In this piece, we delve into the GKG Grid Model's results for Germany in a bit more detail. One revelation is that the bastards pushing expensive nuclear are smarter than you might think. The PDF of this article is here. In that version, the scaling of Figures 2 and 3 better matches the text.
Not unless we can find a sugar daddy with a large computer. The GKG Grid Model is a compute intensive Linear Program. It maxed out my top of the line laptop on Germany, even tho we assumed no imports/exports which really isnt accurate for Germany and certainly won't work for Denmark.
We alos need access to a good interior point solver. Most of the runs we have done used the Mosek solver. Funny thing Mosek is a Danish company started by a guy named Andersen. I had a trial license but when that ran out I could not justify paying for a commerical license. So right now the model is shutdown.
Great stuff Jack and after you've done Denmark can we please get a rush job on Australia! We are currently suffering under the most puerile political debate about keeping the lights on and the fear of nukes.
How much does this analysis chance if one uses more recent prices? Prices for Li-Ion Batteries have more than halved in the last decade an may keep falling for other types of batteries. Solar power also got cheaper. Could you estimate what factor of cost improvement would be necessary on the renewable side that would make it outcompete nuclear?
These runs were done in 2023 using our best guess at 2020 costs. In coming up with these "base case" costs, we tried to err in favor of wind/solar/h2. Here's the overnight CAPEX figures we used in 2020 $/kW.
PV | 1040
On wind | 1436
Off wind| 4070
CCGT | 1000
OCGT | 700
Coal | 1500
Li chg/dsch|71
H2 chrg | 1250
H2 dsch | 1084
and unit storage CAPEX in $/kWh
Li batt | 264
H2 dome | 3
Thanks to Covid, the Ukraine War, and the IRA all these numbers look cheap to very cheap relative to 2024 costs.
Since 2021 wind/solar costs have not improved in real terms. Because of their low energy density, wind and solar are resource hogs. They get clobbered by commodity cost increases. Right now commodity costs are totally discombobolated. But it's hard to see the real cost of basic commodities like copper coming down much in the long run. But we'd be happy to make some sensitivity runs if somebody was willing to pay for them.
What does improve the case for wind/solar is better weather. Germany is a really silly place to make the case for wind/solar. We did do some preliminary runs for Spain. The wind/solar numbers were significantly better, but we ran out of license time before we had a reasonably complete analysis. And all the optimal solutions involved some fossil, even for very high SCC's. I say again for very low marginal capacity factors: wind/solar (and nuclear) are high CO2 sources and OCGT is a low CO2 source.
Ana chance for a similar analyses for Denmark.
Leif,
Not unless we can find a sugar daddy with a large computer. The GKG Grid Model is a compute intensive Linear Program. It maxed out my top of the line laptop on Germany, even tho we assumed no imports/exports which really isnt accurate for Germany and certainly won't work for Denmark.
We alos need access to a good interior point solver. Most of the runs we have done used the Mosek solver. Funny thing Mosek is a Danish company started by a guy named Andersen. I had a trial license but when that ran out I could not justify paying for a commerical license. So right now the model is shutdown.
Great stuff Jack and after you've done Denmark can we please get a rush job on Australia! We are currently suffering under the most puerile political debate about keeping the lights on and the fear of nukes.
Jack,
How much does this analysis chance if one uses more recent prices? Prices for Li-Ion Batteries have more than halved in the last decade an may keep falling for other types of batteries. Solar power also got cheaper. Could you estimate what factor of cost improvement would be necessary on the renewable side that would make it outcompete nuclear?
Soje,
These runs were done in 2023 using our best guess at 2020 costs. In coming up with these "base case" costs, we tried to err in favor of wind/solar/h2. Here's the overnight CAPEX figures we used in 2020 $/kW.
PV | 1040
On wind | 1436
Off wind| 4070
CCGT | 1000
OCGT | 700
Coal | 1500
Li chg/dsch|71
H2 chrg | 1250
H2 dsch | 1084
and unit storage CAPEX in $/kWh
Li batt | 264
H2 dome | 3
Thanks to Covid, the Ukraine War, and the IRA all these numbers look cheap to very cheap relative to 2024 costs.
Since 2021 wind/solar costs have not improved in real terms. Because of their low energy density, wind and solar are resource hogs. They get clobbered by commodity cost increases. Right now commodity costs are totally discombobolated. But it's hard to see the real cost of basic commodities like copper coming down much in the long run. But we'd be happy to make some sensitivity runs if somebody was willing to pay for them.
What does improve the case for wind/solar is better weather. Germany is a really silly place to make the case for wind/solar. We did do some preliminary runs for Spain. The wind/solar numbers were significantly better, but we ran out of license time before we had a reasonably complete analysis. And all the optimal solutions involved some fossil, even for very high SCC's. I say again for very low marginal capacity factors: wind/solar (and nuclear) are high CO2 sources and OCGT is a low CO2 source.