Energy Systems Planning

TIMES · MESSAGE · OSeMOSYS

Multi-decade · Multi-sector · Decarbonization Pathways

Beyond a single country's electricity mix: the energy-systems planning framework models the entire energy system — electricity, heating, transportation, industry, feedstocks — over multi-decade horizons, typically with a 30–50-year planning horizon. TIMES, MESSAGE, and OSeMOSYS are the canonical bottom-up LP/MIP frameworks used by IEA, IIASA, and national ministries to build net-zero pathways, evaluate policies, and stress-test decarbonization scenarios. This is the mathematical backbone of the IPCC's Integrated Assessment Model (IAM) ecosystem.

The problem

Whole-system decarbonization pathways

National net-zero planning can't be done one sector at a time. Electrifying transport adds to power demand; heating electrification doubles peaks; industrial hydrogen requires massive electrolyzer build-out. Every technology competes with every other, subject to resource constraints, lifetime stock-rollover dynamics, and policy targets. Energy-systems models (ESMs) take a bottom-up, technology-rich view: hundreds of individual technologies (ICE car, BEV, gas furnace, heat pump, blast furnace, H2-DRI, wind, gas-CCS, ...), each with capital cost, efficiency, fuel cost, lifetime, and availability. Outputs: capacity and activity of each tech in each year, implied fuel mix, emissions, total discounted cost.

The mathematical structure is a large linear program (or MIP if discrete decisions dominate). Time indexed by year (possibly with time-slices within year for demand-supply matching); commodities (fuels, services, emissions); technologies that convert commodities to commodities. The Reference Energy System (RES) is the flow diagram; the LP is the least-cost solution over the whole horizon subject to demand satisfaction and policy constraints.

Historical note
The MARKAL framework originated at Brookhaven National Laboratory in the 1970s oil-crisis era; TIMES (The Integrated MARKAL-EFOM System) extended it and has been maintained by IEA-ETSAP since 1996. MESSAGE was developed at IIASA for long-horizon global scenarios. OSeMOSYS (Howells et al., 2011) is the modern open-source framework for low- and middle-income country use. PyPSA-Eur (Hörsch & Brown, 2017) provides hourly-resolution European-wide ESM in open-source Python. These tools underpin the IPCC's scenario ensemble and every major national decarbonization plan.

Mathematical formulation

Bottom-up least-cost LP over RES

Core LP

Technologies $k$ convert input commodities to output commodities. Activity $A_{k,t}$ in year $t$; new capacity $NC_{k,t}$; installed capacity $K_{k,t}$.

Objective: minimize discounted total system cost:

$$\min \sum_{t} \frac{1}{(1+r)^{t-1}} \Big[ \sum_k I_k NC_{k,t} + \sum_k FOM_k K_{k,t} + \sum_k VOM_k A_{k,t} + \sum_c C_{c,t}^{\mathrm{imp}} IM_{c,t} \Big]$$

Capacity accumulation: $K_{k,t} = K_{k,t-1} + NC_{k,t} - R_{k,t}$ (retirement at end of life).

Commodity balance: for every commodity $c$ and period,

$$\sum_k \alpha_{k,c,\mathrm{out}} A_{k,t} = \sum_k \alpha_{k,c,\mathrm{in}} A_{k,t} + D_{c,t} \qquad \forall c, t$$

Activity-capacity link: $A_{k,t} \le CF_k \cdot K_{k,t} \cdot 8760$. Emissions constraint: $\sum_{k,t} \epsilon_k A_{k,t} \le E^{\mathrm{cap}}$. Emissions as priced: substitute carbon adder in objective.

Time slices & dispatch coupling

Pure annual LP misses renewable variability and peak-demand events. TIMES adds time slices (representative day-night, seasonal, or hourly within year) so each activity has sub-annual structure, allowing storage and renewables to be modeled with their full operational character.

Stochastic / scenario variants

Uncertain fuel prices, technology learning, climate policy resolved via scenario-based stochastic LP. Robust-LP and multi-objective (cost + emissions + security) variants are common for policy analysis.

Complexity

Typical TIMES or OSeMOSYS LP: 10–30 years × 12 time-slices × 100–500 technologies × 20–50 commodities → millions of variables and constraints. Solves in hours on CPLEX/Gurobi. Hourly-resolution PyPSA-Eur scales to 8760 time-slices, solved via decomposition.

Real-world data

IEA-ETSAP TIMES models

ETSAP maintains national and regional TIMES models for 70+ countries. Referenced by every major national decarbonization study.

IIASA MESSAGEix

IIASA MESSAGEix is the global-scale ESM used in IPCC scenarios including SSP pathways.

OSeMOSYS open-source

OSeMOSYS is the open-source LP ESM used widely in developing-country electrification studies. Python and GNU MathProg interfaces.

PyPSA-Eur

PyPSA-Eur provides hourly-resolution European electricity + sector-coupled network model. Reproducible, open, published data pipelines.

Solution interpretation

ESMs produce mitigation pathways: year-by-year technology mix, fuel use, emissions, and cost to reach a target (typically net-zero-by-2050 or 1.5 °C-compatible carbon budget). Scenarios compare policy choices: a carbon tax, an EV mandate, a nuclear phase-out, a hydrogen subsidy. Each produces a different feasible-region pathway and implies different winners and losers.

Key ESM outputs: shadow prices on binding constraints (the emissions-cap shadow price is the implied carbon price); marginal abatement cost curves by sector; technology adoption S-curves over the planning horizon. These feed macro-economic IAMs (MERGE, WITCH, GCAM) that couple ESMs to GDP/welfare modules.

Modern critique of ESMs: traditional deterministic least-cost LPs underestimate resource-adequacy needs in renewables-heavy systems. Current research integrates hourly dispatch, stochastic weather, and sector-coupling details directly into the ESM (e.g., PyPSA-Eur, SEEDS, TEMOA with sub-annual flexibility).

Extensions & variants

Sector-coupled ESM

Explicit coupling between electricity, heat, transport, and industry. Multi-energy systems at the national or continental scale.

Refs: Brown et al. (2018) PyPSA-Eur-Sec.

Integrated Assessment Models

ESM coupled to macro-economic modules. Used by IPCC: IMAGE, MESSAGE, GCAM, REMIND, WITCH.

Refs: Weyant (2017); IPCC AR6 WG3 Ch. 3.

Stochastic / robust ESM

Uncertain fuel prices, tech learning, and demand growth. Scenario reduction critical.

Refs: Mavromatidis, Orehounig & Carmeliet (2018); Schaefli & Gupta (2014).

Multi-objective (cost + emissions + security)

Pareto frontier of cost vs carbon vs energy security (import dependence) chosen by policymakers.

Refs: Koltsaklis & Dagoumas (2018).

Soft-linked macro-energy

Iterative coupling of ESM with macroeconomic CGE models. Captures rebound effects and price-induced demand response.

Refs: Helgesen & Tomasgard (2018); Fortes et al. (2014).

Open-science ESM reproducibility

FAIR data/code, model intercomparison projects (MIPs) coordinated through ETSAP and IIASA. Standard framework for transparent scenario-based policy analysis.

Refs: DeCarolis et al. (2017); Pfenninger et al. (2018).

Key references

[1]
Howells, M., et al. (2011).
“OSeMOSYS: The Open Source Energy Modeling System.”
Energy Policy, 39(10), 5850–5870. doi:10.1016/j.enpol.2011.06.033
[2]
Loulou, R., Remne, U., Kanudia, A., Lehtila, A., & Goldstein, G. (2005).
“Documentation for the TIMES model.”
ETSAP IEA. iea-etsap.org
[3]
Pfenninger, S., Hawkes, A., & Keirstead, J. (2014).
“Energy systems modeling for twenty-first century energy challenges.”
Renewable and Sustainable Energy Reviews, 33, 74–86. doi:10.1016/j.rser.2014.02.003
[4]
DeCarolis, J., Daly, H., Dodds, P., et al. (2017).
“Formalizing best practice for energy system optimization modelling.”
Applied Energy, 194, 184–198. doi:10.1016/j.apenergy.2017.03.001
[5]
Hörsch, J., & Brown, T. (2017).
“The role of spatial scale in joint optimisations of generation and transmission for European highly renewable scenarios.”
14th International Conference on the European Energy Market (EEM). doi:10.1109/EEM.2017.7982024
[6]
Brown, T., Schlachtberger, D., Kies, A., Schramm, S., & Greiner, M. (2018).
“Synergies of sector coupling and transmission reinforcement in a cost-optimised, highly renewable European energy system.”
Energy, 160, 720–739. doi:10.1016/j.energy.2018.06.222
[7]
Pfenninger, S., Hirth, L., Schlecht, I., et al. (2018).
“Opening the black box of energy modelling: Strategies and lessons learned.”
Energy Strategy Reviews, 19, 63–71. doi:10.1016/j.esr.2017.12.002
[8]
Weyant, J. (2017).
“Some contributions of integrated assessment models of global climate change.”
Review of Environmental Economics and Policy, 11(1), 115–137. doi:10.1093/reep/rew018
[9]
IEA. (2023).
Net Zero Roadmap: A Global Pathway to Keep the 1.5 °C Goal in Reach.
International Energy Agency. iea.org/net-zero
[10]
IPCC. (2022).
AR6 WG3: Mitigation of Climate Change, Chapter 3 (Mitigation Pathways).
Intergovernmental Panel on Climate Change. ipcc.ch/ar6/wg3
Frameworks listed (TIMES, MESSAGE, OSeMOSYS, PyPSA) require commercial or open-source LP/MIP solvers for real analysis. This page introduces the formulation conceptually; running a full ESM requires the actual toolchains.