H2 in GCAM / GCAM-USA

[Youyi77]

Hi team,

A general question regarding the H2 sector.

I want to study the H2 sector in the US on a national level. (Trying to test different subsidies for H2 production techs, different SSPs, etc.)

Since there hydrogen sector in the GCAM-USA is still under development, for now, is it better to use GCAM instead of GCAM-USA even though the latter has a better-tailored dataset (for other sectors) ?

Thanks for any suggestions!!

[pkyle]

The main difference between the two (USA-region, GCAM-USA states) is that the USA-region representation allows hydrogen to be produced from nationally aggregated supply curves of the key resources (wind, solar, CO2 storage, etc), and provided to demands that are similarly represented at the national level. In other words there's no explicit consideration of potential geographic mismatches between least-cost production sites and major demand markets. In contrast, the GCAM-USA (state level) representation doesn't allow any inter-state trade of hydrogen, which will tend to amplify any such sub-national mismatches. Both of these are reasonable and defensible approaches, given the unknowns. While the contained state markets of GCAM-USA are more restrictive than reality (e.g., there are already hydrogen pipelines that cross the Texas/Louisiana border), it's also not realistic to assume that no sub-national geographic constraints will come into play, particularly as most of the wind energy resources are in the center of the country, thousands of kilometers from the major coastal population and industrial centers. In the future there may or may not be hydrogen pipelines transversing large areas in the country, just as there may or may not be schemes devised by which renewable electricity produced in one place could be transmitted thousands of km to produce hydrogen close to a demand center (and still be considered "green"). In my opinion, both modeling approaches produce useful information. We are currently testing out a strategy to have multi-state hydrogen trading blocks in GCAM-USA (following PADD regions) that should be in the core model at some point in the next ~year if all goes well, but I don't expect it to change the results all that much from GCAM-USA at present. The PADD regions still isolate the coastal population centers from the wind resources in the center of the country.

[Youyi77]

Hi @pkyle ! Thank you so much for these detailed explanations. Very helpful!! I actually used GCAM-USA for my analysis, but when I was reading Core Model Proposal 359: Hydrogen and transportation technology update from the website:

"In the core model, GCAM-USA doesn't have a hydrogen module at the state level; hydrogen is represented in the "USA" region, and the electrolysis technologies are disabled because in the GCAM-USA configuration the relevant electricity markets (elect_td_ind, elect_td_trn) don't exist in the USA region."

Q1: If I am understanding correctly, only the grid electricity water electrolysis (both central and forecourt) is turned off in GCAM-USA. (Or solar/wind/nuclear electrolysis are all shut down?)

I ran the reference GCAM-USA scenario to test and added up all the production from each state. It seems that there is still some (but limited) H2 produced from grid electricity production in both the central & forecourt production.

Q2: In addition, since other H2 production pathways (i.e., biomass to H2, NG reforming) also rely on electricity (elect_td_ind) as an input, I think my question is how to understand in the GCAM-USA configuration the relevant electricity markets (elect_td_ind, elect_td_trn) don't exist in the USA region. Electricity has to exist in the model so does it mean in GCAM-USA, there are no differences between these two?

The multi-state H2 trading sounds really interesting! Hope everything goes smoothly for the H2 sector development! Looking forward to it!

[pkyle]

I'm sorry, what I wrote in my prior comment is incorrect. There is a an update to GCAM-USA that creates a full hydrogen module in each state, but it was not merged in time for the June 2023 release, so it's not in the current master branch, or v7 tag, or release. Whenever the next public model update is posted it will include this feature, but I'm not sure when that will be, and for right now I would suggest for any hydrogen market analysis to not use GCAM-USA.

[Youyi77]

Got it! Thanks so much! Looking forward to the new release!!!