Export Control CT and Energy Meter Selection Guide

A current transformer is the eyes of an export control system. Get the spec wrong and the rest of the install — the inverter limits, the controller, the meter — has nothing accurate to act on. This guide walks the decision in the order you actually make it.

Before you size anything, locate where the measurement has to happen. The rule is simple: the CT goes on a section of conductor beyond which power cannot flow back to the grid. For an export control deployment that means somewhere on the service conductors between the utility meter and the main distribution panel.

Physically, this is usually a long run of wire — sometimes a copper busbar — between the meter base and the point where the service splits out into branch circuits. You need eyes on every electron heading toward the meter.

Pull the single line diagram. Find the conductor or panel at your measurement point and read the service rating. North American service sizes are standardized — you'll be looking at 200 A, 400 A, 600 A, 800 A, 1200 A and so on up. That's your CT amperage rating.

This part is straightforward. Either the conductor will be tagged directly on the SLD, or the panel it feeds will be labeled with its capacity. Match the CT rating to the service rating and move on.

This is where most specification errors happen. The amperage rating tells you the electrical spec. The physical conductor configuration tells you what shape of CT you actually need to fit around it.

Split core CT

A split core CT is rigid. It has a fixed window opening — common sizes range from a 1.25-inch round window up to 3 × 5 inches and beyond. The window snaps closed around your conductor. Cheaper, simpler, works with most meters out of the box.

Rogowski coil

A Rogowski coil is a flexible CT. It's a long coil with a clasp on each end. You wrap it around whatever conductor geometry you have — single conductor, busbar, or a bundle — and click it closed. It takes the shape it needs to take.

Procurement decisions get made from a desk. You have the SLD, you have a deadline, you don't have a service truck. Here's the safe play.

The Rogowski coil is the most conservative specification you can make. It will fit whatever conductor geometry shows up on install day. If the SLD gives you the amperage but tells you nothing about whether the service is single-conductor or ganged, default Rogowski.

That conservatism has a cost — both at the part level and downstream in the meter you can pair it with. Which is the next decision.

Rogowski coils and split core CTs do not produce the same output signal. A split core CT outputs a voltage waveform — most commonly 333 mV at full scale — that meters are built to read directly. A Rogowski coil outputs a different waveform: phase-shifted and at a different scale. Most low-cost meters cannot interpret it correctly.

For Sponge deployments, the WattNode is our default recommendation — cheap, reliable, and purpose-built for split core CTs. It is not compatible with Rogowski coils for power and power factor measurement. Plug a Rogowski into a WattNode and you'll get current readings, but the power, energy, and power factor data will be wrong.

If you've specified Rogowski, you have two options:

• →Upgrade the meter. An AccuVim accepts Rogowski directly. Budget ~$1,000 more than a WattNode.
• →Add a signal converter. A digital translation module sits between the Rogowski coil and a standard CT-input meter. Cheaper than upgrading, but adds another device to power, wire, and commission.

A three-phase service requires three CTs — one per phase. Each phase gets its own measurement because the inverter and the building loads aren't necessarily balanced across phases, and the control loop has to know the net.

For export control specifically, the target is net-zero export across all three phases combined. The inverter produces roughly balanced power on each phase, and the building generally draws roughly balanced load — but single-phase loads create imbalance. You don't care if one phase exports a little and another imports a little, as long as the sum at the meter is zero or negative.

Wiring discipline matters here. On the meter, CT input 1 reads phase 1, CT input 2 reads phase 2, CT input 3 reads phase 3 — and the voltage reference inputs need to map to the same physical phases. Cross a phase reference and the meter reports nonsense. Verify the phase mapping at commissioning.

The CT is the cheapest piece of hardware in an export control deployment and the easiest one to get wrong. Spec it deliberately. The control loop is only as good as the current it sees.