Setup

This page explains how EPM's input system works: how files are organized, how the model finds them, and how to configure a run.

---

How it works

EPM uses a layered input system. Each layer has a distinct role:

flowchart TD
    CLI(["<b>python epm.py</b>\n--folder_input data_test"])

    subgraph routing ["Routing layer"]
        CONFIG["<b>config.csv</b>\nMaps parameter names to file paths"]
        SCEN["<b>scenarios.csv</b>\nOverrides baseline paths (optional)"]
    end

    PS["<b>pSettings.csv</b>\nFeature flags &amp; global parameters"]

    subgraph data ["Data layer"]
        CSVS["<b>Data CSVs</b>\nsupply/ · load/ · reserve/ · trade/ · ..."]
    end

    GAMS["<b>GAMS</b>\nOptimization core"]

    CLI --> CONFIG
    SCEN -.->|overlays| CONFIG
    CONFIG --> PS
    CONFIG --> data
    PS --> GAMS
    CSVS --> GAMS

1. You point EPM to an input folder (--folder_input). All data lives inside it.
2. config.csv is the master routing table: it maps every parameter name to its CSV file path within that folder.
3. pSettings.csv is one of those files. It controls model behavior: features, discount rates, reserve margins, CO₂ constraints, etc.
4. All other CSV files contain the actual data: plants, demand, fuel prices, transmission limits, etc.
5. If you pass --scenarios, a scenarios.csv overlays changes on the baseline. Only the files that differ need to be specified.

---

config.csv

The master routing table. Located at epm/input/<folder_input>/config.csv.

metadata,paramNames,file
Global model settings,pSettings,config/pSettings.csv
Planning horizon years,y,config/y.csv
Zone-country mapping,zcmap,config/zcmap.csv
Demand profiles,pDemandProfile,load/pDemandProfile.csv
...

Each row maps a paramName (used internally by GAMS) to a file path relative to the input folder. To use a different file for any parameter, just change the path here.

Category	Parameters
GENERAL	pSettings, y, zcmap, pHours
STATIC	pDays, mapTS
LOAD	pDemandProfile, pDemandData, pDemandForecast, sRelevant, pEnergyEfficiencyFactor
SUPPLY	pGenDataInput, pGenDataInputDefault, pAvailability, pAvailabilityDefault, pCapexTrajectories, pCapexTrajectoriesDefault, pFuelPrice, pVREProfile, pVREgenProfile
RESERVE	pPlanningReserveMargin, pSpinningReserveReqCountry, pSpinningReserveReqSystem
CONSTRAINT	pEmissionsTotal, pEmissionsCountry, pMaxFuellimit, pCarbonPrice
TRADE	zext, pExtTransferLimit, pLossFactorInternal, pMaxAnnualExternalTradeShare, pNewTransmission, pTradePrice, pTransferLimit
H2	pH2DataExcel, pAvailabilityH2, pCapexTrajectoryH2, pExternalH2, pFuelDataH2

pSettings.csv

Controls which model features are active and sets global economic parameters. Located at config/pSettings.csv.

Key flags; see Input Catalog for the full parameter table.

Category	Key parameters
Capacity & Dispatch	fEnableCapacityExpansion · fDispatchMode
Economics	WACC · DR · VoLL · CO2backstop
Trade	fEnableInternalExchange · fEnableExternalExchange · fAllowTransferExpansion
Reserves	fApplyPlanningReserveConstraint · sReserveMarginPct · fApplyCountrySpinReserveConstraint
Policy	fApplyCountryCo2Constraint · sMinRenewableSharePct · fEnableCarbonPrice
Features	fEnableStorage · fEnableCSP · fEnableH2Production · fEnableEconomicRetirement

Custom / Default system

For generation data, availability, and CAPEX trajectories, EPM distinguishes two file types:

• xxDefault.csv: default values indexed by zone · technology · fuel, shared across plants
• xxCustom.csv: plant-level values that override defaults where specified

Case	Behavior
Field missing in Custom	Value from Default is used
Field present in Custom	Value overrides the Default

This lets you populate the Default files once (typically from the CCDR standard dataset) and only specify deviations for particular plants in the Custom files.

Getting started

Copy the Default files from data_test or data_test_region, then adapt them to your zones, technologies, and fuels. Ensure all zone/tech/fuel combinations are covered in at least one file.

Warning

For pAvailabilityDefault.csv and pCapexTrajectoriesCustom.csv, missing values for any zone/tech/fuel combination may cause errors. For pGenDataInputCustom.csv, missing optional fields (e.g., Capex, VOM) silently default to zero.

Scenarios

Scenarios overlay changes on the baseline config.csv. Define a scenarios.csv where each column is a scenario variant — only the files that differ need to be specified. Empty cells inherit the baseline.

paramNames,HighDemand,LowFuel
pDemandForecast,demand/high_demand.csv,
pFuelPrice,,supply/fuel_low.csv
pSettings,pSettings_alt.csv,

Run with:

python epm.py --folder_input data_test --scenarios scenarios.csv

Scenario type	How
Standard	Override file paths in scenarios.csv
Sensitivity	--sensitivity — systematic parameter variation
Monte Carlo	--montecarlo — sample from uncertainty distributions