Matthew Mills, Manasseh Obi, Kendall Cody, Kyle Garton, Amanda Myers Wisser, Sammy Nabahani
©SHUTTERSTOCK.COM/ WRIGHT STUDIO
Electric Vehicle (EV) Adoption is accelerating across the United States, driven by consumer demand, public policy at all levels of government, and industry commitments to electrification. As a result, more than 50% of all new car sales are forecast to be electric by 2030. This article provides a snapshot of how Portland General Electric Company (PGE), a leading investor-owned utility based in Portland, OR, USA, is preparing for the increase in demand from EVs on the power grid.
If not adequately prepared for, rapid EV adoption can pose a challenge for utilities as distribution infrastructure was not originally built with widespread EV charging in mind. EV charging introduces new customer loads that don’t adhere to traditional load shapes, resulting in new or different spikes in peak demand. Consider a neighborhood with high EV adoption, where each Level 2 (240-V) charger doubles the peak demand of each home. When several EVs start charging at the same time, the total peak load that now includes EV charging may push the local service transformer beyond its rated capacity. Overload conditions can lead to accelerated aging or asset failure, resulting in unplanned outages and increasing costs for asset maintenance and replacement.
Opportunities exist to mitigate operational constraints, the cost impact of new infrastructure, and construction lead-time issues by orchestrating EV charging across neighborhoods to manage the total load and the operating requirements of distribution assets, including substations, feeders, and service transformers. One prime example is that utilities are preparing to deploy advanced software solutions to support local distribution and bulk system constraint analysis, planning, and optimizing charging. The utilization of distribution assets and planning thoughtfully for distribution investments are necessary parts of cost-effectively supporting continued and increasingly rapid EV adoption.
This article presents a case study on how PGE is evaluating an innovative approach to mitigating the potential impact of increased demand from EV adoption on the power grid. PGE is proactively designing strategies for advanced managed charging that reduce the need for infrastructure upgrades while delivering additional benefits for grid reliability, resiliency, affordability, and a better customer experience.
To better understand PGE’s work on distribution-optimized EV charging, this section briefly summarizes its system, the customer base, and the underpinning policy and regulatory context. This context is important because investor-owned utilities are highly regulated entities. Therefore, utilities benefit from being innovative, ambitious, and collaborative to address future concerns proactively within a supportive regulatory landscape.
PGE has a large and diverse customer base and electrical grid, which needs to be uniquely considered when designing EV programs. The service territory includes 1.9 million people, representing more than 900,000 m, more than 4,000 mi2, with 188 substations and 695 distribution feeders. As more residential customers and commercial fleets move toward EVs, PGE aims to better understand the impact that the growing number of EVs has on its distribution system so that it can plan accordingly.
In recent years, the Oregon Legislature and the Oregon Environmental Quality Commission (EQC) have adopted policies to support EVs and clean energy. The EQC has adopted the Advanced Clean Trucks rule, requiring manufacturers of medium- and heavy-duty vehicles to sell an increasing percentage of zero-emission vehicles beginning with the 2024 vehicle model. The EQC also adopted the Advanced Clean Cars II rule, requiring 100% of passenger vehicle sales to be electric by 2035. Additionally, in 2021, the Oregon Legislature enacted clean electricity standard legislation (HB 2021), requiring investor-owned utilities to reduce the greenhouse gas emissions associated with serving retail electricity consumers by at least 80% below a 2010–2012 average baseline emissions level by 2030, by at least 90% by 2035, and to 100% by 2040.
Anticipating the clean energy transition, including distributed energy resources (DERs), the Oregon Public Utility Commission (OPUC) began an investigation into historical distribution system planning practices. In 2021, the OPUC set guidelines “for a more transparent, robust, holistic, and community-centered utility planning process for distribution system operations and investments.†This investigation is the origin of the current distribution system planning process for investor-owned electric utilities in Oregon, which includes bottom-up and top-down planning practices to develop more advanced considerations for addressing distribution grid needs now and in the future.
The result of this planning is the Distribution System Plan (DSP), a PGE commitment to a transparent and publicly involved road mapping process. Figure 1 provides a high-level overview of the DSP. This plan is used to reinforce the utility’s ongoing commitments and details the vision, goals, and strategic initiatives currently active and the plans to meet our decarbonization targets. An example of a strategic initiative is identifying the need to find sources of flexibility for the utility’s supply portfolio. For example, the DSP calls out that roughly 25% of the flexible load can come from customers and DERs. Today, PGE has more than 40,000 EVs in its service area. Moreover, today, Oregon’s electricity mix comes from more than 50% of greenhouse gas-free sources of energy, including hydro, wind, and solar power. PGE forecasts that the transportation electrification potential resource is expected to increase from 12 MW average (MWa) in 2022 to 111 MWa in 2030 under the reference case scenario.
EVs can provide a suite of grid benefits as regulators and utilities better understand how EVs can serve the full range of generation, transmission, and distribution value. As the electricity mix decarbonizes, the generation profile also tends to become more variable due to renewable energy generation sources. Meanwhile, transportation electrification is a promising pathway to decarbonizing the light-duty transportation sector. These two trends go hand-in-hand; EVs join the ranks of DERs and can help align electricity demand with clean electricity supply while also managing impacts on the broader network. In other words, effective visibility and control will encourage a virtuous cycle between commitments to renewable energy generation and electric mobility.
To date, efforts around aligning EV charging with the grid have largely focused on bulk system optimization. Such EV programs reduce stress on the grid system by moving customer EV charging schedules to off-peak times. This work can and should certainly advance, but there is an ever greater need surfacing for EV charging optimization focused on the distribution grid.
Managed charging, also known as smart charging, is the optimization of EV charging behavior for financial and/or grid benefits. Effective managed charging prioritizes drivers’ needs and preferences and is typically limited to unidirectional power flow. Managed charging can be passive or active. Passive managed charging uses rates or rebates to influence customer charging behaviors and is often in the form of an EV time-of-use (TOU) rate, meaning that the price of energy varies based on the time of day, but those prices are the same each day (at least within a season). Typically, the objective of passive managed charging is to shift the EV charging load to off-peak times. Active managed charging is where the utility has direct control and influence over customer EV load and can be applied to solve other bulk system challenges. Such applications include reducing or avoiding the curtailment of renewable energy generation, reducing the load on the system in areas that are at risk of wildfires, and shedding load during times of grid stress. Addressing bulk system needs is a critical consideration, and often a first step, in optimizing EV charging. Managing EV charging with just bulk system benefits in mind, however, can have unintended consequences on local distribution systems that will be explored in the remainder of the article. When developing a plan to manage customer charging, utilities should ensure that both the bulk system and the distribution system are considered in optimization.
An example of a PGE program that primarily addresses the bulk system is the PGE Residential EV Smart Charging Pilot Program. PGE manages EV charging schedules against the customers’ TOU rates to shift unmanaged EV load (Figure 2) to avoid peak periods that occur from 5 p.m. to 8 p.m. Monday through Friday. When customers plug in their EVs before 8 p.m. on a weekday, the EVs’ charging schedule is set to begin right at the start of the off-peak period at 8 p.m. (Figure 3). If this schedule is applied to all EVs enrolled in the program, a negative secondary effect of that schedule would be creating a “timer peak†with a large spike in energy demand at 8 p.m. Timer peaks occur when EV customers are enrolled on a utility’s TOU rate and their vehicles start charging simultaneously at the point when the lowest-cost time period starts, which then creates a localized electricity demand peak.
As can be seen in Figure 3, shifting EV charging to capture bulk system benefits can create a separate and more localized peak in demand. Consider again that neighborhood with a high EV adoption rate, where each Level 2 charger doubles the peak load of each home. While this may not be a large-scale issue today, as EV adoption exponentially grows, a simplistic optimization approach such as passive managed charging can cause harm to the distribution grid. A need exists for more sophisticated optimization to solve and prevent this problem from occurring. The risk of creating secondary timer peaks becomes clear when considering how these peaks could impact local distribution systems.
The projections from PGE’s DSP seen in Figure 4 show that EV adoption is not expected to be uniform across the PGE service territory. Thus far, EVs tend to be purchased first in neighborhoods with a higher median household income, as seen in the Bethany Substation area, which leads to concentrations of EVs impacting the distribution system. While this socioeconomic trend is expected to change over time, the impact on the distribution system as EV adoption grows more widespread will only deepen.
As the number of EVs on individual feeders grows, so does the stress placed on those distribution assets, as can be seen in Figure 5. If bulk system optimization creates a timer peak for a feeder that has significant EV adoption, that peak could materially damage equipment on that feeder. Specifically, spikes in energy demand that exceed the equipment capacity ratings could lead to poor power quality, temporary blackouts, accelerated asset aging, and, eventually, asset failure.
For its residential smart charging program, PGE is working with a software platform provided by WeaveGrid to optimize EV charging at the distribution level. To optimize EV charging, this platform collects driver preferences, monitors vehicle telemetry data for all program participants, and then assesses the aggregate charging load requirements for all participating vehicles. The platform will then consider system constraints input by PGE to optimize charging schedules in conjunction with those constraints at various asset levels. Figure 6 demonstrates how EV optimization with system constraints in consideration can enable PGE to capture bulk system benefits while still protecting local grid assets.
Figure 6 demonstrates how more intelligent optimization can create additional value for customers and will provide benefits for PGE, including extending the lifetime of grid assets and avoiding unplanned asset replacement costs. A holistic optimization approach can benefit the bulk system and protect local distribution assets without compromising the needs of customers.
PGE initiated the Residential EV Charging Pilot Program to evaluate how various test groups would react to demand response (DR) and the managed charging of participating EVs. This Residential EV Charging Pilot Program has multiple smart charging programs available to customers, differentiated by whether the charge signals are sent to the Level 2 EV charger or directly to the vehicle via telematics (see Figure 7). PGE leveraged the WeaveGrid platform to utilize the telematics communications capabilities on board the vehicle alongside networked charging equipment. Operationally, PGE provides a forecast of scheduled DR events through the predefined DR seasons (October to March and April to September). Within these DR events, PGE is using WeaveGrid’s software to manage EV charging during scheduled DR event hours for all cars that are plugged in concurrently as well as to optimize charging based on user needs and TOU rates to avoid on-peak charging. This pilot program primarily optimizes based on bulk system considerations by leveraging customer rate data and DR signals that align with PGE’s residential TOU rate to shift load off peak.
To validate and better understand the importance of managing EV charging at the distribution level, PGE has released the next evolution, the Test Bed EV Charging Study (EV Charging Study). The EV Charging Study uses vehicle telematics to further study customer EV charging behaviors and their impacts on the distribution grid. PGE will control the time of EV charging while ensuring that the vehicles meet the drivers’ needs and will evaluate the customer acceptance of charge rate, charge time, and location-based price signals. Research in this project area will focus on improving the understanding of the technical paths for managed charging, costs, performance, and limitations.
For the EV Charging Study, PGE has identified eight distribution feeders that have an advanced level of EV penetration primarily located in the Bethany region northwest of Downtown Portland (Figure 4). PGE will target EV owners who charge their EVs at homes served by the identified feeders to participate in the EV Charging Study. PGE will recruit customers who are already participating in the existing Residential EV Smart Charging Pilot Program, which manages EV charging against EV owners’ TOU rates, in addition to encouraging nonparticipating PGE customers to join both the EV Smart Charging Pilot Program and the EV Charging Study. Customers participating in the EV Charging Study will need to grant PGE permission to manage their EV charging to better understand the distribution impacts of managed and unmanaged EV charging.
In the initial phase of the EV Charging Study, PGE will establish a baseline understanding of EV-specific impact on distribution by monitoring EV charging for approximately eight weeks. This eight-week period will not consider seasonal impacts due to the assumption that driving patterns are the same year-round. Twelve power quality recorders will be installed at predetermined sites within the EV Charging Study boundary on each of the eight feeders to measure and record the impact on feeder operation. These devices will measure power quality parameters such as harmonics, voltage transience, and flicker. The feeders will be monitored in both managed and unmanaged charging scenarios to model the boundaries and operational constraints of EV charging on an individual feeder. Once a baseline understanding of EV charging impacts has been established, the EV Charging Study will transition to its second phase.
In the second phase, PGE will transition to testing real-world use cases to identify how EV charging can be optimized to capture bulk system benefits while still protecting distribution assets. PGE will conduct the testing by inputting various time-series signals into the pilot’s optimization platform. The platform will dispatch the appropriate charging signals to participating vehicles, and PGE’s systems will track the associated impacts on the distribution network. Customers enrolled in the EV Charging Study will be regularly updated via PGE’s appropriate communications and marketing channels. The following use cases will be tested within the EV Charging Study: TOU shifting, bulk system coordination, DER (rooftop solar) integration, distribution feeder loading (asset protection), and power quality support (particularly voltage support) with the goal of determining how best to optimize EV charging for distribution constraints. These use cases are further detailed next.
The TOU optimization strategy focuses on shaping EV load based on a TOU optimization signal: either the default service territory TOU structure or a feeder-specific structure to be developed based on that circuit’s unique characteristics. This strategy will shift EV charging from a period of high demand to a period of low demand (e.g., afternoon load shifted to occur overnight). The format of this use case will include testing the TOU rate used for PGE’s existing Residential EV Charging Pilot Program.
The bulk system coordination strategy focuses on managing EV charging based on wholesale power prices, distribution system capacity constraints, and/or electricity generation attributes (price, fuel type/emission characteristics, etc.). EV charging may be optimized based on a forecast of the wholesale market and/or generation characteristics to achieve various portfolio outcomes, such as minimizing supply costs or minimizing emissions. The EV Charging Study will experiment with different bulk system signals (wholesale price of electricity, forecasted emissions, etc.) and time horizons to understand what factors have the largest impacts on bulk operation.
The DER integration strategy aims to leverage EV charging to address periods of high photovoltaic solar production on the distribution system. If there is enough EV charging on the distribution feeders, PGE will optimize EV charging in areas with excess renewable generation as an alternative to curtailing solar generation to stay within preestablished equipment operational limits.
The distribution feeder loading use case focuses on managing EV charging to control the aggregated loading of a transformer with the goal of maintaining the loading within the acceptable operational tolerance. The total load at the distribution substation transformer can be kept below a predetermined threshold by influencing EV charging load via curtailment, sequencing, or other adjustments to be deployed. This element is a mandatory baseline. Under all use cases, distribution feeder loading will be managed within tolerance.
EV charging systems can negatively impact the power quality of the electrical distribution power system. In this use case, PGE will investigate the impacts that EV charging has on the power quality of its distribution system. It will also evaluate whether EV charging can be optimized for voltage support operations by adjusting charging as needed to maintain steady-state operations at the feeder level.
Once PGE has experimented with the different use cases included in the EV Charging Study, PGE will distill its learnings into a case study on how utility programs focused on EV charging can be holistically designed to address bulk system needs, protect distribution assets, and provide participating customers with a positive experience. As EV adoption grows over the next five to 10 years and results in larger concentrations of EVs, PGE expects that distribution-integrated charging management systems will be imperative to maintaining the reliable operation of the grid. When more EVs come online, defining the optimal time for an EV to charge will become a more complex calculation that depends on a multitude of dynamic factors, such as the following:
The output of this calculation will be specific to each EV and may change day to day for a single EV depending on the conditions during each plug-in event. Today, customers on a simple TOU rate with an overnight off-peak period can easily understand the best time to charge. As EV adoption grows and the calculation for the optimal EV charging time becomes more complex, customers cannot be expected to make that same determination. PGE and its partners will find ways to remove the cognitive burden from EV owners and make it easier for them to charge during optimal times with dynamic system conditions.
As more factors are layered into the optimization calculation and as EV charging programs mature, PGE will maintain a seamless experience for its customers. To accomplish this, PGE is using the EV Charging Study as an opportunity to experiment with optimization design and its impact on the grid while simultaneously building trust with EV owners via the Residential EV Smart Charging Pilot Program. In a future with mass EV adoption, customers will simply plug in their EV when they get home, and PGE will calculate the best time for the vehicle to charge to maximize the benefit to the customer (in the form of electricity cost savings and grid resiliency); capture bulk system benefits, such as lowering the cost of wholesale electricity and integrating renewable energy resources; and maintain grid resiliency by protecting local distribution assets from becoming overloaded.
As residential EV adoption grows rapidly, managing distribution costs is critical for utilities to maintain affordable and reliable power. This challenge represents an opportunity for electric utilities to unlock value in the bulk system without creating unintended impacts on the distribution system. PGE’s Test Bed EV Charging Study aims to determine how to best optimize EV charging for bulk system benefits and distribution constraints. In doing so, PGE leverages intelligent optimization to create more value streams for EVs. Through this study, PGE will gain a better understanding of the impact of EV charging on the grid, learn the best optimization strategies and parameters for their grid topology, and determine strategies to better utilize existing assets and avoid or lessen costly infrastructure upgrades.
“Oregon electric vehicle dashboard,†State of Oregon Data and Reports, Oregon Department of Energy, Salem, OR, USA, 2023. Accessed: Jun. 2, 2023. [Online] . Available: www.oregon.gov/energy/Data-and-Reports/Pages/Oregon-Electric-Vehicle-Dashboard.aspx
“Distribution system planning (DSP),†Oregon Public Utility Commission, Oregon Department of Energy, Salem, OR, USA, 2022. Accessed: Jun. 2, 2023. [Online] . Available: www.oregon.gov/puc/utilities/Pages/EO20-04-UtilityServices-Activities-DSP-Interconnection.aspx
L. Albeck-Ripka, “Amid heat wave, California asks electric vehicle owners to limit charging,†New York Times, Sep. 2022. [Online] . Available: www.nytimes.com/2022/09/01/us/california-heat-wave-flex-alert-ac-ev-charging.html
J. C. Gomez and M. M. Morcos, “Impact of EV battery chargers on the power quality of distribution systems,†IEEE Trans. Power Del., vol. 18, no. 3, pp. 975–981, Jul. 2003, doi: 10.1109/TPWRD.2003.813873.
Matthew Mills is with the Portland General Electric Company, Portland, OR 97204 USA.
Manasseh Obi is with the Portland General Electric Company, Portland, OR 97204 USA.
Kendall Cody is with WeaveGrid, San Francisco, CA 94110 USA.
Kyle Garton is with WeaveGrid, San Francisco, CA 94110 USA.
Amanda Myers Wisser is with WeaveGrid, San Francisco, CA 94110 USA.
Sammy Nabahani is with WeaveGrid, San Francisco, CA 94110 USA.
Digital Object Identifier 10.1109/MPE.2023.3308243
Date of current version: 19 October 2023
1540-7977/23©2023IEEE
