Lewis Hunter, Ryan Sims, Stuart Galloway
Governments across the world are exploring options to transition their population away from internal combustion vehicles toward alternative low- and zero-carbon technologies. For small “light-duty” personal and commercial vehicles, the transition toward battery-powered electric vehicles (EVs) appears to be the candidate solution. In Scotland since 2013, more than £50 million (∼US$63 million) has been invested in a nationwide EV charging network consisting of more than 2,400 charging points. Statistics for the year 2022 indicate that more than 2 million vehicle charging sessions took place on the public charging network in Scotland (not including third-party operators nor private charging points) and delivered approximately 43 GWh of energy to vehicles. As of January 2023, there were 69 public charging points per 100,000 people with 17.3 public rapid charging points per 100,000 people. A rapid charger is a device capable of charging an EV at ≥ 25 kW. Generally, rapid charging points are rated at 50-kW dc and above. Increasingly, the classification “rapid” is being replaced by the term journey charging.
As sales of EVs continue to grow in Scotland, current and future charging infrastructure must provide adequate geographic coverage across the country while also being sized appropriately to service user demand. To facilitate a smooth transition toward a net-zero transport sector, it is necessary for there to be sufficient levels of both public and private charging infrastructure to service the population. Although it is important to make sure that there is adequate provision to support demand, it is also necessary that there is a fully countrywide charging network to create user confidence in nationwide travel. Ensuring provision in rural and islanded communities is an important factor when deploying this national charging network.
To deliver future EV-charging provision, stakeholders must work together to accelerate the rollout of infrastructure in an efficient and cost-effective manner, given the climate emergency declared in Scotland and in other countries around the world. The role of open data published by relevant net-zero-enabling stakeholders (e.g., distribution network companies, governmental transport departments, and so on) is required to derisk infrastructure deployment at an early stage of a project’s lifecycle. Sharing of data will ultimately accelerate the deployment of new low-carbon technologies and avoid unnecessary wasted effort and resources. Additionally, coordinated planning will provide evidence to support ahead-of-need investment cases for power distribution companies to ensure that grid constraints are not seen as a barrier to society’s progress toward net zero.
This article presents the methodology developed to deploy 24 ≥ 50-kW journey charging points across western Scotland as a part of the Facilitating A Sustainable Transition to EVs in the Region (FASTER) project. A key philosophy for developing this approach was that the data underpinning the analysis were open and publicly accessible to allow similar studies to be carried out in the future while also making the process transparent to the public-sector decision makers and the wider population.
As road transportation increasingly decarbonizes via electrification, there will be an increasing requirement for charging stations capable of delivering enough energy to meet societal needs. Distribution companies that manage the “poles and wires,” known in the United Kingdom as distribution network operators (DNOs), are well positioned to help inform low-carbon technology providers as to optimal network connection locations. However, the volume of applications that their connection teams are tasked with evaluating is considerable, and they often struggle to support an “optioneering” process with customers. To help reduce the lead time for connections, it will become increasingly important for connecting customers to try and mitigate against poor sites through analysis before engaging with the distribution company, but the data required to do this are often characterized by the following:
With a first-generation EV charging network already developed across Scotland, the challenge associated with expanding the network with new infrastructure becomes complex. The location of the new journey charging infrastructure needs to
The transition toward net zero will require public and private-sector investment and project delivery, however, a collaborative approach to new infrastructure must be taken to avoid duplication of effort. Open data support electricity network operators and EV-enabling stakeholders to identify candidate locations for future electrified transport infrastructure. The role of open data and decision making will therefore promote
The public EV-charging point rollout in Scotland, for the most part, has been managed by the 32 local authorities and made possible through centralized Scottish Government funding distributed by Transport Scotland, the national transport agency. The charging points deployed through this route are all managed by a single-charge point operator (CPO) called ChargePlace Scotland (CPS). Although this may not be the common model to follow, with many countries opting for private-sector rollout, this centralized approach allows drivers to travel throughout the country using one charging network. All the charging points can be accessed via a single radio-frequency identification card, thus simplifying the experience for EV drivers. A secondary benefit of the approach is the centralized view of infrastructure performance and zero-emission vehicle uptake across the country. Figure 1 highlights all the locations of the publicly owned journey charging infrastructure in Scotland.
To support everyday operation of the network, the publicly owned CPO, CPS acts as the interface between end users and equipment owners, which, in most cases, are local authorities.
The CPO is responsible for
The FASTER project deployed 73 publicly accessible journey charge points throughout western Scotland and the border region between the Republic of Ireland and Northern Ireland. The project is a €6.4 million (∼US$7 million) INTERREG VA project led by East Border Region Ltd in partnership with the University of Strathclyde and the Highlands and Islands Transport Partnership (HITRANS). INTERREG Europe is an interregional cooperation program cofunded by the European Union (EU). HITRANS is one of seven Regional Transport Partnerships in Scotland that were established through the 2005 Transport (Scotland) Act. East Border Region is a local authority-led cross-border organization severing six local authorities along the east coast of Ireland and Northern Ireland. The FASTER project worked with several rural local authorities across western Scotland, as outlined previously in Figure 1, to identify suitable locations for 24 journey charge points, which were installed throughout 2023. More information on FASTER is presented in Table 1.
Table 1. The FASTER project factsheet.
Determining the location for new EV infrastructure is a challenging undertaking for EV-enabling stakeholders and is especially true for public-sector bodies that need to demonstrate that infrastructure is being deployed in a fair, efficient, and just manner. The site identification process developed for the FASTER project uses open and public EV data to underpin the analysis, thus helping support a transparent approach to infrastructure development. A summary of the site identification process developed by the team is outlined in Figure 2 and described in more detail in the subsequent sections. Where possible, infrastructure for the project was to be sited on land owned by local authorities or other public-sector organizations to reduce the number of legal stakeholders and land-sharing agreements required to deliver the portfolio.
Geospatial coverage analysis is used to determine the areas of the road network that are farthest away from the existing journey charging network. Geospatial coverage analysis allows regions that could be considered remote from the charging network to be identified and numerically characterized. The analysis aimed to find the “most remote” charging points in the current network across the study region. Several open and publicly accessible data sources were used to underpin this work, including
For the region, it was determined that a 30-min driving time represented a good starting point for the analysis. This 30-min coverage was selected: 1) as it was deemed to be an acceptable journey time to reach a charging point given the current infrastructure provision in the area and 2) to allow a suitable quantity of gaps in the network coverage to be identified. Note that for more densely populated regions or areas with greater existing infrastructure provision, a reduced driving time may be sought to increase user convenience. To represent driving time on maps, isochrones (iso=equal, chrone=time) were calculated from each existing journey charging point to determine how far a selected vehicle type (in this case, a passenger car) may travel in 30 min using the road network. It is worth mentioning that for this study, driving time was used for the isochrone calculation (commonly referred to as the feature metric). Several valid arguments could be made for using alternative feature metrics such as distance, however, for a rural environment where roads are often single track and slow to traverse, it was deemed that a time-based metric would be more appropriate. This approach was taken primarily due to the inconvenience factor associated with needing to travel farther to charge.
A summary of the developed geospatial process is presented in Figure 3. Note that the study region was expanded by a project-defined time of 15 min to capture the demand served by infrastructure slightly outside the study region.
Figure 4 presents an example of the resulting isochrones for several charge points in an area of the FASTER region. The analysis shows several local authority land assets not covered by the 30-min time isochrone, as indicated inside the dashed circle.
The local authority asset farthest away from the isochrones was identified as the first candidate site for the installation of an EV charger. Recalculating the time isochrones with this new asset installed at this location is outlined in Figure 5, where improved coverage is observed for the region.
Although geospatial coverage could be considered a quantitative problem, the challenge associated with forecasting user demand and behavior, as with any wide-scale population trending, is that analysis is reliant upon many more input assumptions. The concept of demand forecasting was to initially determine the expected energy required to service existing EVs in the area and was based on vehicle registration data, population, and vehicle trends. Based on existing uptake, forecasts of future demand were estimated using EV growth predictions for scenarios published by the electricity system operator (ESO). The ESO is responsible for second-by-second balancing of electricity supply and demand and advising on network investments. In the United Kingdom, the ESO is independent from transmission and distribution network companies. Several publicly available datasets published by various organizations underpin the analysis developed by the team as outlined in Figure 6. The analysis was calculated based on census areas; however, higher- or lower-level analyses (e.g., at a regional or city level) could be considered depending on data availability.
The output of this process was a ranked-ordered list of average daily energy density per square kilometer. The density element of this analysis is important as it captures those living in housing where charging at home is less likely and a greater reliance on public infrastructure is required. Once the daily energy density for each census zone was calculated, the energy served by existing charging infrastructure was subtracted from the calculated value to determine how much additional infrastructure is required to meet demand.
Figure 7 presents the demand analysis results in the form of a heat map for the three local authorities with Figure 8 highlighting the energy densities calculated for the urban area outlined in the dashed area in Figure 7.
The demand analysis helps identify areas where high levels of charging events on public infrastructure are to be expected. In reality, the use of EV hubs in these areas may prove to be the most cost-effective solution going forward.
Although the previous two methodologies aim to determine new sites suited for charging infrastructure, it is important to assess the performance and capacity factors of existing sites to determine where supplementary infrastructure is required to minimize the likelihood of users having to queue at charging points. The data made available by the CPO allowed for the analysis of charging point capacity factors as well as for determining the time between sessions. The time between sessions was categorized into four data bins and analyzed for different times of the day as well as between months of the year to capture seasonal demands. The data points were classified using the following criteria:
These classification times were tuned to the area covered by the FASTER project. If engineers are looking to replicate this approach, they may need to tune the time intervals accordingly. The probability distribution of the dataset helps to provide evidence as to popular charging point locations where the likelihood of needing to wait for a previous user to complete a charging session is high. A high probability of users waiting to charge suggests that additional infrastructure may be required to support EV growth in the area.
An interesting observation from the queuing analysis was the impact that local ferry arrival and departure times had on the likelihood of back-to-back charging events taking place. This observation was apparent across much of the FASTER region, where ferry services that carry vehicles represent lifeline connections from the islands and remote communities to the mainland. Considering ferry operation when planning EV infrastructure leads to interesting engineering tradeoffs for these sites, for example, between the following criteria:
It is worth highlighting that due to the ownership and operating model adopted at present in Scotland, applying this queueing analysis to a portfolio of charging points operated by a single CPO is much simpler compared to a region where several CPOs operate. It is appreciated that this approach may be more challenging to implement when considering several commercial operators.
Engineering methodologies underpinned by open data provide a scientifically rigorous and repeatable approach to deploying EV infrastructure to meet an objective or series of objectives. These approaches can guide enabling stakeholders as to the approximate location for infrastructure to service a customer base; however, the final location of infrastructure is often influenced by the following nontechnical factors:
Before conducting power capacity studies, it was deemed important that appropriate officers in local authorities and wider-area transport partnerships were able to discuss the candidate sites identified through the processes outlined. Local insight allowed knowledge relating to the site and its surroundings to be considered while also offering the opportunity to introduce nearby alternatives that were still supported by the geospatial, demand, and queuing analyses.
This process produced a list of candidate sites to be advanced to the detailed power capacity analysis phase of the process.
The “short list” of candidate locations was assessed from a power capacity perspective to determine whether sites could host a three-phase connection suitable for 50-kW charging equipment. Note that 50-kW chargers were selected due to project funding requirements. The installation of multiple chargers at a site was considered where an evidence-based need for the infrastructure existed. A power capacity analysis was carried out by cross referencing several of the following publicly available documents produced by the distribution company:
A transformer loading database for 11-kV and LV networks
An 11-kV schematic
Generation and demand heatmaps.
An example extract from the assessment tool developed by the authors is presented in Figure 9. Power capacity was assessed for each of the candidate sites using the process summarized in Figure 10.
Combining GIS shapefiles for the network area with a publicly available asset rating and loading library allowed load flow and voltage drop calculations to be carried out using OpenDSS, an open source power system simulation platform. Note that in most cases, the information held in the datasets outlined in Figure 10 was sufficient to build models suitable for balanced three-phase power flows to be conducted. For areas with insufficient information, engineering judgment was applied. A capacity analysis was conducted for the following network components for each of the sites in the short list:
To help disseminate the findings from the power capacity studies, a “traffic light” notation was used with colors scaled against the project’s site budget. Based on the allocated funding for the project, an average budget of ~£46k had to be achieved per site for the portfolio of 24 journey chargers. The following classifications were applied to each “traffic light” category:
These categories were predominantly based on the anticipated reinforcement costs required to upgrade the distribution network in the area. Using cost data published by the distribution company, it was possible to determine ahead of time the extent of the reinforcement required. Note that due to the rural nature of much of the project, upgrading transformer capacity was required for several sites.
During the power capacity assessment process, a notable limitation to the placement of high-powered EV-charging infrastructure was determined and related to the lack of threephase electrical systems in many rural communities as the historic preference of the distribution company was to service these customers with a single-phase system. The lack of three-phase electricity distribution is a considerable barrier to net zero for many rural communities in Scotland. Although single-phase to three-phase conversion systems do exist, they introduce additional complexity and costs to remote EVcharging sites and will introduce additional points of failure in the system. Ultimately, in the context of the FASTER project, it was determined that these systems did not represent a costeffective solution while considering the wider project budget. Regulatory changes that govern the cost of new connections to the distribution network were introduced in Great Britain in April 2023. The aim of these changes is to help socialize grid reinforcement costs and should promote more cost-effective connections for net-zero-enabling infrastructure going forward. Unfortunately, the impact of these changes was not felt by the FASTER project as the project had already committed to sites before the April 2023 implementation.
The following nonelectrical, additional site checks were carried out before final site approval:
Although flood risk and cellular coverage could be checked via online services, additional confirmatory checks were conducted during site visits.
Due to the geographical coastal and rural regions covered by the FASTER project in Scotland, sites needed to be assessed for flood risks to ensure that charge point warranties would not be voided because of locating equipment in an area with a likelihood of flooding. Online flood risk maps were available to access through the environment agency and could be integrated into the GIS modeling system used in the methodologies outlined previously. Visual checks were carried out during the physical site survey to ensure that there was no evidence of localized water pooling.
To communicate with the CPO’s back-office software, it is key that reliable communication systems are in place and readily accessed when installing EV-charging equipment. Reliable communications to the CPO allow for charging point telemetry and billing to be processed and include the ability for the CPO to mark charging units as being defective on EV-charging point maps. Reliable communication between charging points and the CPO is particularly important in rural environments because the probability of EV drivers not having a sufficient range to reach the next available charging point in the event of equipment failure increases as assets are more geographically dispersed. A secondary benefit of reliable communications to remote charging sites allows charging equipment to be interrogated remotely to pre-emptively react and respond to equipment failures. Increasing reliability, or lack thereof, of charging infrastructure is gaining press and political attention. The telecommunications regulator for the United Kingdom, Ofcom, publishes an open coverage map and associated application programming interface on its website detailing coverage for the major mobile telecommunication providers in the country.
Although online mapping and street-level imagery tools allow a portfolio of sites to be evaluated quickly from a desktop, site visits are still an essential component of the site-selection process. Online imagery is usually time stamped to determine the date of capture. Due to the remote nature of many of the sites selected in the FASTER project, some imagery was more than 13 years old, thus increasing the likelihood of unknown changes for an area. During desktop studies, it is also recommended that online satellite and street-level imagery be compared across several data providers as there are often quality and detail differences among services.
When on site, the following measurements and information were collected:
These elements should be recorded in a site survey report. Within the site report, a layout diagram for the proposed infrastructure was also included. An example site layout diagram developed for the project is outlined in Figure 11 for a site with two 50-kW charging points. These bays were designed to accommodate long-wheelbase commercial vehicles while providing step-free access for users with mobility challenges.
The process of developing a large portfolio of sites through one funding mechanism allowed and justified the development of the processes outlined in the previous sections to help promote a data-first approach to site selection. The key learnings and findings discovered through the project are summarized in the following sections, focusing on site design, minor site relocation, regulatory frameworks, “hidden charges,” and the provision and type of data.
Since project conception, the thinking related to accessibility and the built environment adjacent to EV-charging points has progressed significantly. Best-practice guidance for developing accessible EV infrastructure may include consideration of the following factors:
Charging point screen height and charge ergonomics
Communicating power network opportunities and constraints to nontechnical stakeholders is important to help demonstrate why grid connection costs between locations can be highly variable. In many cases, the biggest variable between the delivery costs of charging infrastructure between sites is the underlying connection to the power network. To help communicate these challenges to stakeholders, GIS extracts were provided alongside “traffic light” power capacity assessments for different network elements to help demonstrate where reinforcements may be required and where candidate alternative sites existed.
Figure 12 provides an example where the local authority was considering hosting a charging point at “Site A” due to the proximity to a ferry service. The analysis identified that the following significant reinforcements were required to support this connection:
“Site B” was proposed as an alternative as less extensive network upgrades were anticipated. Note that “Site B” is within close driving range (<2 min) to “Site A.” The budget quotations received from the DNO indicated that this recommendation would save approximately £35k (~US$44k), noting that the average site budget, including the purchase of the journey EV charger (~£25k), was approximately £46k.
With the distribution company for the Scotland FASTER region requiring up to 65 working days to process formal quotations, it was important to derisk connections as far as possible before making a formal connection application. This approach worked well for most of the sites, however, some network connection charges only became visible to the team after receiving the returned quotation for the works: notably, “second-comer charges.”
Second-comer charges were designed to try to ensure that the cost of connecting to the electricity distribution network is shared more fairly between different parties. In Great Britain, any connectee who subsequently connects (i.e., the second comer) to and benefits from infrastructure paid for by an earlier connectee is liable to pay for his or her share of incurred costs under electricity connection charging regulations.
In one example, a second-comer charge of approximately £75k was returned through the formal application connection process. With an average site budget of ~£46k (including the charge point), these unexpected network charges can be a considerable barrier to deploying infrastructure.
Second-comer charges were introduced to the market in 2002, however, in the context of the climate emergency and the transition to net zero, the energy regulator in Great Britain conducted a “significant code review” into distribution network connections. This code review recommended that network upgrades are socialized with a “high-cost cap” to protect consumers from connections that will be too costly to deliver. Unfortunately, the implementation date for the code review did not align with the timescales associated with the FASTER project; however, the sites identified that would benefit from these regulatory changes were listed and passed on to local authorities for consideration at a future date.
Charging infrastructure provision across a country is a constantly changing environment. New charging point installations may be commissioned, while other sites may be out of service or removed from service. The methodologies that assess the demand for charging infrastructure need to be dynamic to allow the ever-changing charging network to be considered in site-identification processes.
Also, a risk exists that several public and/or private entities could unknowingly be competing to secure grid connection capacity for EV-charging infrastructure at the same substation in parallel.
Distribution companies are well positioned to understand the network connection activity in different regions of their operation areas; however, there are challenges with sharing this activity meaningfully with third-party stakeholders. Innovation is likely required here to help pinpoint areas of high connection interest.
Open data and transparent engineering processes underpinning the FASTER methodology have allowed the creation of a process that helps local authorities, transport partnerships, and other EV-enabling stakeholders to understand the steps involved to deliver the infrastructure associated with the project. The availability of power network shapefiles combined with schematics and demand information allows sites to be derisked before formal connection applications, saving time for project stakeholders as well as reducing the burden on connection teams working for distribution companies.
It should be noted that the regulatory arrangements of different countries and regions around the world may expose or restrict different datasets for a variety of reasons (e.g., market structure, security concerns, and so on).
The FASTER process has successfully identified candidate sites for 24 journey EV-charging points that are deliverable both within the project’s budget and timescale. The process designed used open and/or public datasets ranging from tourism statistics to vehicle registration data to underpin the analysis. Summary engineering process diagrams were presented throughout this article to demonstrate the data transformations designed to estimate the requirement for EV-charging points in an area.
To facilitate the transition toward net zero, the role and requirements of open energy data will likely be increased. Collaboration among stakeholders and sharing of data helps to avoid duplication of effort and, therefore, will accelerate the transition toward net zero. With the increasing number of low-carbon technologies looking to connect to distribution networks, automated “self-serve” connection tools informing potential customers as to the cost of connecting at a location are becoming increasingly common. Self-serve tools provide an immediate estimate of the cost to connect to the network while also helping to manage the workload of connection teams within distribution companies, given the increase in activity expected as the transition toward net zero accelerates. Self-serve tools are only valuable if the underpinning data are accurate and updated regularly. The use of confidence intervals could be a useful addition to the output of tools and may help manage the expectations of customers at locations where multiple projects are competing for grid access. Until a formal time-bound quotation is received by infrastructure installers, self-serve budgets should only be considered for initial site validation.
Although second-comer charges prevented a small number of sites from moving forward to construction, the remaining sites identified through the process returned costs that were within the project’s budget allocation.
As the world transitions toward net zero and, in the context of the climate emergency, the role of open data and transparent engineering processes will help support engineers as they develop an EV-charging network that is fit for the future. Although much of the existing charging network has been established in an ad hoc manner, the challenge associated with further decarbonization of road transport will likely require an increased focus from all stakeholders to ensure the delivery of a charging network that has the following characteristics:
Open and transparent data and processes are one means to help facilitate the transition.
This work was supported by the EU’s INTERREG VA program and managed by the Special EU Programmes Body.
“A network fit for the future: Draft vision for Scotland’s public electric vehicle charging network,” Transport Scotland, Glasgow, Scotland, Jan. 2022. [Online] . Available: https://www.transport.gov.scot/publication/a-network-fit-for-the-future-draft-vision-for-scotland-s-public-electric-vehicle-charging-network/
“Electric vehicle charging infrastructure report,” Transport Scotland, Scottish Futures Trust, Glasgow, Scotland, Jul. 2021. [Online] . Available: https://www.transport.gov.scot/publication/report-on-public-electric-vehicle-ev-infrastructure-in-scotland-opportunities-for-growth/
“Electric vehicle infrastructure guide,” Energy Saving Trust, Transport Scotland, Glasgow, Scotland, Jul. 2022. [Online] . Available: https://evinfrastructureguide.com/
“Common requirements and good practice for the charge place Scotland network,” Transport Scotland, Glasgow, Scotland, 2019. [Online] . Available: https://www.transport.gov.scot/media/45830/common-requirements-and-good-practice-for-chargeplace-scotland-network-switched-on-towns-and-cities.pdf
L. Hunter and R. Sims, “FASTER: A site-selection methodology for journey EV chargers,” in Proc. CIRED Porto Workshop, E-Mobility Power Distribution Syst., Hybrid Conf., 2022, pp. 369–373, doi: 10.1049/icp.2022.0729.
Lewis Hunter is with the Power Networks Demonstration Centre, the University of Strathclyde, G68 0EF Glasgow, Scotland.
Ryan Sims is with the Power Networks Demonstration Centre, the University of Strathclyde, G68 0EF Glasgow, Scotland.
Stuart Galloway is with the Department of Electronic and Electrical Engineering, University of Strathclyde, G1 1XW Glasgow, Scotland
Digital Object Identifier 10.1109/MPE.2023.3308221
Date of current version: 19 October 2023
1540-7977/23©2023IEEE