We build the control layer — one that stays with you, including when you change aggregator or supplier. The system plans its operating strategy, corrects it as conditions change, and shows afterwards why the site did what it did.
Threshold: a grid connection from roughly 200 kW, storage from roughly 500 kWh.
Solar, battery and grid report their state. The node decides what to do with the energy. Setpoints travel back to the devices and the loop starts again.Problem
A battery does not earn money. The decision when to charge it does
A battery attached to a solar farm without a decision layer does one thing: it charges on surplus and discharges when the surplus ends. It does not look at the price, the forecast, or the grid connection limit. You see the consequences on the invoice.
Energy sold when nobody wants it
On days with negative prices, export is not revenue — it is a cost. A controller that does not check this pays for every kilowatt-hour it produces.
Grid power bought to fill the battery
A regulator that pushes charging power up against current production drops below zero at every passing cloud. On paper you are charging from solar. The meter shows imports.
Capacity left idle
A battery charged at a fixed power rating is more mismatched the larger it gets. The problem grows with the investment rather than shrinking.
Revenue handed over for market access
Reaching the balancing market through an aggregator typically costs 20–30 % of the revenue from system services. The harder part is what remains when you change aggregator: the controller was theirs, the data was theirs, the integrations were theirs.
No answer to “why”
When a day goes badly, without a record of decisions there is no way to tell whether the market, the weather, or the controller was at fault.
How it works
Three parts, one decision
You supply the battery, we supply the system that runs it.
On site
SolverGrid Node
The autonomous unit that runs the installation. It integrates meters, inverters, the battery and the panels. It plans the operating strategy from energy prices, their forecasts and the weather forecast, corrects it as conditions change, and accounts for operating, transmission and loss costs. It sits on site, so a dropped link does not stop the control.
control and strategy in real timeSCADA for the operator
SolverGrid Console
A full SCADA: live power figures, the state of every device, the operating plan, alarms and history. After any day you can see why the farm did what it did.
operating history · decision trailIn development
SolverGrid Pool
A virtual power plant: individual sites add up to one portfolio. An installation too small on its own to reach the markets reserved for large players reaches them together with others — and shares in that revenue.
building it — not selling it yet
Reference site
Reference site
The system has been in production control since August 2026 on a farm in central Poland. The control is ours — we are not a layer on top of someone else's regulator.
985 kWsolar, AC side
510 kWhstorage
1 MWgrid connection
Price arbitrage
The battery is charged and discharged against the price curve, not against production surplus alone.
Planning ahead
The plan for the next day exists before the day starts, and it is stored. This is not hindsight.
Curtailment at negative prices
When export stops paying, the system curtails production rather than paying to sell. The same mechanism works on a farm with no battery at all.
Import discipline
Charging is limited so that the connection point never goes below zero. Drawing from the grid during a charging window is treated as a fault, not an acceptable side cost.
A decision trail
Every setpoint carries its reason, its timestamp, and the state of the inputs behind it.
We do not name the site, the owner, or any settlement figures. A client's financial data is not ours to publish.
Who this is for
When it is worth talking to us
Size threshold: a grid connection from roughly 200 kW, storage from roughly 500 kWh. We are strongest in the 0.5–10 MW band. Below that scale our system is overkill and it is better to say so at the first call.
Priority
Renewable investors
You own a farm or a project, you are considering storage, and you want to know what it will return before you buy it. We compute that against your production profile and your connection limits.
farm, project, SPVChannel
EPC and installers
You are building the installation and need a control layer that will not become your service problem — and that will not require you to hire a developer. We come in as the EMS supplier and leave the client relationship, and room for your margin, with you.
closing the offerLonger cycle
Industrial plants
You have your own load, large draws, and a connection that is sometimes the bottleneck. Storage with real control is a tool for shaving peaks, not only for trading energy.
contracted capacity, peaks
Safety and limits
What the system is forbidden to do
This section is for whoever is responsible for operations on site.
Limits are hard, not advisory
The export limit at the connection point, the price floor below which we do not sell, and the permitted grid draw are parameters of the optimizer. A plan that violates them is never produced — not produced and then corrected.
Control is deterministic
Given the same inputs, the system returns the same setpoint. There is no component that learns in production and behaves differently a month later without a configuration change.
There are no overrides
The operator's manual mode is subject to the same limits as the optimizer. You cannot enter a setpoint that exceeds the connection.
Every decision is recorded
The setpoint log, the optimizer's plan, and measurement history. After an event there is something to reconstruct it from — including when the reconstruction does not favour us.
Contact
Let's talk about a specific site
The shortest route: tell us what is installed and what your grid connection is. We will tell you whether it is our scale, and what could be gained from it.