Sequence Fault Visualizer

How fault currents split into positive, negative and zero sequence across a transformer

1

Fault

Fault type
Faulted phase
The fault touches

Where on feeder 1

2

Transformer

High-side neutral
Low-side neutral

This view is a delta – grounded wye station transformer, both neutrals solidly grounded, to a 12.47 kV bus.

More transformer data
3

Utility source

Source neutral
More source data
4

Low-side feeders

Low-side fault

Starting estimates, per phase to ground: shielded underground cable about 1.0 µF/mi for 5 kV class (roughly 0.7 to 1.5) and 0.6 µF/mi for 15 kV class (roughly 0.3 to 0.8), by size and insulation; overhead line about 0.01 µF/mi (zero sequence, roughly 0.006 to 0.01). Capacitance matters most when the low side is ungrounded.

5

Feeder 1 to the fault

Overhead construction
Lengths, from the bus to the fault
Conductors

Grounds
Return path

Grounds as built: one resistance and one spacing for every ground.

Feeder 1 Details

Feeder 1 impedances

Scroll sideways to see every column.

6

Ground relay and branch fuse

51N/G ground relay

Distribution transformers

With every transformer connected phase to phase, load puts no residual current in the feeder CT on a 3-wire or a 4-wire line. Phase-to-neutral transformers, possible only where there is a neutral, put their unbalance there.

Time to trip

Branch-line fuse

Fuse link

Per unit on the transformer rating. Prefault voltage 1.0 per unit at the source, no load; the cables' charging current flows before the fault. Angles are referred to the source phase A voltage.

Scenarios

Each scenario adds to the others; click it again to undo.

What the base case is

12.47 kV feeder 1: 4-wire, hybrid and 3-wire side by side

The same station, feeder, grounds, fault and settings on each construction. Click the feeder to move the fault.

Scroll sideways to see the whole diagram.

Getaway cable mi
Overhead trunk mi
Overhead lateral mi
Underground dip mi

    COG (IEEE coefficient of grounding): the highest line-to-ground voltage on an unfaulted phase during a fault to earth at a location, as a percentage of the line-to-line voltage there before the fault (IEEE C62.92.1). 100 % would put an unfaulted phase a full line-to-line voltage above the reference; with no fault it is 57.7 %. The kilovolts are shown beside it. IEEE C62.92.4 calls a system effectively grounded where the COG is 80 % or less; GO 95, GO 128 and the NESC use that term for the grounds themselves, so the tool gives the number, not a verdict. The tool takes the worst single or double line-to-ground fault there, through any resistance from 0 to unlimited, measured to remote earth. Inside a cable the fault goes to the cable's neutral, so there the COG is measured to the cable neutral. Arrester voltage (TOV) is the same kind of figure for the worst fault to what an arrester at the fault is bonded to, measured to it: the neutral, or the pole's or object's own ground. It is not the COG: it is the voltage across the arrester, and with how fast the fault clears it sets the arrester rating. Both are percentages of the line-to-line voltage, so 80 % of 12.47 kV is 9.98 kV, 1.39 per unit of the line-to-neutral voltage.

    Source, transformer and feeders

    Source, Transformer and Feeders Details

    The same fault on each construction

    Scroll sideways to see every column.

    IEEE C62.22's commonly applied ratings at 12.47 kV are 9 or 10 kV on a four-wire multigrounded system and 15 kV on a three-wire solidly grounded one.

    The rise is the voltage of the neutral or the object to remote earth while the fault lasts, until the first device to clear opens. It is not the touch or step voltage a person would bridge.

    Equipment ratings to check

    Scroll sideways to see every column.

    Each row is this case's duty against a published limit, not a choice of approved equipment. "Meets" and "exceeds" check that duty only; "not assessed" marks what a 60 Hz fault calculation cannot settle, such as impulse, switching, contamination and a maker's temporary-overvoltage curve. Two constructions can need the same equipment class with a different margin. A cable, its insulation level and its elbows and splices go together: another insulation level changes the diameter over the insulation, and so the accessories' range.

    Requirements by construction

    What the codes ask that differs between the 4-wire, the hybrid and the 3-wire, with this case's values where the tool has them.

    Scroll sideways to see every column.

    Summaries of GO 95, GO 128, the NESC (2023 edition) and IEEE 1547, not the rules' full text or a compliance review. Only the values set against a limit are checked; the rest is not assessed. The governing edition and the utility's own standards decide. The COG does not settle any of these rules.

    Along feeder 1: voltages and currents

    The same fault moved from the bus to the end of the feeder, on each construction. The dotted orange line and the triangle mark where it is now.
    The numbers

    Clearing times: branch fuse and ground relay

      Fuse selection

      Scroll sideways to see every column.

      The numbers along feeder 1

      Three-line diagrams

      One for each construction. Line width follows current. Click a set of CTs, the windings or the fault to see its phasors.

      Scroll sideways to see the whole diagram.

      Phase APhase BPhase C No currentFault● polarity dotValues beside the coils are winding currents

      Sequence networks

      Per unit. The fault connects the three networks at their F and N terminals.

      Phasors

      The same point on each construction.
      Show

      positivenegativezero

      Not in this comparison

      For explaining and testing sensitivities; it does not rank the constructions.
      • On a lateral fault, the trunk beyond the tap (drawn on the one-line as "trunk continues") has no neutral or grounds in the model.
      • Touch and step voltage: the rise shown is to remote earth, not what a person bridges.
      • A conductor broken on the load side and fed back through phase-to-phase transformers.
      • Ferroresonance when phase-to-phase transformers on cable are switched one pole at a time.
      • Neutral-to-earth voltage where there is phase-to-neutral load; only its standing residual's effect on the ground pickup is here.
      • A transformer winding failing to its secondary neutral, and what a primary contact does on customer premises beyond the object's own ground (the preset "Primary on the secondary neutral or a transformer tank" models the contact itself).
      • Lightning performance, and the insulation margins that follow from the arrester rating.
      • Resonant grounding, which needs a 3-wire system.
      • Equipment, construction and operating cost.

      How the comparison is built, and what the tool leaves out on purpose, is in How this works.

        Calculation sheet: each step, its equation and its value

        Print report prints it after the results.

        Scroll sideways to see every column.

        Basis. Symmetrical components (Fortescue, 1918), worked in per unit on the station transformer's rating with the nominal line-to-line voltages as bases; complex quantities are written a + jb, or as magnitude ∠ angle in degrees. Prefault voltage 1.0 per unit at the source, no load current. Each feeder's capacitance to ground is lumped at the low-side bus. Line and cable impedances from Carson's equations with 100 Ω·m earth, for the conductors chosen, on the pole of Kersting's Example 4.1 and the cable arrangement of his Example 4.2, averaged as if transposed; ACSR resistance at 50 °C and GMR from the conductor table Kersting uses (Westinghouse T&D Reference Book), cable dimensions from 15 kV URD catalogs. Effective grounding to IEEE C62.92.1; arrester ratings to IEEE C62.11 and their application to IEEE C62.22; relay curves to IEEE C37.112; fuse times from S&C Positrol curves (TCC 165-6, 165-6-2, 170-6 and 170-6-2). The values marked as found by the computer, the ladder of grounds and the worst coefficient of grounding by each of its two measures, are exact solutions of the circuit drawn, not estimates. All other values are worked from the equation on their row, and the tool checks that they agree with its own solution. What the model leaves out is listed under How this works.

        Three-line diagram

        Line width follows current. Click a set of CTs, the windings or the fault to see its phasors.

        Scroll sideways to see the whole diagram.

        Phase APhase BPhase C No currentFault● polarity dotValues beside the coils are winding currents

        Sequence networks

        Per unit. The fault connects the three networks at their F and N terminals.

        Phasors

        Show

        positivenegativezero

        Try it: drag three phasors and see their sequence parts

        Drag the tips of a, b and c. Any three phasors can be split into a balanced positive set, a balanced negative set and a zero set that is the same in all three phases.

        I0 = (Ia + Ib + Ic) / 3  ·  I1 = (Ia + a·Ib + a²·Ic) / 3  ·  I2 = (Ia + a²·Ib + a·Ic) / 3  ·  a = 1∠120°

        How this works

        The tool solves one radial system: a utility source, a two-winding transformer and a fault at the high-side or low-side bus. It uses symmetrical components (Fortescue, 1918). The three sequence networks are built in per unit on the transformer rating. They are connected at the fault as the fault type requires, and the sequence currents are turned back into phase currents at every point in the diagram.

        • Source: a Thevenin equivalent with Z1 = Z2 from the three-phase short-circuit MVA and X/R, and Z0 = (Z0/Z1)·Z1 when the source neutral is grounded. An ungrounded source has no zero-sequence path.
        • Transformer: leakage impedance Z1 = Z2 = %Z at the given X/R, and Z0 = (Z0/Z1)·Z1. A resistor in a grounded neutral appears as 3R in the zero-sequence network: the neutral carries 3I0 while the per-phase network carries I0, so the drop 3I0·R is drawn as 3R.
        • Zero-sequence path by connection: grounded wye – grounded wye passes zero sequence straight through. A grounded wye with a delta gives a path to the reference on the wye side only: zero-sequence current circulates in the delta and never reaches the lines on the delta side. Seen from its own terminals, a delta or an ungrounded wye gives no path, so a delta-delta bank passes no zero sequence in either direction.
        • Phase shift: IEEE C57.12.00 (C57.12.01 for dry-type) requires the low side of delta-wye and wye-delta transformers to lag the high side by 30°; in IEC 60076-1 notation these are Dyn1 and YNd1. Positive-sequence quantities shift by −30° from high side to low side and negative-sequence quantities by +30°. Dyn11 is the reverse. YNyn0 and Dd0 have no shift.
        • Feeders and cable capacitance: each feeder on the low-side bus is its capacitance to ground per phase (length × µF per mile), lumped at the bus; its series impedance is left out. The capacitance is a shunt in all three sequence networks (phase-to-ground capacitance, so C1 = C2 = C0), and its charging current flows before the fault. Leaving out the feeders' series impedance for the capacitance is fair: a feeder's capacitive reactance to ground is thousands of ohms against an ohm or two in series. Feeder 1's path to the fault is modeled (below). On an ungrounded system the capacitance is the only return path for ground current: a bolted fault draws about 3ωC0·V_LN, the faulted feeder's CT sees the charging current of all the other feeders, and each healthy feeder's CT sees its own, in the opposite direction. The direction shown at each feeder CT is the sign of the zero-sequence reactive power, Im(3V0·3I0*), the quantity a sin φ (varmetric) directional ground element uses; it is shown only when the low side has no grounded neutral.
        • Feeder 1, from the bus to the fault: up to four sections in this order: an underground getaway, an overhead trunk, an overhead lateral and an underground dip to the fault. The overhead construction is 4-wire multigrounded, 3-wire unigrounded, or a hybrid with a 4-wire trunk and 3-wire laterals. The conductors are chosen for each section. The overhead line is worked on the pole of Kersting's Example 4.1 (its own conductors, 336.4 ACSR phases and a 4/0 ACSR neutral, are the defaults) with ACSR from 795 kcmil to #2; the cable as Kersting's Example 4.2 (its own cable, 250 kcmil aluminum with a one-third neutral of 13 No. 14 copper, is the default) in 1/0 AWG (full neutral), 250, 500 and 1,000 kcmil aluminum, averaged as if transposed, with the three concentric neutrals acting as one bonded neutral. A neutral size applies where a section is 4-wire, and in the 4-wire columns of the comparison. Soil 100 Ω·m. Line capacitance stays lumped at the bus.
        • Ideal return: as in earlier revisions. Each neutral is grounded continuously, so its effect is folded into Z0 by Kron reduction, and the earth contact and the station grid are 0 Ω. On a 4-wire section the neutral carries |zn/znn| of the ground current: about 46 % with the default 4/0 ACSR neutral (the earth about 58 %), from 32 % with a #2 neutral to 53 % with 336.4; on a cable the concentric neutral carries 92 % to 99 %, by size. The two parts add as phasors. On a 3-wire section the ground current returns through the earth alone.
        • Grounds as built: each neutral on the path is grounded through the same resistance (each ground) at the same spacing, and the transformer neutral reaches remote earth through the station ground grid. The neutral and its grounds form a ladder. The tool solves it, exact for the circuit drawn, replacing each node from the far end inward by the equivalent of everything beyond it, and adds it to the line's zero-sequence impedance as three times the ground-return impedance. So the neutral's share of the ground current changes along the line: largest near the fault, where the current enters it, and smaller toward the station, as the grounds pass part of it to the earth. A cable bonded to an overhead neutral joins that neutral; a cable cut off from the station by a 3-wire section is an island that returns its current through its own grounds. The path is cut into at most 400 segments: on a longer path the spacing is widened, and the result says so. In the 12.47 kV view the feeder stays as built beyond a fault part-way along: its neutral and grounds stay where they are (moving the fault never moves a ground), the fault falls on a node with no ground, and the phases carry fault current only up to the fault, so their coupling to the neutral counts only there. In that view the other feeders at the station (the number of feeders less one) are copies of feeder 1 in each column, and their neutral runs from the station, with their grounds, join the station grid: the station's path to earth is the grid's resistance in parallel with them, Rsub ∥ Zin/(n − 1), where Zin is one run seen from the station. That run is the getaway's, the trunk's and the lateral's neutral on the 4-wire column, the getaway's and the trunk's on the hybrid, and on the 3-wire only the getaway cable's neutral to its ground at the riser (none when there is no getaway). They carry no fault current in their phases and are not coupled to feeder 1. The General view keeps one feeder, ended at the fault.
        • What the fault touches: with the grounds as built, a ground fault on feeder 1 is either to a grounded object (a pole ground, a crossarm, a riser) or a conductor lying on the ground. On a 4-wire line or a cable the object is bonded to the neutral, so the fault current enters the neutral. On a 3-wire line the object has only its own ground, in series with Zf and the station grid. A conductor on the ground touches no neutral: Zf is its contact resistance, and the current enters the earth. A fault on a cable always reaches its concentric neutral.
        • What each voltage is referred to: at the fault, the neutral there when the fault current enters it, otherwise remote earth; at the low-side bus, the station ground. The rise of the neutral, the object or the grid is its voltage above remote earth while the fault lasts. For an object on a section with no neutral the unfaulted phases are given to the object's own ground, which is what equipment on that structure sees, and to remote earth beside it. The coefficient of grounding is given by two measures, each to its own reference (below), and constructions are compared measure by measure.
        • Fault surfaces: typical currents for a conductor down on each surface on a 12.5 kV feeder (7.2 kV to ground), from IEEE PSRC Working Group D15, High Impedance Fault Detection Technology (1996), turned into a resistance. Real high-impedance faults arc, flicker and change; the tool shows their steady 60 Hz equivalent. Each label gives the current that surface would draw at the present voltage, leaving out the system's own impedance.
        • Voltage rise and arresters: the highest unfaulted-phase voltage to ground during a ground fault, at the fault and at the bus, in per unit of the nominal line-to-neutral voltage and to what the fault is on. The coefficient of grounding at the fault location is the highest unfaulted-phase voltage for the worst single or double line-to-ground fault there, over fault resistance from 0 to unlimited, divided by the line-to-line voltage there before the fault (IEEE C62.92.1); kilovolts stay absolute. IEEE C62.92.4 calls a system effectively grounded where it is 80 % or less; the tool shows the COG and its kilovolts, not a verdict, because GO 95, GO 128 and the NESC use the term for the grounds themselves. On a real feeder ground is not one thing, so two figures are worked. The COG, IEEE's coefficient of grounding: a fault to earth, measured to remote earth, with the zero-sequence impedance through the earth. Each card and the table show it with the kilovolts it means. The arrester voltage (TOV): a fault to the local ground, measured to it, which is the voltage across an arrester bonded to that ground: the neutral or the cable's neutral where there is one, with the zero-sequence impedance to it, otherwise the pole's or object's own ground, through its resistance (the input Ground of a pole or object with no neutral). It is not the COG; with how fast the fault clears, it sets the arrester (below). Where there is a neutral the two are close; with no neutral, the arrester voltage is higher, because the object's ground rises with the fault. A cable fault goes to its neutral, so in a cable the COG is measured to the cable neutral, and the two figures are the same. The constructions are compared COG with COG and arrester voltage with arrester voltage, never one construction's COG against another's arrester voltage. As the fault resistance changes, each unfaulted-phase voltage moves on a circular arc, so the highest point is found exactly, not from a few sampled resistances. Near 80 % the value is shown with enough decimals to fall on the right side of 80 %. The coefficient depends on Z0 at the location, so it carries the transformer's zero-sequence model (below). Also shown is the approximate test X0/X1 ≤ 3 and R0/X1 ≤ 1, which assumes R1 is small against X1. The arrester is sized from the voltage across it, as described below. Ratings shown are the IEEE C62.11 distribution ratings, for 2.4 to 34.5 kV systems. On an ungrounded system a ground fault can stay on until it is found, so MCOV is usually chosen at least equal to the line-to-line voltage; a fault through some resistance can even lift the leading unfaulted phase a little above line-to-line voltage (to about 1.82 per unit in this lossless model, a coefficient of grounding near 105 %).
        • Arrester suggestion: with the grounds as built, every line-to-ground and double line-to-ground fault at the location, through any resistance from 0 to open, is taken in two families: to the local ground, measured to it, always; and to earth, measured to remote earth, where there is no neutral at the fault. For each fault the tool finds the voltage across an arrester there and whether the modeled protection clears it: the branch fuse on its published total-clearing curve, or the ground relay and the breaker. A fault below the relay's pickup that the fuse does not melt, or melts only off the end of its curve, stays on. The MCOV needed is the largest of 1.05 × V_LN, 1.05 × V / k over the faults that are cleared and 1.05 × V over the faults that stay on, where k is the temporary-overvoltage capability entered as a multiple of MCOV (IEEE C62.22); the suggestion is the smallest standard rating with that MCOV, shown with the fault that sets it, its voltage, and whether and when it is cleared. 1.25 is a placeholder: read the capability from the maker's curve at the clearing time shown. With the ideal return, or on an ungrounded system, there is one reference and no clearing model, so the suggestion is the smallest rating whose MCOV covers 1.05 × V_LN and whose capability covers COG × 1.05 × V_LL, and its row says so; above 80 % the rule MCOV ≥ 1.05 × V_LL is shown as well, and on an ungrounded system the suggestion is that rule.
        • Ground relay: a residual (3I0) ground pickup at feeder 1's CT. It sees a fault when 3I0 there is at least the pickup. The largest fault resistance it sees at this location is for a conductor on the ground: the resistance at which the line-to-neutral voltage divided by the magnitude of (Z1 + Z2 + Z0)/3 + R equals the pickup, with Z0 as a conductor on the ground sees it. The connection of the distribution transformers sets the standing residual: with every transformer connected phase to phase, load puts no residual current in the CT on a 3-wire or a 4-wire line; phase-to-neutral transformers put their unbalance in the neutral and so in 3I0, and the pickup has to stay above about 1.5 times that residual. Directional elements and 3V0 are not covered here.
        • Branch-line fuse and the relay's time: the branch fuse sits where the lateral leaves the trunk, or at the riser to the dip when there is no lateral. For a fault beyond it, the fuse carries the fault's largest phase current and the relay sees 3I0 at the feeder CT; for a ground fault on a radial feeder with no load the two are nearly equal. The fuse times come from S&C's published curves for Positrol fuse links (K speed, TCC 165-6 and 165-6-2; T speed, TCC 170-6 and 170-6-2), read by straight-line interpolation in log-log: minimum melting is the earliest the link can melt, total clearing the latest it clears. Below the current at the top of the curve (300 s, or 600 s for 140 and 200 A) the link does not melt. The relay's inverse-time element follows t = TD × (B + A / (Mp − 1)) with M = 3I0 / pickup, for the U.S. curves U1 to U5 (time dial 0.5 to 15): U1, U3 and U4 have the IEEE C37.112 moderately, very and extremely inverse shapes, U2 and U5 the inverse and short-time inverse shapes of induction-disc relays. The instantaneous element, when set, trips in about one cycle, and the breaker adds its interrupting time. The fuse clears first when the relay would still need at least the coordination margin after the fuse has cleared. The relay is first when the breaker opens before the fuse can melt. In between, both can operate and the whole feeder goes out. The comparison table, the typical links and the curve along feeder 1 use the same rule.
        • Comparison and curves: the comparison table solves the same fault with the same station, lengths, grounds and fault on each overhead construction (the hybrid only when there is both a trunk and a lateral). The curves move the fault along the path from the bus to where it is now, cutting the path at each point, on each construction.
        • 12.47 kV: 3-wire vs 4-wire: the second view holds the station to a delta – grounded wye bank to 12.47 kV with solidly grounded neutrals, the grounds as built and a ground fault (L-G or L-L-G) on feeder 1, and always solves the 4-wire multigrounded, hybrid and 3-wire unigrounded constructions side by side. Feeder 1 is entered as built, getaway, trunk, lateral and dip, and the fault is placed anywhere along it. The series path is cut at the fault; the feeder beyond it stays in place with its neutral and grounds, and the other feeders are built as feeder 1 is (Grounds as built, above). So the view and the General view differ for the same path: for a fault part-way along the view gives a little more current on a line with a neutral and a lower rise, and at the station the other feeders' neutrals share the grid's current. Ahead of the fuse the hybrid is nearly the 4-wire line: its trunk is the same, but its other feeders have less neutral. A named location follows the lengths when they change; a distance typed or picked on the drawing stays where it is. The view takes the whole width: the scenarios run across the top, the summary of each construction sits under the one-line's length sliders and above the fault, and each section of feeder 1 on the one-line carries its own conductor drop-downs and a length box, with the four length sliders side by side under it (the same values as under Feeder 1, so either place can be used). Under the one-line, each construction is drawn as a side view on the same distances: the station (the transformer and a metal-clad switchgear lineup, feeder 1's cubicle marked with its 51N ground relay and the other feeders' cables leaving the others), the getaway cable under the ground, the poles with the phases and, where the construction has one, its neutral in the construction's colour with its grounds, the fuse's cutouts on the tap pole, and the fault as it is (a conductor on the ground, a conductor on a grounded object, bonded to the neutral where there is one, or a cable fault); clicking a side view moves the fault too. The pictures are not to scale. The ground-fault current coming back is drawn from the solution: on the neutral, span by span from the ladder of grounds, and through the earth, the rest of the ground current (across any cut between the station and the fault the two carry it together), each with a width to one scale for the three constructions and arrows toward the station (beyond the fault, where a neutral continues, it carries current away from the fault and the earth brings it back), with the values beside the fault and at the station. Under the sliders come the fault, feeder 1 (lengths, conductors and grounds; Feeder 1 Details opens a table of each section's impedances), the source, the transformer and the other feeders in one panel whose settings open with Source, Transformer and Feeders Details, and Protection, with the 51N/G ground relay and the branch-line fuse. The base case's station is a 1,000 MVA source and a 30,000 kVA transformer at 13.33 %Z; the General view's 12.47 kV examples keep their own 800 MVA source and 20,000 kVA transformer at 8 %Z. The chips set only the fields they name: a fault type sets the type, Zf and what the fault touches; a location places the fault; a scenario sets its own fields and, turned off, returns them to the base case. The base case's relay settings (U4, time dial 3, 240 A, no instantaneous) and its 65T fuse are typical starting values, not a recommendation. The curves along the feeder run from the bus to the end of the feeder on each construction, with the fault marked where it is. The view keeps its own inputs, separate from the General view, and both are remembered in this browser.
        • Calculation sheet and printed report: the sheet works the fault on screen by hand, step by step, from the equations it shows, in per unit on the transformer rating, with the numbers substituted under each value. For a ground fault on feeder 1 it works the 4-wire, hybrid and 3-wire constructions side by side. Values that come from the computer are marked: the ladder of grounds and the worst coefficient of grounding by each measure, all exact for the circuit drawn; for an object's own ground the sheet also shows the formula. Each time the sheet is shown it is compared with the tool's own network solution, and the line under its title says whether they agree. Print report prints the results and then the sheet on letter pages, with the project and case typed in the toolbar; the diagrams and these notes can be added.
        • Fault: prefault voltage 1.0 per unit at the source and no load current. Zf is in each phase to a common point for a three-phase fault, from the phase to ground for L-G, between the phases for L-L, and from the joined phases to ground for L-L-G.
        • Directions: source currents toward the high-side bus, high-side CT currents into the transformer, low-side CT currents out of the transformer, fault currents into the fault. Winding currents are drawn into the polarity dot on the high side and out of it on the low side, so the two windings of a unit are in phase. 3I0 at a set of CTs is the residual, the sum of the three phase currents at that point; a neutral current is what flows in a transformer neutral or a line neutral; the ground current at the fault is the current into the ground connection. They are the same current only when it has no other return path: feeder capacitance, a multigrounded neutral and the earth each take part of it.
        • Inputs: a blank field that the case uses, or a value outside the range the calculation handles, gives a message and no result. That range is up to 1012 in each field (1015 MVA for the source), with at least 10−9 for voltages, ratings, %Z, X/R and Z0/Z1, and at most 50 feeders. An X/R of 0 (no reactance at all) is refused rather than computed. Outside it the arithmetic can overflow; it is far wider than any real system. The ground relay's settings, the branch fuse, the coordination margin and the arrester's TOV capability take no part in the fault solution: an invalid value there is marked and reported in its own panel, and the rest of the result stays.

        Left out on purpose: capacitance on the high side and of the bus, transformer and surge capacitors; cable insulation losses (so feeder 3I0 angles come out exactly ±90°, where real relays need some angle margin); series impedance of the feeders other than feeder 1; arcing and restriking transients, which can reach several per unit on an ungrounded system; grounded-wye VTs and capacitor banks, which add their own zero-sequence path; cross-country faults on two feeders; load current; motor and generator contribution on the low side; DC offset and decrement; three-winding transformers; and the zero-sequence flux path of a three-legged core. Every transformer here is a bank of single-phase units. A three-legged core-type unit has a zero-sequence flux path through its tank that acts like a high-impedance delta winding, so, for example, a wye-wye unit with one neutral ungrounded passes some zero-sequence current where this model shows none; a five-legged core behaves much like the bank. For a delta-wye or wye-delta unit, enter its tested Z0/Z1. With the grounds as built: the model covers a delta – grounded wye station transformer with a grounded low-side neutral, and for any other connection the return stays ideal; in the General view the feeder ends at the fault, so grounds beyond it are left out, which makes the rise at the fault a little high, and the other feeders' neutrals are left out (the 12.47 kV view has both); one spacing and one resistance serve every ground on the path, so service grounds and metallic water pipes tied to the neutral are represented by a shorter spacing and a lower resistance; the feeders' charging current returns to the transformer neutral directly, a few amperes against hundreds or thousands on a solidly grounded system. On a 3-wire line the worst coefficient of grounding for a fault to earth (measure A) still starts at a fault resistance of 0; away from the station a 0 Ω earth contact is seldom met, so that figure is an upper bound there. Also not modeled: an open conductor and the backfeed through phase-to-phase transformers, ferroresonance, and distributed generation as a ground source. For the fuse and relay times: load current through the fuse and its preloading, ambient temperature, the fuse's damage curve (often taken as 75 % of the minimum melting time), reclosing and fast-curve (fuse-saving) sequences, relay and CT tolerances and overtravel (the coordination margin is meant to cover them), a fault current that changes while an arc burns, the feeder's phase overcurrent element (it also sees a large ground fault, and it is the only one for a fault without ground), and reclosers or sectionalizers on the line.

        Checking: the solver was compared with an independent phase-by-phase circuit model of the same system for every combination of connection, neutral grounding, source grounding, fault location, fault type and phase, with and without fault impedance.

        References: C. L. Fortescue, “Method of Symmetrical Co-Ordinates Applied to the Solution of Polyphase Networks,” AIEE Transactions, vol. 37, pp. 1027–1140, 1918. J. L. Blackburn, Symmetrical Components for Power Systems Engineering, 1993. IEEE Std C57.12.00-2021, angular displacement of three-phase transformers. IEEE Std 551-2006 (Violet Book), short-circuit calculations. IEC 60076-1:2011, vector group notation. W. H. Kersting, Distribution System Modeling and Analysis, Examples 4.1 (overhead line) and 4.2 (cable). IEEE PSRC Working Group D15, High Impedance Fault Detection Technology, 1996. IEEE Std C62.92.1 (effective grounding), IEEE Std C62.92.4 (neutral grounding, distribution), IEEE Std C62.22 (arrester application), IEEE Std C62.11 (arrester ratings), IEEE Std 80 (station ground grid). S&C Electric Company, TCC 165-6 and 165-6-2 (Positrol fuse links, K speed, curve data dated January 20, 1986) and TCC 170-6 and 170-6-2 (T speed, December 17, 1984). IEEE Std C37.112 (inverse-time overcurrent relay equations). IEEE Std C37.41 and C37.42 (distribution cutouts and fuse links).

        For learning and quick checks. For design work, confirm results with a full short-circuit study.