Modelling flexibility: how storage and demand response change price dynamics

Flexibility modelling enables analysts to understand how these new resources behave, how they move across markets, and how they absorb or amplify volatility. For traders, system planners, and operators, the ability to simulate flexible behaviour is now crucial for interpreting spreads, valuing assets, and predicting turning points in price dynamics.

The growing importance of flexibility in power systems

Renewables' variability now defines modern power systems. Wind droughts, high-solar-output periods, and rapid technology ramps all demand balancing resources that can respond quickly. Simultaneously, market designs across Europe and beyond increasingly favour rapid responses, energy shifting, and system support; whether through wholesale arbitrage, balancing markets or system services.

This shift has propelled the growth of energy storage and demand response. Batteries have transitioned from niche demonstration projects to large-scale, commercially optimised assets. Industrial loads, EV smart-charging, and flexible heat demand are emerging as new tools for system balancing. The outcome is a market ecosystem where flexibility influences volumes and prices over hours, days, and even weeks.

Understanding these dynamics requires models that capture operational constraints, incentives, and cross-market interactions, rather than relying on static merit-order logic.

How storage and demand-side assets operate

Battery optimisation: state of charge, cycle limits and degradation

A set of tightly coupled constraints governs battery behaviour:

• State of charge (SoC) limits

• Maximum charge/discharge rates

• Round-trip efficiency losses

• Cycle depth and degradation costs

• Minimum revenue thresholds for cycling

Optimisation models generally view storage as an arbitrage engine that purchases energy during low-price periods and sells it during peak times. However, the reality is more complex: batteries respond to intraday signals, balancing prices and system-service opportunities simultaneously. A realistic model must account for the opportunity cost of using capacity in one market versus another, particularly when ancillary revenues are high.

Degradation is significant. Each cycle carries an implicit cost, meaning storage is selective: it only acts when spreads justify using up part of its lifetime. This behaviour directly affects price formation because batteries do not neutralise every peak or trough; they flatten only those supported by value.

Demand flexibility: industrial load, EVs and heat pumps

Demand-side response operates differently. Industrial loads may be shifted or reduced when prices exceed a certain threshold, while EV charging and heat pumps can be scheduled over several hours without compromising user comfort.

Demand flexibility models often incorporate:

• Time windows within which the load can be moved

• Cost or comfort penalties for changing consumption

• Baseline consumption profiles

• Aggregator behaviour and participation rates

These assets help smooth demand peaks, lessen price spikes, and shape net load curves. However, their impact heavily relies on participation levels and the sophistication of the control strategies used.

Building flexibility into market models

Representing dispatch constraints and opportunity costs

Flexibility modelling involves incorporating operational constraints directly into the market simulation. For batteries, this encompasses SoC development, ramp rates, efficiency, and degradation. For demand response, it covers shiftable loads, curtailment limits, and rebound effects.

Opportunity cost is central. A battery holding energy for a potential evening peak might forego earlier intraday spreads. Demand response might choose one event window over another. Realistic modelling treats these as optimisation problems, not simple “if price > threshold, act” rules.

Co-optimising between wholesale and ancillary markets

Modern storage assets engage in various value streams, including wholesale arbitrage, balancing energy, frequency response, reserve services, and sometimes capacity markets. Co-optimisation is crucial to prevent artificially inflating or suppressing revenue.

A robust modelling framework should:

• Allocate capacity dynamically across markets based on anticipated revenue.

• Respect delivery obligations for committed system services.

• Capture the interaction between energy prices and service availability.

Without co-optimisation, models often overestimate arbitrage opportunities or underestimate the value of system services, resulting in inaccurate price impact predictions.

Forecasting the impact on prices and volatility

Modelling dampening of peak prices and volatility

As flexibility enters the system, its primary effect is to smooth out peaks and fill in troughs. Battery discharging during peak hours tends to lessen extreme price fluctuations, while charging during low-price periods raises the floor slightly. Demand response similarly reduces consumption during tight system conditions.

But flexibility does not completely eliminate volatility; instead, it redistributes it. Modelling this involves probabilistic forecasting and sensitivity analysis that capture:

• Renewable variability

• The availability of flexible capacity at different times

• The strategic behaviour of storage making forward decisions

• Local network constraints and congestion

By modelling extensive sets of renewable and demand scenarios, analysts determine when flexibility reaches its limit - times when prices may still rise rapidly.

Storage as an arbitrage engine between hours and days

Storage not only redistributes energy within a single day but also over several days during weather events clusters. As long-duration storage technologies develop, multi-day arbitrage gains greater significance.

Key modelling considerations include:

• Consecutive-day SoC management

• Anticipating future scarcity events

• Interactions with balancing markets during prolonged renewable lulls

These behaviours lead to new price patterns, characterised by wider spreads during windy versus calm periods and more abrupt shifts between system-tight and system-loose days.

Trading implications for a flexible system

How flexibility alters price duration curves

As batteries and demand response grow, price duration curves tend to flatten. This leads to fewer extreme high prices, although the mid-range values change based on the timing and method of storage dispatch. Analysts must incorporate these shifts into their trading strategies.

Important signals include:

• Decrease in the occurrence of scarcity prices

• Emergence of new mid-range price clusters

• Increased correlation between renewable conditions and price spreads

These modifications affect the optionality value, the profitability of strategies spanning multiple hours, and the risk profile of positions.

New opportunities: intraday spreads and imbalance arbitrage

Flexibility creates opportunities as well as dampens risk. Traders increasingly focus on:

• Intraday volatility driven by renewable forecast error

• Balancing market spreads, especially when storage is near SoC limits

• Price reversals caused by competing batteries responding to the same signals

• Evening and morning ramps amplified by EV and heat behaviour

Analysts capable of modelling the interaction of these factors gain a significant advantage in pinpointing where volatility, and value, will persist.

Conclusion

Flexibility doesn’t eliminate volatility; it redefines it. Batteries, demand response, and system services alter the timing, magnitude, and drivers of price movements. Analysts who account for operational constraints, co-optimised behaviours, and cross-market incentives will be best positioned to interpret the evolving nature of price dynamics. As markets continue to develop, those who model flexibility accurately will lead in tomorrow’s trading strategies and system planning decisions.