Access all areas

Summary

Energy system modelling is a powerful tool in the quest to design secure, affordable, clean grids of the future

Current energy models and datasets are either proprietary or require advanced programming skills, making them inaccessible to most stakeholders

TransitionZero’s flagship product, Model Builder, will make cutting-edge energy models and data accessible to all

Universal access to best-in-class modelling and data will improve the understanding of assumptions and inputs that underpin energy planning

Model Builder will improve trust between investors, donors and host countries, facilitating climate finance discussions

“Give a man a fish, and you feed him for a day. Teach a man to fish, and you feed him for a lifetime.”

This age-old piece of social wisdom has stood the test of time. Its message of empowering individuals through self-sufficiency is gaining fresh relevance in today’s hyper-modern, hyper-complex world.

Planning an energy system that is fit for the future is among the most complex and urgent undertakings currently facing humanity. Thankfully, technological advances are providing some of the tools needed to attack this challenge head-on.

One of the most sophisticated tools at our disposal is energy system modelling – the process of creating mathematical or computational representations of the behaviour of energy systems in different scenarios. Models can help us to predict future performance under different conditions, to optimise energy systems planning and build out by finding the most efficient and cost-effective combination of sources and technology, and to assess the impact of policy measures and regulations on supply and demand.

High profile examples of mainstream energy modelling include the International Energy Agency, which uses its Global Energy and Climate (GEC) Model to produce scenario-based insights for its flagship World Energy Outlook report. Oil major Shell uses two proprietary models to produce its own scenarios for planning and advocacy purposes.

Energy systems modelling has improved significantly over time, driven by advancements in computing power, data availability and modelling techniques. But the barriers to entry are high as the skills and costs associated with systems modelling remain relatively inaccessible, putting the power of modelling in the hands of a few technical experts.

Gatekeepers and black boxes

Academics and consultants are best placed to design, interact with, and interpret the outputs from today’s energy systems models. These highly complex models are constructed with a different language, which necessitates an understanding of not only a model’s inputs and assumptions but also its limits.

While this is understandable, it creates a ‘gatekeeper’ problem: only specialists are positioned to provide the deep analytical scenario-driven insights that investors, asset managers, energy planners and policymakers the world over are clamouring for in the quest to design secure, affordable, clean grids of the future.

Another structural barrier preventing stakeholders from leveraging the power of energy systems models is the ‘black box’ problem. The closed source nature of most models means audiences cannot interrogate the assumptions used, which can run to hundreds or even thousands of data points.

The two cases cited above – the IEA and Shell – are classic examples of high-impact, closed-source modelling. The insights gleaned from these models are often widely publicised and can influence public policy debate. But stakeholders cannot see the data that goes into them, nor the assumptions used to generate the findings.

This cultivates a tendency for audiences to grant undue authority on the model’s outputs. Ignoring inputs overlooks the core value of system modelling: to explore the potential impacts of today’s choices on the energy system across a range of uncertain futures. How can an energy planner, investor, corporation or civil society organisation hope to examine how different policy decisions or physical constraints alter systems behaviour over time if they do not know – or cannot test – the assumptions used to model it?

The energy systems modelling landscape.

The ‘black box’ problem is endemic in the world of commercial energy modelling. Too often, time and effort are spent on modelling workstreams that are at best iterative and at worst duplicative because stakeholders do not have clear visibility of the work already undertaken. This leads to sclerotic policy development, particularly in low-income countries, where the capability to test ambitious targets or cost creative climate solutions lies out of reach.

Open source is often held aloft as a panacea to the ‘black box’ problem, with its use in energy systems modelling gaining momentum. For example, the European Scientific Advisory Board on Climate Change called for greater transparency of energy market and network models and calculations to facilitate “public scrutiny of political investment decisions”. In a written testimonial to the European Commission, the council said:

“The traditionally closed and proprietary nature of energy system planning at national levels is no longer fit for purpose, given the interconnected market fundamentals and the need for rapid reductions in GHG emissions and integration of wind, solar and storage technologies.”

Open source alternatives address the major issue of proprietary modelling tools by improving transparency, but they create issues of their own. For example, they require users to have programming knowledge, which increases the ‘gatekeeper’ problem. Also they struggle to ensure reliable maintenance of tools and tend to lack direct support, both of which compromise usability.

Data gaps

Not all data is created equal. This is another area hindering uptake: the availability, timeliness, and auditability of data and its sources.

Imagine if the latest energy systems modelling capability were suddenly made freely available and accessible to all. What then? The hope is that the focus would quickly shift to data quality. After all, the output from a systems model is only as good as the input data – and the user’s ability to validate both.

Energy data in emerging economies is often too hard to access, updated infrequently and too high-level (e.g. country-level, rather than at the grid node or asset-level) to assist with crucial energy system planning issues. The value of systems modelling is a function of the quantity and quality of input data, and there is significant room for improvement on what is publicly available for use on both fronts.

In many countries, data is often privately held. For commercial providers, controlling entire databases is a necessity for business and legal reasons – and the data rentier business model has proven to be extremely profitable. Nonprofits are often not incentivised to embrace open source or open data frameworks, as it is not a requirement to get funding. Also, many lack the appropriate resources and expertise. These structural issues result in data gaps.

Missing, outdated or flawed datasets result in a mismatch between model outputs and their relevance to informing real-world outcomes. Open-access data products can be produced in the public sector, but this requires funders who see its value and commit to it long-term. A systematic approach to data gathering and model validation can improve trust in both the model itself and modelling per se.

Why this matters

The barriers hindering access to energy systems data and modelling capabilities are not theoretical – they have real-world implications, especially at this crucial time when energy systems must transition rapidly towards net zero amidst the threat of climate change.

Fundamentally, countries’ energy policies and roadmaps are informed and built on models. As a result, we face a situation where a lack of data transparency and trust is hindering the flow of climate finance capital from donor countries and institutions to host countries because funders cannot validate key assumptions about market-level energy transition fundamentals.

New policies will be key to building bridges and opening the finance floodgates – but only if decarbonisation plans are transparent, cost-effective, and investable. Open source data and modelling tools have the potential to accelerate the development of such policies, so making them accessible to participants on both sides of the climate finance ledger is of critical importance.

This is a bottleneck in decarbonisation – particularly in low and middle-income countries, which are poised to see the largest increases in population growth and energy demand over the rest of this century.

To this end, TransitionZero is developing Model Builder – an open source energy systems modelling suite and data platform that seeks to resolve the ‘gatekeeper’, ‘black box’ and ‘data gaps’ problems in one fell swoop.

A product-led approach

Model Builder is TransitionZero’s flagship product, and in development at time of writing. At its simplest, it is a tool to help policymakers, investors and other stakeholders make better energy-related decisions. The web-based platform will allow both technical and non-technical users to test assumptions and explore trade-offs in energy transition planning.

Users can design models that project how power production and generating capacity will grow in the future, using data and models by TransitionZero and other users already on the platform, or by bringing their own. These models can be used to explore different scenarios based on new climate policies, technology costs and pathways, and infrastructure investment plans.

TransitionZero’s team of data scientists and engineers is developing an intuitive web app user interface and building on the most widely-used open source systems models and input data, while improving them with deeper spatial granularity and higher temporal frequency. Such improvements contribute to the open source frameworks – a feedback loop that allows for continuous improvements to the ecosystem of open source data, which will be incorporated back into Model Builder to offer best-in-class systems modelling capability to users.

Energy market and power generation data is sourced from automated web scrapers. Asset-level data is gleaned from satellites, with machine learning to improve data accuracy and infer asset properties and activities.

In parallel, our team of in-country analysts is working with local partners, regulators and government officials to ensure data quality and completeness, and to validate model assumptions.

Our direction of travel (at time of writing!)

Global ambition

We aim to offer unparalleled coverage of electricity and energy markets globally. Our platform and its geographic coverage are being rolled out in phases, with an initial focus on Southeast Asian electricity markets.

The beta launch covered Indonesia with model runs that compare how emissions and system costs evolve out to 2050 under differing scenarios with and without early retirement of coal-fired power plants. (Read more in Two billion reasons: how Indonesia can get ahead of the net zero curve.) Asset-level data was incorporated from our Coal Asset Transition (CAT) tool to provide granular estimates of individual coal plant costs to inform asset retirement decisions in coal-heavy electricity markets.

Building capacity, not reliance

Like a novice fisherman taking to open waters for the first time, non-technical users of with no prior experience of systems modelling will require guidance and technical support. Hosting workshops and creating partnerships with aligned organisations are all part of the TransitionZero plan.

Building capacity will also require a change in mindset. The underlying data and assumptions behind the insights and outputs driving decision making and investments should be subject to critical scrutiny, rather than unquestioning acceptance. Without access to the underlying data and model documentation, the validity of outputs should be questioned.

Model Builder encourages such transparency and will empower users with the confidence to question and stress test assumptions and data. Considering the immense complexity and range of assumptions that go into long-term scenario forecasting, there is a pressing need to move beyond blind acceptance of modelling results in ‘big brand’ reports, especially if they have outsized influence on policy formation and capital flows. Models should empower decision-makers to make more informed policy and investment choices, rather than cultivating a culture of dependency. Transparency is vital to achieving this.

Limitations of modelling

Building critical awareness will also help tackle common misconceptions about system modelling in general. Chief among these is the view that models can provide precise or deterministic predictions about the future. This leads to the mistaken judgement that a model can be ‘right’ or ‘wrong’.

In reality, model results should be seen as insightful and informed guidance rather than definitive predictions. A ‘wrong’ projection can be valuable when a user is adequately informed and empowered to understand the criteria and variables that produce undesirable or erroneous outcomes.

Models should be seen as decision support and research tools, providing insights and aiding in the complex decision-making process. In this regard, Model Builder is no different to other systems models. Where it stands out is its ability to deliver more accessible, best-in-class depth and granularity of insights that lead to better decisions.

We are also very much aware that modelling does not automatically lead to perfect decision-making. The exercise of scenario building and systems modelling provides useful information, but decisions also involve values, trade-offs, and considerations beyond what a model can capture. Stakeholder engagement, policy context, and ethical considerations are important factors in real world decision-making, along with model results.

Climate finance network effects

Broad adoption will be key to success. The more people use a common open source data platform, the greater the emphasis on improving the scope of models and the quality of the data in a way that benefits the rest of the community. This lowers the risk of duplication of effort, as awareness spreads that there is no need to reinvent the wheel.

Network effects are beneficial in other ways. When people from diverse backgrounds and disciplines work from the same open data source and modelling framework that can be shared, replicated, and interrogated, they will start to speak a common language. When there is a common base of knowledge and understanding for how energy transition opportunities, trade-offs and risks are assessed and factored into modelling, groups with competing priorities can more easily find common ground on the design of more cost-effective and efficient policies, markets and systems.

Our ambition is to make the most comprehensive, powerful and accessible open source energy data and modelling suite available, with global coverage. Achieving this will create a common data standard for energy system planning, facilitating transactions between donors, commercial funders, and recipients of climate finance and investment. This is core to delivering a future where clean energy is dependable and affordable for all.