Site selection for an EV charging hub with many fast chargers
A charging hub needs drivers within a short drive, few fast chargers already close, roads that flow and something to do while charging. How to score candidate sites.
Part of the guide AI employee for site selection
A good site for an EV charging hub, a lot with many fast chargers in one place, sits within a short drive of many people or busy roads, has few fast chargers already nearby, can be reached on roads that are not jammed at the times drivers charge, and offers something to do while a fast charge runs. Score each candidate on those four points with the same data and the same distances, and the shortlist sorts itself.
This article is for charge point operators, fuel retailers adding charging, landlords of retail parks, fleet operators planning a depot hub, and city teams. It is about the hub, a destination with many fast bays, rather than a single charger in a car park. It covers the four tests, a made-up comparison in Toronto, the mistakes that lose money, and an AI employee that scores sites every month.
Test 1: drivers within reach
A hub lives on drivers who can get there quickly. Draw the area within a 10- or 15-minute drive of each candidate with Travel area, which follows the roads rather than a circle. Then count what is inside: in countries with population figures, Demographics gives the residents in that drive area; everywhere, Nearby places can count the offices, apartments, hotels and taxi bases that bring repeat charging.
Test 2: few fast chargers already close
EV charging counts the public charging stations around each place from a worldwide open register: how many stations and charge points are within the distance you pick, how many are fast (50 kW and more), the nearest one and the main operators. Set the power to fast only and the distance to 2 or 3 km for an urban hub, wider on a highway. A candidate with no fast station within 3 km has a gap; one with four fast sites already is a price fight.
Coverage of the register varies by country. When a search finds very few stations in a city area, the step says the coverage may be thin, so check a sample on the ground before you treat an empty map as an opportunity.
Test 3: roads that flow when drivers charge
A hub beside a jammed junction loses drivers who do not want to queue to get in and queue to get out. Traffic Flow gives the live speed, free-flow speed and jam factor on the roads around each candidate. Run it at the hours you expect most charging, such as the evening commute or Saturday morning, for a few weeks. It does not count cars, so it tells you about access, not demand.
Test 4: something to do while charging
Drivers charge where they would stop anyway. A hub beside a supermarket, a gym, a mall or a well-rated cafe turns waiting into errands. Nearby places can list those around each site within 300 m, with their rating, and how long people tend to stay at each (avg_stay_min), which is a useful hint of whether a stop lasts as long as a charge.
A made-up comparison in Toronto
Invented figures, for illustration only. An operator has three candidate lots in Toronto: a retail park by a highway exit, a downtown parking structure and a suburban plaza. For each it measures a 10-minute drive area, residents inside it, fast stations within 3 km, the evening jam factor over three weeks, and amenities within 300 m.
| Candidate | Residents in 10 min drive | Fast stations within 3 km | Evening jam factor | Amenities within 300 m |
|---|---|---|---|---|
| Retail park by highway exit | 210,000 | 1 | 3.1 | Supermarket, cafe, gym |
| Downtown parking structure | 340,000 | 6 | 7.8 | Many |
| Suburban plaza | 150,000 | 0 | 2.4 | Pharmacy, cafe |
Invented figures for illustration; not real data for any site in Toronto.
Downtown has the most people but already six fast sites and jammed evenings. The plaza has no competition but fewer people and little to do. The retail park balances all four tests and goes to a site visit first. Nothing in the table settles grid capacity or the lease, which still decide whether any of them can be built.
Mistakes to avoid
- Counting all chargers as rivals. Slow chargers at a hotel do not compete with a fast hub. Filter the count to fast stations.
- Trusting an empty map. The register is open and grows, but a city with thin coverage can look empty when it is not.
- Reading speed as traffic volume. Free-flowing roads say the site is easy to reach, not that many cars pass. There are no car counts in Traffic Flow.
- Forgetting the grid. Location data cannot tell you whether the site can get the power a hub needs. Ask the utility early.
- Scoring with different distances. Compare every site with the same drive time and the same radius, or the ranking means nothing.
An AI employee that scores candidate sites
- Schedule: on the 1st of every month.
- My data: your candidate sites with positions.
- EV charging: Each incoming place, with counts, fast only, within 3 km.
- Nearby places: Each incoming place, with counts, supermarkets, gyms and cafes within 300 m.
- Traffic Flow: roads within 500 m.
- xMap LLM: score each site from 0 to 100 on your four tests, with a reason.
- Sort and limit, then Email and Show on a map.
Each month you get your candidate sites ranked, each with the fast stations and charge points around it, the nearest fast station and its operator, the amenities within a short walk, the road speed and jam factor at the time of the run, a score and a one-line reason. When a new fast station opens near a candidate, its count rises in the next email, and the score moves with it.
Questions
How close is too close to another fast charging site?
There is no fixed rule. Count the fast stations within the same distance for every candidate, often 2 to 3 km in a city, and compare.
Does Traffic Flow tell me how many cars pass a site?
No. It gives live speeds, the free-flow speed and the jam factor, never car counts. Use it to judge access at busy hours.
Where do the charging stations come from?
From a worldwide open register of public chargers. Coverage varies by country, and the step warns when it looks thin.
Can I see how many people live near each site?
In countries with population data, Demographics counts residents inside a drive area. The United Kingdom and the United States are not covered.
What about power from the grid?
Location data cannot answer that. Grid capacity and connection cost come from the local utility and often decide the final choice.
Templates that do this
Read next
More on site selection and expansion · Every article · Templates