How to build a fundamental power market model

What a fundamental model is and why it matters

A fundamental power market model seeks to replicate real-time electricity price formation by integrating physical system constraints with economic decisions. Instead of depending solely on historical data, it constructs the market from scratch, modelling the interactions among generation, demand, and interconnections across various scenarios.

This approach differs from statistical models that rely on regression or machine learning to identify correlations in past price data. While statistical methods are effective for short-term trading and spotting patterns, they tend to falter when market fundamentals change, such as during rapid renewable energy expansion, fuel price shocks, or the introduction of new policies.

Fundamental models, by contrast, reflect the physical realities of the system and can be adapted to new conditions. They are used for price forecasting, asset valuation, policy analysis, and the development of trading strategies. In an increasingly complex power system, these models provide the structure needed to interpret volatility and anticipate change.

Key inputs: supply, demand, and interconnections

Building a strong fundamental power market model begins with comprehensive, precise inputs. Each component - demand, supply, and interconnection - influences the dispatch outcome and price formation.

Demand

Demand modelling encompasses both the structural and behavioural factors of electricity consumption. It primarily involves predicting load curves, which represent hourly usage patterns throughout the day and across different seasons. More sophisticated models also account for temperature sensitivity, electrification trends like EV adoption, and behavioural reactions to price signals or demand-side flexibility. The main objective is to create accurate hourly demand profiles tailored to each scenario.

Supply

On the supply side, the model needs to include the full generation fleet, comprising thermal, renewable, and storage units. Each unit is characterised by technical and economic parameters like capacity, availability, efficiency, and cost structure. Marginal costs are usually influenced by fuel prices and carbon costs, shaping how units are ordered in the merit stack. For renewables, output generally depends on weather-dependent factors such as wind speeds or solar irradiance patterns.

Interconnections

Cross-border flows can significantly influence local price formation. Interconnectors allow electricity to move between markets, helping to equalise prices when capacity is available. Modelling interconnection requires data on transmission capacity, flow allocation rules, and policy assumptions such as market coupling or export restrictions. These inputs collectively prepare the market simulation, the process that converts physical data into economic outcomes.

Building the merit order and dispatch simulation

Once inputs are assembled, the next step is to construct the merit order: a ranked list of generation units by marginal cost.

Merit order construction

Each generating unit is arranged in increasing order of variable cost, including:

• Fuel and carbon prices

• Efficiency factors and heat rates

• Start-up or variable operating costs

• Unit availability and maintenance schedules

Renewable and nuclear assets usually occupy the lowest position with nearly zero marginal cost, succeeded by progressively costlier thermal units. This hierarchy determines which plants operate at different demand levels and sets the clearing price.

Dispatch simulation

For each hour of the simulation period, demand is matched to available supply by “dispatching” units in merit order until the load is satisfied. The last plant needed to meet demand determines the marginal price and is often referred to as the system marginal price.

A realistic dispatch model must also consider renewable output variability, unit outages, transmission bottlenecks, and reserve requirements. It produces a series of hourly or half-hourly clearing prices, which can then be grouped into daily, monthly, or yearly averages.

Running scenarios and sensitivity analysis

Once the base case is calibrated, scenario analysis enables users to examine how the market responds under various assumptions. Common methods include weather variability, fuel price sensitivities, and policy changes such as carbon caps or renewable targets.

Stress testing is vital for understanding market stability. Extreme scenarios, such as prolonged cold winters or sudden fuel price spikes, aid in evaluating resilience and identifying volatility triggers. This analysis underpins risk management and strategic planning.

Validating results and interpreting outputs

No model is complete without thorough validation. Backtesting compares simulated prices with historical data to evaluate accuracy and identify structural gaps. For instance, if the model consistently underestimates peak prices, it may indicate missing flexibility constraints or underestimated scarcity pricing. Calibration then enhances the model’s realism by adjusting outage rates, fuel correlations, or load forecasts.

Once validated, results can be interpreted through several analytical lenses:

• Price spreads between peak and off-peak hours reveal volatility and scarcity risk

• Renewable capture rates assess how well generation assets monetise production

• Cross-border flows help evaluate the contribution of interconnectors to price convergence

• Marginal plant frequency indicates which technologies are most often price-setting

For traders and strategists, these outputs translate into actionable intelligence, supporting curve construction, hedging, and asset optimisation.

Conclusion

Fundamental power market modelling introduces clarity and organisation into an increasingly complex system. By connecting supply, demand, and policy factors, it creates a consistent framework to evaluate assumptions, assess risks, and predict market behaviour.

Whether for asset valuation, long-term planning, or trading strategies, these models convert data into understanding, turning complexity into clarity, and clarity into profitable foresight.