This interface allows you to run custom optimisations for a sector-coupled model of the European energy system, PyPSA-Eur-Sec. You can, for example, explore different scenarios to reach net-zero carbon dioxide emissions across electricity, heating, transport and industry. There are more details and context on the underlying model in a slide deck from October 2021.
The model minimises the costs of the energy system assuming all capacity investments in generation, storage, energy conversion and energy transport can be re-optimised. Energy services (electricity, heating, transport, industrial demand) are provided at today's levels by default, but they can also be altered. Default cost assumptions are taken from forecasts for 2050, mainly from the Danish Energy Agency Technology Data. A weighted average cost of capital of 7% is applied. 45 regions are assumed. A full year of representative weather and load data is used, but sampled n-hourly. Please carefully read the limitations below.
The user interface, optimisation model construction, solver and
post-processing are all built with free software and open data
only. The code can be downloaded from the GitHub
repository pypsa-server. It
uses the Python for
Power System Analysis (PyPSA) energy optimisation
framework, open weather data from the European Centre for
Medium-Range Weather Forecasts
(ECMWF) ERA5
reanalysis dataset,
the atlite
library for converting weather data to generation profiles,
and free software for the server infrastructure (GNU/Linux,
nginx, Flask, gunicorn, Redis). The optimisation problem is
solved by the high-performance open source solver HiGHS.
Limitations
There are many limitations to the model. Only a selection are given here:
For computational reasons, time series are only sampled n-hourly with maximum resolution of 25-hourly for Europe and 3-hourly for single countries. This means that not all extreme events of high residual load are seen. It also leads to an underestimation of short-term flexibility requirements (e.g. for stationary batteries and demand-side management in battery electric vehicles) but correctly captures more costly medium (weekly) and seasonal flexibility needs. Some 3-hourly results for Europe are available for comparison.
The model assumes Europe is completely self-sufficient in clean energy and does not consider the import of clean electricity, hydrogen, ammonia, steel or hydrocarbons from outside Europe. Imports of green fuels from renewables-rich regions could reduce costs and, for example, the need for extensive hydrogen networks.
Some assumptions about the ratio of primary and secondary production for steel, aluminium and petrochemicals, as well as process switching, are made exogenously with a net-zero emission energy system in mind. For example, it is assumed that 70% of steel is produced via the secondary route (electric arc furnace (EAF)) and the remaining 30% is produced via a primary route with direct reduction of iron ore using hydrogen followed by EAF. The fraction of aluminium recycled via the secondary route is raised to 80%. 60% of high value chemicals (HVC) come from the primary route, while the rest is recycled.
The model meets primary petrochemical feedstock with naphtha only and does not consider other pathways such as methanol-to-olefin processes.
Exogenous decisions are also met in the transport sector with net-zero emission systems in mind. For example, shipping moves to hydrogen (either liquid hydrogen or ammonia), all aviation is still supplied with liquid hydrocarbons, while the ratio of land vehicles between electric, hydrogen fuel cell and liquid hydrocarbon internal combustion engine can be set via the submission form.
The model may build infrastructure that does not meet with societal acceptance.
Infrastructure may not be feasible to build given supply chain and personnel training limitations, which are not included in the model. For example, the implementation of efficiency measures in buildings requires a large, trained workforce.
In this simplified 45-node model there are not enough model regions to simulate the energy networks properly.
There is not enough spatial and temporal resolution to see the full variability of wind, solar and demand.
Ideally decades of weather data would be used to sample historical weather patterns, as well as data that takes into account future climate change.
The pathway of asset investment from today until a low-carbon future should be optimised to see what needs to be done and when, and to minimise stranded assets. Pathway analysis with this model can be found in Victoria et al, 2021. Soon we will include pathway analysis on this website.