Demand Charge Management

The foundational layer of any demand charge management system is sub-second power metering. Unlike net energy metering, these systems sample kilowatt (kW) draw at intervals of 1 to 15 minutes to align with the utility's demand interval window.

• Rolling Window Logic: The system continuously calculates the maximum average power over the utility's specific interval (e.g., 15-minute sliding window).
• Predictive Thresholding: If the current trajectory predicts a new peak within the interval, the controller triggers a corrective action before the interval closes.

When a demand threshold is approached, the system executes automated load shedding without human intervention. This is not a simple on/off switch but a granular power reduction strategy.

• Proportional Curtailment: Instead of stopping a session, the system reduces the C-Rate across multiple vehicles simultaneously, lowering aggregate kW while maintaining charging continuity.
• Priority-Based Queuing: Algorithms factor in State of Charge (SoC) and Depth of Discharge (DoD) to prioritize vehicles with urgent departure schedules, curtailing only those with flexible dwell times.

To avoid curtailment entirely, advanced systems integrate a Battery Energy Storage System (BESS) as a buffer. The control logic treats the grid connection as a fixed-rate pipe and the battery as a surge tank.

• Discharge Triggers: When EV load exceeds the predefined demand cap, the BESS discharges to cover the deficit, ensuring the grid meter never registers the spike.
• Recharge Strategy: The BESS recharges during low-load periods or when solar generation is abundant, preparing for the next peak event without creating a secondary demand charge.

A statistical approach to managing the probability that multiple chargers will draw maximum power simultaneously. The system actively enforces a maximum coincidence factor to flatten the aggregate load profile.

• Staggered Sequencing: Instead of simultaneous startup, chargers are sequenced with micro-delays to avoid inrush current spikes that artificially inflate demand readings.
• Group Power Limits: A static or dynamic cap is placed on a group of chargers, forcing the local controller to solve a Mixed-Integer Linear Programming (MILP) problem to allocate power without exceeding the group limit.

Demand charge structures are complex, often involving time-of-use (TOU) multipliers, ratchet clauses, and non-coincident demand peaks. The management system must parse the digital tariff.

• Ratchet Avoidance: The system models the long-term financial impact of setting a new peak, as many utilities enforce a ratchet clause that locks a percentage of that peak as the minimum billable demand for the next 11 months.
• TOU-Aware Shaving: The demand limit is dynamically adjusted based on the time of day, allowing higher peaks during off-peak periods when demand charges are lower or waived.

Rather than reactive threshold braking, Model Predictive Control uses a forward-looking optimization model to plan charging schedules hours in advance.

• Input Variables: The model ingests charging load forecasts, vehicle departure schedules, and weather data to predict future demand peaks.
• Optimization Horizon: The controller solves a finite-horizon optimization problem to minimize the sum of energy and demand costs, outputting a dynamic power limit that preemptively avoids peaks rather than reacting to them.

A load management strategy that reduces grid power consumption during periods of highest electricity demand by utilizing stored energy from batteries or curtailing flexible loads. In the context of EV fleets, peak shaving directly targets the highest 15-minute interval of power draw to slash demand charges.

• How it works: A Battery Energy Storage System (BESS) discharges to cover load above a set threshold.
• Key metric: Reduces the maximum kilowatt (kW) draw recorded by the utility meter.
• Example: A depot with a 150 kW unmanaged peak can shave 50 kW using a stationary battery, keeping the billed demand at 100 kW.

A centralized software platform that monitors, schedules, and optimizes the charging of multiple electric vehicles simultaneously while respecting operational route schedules and local grid constraints. A FEMS is the algorithmic brain that executes demand charge management.

• Core functions: Aggregates State of Charge (SoC) data, forecasts departure times, and dispatches charging setpoints.
• Optimization engine: Often uses Model Predictive Control (MPC) to solve for the lowest-cost charging schedule.
• Constraint handling: Ensures that the aggregate site load never exceeds a pre-defined power cap to prevent demand charge spikes.

An advanced process control algorithm that solves a finite-horizon optimization problem at each time step to determine optimal charging schedules based on forecasted energy prices and load. MPC is the gold-standard mathematical framework for demand charge management.

• Mechanism: The controller predicts future system states using a model of the battery and charger dynamics.
• Receding horizon: Only the first step of the computed control sequence is executed before the optimization is re-solved with new data.
• Advantage over simple rules: Proactively pre-cools or pre-charges assets to avoid predictable peaks rather than reacting after the peak has already occurred.

A mathematical optimization technique used to solve discrete charging scheduling problems where variables like on/off status are integers and power flow is continuous. MILP is often the computational engine inside a FEMS for exact demand charge minimization.

• Binary variables: Represent whether a charger is active (1) or idle (0) at a specific time interval.
• Objective function: Typically minimizes the sum of energy costs plus the peak demand charge component.
• Solver requirement: Requires powerful commercial solvers like Gurobi or CPLEX to find globally optimal solutions for large fleets within operational time limits.

The active monitoring and algorithmic control of distributed energy resources to prevent thermal overload and accelerated aging of distribution transformers caused by coincident electric vehicle charging. This is the physical infrastructure protection layer beneath demand charge management.

• Thermal dynamics: Transformers have a thermal time constant; sustained high loads degrade cellulose insulation, halving life for every 6-8°C rise.
• Coincidence factor: The probability that multiple EVs charge simultaneously; unmanaged fleets have a factor near 1.0, guaranteeing peak stress.
• Control strategy: The FEMS enforces a dynamic apparent power (kVA) limit to keep transformer hotspot temperature below rated limits, preventing unplanned outages.

A stationary electrochemical storage asset that acts as a buffer between the grid and EV chargers. A BESS is the most effective hardware tool for demand charge management, decoupling the timing of grid power draw from vehicle charging.

• Sizing: Typically sized to cover the difference between the maximum fleet load and the desired grid import limit for the duration of the peak window.
• C-Rate: The battery must support high discharge rates (e.g., 1C or 2C) to effectively shave sharp, short-duration peaks.
• Degradation trade-off: Each discharge cycle for peak shaving imposes a marginal Depth of Discharge (DoD) cost that must be factored into the total cost of ownership calculation against the demand charge savings.