Robbie Morrison has been working on high‑resolution national energy system models since 1995 and open source variants since 2003.  He now contributes regularly to the Open Energy Modelling Initiative, first becoming active in 2016.  Robbie began advocating for genuinely open data to underpin energy policy development in 2017.  Robbie holds an MSc in Energy Management from Otago University, Aotearoa/New Zealand.

Icebreaker One aims to open up discussion about how data access can speed up our progress towards net zero, so invited Robbie to share his perspective on open data and the IEA.

The International Energy Agency (IEA) took a first step toward open data on 24 March 2022 as an outcome of its 2022 Ministerial Meeting in Paris. But the destination remains unclear. Is the final objective simply to make the data the IEA collects from governments public, while leaving its legal status dangling? Or is the objective to provide genuinely open data, ideally under Creative Commons Attribution licensing, to ensure the information being made public is robustly usable and reusable by anyone and for any purpose? I favour, as does the open energy modelling community, the more ambitious journey of course, albeit one with doubtless more obstacles en route. 

IEA data is valuable

The IEA collects energy statistics from its member countries and from other nations and then collates and processes this information. These datasets are unique because no other organisation is in a position to readily assemble such data from governmental sources under relatively consistent technical parameters. The datasets are valuable because of their consistency and reach, both geographically and historically. And the datasets are necessary because citizens — on the presumption that democratic processes are to be valued — should by default be able to source the information being used to form and refine policy options free of cost and legally unencumbered. 

The current and tragic Russo-Ukrainian War and ensuing energy crisis only seeks to highlight the urgency of access to the kind of high quality energy sector information that can underpin quality political discourse and robust public policy development. This necessity predates the Ukrainian conflict: the current attempts to liberate the data the IEA assembles began in October 2021 with an open letter from Hannah Ritchie published in the respected journal Nature. Hannah is head of research at Our World in Data based in Oxford, UK. In her letter, Hannah stresses that the process of scrabbling together statistics on the progress of the Covid pandemic and subsequent immunisation rates must not be repeated for energy and sustainability data. 

IEA data and the transition to net-zero

The IEA datasets can serve many potential roles. The application we highlight here is the provision of data that can be used to populate energy system models. By exploring a wide range of scenarios, these computer models can help shed light on a myriad of questions about future energy systems — on how they might work, what their overarching characteristics might be, and indeed whether a set of suggestions being made are even collectively feasible. That final point being all the more important as we turn to fully renewable and operationally agile systems with greater levels of potentially erratic demand-side participation — including asset purchasing decisions, usage patterns, and demand responsiveness (and noting thankfully that social scientists are now contributing expertise). 

So the question of how best and how rapidly to pivot from western dependence on Russian coal, oil, and natural gas is now paramount. Together with the question of how to react if Russia unilaterally elects to curtail supply. Some modelling projects have recently extended their European energy system models to include Ukraine and Moldova (PyPSA for instance) but the availability of suitable national data remains problematic nonetheless. 

What’s next?

The larger issue is, of course, not how to pivot from Russian supply, but how to pivot from fossil fuels entirely. And recently, 50 or so system modellers signed another open letter to the IEA to point out that the cost of failing to identify high performing system trajectories because of incomplete or less accurate information could be astronomical. And in complete contrast to the very modest revenues that the IEA currently receives from selling such information to commercial organisations. 

The IEA’s first attempt at the public licensing of nil cost data was a shambles. In 2021, the IEA published some non-primary datasets under Creative Commons CC-BY-NC-ND-3.0-IGO licensing. This particular license is not open, not suited for data (fails to address European database rights), not international, not compatible with material under other licenses, and not current. A bizarre choice really. The Creative Commons CC-BY-4.0 addresses these issues head on and is the license most practitioners recommend in this context.

Key advantages of the CC-BY-4.0 are that it grants permission to both use and republish in original or modified form by anybody and for any purpose. And it has also become a de facto standard that then avoids the creation of legally walled‑off data silos through incompatible licensing provisions.

Hopefully the IEA will appoint a navigator for the next steps in its journey toward genuinely open data. What you too can do to help speed this journey: 

  • raise the issues with the IEA via social media or other channels 
  • raise the issues with your elected representatives 
  • support organisations advocating for open data in the energy sector and elsewhere