Plan generation, grid and demand together and stop paying for lost power.

Plan every piece of the system together

A plan can only find the least cost if every piece of the system is in the model. Generation, transmission, distribution and demand are often planned by different bodies on different cycles. The result can be a solar farm waiting for a grid connection, or a district whose old cables lose more power than they should. The UN's 2026 SDG report says renewables supplied over 30% of global electricity consumption in 2023, and wind and solar output moves with the weather, which makes planning each piece alone harder. World Bank authors in the same report call least-cost electrification planning essential for using scarce public resources well.

Start from open data, then add the utility's records

A first model can start from open data and cover plants, storage, lines and hourly demand by area.

• Power lines and substations mapped in OpenStreetMap, which PyPSA-Earth uses to build the grid topology, meaning the map of how lines connect
• Weather data turned into hourly wind and solar output with atlite
• Plant capacity and demand statistics published by national statistics offices and regulators
• Meter, outage and loss records from the utility, added in a second stage
• Electrification and building plans from the city or ministry, used to forecast new demand

Match today's grid before planning tomorrow's

We would build the network and run a power flow, the standard calculation of how current moves through the lines, to check that the model matches the flows and voltages the utility measures. Then we would replay a past year hour by hour. Where the model's flows or losses differ from recorded values, we fix the data or state the gap. Only then would the optimization search for the cheapest plan that meets demand and reliability limits. Each plan is rerun under several weather years, so planners see which choices hold up.

Let every investment compete on the same terms

Cutting losses or moving demand should compete with new plants and lines on cost and power saved. A ministry, regulator or utility would get evidence it can take to a board.

• A ranked list of investments with cost, losses avoided and emissions for each
• Lines and substations near their limits, and where efficiency programs or moving demand to other hours would delay an upgrade
• An estimate of where losses are physical, such as heat in old cables, and where they may come from theft or billing errors
• A plain-language summary of each scenario, checked by an engineer
• The full model and data, so the utility's engineers can rerun it

A lower average cost can still raise someone's bill

A tariff or investment that lowers average cost may still raise bills for one group, and the report shows when it does. The model works with demand totals by district or by feeder, the line that supplies a neighborhood. Household meter readings stay in the utility's systems. Every run is logged with its inputs and versions, and engineers and officials approve every decision.

Open tools let regulators check the plan

A grid plan that the public pays for should be open to regulators and the public. Each tool below is open source or open data.

• PyPSA, an open framework for simulating and optimizing power and energy systems, including investment planning and dispatch, the choice of which plants run each hour
• PyPSA-Earth, an open optimization model for energy system futures around the world, with data drawn from open sources
• pandapower, a network calculation program for the analysis and optimization of power systems
• atlite, which turns weather data into renewable power potentials and time series
• HiGHS, an MIT-licensed solver for large linear and mixed-integer optimization problems
• OpenStreetMap, for mapped power lines and substations