Argus EMS · 26.10.2026 · 6 min read
Every kW, kWh and power factor value an energy analyzer reports depends on the current transformers (CTs) in the panel being correctly sized, configured with the right ratio and installed facing the right way. However precise the analyzer is, a single reversed CT or a wrong ratio typed into the setup menu produces numbers that look plausible on screen yet do not reflect reality. This article explains how to choose a CT ratio, how ratio, polarity and phase mapping errors show up in the measurements, and how to verify them in the field.
A current transformer converts the high current flowing through its primary into a small, proportional secondary current that the analyzer can read. A nameplate reading of 400/5 A means that 5 A flows in the secondary when 400 A flows in the primary, a transformation ratio of 80. The rated secondary current is usually 5 A or 1 A. The analyzer measures the secondary current and multiplies it by the configured ratio to obtain the primary current; power and energy values are then derived from that calculated current.
The ratio is not the only selection parameter. The accuracy class defined under IEC 61869-2, the rated burden (VA), which is the total load the secondary circuit may carry, and the physical fit of the CT around the busbar or cable all matter. Metering CTs deliver their best accuracy close to rated current.
One of the most frequent field errors is a mismatch between the CT installed in the panel and the ratio configured in the analyzer. The CT is replaced during a panel retrofit but the analyzer setting is left unchanged, or the factory default is never edited during initial setup. The error is insidious because the readings are neither zero nor negative; they are simply shifted by a constant factor:
True value = Displayed value × (Installed CT ratio ÷ Configured CT ratio)
For example, if a 400/5 A CT is installed but 200/5 A is configured, every current, power and energy value appears at half its true magnitude. Power factor and voltage are not affected by the ratio and keep looking correct, which makes the error harder to spot. In medium-voltage metering the same logic applies to the voltage transformer (VT) ratio. How the accumulated energy register behaves after the ratio is corrected depends on the analyzer model, so recording the date and time of the change is the most reliable way to explain the break that will later appear in the consumption chart.
Checks that can expose a scaled error:
P ≈ √3 × U × I × cos φ, with U as line-to-line voltage) and compare it with the displayed kWEvery CT carries P1/P2 markings on the primary side (K/L in older notation) and S1/S2 (k/l) markings on the secondary side. The rule is simple: P1 faces the source, meaning the transformer or grid side, and S1 goes to the start terminal of the matching analyzer current input. When a CT is mounted backwards on the busbar, or S1 and S2 are swapped at the terminal block, that phase current is measured rotated by 180 degrees and the analyzer calculates negative active power for that phase.
The symptoms depend on how many phases are affected. With one reversed CT, total power stays above zero but is clearly lower than it should be; on a balanced load the reversed phase is subtracted from the two correctly wired phases and the total appears at roughly one third of the true value. If all three CTs are reversed, total power turns fully negative, the analyzer accumulates energy in the export register, and the import register stops advancing under load.
A harder error to spot is a CT installed in the right direction but mapped to the wrong phase: the secondary of the CT on the L1 busbar is wired to the analyzer's L2 current input. Because the analyzer calculates each phase power by multiplying that phase voltage with that phase current, an artificial 120-degree angle is introduced between them. Per-phase power factors become meaningless, some phases show negative or near-zero active power, and total power no longer tracks the load in a sensible way.
| Symptom | Likely cause | Field check |
|---|---|---|
| One phase shows negative kW, total is low | That phase CT is mounted backwards or S1/S2 are swapped | Check P1 direction and terminal order |
| Total kW negative, export register growing | All CTs reversed or a collective reversal at a common terminal block | Verify CT mounting direction and S1/S2 wiring together |
| Per-phase power factors differ widely | CT and voltage input mapped to different phases | Trace each CT together with the voltage tap of the same phase |
| Current and power off by a constant factor, power factor correct | Configured ratio differs from the installed CT | Read the nameplate, compare with a clamp meter |
| Power factor shown above 1 | Register scaling or decoding error | Check the register map and scale factor |
Argus EMS reads energy analyzers such as the Schneider PM5350 over Modbus. On the sites the platform monitors, energy flows in one direction only and export is not expected. Negative total kW is therefore not interpreted as generation but as a CT polarity or phase mapping fault, and a CT polarity alarm is raised. The value is never turned positive with an absolute function; the fault stays visible so that the field team corrects it. Because flow direction cannot be measured reliably at very low load, the alarm opens only when the negative value exceeds the 1 kW noise floor and persists over at least two consecutive checks. An hourly wiring fault scan also flags a growing export register and an import register that does not advance while power is flowing, and the transformer health card shows a reversed phase sequence as a critical condition.
Consumption is calculated only from the difference of the import kWh register, so a reversed meter shows zero consumption instead of silently distorting reports. The field agent keeps readings in a 30-day local buffer and does not delete them before the server acknowledges receipt, and the server retains raw readings for 13 months, so the moment a fault began can be traced back. A wrong CT ratio, however, is a scaled error that software cannot distinguish on its own; an engineer needs to confirm it with the main meter versus submeter comparison described above.
Related Content
Explore the system with your own data in a demo session.