This data-file is a global energy market model for the energy transition. It contains long-term energy supply-demand forecasts by energy source; based on a dozen core input assumptions. Total useful energy consumed by human civilization rises from 80,000 TWH pa to 140,000 TWH pa by 2050. The 2050 mix is 30% gas, 30% solar, 19% oil, 8% coal, 5% wind, 5% nuclear, 4% hydro. Numbers can be stress-tested in the data-file.
This model captures our forecasts for global energy supply, by source, through 2050. I.e., it is what we think will happen. The world is not currently on a path to limit global warming to 2C, and will more likely achieve a 2.5-3C climate change scenario. But we increasingly think a predictive model is more useful than an aspirational fantasy.
Global useful energy surpassed 80,000 TWH in 2021, having risen at 2.5% per year in the past decade. It will continue rising to 140,000 TWH pa by 2050, per our breakdown of global energy demand by region.
The largest gains in the energy mix are seen in solar and gas, both reaching around 30%, as these two resources are both very large and low cost. Conversely, coal plateaus and then peaks due to coal resource depletion.
Oil demand moves sideways, rising gently to a peak of 104Mbpd in 2030, which is driven by the emerging world, then slowly declining back to 90Mbpd in 2050.
Electricity rises from 40% to 65% of the world’s total useful energy by 2050, from 30,000 TWH generated in 2023 to 90,000 TWH, per our model of global electricity supply-demand, and the remaining non-electric energy is used for heat, motion, materials.
Total global investment in energy steps up from around $2trn pa to $4trn pa by 2050. The numbers are broken down in the data-file, but the largest increases are in solar and power grids, the latter comprising the biggest bottleneck in energy transition.
Other inputs include our modelling of wind and solar capacity additions, a long-term oil demand model, gas market models, coal supply forecasts and an increasingly favorable outlook on nuclear.
Annual data are provided back to 1750 to contextualize the energy transition relative to prior transitions in history.
At the end of 2024, we began considering whether the future might hold vastly more solar than previously envisaged, and a variant of our model, as discussed in this research note, is also available in the download.
You can ‘flex’ different assumptions, to see how it will affect future oil, coal and gas demand, as well as global carbon emissions.