modeling of resource extraction technologies

[SongminYu]

While understanding the fossil resource tutorial, I found the concept of "resource" is a bit special. On one hand, it is a commodity and can serve as input for technology. On the other hand, it also has its own volume and cost, which is quite special:

• A normal commodity in MESSAGE does not have volume directly, but can only be calculated by multiplying the ACT of technologies and output.
• A normal commodity in MESSAGE does not have cost, which is the parameter of technologies.

Following this, I have two questions:

• First, I want to confirm if my understanding is right: resource is a special commodity on the edge of the system boundary and it is scarce in the system covered by the model, so it has to have volume, as well as cost. Are there any other considerations about why resource is designed in such a way?
• Second, is it possible to model resource extraction technologies (or mining sectors in general) in MESSAGE? If so, are there any suggestions?

[khaeru]

To your second question only, please note there is some methodological (but not code-focused) discussion in the docs of message-ix-models (for the IIASA family of global and country-scope models) on resources; see https://docs.messageix.org/projects/models/en/latest/global/energy/resource/fossilfuel.html and the following pages.

[SongminYu]

Hi @khaeru, thanks for your swift reply! So in MESSAGE, the convention is that the extraction technologies are embedded in the assumptions of volume and cost of the resource. They are not modeled explicitly.

So, in a national model, taking coal as an example, can I have another technology coal_import --- only with output and var_cost meaning external import (as the last part of this tutorial) --- to compete with coal extraction?

[khaeru]

• MESSAGE (the generic framework) lets one do many things. Your first question is about the generic framework and its many possible uses.
• "MESSAGEix-GLOBIOM" is our term for a particular set of models that use the generic MESSAGE features in a few particular ways. (I know the naming is confusing, sorry—I would have chosen something different.) I wouldn't call this "convention", just our way of using the framework. Some of the ways we currently use the framework are for legacy reasons and I would specifically counsel new users to avoid emulating.
• The link I gave above describes MESSAGEix-GLOBIOM, but per my comment it only describes the methods in a textual way, not specifically what is done in code.
• The other docs of message-ix-models describe some (not all) of the code/data that implement those methods.
• In particular, looking at Technologies (technology.yaml), there are many technology codes with _extr_ in the ID.
• For example, there are bio_extr_1 through bio_extr_6. I would assume (there are no specific docs on this point, so I cannot check) that these are used to represent different extraction technologies with different techno-economic characteristics that all (per the YAML file) output: [biomass, primary].
• So I would say, in the specific case of MESSAGEix-GLOBIOM, that it's not true that "they are not modelled explicitly". They are.
• However I would guess that a modeler could choose to use the underlying/generic MESSAGE features to assign extraction costs (as opposed to other resource-related costs, like the costs of mineral rights or permits, etc.) to resource_cost. Nothing seems to prevent this.
• Maybe this is what happens in the fossil_resource tutorial; I didn't write it and am not so familiar, so can't say.

Per your new question, to take the example above, the bio_extr_[123456] technologies output to commodity=biomass, level=primary. Indeed any other technology, like an import/export technology, can be parameterized to output to the same (commodity, level), and would thus compete with bio_extr_*. However, I don't know how the input parametrization of bio_extr_* relate to the resource_{cost,volume} parametrization of coal or other commodities. I have asked colleagues who have worked directly with these details to look at your question here and chime in.