Electric vehicles present all kinds of challenges to the traditional ways of understanding cars. From design to differentiation, from range to refueling, EVs simply act different than the internal combustion-powered cars we’ve been refining for centuries now. And yet, through consumer incentives and subsidized charging stations, governments seem to be barreling headlong towards the goal of simply replacing our gas cars with electric ones, as if the two were fundamentally interchangeable. Sadly this is not the case, and a study by Project Better Place and PJM Interconnection [PDF] illustrates in stark terms just how costly an unplanned, uncoordinated rush to electric cars can be.
PJM and Better Place open their study with a question that some might find slightly absurd: what would happen if a major metropolitan area suddenly had a million EVs? The question is only absurd from a pure market perspective, as global EV sales volume projections are generally low enough to keep the possibility of a single million-EV metropolis squarely in the realm of science fiction. From a policy perspective, however, the study offers profound insights into issues that the governments who are currently promoting EVs absolutely must consider. Without an understanding of the unintended consequences of a rush to EVs, governments risk spiraling costs, misplaced investments, and market failures.
To understand the potential effects of a million-EV metropolis, PJM and Better place have created a complex computer model which
considered a distribution of 1 million EVs in the Washington-Baltimore Metropolitan Area and modeled the impact of charging the EV batteries in three scenarios: unmanaged charging, consumer-price-incentivized charging, and managed charging via a Central Network Operator (CNO).
With a million EVs in one metropolitan area, a huge percentage of grid energy would be diverted towards transportation that was once powered by gasoline, and these three scenarios represent different approaches to managing the grid impact. The first, or “unmanaged” scenario is essentially the status quo, a market-driven pricing system in which cars are simply powered off of a standard electrical grid using home chargers and the public fast chargers that some cities are already installing (called Battery Quick Chargers or BQCs). The “Time Of Use” (TOU) scenario used a two-tier pricing scenario, modeled on the pilot EV tariff developed by Southern California Edison, which uses advanced home meters to distribute energy for (theoretically) lower grid impacts and electricity prices (as well as public BQCs). The “Central Network Operator” (CNO) scenario models a single EV services provider responsible for all charging and infrastructure, using Better Place’s in-house network models and experiences. In this scenario, the BQCs are replaced by BSSs, or Battery Swap Stations, another unique Better Place offering.
Without going into too much complexity in describing the simulation (check out the PDF for more), it starts with a transportation model which maps EV distribution, trips and charging behavior. That model is then run through each of the three different scenarios, and the results of each is then sent through PJM’s grid market model and assessed for impacts on grid load and energy prices (assuming no fundamental changes in generation and transmission techniques). The results are dramatic, and graphically illustrate the problem with a vehicle-centric approach to EV stimulus.
As the very first chart in this post shows (also shown here in grey), the unmanaged scenario causes huge peaks and valleys in grid load, as commuters follow regular schedules and charge their vehicles at roughly the same times, charging them until full as soon as they are plugged in. The red line in that chart tracks “Locational Marginal Prices” (LMPs), which are at their highest when the grid faces its highest draws. This results in $786.3m in wholesale energy increases per year, a number that the TOU scenario (shown above) actually makes worse by 4%. Where TOU does help is in the annual energy costs aggregated to EV owners (thanks to fixed prices), but it is only shown to help by a mere 3.7%.
If you replace the haphazard system of home-charging and public BQCs with Better Place’s battery swap stations (BSSs) and network management system, the peaks and valleys in the grid draw are dramatically leveled out compared to the unmanaged and TOU scenarios. And though localized marginal prices are higher at times than in the TOU scenario, on aggregate they offer 22% savings compared to the unmanaged scenario. That’s over $35m annually (in one city) that’s not coming out of consumer’s pockets. More importantly, wholesale energy prices enjoy a whopping 45% savings compared to the unmanaged scenario for a staggering $350m in annual savings. Now imagine those results multiplied across every American metropolis with a million vehicles, and the impacts of not committing to a central network operator are impossible to ignore on a national policymaking level.
In essence, only a single central network operator can manage the chaos of individual transportation without restricting mobility or causing regular stress on the grid. I personally tend to favor bottom-up, market driven solutions, and at first glance putting a single operator in charge of managing the distribution of energy for private transportation does not seem to be that. But when you go through the model it becomes clear that this single central switchboard and distribution system is actually necessary for efficient market function, allowing for constant response to localized marginal prices and constant mitigation of naturally clustered usage patterns. In light of this reality, the study’s policy implications are less shocking:
This joint study firmly concludes that the increases in wholesale energy cost due to the additional load of 1 million EVs in the Washington-Baltimore Metropolitan Area can be reduced by hundreds of millions of dollars per year if the charging is managed by a CNO responding to real-time LMPs. These savings are without considering the value from various ancillary services and of large-scale dispatchable load for increasing the penetration of renewables, economic dispatch efficiency, and heat-rates for environmental considerations. Existing mechanisms do not necessarily allow CNOs to capture any of this value, which could be used for infrastructure deployment. Based on these conclusions, we emphasize how critically important both the presence of real-time LMPs and of CNOs are to reducing the impacts to the electric power system. Therefore, we recommend that incentives be developed for advancing the power system such that PRD incorporates LMPs and for EV incentives to reach beyond the consumer to CNOs so that intelligent charging networks can be quickly constructed.
By simply giving consumers credits to buy EVs, the government is setting up the same consumers to overpay massively for their electricity, grids for overstress and utilities for waste and inefficiency. Rather than encouraging these negative outcomes, perhaps governments should consider investing in Better Place’s holistic network management approach. The upfront costs of a Better Place-style CNO are indeed large, but the alternative is well-over $350m in annual increased wholesale energy costs (in one city alone)… waste without end. Throughout history economists have found so-called “natural monopolies,” in which markets are unable to provide a service as efficiently as a single actor. With the problem of EV grid management, we seem to have found another. And because the battery-swap model also fixes the major micro-level problems with EVs, namely lack of range and battery depreciation costs, Better Place is looking more and more like a no-brainer to me all the time.