EV Load Detection and Scenario Modeling Using AMI (Smart Meter) Data

Client: Veitur (Utility & Distribution System Operator)

Timeline: 2023-2024
Region: Reykjavik, Iceland

Publication: Smart Meter Insights: Identifying and Projecting Electric Vehicle Loads in Reykjavik

Trends in new vehicle registrations

EV adoption was steadily increasing in Iceland in 2023.

Normalized residential power draw

Data shows EV demand peaks are higher and at the same time of day, placing high stress on the system.

Max power vs annual energy consumption

Customers with EVs have higher power requirements but also generate more energy-based revenue from charging their vehicles.

Ratio of consumer peak power to feeder transformer ratings under high EV adoption scenario

Red indicates overloading.

Extra transformers needed to service consumer demand under high EV adoption scenario

Yellow indicates only one extra transformer is needed. Red indicates more than one is needed.

Project Summary

In response to Iceland’s rapid shift toward electric mobility, Veitur—the DSO for the Greater Reykjavik Area—sought to assess future grid impacts of widespread Electric Vehicle (EV) adoption. With data collected from over 24,000 customer smart meters and EV adoption steadily increasing, the challenge was clear: What will "business-as-usual" demand look like if everyone drives an EV—and where will the city’s grid infrastructure feel it first?

We analyzed time-series data from Veitur’s smart meters to:

• Detect EV charging behavior in unlabeled data,

• Estimate EV energy demand per connection point, and

• Model transformer-level peak impacts under a high transport electrification scenario.

Client Motivation

1. EVs already constituted 15%+ of the national vehicle fleet, with adoption increasing.

2. New Advanced Metering Infrastructure (AMI) provided a wealth of new data via smart meters, yet to be used.

3. No existing labels were available to identify EVs serviced behind the meter.

4. Planning upgrades and prioritizing investment requires knowing when and where grid stress will emerge under continued EV adoption.

Our Methods

1. EV Load Identification:
   Developed a multi-stage algorithm to detect EV charging profiles using characteristic features, benchmarked against known AC-charging behavior for EVs.

2. EV Energy Consumption:
   Calculated annual EV energy consumption, estimating the number of EVs served at each point of connection.

3. Synthetic Load Modeling:
   Scaled EV ownership to target a 100% electrification scenario using zoning and population data.
   Created synthetic time series demand profiles for meters not yet upgraded with smart meters.

4. Substation Stress Mapping:
   Aggregated synthesized profiles to secondary substations.
   Compared peak apparent power to transformer ratings to flag potential overload areas.

Key Deliverables

1. EV detection algorithm utilizing unlabeled smart meter time series data

2. Forecasted transformer stress maps under high EV adoption scenario

3. Technical documentation and reproducible modeling workflow

4. Recommendations for monitoring, investment, and scenario planning

Results & Impact

1. Detected 3,000+ EV equivalents in current data—closely matching expectations based on vehicle registration data

2. Identified transformer overload risk areas under high EV adoption, enabling targeted investment planning

3. Developed a replicable framework for data synthesis and EV detection usable in other jurisdictions

Client Testimonial (TBD)

"This work gave us a critical forward-looking lens into how EV adoption will shape demand across our network. The ability to map future load scenarios using real meter data was a game changer."

— Distribution System Planning, Veitur ohf.

Anton & Company is proud to continue supporting Iceland’s third energy transition—delivering data-driven tools that help stakeholders make confident, forward-looking infrastructure decisions in the face of electrification.

---

Wondering how EV adoption will impact your grid?

Whether you're dealing with unlabeled AMI data or planning for large-scale electrification, we can help you uncover insights and prepare your grid for the future.

→ Get in touch today