EV Charging Station ROI: How to Model Payback for a CPO Network

Payback on a charging station swings from 2 years to 12 on the same hardware. Utilization is the variable that decides it, and it is the one most business cases guess at.

The same 120 kW charger can pay for itself in two years or never. The hardware is identical, the difference sits almost entirely in one input: how much energy each gun delivers per day. Most business cases that fail do not fail on capital cost. They fail because utilization was assumed rather than measured.

This is a working model rather than a forecast. Build it in a spreadsheet, change the inputs to match your market, and treat the output as a range. Every figure below is typical for a European or North American public site.

What Payback Measures, and What It Misses

Simple payback is capital expenditure divided by annual net cash margin. It ignores the cost of capital, residual value, and everything after the payback date. It is still the right first filter, because it is fast and it exposes the assumptions that matter. If simple payback exceeds seven or eight years, a discounted model will not rescue the project.

Capital Expenditure per Charger

Capital cost is the easiest input to get right, and most operators underestimate it because they count the charger and forget the site. A 120 kW dual gun DC unit delivered to site typically lands between 22,000 and 36,000 dollars. The DC fast charging station class is the usual starting point for a public site.

• Hardware, freight, and duty: the visible line, and often only half the total
• Civil works, foundations, trenching, cabling, and protective devices
• Grid connection or transformer upgrade, allocated across the chargers on site
• Backend setup and payment terminal integration, a one off cost per site

For AC destination charging the picture is different, because the hardware is cheaper and the site work dominates. A wall mounted AC charger at 7 to 22 kW costs a fraction of a DC unit, but the electrician, the cabling, and the parking bay work do not scale down with it. Our deployment cost breakdown covers those lines.

Utilization: The Variable That Dominates

Utilization is measured in kWh delivered per gun per day, not in sessions, because a 30 minute session delivering 20 kWh and a two hour session delivering 40 kWh are not comparable. On a public DC site, 60 to 120 kWh per gun per day is a realistic range once a site matures. Busy corridor and fleet sites run 200 to 400.

What drives it is location and dwell time. A retail site with 40 minute average dwell captures charging that a highway site with 20 minute dwell cannot, and a fleet depot with fixed overnight routes carries a predictable load. Utilization also grows: most sites take 12 to 24 months to mature, so a model built on month one data will understate reality.

Energy Margin: Tariff Against Retail Price

The margin between what you pay for electricity and what you charge the driver is the engine of the model. Tariffs vary widely by market and contract, and demand charges add a fixed monthly cost that is easy to omit. A spread of 0.20 to 0.35 dollars per kWh between cost and retail price is typical.

Two things erode that spread. First, energy prices move: if your retail price is fixed and your cost is indexed, the margin compresses without warning. Second, demand charges scale with peak power rather than energy delivered, so a site running two short high power sessions at once can pay the same demand charge as a site delivering three times the energy.

Host Fees, Payment Fees, and Backend Costs

These line items are small individually and material together. They typically remove 20 to 30 percent of gross energy margin on a host site, the difference between a four and a six year payback.

• Host or land fee, commonly 10 to 20 percent of gross revenue at retail and hospitality sites
• Payment processing, roughly 1.5 to 3 percent of transaction value
• Charging network backend and OCPP platform, often a per charger monthly fee plus roaming fees
• Roaming and e-mobility service provider commissions, which can take another 5 to 15 percent

Maintenance and Downtime Reserve

Budget 2 to 4 percent of hardware capital expenditure per year for maintenance across a mixed fleet. That covers preventive service, filters and fans, cable and connector replacement, and the occasional power module. It does not cover revenue lost while a unit is down.

Availability is a financial input, not just an engineering metric. A site running at 95 percent availability against 99 percent loses roughly 15 days of revenue a year. On a charger earning 30 dollars a day in net margin, that is around 450 dollars, often more than the annual maintenance reserve.

A Worked Payback Model

The table below models one 120 kW dual gun charger on a retail site. It assumes a retail price of 0.45 dollars per kWh, an energy cost of 0.20 dollars per kWh, a 15 percent host fee, 2.5 percent payment processing, a 90 dollar monthly backend fee, and a 3 percent annual maintenance reserve. Tax, financing, and demand charges are excluded.

Sensitivity: Utilization Against Payback

Change one input and hold everything else. The table below shows why utilization deserves more diligence than any other number, and why a conservative assumption is correct without traffic history.

Two conclusions follow. Below roughly 80 kWh per gun per day, a 55,000 dollar charger is a poor investment on energy margin alone, and the site needs another revenue reason to exist. Above 180 kWh per gun per day, payback compresses fast enough that a second unit at the same site is usually the better use of capital than a new site.

Hardware price matters less than most buyers expect. Cutting 20 percent from the hardware line lowers total capital by roughly 11 percent, moving payback from 4.5 to 4.0 years. Doubling utilization does far more. Our DC fast charger price guide covers where hardware cost sits.

When the Model Says No

A payback longer than seven years is a signal, not necessarily a rejection. It usually means one of three things: the site is early, the tariff is wrong for the load profile, or the capital scope includes something that other chargers would also use.

The fix is rarely a cheaper charger. It is usually a better tariff, a higher margin mix such as fleet contracts, or a staged build where the civils and grid connection are sized for four bays and only two are equipped at first. Staging reduces payback risk more than anything else, because the second pair of chargers costs less and earns on proven demand.