Undersized DC fast charging sites fail in a very specific way: not by breaking down, but by making drivers wait so long during peak hours that they stop coming back. Oversized sites fail more quietly, through capital sitting idle most of the day on hardware that never needed to be that large. Both failures share the same root cause: sizing a site around average daily demand instead of peak concurrency, which is the only number that actually determines whether a driver gets a charger in under five minutes or circles the lot during the evening rush.
Why Average Demand Is The Wrong Number
Average demand tells you how much energy a site delivers across a full day, which is useful for revenue forecasting but close to useless for sizing hardware. A highway plaza might average four vehicles per hour across 24 hours, a number that suggests a single charger would suffice, while actually seeing twelve vehicles arrive within a single evening peak hour as weekend travellers pass through. Size for the average and that site queues badly every Friday evening even though its daily utilisation numbers look healthy on paper. Peak concurrency, not average throughput, is what determines whether drivers experience your site as reliable or as somewhere to avoid.
Modelling Peak Concurrency For Your Site Type
Different site types have fundamentally different peak shapes. Highway corridors see sharp peaks around holiday weekends and typical commute hours, with demand that can be four to six times the daily average during the worst sustained hour. Cab and fleet hubs see more predictable peaks tied to shift changes, often twice daily, where every vehicle needs to turn around within a tight window. Public charge-point-operator sites in dense urban areas tend to have flatter demand but higher absolute volume throughout the day. Model your specific site type against real traffic or fleet schedule data, not generic industry benchmarks, since the gap between a highway plaza's peak and a fleet depot's peak is wide enough to produce completely different hardware specifications from the same average-demand number.
Choosing Power Output Within The Tejas Range
Once peak concurrency is known, power output per port follows from the vehicle mix you expect and how much dwell time you can tolerate during that peak. The Tejas Series scales from 30 KW, suitable for lower-traffic sites and fleet vehicles with smaller batteries, up to 240 KW for sites where minimising dwell time per vehicle matters most, such as high-traffic highway corridors serving a mix of passenger and commercial EVs. Specifying higher output than your typical vehicle mix can actually use is wasted capital, since most current passenger EVs cannot accept charge rates anywhere near 240 KW; matching port power to realistic vehicle acceptance rates, not just maximum theoretical demand, keeps the specification grounded in what will actually be used.
Dual-Port Designs And Installed Cost Per Bay
Dual-port cabinet designs change the sizing math by letting two vehicles share the power electronics of a single unit, splitting output dynamically between them when both are charging. This lowers the installed cost per charging bay significantly compared to two fully independent single-port units, since civil work, the grid connection and much of the cabinet hardware is shared. The tradeoff is that both vehicles share total available power when charging simultaneously, which is a reasonable compromise at sites where full-power single-vehicle sessions are less common than moderate, overlapping sessions, but a poor one at sites where every vehicle genuinely needs maximum power without compromise.
Grid Capacity As The Real Constraint
On many sites, grid capacity decides the achievable specification well before cost or demand modelling does. A site with limited available power from the local utility may simply be unable to support the DC output a pure demand model would recommend, without a transformer upgrade that can add months to a project timeline and meaningfully change its economics. Confirm available grid capacity early in site selection, not after hardware has already been specified, since redesigning a charging plan around a lower available power budget after equipment has been ordered is one of the most expensive and avoidable mistakes in a fast-charging rollout.
Building In Headroom For Growth
The sites that perform best over a five-year horizon are rarely sized exactly to day-one demand. EV adoption curves mean that a site sized precisely for today's peak concurrency will likely be undersized within two to three years as the local EV population grows. Where budget allows, oversizing the electrical infrastructure, conduit, transformer capacity and grid connection, while phasing in actual charger units as demand materialises, is usually cheaper than a second disruptive civil works project later. A manufacturer who can supply the full Tejas range from one platform makes that phased approach straightforward, since adding capacity later means adding compatible units to an existing site rather than redesigning it from scratch.
Building an EV Charging Offering?
Talk to our engineering team about OEM, ODM and turnkey deployment.
