SEPA report demonstrates benefits of AI for e-mobility
Image courtesy 123rf
The use of AI for the EV sector can offer up multiple benefits, key among them assistance with load growth prediction and grid assets planning, a new study by SEPA has found.
The report, Insight Brief: AI for Transportation Electrification, released by the Smart Electric Power Alliance (SEPA) and Bidgely, outlines these benefits of AI for the EV sector in the US, namely its impact on the power grid and how to better facilitate management of EV demand.
Tapping case studies from the likes of utilities Hydro One and NV Energy, the report outlines the key advantages utilities gain by disaggregating advanced meter infrastructure (AMI) data to gain insights into EVs and EV users in their service territories.
Have you read:
How to stimulate Europe’s bidirectional charging market
Australia gets a bidirectional charging roadmap
Key among these are:
• Easier to find and engage EV-driving customers
According to the study, utility staff can use AI analytics to find more EV-driving customers than self-report or basic analysis and can more efficiently target those with high potential to help meet load flexibility goals. Outcomes for utilities include more efficient customer analytics, customer outreach, programme implementation, and more effective programmes.
• Access to a novel source of EV charging data
This includes differentiation of Level 1 and Level 2 chargers; hourly charging patterns (time of day, duration, and intensity); and amplitude of charges.
The study adds that utilities with AMI can use AI analytics to identify EV charging sessions from hourly or 15-minute meter data. For some utilities, this will open a new door to EV detection analytics (e.g., those without a managed charging programme or third-party access to EV telematics or EVSE data).
• Higher-quality EV charging characteristics
Additionally, by using more granular data, utilities can strengthen their understanding of (and better account for) local variability in how, when, and how much customers charge their EVs. As EV adoption rises and driving and charging patterns diversify, this granularity becomes crucial to pinpoint EV-grid integration challenges and devise programmes, services, and grid upgrades that reflect this diversity and dynamism.
• Planning the grid
Through the use of AI, EV load growth can be forecast on individual grid assets, enabling utilities to better manage future infrastructure planning.
Specifically, deep learning analysis of AMI data can help utilities set goals for load shift initiatives, improve the management of such initiatives, determine EV charging peak coincidence, forecast EV load growth by grid segment and integrate learnings in future grid planning scenarios.
Hydro One and NV Energy
In support of their findings, SEPA and Bidgely’s study details the application of AI software by two North American utilities, Canadian transmission and distribution utility Hydro One and Nevada generation, transmission and distribution utility NV Energy.
Hydro One identified 20,000 EVs charging on its grid via AMI data disaggregation—10 times more than were self-reported through customer surveys. The utility further refined its customer engagement strategy using AI-powered consumption insights to personalise messages for enrollment in a pilot EV demand response programme, resulting in 300 signups within 24 hours.
NV Energy leveraged AI-powered data disaggregation to gain a holistic understanding of how often EV drivers charge on-peak and how their behaviour contributes to overall electricity demand. By using AI to identify certain customer use profiles and then engage only customers with high-value baseline charging behaviour, NV Energy achieved a load-shift potential of 2 – 4 kilowatts (kW)/vehicle per managed charging event as opposed to typical load shifts of 0.2 – 0.8 kW/vehicle per event— 2.5 times to 10 times greater load-shift on average.
Targeted load shifting initiatives like these, says the study, can enhance utilities’ system resilience capabilities as EV charging increases, while yielding cost efficiencies for utilities and customers.
