GEOH2

Geospatial analysis of hydrogen production costs

GEOH2 calculates the locational cost of green hydrogen production, storage, transport, and conversion to meet demand in a specified location. These costs can be compared to current or projected prices for energy and chemical feedstocks in the region to assess the competitiveness of green hydrogen. Currently, different end-uses, such as fertilizer production, export shipping, and steel production, are not modeled.

The model outputs the levelized cost of hydrogen (LCOH) at the demand location including production, storage, transport, and conversion costs.

In the code provided, the specific use case of Namibia is investigated. Parameter references for this case are attached. However, as the code is written in a generalized way, it is possible to analyse all sorts of regions.

GeoH2 builds upon a preliminary code iteration produced by Leander Müller, available under a CC-BY-4.0 licence: https://github.com/leandermue/GEOH2. It also integrates code produced by Nick Salmon under an MIT licence: https://github.com/nsalmon11/LCOH_Optimisation

Setup instructions

Clone the repository

First, clone the GeoH2 repository using git.

... % git clone https://github.com/ClimateCompatibleGrowth/GeoH2.git

Environment setup

The python package requirements are in the environment.yaml file. You can install these requirements in a new environment using mamba package and environment manager (installation instructions here):

.../GEOH2 % mamba env create -f environment.yaml

Then activate this new environment using

.../GEOH2 % mamba activate geoh2

CDS API setup

The get_weather_data rule downloads the relevant historical weather data from the ERA-5 reanalysis dataset using Atlite to create a cutout. For this process to work, you need to register and set up your CDS API key as described on the Climate Data Store website.

Note: Ensure the API key and URL are affiliated with CDS-Beta.

Solver setup

For the optimize_hydrogen_plant rule to work, you will need a solver installed on your computer. You can use any solver that works with PyPSA, such as Cbc, a free, open-source solver, or Gurobi, a commerical solver with free academic licenses available. Install your solver of choice following the instructions for use with Python and your operating system in the solver's documentation.

In Scripts/optimize_hydrogen_plant.py line 160, the solver is set to gurobi. This must be changed if you choose to use a different solver.

Note: Snakemake uses Cbc, which will be installed upon environment setup. To check, activate your environment and enter mamba list in your terminal for the environment's list of packages.

Preparing input data

Hexagons

To analyse a different area of interest, the input hexagon file needs to be changed, but needs to follow the logic of the one provided.

A full walkthrough on all the tools to create these hexagons are in the GeoH2-data-prep repo.

An explanation of how to create a H3-Hexagon file can be found in the following repo:

https://github.com/carderne/ccg-spider

The hexagon file needs to filled with the following attributes:

waterbody_dist: Distance to selected waterbodies in area of interest
waterway_dist: Distance to selected waterways in area of interest
ocean_dist: Distance to ocean coastline
grid_dist: Distance to transmission network
road_dist: Distance to road network
theo_pv: Theoretical PV potential --> Possible to investigate with: https://github.com/FZJ-IEK3-VSA/glaes. Note that this value should be in MW.
theo_wind: Theoretical wind turbine potential --> Possible to investigate with: https://github.com/FZJ-IEK3-VSA/glaes. Note that this value should be in MW.

Once you have created a hexagon file with these features, save it in the Data folder as hex_final_[COUNTRY ISO CODE].geojson.

Note: COUNTRY ISO CODE is the country's ISO standard 2-letter abbreviation.

Input parameter Excel files

Required input parameters include the spatial area of interest, total annual demand for hydrogen, and prices and cost of capital for infrastructure investments. These values can be either current values or projected values for a single snapshot in time. The parameter values for running the model can be specified in a set of Excel files in the Parameters folder.

Basic H2 plant: in this folder, there are several csv files containing the global parameters for optimizing the plant design. All power units are MW and all energy units are MWh. For more information on these parameters, refer to the PyPSA documentation.
Conversion parameters: conversion_parameters.xlsx includes parameters related to converting between states of hydrogen.
Country parameters: country_parameters.xlsx includes country- and technology-specific interest rates, heat and electricity costs, and asset lifetimes.
- Interest rates should be expressed as a decimal, e.g. 5% as 0.05.
- Asset lifetimes should be in years.
Demand parameters: demand_parameters.xlsx includes a list of demand centers. For each demand center, its lat-lon location, annual demand, and hydrogen state for that demand must be specified. If multiple forms of hydrogen are demanded in one location, differentiate the demand center name (e.g. Nairobi LH2 and Nairobi NH3) to avoid problems from duplicate demand center names.
Pipeline parameters: pipeline_parameters.xlsx includes the price, capacity, and lifetime data for different sizes of hydrogen pipeline.
Technology parameters: technology_parameters.xlsx includes water parameters, road infrastructure parameters, and whether road and hydrogen pipeline construction is allowed.
Transport parameters: transport_parameters.xlsx includes the parameters related to road transport of hydrogen, including truck speed, cost, lifetime, and capacity.

Snakemake

This repository uses Snakemake to automate its workflow (for a gentle introduction to Snakemake, see Getting Started with Snakemake on The Carpentries Incubator).

Wildcards

Wildcards specify the data used in the workflow. This workflow uses two wildcards: country (an ISO standard 2-letter abbreviation) and weather_year (a 4-digit year between 1940 and 2023 included in the ERA5 dataset).

Config file

High-level workflow settings are controlled in the config file: config.yaml.

Multiple wildcard values are specified in the scenario section. These can be changed to match the country and weather_year you are analysing.

Renewable generators considered for hydrogen plant construction are included in the generators section.

In the transport section, pipeline_construction and road_construction can be switched from True to False, as needed.

Note: country and weather_year can be a list of more than one, depending on how many countries and years you are analysing.

Rules

Rules can be run multiple ways using Snakemake. Below, you will be able to run rules by entering the rule name or their output in the terminal. Snakemake will run all necessary rules and their corresponding scripts to create an output. While all rules are discussed here for completeness, you do not need to enter each rule one-by-one and can simply enter the output you're interested in or one of the run all rules. Rules are defined in the Snakefile.

Snakemake requires a specification of the number of cores to be used; this can be up to 4.

Run time

The get_weather_data rule, depending on country size and your internet connection, could take from a few minutes to several hours to run. Ensure that you have space on your computer to store the data, which can be several GB.

The optimize_hydrogen_plant rule, depending on country size and the number of demand centers, could take from several minutes to several hours to run.

The optimize_transport_and_conversion rule, depending on country size, should take a few minutes to run.

All other rules take a few seconds to run.

Rule to remove all files

Note: This rule does not work on Windows, as of yet. Please manually remove the files you need to.

This rule is important to know first, as it will remove all the files that the below rules will create as well as the file you initially saved into the Data folder as hex_final_[COUNTRY ISO CODE].geojson.

This is to allow for a quicker transition to analyse more data and to clear up space. Make sure you save the created files that you need elsewhere before running the following rule into the terminal:

snakemake -j [NUMBER OF CORES TO BE USED] clean

Run all rules

This section can be used to run all rules, without having to run exact output files. If any files are changed after a completed run, the same command can be used again and Snakemake will only run the necessary scripts to ensure the results are up to date.

The total hydrogen cost for all scenarios can be run by entering the following rule into the terminal:

snakemake -j [NUMBER OF CORES TO BE USED] calculate_all_countries_and_years_total_hydrogen_costs

Similarly, you can map hydrogen costs for all scenarios with the following rule:

snakemake -j [NUMBER OF CORES TO BE USED] map_all_countries_and_years

`assign_country` rule

Assign country-specific interest rates, technology lifetimes, and heat and electricity prices from country_parameters.xlsx to different hexagons based on their country.