HEVI-LOAD Software: MHDEV Charging Simulation

• HEVI-LOAD is a software platform developed by LBNL to simulate and analyze medium- and heavy-duty electric vehicle charging impacts on power grids.
• It employs M/M/c queuing models to generate operationally plausible, high-resolution load profiles from real-world charging data.
• The tool integrates with IEEE 33-bus system models to conduct AC power flow simulations, assessing voltage stability and infrastructure upgrade needs.

HEVI-LOAD is a software platform developed by Lawrence Berkeley National Laboratory (LBNL) for the simulation, projection, and analysis of charging infrastructure requirements and grid impacts associated with medium- and heavy-duty electric vehicles (MHDEVs). The tool provides high-resolution, real-world charging load data, enabling researchers and practitioners to investigate the implications of megawatt-scale electric vehicle charging on power distribution systems, with a particular emphasis on the operational challenges and infrastructure upgrades necessitated by widespread fleet electrification (Hassan et al., 23 Jul 2025).

1. Software Capabilities and Data Outputs

HEVI-LOAD (Medium- and Heavy-Duty Electric Vehicle Infrastructure – Load, Operations, and Deployment) is designed to project the deployment needs of charging infrastructure for MHDEVs and generate realistic load profiles for subsequent use in grid impact studies. The software supports:

• Data Collection: Provision of detailed charging sessions from multiple locations, with raw data encompassing Time (minutes), Vehicle ID, Charging Power (kW), and Charging Time (minutes).
• Operational Realism: Application of queuing theory (specifically M/M/c models) to model realistic usage constraints at high-power charging stations (HPCS), such as finite port availability and first-come, first-served service discipline.
• Granular Load Profiles: Output of time-resolved megawatt-scale power profiles that capture the stochastic and operationally constrained behavior of MHDEV charging ecosystems.

Component	Description	Example Use
Raw load data	Charging events with timestamps, IDs, and power	System modeling
Queued load model	M/M/c application for port limitations	Queue discipline
Location diversity	Data from several HPCS across sites	Geographical spread

The combination of empirical data collection and statistical queuing post-processing ensures that the resulting load profiles are both high-resolution and operationally plausible.

2. Queuing Model Application and Data Preprocessing

Initial raw outputs from HEVI-LOAD may exhibit unrealistic concurrency in vehicle charging—i.e., more vehicles charging simultaneously than the HPCS physically supports—due to overlapping vehicle IDs. To enhance operational realism:

• Queue Model: The M/M/c model is used, where arrivals follow a Poisson process (rate $\lambda$), service times are exponentially distributed, and $c$ represents the number of charging ports. This structure achieves:
  • Enforced hard limit on simultaneous charging sessions
  • A realistic temporal distribution of queued and active vehicles
  • Properly shaped aggregate load curves reflecting real-world facility throughput constraints

The adoption of queuing theory ensures that downstream grid studies use plausible input—critical when assessing secondary distribution system impacts.

3. Integration with Electric Distribution System Models

Processed HEVI-LOAD profiles serve as input to power system analysis platforms. The benchmark IEEE 33-bus distribution system is frequently employed for this purpose. Key aspects include:

• System Topology: The IEEE 33-bus system is a radial, 12.66 kV network with standard residential and commercial loads.
• Time-Synchronized Loads: Residential base loads are temporally interpolated to the HEVI-LOAD profile’s resolution, yielding a composite load scenario for each bus.
• Spatial Load Assignment: Charging profiles are injected at selected bus nodes, reflecting possible siting of HPCS for MHDEVs.

This enables a temporally accurate and spatially distributed representation of future electric vehicle charging within actual power networks.

4. Power Flow Simulation Methodology

Impact analysis is conducted using AC power flow simulations, often implemented in Panda Power, an open-source Python toolkit for modeling and analyzing electrical distribution networks. For each simulation time step:

• Active and Reactive Power Calculations:

  $$P_i = V_i \sum_{j=1}^{n} V_j \left( G_{ij} \cos \theta_{ij} + B_{ij} \sin \theta_{ij} \right)$$

  $$Q_i = V_i \sum_{j=1}^{n} V_j \left( G_{ij} \sin \theta_{ij} - B_{ij} \cos \theta_{ij} \right)$$

where $V_i$, $V_j$ are voltage magnitudes at buses $i$ and $j$, $G_{ij}$, $B_{ij}$ are elements of the network’s admittance matrix, and $\theta_{ij}$ are voltage angle differences.

• Observables: The simulation determines bus voltages, power flows, and network losses throughout the diurnal cycle at the resolution dictated by the HEVI-LOAD output.

Inclusion of accurate, temporally and spatially resolved MHDEV charging loads is critical for revealing nodal impacts on the grid.

5. Key Findings Regarding Distribution System Impacts

Simulation results demonstrate substantial impacts of MHDEV charging on network stability:

• Voltage Stability: Nodes proximal to the feeder (e.g., Bus 1, Bus 10) maintain voltages within the $0.95$ to $1.05$ p.u. band. In contrast, distal nodes (Bus 15, Bus 20, Bus 30) experience voltage depressions below $0.95$ p.u., especially at high aggregate charging events (e.g., $1.4$ MW contemporaneous load at specific times).
• Load-Driven Violations: Voltage drops are most acute during periods of coincident charging demand—typically the morning window, around 570 minutes (9:15 AM).
• Infrastructure Implications: The findings reveal the inadequacy of existing radial distribution infrastructure to absorb projected MHDEV charging loads without significant investment in line reinforcement, transformer upgrades, or advanced voltage support.

A plausible implication is that unmanaged proliferation of high-capacity EV charging will precipitate widespread voltage violations in legacy systems, necessitating preemptive grid planning.

6. Recommended Mitigation Strategies

To address the identified operational and stability challenges, several strategies emerge:

• Grid Reinforcement: Upgrading conductors, transformers, and voltage regulators in stressed sections.
• Smart Charge Management: Implementation of coordinated charging algorithms to spread demand, reducing temporal peaks.
• Integration of Distributed Energy Resources (DERs): Incorporation of local energy storage or renewable generation at or near charging sites to partially offset grid loading and provide voltage support.

These measures are suggested to be essential for accommodating rapid MHDEV electrification in accordance with policy directives.

7. Significance for Electrification Planning and Policy

HEVI-LOAD’s empirical, operation-informed outputs underpin quantitative planning for transportation electrification, supporting both technical grid studies and high-level infrastructure policy:

• Planning Support: Provides analytically justified load scenarios for utility engineers and regulators.
• Policy Foundations: Lays the groundwork for policies requiring grid upgrades and demand management as prerequisites for mass MHDEV adoption.
• Research Utility: Serves as a reference dataset for further academic investigation into charging impacts, technical mitigation, and system design.

This integration of data-driven modeling, statistical queuing, and power systems engineering ensures that HEVI-LOAD remains a foundational resource for the evolving landscape of electric transportation infrastructure studies (Hassan et al., 23 Jul 2025).