Electricity CT Metering Accuracy

During the commissioning of a ground source heat pump it became evident that the current transformer (CT) metering was only recording about 75% of the energy, compared with instantaneous readings taken from a recently calibrated multi-meter. The CT meter was showing a current of 3.6 Amps and the multi-meter 4.5 Amps. This clearly would substantially affect the conclusions of data analysis in respect of COP/SPF calculated from, for instance, the electricity meter kWhe and a heat meter kWhth; in fact it would lead to the conclusion that the GSHP efficiency was better than it actually was. For example, if the average output of the heat pump was 3.5kWth, at 230V the electricity meter reading would lead to a calculated COP/SPF of 4.23, when in actual fact it would be 3.38! 

The project required most sensors, water/thermal metering and electricity metering to provide +/-1% (class 1) accuracy and current transformer (CT) electricity metering was used to reduce the  invasive effect of installation and provide more detailed data, such as reactance, Amps and power factor, than provided by whole current (kWh) electricity meters.

The equipment was supplied as a package from the manufacturer of the monitoring equipment, which included third party temperature/humidity sensors, CO2 sensors, electricity meters and current transformers. The current transformer supplied for the GSHP meter was a 40:5 Amp solid core type which was the lowest rated CT, and considering the heat pump and ground pump total FLC was less than 5 Amps would seem to be the obvious choice.

However, following the discovery of inaccurate data from the GSHP electricity meter the current transformer selection was analysed in more detail. It became evident that the selection of metering current transformers must not only consider the burden (VA), which is the sum of the cable circuit length between the meter and the CT plus the impedance, across the CT terminals of the meter but, also the accuracy class of the CT. The table below immediately highlights that the 40:5 Amp CT cannot meet class 1 accuracy required by this project. Also, it can only support a burden of 1VA.

  The problem at commissioning was that the wiring of sensors and CT cabling had been completed during refurbishment of the property and therefore up-rating the CT cable to reduce the burden wasn’t an option. The selection table above showed that a 100:5 CT would give class 1 accuracy at a maximum burden of 2.5VA. By calculation, the maximum total impedance at the secondary winding of the CT should be no greater than 0.1Ω.

The electricity meter technical data sheets stated a terminal impedance of 0.012Ω and therefore cable resistance using R=ρl/A, the maximum circuit length with the existing cable should not exceed 4 metres with a burden of 2.08VA. Although slightly outside the threshold, a 5 metre circuit length had a calculated secondary impedance of 0.101Ω. 

The 40:5A current transformer was replaced by a 100:5A CT and similar Amperage comparison between the electricity meter and a calibrated multi-meter proved that acceptable tolerances were being met.

It is worth noting that during discussions with the current transformer manufacturers they commented that for cables >=10m at least a 6mm². The DIN rail electricity meter terminals used on the project will take a maximum cable of 2.5mm²!

Important conclusions may be drawn from this discovery.

1. When selecting current transformers for use with electricity meters it is important to meet both the requirements in respect of burden (VA) and accuracy class.
2. Do not blindly accept accuracy claims made by equipment manufacturers.
3. During commissioning, wherever possible, check other sensor accuracies with independent calibrated meters. For instance, mA data calibrated to units of measurement (UOM).