Dynamic Load Balancing

The system continuously measures the aggregate current draw at the main feeder or sub-panel level using current transformers (CTs) or smart meter data. When total demand approaches the maximum rated capacity of the circuit breaker or transformer, the algorithm intervenes before a trip occurs.

• Sampling rates typically range from 1 to 10 seconds
• Monitors per-phase loading to detect phase imbalance
• Compares real-time consumption against a configurable static or dynamic limit

Rather than simply turning chargers on or off, the algorithm distributes the available headroom among all active charging sessions. Each connected vehicle receives a proportional share of the remaining capacity, ensuring fair access.

• Uses max-min fairness algorithms to prevent starvation
• Allocations recalculate instantly when a vehicle plugs in or departs
• Prevents the first-come, first-served problem where early arrivals monopolize capacity

Not all charging sessions are equal. The system assigns priority tiers to vehicles based on business rules such as departure time, battery State of Charge (SoC), or fleet role. Emergency vehicles or those with imminent departure deadlines receive preferential current allocation.

• Integrates with Fleet Energy Management Systems (FEMS) for schedule awareness
• Supports static priority (VIP users) and dynamic priority (low SoC first)
• Ensures operational readiness for mission-critical fleet vehicles

In three-phase installations, single-phase EV chargers can create dangerous phase imbalances that cause neutral conductor overheating and voltage instability. The algorithm actively shifts charging loads across phases to maintain symmetry.

• Monitors L1, L2, and L3 currents independently
• Redistributes single-phase charger assignments to minimize neutral current
• Complies with utility requirements for voltage unbalance below 2%

The system can interface with external signals from OpenADR or utility Demand Response programs. When a demand limit event is triggered, the algorithm temporarily reduces aggregate charging load to avoid peak demand charges or support grid stability.

• Responds to price signals and direct load control commands
• Implements configurable load shed strategies (proportional vs. sequential)
• Critical for commercial sites facing demand charge management requirements

If communication with the central management system is lost, the local controller defaults to a conservative safe mode. It enforces a pre-configured static limit well below the circuit breaker rating to guarantee physical safety regardless of software state.

• Implements watchdog timers to detect communication loss
• Defaults to a minimum guaranteed current per charger to prevent complete shutdown
• Stores the last known valid schedule in non-volatile memory

A unidirectional control strategy where the charging rate of an EV is dynamically modulated by an external signal—typically from a utility or charge point operator—to optimize grid load. Unlike dynamic load balancing, which distributes capacity across multiple charge points, V1G focuses on time-shifting demand at individual stations.

• Adjusts charging power between 0% and 100% of rated capacity
• Responds to OpenADR or OCPP signals
• Reduces peak demand without requiring bidirectional hardware
• Foundational precursor to full dynamic load balancing deployments

The active monitoring and algorithmic control of distributed energy resources to prevent thermal overload and accelerated insulation aging in distribution transformers. Dynamic load balancing serves as the primary execution mechanism for transformer load management strategies.

• Monitors winding temperature and top-oil temperature in real time
• Calculates loss-of-life metrics based on IEEE C57.91 standards
• Enforces dynamic ampacity limits across EV charging clusters
• Prevents coincident peak loading from fleet charging events

An advanced process control algorithm that solves a finite-horizon optimization problem at each time step to determine optimal charging schedules. MPC is the mathematical backbone of many dynamic load balancing implementations, incorporating forecasts of energy prices, renewable generation, and expected EV arrivals.

• Uses a receding horizon approach—recalculates every 5-15 minutes
• Incorporates constraints like transformer capacity and vehicle departure times
• Outperforms rule-based load balancing in variable tariff environments
• Requires accurate charging load forecasting inputs

An open-source communication standard governing interoperability between EV charging stations and a Central System Management System (CSMS). Dynamic load balancing commands—such as per-connector power limits—are transmitted via OCPP's SmartCharging profile.

• OCPP 2.0.1 supports dynamic setpoint adjustments per connector
• Enables vendor-agnostic load management across mixed hardware fleets
• Communicates real-time State of Charge (SoC) data for optimization
• Critical for avoiding proprietary vendor lock-in at multi-station sites

A load management strategy that reduces grid power consumption during periods of highest electricity demand by utilizing stored energy or curtailing flexible loads. Dynamic load balancing implements peak shaving by capping aggregate site power below a defined threshold.

• Targets demand charge reduction for commercial fleet operators
• Typical thresholds set at 70-80% of site maximum capacity
• Integrates with stationary battery energy storage systems
• Can respond to day-ahead or real-time pricing signals

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 solvers underpin many dynamic load balancing engines that must respect binary constraints.

• Handles combinatorial complexity of multi-vehicle scheduling
• Respects minimum charging power thresholds of EVSE hardware
• Often combined with heuristics for real-time performance
• Commercial solvers like Gurobi and CPLEX commonly deployed