Hello,
I’m a master student at the TU Munich, currently writing my master’s thesis. My goal is to assess the impact that the use of electric vehicles would have on electrification strategies in Africa. I am focusing in particular on e-mobility (2/3 wheelers, vans, buses) and agriculture using electric tractors. I am thus trying to get more familiar with the OnSSET tool and some questions have emerged:
-
I found two different repositories (https://github.com/global-electrification-platform and https://github.com/OnSSET) that seem very similar. Although the former seems to be a newer version of the tool (takes into account hybrid solutions + productive demand), the latter seems to be more regularly maintained. Which one would you recommend to use ?
-
Are primary input csv files for all African countries available somewhere (on the same model as the ones provided for Sierra Leone) ?
-
Regarding agricultural demand estimation : if I understood the code correctly, in the v2.0 of the tool, the agri demand is defined as 10% of the household demand. Are you planning to estimate it more precisely in the future ? If yes, what would be your method ? I’m currently having some thoughts on how this could be done based on GIS, and I would be happy to discuss that in more detail with you if you’re interested.
Thanks for your help !
Irénée Froissart
Hello Irénée,
My name is Fhazhil, a student.
I saw your question and thought I should make a comment on it. I am not a developer of ONSSET, but I have some thoughts on the issue of agricultural electricity demand. In my view, the developer of the tool appreciates the fact that agricultural electricity demand is highly variable depending on the exact activity being considered (crop and animal farming, primary and post-harvest demand). On the part of the crop, it may depend on cropland cover, crop type and cultivar, soil characteristics, number of cropping seasons, and climatic conditions of the spatial area of interest. Therefore, the 10% of residential demand unused to represent agricultural electricity demand, I believe is considered as a good starting point (not necessarily as a true north estimate); perhaps the ratio is arrived at based on some ground truth data and observations from the field that the developers of ONSSET have. But one can do sensitivity analysis around that demand estimate to understand the impacts of changes in agricultural demand on electrification planning with Agriculture.
I have not used ONSSET for any meaningful academic work, but I have simulated some scenarios of some countries like Zimbabwe, Kenya, and Ethiopia out of my own learning interest. I do spatial modelling of crop-yield and groundwater-fed irrigation water requirement for various crops in the entirety of sub-Sahara (currently considering 46 countries), under different production conditions. This is based on the available cropland data, soil characteristics and climate data spanning decades (I use a lot of GIS processing to get around this). Of course, irrigation water requirement is different for each crop. So, it will be interesting to understand the impact of irrigation water requirement for various crops on electrification planning if one focused mainly on primary production (irrigation farming). But one can also look at other aspects including post-harvest electricity demand.
I use a very course resolution (leve2 admin in my models) but one can downscale/down-sample such data and use it spatial electrification planning in ONSSET. Various interesting scenarios can be considered here. In short, I believe the options are endless, perhaps I should consider one these interesting studies.
I am keen on learning about your thoughts on integration of agricultural electrical energy demand in the model. Of course, I am also keen on learning from the tool developer. They are great people and I have a lot of confidence in them.
Regards,
Fhazhil FSW
Dear both,
The https://github.com/global-electrification-platform GitHub repo is a particular application of OnSSET, for the Global Electrification Platform - GEP (https://electrifynow.energydata.info/). There is also a training course on the first version of the GEP application code available here: https://www.open.edu/openlearncreate/course/view.php?id=8393. I think you can find most input information from the GEP application for each country here: https://energydata.info/dataset?q=global+electrification+platform
We will soon update the main repo (https://github.com/OnSSET) to include hybrid systems as well.
For agricultural irrigation demand, that should ideally be calculated for each location as Fhazihil mentions. Here is an example that has been used in some country application, but not for the whole african continent: https://openknowledge.worldbank.org/server/api/core/bitstreams/30d60532-a67a-5664-a971-385ff48045a2/content
Andreas, Fhazhil,
Many thanks for your explanations and insights, and for the very interesting paper. Indeed, I’m trying to calculate the agricultural demand for each settlement of the OnSSET model as precisely as possible. As I’m focusing on the use case of an electric tractor, I probably need less various input information than you @Fhazhil to get to first estimates. Basically, what I need to get from GIS is the cropland extent map, and the crop type map, to have a first idea of the places where the tractor would be needed. While cropland extent maps are already available at a satisfying level of detail, a precise mapping of crop types across Africa has not been conducted so far, as pointed out in the paper (if one of you know about good partial crop types datasets, I would be happy to hear from it).
My second challenge then is to find a meaningful way to allocate cropland to the settlements produced by the population clustering (basically to know which farm cultivates which field). So that the additional demand (obtained through a tractor model) can be allocated to the right settlement. For this, several approaches may be possible, e.g. allocating each field to the closest settlement, and applying some corrections depending on the population or GDP per capita in the settlement. In my opinion, this allocation step is essential to integrate the agri demand in the OnSSET framework, which is based on settlements. But maybe there is a workaround…
Best regards
Irénée