Electric Vehicles are not air conditioners and why their load can’t be viewed the same way
June 5, 2019
June 5, 2019
Electric vehicles are not just another appliance on the horizon that will consume more energy. They can generate more demand than a house, their demand moves and their potential adoption rate is still unclear. What makes these challenges even more difficult for someone managing load is that there isn’t one standalone solution that will work for everyone. Every service territory is going to have to look at their own unique EV ecosystem and with the evolution of this technology problems may be approaching quicker than they realize. However, before utilities can come up with a solution for this problem, they need to better understand it.
Pricing signals, like TOU rates, are a common way for utilities to shift load to their off-peak hours however when applied to EVs they can create a new problem.
Changes with EV technology and their growing place in the market
It’s no surprise that charging electric vehicles is becoming a growing concern for anyone who is managing electrical load. Bloomberg New Energy Finance announced that over 2 million electric vehicles were sold in 2018 and they estimate that number will grow to 56 million by 2040, representing 57% of all passenger vehicle sales.
While 2040 is still a ways off, it’s important for utilities to start preparing for this increase in demand now. It’s not just a question of how many EVs there will be eventually, but also what type there are currently. With the increases in battery technology, grid disruptions can now occur with fewer vehicles. New long-range vehicles, such as the Tesla Model X, can have a battery capacity of up to 100 kWh, more than triple the average daily energy consumption of a household in the United States (28.5 kWh). Given that battery range is a primary selling feature for these vehicles, these capacities are going to continue to grow. You also need to consider new vehicle types, like the planned Rivian R1T electric pickup truck with its 180 kWh battery, that will enter the market in the near future. The concerns brought on by these batteries are further compounded when you realize you will have numerous vehicles charging on the same feeder. During our study on EV charging load it was concluded that these electric feeders would overload from the simultaneous charging of as few as five long-range battery electric vehicles (BEV). Once you add the fact that current solutions for managing load are less effective when applied to EV charging you come to the realization that profiling the EV ecosystem of your territory is critical.
The difference aspects of managing EV load
EV load presents many unique challenges that make it necessary to approach it differently. It’s dynamic and will continue to evolve with the previously mentioned changes in technology as well as the overall increased adoption. Battery efficiencies and the availability of charging stations will greatly affect driver habits. This poses the next challenge: there is no static location as drivers can charge at home, work or public charging stations. While it’s been well-established that most EV charging happens at home, this trend is now changing. This reduction in home-based charging coincides with the increase of workplace and public charging, an important distinction. As EV market share increases so does the likelihood that there would be more than one EV charging at these locations. The rise in EVs being used as fleet vehicles may also create concerns as they will most likely charge at a single location and are less likely to be influenced by curtailment initiatives like time-of use rates. These clustering effects are also not strictly limited to workplace charging as socioeconomic factors can also result in a larger amount EVs in specific neighborhoods and public charging stations are becoming more common in places like retail outlets, universities or hospitals. All of this can create pockets of high demand, making it difficult to plan for infrastructure upgrades. Lastly, unlike other high demand residential devices such as air conditioners, EVs are used year round. Seasonal weather does affect battery performance but the peak coincident load on distribution infrastructure may not be seasonally impacted. During the colder months vehicles have to be charged more often comparatively, however they typically are driven more overall in the warmer months. When you consider all of these factors it’s easy to say that EV load is unlike any other.
While infrastructure-side profiling can provide you with EV charging data, it’s going to be fragmented.
Service territories are not as similar as you think
Another challenge when planning for EV load is that no two service territories are alike. At first glance you might think you can compare your territory to another with a similar population or climate. However, this would be a mistake as there are many factors that are being overlooked. The physical characteristics of an area, such as size, topography or population density, will greatly vary from one another. What is the make-up of city versus highway driving? How many commuters do you facilitate and what are their cycles? Every territory’s total EV market share, as well as model specific market share, will be different and this may influence the number of long-range BEVs versus short range PHEVs. How these drivers charge may also be influenced by the local factors as service providers may use TOU rate structures to encourage charging off-peak. However many workplaces that have charging stations offer free charging to their employees, making these TOU rates a non-issue. Simply put you can’t accurately assume what is in your territory or how it’s being used.
Managing EV Load
Since EV load is different from other types it needs to be managed differently as well. Pricing signals, like TOU rates, are a common way for utilities to shift load to their off-peak hours however when applied to EVs they can create a new problem. While a static residential TOU rate structure can reduce the amount of morning and midday charging it can also create a new second peak. This unintentional peak often begins immediately after peak-pricing ends and coincides with other shifted load such as cooking, laundry, etc. Some utilities may decide to monitor EV charging by installing a submeter, which would then allow them to create a separate EV-specific TOU rate structure.
The issue with this is that it’s an expensive and time-consuming process which will require EV owners to have a networked EVSE to meet regulatory standards. These pricing signals also do not take into consideration mid trip (opportunity) charging, which cannot be shifted, and the previously discussed workplace charging. Utilities may also be considering using these EVSEs for direct load control, which is not a great option either as ESVEs cannot detect the vehicle’s battery state-of-charge (SOC). Without this data, in the event of a demand response call, they cannot guarantee the driver will have the amount of charge they need for their next trip. Managing EV specific load is going to become a large component of load management, but before figuring out this solution you need to fully understand your problem by profiling your territory.
Profiling with vehicle-side data
When collecting data on the EVs in your service territories you have two options: infrastructure-side profiling or vehicle-side profiling. While infrastructure-side profiling can provide you with EV charging data, it’s going to be fragmented. Firstly you will only be capturing charge events from networked charging stations and there are numerous competing networks, so you would need to have agreements for each one. Most infrastructure-side profiling only takes home charging into consideration so any public or workplace charging wouldn’t be collected. Even if you collected all home and public networked charging data you still wouldn’t have an accurate representation since is doesn’t factor in Level 1 charging, non-networked Level 2 charging or DCFC charging which collectively represent 56% of charging. Vehicle-side profiling ignores all of these issues by collecting from the EV itself. It’s charge station agnostic from both a brand and type perspective. It also collects data regardless of the station’s location, although you can limit this through geofences so you aren’t including events that occur outside of your territory. Another benefit of this type of profiling is that you can gather data on the specific make and models of the EVs charging in your area as well as trip related information. Finally, you can also collect battery SOC data which was mentioned earlier in this article. This data is what allows for the only direct load control solution which guarantees participants will have sufficient range for their next trip.
Everything electric utilities need to integrate electric vehicles with the grid
FleetCarma, a division of GEOTAB, has helped numerous utility companies prepare for the integration of electric vehicles into the grid with through pilot programs. This is accomplished with their SmartCharge Platform which includes vehicle-side profiling, behavioural load shifting through a rewards program and a direct load control software that can operate as a standalone solution or be incorporated into a DRMS. To see an example of one of these programs download the case study included at the end of this article.
Summary
Electric vehicles are not only disrupting the automotive industry but are forever altering the way utilities manage electrical load. The increasingly faster adoption rate and the advancing technology is not only a problem for the future but has potential to create one now. They are vastly different from other types of electrical load and need to be treated as such. The first and most crucial step for utility companies is to accurately profile their service territory.
More resources and information
