Many electric vehicle and renewable energy enthusiasts are enthralled by the vast potential of a technology termed V2G. It’s argued that this technology could allow for widely adopted electric vehicles to provide storage for the electric grid when not in use, paving the way for a cleaner and more resilient grid. But is the technology economically viable? And how efficient is electric vehicle charging and discharging? This is the subject of two recent (and freely available) articles in the journal, Energy.
Vehicle-to-grid (V2G) – also known as vehicle-grid-integration (VGI) or grid-integrated vehicles (GIV) – is a widely discussed technology option for energy storage in electrical grids. Essentially, V2G recognizes electric vehicles as stationary storage assets when not being used for transportation purposes. Charging and discharging of parked vehicles, aggregated into blocks of 100 kW or more, can be used for grid services like frequency regulation, peak shaving and buffering variable renewable generation.
Academics have accepted the research challenges of V2G with a near avalanche of peer-reviewed articles on the topic in recent years. Google Scholar yields over 14,000 results for the term ‘V2G’ since 2010. (The search term ‘V2G’ captures a small proportion of erroneous results, but also doesn’t capture those research articles using the similar terms ‘VGI’ or ‘GIV’.) Findings from these analyses generally presume or conclude a bright future for V2G. Detractors, however, argue that V2G is based on overly-optimistic assumptions and is not the best energy storage option in a future smart grid.
One major shortcoming of V2G research is the dearth of empirical performance data to inform cost-benefit analyses. Roundtrip electrical efficiency (defined as the amount of energy provided to the grid divided by the amount of energy drawn from the grid) remained unreported until very recently despite several years of V2G trials across the country. Accurate roundtrip efficiency assumptions are crucial for determining the economic viability of this technology.
It is sometimes believed that the roundtrip efficiency of the li-ion battery is the appropriate metric for analysis. In reality, most losses during EV charging and discharging occur in components other than the li-ion cells, such as the inverter, EVSE, transformer, and parasitic loads. Thus, the appropriate measure for storage is the efficiency of the storage system.
As we highlight in our paper (described below), roundtrip losses erode V2G economics in at least five key ways. Roundtrip losses 1) Increase the billed kWhs from the power provider; 2) Increase the energy throughput of the battery, thereby accelerating battery degradation; 3) Decrease the maximum energy available for the grid; 4) Decrease the maximum power available to the grid; 5) And for frequency regulation specifically, reduce the maximum achievable accuracy of a vehicle’s response to the automatic generation signal, resulting in proportionally reduced compensation.
Without the benefit of prior empirical data, published V2G analysis relied on assumptions for roundtrip efficiency values (or omitted the concept altogether). Recently published data suggest actual roundtrip efficiency may be substantially lower than prevailing assumptions.
Apostolaki-Iosifidou, Codani and Kempton (2017)
A recent study conducted at the University of Delaware and published in the journal Energy, reports the first empirical data on this subject (freely available here). The study measured losses from a grid-connected electric vehicle at five nodes along the charging and discharging pathway; transformer, breakers, EVSE, onboard power electronics unit (PEU), and li-ion battery.
Losses for the system are reported at two currents. At 10A (~2kW), losses were 17% during charging and 36% during discharging. At 40A (~10kW), losses were 12% during charging and 30% during discharging. These values are shown in Table 7 of the paper and feature in the paper’s abstract.
In the 40A scenario, which represents a more likely real-world scenario, the onboard power electronics unit (PEU) is the largest source of loss, yielding charging losses of 6% and discharging losses of 19% (Table 7). The second largest source of loss is the transformer, yielding charging losses of 3% and discharging losses of 7%. The authors note that the transformer employed in the experimental design was over-sized for the relatively low currents in the experiment, a factor which likely inflated transformer losses.
Shirazi and Sachs (2018)
David Sachs and I published a discussion piece in the same journal (freely available here) exploring the implications of the above findings for V2G economics. Our first step was to convert the data of losses into the more familiar measurement of efficiency.
The mechanics of this conversion are straightforward. Efficiency is equal to 100% minus losses. Roundtrip efficiency is calculated by multiplying the efficiencies of each step in the energy conversion process. For example, a component with 90% efficiency during charging and 90% efficiency during discharging has a roundtrip efficiency of 81%.
Therefore, the 40A scenario has an efficiency of 88% during charging and 70% during discharging. This equates to a roundtrip efficiency of 62%. (Note: We agree with the authors that their choice of transformer likely decreased efficiency, but we also note that measurements were conducted only at ideal temperatures, which likely increased efficiency. We suspect that these two factors partially or fully offset each other and encourage future studies to take these factors into account).
For comparison, we identified 12 published V2G economic analysis with transparent methodologies. Of those we identified, half assume roundtrip efficiencies ranging from 73% and 86%, while the other half assume 100% roundtrip efficiency – meaning no losses whatsoever!
As we demonstrate in our paper, combining 62% efficiency with recent PJM average frequency regulation of $13/MW-h mean that electricity losses offset over 50% of total V2G revenue.
Authors of the original research replied to our work with a comment of their own (freely available here).
This is merely the beginning of an overdue discussion regarding electrical efficiency of V2G and its impacts on V2G economics. Many peer-reviewed studies ignore this important cost element, indicating the degree to which the peer-review process can improve. Existing V2G economic analyses should be re-specified with empirical estimates of roundtrip efficiency for more accurate conclusions. In an upcoming paper, we propose that V2G analyses also fail to incorporate other major costs elements, such as demand charges, aggregator fees, convenience losses and voided warranty costs, among others.
While providing a solution to renewable intermittency is extremely important for a grid of the future, we are not certain that V2G is the best solution in most cases. Other alternatives, like stationary storage technologies, demand response and others may provide similar benefits at lower costs to society.
Co-authored with David Sachs
Apostolaki-Iosifidou, E., Codani, P., & Kempton, W. (2017). Measurement of power loss during electric vehicle charging and discharging. Energy, 127, 730-742.
Shirazi, Y. A., & Sachs, D. L. (2018). Comments on “Measurement of power loss during electric vehicle charging and discharging”–Notable findings for V2G economics. Energy, 142, 1139-1141.
Apostolaki-Iosifidou, E., Kempton, W., & Codani, P. (2018) Reply to Shirazi and Sachs comments on “Measurement of Power Loss During Electric Vehicle Charging and Discharging”. Energy, 142, 1142-1143.