Vehicle-to-grid project reveals challenges of the early days

Los Angeles Air Force Base became the first federal facility to replace its entire general-purpose fleet with plug-in vehicles in a two-year V2G research project

Vehicle-to-grid (V2G) is a technical and policy-based effort to optimize the way EVs interact with the electrical grid. V2G developers view EVs as “resources” within an electrical network. In V2G literature, the word resource is often used instead of cars or vehicles.

The notion of a resource is based on the idea that a vehicle linked to a charger can (and should) do more than just take on fuel. At night or during the work day, the car’s battery can store power when overall electrical demand is low but wind and solar facilities are generating inexpensive power. When demand and generation shift, the battery can send power to the grid, offsetting the need for additional generation. EVs can also provide ancillary services, helping to regulate the stability of the grid.

Obviously, such benefits will not be delivered by just one EV, perhaps not even by 1,000 or 10,000. But eventually, it’s expected that a grid/EV dynamic will develop that is dependable and mutually beneficial for vehicles and the network.

Also obviously: this ain’t easy. Right now, V2G is still pretty theoretical. It could happen, if…

California, as in all things EV, leads in V2G research. For state policymakers, V2G is a must, not a want, required by Governor Brown’s Executive Order B-16-2012, which directs that “electric vehicle charging will be integrated into the electricity grid” by 2020, less than two years away. Policymakers need this working sooner, not later.

In December, a major V2G research project concluded in California, closely watched because of its scope and its utility market interaction. The project, which ran from October 2015 to December 2017, was a partnership between Southern California Edison (SCE) and the Department of Defense (DOD), specifically Los Angeles Air Force Base, part of a $20-million DOD research investment. The base became the first federal facility to replace its entire general-purpose fleet with 41 plug-in vehicles, including automobiles, pickup trucks and a passenger van.

In December, SCE filed its V2G report with the California Public Utilities Commission. SCE calls the project successful because the work supported “a pioneering customer in the direct participation space.” SCE writes that “the greatest accomplishment of the V2G Pilot is that it has paved the path for smaller resources to participate in the ancillary services market going forward.” On its web site, SCE writes that a successful pilot could be the “proof of concept” that helps establish the viability and scalability of V2G technology.

From SCE’s December report, viability and scalability seem to advance somewhat, but unevenly – the phrase “bits and pieces” comes to mind. SCE’s report makes it hard to assess how much this project (pardon the sports metaphor) moved the ball – 3 yards, 30 yards, or perhaps even close to the goal line.

One thing is certain: the project required a lot of remedial work. SCE and Air Force engineers started almost from scratch, expending a lot of effort just to get to the work at hand. SCE and the Air Force repeatedly confronted challenges which, taken together, belie any notions that V2G is at the almost-ready stage. SCE writes of shepherding “the pilot through myriad testing, qualification and participation wickets.”

For example, engineers hoped to take advantage of V2G work from the University of Delaware (UD), work that was favorably referenced among V2G proponents. It turned out, though, that UD’s components were proprietary, and couldn’t be used for the new project. Furthermore, UD communications protocols did not comply with SAE standards. Surprisingly, UD’s inverter was not utility-certified, something which is required throughout the US. In sum, UD’s work didn’t provide any advantage for SCE/Air Force. This is more than disappointing, and it’s not only about technology. V2G’s success depends on an open, common body of knowledge – secrets are a setback.

Another unexpected problem: even in California, there were no commercially available vehicles and electrical supply hardware, other than for the Nissan LEAF, capable of meeting all of the tasks and functions required by the project.

The project started with the Nissan LEAF, a bus from Phoenix, and trucks from EVAOS, EVI and VIA Motors. None of the equipment could pass initial testing in its manufactured configuration. Issues included, but were not limited to: limited charger functionality; standby load that was higher than SCE’s recommendations for charging infrastructure; heavy parasitic loads that reduced the reliability of the vehicles; and inverter issues. Each vehicle had to be recalibrated and all compliance and functionality issues readdressed. Interestingly, the manufacturers couldn’t perform the requisite tests, which eventually took over two years. SCE expected testing to take six months.

In fact, the trucks were never used – they were shipped to Texas. That left 29 vehicles – a somewhat puny fleet, in retrospect, considering this was supposed to be the largest V2G project in the world. Among the vehicles, quality and performance issues were troubling. Some vehicles that passed the certification test in October 2015 were no longer reliably operational just one year later.

One important project goal was to show that EVs could participate in CAISO’s wholesale market. (The California Independent System Operator is the entity that organizes and oversees generation and transmission within multiple utility service areas.) SCE calls this a success: “The Pilot was successful in providing frequency regulation to the CAISO market for a total of 243 MWh of Regulation Up and 102 MWh of Regulation Down from May 2016 to September 2017.” (Up/Down refers to receiving or sending power from or to the grid.)

However, market interaction had problems. The vehicle fleet could not provide accurate information to CAISO. This impacted day-ahead market awards, and it impacted vehicle performance because it’s related to battery issues. New tools are needed, SCE writes, to automate energy bidding but still track power. Consequently, the Air Force had to reduce the number of hours its vehicles were in the market, a move completely contrary to V2G’s potential benefits.

It’s important to note that a fleet presents different V2G characteristics than do individual EVs. The Air Force’s fleet needed to be ready on demand, which conflicts somewhat with the idea of using an EV as a grid resource. An individual EV owner would likely take advantage of a more predictable V2G schedule, staying connected at night, for example, or during the workday when the vehicle is parked.

There were other challenges. Communication software didn’t work, which meant the vehicles’ market performance wasn’t stabilized until early 2017. Problems compounded: CAISO had to decertify some of the EV resources from certain activities because of inaccurate and insufficient data. Again, a big hit when thinking of EVs as assets.

The project used two communications standards: Open Charge Point Protocol 1.5 (OCPP) and the SAE Smart Energy Profile 2.0 (SEP2). They didn’t work. It took months for engineers to make corrections and build new operating features. Some engineers questioned why two standards were used, when one was likely sufficient. OCPP was chosen very early in the project’s development by the DOE’s Lawrence Berkeley National Laboratory; it was retained, even though engineers knew early on that its performance was questionable.

The re-tooled software got the job done. But again, to what end? SCE’s report does not say whether these fixes just kept this particular project going or whether they are the kind of advances that could lead to standardized applications. When asked about a metric for progress, an SCE spokesperson said (via email) that “the V2G pilot moved knowledge forward towards COTS (commercially available off the shelf) products. Additional insight would need to come from the customer, the Department of Defense.”

One reason that every electron needs to be accounted for within a V2G system is because money is involved – for the resource owner, the utility, ratepayers, taxpayers, grid operators and new entrepreneurs who want to provide services in “this direct participation space.”

Importantly, the Air Force did realize income from its energy trades, despite the extensive communication-software problems. Revenue per vehicle ranged from $25 to $72 per month, with an average of $41 per vehicle-month.

Aggregated gross revenue was $7,639. Unforeseen, however, were high monthly participation fees. SCE had to charge a manual billing fee of $118 and a meter data feed fee of $216. The biggest hit was a $1,000 monthly fee for a CAISO scheduling coordinator, a person certified to participate in trades. In total, the project lost about $17,000.

SCE’s report suggests that the fees could possibly be reduced by “utilizing a significantly larger resource,” i.e., more EVs linked together, lowering the cost per vehicle. How many vehicles would that be? SCE’s answer is a bit of a riddle: “By ‘larger resource’ we mean more vehicles included in an aggregation. The order of magnitude question depends on the desires of the EV owners.”

CAISO’s $1,000/month charge took project planners by surprise. When asked about the coordinator’s role, SCE explained that “Scheduling coordinators deal directly with the CAISO for bidding, self-scheduling and financially settling participatory resources. In short, it would be costly for each EV owner to be his or her own scheduling coordinator. Scheduling coordinator services can be purchased that would support aggregation of many EV owners’ vehicles into a resource through aggregation and thus share fees.” Obviously an issue needing further attention.

DOD’s V2G work will continue, including testing vehicle-to-building technologies. DOD will present its own final report in early 2018.


This article originally appeared in Charged Issue 35 – January/February 2018 – Subscribe now.