Plenty of commercial sites end up with a connection agreement that limits export to a figure well below what the roof could produce. Sometimes it is offered as the fast alternative to a long queue; sometimes it is simply what the network can accommodate.
A cap is not the end of the project. It changes what a good design looks like.
Export limitation caps power flowing from your site to the grid. Self-consumption is unaffected. If you are consuming 400 kW and generating 600 kW with a 100 kW export cap, the system must reduce output to 500 kW.
Generation limitation caps total generation regardless of consumption. Much more restrictive and less common, but it exists – check which one your agreement actually says, because the design implications differ completely.
A meter at the connection point measures net flow and feeds a controller that instructs inverters to reduce output when export approaches the limit.
Two parameters matter and are often left to default:
Response time. The system must react before a sustained breach. Network operators usually specify a maximum response time – commonly a few seconds. Faster response allows operation closer to the limit; slower response requires a wider safety margin, which costs you energy every day.
Set point margin. Systems are typically configured to target somewhat below the actual limit for safety. A 10% margin on a 100 kW cap means you routinely export only 90 kW. Over a year that adds up, and it is worth asking whether the margin can be tightened with better control.
This is the counterintuitive part. When export is capped, adding more DC capacity does not increase your peak – you were already capped at peak. It increases output during the shoulder hours, mornings and evenings and overcast days, when you were nowhere near the cap.
The result is a flatter, wider production curve that spends more of the day at the cap rather than a sharp peak that touches it briefly. Total annual energy rises significantly even though peak power does not.
DC:AC ratios of 1.4 or higher can be rational under a firm export cap, where they would be wasteful on an unconstrained connection. The clipping that would normally be a loss is largely irrelevant, because you could not have exported that energy anyway.
Export limits make storage more attractive, sometimes decisively. Energy that would be curtailed at midday can charge a battery instead and discharge in the evening – when you are importing, and when export headroom exists anyway.
DC-coupled storage is particularly effective here, since it can capture energy before the inverter's export-limited output stage.
Under a cap, the value of your system comes overwhelmingly from self-consumption. That makes your load profile the design driver.
A site consuming 500 kW steadily through daylight hours can host a large array under a 50 kW export cap, because almost everything is consumed onsite. A site consuming 80 kW cannot, regardless of roof area.
Model self-consumption hour by hour against real interval data. The optimal size under a cap is usually the point where marginal generation starts being curtailed rather than consumed, and that point is entirely determined by your load.
Network constraints change. Reinforcement projects complete, other generators disconnect, load grows locally. Some operators offer flexible connection agreements where the cap varies with actual network conditions rather than being fixed at the worst case – often giving substantially more export in practice.
Ask whether flexible terms are available, what the historical availability has been on your circuit, and whether there is a defined route to review the limit later. A cap that can be revisited in five years is a different asset from one fixed for twenty.