Selecting Current Transformers for Factory Power Monitoring: Meter Fit, Burden, FAT and SAT

Selecting a CT for factory power monitoring takes more than matching its primary rating to a breaker. The meter input, the lowest operating current, the resistance of both outgoing and return secondary conductors, available panel space and safe meter replacement all belong to the same measurement design. Otherwise a plant may see plausible numbers while missing standby energy, reversing one phase of power or leaving no safe way to service a meter. This guide gives engineering, maintenance and procurement teams in Thai factories a practical specification and acceptance route for CTs and energy meters.

Define the measurement purpose and point first

Start with the decision the reading will support: invoice reconciliation, peak demand, compressor standby load or energy per production unit. These uses demand different low-current performance and recording intervals. Record measured current at idle, ordinary production and peak operation; a breaker nameplate is useful for electrical limits but is not evidence of normal operating current. For each point, record panel and circuit IDs, single-line drawing, phase and wiring system, conductor dimensions, access, meter location, one-way cable route, current range, existing meter and communications interface.

If accumulated energy is the only requirement, an existing pulse or communication output may avoid another CT. See our factory power monitoring system guide and power meter data collection guide for the wider architecture. The decision here concerns points that do need a CT and meter.

Do not select a CT ratio merely because it is larger

Select primary current from the actual operating range and documented expansion plans, rather than copying the breaker frame rating. A needlessly high primary rating pushes idle current into a smaller percentage of the CT rating. Accuracy claims have defined current and burden conditions; a class stated at rated current cannot automatically be applied at every lower current. ABB’s low-voltage CT guidance, based on IEC 61869-2, shows error limits at specified percentages of rated primary current and distinguishes ordinary classes from S classes at low current. Put the lowest current at which useful accuracy is needed in the request for proposal and check the exact candidate’s data sheet.

For example, a circuit peaking around 430 A and expected to approach 500 A after expansion may justify a 500/5 A or 500/1 A candidate. That does not mean every excursion beyond 500 A immediately makes measurement impossible: permissible continuous current, overcurrent behavior and meter input limits are product-specific. Conversely, on a circuit peaking at 150 A, a blanket 600 A primary rating places a 30 A standby current at 5% of rating. Verify the selected model’s accuracy conditions at that point if standby energy matters. These figures illustrate the selection method; site measurements govern the purchase.

For three-phase power, CT count and locations follow the meter’s supported three-wire or four-wire connection and its voltage inputs. Do not assume that every three-phase scheme uses exactly three CTs. Attach the selected meter’s wiring diagram to the RFP and identify CT and voltage pickup by phase. A correctly wired CT with a wrong ratio or system type in the meter can still produce convincing but incorrect readings. Schneider Electric lists CT ratio, VT ratio, system type, frequency, polarity and wiring as checks for inaccurate meter readings.

Choose 1 A, 5 A or a dedicated sensor from the meter input

Confirm whether the meter supports 1 A, 5 A or a proprietary sensor before ordering CTs. A 5 A output is not interchangeable with a 1 A-only input. Some split-core devices provide a voltage output and have different burden, wiring-distance and open-circuit rules. Check compatibility in both the CT and meter documentation, even when model names look similar.

For conventional current-output CTs, secondary cable resistance creates burden. With equal loop resistance, wire loss follows I²R, making 5 A loss 25 times 1 A loss at rated secondary current. Schneider Electric advises including instrument burden and wiring loss, and identifies 1 A secondary or a larger conductor as options where burden is excessive. That does not make 1 A universally better: a short run to an existing 5 A meter can favor 5 A. Decide on the complete circuit.

Calculate burden using the round-trip cable

Estimate total CT secondary burden as meter input VA plus cable loss VA, accounting for any series test block and connection effects. Let I_s be rated secondary current and R_loop be the resistance of both conductors from CT through the meter circuit. Cable loss is I_s² × R_loop. A 20 m one-way route normally means roughly 40 m of conductor in the loop. Determine design resistance from actual wire size, material, temperature and terminations; a one-way distance must not be mistaken for total loop length.

An arithmetic illustration uses 0.40 Ω loop resistance, 0.30 VA meter input and zero extra terminal loss. At 5 A the cable contributes 25 × 0.40 = 10.00 VA and the illustrated total is 10.30 VA. At 1 A the figures are 0.40 VA and 0.70 VA. These are assumptions, not a product claim. Compare the computed total with the selected CT’s rated burden at the required accuracy class. If it exceeds the relevant limit, revise wire size, route, secondary rating or CT choice. Schneider Electric explicitly relates a CT accuracy class to the combined VA of meter and cable. Also check any applicable minimum burden or operating range in the selected product documentation.

Recalculate with actual cable and meter values before ordering. CT burden VA is not the same quantity as the power supply consumption of the meter or gateway. Plenty of power-supply capacity does not fix an overloaded CT measurement circuit.

Accept the measurement chain, not the CT label alone

The claim “Class 0.5 CT means kWh is within 0.5%” is unsound. A CT has ratio and phase errors; the meter has its own active-power and energy accuracy. Voltage sensing, power factor, system configuration, settings, integration and displayed rounding have conditions too. Specify the intended operating range and an end-to-end comparison method or agreed tolerance. Do not add unrelated accuracy labels and present the sum as a guaranteed system result; request the suppliers’ documented guarantees.

Calculate idle, normal and peak current as percentages of the CT primary rating. If idle is 1%, confirm that the selected CT and meter specify performance there. ABB’s IEC 61869-2 table shows test points such as 5%, 20%, 100% and 120% for ordinary examples, while S-class examples also cover 1%. A 30 A idle load on a 600 A CT therefore needs attention at 5%; 5 A on a 500 A CT needs attention at 1%. Check the exact product class, rated burden and meter low-current specification rather than treating the table as a blanket guarantee.

An internal submeter for operating decisions is not automatically an approved meter for commercial settlement. If the reading will determine a bill or contractual payment, ask the utility, metrology owner and equipment supplier for the applicable local approval, calibration and sealing requirements. This guide concerns plant monitoring.

Split-core or solid-core construction

In an existing panel, whether conductors can be moved and whether a shutdown is available may favor a split-core unit. In a new panel, a solid-core unit may be easier to integrate. Split-core does not mean an energized installation is safe for any worker. An authorized electrician must assess live-work restrictions, cover removal, clearances, voltage connections and the manufacturer’s instructions. On a split-core CT, closure, latch, conductor position and orientation require inspection; on a solid-core CT, consider how conductors are routed and how future replacement would work.

Check more than the nominal window diameter. Busbar width and thickness, cable insulation, adjacent phase clearance, room to open the latch, fixings and cable bend radius affect whether the device can actually be installed. Schneider Electric’s selection guidance starts with primary current range and window fit, then application, accuracy and burden. A CT that fits around the conductor but cannot be closed in the panel can cost more in rework than its purchase price.

Prevent polarity and phase errors in the drawings

Match primary orientation, secondary polarity and meter phase assignment as one drawing. Where the manufacturer marks H1 or an equivalent source side, place it and connect the corresponding secondary terminals according to that product’s instructions. Mark the source direction for phase A, B and C and give the wire numbers from CT terminals to I1, I2 and I3. Correct CT ratio settings cannot compensate for pairing voltage from one phase with current from another.

Schneider Electric’s installation guidance calls for polarity and orientation checks, sound terminations, correct split-core closure, and post-installation comparison of phase readings and expected power direction. Some meters offer a reversal diagnostic, but its meaning is model-specific. A plausible current display alongside unexpected power sign or phase angle calls for a phase and polarity check. Record the time and machine state for each SAT comparison so a changing load does not masquerade as an instrument fault.

Specify safe CT secondary handling

Never leave a current-output CT secondary open while primary current flows. Schneider Electric warns that hazardous voltage can appear and calls for an appropriate shorting block during installation or service; the CT secondary must be shorted or connected to a suitable burden. Specify a CT-rated shorting/test terminal and an approved service procedure before a meter is disconnected. A generic terminal strip is not evidence of a shorting function.

This paragraph is a design requirement, not a field switching instruction. Trained and qualified personnel must use local isolation, verification and manufacturer procedures. Do not perform energized CT work from a blog checklist. Voltage-output sensors may have different internal burden and service rules; use their own manual rather than applying the current-output CT shorting rule indiscriminately.

Place shorting terminals where future meter maintenance can be done without reaching deep into an inaccessible panel. Preserve secondary circuit drawings, terminal numbers, warning labels and CT and meter nameplate records. Procurement should include the service method because the next meter replacement is part of the system life cycle.

Put comparable CT and meter conditions into the RFP

Ask all bidders to quote against the same measurement-point table. Mark unmeasured values as estimates and name who will confirm them at the site survey. Compare complete installed scope rather than one CT price.

Separate CT, meter, cable, panel work, tests, configuration, drawings and shutdown costs. Otherwise a low-priced equipment quote may simply omit shorting terminals or SAT. Treat software and cloud fees separately and define data retention and recovery from communication loss.

Confirm quotation assumptions during the site survey

If a panel survey cannot happen before the quotation, make the post-award survey more than a window-size check. Reconcile single-line drawings with panel labels; record existing CTs, meters and test terminals, conductor shapes, access with the door open, routes for secondary and communication wiring, and the available shutdown window. Take controlled photographs of the whole panel, proposed CT position, meter terminals and nameplates under site information rules. If measuring current, observe representative production and peak periods rather than calling a single snapshot the maximum. A longer secondary route changes the burden calculation. Specify how discoveries change price and who approves revised calculations, drawings and bill of materials together.

FAT: verify what can be checked before delivery

Factory acceptance testing often cannot reproduce the site’s actual load. It can still compare CT models, ratio, secondary rating, class and burden documentation, phase and terminal labels, shorting blocks, meter ratio settings, communication mappings and drawings against the purchase specification. If test injection is used, record whether the test includes the actual CT or only the meter. A changing number on the display alone does not prove CT ratio, polarity or burden.

Keep serial numbers, drawing revisions, meter configuration exports, test equipment identification and calibration status, injected and indicated values by phase, open issues and corrective owners. Agree limits before testing and tie them to product specifications and test method. Do not create an intentionally open CT secondary to test an alarm.

Use an acceptance matrix that supports fault diagnosis

For each FAT and SAT test, record test number, point ID, wiring system, configured CT primary and secondary ratio, injected current or operating load, expected power direction, actual reading, result and corrective action. Include a field confirming that voltage and current inputs correspond to the same phase. Schneider Electric’s meter guidance names CT and VT ratios, system type, frequency, wiring and polarity as checks for inaccurate readings. Test low, representative normal and feasible high-load conditions separately. At low load, record the current as a percentage of CT rating and compare both CT and meter specifications; ABB’s IEC 61869-2 table shows why a class label alone cannot represent every current. For reference-meter comparisons, align measurement boundary, start and end times, load condition, reference calibration and energy increments. When a test fails, log the wiring or setting cause and the retest, rather than changing a ratio just to make the dashboard number match.

SAT: verify real loads through to the data system

At site acceptance, reconcile as-built wiring with drawings. Inspect CT orientation and closure, terminals, shorting-block state, meter ratio settings and voltage/current phase matching. Under live operating conditions, observe phase current and voltage, power factor, power sign, total kW and incremental kWh. When comparing with a reference meter or an upstream meter, align measurement boundary, time interval and operating state. Differences between an incoming meter and summed subcircuits can reflect unmetered auxiliaries or timing before they indicate a CT fault.

Test the operating conditions that justified the investment: idle and standby if standby energy is the target, normal production and a relevant rise in load. Define acceptance in observable terms: readings and integration across the required current range, expected phase direction, consistent units and register scaling through the dashboard, and recorded communications interruption and recovery. Set numerical tolerances from equipment guarantees and project needs, with evidence in the acceptance plan.

If readings fail, use a structured diagnosis. Zero may call for checking CT closure, terminals and meter input. A single negative phase suggests direction, secondary polarity or phase assignment. A common fixed multiplier points to CT ratio, voltage ratio or unit configuration. Errors concentrated at low load call for reviewing selected ratio, low-current specification and burden. Any disconnection or rewiring belongs to qualified personnel under the site safety procedure; do not hide a wiring error by changing settings just to match a desired total.

Maintain a measurement-point register

After acceptance, preserve point ID, panel and circuit, CT model and ratio, secondary rating, class, burden calculation, meter model and settings, network address, drawing revision and SAT evidence. Circuits and machines are moved; CTs, meters and gateways are replaced. Without a register, a step change in energy data can be mistaken for a process improvement. Log changes to CT ratio with time and old and new values so historical data remain interpretable.

Value starts when trustworthy measurements support action, not when the display first lights up. Include the intended data users in procurement decisions on minimum load, aggregation and alarm conditions. The power meter data collection guide covers the downstream path.

Frequently asked questions

What CT primary rating should a factory power monitoring project choose?

Use measured idle, normal and peak current, documented expansion and the candidate’s guaranteed current range and burden conditions. A breaker rating alone is insufficient. Avoid excessive primary rating where standby accuracy matters.

Is a 1 A or 5 A CT better for an energy meter?

Start with the meter input. At equal loop resistance, rated-current cable loss at 1 A is one twenty-fifth of that at 5 A, but an existing 5 A-only meter or a voltage-output sensor changes the choice. Quote CT, meter and wiring together.

Does CT burden include the secondary cable?

The total burden compared with the CT’s accuracy rating includes meter input and secondary cable loss. Calculate resistance for the outgoing and return conductors, and account for actual connections and conditions.

May a meter be disconnected with the CT secondary open?

Not for a current-output CT while primary current flows. Hazardous voltage may result. Only qualified personnel following product-specific and site safety procedures should manage the shorting and isolation arrangement.

What belongs in FAT versus SAT?

FAT checks the specified components, drawings, terminals, configuration and bench signals. SAT confirms installed phase relationships, real-load readings including the low-load case, energy integration and transmitted data. Agree pass criteria at RFP stage.

Summary

A factory-monitoring CT is selected as part of a whole circuit: primary range, meter input, low-load accuracy, round-trip cable burden, panel installation, polarity, safe service and FAT/SAT. A measurement-point table and wiring drawing let bidders quote and test on equal terms. The two figures most often missing from early estimates are the measured minimum current and secondary loop resistance.

If you are planning CT and meter replacement and need help defining measurement points or an RFP, contact TOMAS TECH. A discussion can begin with the existing panels, data path and acceptance requirements before a model is fixed.

Sources

• Schneider Electric: CT selection inputs.
• Schneider Electric: accuracy class and burden.
• Schneider Electric: remedying insufficient burden.
• Schneider Electric: CT-to-meter distance.
• Schneider Electric: CT installation and polarity.
• Schneider Electric: CT shorting blocks.
• Schneider Electric: checking inaccurate meter readings.
• ABB: Low-voltage current transformers, IEC 61869-2 accuracy table.