Most EV charging strategy imported into African cities is built around an assumption that does not hold here: that the vehicle being charged is a private car, parked overnight at a home with reliable power, driven by an owner who values convenience over cost per kilometre.
Nairobi's electrifying fleet is overwhelmingly two- and three-wheelers. Boda-boda riders are commercial operators running thin margins on high daily mileage. Their charging behaviour is driven by economics and time-off-road, not convenience. Design for a car owner and you build infrastructure in the wrong place, at the wrong power rating, on the wrong business model.
Downtime is the real currency
For a commercial rider, a charging stop is lost earning time. That single fact reorders every design priority. It is why battery swap is compelling for this segment in a way it never was for passenger cars in Europe: an exchange takes minutes, and the depleted pack recharges on the operator's schedule rather than the rider's.
Swap moves the economics too. The rider no longer finances the most expensive component of the vehicle. The operator holds the pack, manages its lifecycle, and can charge when electricity is cheapest or solar generation is strongest. It converts an ownership problem into a logistics problem, which is generally easier to solve well.
A rider does not buy kilowatt-hours. They buy the shortest possible interruption to their working day.
Siting is the whole game
Leading site-selection modelling for distributed fast-charging along Nairobi's commuter corridors, the pattern that emerged was clear: charging belongs where riders already stop. Stage points, market edges, fuel stations, transport hubs. Not purpose-built destinations that require a detour, because a detour is unpaid distance.
GIS analysis makes this tractable. Overlay commuter flow, existing informal stage locations, grid availability and land access, and viable sites narrow quickly. The constraint that eliminates most candidates is rarely demand. It is the connection: whether the feeder at that location can support the load without an upgrade that destroys the project economics.
Solar is not decoration here
Pairing charging with solar PV and second-life battery storage is often treated as a sustainability gesture. In this context it is a grid-constraint workaround and a tariff strategy.
A site with local generation and storage can present a far smaller peak demand to the network than its charging throughput implies. That can be the difference between an achievable connection and a prohibitive one. Storage buffers the mismatch between when riders arrive and when the grid or the sun is cheapest. The hub becomes a managed energy asset rather than a load.
- Size for peak concurrent riders, not theoretical daily throughput
- Model against real tariff structure and demand charges, not average unit cost
- Confirm feeder capacity before committing to a site, not after
- Design swap logistics and charging logistics as one system
- Assume dust, heat and heavy cycling as the operating baseline
Regulation is a design input
Engaging EPRA and EMAK on compliance readiness and technical risk mapping for charging networks taught me to treat regulation as an engineering constraint rather than paperwork. Standards for installation safety, metering and grid interconnection shape what is buildable, and the requirements are still consolidating.
That fluidity cuts both ways. It creates uncertainty, and it creates an opening: operators engaging early help shape workable rules instead of inheriting rules written without operational input. Supporting stakeholder consultations that attracted a substantial investment pipeline, the projects that held up under scrutiny were the ones that could evidence regulatory pathway alongside technical viability. Investors price unresolved compliance risk aggressively.
Build measurement in from the start
The reason I built a full-stack platform pairing real-time energy simulation with Kenya-specific tariff logic and location-aware irradiance modelling is that charging infrastructure without instrumentation cannot be optimised, financed or defended.
Utilisation by hour, energy cost per session, solar contribution, battery state of health across the fleet: these determine whether a site is profitable and whether the next one should be built. Retrofitting measurement onto deployed infrastructure is far harder than designing it in. The dashboard is not a reporting layer. It is the thing that makes the asset legible to whoever funds the next ten sites.
The model that fits
What works here is a dense network of modest, solar-backed, swap-capable points positioned along corridors riders already travel, instrumented well enough to prove their own performance. Not a thin scatter of high-power stations built for vehicles that make up a small share of the fleet.
That is a less impressive press release and a considerably better piece of infrastructure.
Working on something in this space? Start a conversation.