Defining region-specific technology cost about gcam-core HOT 6 CLOSED

Region-specific deviations from the default values in the global-technology-database can be assigned by putting the information into the stub-technology objects located within each region. The stub-technology's values are used in the case of conflicts between the two (i.e., a cost specified within the stub-technology will over-ride the corresponding cost specified in the global technology database). The existing electricity_water.xml file already has stub-technology objects for the electric generation technologies, but there's generally only information read to the stub technologies in the model base years, where calibration values and calibration-year share-weights are assigned. Renewables have region-specific capacity factors indicated for all periods, so those have stub-technology information assigned to all periods.

from gcam-core.

Region-specific deviations from the default values in the global-technology-database can be assigned by putting the information into the stub-technology objects located within each region. The stub-technology's values are used in the case of conflicts between the two (i.e., a cost specified within the stub-technology will over-ride the corresponding cost specified in the global technology database). The existing electricity_water.xml file already has stub-technology objects for the electric generation technologies, but there's generally only information read to the stub technologies in the model base years, where calibration values and calibration-year share-weights are assigned. Renewables have region-specific capacity factors indicated for all periods, so those have stub-technology information assigned to all periods.

Hi pkyle,

Thanks for the suggestion. I have tried to do so by including the information as seen below.

However, I ended up with this when trying to run GCAM. It seems the parsing failed resulting in the GCAM failing to commence

Do you see any mistakes in putting the information (supplysector/subsector/stub-technology/)?

from gcam-core.

Well, if your goal is to switch hydro from a fixedOutput technology to an endogenous one, there are a number of steps that ought to be performed including:

• re-set the share-weight from 0 to 1 at the subsector level and at the technology level
• assign a subsector share-weight interpolation rule (see other subsectors)
• move the historical calibration values (currently in fixedOutput) to CalDataOutput/calOutputValue
  That's sort of the minimum set of steps, but really to do endogenous hydro you'll also need to assign a supply curve that characterizes the scarcity of the resources in the region, capturing (1) the increase in the levelized costs of electricity generation as a function of deployment, and (2) the limits on the deployment. See e.g. the representation of geothermal for a simple bounded supply curve which is read in as an input to a technology.

from gcam-core.

Hi pkyle,

Thanks for the swift response, and the suggestions.

I didn't really aim at switching hydro to be an endogenous. Rather, relevant the main question here, the goal is to include a different technological costs (not just for hydro, but for all generation technologies, both RE and non-RE) than the ones used in the global database. The costs I intended to use are tailored specifically for the region of Indonesia.

In the case of hydro, I have tried to leave everything as default and add the information on the costs only to each year. But, the simulation still wouldn't start. Does this mean that the fixedOutput is incompatible with the redefinition of the technology cost?

from gcam-core.

I can't promise that this is why the model crashed (didn't seem to leave an error message), but the use of fixedOutput is incompatible with tracking of costs. Fixed output technologies should have share-weights of zero, and they aren't included in any logit share or price calculations. They do not contribute to the weighted average price of their parent sector (commodity). The model possibly got jammed up trying to track the capital investment of a fixedOutput technology.
I don't really know what to recommend because I don't know what the aim is here. If you're just trying to estimate the capital investment into hydro from the scenario, that could be computed post-hoc based on the assumed fixed output, along with assumptions about survival curves, the ages of existing assets, and capacity factors. If you want the model to have hydro as an endogenous choice, I don't really see how to get reasonable results without some sort of supply curve, but you could just try it and see. If you're OK with just holding the 2015 output constant going forward, you can assign a long lifetime and then set the share-weight to zero in future periods, starting in 2020. But there's no way to exogenously fix the output of a technology while still tracking its costs in the model.

from gcam-core.

Okay. Thanks Page for your explanation!

from gcam-core.