In short. A coordination study chooses settings so that, for any fault, the device nearest the fault trips first and the rest wait, while cables and equipment are protected. The workflow runs from verified data through fault levels and curves to documented settings and field verification.
Workflow
- Scope and criteria. Decide the boundary, standards, grading margins and objectives. See setting philosophy.
- Collect data. Single-line diagram, device types and models, existing settings, CT ratios, cable and transformer data, motor data.
- Verify on site. Compare nameplates and settings to the drawings. See inputs and validation.
- Fault levels. Maximum and minimum for each operating mode. See maximum and minimum.
- Load and inrush data. Full-load, starting, transformer inrush and cold-load pickup currents.
- Equipment damage curves. Cable thermal limits, transformer damage curves, motor stall and starting curves, generator damage curves.
- Plot curves. Time-current curves of devices from load to source. See TCC engineering.
- Set devices. Starting from the load end, set each device with margin above downstream devices and below damage curves.
- Resolve conflicts. Change devices, add zone interlocking or differential, or accept partial selectivity with a recorded reason.
- Review modes. Repeat for utility, generator and parallel.
- Document. Settings tables, curves, assumptions and notes.
- Implement and verify. Apply settings, test by injection, and record as-left values.
Deliverables
- Time-current curve sets per path, per mode.
- Settings schedule by device, with tags.
- List of coordination issues and decisions.
- Arc-flash input data. See arc-flash study.
Grading margins
Margins reflect the downstream device’s clearing time, the upstream relay’s tolerance and overshoot. Use manufacturer data and the philosophy document rather than a fixed rule.
Typical findings on generator sites
- Instantaneous settings above generator fault current.
- Overlapping curves between main and feeders.
- Earth-fault elements set for utility fault levels.
- Cable protection not met at the minimum fault.
- Undamped inrush trips.
Illustrative example
A site with a main breaker and eight feeders trips the main when any feeder faults. The study finds the main’s short-time delay is zero and its pickup is below the feeder fault level. Setting a short-time delay and a higher pickup restores selectivity, and the study checks that the delay still protects the busbar and the upstream cable.
A coordination study in eight steps
- Obtain the verified single-line diagram and device data.
- Collect equipment damage curves: transformer, cable, motor and generator.
- Run load flow and short-circuit studies for load, inrush, maximum and minimum faults.
- Plot device curves on a common current base.
- Set the device nearest the load first, then work towards the source.
- Apply time margins between devices, commonly 0.2 to 0.4 seconds depending on device type.
- Check each setting against the damage curve and the minimum fault.
- Record settings and issue relay files.
Data you must have
| Data | Why |
|---|---|
| Device make, model, curve family | Curves differ between makers |
| CT ratios | Convert primary to secondary settings |
| Transformer inrush | Avoid false trips on energisation |
| Generator decrement curve | Verify pickup on decayed fault current |
Operating scenarios
Coordinate for every state that matters: mains, generator, parallel. A setting that is selective on the mains may lose selectivity on the generator because the fault current is much lower.
Deliverables
- TCC plots for each path.
- Setting table for every relay and adjustable breaker.
- List of unavoidable miscoordination and the reason.
See TCC engineering and coordination and selectivity.
Common mistakes
- Copying settings from similar sites.
- No site verification of existing settings.
- No record after implementation.
What to send us
Send relay and breaker data. Send your application details. See also coordination and selectivity.
Questions engineers ask
How long does a coordination study take?
It depends on size and data quality. Data collection and verification often take longer than calculations.
Can I coordinate without software?
For simple radial systems, yes with manufacturer curves. Larger systems benefit from software.
Is full selectivity required?
For critical loads, it is the goal. Partial selectivity may be acceptable elsewhere with documented reasons.
When should the study be updated?
After changes in sources, loads, protection devices or settings.
This guide gives general principles. Settings must be verified on site.