Distance Relay Reach Testing: Verifying the Characteristic, Not Just the Pickup

A distance element does not decide by how much current is flowing. It decides by the ratio of voltage to current, because that ratio is a proxy for distance along the line and is largely independent of how strong the source behind the relay happens to be. Testing one therefore means testing a shape on a plane, not a threshold on an axis — and that changes what a competent test plan looks like.

This article covers what the 21 element measures, the characteristics you will meet, what “reach testing” means as distinct from a pickup check, how to choose test points, the secondary-versus-primary ohm trap, and — importantly — what a steady-state sweep does not prove.

Key takeaways

• Impedance is the discriminant because it stays roughly constant for a fault at a given point regardless of source strength, whereas fault current does not.
• Reach testing locates the boundary of the characteristic. Confirming the element picks up somewhere inside the zone tells you almost nothing about the boundary.
• Two boundary methods cover most needs: constant-angle radial searches to map the edge, and fixed-point shot tests inside, outside and on it.
• Secondary ohms equal primary ohms multiplied by the CT ratio and divided by the VT ratio. Testing in the wrong basis puts every result out by that factor, and it is a common error.
• A boundary sweep proves the characteristic. It does not prove the scheme: SIR effects, fault resistance, load encroachment, mutual coupling and CVT transients need dynamic or system-based testing.

What the element measures

For a fault at distance d along a line of impedance Z per unit length, the voltage at the relay is approximately the fault current multiplied by d times Z. Divide the measured voltage by the measured current and the fault current cancels: what remains is the impedance to the fault. Double the source strength and the current doubles, the voltage doubles, and the measured impedance is unchanged.

That independence is the whole reason distance protection exists. It gives a relay a discriminant that is a property of the network geometry rather than of the operating condition, which allows fixed reach settings and predictable coordination between substations. Overcurrent protection has to be re-examined every time the source impedance changes; a distance zone does not.

The independence is approximate. Fault resistance adds a real component that was not in the line, load current flowing during the fault shifts the apparent impedance, and a weak source makes the relay voltage very small. Those confounders are discussed later.

The R–X plane and characteristic shapes

Plot the measured impedance as a point with resistance on the horizontal axis and reactance on the vertical. A fault on the protected line lies close to a ray from the origin at the line angle — the argument of the line’s positive-sequence impedance, typically in the region of 70 to 85 degrees for a transmission circuit, and lower for cable or distribution feeders. The relay’s reach setting is a distance along that ray; the characteristic is a closed region around it.

The shapes you will meet:

Characteristic	Shape	Practical note
Mho	Circle through the origin, diameter along the reference angle	Inherently directional; resistive coverage is limited by the diameter
Offset mho	Circle displaced to enclose the origin	Non-directional; used for backup and some power-swing applications
Quadrilateral	Four independent boundaries: reach, resistive limits, directional line	Resistive reach set independently of reach, so better for resistive faults
Reactance line	Horizontal-ish boundary limiting reach	A building block, tilted to avoid overreach on resistive faults
Directional line	Boundary through the origin separating forward from reverse	Determines which faults the zone will see at all
Blinder	Boundary limiting resistive extent	Keeps heavy load out of the characteristic

A practical numerical relay zone is usually the intersection of several of these elements rather than one textbook figure, and the reactance boundary is often deliberately tilted by a few degrees so that fault resistance pushes the apparent impedance out of, rather than along, the characteristic. When your measured boundary does not match the ideal shape you drew, check whether a tilt or a load-encroachment cut-out is configured before concluding the relay is out of specification.

The reference angle (variously called the line angle, characteristic angle or relay characteristic angle depending on the manufacturer) sets the orientation. It is a setting, and it may not equal the actual line angle — some engineers set it to the line angle, others to a compromise across several line sections. Test against the setting.

Zones and what they are for

• Zone 1 trips with no intentional time delay and is set deliberately short of the remote busbar. A margin below full line impedance absorbs line-parameter uncertainty, instrument transformer error and relay measurement error, so that no combination of them lets an instantaneous zone reach into the next circuit. A commonly cited starting point is around 80 percent of the line, but the actual margin is an engineering decision that depends on how well the line parameters are known and on whether the line is series compensated or mutually coupled — treat 80 percent as a convention, not a rule.
• Zone 2 overreaches the remote busbar, covering the remainder of the line that Zone 1 gives up, and is time delayed so that it coordinates with the remote end’s Zone 1. Its reach must be long enough to cover the whole line with margin and short enough not to overreach the shortest adjacent circuit at the remote bus, which is not always simultaneously achievable — that tension is where teleprotection schemes earn their place.
• Zone 3 and reverse zones provide remote backup or supply the directional logic that blocking and permissive schemes need. Reverse-looking zones are set to see behind the relay, and they exist to tell the far end something rather than to trip locally.

Timing grades between the zones, typically in the range of a few hundred milliseconds per step, must accommodate the remote breaker’s clearing time plus relay reset plus a margin. IEEE C37.113 is the application guide that develops this reasoning properly for transmission lines.

What reach testing actually means

Confirming that the element operates for an impedance somewhere inside Zone 1 proves the element is alive. It does not tell you where the boundary is, which is the only thing a reach setting specifies. Reach testing means locating that boundary and comparing it against the setting.

Two methods cover the ground.

Constant-angle (radial) searches

Fix the impedance angle and vary the magnitude along that ray until the operate/no-operate transition is found. The mechanics are a bisection: establish a magnitude that operates and one that does not, then repeatedly test the midpoint and discard the half that does not contain the transition.

lo = magnitude known to operate      (inside)
hi = magnitude known not to operate  (outside)

repeat:
    mid = (lo + hi) / 2
    apply mid at the fixed angle
    if operate:  lo = mid
    else:        hi = mid
until (hi - lo) <= tolerance

reported boundary = (lo + hi) / 2, with uncertainty +/- tolerance/2

Each halving buys one more bit of resolution, so ten iterations narrow the bracket by a factor of about 1000 — the search cost is low and there is no excuse for a coarse result. Run the search at the reference angle first, since that is the point the reach setting names directly, then at a spread of offset angles to map the shape.

Constant-impedance shot tests

Apply defined impedance points and record operate or no-operate, without searching. Choose points clearly inside the characteristic, clearly outside it, and just either side of the boundary at the tolerance limit. This is faster than a search, gives a clean pass/fail record, and is the natural form for the points that matter as no-operate results — load impedance, for example.

In practice a plan uses both: searches to establish the boundary geometry, shots to demonstrate margin at the points that carry a decision.

Choosing test points

• At the reference angle, for every zone. This is the point the reach setting names.
• At offset angles either side of the reference angle, enough to establish the shape. For a mho circle, points at the reference angle and roughly plus and minus 30 to 45 degrees from it already distinguish a circle from a quadrilateral.
• At the resistive extremes for quadrilateral characteristics, where the resistive reach setting is verified independently of the reactive reach.
• Near the reactive boundary at low resistance, to catch a tilt that has been set differently from what you assumed.
• In the load region — a no-operate test. Take the maximum load impedance the circuit will see, at the load angle, and demonstrate that no zone picks up. This test finds load encroachment problems and mis-set blinders, and it is worth more than several redundant points on the reach ray.
• Around the directional boundary, forward and reverse, at small impedance magnitudes. Directional behaviour close to the origin is where memory-voltage polarising and minimum-voltage logic live, and it is where relays differ most from one another.

For each point, record the applied impedance magnitude and angle, the fault type applied (phase-to-phase, phase-to-earth, three-phase), the expected result, the measured result and the operate time where relevant. Phase-to-earth loops involve the earth-return compensation factor, so a phase-to-earth test at the same impedance as a phase-to-phase test is not the same test — record which loop you exercised.

Secondary versus primary ohms

Reach settings may be expressed in primary ohms (the impedance of the actual line) or in secondary ohms (the impedance the relay measures at its terminals). The conversion:

Z_secondary = Z_primary * (CT ratio) / (VT ratio)

Example: 132 kV line, CT 600/1, VT 132000/110 (ratio 1200:1)
  CT ratio = 600
  VT ratio = 1200
  Z_sec = Z_pri * 600 / 1200 = Z_pri * 0.5

Line impedance 30 ohm primary at 80 degrees
  Zone 1 at 80 percent of line = 24 ohm primary
  Zone 1 in secondary ohms     = 24 * 0.5 = 12 ohm

The factor here is 0.5, so testing a 12 ohm secondary setting with 24 ohm values is a factor-of-two error — obvious once seen, easy to make when the setting file, the line data sheet and the test plan were each written by different people. It is more insidious when the conversion factor is close to 1, because the results look nearly right and get accepted.

Two guards are cheap. State the basis explicitly in every row of the test plan, not once in a header. And check one point by hand: the reference-angle Zone 1 boundary should come out at the setting value in the basis you claim to be using.

Confounders the sweep does not capture

A boundary sweep injects a steady-state voltage and current pair that produce the impedance you asked for. Real faults do not arrive that way.

• Source impedance ratio. With a strong source impedance relative to the line, the relay voltage during a remote fault can be very small. Relay measurement accuracy and the ability of the CVT to reproduce that voltage both degrade as SIR rises, and a characteristic verified at comfortable voltage levels may behave differently at the extremes. IEC 60255-121 addresses how distance performance is to be specified and tested including such influencing conditions.
• Fault resistance. Arc and tower-footing resistance add a real component, moving the apparent impedance to the right on the plane. For a short line with a long resistive excursion, the apparent impedance can leave a mho characteristic entirely even though the fault is on the protected line. This is the main argument for quadrilateral characteristics with independently set resistive reach.
• Load flow during the fault. Pre-fault load current combined with fault resistance skews the apparent impedance depending on power-flow direction, causing underreach in one direction and overreach in the other.
• Load encroachment. Heavy load at a low angle can enter a long Zone 3 or a wide quadrilateral. Blinders and load-encroachment cut-outs exist for this, and the no-operate load test verifies them.
• Mutual coupling on parallel lines. Zero-sequence coupling between circuits on the same towers changes the earth-loop impedance the relay measures, causing overreach or underreach depending on the coupled circuit’s loading and earthing state. A single-circuit steady-state test cannot show it.
• Series compensation. A series capacitor can make the apparent impedance to a fault look capacitive, producing voltage reversals and direction-measurement problems that steady-state injection at a fixed angle will never reproduce.
• CVT transients. A capacitive voltage transformer does not reproduce a step collapse in voltage instantly. For a close-in fault with high SIR, the transient in the CVT output can cause a transient overreach in the relay’s measurement. This is a waveform phenomenon; it is invisible to a sinusoidal test.

Be explicit about this in the report. A boundary sweep proves the characteristic. It does not prove the scheme. The distinction is not pedantry — it is the difference between a test record that supports a post-event review and one that does not.

What does exercise these effects is dynamic testing with computed pre-fault, fault and post-fault states, and system-based testing driven by simulated network models with realistic source impedances and transients. CT behaviour under those conditions is its own subject, covered in CT saturation and relay testing and in IEEE C37.110.

Timing and scheme tests

Beyond the characteristics: verify each zone timer at a point well inside its zone, so that timing is not entangled with boundary behaviour. Verify the reset behaviour too, since coordination assumes it.

Teleprotection logic needs its own tests. Permissive schemes (a remote signal permits a tripping decision the local relay has already made) and blocking schemes (a remote signal prevents one) each have a defined behaviour when the channel fails, and that behaviour — trip-on-loss or block-on-loss — is a design decision that must be tested, not assumed. Current-reversal and echo logic add states that only appear with specific sequences of signals.

End-to-end tests, where synchronised test sets at both line ends apply a coherent fault scenario, are how the full scheme including the channel is proven. They need time synchronisation between the sets, normally from GPS or a PTP-based source, so that both ends apply their part of the scenario at the correct relative instant. For ProtectionAI, synchronised end-to-end testing is on the roadmap and is not available today; IEEE C37.233 is the guide to consult for how such system tests are structured.

A practical software workflow

The repeatable core of zone verification is a fixed pattern of points, computed from the settings, executed the same way each time. A nine-point automatic zone sweep is a reasonable default: three points spread across the boundary of each of three zones, run at the reference angle and at offsets, with the swept points plotted on the impedance plane against the nested characteristics computed from the setting file. Seen as a picture, a wrong reference angle or a mis-entered resistive reach is obvious in a way that a column of numbers does not make obvious.

In published GridAPM ProtectionAI v1.3 this sweep runs against the built-in simulator — the characteristic is computed from the settings, the search logic executes, and the plot and report are produced, so the plan can be validated and corrected at a desk. The v1.4 source candidate’s bounded OMICRON path does not yet implement or physically qualify this distance-sweep workflow. See how ProtectionAI works for the mechanics, and relay testing software for how this sits alongside the test-set manufacturers’ own software.