©istock/Varijanta

A key challenge in achieving a successful transition to a low-carbon Europe is implementing the correct suite of policy measures that are based on robust evidence. Today policy-makers across Europe draw on integrated energy system models to inform long range climate mitigation and energy policy choices. Established European models such as PRIMES, TIMES, MESSAGE, EnergyPLAN and newer models such as POTEnCIA and OSeMOSYS consider all modes of energy (electricity, heating and transport) across all sectors of the economy in an integrated fashion, rather than treating individual modes in isolation which can lead to poorly informed insights.

Our research in integrated modelling started with a specific focus on the wider energy system in Ireland. We use the TIMES integrated model,[1] which is a techno-economic optimisation framework developed over the past 40 years through the International Energy Agency’s (IEA) Energy Technology Systems Analysis Program (ETSAP). The model allows users to generate future energy system pathways to meet energy needs at least cost, subject to user defined constraints. TIMES optimises for energy service demands (i.e. the utility we get from energy use) such as lighting, heating, passenger kilometres, tonnes of steel and cement etc. This is significant because as a society we don’t intentionally use energy, but rather have requirements for mobility, lighting, goods etc. In all, TIMES considers a wide range of over 1,300 technologies in the timeframe to 2050 from light bulbs, cars, fridges, heaters, boilers, power plants, bio refineries etc.

©istock/chombosan

Early in our research we recognised a number of limitations to our modelling techniques and identified key areas that required improvement. One of these areas was how integrated models like TIMES dealt with variable renewable generation such as wind and solar power. Many integrated models have a simplified temporal and technical resolution of the power system in order to keep problems computationally manageable, however this comes with the trade-off of poor representation of variability within the models. To resolve this issue, we developed soft-linking techniques to link the energy system model to dedicated power system models. This allows us to leverage the strength of high-resolution technical and temporal power system models. In doing this we could account for greater temporal resolution (15 minute or hourly simulations) and also capture important technical characteristics of the power plants such as ramp rates, start costs etc. We recently expanded these techniques to include the full EU 28 power and gas systems with water as our next target for development. The geographical expansion of the research was partially driven by the need to model greater interconnected markets (both gas and electricity) within the EU and also to understand the distribution of effort for decarbonisation across all EU Member States. Soft-linking techniques has the advantage that it allows us to verify the technical robustness of simulations but comes with the challenge that an extra model must be maintained and it requires modeller judgement on feedback to the energy system model.

©istock/SundayMorning014

Another challenge was the integration of land-use and agriculture into our models. Ireland is unique in Europe as over 30% of GHG emissions come from agriculture. This research required us to work closely with agricultural scientists to develop a framework where we could account for these emissions and model the interactions, particularly for land use competition, between the energy sector and agricultural sector. Our current research on land use and agriculture has an important focus on the role of bioenergy and the implications of indirect land use change (ILUC) and recent amendments to the Renewable Energy Directive (so called ‘ILUC directive’). While the science of ILUC is at an early stage our initial results point to increased costs for decarbonisation when ILUC is considered.

Like many research groups across Europe, we have seen our integrated models expand in size and complexity. Complexity is unavoidable in such large models due to the multi-dimensional and intricate nature of energy systems, but complexity has to be balanced with the inherent uncertainty in long range inputs such as fuel prices and macroeconomic estimations. The challenge of making models computationally manageable has often forced us to look at our simplifications and heuristics and ask the question, “are we making our models better or just getting the wrong answer quicker?” A recent focus of our research is trying to understand what level of complexity is appropriate in long term models given the uncertainty in inputs, and trying to understand how this value of complexity diminishes as we look into the future. We have found it beneficial to explore multiple pathways and seek out commonality between pathways rather than focus on deterministic solutions.

High-performance computing offers exciting possibilities for further development of integrated modelling, however many of the current architecture processes are challenging to parallelise. Projects like ‘BEAM-Me’ in Germany are investigating the potential for high-performance computing to enhance energy system models, and it will be interesting to see what developments occur.

Above all we have learned that modelling the future is a humble science and great care must be taken not to confuse model insights for predictions. Human behaviour, economic volatility and technology readiness are but a small section of elements that have big influence on resulting pathways from models. We must be aware that the boundaries of the energy system don’t stop at the end of the pipeline or cable; they extend in to our lives, communities, and wellbeing. Current modelling efforts primarily have a techno-economic focus, however the challenge of decarbonisation, and more recently the greater level of decarbonisation required by the Paris Agreement will require our modelling community to look outward to other disciplines to inform pathways that we as a society are willing to travel on together.

Dr Paul Deane

Dr Paul Deane is a research fellow with the Energy Policy and Modelling Group in University College Cork in Ireland. He has been working in the energy industry for approximately 15 years in both commercial and academic research. His research activities include integrated energy systems modelling to assess holistic pathways to low carbon energy futures. Paul is also a member of the Insight_E group which is a European, scientific and multidisciplinary think-tank.