Ljubljana trial uses day-ahead prices for charging schedules
A demonstration in Ljubljana involving Avant Car and Kolektor sETup applied automated scheduling to shift EV charging into lower-cost electricity periods while keeping vehicle charging requirements met. The project used day-ahead electricity-price signals to decide when vehicles should charge. The trial ran for two months and covered six charging stations .
Cost results were benchmarked against a theoretical least-cost scenario. Before optimisation, actual charging costs were 18.9% above the least-cost reference. During the demonstration, that gap narrowed to 5.99%, an improvement of around 68% aligned with the cheapest available charging periods .
Charging flexibility created by connection and departure windows
The commercial significance described in the project relates to how EV fleets interact with power systems. EV fleets combine large electrical loads, predictable periods when vehicles are connected, and flexibility over when charging occurs. A vehicle can require electricity before its next journey but typically does not need to consume every kilowatt-hour immediately after plugging in.
The hours between connection and departure form a flexibility window that can be monetised through software. A fleet-management platform can determine which vehicles need immediate charging, which can wait, and how much aggregate consumption can be moved between different electricity-market periods. Charging is therefore treated as an optimisation problem rather than a simple transaction between charger and vehicle.
From procurement savings to controllable portfolios for aggregators
For fleet operators, the immediate benefit highlighted is reduced electricity cost. For aggregators and electricity suppliers, the larger opportunity is combining hundreds or thousands of chargers into a controllable portfolio. The project notes that a fleet with hundreds of vehicles could shift several megawatts of electricity demand from one period to another without changing the transport service delivered to customers.
This approach is described as relevant beyond energy procurement, extending toward demand response, balancing and local flexibility markets. The Slovenian demonstration focused primarily on day-ahead price optimisation, shifting charging sessions toward less expensive periods, particularly after midnight. Automated optimisation was reported to narrow the difference between actual charging cost and the theoretical optimum .
Scaling smart charging requires algorithms that reflect network limits
The material also describes a scaling challenge if large numbers of vehicles respond to the same price signal. If thousands of vehicles automatically move charging into the same cheap hour, it could create a new demand peak. It also states that what is optimal for electricity consumers may not be optimal for the network.
As a result, it says smart-charging algorithms need to consider at least two signals simultaneously: wholesale electricity price and the physical condition of the local network. It adds that a third signal may come from balancing or flexibility markets, allowing an EV fleet to respond differently depending on which service has the highest value .
Centrally managed fleets use schedule information to optimise charging
The business model is described as particularly attractive for centrally managed fleets where schedules are easier to forecast. Car-sharing operators, delivery companies, municipal fleets, taxis, corporate vehicles, buses and logistics companies are listed as having better information about vehicle schedules than individual residential customers.
A fleet operator may know which vehicles must leave at 06:00, which remain parked until noon, and how much energy each one requires. An optimiser can use those constraints to determine the cheapest or most valuable charging schedule automatically. In this setup, the physical charger is described as only one part of the service.
Software-led value creation and implications for infrastructure procurement
The material states that more value may sit in the software layer controlling thousands of chargers. It identifies opportunities for companies providing fleet-management systems, aggregation platforms, automated trading, charging optimisation and electricity-market access . It also says infrastructure procurement could change as buyers evaluate a platform’s ability to optimise electricity cost and earn flexibility revenue over an equipment lifetime rather than comparing chargers mainly by hardware price and maximum power.
The text also notes that vehicle-to-grid technology could extend opportunities by allowing electricity to flow back from EV batteries. However, it states bidirectional charging is not necessary for the first stage of the market because controlling when vehicles consume electricity already creates substantial flexibility . It adds that smart charging becomes commercially relevant sooner than full V2G deployment.
Limits of current demonstrations relative to national flexibility markets
The Slovenian demonstration is described as small compared with the scale required for a liquid national flexibility market. Still, it reports that smart charging of a shared fleet can materially reduce the gap to optimal charging costs under real operating conditions . The larger opportunity is presented as scaling the model as EV adoption grows.
The material warns that unmanaged charging risks becoming another source of peak electricity demand as adoption increases. It contrasts this with managed fleets described as controllable loads for power systems rather than an added burden on the grid . It further states that an EV fleet’s value may increasingly include not only kilometres travelled but also flexibility created while vehicles are standing still.
