Protection, safety & stability

This page checks that the protection chain works: that each device trips, that it trips before the device upstream of it, and that it is sensitive enough to see the smallest fault it has to clear. It is built on studyOptions.enableProtectionCoord plus the protection study, with the trip-curve library engines/tcc-library.js as the single source (IEC 60255-151 inverse-time families, electronic trip units with their I²t long-time law, and gG / aM fuse tables). The page shows the time-current curves, the coordination time interval ladder between adjacent devices, the selectivity and sensitivity verdicts, the printable setting sheet, the ZSI zone-interlock check per IEC 60947-2 and the negative and zero-sequence overcurrent sensitivity check that reads the sequence networks, with the arc-flash analysis sharing the same curve library for its clearing time. Standards referenced: IEC 60255-151, IEC 60947-2, GB/T 14048.2 and IEEE 242.

Protection that does not coordinate fails in two directions: too slow or too high a setting and the fault is not cleared, too much overlap and the upstream device trips first, taking out a healthy part of the installation. Selectivity is what keeps a single fault from becoming a site-wide outage, and sensitivity is what keeps a high-impedance earth fault from being missed entirely — which is also the link to the personnel safety page, because the required trip current and the clearing time come from the same curves. The setting sheet is the document the commissioning team actually works from, and the coordination curves are what a design review asks to see.

Input: the device chain from source through transformer, switchgear, breakers and loads, the rated current and settings of each device (thermal setting multiple, long-time delay, magnetic trip multiple, short-time delay), and the fault currents from the short-circuit study → chain: the curve library returns the trip time for each current (the thermal inverse-time form is the I²t law evaluated up to a cap, fuses are interpolated from their log-log tables), the coordination time interval is the difference between the upstream and downstream operating times and is judged against the minimum interval, the sensitivity check compares the minimum fault current with the setting and judges against the minimum sensitivity ratio, the setting ladder is derived for devices that have no explicit configuration, the sequence protection check compares the negative and zero-sequence currents from the sequence networks with their settings, and the ZSI check verifies that the zone-interlock signals do not create a gap → output: the time-current curves as vector charts, the coordination ladder table with the margins, the selectivity and sensitivity verdicts, and the protection setting sheet that can be printed or exported. Linkage: the short-circuit current (and therefore every section, rating and Ssc change) moves the sensitivity and the coordination margins; a setting change feeds back into the personnel-safety required current and into the arc-flash clearing time, because all three read the same curve library; the transformer capacity and the device chain decide the default setting ladder; and the negative and zero-sequence sensitivity is tied to the earthing arrangement, so changing the transformer vector group or the neutral earthing moves it. Approximations, all disclosed: the thermal element uses the simplified I²t inverse-time law rather than a full thermal model, and the curves are standard families interpolated from the library.

On the example scheme the engine grades 9 cascaded pairs and 3 of 9 pass: MV-SWGR releases at 0.888 s while LV-MAIN releases at 0.592 s, a CTI margin of 0.30 s (full selectivity), but QF-LOAD-1 to QF-LOAD-2 shows a 0.00 s margin and must be re-graded. The same run gives arc-flash incident energy 9.56 cal/cm2 (PPE class 3, boundary 231 mm) at 22.67 kA and 0.2 s, N-1 where 23 of 24 outages lose load (worst 6250 kW, SAIDI 9.6 h, availability 99.89%), and a cable thermal check requiring 157 mm2 against 50 mm2 installed.

IEC 60255-151 (IDMT curve families SI/VI/EI/LTI), IEC 60909-0 (fault current feeding the coordination and thermal checks), IEEE 1584-2002 (incident energy, arc boundary, PPE class), IEEE 80 / GB/T 50065 (allowable touch and step voltage), IEC 60364-5-54 (earthing), GB/T 16895.5 and IEC 60364-4-43 (S ≥ I·sqrt(t)/K, k = 143 for copper), IEC 60947-2 / IEC 61439 (Icw and peak withstand tiers 25/31.5/40/50/65 kA) and GB/T 14549-1993 for harmonic limits (shown on the power-quality page).

Why is the protection verdict on this page the same as in the report?

Because both read one source: engines/protection-coordination.js coordinate(). The card shows the study response (levels, curve samples, CTI checks) and the report reads design.prot.checks from the same run, so the pass count and the failing pairs cannot drift apart.

What is the difference between the withstand card and the thermal-stability card?

The withstand card classifies every device (switchgear, transformer, cable) against its rated value — required Icw / dynamic peak / thermal section versus rated, with the margin and the recommended standard tier, and it says whether the rating came from the configurator, from a value typed on the page, or is not available. The thermal-stability card is a narrower cable-only check: S ≥ I·sqrt(t)/K with k = 143 for copper, applied to the actual fault current and clearing time.

Where does the energization inrush on the sequence card come from?

From the energize event in the scenario sequence: the engine estimates Ip_peak = k·sqrt(2)·In with k clamped to 3 to 12 from the closing angle, remanence and vector group, decays it as Ip·exp(-t/tau) with tau 0.1 to 1 s, and reports the 2nd-harmonic share that decides whether differential protection needs harmonic restraint. In the example, energizing the 2500 kVA transformer gives 45.93 kA peak, tau 0.619 s and 39% 2nd harmonic, so restraint is required.