July | 2018 | PassiiviIdentiteetti

I have earlier expressed my dissatisfaction with the integrated assessment modeling in the context of climate change. Way too many modelers hide insane and poorly justified assumptions into models and then pretend the outcome is “science”. Recent Nature energy paper by Grubler et al. titled “A low energy demand scenario for meeting the 1.5 °C target and sustainable development goals without negative emission technologies” seems to provide another example of this. I read the paper with interest, but then noticed something in small print on the last page.”Estimates for present-day and mature technology costs are from the GEA and World Energy Outlook. Assumptions for granular technologies, which include solar PV, small-scale hydrogen production, fuel cells and heat pumps, and distributed energy storage, such as batteries or fuel cells, were updated from SSP2 to reflect the more dynamic storyline of the LED scenario.”

Hmmm…granular…dynamic storyline…sounds suspicious. Maybe I have to read the supplementary as well. Then from the page 80 of the 122 page supplementary I find the actual cost assumptions. Grubler et al. assume solar PV installed cost of 50$/kW post 2050! This is around 30 times lower than the current costs. (Who cares if its a factor 20 or 40?) Obviously if you wish to make a claim that scenario with lots of solar is economical it helps to invent your own costs so that you get whatever you desire as an outcome. To illustrate how extreme their assumption is, below I compare it with some other projections. (“Breyer” refers to numbers Christian Breyer, a solar advocate from Lappeenranta University of Technology, uses. Greenpeace report was written together with European Photovoltaic Industry Association.) Grubler et al. assume costs that are an order of magnitude lower than even EPIA+Greenpeace project. Even Breyer’s (finnish equivalent of Mark Jacobson) most optimistic dreams imply around 6 times higher costs.

I am sorry. You assumed what?

To me the cost assumption does not even seem consistent with learning curves (which should be treated with caution in any case).

If one makes such extreme assumptions one would expect extensive discussion and justifications for this. Certainly one should present result with and without such assumptions to see how sensitive results are for those funny assumptions. However, in the Grubler et al. paper this was done quietly and hidden in the supplementary where it is justified with…”The technology portfolio choice in MESSAGE is informed by modifying particular granular and economies‐of‐scope technologies for the LED scenario (Supplementary Table 28) whose stationary cost trends in the original SSP2 scenario was judged non‐compliant with the LED scenario storyline. All other technologies not listed in Supplementary Table 28 have been retained at their original (quite conservative) SSP2 values (e.g. for the year 2050: wind 500 $/kW, nuclear 2600 $/kW, biomass power plants 1200 $/kW, etc.), an assumption in line with keeping LEDs emphasis on efficiency and demand, and granular, decentralized supply options and new organizational IT and digital economy models of combining supply and demand, e.g. in grid‐to‐vehicles but also vehicles‐to‐grids options or other distributed storage options (e.g. hydrogen based).”So basically there is no other reason than authors narrative desires. Adding buzzwords “granular”, “organizational IT” etc. does nothing to strengthen the argument. If someone would make a scenario where nuclear power in 2050 would cost 170$/kW (from about 5000$/kW today), he would be laughed at. Probably most outraged would be nuclear engineers with actual understanding of the matters. You do the same with renewables and instead facing ridicule you land your paper in Nature Energy. Hopefully this example does not reflect the intellectual standards of the IAM community.

P.S. I encourage you to have a look at the numbers in their database as well https://db1.ene.iiasa.ac.at/LEDDB. I do not understand those cost assumptions. Sometimes coal prices are negative, sometimes positive, sometimes they are positive in the south, negative in the north…sigh. Also, 4Gt sink seems to magically appear from afforestation. Big if true.

Added 19.10.2018: Bizarre, but this scenario has now become one of the four illustrative pathways in the IPCC Global warming of 1.5 degrees special report. If we were to use 20% cost reduction for each doubling of capacity, reaching 50$/kW level would require about 15 doublings. With current installation rate that would take approximately 100000 years. (If we use the 11% rate which is maybe more defensible as it contains also the currently dominant balance of system costs, it takes more than billion years.) The IPCC scenario database sadly doesn’t contain capital costs for most 1.5 degrees scenarios. Here it is hidden in the 122 page supplementary of the article.

Added 23.10.2018: Accidentally clicked also at the forestry section of the scenario database. What is going on with those residues? First collapse and then huge increase when underlying forestry production increased only by about 40%.