Skip to content Skip to navigation


GeoH2 model: Geospatial cost optimization of green hydrogen production including storage and transportation is a new journal article written by Claire Halloran, Alycia Leonard, Nicholas Salmon, Leander Muller and Stephanie Hirmer. It has been published on the ScienceDirect website and here, Alycia Leonard explains what the article aims to do.

What’s the significance of this paper?

The Hydrogen paper that just came out is the descriptor of the model that we’ve produced. It explains what the model does and how it can be used.  We feel that the model will be particularly interesting and useful to technocrats such as policymakers, as well as academics in the sector. It’s deliberately open-source and we want people to use it freely.

What is the origin of the project?

This project has been in development for a few years and started in collaboration with Aachen University in Germany.  The initial idea was to look at what the costs would be for producing green hydrogen in Kenya, taking into account where you would produce it versus where the demand would be. For example, if you produce it up in Turkana, but then you need to ship it from Mombasa, how does that affect things, including the form of hydrogen you need to produce?

Since we first ran the model in Kenya, we’ve made the model location agnostic so it can be applied to any country. We’ve run cases in a number of different locations, and they’ve worked successfully. We’ve added some functionality where you can set a demand level and type for the hydrogen you need as well as a location. The model will then look at the cost to implement all of the infrastructure you would need to produce that hydrogen, including the solar, wind and storage combination.

There is a very clear need for this type of model because a lot of countries are signing MOUs to export hydrogen. The model enables us to evaluate if these deals are going to be beneficial to the lower income countries that are likely to produce the hydrogen using their high renewable potential for export to higher income countries.

There are some forthcoming papers about that, but that’s the motivation for the model.

What are the criteria for a good location for producing green hydrogen through electrolysis, using renewable electricity?

Right now, it’s wind and solar that we’re looking at, but we also take into account things like land availability. You are constrained based on which areas are built up, which areas are protected, where the transport network is and so on. We can look at shipping considerations and if you need to build more infrastructure to be able to ship the hydrogen, we look at what those costs would be.

And so, is there a certain set of criteria or elements that you’re looking for?

Yes. So far, what we’ve seen is that the locations with more consistent availability of wind come out cheaper; wind is a big factor in the overall viability costing, along with land availability to build the wind turbines. The transport infrastructure does also matter, but it’s not the major factor. Often, you’ll see areas closer to a port come out slightly cheaper if they have similar wind and solar resources to other places.

Electrolysis involves water, so you take that into account as well?

We allow for both sea water and fresh water.  If there’s access to fresh water, the costs will be both for the water and the treatment.  If there’s access to the coast and sea water, we would include the cost of the desalinisation processes, and take whichever one is cheaper.

What are your plans for other papers on this topic?

As well as this first paper explaining how the model works, we have another paper submitted to a journal that is being considered right now. This will feature the findings from us running the model for every coastal country in sub-Saharan Africa and looking at all the different implications from that.

Did you do this project ‘from your desk’ or did you visit certain locations?

It was a combination.  We started off in Kenya, working with a master’s student at Aachen and we did a number of stakeholder interviews to determine some of the assumptions that we were starting with. We also thought about which factors to include, which processes to use and whether to look at smaller demand for in-country use or large export scale demands.

For our next (forthcoming) paper, we ended up taking data from a lot of different sources to come up with more detailed estimations for cost of capital in different countries, to better inform the pricing and other aspects.

We presented the model at the recent CCG workshop in Kenya and got good feedback on it from the folks who were there from the Ministry of Energy and other departments.  It was great to present it and get that validation.  And it will be another milestone for us when the next paper comes out after the review process.  We’re just at the start. 

You can find the full article here.