Time-of-Use (TOU) Rate Arbitrage

The core operational loop of TOU arbitrage consists of two distinct phases executed daily. During off-peak periods (typically overnight or midday when solar generation depresses prices), the battery management system draws power from the grid to charge storage assets at low cost. During on-peak windows (early evening when demand peaks and solar ramps down), stored energy is discharged back to the grid or behind-the-meter loads. The round-trip efficiency—typically 85-95% for lithium-ion systems—determines the net energy delivered per unit charged. A 100 MWh charge at $20/MWh yields approximately 90 MWh of dischargeable energy, requiring a price spread exceeding $22.22/MWh just to break even on energy losses alone.

The economic viability of TOU arbitrage depends entirely on the peak-to-off-peak price differential. Key components include:

• Wholesale energy arbitrage: Capturing the spread between locational marginal prices at different hours
• Retail rate arbitrage: Exploiting differences in utility TOU tariff structures behind the meter
• Demand charge avoidance: Discharging during a facility's peak demand intervals to reduce monthly demand charges, which can constitute 30-70% of a commercial electric bill

The spark spread equivalent for storage—the minimum price differential required for profitability—must account for degradation costs, round-trip losses, and capital amortization. Typical breakeven spreads range from $15-40/MWh depending on battery chemistry and cycle life.

Every charge-discharge cycle imposes a marginal cost of degradation on the battery asset. Lithium-ion cells experience capacity fade through multiple mechanisms:

• Calendar aging: Time-dependent electrolyte decomposition independent of cycling
• Cycle aging: Loss of active lithium inventory proportional to depth of discharge and C-rate
• State-of-charge stress: Accelerated degradation when stored at extreme SOC levels (>90% or <10%)

Advanced dispatch algorithms integrate electrochemical degradation models into the optimization objective function. Rather than simply maximizing revenue, they solve for the dispatch schedule that maximizes net revenue after degradation cost. This often means forgoing small arbitrage opportunities where the marginal degradation cost exceeds the price spread.

TOU arbitrage is fundamentally a stochastic optimization problem dependent on accurate forecasts:

• Price forecasting: Day-ahead and real-time LMP predictions using neural networks trained on historical nodal prices, weather, and generation stack data
• Load forecasting: Behind-the-meter demand predictions to determine residual capacity available for arbitrage
• Solar generation forecasting: Critical for commercial buildings with rooftop PV, as excess solar may charge the battery for free, altering the arbitrage calculus

Model Predictive Control (MPC) frameworks re-optimize the dispatch schedule at each timestep as new information arrives. This closed-loop approach corrects for forecast errors and captures intraday price volatility that static day-ahead schedules miss. Typical re-optimization intervals range from 5 to 15 minutes.

Pure energy arbitrage rarely justifies storage investment alone. Modern systems stack multiple revenue streams within a single dispatch framework:

• Frequency regulation: Reserving a portion of capacity for fast-responding regulation signals (e.g., PJM RegD) while arbitraging with the remainder
• Spinning reserve: Holding capacity in reserve for contingency events, earning capacity payments while still capturing some energy arbitrage
• Resource adequacy: Bidding into capacity markets to guarantee availability during system peak hours
• Voltage support: Providing reactive power services at the distribution level

The optimization engine must co-optimize across these stacked services, respecting the physical constraints of the battery while maximizing total portfolio revenue. Conflicts arise when, for example, a regulation signal requires charging during what would otherwise be a high-price discharge window.

TOU arbitrage operates within specific market constructs that vary by jurisdiction:

• ISO/RTO wholesale markets: FERC Order 841 mandates that storage resources can participate as both generation and load, setting the market-based price floor for arbitrage
• Utility TOU tariffs: Regulated retail rates with defined on-peak and off-peak periods, such as California's TOU-D-4-9PM rate with a 4-9 PM peak window
• NEM 3.0 solar-battery pairing: California's net billing tariff creates strong incentives for pairing batteries with solar to export during high-value evening hours rather than midday
• Local flexibility markets: Emerging constructs where distribution utilities procure localized flexibility services, creating locational value differentials beyond wholesale prices

Market participation requires telemetry and settlement infrastructure compliant with ISO metering standards, including revenue-quality metering and real-time telemetry to the market operator.

A control logic that dispatches battery energy storage to discharge during periods of highest site load, thereby reducing the maximum demand charge levied by the utility. Unlike pure TOU arbitrage which targets energy cost differentials, peak shaving specifically targets the kW demand component of a commercial bill. The algorithm must forecast load and decide when to reserve capacity for demand reduction versus energy arbitrage.

The calculated cost of delivering an additional unit of energy to a specific node on the grid. TOU rates are a blunt temporal signal, while LMP provides granular, locational price differentiation that changes every 5-15 minutes. Advanced arbitrage strategies layer LMP forecasting on top of static TOU schedules to capture congestion-driven price spikes at specific substations.

A statistical methodology that estimates what a customer's energy consumption would have been without a demand response event. For TOU arbitrage, accurate CBL is critical when the battery participates in utility programs—if the baseline is miscalculated, the arbitrage revenue from perceived load reduction may be overstated. Common methods include the 10-day average and regression-based baselines.

An advanced optimization strategy that uses a dynamic model of the microgrid to forecast future states and determine the optimal dispatch schedule over a receding time horizon. For TOU arbitrage, MPC outperforms simple rule-based charging by anticipating tomorrow's solar generation, load, and price peaks, ensuring the battery is never caught empty during a high-price window.

A mathematical optimization technique used to solve the unit commitment and economic dispatch problem for DER fleets. MILP handles the discrete on/off charging decisions and continuous power output levels simultaneously. For TOU arbitrage, the solver maximizes net revenue subject to constraints like battery state of charge limits, cycle aging costs, and round-trip efficiency.

A billing mechanism that allows a single customer with multiple meters on contiguous property to offset total load with total generation. NEM aggregation fundamentally changes the TOU arbitrage calculus—exporting stored solar energy during peak periods may earn retail rate credits, dramatically improving the economics compared to wholesale net metering or standalone battery arbitrage.