How to choose a site for an EV charging station

Site selection conversations almost always start with footfall. How many cars pass, what is nearby, is it on the highway. Those are reasonable questions and they belong in the process — but they belong last, and putting them first is the most expensive sequencing error in this business.

The logic is simple. Demand analysis is slow and costs real money. Electrical and legal screening are fast and nearly free. If a plot is going to fail, you want it to fail on the cheap test, in week one, not after you have paid for a traffic study.

The four tests, in order

1. Can the site get the power?

   This rejects more plots than anything else, and it is the test people skip because it requires talking to the distribution utility rather than looking at a map.

   What you are establishing: the sanctioned load on the existing connection versus what your intended charger configuration needs; whether a transformer must be added or upgraded, and who pays for it; the distance to the nearest feeder with adequate capacity; and the DISCOM’s schedule of charges for that connection work.

   A plot that is perfect on every other measure and needs a new transformer plus a long cable run can carry a connection cost in the same order as the chargers themselves — and that cost lands entirely on your capex before you have sold a single unit. Find this out first.

2. Can you lawfully build and operate on it?

   Land status, title, and the terms on which you hold it. If it is leased, the lease term needs to be longer than your payback horizon — a five-year recovery on a three-year lease is not a plan. Check permitted use, any statutory approvals your state requires, and setback or access rules where the plot fronts a highway.

   If you are using public land, note that a revenue-share reference of ₹1/kWh applies under the national guidelines — a real cost line that belongs in your model from the start, not a detail to discover later.

   Requirements here are decided locally and vary by state. Our state-wise rules reference sets out what applies where, with the source for each.

3. Does the physical plot work?

   Geometry is not a detail; it determines how many vehicles you can actually serve. What matters: entry and exit that a vehicle can use without a difficult manoeuvre, turning circle for the largest vehicle you intend to serve, space to queue without blocking the road or the site’s other users, bay layout and the cable reach that layout implies, and somewhere to put the equipment that is not in the way.

   The common failure is a plot with adequate total area but the wrong shape, where the usable configuration turns out to be two bays rather than the four the area suggested.

4. Will anyone use it?

   Only now does demand become worth spending on. Who is nearby, what they drive, dwell time at that location, what competing stations already exist and how they are performing, and whether the trip pattern makes your site a natural stop rather than a detour.

   One caution we apply to ourselves as much as to anyone else: published utilisation benchmarks by site type — urban, highway, mall, office, fleet depot — are largely estimates rather than measured data. The ones in our own calculator are internal advisory estimates and are labelled as such in the tool. Validate against operating stations near you before letting any such figure carry your business case.

The mistake underneath most bad sites

Almost every genuinely bad site we see was chosen because it was available — someone owned it, or it was cheap, or a relationship made it easy — and the analysis was then constructed to justify it.

That is a hard habit to notice in yourself, because the analysis is real work and feels like diligence. The tell is the direction of travel: if you started with a plot and went looking for numbers that supported it, rather than starting with criteria and going looking for a plot that met them, you were building a justification.

The defence is to write down your disqualifying thresholds before you evaluate a specific site — the minimum sanctioned load, the minimum lease term, the minimum usable bay count — and then hold to them.

Screen a specific plot

The first three tests are largely deterministic: given coordinates, plot area and your intended configuration, geometry and electrical sizing can be worked out rather than guessed.

That is what GeoSite does — pre-feasibility screening for one specific plot from GPS coordinates and area, covering geometry, electrical sizing, transformer selection and a go/no-go, with confidence-graded evidence behind each finding rather than a single unexplained score.

Sources and basis

Screening is not a substitute for a licensed electrical contractor, a DISCOM sanction, or statutory approval — the requirements that matter most are decided locally. Verify anything you are about to spend money on. Found an error? Tell us.