|Fig. 1: A Tesla Model S and a Tesla Model X charging side by side. (Source: S. Jurvetson, reproduced under terms of the Creative Commons 2.0 License).|
Since the arrival of highway-legal commercial plug-in electric vehicles in the United States in 2008, their market share have seen tremendous growth. In California, plug-in hybrids and electric vehicles such as the Teslas shown in Fig. 1 already make up 3.1% of all vehicles registered on the road in 2015, an astounding growth considering it has only been a decade since the introduction of these vehicles.  While the trend towards zero-emission vehicles is a great stepping stone in our goal of reducing the production of green-house gasses, their popularity does bring about new challenges to our existing infrastructure. One particular challenge is the issue of the extra load that plug-in vehicles place on our existing electrical grid infrastructure. Due to the speed of electrical vehicle adoption, this phenomenon must be addressed in order to ensure the smooth adoption of plug-in electric vehicles as well as maintaining the stability of our grid system.
In order to understand the issue of plug-in electric vehicle adoption as well as the potential impact to our grid, we must understand the sheer scale of the problem. Plug-in electric vehicles include both pure electric cars such as Tesla shown in Fig. 1, as well as plug-in hybrids that can charge from a charging station such as those shown in Fig. 2. According to the U.S Energy Information Administration, 28% of all energy consumed in the U.S was used for transportation, and 77% of the energy used in transportation comes from gasoline and diesel alone (not including jet fuel and other liquid fuels).  The sheer scale of the energy we use in transportation outlines our need to understand how this might impact our grid when we gradually move over from fossil fuels.
|Fig. 2: An example of plug-in hybrid vehicle. (Source: Wikimedia Commons)|
To get a more precise estimate of the excess load that might be placed upon our grid, we need to understand how much of our energy is consumed by personal transportation. We can do a quick back-of-the-envelope top-down calculation based on current vehicle usage statistics. In 2015, the EIA estimated that 14,152 trillion BTU of fuel was consumed by personal vehicles.  If we convert that into watt-hours, and divide that number by the number of days in a year (365) and the number of households in the United States (116,211,092 according to the 2014 ACS community survey), we get 98 Kilowatt-hours per day of energy usage for personal transportation per household.  Using the fact that electric vehicles averages around 100 mpge in terms of energy efficiency compared to the average of 22 mpg for gasoline vehicles, we can assume that electric vehicles will use 5 times less energy to cover the same distance, thus using approximately 20 kilowatt-hours per day per household.  A level 2 charger currently charge at 6.6 kw, thus each household will gain an additional 6.6 kwh load for approximately 3 hours. This top-down estimate does not include the fact that most commuters with longer trips would opt for non-electric vehicles due to range limitations as well as psychological biases such as range anxiety when it comes to electric vehicles for longer trips. That means this estimate would likely be an upper-bound estimate on the extra load demand that a household will place on the grid if they adopt the plug-in electric vehicle as their primary mode of personal transportation.
Other studies have attempted to estimate the impact on the adoption of electric vehicles on the grid as well. In a study conducted by Northumbria University in the UK, the researchers estimated that a 30% adoption of electric vehicles would increase the residential peak demand between 6pm - 10pm by an additional 45%.  Without proper load-balancing, the introduction of a large number of plug-in electric vehicles into the residential grid can cause a series of problems.
The large introduction of load sources into the residential grid system can cause a myriad of problems. The two biggest problems comes from the delivery and the generation needed to match a sudden demand peak. When an electrical circuit is installed, it has a maximum load capacity; any electrical circuits has a power delivery limitation. Exceeding the maximum designed power capacity causes degradation as well as dangerous conditions around the circuit. Without load-balancing, the increase in peak power delivery would necessitate the reconstruction or reinforcement of the power delivery system at every stage of the delivery infrastructure. The generation of peak power itself is a major issue. Peaker power plants often have terrible capacity utilization since they are only turned on during the peak hours of the day, and most of the peaker power plants are generated using natural gas and other non-renuable fossil fuel resources due to the easy of ramp up and ramp down. The number of natural gas peaker plants have actually grown in the recent year despite the overall gradual trend towards renewable electricity generation. 
The primary method to remediate this problem would be the adoption of load-balanced residential charging units.  Residential electricity demand typically peaks during the hours of 6pm - 10pm, which there is a significant drop-off during the hours of 12am - 5am. If the charging of the electrical vehicle can be shifted towards those hours, there will be less need to implement higher peak power infrastructure. In order to encourage load-balancing, time-specific electrical pricing might need to be implemented. By displacing the charging to off-peak hours, one interesting positive externality is the improved overall load balancing. While the overall electrical demand is higher, the relatively more even power consumption can lead to a more stable base power demand, which can allow better deployment of renewable energy generation sources.
Another possible solution would be to use a distributed energy storage system in conjunction with a smart-grid in order to maintain a constant balance of electric power.  Instead of simply having electricity consumers, a distributed energy storage network will allow for the electric consumer to reduce their power consumption during peak hours. The system functions by having battery storage system in residential households. These batteries would charge during off-peak hours or, if the homes are equipped with distributed energy generation sources, during hours when the energy generated is higher than consumption. The batteries would then discharge during the peak electric hours so that the load from the individual residential unit is lowered during peak hours.
Finally, perhaps the most surprising solution could be the electric vehicle themselves. The vehicles themselves present a large capacity for energy storage.  Residential vehicles are not constantly in operation, and in-fact, often have a relatively low utilization rate. That means for the overwhelming amount of time, the electric vehicle will be in idle at a certain location. These idle vehicles can be used as a distributed energy storage system. Using machine learning algorithm, we can effectively predict the commuting habits of residential vehicles, and thus effectively use them as load-balancing batteries without interfering with their owners driving tendencies.
Electric vehicles present perhaps the best opportunity to reduce our dependency on green-house gas producing energy sources without a complete reinvention of our current infrastructure. However, their adoption will inevitably bring about challenges to our electrical grid. By understanding the scale and potential impact of millions of electric vehicles hitting the road, we can effectively design infrastructure as well as incentive programs that will maximize the benefit of electric vehicles. Through solutions such as time delayed load balancing as well as distributed energy storage systems, we can not only supply the power needed for these electric vehicles, but also increase the reliability of our existing grid infrastructure.
© Peter Wang. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
 "California Auto Outlook," California New Car Dealers Association, 12, No. 4 (2016).
 "November 2016 Monthly Energy Review," U.S. Energy Information Administration, DOE/EIA-0035(2016/11), November 2016.
 "Annual Energy Outlook 2016," U.S. Energy Information Administration, DOE/EIA-0383(2016), August 2016.
 J. Vespa, J. M. Lewis, and R. M. Kreider, "America's Families and Living Arrangements: 2012," U.S. Census Bureau, P20-570, August 2013.
 "Fuel Efficiency Guide," U.S Department of Energy, DOE/EE-1249, November 2016.
 G. A. Putrus et al., "Impact of Electric Vehicles on Power Distribution," IEEE 5289760, 7 Sep 09.
 M. Nyberg, "Thermal Efficiency of Gas-Fired Generation in California: 2014 Update," California Energy Commission, CEC-200-2014-005, September 2014.
 A. Mishra et al., "Scaling Distributed Energy Storage for Grid Peak Reduction," in Proceedings of the Fourth International Conference on Future Energy Systems (ACM, 2013), p. 3.
 Z. Wang and S. Wang, "Grid Power Peak Shaving and Valley Filling Using Vehicle-to-Grid Systems," IEEE Trans. Power Del. 28, 1822 (2013).