EV Managed Charging: Lessons from Utility Pilot Programs

As of May 2019, over 1.22 million EVs had been sold in the United States. Predictions now estimate 20 million EVs on U.S. roads by 2030, a huge load growth opportunity for utilities if integrated successfully, or if not, a significant challenge to operate and balance grid loads in the future.

To address potential issues and evaluate industry progress, SEPA recently published A Comprehensive Guide to Electric Vehicle Managed Charging, providing a detailed overview of managed charging capabilities, its potential benefits and barriers, the current state of the industry, and insight into the requirements for managed charging programs to become a viable and scalable demand response option in the future.

According to the report’s author and SEPA principal, transportation electrification, Erika Myers, the rapid adoption of EVs represents the most significant new electric load for the grid since the rise of air conditioning in the 1950s. As such, many stakeholders today contend that an electric vehicle future “must include some form of managed charging in order to reap the maximum benefits for consumers, the grid, and society as a whole.”

Overview of Managed Charging

SEPA’s report identified 38 utility-run managed charging pilot and demonstration projects across the U.S., and showed that interest is growing among utilities with 74% of survey respondents reporting interest or participation in utility-run managed charging programs.

Managed charging relies on communication signals from a utility or aggregator to be sent to a vehicle or charging device to control charging events. Managed charging programs fall into two categories: passive and active. Passive programs focus on altering customer charging behavior. One way to achieve this is using time-varying rates to incentivize customers to charge during less expensive off-peak hours. To dive deeper into the capabilities and efficacy of behavioral demand response programs, SEPA will be releasing an extensive report on time-based rates this fall.

Active managed charging programs provide utilities, or a market aggregator working with charging networks, with the capability to determine and/or control charging time, scale, and location, “in order to achieve a variety of outcomes, such as managing peaks, absorbing excess renewable generation or supplying some ancillary services to a structured market.” These programs rely on a reliable two-way flow of information that includes 1) a transport layer which relies on a communication signal via  Wi-Fi, cellular, vehicle telematics, etc. to send the charging instructions and 2) a messaging protocol or standard that can help the device understand and execute the instructions.

Barriers to Implementation & Adoption

As with other elements of demand response (DR) and grid modernization, making improvements in network communication and equipment interoperability is key to the success of managed charging. Stakeholders, including automotive original equipment manufacturers (OEMs), electric vehicle supply equipment (EVSE) manufacturers, and network service providers have yet to converge around a common managed charging open protocol or set of protocols for controlling EV charging.

If managed charging standards are not adequately advanced and deployed, the industry will struggle to meet the required load produced by millions of EVs on the road. As explained in the report, “this could lead to grid constraints and increased transmission and distribution peaks that prompt the construction of more peaker plants, unplanned grid upgrades, and other costly solutions.”

Lessons from Utility Pilot Programs

In June, SEPA hosted a webinar that featured speakers from three utility-run managed charging pilot programs at Southern California Edison (SCE), Avista, and Pacific Gas & Electric (PG&E). Program managers from each pilot shared the challenges they faced in implementing their programs with customers, and their successes in managing EV loads. Below are some of the key findings.

Southern California Edison

Mauro Dresti, senior advisor, transportation electrification technology strategy at SCE detailed their Charge Ready DR Pilot Program, which initially targeted commercial and multi-unit dwelling customers. Via this pilot, SCE demonstrated the capability to shift and reduce loads by up to 50%. SCE provided equipment rebates and covered the cost to install the equipment. In exchange, customers agree to participate in DR events, offsetting some or all of the installation and equipment costs.

Key Takeaways

• The program successfully reduced load by an average of 42% at participating sites during DR events in 2018.
• They used an automatic opt-in system, notifying and allowing customers to opt-out. This ensured the opt-in “had absolutely no thinking for the customer” to minimize inconvenience while increasing participation.
• Commercial charging was viewed as a better avenue for absorbing renewable overgeneration (to prevent curtailment of excess supply), as residential EV owners would often be at work when overgeneration occurred.

Room for Improvement

• The program was insignificant at shifting load to periods of overgeneration (11am – 3pm) during DR events.
• Prohibitively expensive networking costs (i.e., monthly cellular fees) paid by customers who own the charging devices.

Avista Utilities

Mike Vervair, engineer at Avista outlined their EVSE pilot program in Washington State. Avista owned, installed and maintained the chargers used in the program for residential and workplace charging applications. During anticipated peak hours, Avista tracked meter data to determine when to curtail charging load via signals sent to networked chargers. Customers were notified a day in advance via a phone app to allow for opting out.

Key Takeaways

• Overall, there was an 84% average opt-in rate despite being a completely voluntary program with no incentives for participating in individual DR events.
• Decreased average load per EV by 69.2% during curtailment periods.
• EV loads are often geographically clustered, with several households having more than one EV. This further incentivizes the need for managed charging to avoid overloading and upgrading transformer infrastructure.

Room for improvement

• Reliance on customer Wi-Fi networks proved unreliable, with 30-45% of systems losing connection with charging devices.
• Customers preferred a simple way to opt-out of DR events.
• It is still difficult to estimate at what level of EV penetration these programs will make fiscal sense.
• The next steps will expand the program to test different control groups to determine customer impacts, add additional EVSE models to the testing group, and explore using the utility’s AMI network for communications instead of customer Wi-Fi.

Pacific Gas and Electric

Karim Farhat, Expert Product Manager, Vehicle Grid Integration at PG&E detailed their ongoing ChargeForward pilot with BMW. In its first phase, 96 BMW i3 drivers were enrolled with a $1,000 incentive. BMW created its own proprietary aggregation software and app, which could delay charging via cellular telematics on the vehicle. BMW also tested 2nd-life EV batteries for grid services to meet DR requirements. In its second phase, the program expanded to 350 participants, focused on maximizing renewable energy intake, accounting for residential and away-from-home charging, and offering DR grid services.

Key Takeaways

• Customers were highly satisfied, with 93% of participants willing to recommend the program to others.
• EV charging optimized with renewable energy overgeneration through day-ahead forecasting from PG&E and BMW.
• BMW met 90% of load requirements for DR events with an average 20% contribution from EVs and 80% from the 2nd life battery system.
• Customers shared that the program educated them about the impacts their EV had on the grid, and expressed enthusiasm in knowing their EVs were charging from excess renewable energy supply.

Room for Improvement

• Most participation in EV overgeneration optimization took place during the weekend when more residential EVs were at home.
• There was low contribution to demand response events from participating EVs, with a strong reliance on second-life batteries.

Ensuring EVs Become Grid Assets

Managed charging technology offers a solution for the mass adoption of flexible EV loads. As shown in the three case studies, wherever possible, decreasing direct customer involvement and facilitating simple and seamless implementation drives improved results and customer satisfaction. Managed charging programs can educate customers about the grid and how their EV interacts with the larger system, driving greater utility engagement.

All three programs expressed difficulty with networking managed charging systems. Avista encountered reliability issues which complicated charging behavior forecasts. SCE conveyed that current costs for networking services were too high and were a barrier to customer implementation. Since PG&E used cellular auto telematics, direct customer networking barriers were less of an issue. However, since these products are proprietary to the automaker, utility access to information regarding charging habits can be limited. The industry still has much work to do to make managed charging programs the most customer-friendly, inexpensive, and reliable.

Despite the challenges, utilities play a key role to help the industry converge around a common set of messaging protocols and standards, identify best practices to decrease costs, and proactively engage with customers to ensure EVs become grid assets and not grid liabilities.