Method

A unique feature of our work is the quantification of our narratives.  These are quantified using a model known as “TIMES”, an integrated energy-systems model. TIMES simultaneously models all components of the energy system, ensuring any interdependencies are reflected.

TIMES is a technology rich, bottom-up model generator, which uses linear programming to produce a least-cost energy system, optimised according to a number of user-specified constraints, over the medium to long-term. It is used for the exploration of possible energy futures based on contrasted scenarios.

TIMES is an advanced version of the MARKAL modelling family. TIMES is a linear program optimization, meaning that it minimizes the total discounted costs, through time, of meeting all energy service demand.

This modelling approach has several advantages.

Firstly, TIMES is an integrated energy-systems model, meaning that it simultaneously models all components of the energy system, ensuring any interdependencies are reflected – the impact of gas exploration and supply, or the impact in changes in freight transport technology, on electricity generation, and the consequential impact on electricity costs for what technologies residential consumers may choose to buy.  This way we can always trust that the results from the TIMES model are internally consistent at every point in time.

Secondly, as much as possible, TIMES requires the services demanded of the energy system as inputs, not simply forecasts of energy demand.  The services are expressed as, for example, vehicle kilometres travelled and the space required for heating and lighting, or the demand for heat at an industrial site.  It is then up to the model to determine which are the optimal technologies to use to supply these service demands – which type of car, or which type of heating device.  Simultaneously, it determines the optimal fuel for those energy-consuming technologies will be procured and delivered.

While this relieves us of having to form specific forecasts for energy demand, it does require us to provide the model with forecasts of service demand.  The BEC2060 project used inputs from a variety of sources – using two of MoT’s transport outlook scenarios to form projections of the need for passenger and freight transport, using sub-sectoral GDP forecasts project the future service demand from the commercial, agriculture and industrial sector, and population to form the basis of the residential service demand projections.

Finally, a feature only available to linear programming optimization models, is that TIMES produces a rich array of economic information as part of its solution.  Rather than simply tell us what the optimal quantities of different fuels and technologies are for each scenario, it also tells us what the implied commodity prices are, and how far away technologies are (economically) from becoming “optimal”.   In our view, this has been a missing piece of the dialogue in so many scenario-based discussions. The precise estimates of energy demand and supply become decreasingly useful the future into the future we peer; but knowledge of how the relative economics of different technologies are playing out is far more useful.