I will shortly submit the following three abstracts in rank order. And request 20 minutes for each (the maximum on offer being 55 minutes).
If someone else has submitted something covering much the same theme at my third abstract, I would be quite happy to have it bumped.
And please offer comments, either below in public, via private messaging using this forum, or by email. The work outlined is all somewhat speculative and could certainly be improved upon.
1 : Update on open‑source energy system modeling in the global south and including Africa
Energy system models are simulations of future energy systems that can be used to test scenarios. More specifically, such models can explore a range of net‑zero options in an integrated fashion, determine which scenarios are indeed feasible, and then report on system development trajectories, detailed and aggregate costs, and related attributes for further consideration. Many of the underlying modeling frameworks are now fully‑fledged open‑source projects. In addition, there are several nascent initiatives to develop coherent databanks and also the overarching data standards they require, with both endeavors suitably open licensed.
These various efforts are now starting to spill into the global south generally and sub‑Saharan Africa in particular. A number of potential benefits then arise from this kind of open analysis. The first is the zero monetary cost of course. The next is organic knowledge transfer both northward and southward within the various project communities. A third is doubtless that a greater range of scenarios will be placed on the table — indeed I sense that the multilateral agencies working in Africa have settled on a selected set of solutions and that suggestions that fall outside the prevailing orthodoxy are unwarranted and unwanted. A fourth potential advantage is local engagement, and further, the prospects of improved local autonomy — and while there are no examples of model‑mediated public processes in the global south as yet, that concept is being trialed in the global north.
The use of open analysis in the global south will offer distinctive challenges nonetheless. The most obvious difficulty is data availability and a number of proxy solutions have been developed. The next is how best to channel these efforts into public policy formation and then on to live projects. Also critical will be the necessity of finding new ways of interacting between official agencies and these clearly informal modeling communities.
Two of the leading open‑source framework projects, OSeMOSYS and PyPSA, have begun significant efforts to broaden into the global south. These two initiative will be reviewed (I am not directly involved in either).
Clearly early days still but sufficient progress has been made to warrant an update at FOSDEM’23.
2 : Energy systems data, overarching standards, and a knowledge commons — from a legal perspective
Energy system models are simulations of future energy systems that can be used to test scenarios. As such, they are extremely data intensive and critically dependent on both data availability and the legal right to use and reuse that information.
To set some context for this session. Only information that has been or can be legitimately made public is in scope — and only that subset that is of some public interest. Then to further restrict focus to structured data and the data models that govern that structured data. Such data may comprise: lists of generation assets, timeseries covering demand, market clearance information, parameters characterizing engineering plant, graph structures representing network connectivity, geospatial information for various purposes, and representative past and future weather years. Data models may include glossaries, database schemas, reference architectures, and ontologies. When a broad cooperation is sought, these data models need to be both consensus‑driven and suitably open licensed.
The legal status of these overarching data models has not received sufficient attention to date. In the energy sector historically, such models are typically incorporated as industry standards and distributed under FRAND (fair, reasonable and non-discriminatory) licensing terms and with four‑figure cover prices. The problem is not simply one of cost and availability however. The copyright in non‑open data models may potentially pass across to conforming datasets, even though the collated data itself is able to be open licensed. And that same copyright may also transmit to any energy systems framework that embeds the articulated semantics. The best solution is simply to open license these data models — which then resolves both issues in a single stroke.
The open licensing of the datasets themselves needs to avoid the creation of legally‑determined data silos. At present, the Creative Common CC‑BY‑4.0 license offers the best general solution in terms of being well known, international, in common usage, and with at least some government licenses already inbound‑compatible. The use of other instruments will likely hinder the development of a much needed knowledge commons spanning the energy systems domain.
This session will also review a number of problematic examples, including the use of bespoke (one‑off) public licenses by key research institutes and the use of non‑disclosure by pivotal agencies.
3 : What are energy system models — and why make them open?
Energy system models are simulations of future energy systems that can be used to test scenarios. The simulations themselves offer high spatial, temporal, and topological (network) resolution and may employ clustering techniques to improve numerical tractability. Aspects of abutting systems are sometimes included, such as land usage, water resources, and industrial materials flows. Time horizons typically run to 2050. And climate modified future weather sets are now routinely applied.
The simulations themselves capture both systems operations and capacity expansion and retirement. They embed various forms of decision‑making and differing levels of foresight. Carbon‑constrained cost optimality is usually the goal, with near‑optimal solutions now sometimes also explored. These formulations are well suited to variable renewables generation and synergistic storage. When a scenario proves feasible, reported outputs include the system development trajectory, detailed and aggregate costs, and related attributes. Techniques to assess both model robustness and system resilience are now also starting to be developed and deployed.
Individual simulations can also act as a test bed for hypothetical future technologies, such as green hydrogen, emobility, and incentivized load management — as well as left‑field solutions like consumption sufficiency.
A number of the underlying modeling frameworks are now fully‑fledged open‑source projects with active and rapidly growing communities.
An emerging theme is the uptake of such models in the global south generally and in sub‑Saharan Africa specifically. Open methods offer analysts in the global south comparative ease of entry and informal knowledge transfer. While the projects themselves should benefit from improved uptake and diversity. And the greater analytical scope should enhance the available solution space, particularly as energy systems are essentially global.
One novel idea is the application of commons‑based peer production to public policy development, in part driven by the simple expedient that public agencies alone do not have the capacity to search the potential scenario space — and nor the creativity for that matter (with neither aspect intended as a criticism of government analysts).
This session will briefly cover the more notable framework projects and review the benefits of adopting open‑source development, open data, and data models that are both widely negotiated and genuinely open.