Energy Trading & Battery Management

Turn energy forecasts into governed asset decisions

Decision Deck brings together price, demand and generation forecasts with asset condition, trading policy, grid constraints and human authority.

Charge, discharge, hold, bid and buy decisions reflect the value available in the market and the physical, commercial and contractual limits of the asset.

Stronger revenue capture · protected asset life · clearer trading control.

The decision surface

Trading value, with the asset in view

A forecast informs the trade. The decision also needs to account for state of charge, cycle cost, degradation, export limits, grid commitments, warranty conditions, market rules and the value of waiting.

Decision Deck creates a controlled decision path around every trading interval. Each action reflects the evidence available, the policies in force and the consequences for the asset.

  1. 01

    More value captured

    Trading actions reflect expected value across price, demand and generation conditions.

  2. 02

    Asset limits held

    Battery, grid, contract and warranty constraints shape the decision before an action is taken.

  3. 03

    Decision authority retained

    Routine decisions can run automatically while higher-value, uncertain or unusual cases reach the right person.

Choose the trading action that fits the moment

Price movement alone does not determine whether a battery should charge, discharge or wait. The best action depends on future opportunity, current asset condition, available capacity, market timing and the risks attached to committing now.

What Decision Deck does

Decision Deck evaluates those conditions together. It selects, recommends or routes the next action with a clear record of the signals, trade-offs and policy conditions behind it.

Components used

  • Forecasting

    Predicts price movement, demand, renewable generation, volatility and potential value windows.

  • Predictive Models

    Estimate expected revenue, forecast confidence, dispatch value and trading risk.

  • Recommendations

    Selects or recommends the next action: charge, discharge, hold, bid, buy, curtail, defer or seek approval.

  • Policy Dynamics

    Applies trading rules, risk appetite, gate closure times, market constraints and owner instructions.

  • Evidence Vault

    Retains the evidence behind each recommendation and executed action.

Protect the asset while pursuing value

A high-value trade can still create the wrong outcome when it accelerates degradation, breaches a grid condition or reduces the asset's ability to capture a stronger opportunity later.

What Decision Deck does

Asset condition becomes part of the trading decision. Revenue, cycle impact, degradation cost, constraint exposure and future optionality are assessed together rather than in isolation, creating a clearer balance between short-term value and long-term asset economics.

Components used

  • Multi-output Models

    Estimate revenue, degradation cost, cycle impact, constraint risk and future optionality in the same decision context.

  • Policy Dynamics

    Enforces battery limits, warranty conditions, export constraints, grid obligations and risk thresholds.

  • Risk Stratification

    Separates low-risk dispatch from constrained, high-degradation or approval-required actions.

  • Exception Handling

    Stops the usual route when telemetry, constraint or asset-condition signals make an action unsuitable.

  • Evidence Vault

    Records the trade-off considered and the final asset decision.

Keep traders in control where judgement matters

Energy trading has clear boundaries for automation. Some decisions are routine and time-sensitive. Others carry higher asset exposure, uncertainty or operational consequence.

What Decision Deck does

Decision Deck routes each decision according to its materiality. Low-risk actions can follow approved operating rules. Higher-value or unusual cases can require trader, operator or risk approval before execution, with a clear record of when human judgement was used and why.

Components used

  • Case Triage

    Routes decisions by value, uncertainty, time pressure, operational consequence and approval requirement.

  • Risk Stratification

    Distinguishes routine actions, medium-risk recommendations, unusual conditions and high-impact approvals.

  • Recommendations

    Presents the proposed action alongside the reasons for approval, rejection or override.

  • Policy Dynamics

    Defines what can execute automatically and where trader, operator or risk approval is required.

  • Evidence Vault

    Records the recommendation, decision, override rationale and final action.

Understand where trading value was won or lost

Revenue outcomes can reflect forecast accuracy, market movement, dispatch quality, grid constraints, policy choices and human intervention. Those effects need separating before a trading strategy can improve.

What Decision Deck does

Decision Deck retains the context around every interval. Performance can be assessed against the forecast available at the time, the constraint state, the policy in force, the action selected and the outcome realised, creating a usable basis for improving forecasts, trading logic, operational policy and asset strategy.

Components used

  • Monitoring

    Tracks forecast performance, dispatch outcomes, revenue impact, override rates and constraint-related losses.

  • Evidence Vault

    Connects forecasts, constraint state, policy context, recommendation, human action and final outcome.

  • Predictive Models

    Compare expected and realised value across trading intervals.

  • Noise Resilience

    Distinguishes normal market volatility from material data, model or decision degradation.

  • Simulation

    Tests live decisions against baseline strategies and alternative decision paths.

Apply local market rules without rebuilding the decision system

Energy markets differ by geography, asset type and market participation model. Settlement structures, ancillary services, grid restrictions, gate closure times and operational obligations all shape the actions available.

What Decision Deck does

Decision Deck keeps the core decision method consistent while allowing rules, policies and permitted actions to vary by market, country, asset portfolio or owner mandate. Trading logic stays connected to the controls that govern it.

Components used

  • Policy Dynamics

    Versions market rules, country rules, asset rules, owner instructions and trading risk appetite.

  • Recommendations

    Selects actions from the permitted options for the relevant market and asset.

  • Exception Handling

    Catches trades or dispatch actions that fall outside available authority, evidence or market conditions.

  • Evidence Vault

    Records the policy version and market rules that applied at the time.

  • Monitoring

    Tracks performance and policy impact by market, asset and trading strategy.

Treat missing telemetry as a decision condition

Energy decisions rely on live operational evidence. Telemetry can arrive late, conflict with another signal or fail to reflect the current state of the asset. Market conditions can also move before the next decision interval.

What Decision Deck does

Decision Deck evaluates whether the evidence is sufficient for the action under consideration. It can continue under approved fallback rules, hold position, request confirmation or route the case for review, making data quality an explicit part of trading control.

Components used

Give owners and investors a clearer account of performance

Asset performance is more than a revenue figure. Owners need to understand what created value, what constrained it, what was sacrificed to protect asset life and where future performance can improve.

What Decision Deck does

Decision Deck creates interval-level decision evidence across bids, dispatches, constraints, policy conditions, human interventions and realised outcomes, supporting stronger operational reporting, asset management and investor dialogue.

Components used

  • Evidence Vault

    Records each bid, dispatch action, forecast, constraint, policy condition, override and outcome.

  • Monitoring

    Tracks revenue by market, strategy, asset, constraint type and decision path.

  • Simulation

    Compares actual decisions with baseline strategies and viable alternative actions.

  • Predictive Models

    Estimate expected value, lost value, degradation cost and opportunity cost.

  • Recommendations

    Identifies where strategy, policy or constraint settings may warrant review.

Test policy before it controls the asset

Small changes to risk appetite, degradation assumptions, cycling limits, gate closure rules or market participation policy can materially change revenue and asset exposure.

What Decision Deck does

Decision Deck supports controlled re-evaluation before policy changes reach live trading workflows. Historical intervals, current asset states and alternative policy versions can be assessed before deployment, and each live decision remains linked to the model and policy versions active at the time.

Components used

  • Simulation

    Tests policy, strategy, risk appetite and constraint changes against historical intervals and current asset states.

  • Policy Dynamics

    Versions trading rules, asset constraints, risk thresholds and approval requirements.

  • Predictive Models

    Estimate the likely effect on revenue, degradation, volatility exposure, constraint losses and approval volume.

  • Monitoring

    Shows where live outcomes diverge from expected performance after a change.

  • Evidence Vault

    Preserves the policy and model lineage behind each trading decision.

Better trading decisions, with the evidence to stand behind them

Decision Deck gives renewable asset owners, battery operators and trading teams a governed decision layer for market-facing asset decisions.

Trading actions reflect market opportunity, asset economics, operating constraints and the level of human authority required. Each decision carries the evidence needed to understand what happened and improve the next one.