Financial Incentives to Shift Electric Vehicle Charging: Time-of-Use versus Managed

Phase 2: The treatments will include: i) Time-of-use (TOU), monetary incentives (payments) based on kWh charged during "off-peak" hours and ii) Managed, monetary incentives (payments) based on kWh charged when the vehicle is available to have its charging managed centrally.

Phase 3: The treatment will be a one-time randomized incentive for participants to opt-in to an incentive program for shifting vehicle charging for a defined additional period (e.g. 6 months).

Phase 2: treatments are at the virtual transformer group-level

Phase 3: treatments will be at the individual level, independent of virtual transformer group

2023-07-05

2024-06-30

Phase 2: Distribution group hourly charge, distribution group daily peak/off peak/total charge, household and vehicle hourly charge, household and vehicle daily peak/off peak/total charge, distribution group violation indicator (hourly and daily) and violation magnitude by hour (in kWh), charge indicators for whether cars are plugged in (hourly and daily), distribution group transformer charging room (hourly and daily), compliance with treatment

Phase 3: Same as Phase 2 outcomes plus household-level indicator for whether it opts into the offered incentive program for shifting charging

We will also collect data on costs of distribution system upgrades and distribution network outages from the utility to quantify the benefits of shifting EV charging times.

Phase 1: Our utility partner will recruit willing customers to participate in the trial. This will be done through a media campaign, where willing customers can elect to participate in the trial. Participants will enroll in a 3rd party software that can monitor and managing EV charging.
Phase 2: Our utility partner and UCalgary/UAlberta researchers will randomize participants into treatment and control by distribution group. The treatments will include: i) Time-of-use (TOU), monetary incentives (payments) based on kWh charged during "off-peak" hours and ii) Managed, monetary incentives (payments) based on kWh charged when the vehicle is available to have its charging managed centrally.
Phase 3: All participants will be asked if they wish to opt-in to an incentive program for shifting charging for defined additional period (e.g. 6 months or a year) in exchange for a one-time financial reward. The utility partner and UCalgary/UAlberta will randomize financial incentives offered to households independent of Phase 2 treatment status.
All participants will receive a monetary payment ($150) for participating in the study, regardless of treatment or control group status. This payment will be 50% upfront, and 50% at the end.

Details on Randomization Process for Phase 2

Phase 2: 1) Fortis and UCalgary/UAlberta researchers will place participants into treatment and control groups, using household-level randomization. The treatments will include: i) monetary incentives (payments) based on kWh charged during "off-peak" hours and ii) monetary incentives (payments) based on kWh charged when the vehicle makes itself available to have its charging managed centrally.
Virtual Transformer Randomization: To mimic the types of transformer constraints faced by utilities in the future with increased EV vehicle penetration, we assign households to virtual transformers in groups of approximately 10 home-charging cars. The randomization follows a sequential process. First, non-controllable EVs are allocated to their own virtual transformer (plus any controllable EVs that are part of a non-controllable household). These non-controllable EV’s are not eligible to be managed and households with them are separated from the remaining EVs. We will do analyses with and without these homes as part of the control group. Second, eligible EVs from multi-EV homes are randomly allocated to one of the three treatment groups. Then, within each treatment group, these EVs are randomly allocated to a transformer at the household-level. This ensures that cars from multi-EV homes are in the same transformer. Third, the controllable single EV homes are randomly allocated to the remaining group-transformer slots across the treatment and control groups.
We randomize maximum charging capacity on a given virtual distribution transformer at the transformer-day level. For the managed charging group, the constraint is used by the 3rd party software company as a maximum to inform managed charging throughout the day, subject to EVs still being able to obtain needed charging. For all groups, the constraint will give us our violation indicator and violation magnitude outcomes. Constraints were crafted using pre-intervention charging data, power calculations from simulated outcomes, and data on non-EV daily demand patterns.
Numbers at the start of the trial:
o 216 Eligible EVs randomized
o 13 non-controllable EVs (one non-controllable household has 2 EVs, one of which is controllable)
o 46 EVs are part of multi-EV homes (one of these is part of a home with a non-controllable EV)
 Two 3 EV homes
 Twenty 2 EV homes
o 158 EVs are single eligible EV homes and are controllable.

Office computer

Phase 2: distribution group level randomization
Phase 3: household level randomization

Yes

20 (start of the study) to up to 28 (if additional vehicles enroll) distribution groups

The number of observations will vary by vehicle according to when they downloaded the app and enrolled in the trial. We anticipate 216-275 vehicles. Daily observations could be up to 714 per vehicle (17,136 hourly observations). These vehicles make up the 20-28 distribution groups

Managed: 7 distribution groups (70 cars)
Time-of-Use: 7 distribution groups (70 cars)
Control: 8 distribution groups (76 cars)