Open Cermouth opened 2 years ago
Hi @Cermouth,
Technology costs are calculated at the technology level as the sum of exogenously defined levelized non-fuel costs, endogenously calculated fuel + water costs, CO2 storage costs, CO2 emission prices, and any other policy-related subsidies or taxes.
At the subsector level, costs are calculated as the weighted average cost of all technologies in that subsector (i.e., cost x output of each tech, divided by overall subsector output). The same applies at the supply sector level (weighted average cost for all subsectors in that supply sector). At the supply sector level, prices and costs can be used interchangeably.
In the case of DACCS, the "airCO2" cost should be negative - opposite in sign and equal in magnitude to the CO2 emissions price (after converting units and inflation-adjusting to a common year). This is equal to the subsidy paid to capture CO2 from the atmosphere with DACCS. If this subsidy exceeds the levelized non-energy + fuel + water + CO2 storage market cost of DACCS (plus a small emissions penalty for uncaptured natural gas combustion emissions for the high-temperature DACCS process), the technology becomes becomes financially viable and begins to be deployed. So, an aggressive emissions constraint resulting in a very high CO2 price could reduce the total cost of DACCS once the subsidy is accounted for and lead to large deployments in the model. Especially if other mitigation and negative emissions alternatives (e.g., biomass, nuclear) are limited or more expensive.
I'm not sure what the set-low, set-high scenario formulation is so I'd need more information to answer this more specifically.
Hope this helps!
@jayfuhrman Thanks for your answer! Sure it's so beneficial. Sorry for the lack of details in my question. Specifically, I used the three different non-energy costs of DAC in the default folder input/gcamdata/xml: dac_ssp2, dac_ssp3, dac_ssp5. I also set CO2 emissions constraints to reach a global net zero in 2050 (attached Zip file). xml_Cermouth.zip
I thought the non-energy costs of DAC in the three files are in order of medium (ssp2), high (ssp3), and low (ssp5). But in the output, i.e., the total technology costs show various orders among regions and years (attached png shows DAC costs change for the USA).
P.S. I also tried to change the non-energy costs by less than 10% (based on SSP3) and got output tech costs in the same order as I set. But larger change will lead to messy costs.
I want to focus on the net-zero year but find it common that DAC costs surge in 2050 but decline afterward, with different orders of DAC costs that are different from what I assumed. So I just call them DAC_cost_ssp_x instead of DAC_cost_low etc.
Given the multilevel calculation of the technology costs, I guess it may be complex if I want to "set" a certain DAC technology cost because it requires many sectors' change. Then does it mean I can only accept what the model gives?
I am very grateful for your help!
Are these results for the sector or technology level?
I'd recommend looking at the costs for the individual technologies, as the sector level results take into account weighted averages. It's possible that the endogenously solved shares of each technology, and their energy costs are leading to non-intuitive results between the scenarios if DAC is considered as a single sector. You might try taking more of a bottom up approach to calculating the costs to each of the 3 DACCS techs i.e., the exogenously defined non-energy cost + the exogenously defined energy input coefficients (e.g., natural gas, electricity) * the price of that energy input in a given region.
In that sense, it would be difficult to set a total cost for DACCS a priori since as you mentioned it depends so heavily on the cost of the input energy that is solved for endogenously. This could be done, in theory, by iteratively tweaking the scenario parameters until you get the total cost you want (again, for an individual technology, not the sector), but recognize that it would likely change somewhat between scenarios given the endogeneity.
@jayfuhrman Many thanks for your answers! The results shown are for the sector level. I agree with your suggestions and am looking into costs for each technology.
I set the DAC costs based on the SSP3 level in the inst/extdata and run gcamdata to get new xml files of DAC costs, but find that costs change differently from expected. In the costs by tech query, the high-medium-low order is correct. But in the prices by subsector query, the order is inverse, where the DAC cost in the set-low scenario is higher than that in the set-high scenario... I think maybe it's related to the aggregation from tech level to the subsector level, does anyone have an idea how to find the causes? Another point is that I want to set net-zero emission constraints, will the emissions constraints also influence the costs or prices of the negative emission technologies?
Thanks for any suggestions!