Table of Contents

A PV+ESS+EV charging station is three supplies and a car, joined at the site switchboard. The panels produce DC, and an inverter turns that into AC. A stationary battery charges or discharges through its own power converter. The grid covers whatever is still missing, or accepts an export if the connection allows it. The charger is a controlled load with a plug. It does not take a private cable from the module to the vehicle, and it does not follow the sun unless a separate controller tells it to.

That is a different picture from a four-step story in which the panels generate, an inverter “matches” the car, the car charges, and a battery is mentioned afterwards. At a busy bay the three supplies are available in the same minute. Which one actually feeds the car is a setting: solar first, a cap on grid import, or a full charge before the driver leaves. The setting is not built into sunlight.

Where PV, storage and the charger meet

On a station built from ordinary products, the meeting point is AC. A rooftop inverter, a battery cabinet and an AC charger each have their own connection to the building. The single-line diagram is what tells you the order, not the logos on the boxes. Add the building’s own lights and HVAC on that same board and the car is no longer the only load the solar has to serve.

A second architecture shares a DC bus: panels, a battery and a DC charger sit on one DC link, and one converter faces the grid. Research stations and some integrated posts are built this way. The IEA PVPS Task 17 report on PV-powered charging stations (T17-04:2025) studies microgrid stations of that family, including vehicle-to-grid as a possible service. It is not the wiring you get by putting a standard roof inverter next to a standard AC post. If the diagram does not show a common DC bus, the solar DC is converted to AC before the car ever sees it. A DC charger then converts it back.

One hour when the cars overlap

Power adds on the board. A 7 kW AC charger held at full output for one hour delivers 7 kWh. Two of them are 14 kW while both are unrestricted. If the array is producing 6 kW at that moment, 8 kW still has to come from the battery, the grid, or a decision to slow the cars. Those three numbers are arithmetic, not a mode in a brochure.

The same report describes one station in the north of France: five parking spaces, 28 kWp of panels (70 modules on 124 m²) and 22 kWh of stationary storage. Five bays and 28 kWp of modules sit in the same range as a handful of slow AC posts. They are not in the range of a row of 30 kW DC posts, which would be 150 kW if five of them ran together. The 22 kWh battery, taken as nameplate and ignoring losses and reserve, is a few hours at 7 kW and well under an hour at 30 kW. The grid connection is what makes the busy hour possible. The panels and the cabinet do not.

Storage is for that overlap, and for the hours after the array has fallen off. It is not a bin for “whatever is left once the cars are full.” If the cars arrive as the sun drops, the battery is the supply, and it only has what it stored earlier. A vendor flowchart that waits until the battery is down to a fixed percentage and then jumps to the grid is describing that vendor’s setting. The percentage is not a property of PV+ESS+EV.

What the charger converts

An AC charger delivers AC to the car. The car’s onboard charger makes the DC that its battery can take, so a 7 kW or 11 kW post is limited by that onboard unit as well as by the post. A DC charger makes DC in the station and feeds the car battery more directly. That is why a 30 kW DC post is a different service, a different cable and a different load on the switchboard from a 7 kW wall unit.

If the solar side is microinverters, the array is already AC at the trunk. Feeding a DC post means a second conversion inside the charger. The charger’s own efficiency applies to that step only. It is not the efficiency of the panels, the battery or the round trip through storage. Keeping the solar on DC all the way to the car requires the shared DC bus above. Buying a DC charger does not create that bus.

Vehicle-to-grid is another function again. The car must be allowed to discharge, the charger must be bidirectional, and the site must be allowed to export or to feed the building. A charge-only post does not do it because the acronym PV+ESS+EV appears on the drawing. The IEA report treats bidirectional transfer as a service people have to accept, not as the default path of every station.

Kilowatt-hours are the easy half

The same IEA report’s grid discussion, drawn from French system data and offered as a wider lesson, is that the energy of a day’s charging can fit the network more readily than the power of many cars charging at once. Smart charging has to do more than turn the posts down. It has to decide when a car waits for the sun and when it must not.

Two of the report’s field results show what that decision is worth, and where it stops.

  • At CEA Cadarache, a supervisor tried to fill the cars and, at the same time, line up 24 charging points with a solar plant. The experiment raised the self-production ratio by as much as 60 percentage points while still meeting the drivers.
  • At a park-and-ride solar car park outside Lisbon, a year of real occupancy plus smart charging cut electricity imported from the grid by more than 35%. The same section of the report is blunt about the clock: solar peaks and parking peaks do not match, weekends overproduce, and adding smart charging or a battery did raise self-consumption but, at the prices they used, earned less than simpler energy management.

A further result in the report’s real-time control work is that, when the controller knows when the driver will leave, most of those cars were charged with more than 80% PV energy. Without a departure time, “as fast as possible” pulls the grid at the moment the plug goes in, which is often not noon. The 80% figure is that control study, not a promise for every car park.

On carbon, the report’s comparison with a grid-only public charger is a range of about 1.5 to 10 times lower for the PV station, depending on the grid mix. The French case above was assessed at 40.7 gCO₂eq/kWh, against the report’s averages of 275 gCO₂eq/kWh for a grid-powered public station on the EU mix and 368 on the US mix. A grid that is already low-carbon leaves less for the panels to improve. The report’s own cost line for that French station is an initial outlay near €65,000 and about €150,000 over 30 years, with maintenance a large share of the lifetime cost. None of those euros is a ShineTong price.

The protocol is not the energy manager

Three different jobs get sold as one “smart” feature.

OCPP, in the 1.6J form stated on charger products, is how a post talks to a charging-network back office: sessions, users, a remote start. It does not measure the array and it does not discharge a stationary battery. A site can have OCPP and still have no idea what the roof is producing.

Load balancing shares a limited supply across posts so the main breaker holds. If two outlets are told to split 11 kW, they split 11 kW. They do not look up the solar curve unless some other device keeps changing that 11 kW limit.

The energy manager is the piece that sees PV output, battery state, the import cap and the cars, then writes the limit. The IEA report’s human interface exists so a driver can see whether the kilowatt-hour came from the roof, the cabinet or the grid, and so the site can use a departure time. If that screen and that controller are not in the scope, the station is three products side by side. It will still charge cars. It will not prefer the sun.

What ShineTong supplies on this station

ShineTong sells the pieces, not the dispatcher. We do not build the hybrid inverter, the site energy manager, or a single cabinet that runs PV, storage and charging as one loop. A line that says the charger syncs to solar is not a control diagram, and this article will not treat it as one. Load balancing, where a model lists it, is the shared-supply limit above.

The EV chargers on the current page:

OutputPlugWhat the card adds
ST-AC077 kW, 32 AType 2IP65, Wi-Fi and an app. A home post. One of these is a 7 kW load for as long as the car will take it.
ST-AC1111 kW, 16 AType 2 socketIP65. RFID and OCPP are listed on this card. The page states OCPP 1.6J, not a later release.
ST-AC2222 kW as 11 kW × 2Two Type 2 outletsIP54. Load balancing is listed here. It is two posts sharing a cap, not one 22 kW gun and not a solar tracker.
ST-DC3030 kW DCCCS2 and CHAdeMOEfficiency above 95%, 7-inch touchscreen. This is the post that makes DC in the station. It is an order of magnitude above one quad microinverter.

Microinverters (ST-500M, ST-1200D, ST-2800Q) stop at 2.8 kW AC on the largest series, at 220–240 V, with a 60 V DC maximum. They can feed the same board as the chargers. They are the wrong tool if the drawing assumes one box will run a 30 kW DC island. LiFePO4 storage is the stationary battery we quote, for grid-tied or off-grid projects, with no public kilowatt-hour menu and no NMC pack. It still needs a converter we do not make, and a charge profile written for LFP. Whether that battery belongs on a house is a separate question, answered in are LFP batteries a good fit for a house solar system. How the inverter class is chosen is in how to pick a PV inverter.

Send the array kilowatts, the import limit, the charger count and whether anyone is actually buying the energy manager. We will say which of the three catalogues we can fill, and which box you still have to get elsewhere.

Send the array, the import limit and the charger list

Frequently asked questions

How does a PV+ESS+EV charging station work?

Panels, a stationary battery and the grid can all feed the same switchboard, and the EV chargers draw from that board. Solar is converted to AC unless the design really is a shared DC bus. The battery charges when there is surplus and discharges when the cars exceed what the panels are making. The grid covers the rest, up to the import limit. A controller has to choose that order. The hardware does not choose it by being installed in the same car park.

Does the solar array charge the car directly?

Only on a design with a common DC bus from the modules to the charger. On a normal AC-coupled site the modules go through an inverter first. An AC charger then leaves the last conversion to the car. A DC charger converts again inside the post. “Solar charging” on that site means the board was importing less because the roof was producing, not that a DC cable ran from the module to the vehicle.

What is the stationary battery for?

It covers the minutes when several cars want more power than the array is producing, and it can move midday surplus to a later arrival. It does not enlarge the grid connection, and a small cabinet cannot stand in for a 30 kW post for long. In the IEA report’s Lisbon car park, storage and smart charging raised self-consumption, and the same study found the extra equipment a weaker financial return than simpler management at the prices they used.

Will an OCPP charger follow the solar?

Not by itself. OCPP 1.6J is the back-office language for sessions and users. Load balancing shares a fixed cap across posts. Following the roof requires an energy manager that keeps rewriting that cap from measured PV, battery state and the driver’s departure. Without the departure time, fast charging uses the grid whenever the car plugs in.

Does ShineTong sell the whole station as one system?

No. We supply microinverters, LiFePO4 storage and the four EV charger series above. We do not supply the site controller that dispatches them, the hybrid inverter in the battery cabinet, or vehicle-to-grid. If the project needs those, they have to be specified as their own equipment and shown on the single-line diagram.

Sources