Commercial EV Charging Infrastructure: The Complete Buyer's Guide

SINELINK / EV CHARGING GUIDE

SUMMARY: Plan commercial EV charging around vehicle dwell time, available grid capacity and long-term operating needs. This guide explains AC and DC selection, charger architecture, load balancing, project costs, utilisation, market compliance and service contracts, helping buyers turn site requirements into a clear specification before ordering.

Who this guide is for

Charge point operators (CPOs) building or extending public networks, particularly those moving into a country whose standards differ from the ones they already deploy.

Fleet operators converting depots, where the constraint is usually the overnight window rather than the charger.

Property owners, retailers, hotels and workplace operators adding charging as an amenity or a revenue line, where dwell time - not charging speed - is the governing factor.

EPC firms and consultants who need a defensible basis for a specification they will later be asked to justify.

If you read nothing else

Match charger power to dwell time, not to ambition. Most overrated sites end up delivering a fraction of their rated output for reasons explained in H2-3.

Grid capacity is decided before the hardware is. Whatever the site's service capacity is, that number - not the charger nameplate - caps your revenue.

Rated kilowatts are not delivered kilowatts. Output is bounded by whichever limit is hit first: cabinet power, connector current, battery voltage, vehicle request, or thermal state.

Hardware is usually the smaller part of a project's capital cost. Civil works, grid upgrade, software subscriptions and payment rails are where budgets break.

Utilisation, not tariff, decides payback. Below roughly mid-single-digit utilisation, a DC fast site struggles to make commercial sense on charging revenue alone.

Compliance is a purchase-time problem, not an import-time one. A unit that cannot meet the rules of the market you sell into has no resale value, whatever its price.

The five decisions that shape every charging project

Every charging project, whatever the site type, resolves into five questions. They are sequential: a wrong answer to an early one cannot be corrected by a good answer to a later one.

A note on sequencing. Most projects start at decision 2 because that is where the product catalogues start. Starting at decision 1 costs a week and routinely saves six figures; it also tells you whether the site is viable at all before anyone commits to a layout.

To read each of these decisions in depth, here is where each one lives: power and sizing , architecture , module-level serviceability , cost , utilisation , grid timeline , compliance , certificates , contract terms .

Technology basics you must know before talking to suppliers

You do not need to be an engineer to run this project, but you need enough vocabulary that a supplier cannot hide a gap behind a specification sheet. Four topics cover it.

Charging levels: Level 1, Level 2 and DC fast charging

Charging is conventionally described in levels, and the levels describe where the conversion from AC to DC happens and how much power is available.

The part buyers most often miss is that Level 2 power is set by the vehicle's on-board charger, not by the wallbox. Installing a 22 kW three-phase unit on a fleet whose vehicles only accept 7 kW single-phase buys you nothing. This is why our guide to charging levels leads with how buyers mix up upstream and downstream limits .

Where AC charging genuinely wins is cost per installed bay and simplicity of the grid connection. Our view is stated plainly in the comparison: many sites have no business installing fast chargers at all.

Connectors by region: what to order for each market

Connector choice is not a preference; it is set by the vehicle population at the destination.

Ordering the wrong connector type is one of very few decisions in this list that cannot be rescued in the field. Our regional breakdown covers the per-market decision in detail .

Communication protocols: OCPP, ISO 15118 and why the version is a hardware question

Two things get confused constantly, and conflating them costs weeks:

OCPP (Open Charge Point Protocol) runs between charger and back-office - sessions, authorisation, status, firmware. Version 1.6J is the long-standing baseline; 2.0.1 adds richer device management, stronger security and better support for smart charging.

ISO 15118 runs between charger and vehicle - notably Plug & Charge. Edition 2 covered the first practical automated identification; edition 2018/20 broadened it with newer capabilities including bidirectional operation.

The operational point: moving from OCPP 1.6J to 2.0.1 is not guaranteed to be possible later. Whether a unit can support 2.0.1 depends on its compute platform and firmware architecture, decided at manufacture. If your destination market or your network operator will require advanced capability - and European procurement increasingly does - verify it before the purchase order, not after the container ships. Our existing deep dive on the communication layers inside a charging project sets out where each layer sits; the version-choice question that complements it is here .

Smart charging - scheduling sessions to respect site or grid limits - has moved from a feature to an expectation in several markets, especially where grid connection capacity is constrained.

AC or DC: choose by dwell time, not by specification sheet

The honest test is simple: how long does a vehicle stay parked, and how much energy does it need during that stay?

If the answer is "several hours" - hotel guests overnight, employees during a shift, fleet vehicles on a depot apron until morning - AC Level 2 almost always wins. It is far cheaper per installed position, lighter on the grid connection, and simpler to maintain. Our H Series smart AC wallbox exists for exactly this: destination AC with Type 2 interface, certified to IEC 61851-1 and carrying EU radio-equipment conformity assessment, and we say plainly where it is *not* certified rather than leaving buyers to assume.

If the answer is "twenty to forty minutes," DC is the only answer - see the DC charging range - and for this you want high power and a site configuration that lets multiple bays share it.

A useful hybrid - and one that most fleets arrive at eventually - is mostly AC with one or two DC units covering exceptional cases. That pattern is the one we recommend most often, and it is argued out in full in the AC-vs-DC comparison .

Choosing the right hardware architecture

This section is where most of the money is decided, and it is also where the gap between what buyers think they bought and what they can actually deliver is widest.

Sizing: why the bigger charger is usually the wrong answer

Here is a concrete and frequently misread case. Our 60–120 kW integrated DC station is sold in three ratings sharing a platform. All three carry a single-connector current limit of 250 A. That limit interacts with battery voltage:

At the extreme: at 200 V, a 250 A limit delivers around 50 kW regardless of what the cabinet says. A 120 kW cabinet and a 60 kW cabinet look identical to that vehicle. This is not a defect; it is physics, and it is disclosed on the page as a power envelope because it should be. View the range here: 60–120 kW integrated DC charging stations.

The buyer lesson: a higher nameplate does not buy you throughput unless the vehicles that will use it can accept it, and the grid can feed it. Size from dwell time and expected vehicle mix. Full method .

Integrated or split: decide by what you think will happen in year three

An integrated charger puts conversion and the user interface in one cabinet. Simple to install, simple to permit, easy to understand. Its limit: power is effectively locked into the cabinet, so growing throughput usually means another full unit and another grid connection point.

A split, or distributed, system separates a central power cabinet from the dispensing terminals. It costs more to install and takes more design work, but capacity becomes a shared pool: terminals are added later without touching upstream capacity, and importantly, one power cabinet can back up others.

Our 1440 kW flexible shared supercharging system is the extreme end of that trade - one power source serving multiple bays with allocation changing minute to minute, aimed at locations whose demand is both high and unpredictable. Browse the split-system options here: distributed DC fast charging systems. The full argument, including why we think many buyers should *not* choose split, is here .

Power modules: the component that decides whether the site is open tomorrow

Commercial perception of network quality is driven by whether the charger works when the driver arrives. Contributor number one is how the unit behaves when a component fails.

Modern high-power chargers build their output from repeating power modules rather than one monolithic conversion stage. Two consequences follow that most buyers never ask about:

Our SER1000 series covers 20 kW, 30 kW and 40 kW modules in IP55 / IK10 enclosures, and the 40 kW module is the building block used in our own 240–400 kW unit. Review the modules: EV charger power modules. Everything this means for availability budgets is written up here .

Two questions to put to any supplier on this topic: *what happens to output when one module fails*, and *how many minutes does a swap take, and who performs it*.

Dynamic load balancing: more chargers on the connection you already have

Load management is the single highest-leverage lever for sites where grid capacity, not demand, is the constraint.

The mechanism: instead of sizing the connection for every charger running simultaneously at full output, the site allocates available power dynamically across active sessions as vehicles come and go. Because vehicles rarely all draw peak at once, and can be told to throttle, the same site performance is achieved on a much smaller connection - avoiding an upgrade whose cost often exceeds that of the chargers, and reducing standing capacity charges thereafter.

The most commonated form: real-time allocation across a group of chargers, prioritising by session need or by pre-set business rules. Our own work on this covers overlap with site PV and storage, where the same controller arbitrates between grid supply, local generation and battery discharge - see the deployment write-up here: intelligent charger group control.

Our Dynamic Load Balancing Controller handles this without data cabling to every unit: it links wirelessly. Two practical cautions, stated because they are true rather than because they flatter us - (1) metal enclosures, walls and nearby electrical equipment can degrade a wireless link, so site layout should be checked during design, not on installation day; and (2) the achievable reduction depends on how many units you run and how your vehicles behave, so it must be calculated per project, not estimated from a brochure. Full method .

What it actually costs - and what decides whether it pays back

Every cost line buyers forget

Hardware quotations are the visible part. The pattern of project overruns is remarkably consistent: the cost lines nobody priced at quotation stage.

The recurring finding across the industry - and the reason we discuss this before hardware at all - is that hardware frequently represents less than half the installed cost of a fast-charging site. Exact proportions are site-specific; get your own figure before you approve a budget. Item-by-item detail .

Utilisation: the threshold that decides payback

Payback is dominated by utilisation far more than by the electricity tariff. Below is transparent arithmetic - insert your own capital cost and margin rather than trusting ours.

Assume a 120 kW unit. Energy delivered per year = rating × 8,760 h × utilisation:

To turn that into money, multiply by your gross margin per kWh - what you sell energy for, less what you pay for it and less payment processing. Simple payback is then installed cost divided by annual gross margin, before operating cost, demand charges and maintenance.

Worked with illustrative inputs only - replace them - installed cost of 700 currency units per kW (84,000 for a 120 kW unit) and 0.12 per kWh margin:

Two caveats matter more than the numbers themselves. First, these figures exclude operating cost, demand charges, payment fees, and maintenance - all of which lengthen payback, sometimes drastically. Second, the direction of the result, which is the part that generalises: below roughly mid-single-digit utilisation, commercial DC fast charging struggles on charging revenue alone, and needs another reason to exist - a regulatory obligation, a loyalty effect, a lease requirement, or fleet necessity. Model it in full .

Demand charges: the line that quietly kills margins

Many commercial electricity tariffs bill not only for energy consumed but also for the highest rate of consumption in the billing period - commonly the peak power drawn in any interval. A single short peak sets that charge for the whole month.

Illustration, again with illustrative inputs: a 350 kW monthly peak at 15 per kW lands around 5,250 per month in demand charges alone, *before* a single kilowatt-hour is counted. Annualised, that is a fixed cost that must be covered before anyone talks about profit - and unlike energy cost, it does not fall when utilisation falls.

This is why peak-shaving enters conversations about fast charging: adding storage, or simply controlling the site's aggregate draw, caps the peak and converts a lumpy fixed cost into a managed one. Where a site has roof or land available, the solar-storage-charging configuration becomes genuinely economic rather than merely attractive - see our system write-up here: integrated solar-storage-charging stations. The economics are worked through .

Grid connection and site selection: the part measured in months

Two realities buyers consistently underestimate.

First, connecting significant new load is not a form-filling exercise. There is an application, a utility study, often reinforcement works, then installation and energisation - frequently far longer than the hardware lead time, so orders and designs can end up waiting. Second, site qualities that look fine on a plan - available capacity, cable route, transformer condition - determine feasibility, and the least flexible of these is usually the existing supply, not the land.

Practical guidance: treat grid connection as the project's critical path from day one, start the utility conversation before committing to a layout, and never sign a lease assuming a given upgrade a landlord informally suggests may be possible. Stage-by-stage, with who owns what .

Compliance checklist by destination market

This is the section buyers most frequently come back to, and it should be treated as a checklist rather than read once.

The European Union

AFIR shifts compliance from a shipping question to a specification question. Public tenders and CPO procurement now routinely require capabilities that cannot be added later - buying hardware that will not satisfy them has no resale value. Our European-standard SEC units ship with CE and TÜV certification - see the SEC 60 / 80 kW European DC charger. Timeline detail .

North America

Our North American equipment ships with this stack behind it, including Energy Star where applicable - see the SEC 160 kW North American standard unit. Full verification method .

Two cross-market traps

Trap one: assuming OCPP 1.6J will satisfy tomorrow's buyer. Some markets and tenders now expect more modern capabilities; whether a unit can reach them is a hardware question. Confirm against the Open Charge Alliance certificate list before you commit.

Trap two: assuming a test report exists because a certificate exists. Certificates reference specific models and specific editions. Ask for the report, check the model number on it against the model you are buying, and file it - this is the document an inspector or insurer will ask for first.

Requirements to write into your purchase contract

What follows protects you after installation, when leverage is at its lowest.

Availability, quantified

Uptime percentages look similar on paper and are very different in operation. Assume a 100-position site:

The three-point difference between 96% and 99% removes roughly 26,000 position-hours of availability a year - before counting the driver who arrives during an outage and does not return. Translate committed availability into lost energy opportunity using the arithmetic in H2-4 and the number stops being abstract.

The clauses themselves

Write the availability number as the metric you will actually be able to see. A supplier who will not commit to how availability will be measured has told you something. Detail plus drafting language .

Frequently asked questions