Will adoption cluster?
Income, education, dwelling type, commuting, garage access and housing patterns cause EV ownership to concentrate geographically.
EV Load Forecasting for Utilities
Forecast where, when and how much EV charging load will occur.
Systemwide EV counts are not enough for distribution planning. Utilities need localized adoption forecasts and hourly charging profiles that reveal clustering, peak coincidence and potential grid exposure.
Peak-Day EV Charging LoadHOURLY VIEW
Unmanaged peak-period loadManaged/off-peak load
Time resolution8,760 hours
Local detailBlock group
Scenario typesManaged + unmanaged
The essential forecast
Distribution planning requires three answers—not one adoption number
The same number of EVs can produce very different utility impacts depending on location, charging time and peak coincidence.
Will charging occur?
Arrival times, charging power, energy requirements, season and customer behavior determine the hours that create system and local peaks.
Load will be added?
Customer-level charging profiles and diversity determine coincident kW at the transformer, feeder, block-group and service-area levels.
Why forecast direction matters
Top-down forecasts can hide the local problem
A system forecast may show modest average EV growth while individual neighborhoods experience much higher adoption and peak-hour loading.
Forecast the total, then allocate it downward
Useful for energy and broad resource planning, but potentially weak for identifying concentrated distribution impacts.
- Starts with service-area or regional EV totals
- Often uses average charging profiles
- Spreads adoption across large geographic areas
- Can dilute neighborhood clustering and peak coincidence
- Requires separate assumptions for local engineering studies
Forecast households, loads and locations first
Builds localized results from representative customers and then aggregates upward to the planning area required.
- Estimates household EV ownership probabilities
- Applies customer-level hourly charging profiles
- Allows clustering to emerge from household characteristics
- Produces block-group, ZIP and service-area results
- Exports scenario profiles for engineering analysis
Customer digital-twin forecasting
From household adoption probability to localized hourly load
GIM combines identity-protected customer records with housing, demographic, transportation, weather and metered end-use information.
Assemble representative customers
Match the utility service area with identity-protected MAISY household records.
Estimate EV adoption probability
Apply income, education, dwelling, commuting and other household characteristics.
Assign hourly charging loads
Reflect travel, arrival, charger power, energy needs and charging behavior.
Run utility scenarios
Test current, 2030 and 2035 adoption, weather and managed charging assumptions.
Aggregate and export
Summarize by customer, block group, ZIP, feeder or service area.
More than a single peak estimate
Hourly profiles reveal duration, seasonal variation, coincident peaks and the hours available for load shifting.
UNMANAGED PEAK
8,760 hourly charging forecasts
See the timing and duration of EV grid impacts
A peak value alone cannot show whether a constraint occurs for one hour, several consecutive hours or under a specific weather and charging combination.
Weekday and weekend patterns
Seasonal charging differences
Peak-hour coincidence
Charger-power alternatives
Managed charging schedules
Customer participation scenarios
Recovery and load-shift effects
Engineering-ready exports
Use the geographic level the decision requires
Forecast locally—then aggregate consistently
System planners, distribution engineers and program managers can work from the same underlying customer and hourly-load assumptions.
For utilities with circuit models, GIM outputs can be mapped to feeders, transformers or GIS service areas. Without detailed circuit models, block-group results provide a practical first screen for prioritization.
Service areaSystem energy, peak and total program economics
ZIP codeBroad market, adoption and load differences
Block groupNeighborhood clustering and priority-area identification
Feeder / transformerMapped forecasts and engineering-study inputs
Representative customerHousehold characteristics and complete hourly profiles
Planning applications
Turn EV forecasts into utility decisions
Localized adoption and hourly-load forecasts create the common foundation for distribution planning, program design and financial analysis.
Prioritize transformer and feeder review
Identify neighborhoods and block groups where EV growth is most likely to create planning pressure.
Screen transformer exposure
Apply user-defined size, loading, planning limits and households-per-transformer assumptions to selected block groups.
Support capital timing
Compare future-year unmanaged and managed charging loads before committing to detailed studies or upgrades.
Design managed charging
Test participation, engagement, charging shifts and technology alternatives against system and local peak objectives.
Quantify the business case
Connect G&T demand savings, EV revenue margins, incentives and program costs to payback, NPV and benefit/cost metrics.
Evaluate flexible-load alternatives
Compare managed charging with DSM, DER, battery and VPP strategies under consistent hourly assumptions.
EV forecasting is the starting point—not the final decision
The expanded Grid Impact Model connects these forecasts directly to block-group transformer screening and service-area managed charging economics.
Practical implementation
No AMI data or customer contact required
GIM uses curated, identity-protected MAISY customer records and supporting datasets to estimate representative hourly loads and future scenarios.
Utility AMI, GIS and equipment information can strengthen or refine applications where available, but a co-op does not need to complete a major data-integration project before beginning planning-level analysis.
GIM forecasting and transformer worksheets support planning-level screening and prioritization. Detailed asset decisions should incorporate utility measurements, equipment records and engineering analysis.
7+ million identity-protected customer recordsResidential energy, equipment, building, income, demographic and location characteristics.
Detailed end-use meteringWhole-building and end-use hourly and 15-minute information supporting customer load profiles.
Transportation and commuting informationInputs supporting travel, availability and charging requirements.
Weather and housing characteristicsLocal climate, dwelling stock, growth, demolition and rebuilding assumptions.
Visible utility-controlled scenariosAdoption, charging, participation, engineering and program assumptions remain adjustable.
Frequently asked questions
EV load forecasting questions
What is EV load forecasting for utilities?
It estimates where, when and how much electric-vehicle charging load will occur. For distribution planning, the forecast should address localized adoption, hourly charging behavior and coincident load at the geographic or grid-asset level relevant to the decision.
Why is customer-level forecasting useful?
Household income, dwelling type, commuting behavior, garage access and other characteristics affect both EV adoption and charging. Modeling representative households allows geographic clustering and load diversity to emerge rather than assuming uniform adoption.
Does GIM require AMI data?
No. Curated identity-protected customer data and supporting datasets provide the initial modeling foundation. AMI data can be incorporated or used for calibration and validation when it is available and suitable.
Can the forecast support managed charging?
Yes. GIM compares unmanaged and managed hourly charging, including participation and engagement assumptions, and connects the results with avoided G&T demand charges, program costs and utility financial metrics.
Can forecasts be used with engineering models?
Yes. Localized hourly profiles can be mapped or exported for CYME, Synergi, WindMil, OpenDSS, GIS and other planning workflows. GIM can also provide block-group screening when detailed circuit models are not yet available.
See localized EV forecasting in action
Review a representative customer, block-group and service-area analysis in a guided online demonstration.