
Since we don't have the latest globally aggregated production numbers, we can use the ratio WEO has established for us between cap/gen and carry it forward as a point of departure. Remember, the 1200GW is only solar and wind, thus we have some degree of buffering as the mix changes and small scale hydro , geothermal, storage etc. increase their effect on the overall capacity factor. (We'll leave out the recent crash in solar costs that is expanding the unsubsidized solar PV market to between 600 - 1000GW by 2020.)
That would make your final projected data point for renewables about 3676TWh, wouldn't it?
79/1200 = .065
239/.065 =3,676
What this argues for is that in reviews like this the optimistic trajectories, which DO focus on what happens when there is an evolution in the system, are more likely correct than those that are shackled to EIA type assumptions about the future value of the present infrastructure.
IPCC on Renewable Energy
Close to 80 percent of the worlds energy supply could be met by
renewables by mid-century if backed by the right enabling public policies a new report shows.
http://www.democraticunderground.com/112715489