Substation Grounding Touch Voltage Fault Case Study

Substation grounding is not proven by a low resistance number alone. During a ground fault, part of the fault current enters the earth grid, raising the local ground potential. A person touching a fence, gate, cable tray, transformer tank, control cabinet, or other exposed metal can be exposed to a touch voltage that depends on ground potential rise, bonding, surface layer, clearing time, soil conditions, and the current path through the body.

This case study follows a medium-voltage substation where commissioning measurements show an unacceptable touch-voltage risk at a personnel gate. The feeder protection works, but the ground fault clears more slowly than the original grounding study assumed. The result is a safety problem even though the ground grid resistance looks reasonable.

The values are realistic screening data, not a substitute for a project-specific grounding study or the governing electrical safety standard. The purpose is to show how an engineer should connect fault current, grid-current split, ground potential rise, clearing time, touch and step voltage, corrective bonding, and validation evidence.

Case Context

An industrial 13.8 kV substation feeds process loads through a main transformer and metal-enclosed switchgear. During pre-energization review, the construction team performs grounding continuity checks, a ground-grid resistance test, and low-current touch-potential measurements scaled to the studied fault current.

Item	Value
maximum line-to-ground fault current at the yard bus	I_{LG}=7.5\ \text{kA}
fraction of fault current entering the earth grid	S_g=0.58
measured ground-grid resistance	R_g=0.48\ \Omega
original study clearing time	t_{design}=0.45\ \text{s}
measured protection clearing time	t_{clear}=0.72\ \text{s}
project touch-voltage limit at 0.45\ \text{s}	E_{touch,0}=650\ \text{V}
project step-voltage limit at 0.45\ \text{s}	E_{step,0}=1500\ \text{V}
worst measured touch transfer factor	k_{touch}=0.36
worst measured step transfer factor	k_{step}=0.18

The touch and step limits above come from the project safety study for the selected surface layer, footwear/body model, soil model, and clearing-time basis. A real project must use the approved standard and site-specific assumptions. This case uses the limits only to show the engineering workflow.

Field Evidence

The commissioning evidence points to a localized touch-voltage problem, not a general failure of every grounding component.

Evidence	Engineering interpretation
ground-grid resistance is lower than 0.5\ \Omega	grid resistance alone appears reasonable
gate post bonding jumper is missing	fence metal may sit at a different potential than the grid
crushed-rock layer is thin near the gate	surface-layer protection may be weaker than the study assumed
relay event simulation shows 0.72\ \text{s} clearing	touch-voltage limit must be checked at the real clearing time
low-current injection shows the worst touch point at the gate latch	the hazard is localized and measurable
step-potential measurements remain comparatively low	touch voltage, not step voltage, governs the corrective action

The key lesson is that grounding must be verified as an installed system. Drawings, resistance tests, relay files, and field bonding details all affect the answer.

Grid Current and Ground Potential Rise

Only part of the ground fault current enters the earth grid. Some current returns through cable shields, neutrals, overhead ground wires, utility grounding, structural steel, or other metallic paths. The grid current used for the grounding screen is:

Substitute the data:

Ground potential rise is:

Use kiloamperes and ohms:

So:

This number is not the touch voltage by itself. Touch voltage depends on where the person stands, what metal is touched, how the metal is bonded, the surface layer, and how the potential gradient spreads around the grid. Ground potential rise is the driving quantity that makes the hazard possible.

Clearing-Time Adjustment

The original project limit for touch voltage was based on:

but the measured protection clearing time is:

For a first-pass screen using the same body and surface-layer assumptions, tolerable voltage is often treated as inversely proportional to the square root of exposure time:

The adjusted touch-voltage limit is:

The adjusted step-voltage limit is:

The exact allowable voltage must come from the applicable standard and approved design basis. The screening point is still important: a slower relay clearing time reduces the acceptable touch voltage.

Touch and Step Voltage Estimate

The measured transfer factor at the worst touch point is:

Estimate touch voltage:

Touch-voltage utilization is:

The touch-voltage screen fails.

For step voltage:

Step-voltage utilization is:

The step-voltage screen passes. The governing hazard is a person standing near the gate and touching bonded or poorly bonded metal during a ground fault.

Why Low Ground Resistance Was Not Enough

The grid resistance of 0.48\ \Omega looked acceptable as a single number. It did not prove safe touch voltage because:

1. fault current was high enough to create a ground potential rise above 2\ \text{kV};
2. the real clearing time was longer than the original grounding study assumed;
3. the personnel gate had weak bonding and a high local touch transfer factor;
4. the crushed-rock layer near the gate did not match the assumed surface condition;
5. touch voltage is a local exposure problem, not only a global resistance problem.

This is a common commissioning trap. A grounding grid can have low resistance and still produce unsafe touch voltage at a fence, gate, transformer, cable rack, or remote metallic object if bonding and potential gradients are poorly controlled.

Engineering Decision

The substation should not be released for normal operation with the personnel gate in service. The decision is:

Hold energization for unrestricted access, correct the localized touch-voltage hazard, verify relay clearing time, and repeat scaled touch-potential measurements before handover.

The immediate controls are:

1. keep the affected gate locked out or administratively restricted;
2. install temporary warning and access controls during commissioning;
3. verify all fence, gate, cable tray, switchgear, transformer tank, and structural-steel bonding;
4. review ground relay timing against the grounding study clearing-time basis;
5. repair the crushed-rock surface layer where it is missing or contaminated;
6. add gradient-control conductors where measurements show excessive touch transfer factor;
7. retain all test locations and scaling calculations in the commissioning record.

This is a safety release decision. The site should not rely on normal operating discipline to control a hazard that appears during a credible fault.

Corrective Package

The corrective package includes both grounding work and protection work:

• bond the gate leaf, hinge post, latch post, and adjacent fence sections to the ground grid;
• install a gradient-control conductor around the gate access path;
• restore the crushed-rock layer thickness and remove conductive soil contamination from the surface;
• add supplemental grid conductor and rods where the grounding study shows useful reduction in grid resistance;
• correct the ground relay settings so the same fault clears in the intended time window;
• update drawings, test points, relay files, and commissioning acceptance records.

After the correction, measured and reviewed values are:

Item	Corrected value
fault current entering earth grid	I_g=0.55(7.5)=4.125\ \text{kA}
measured ground-grid resistance	R_g=0.33\ \Omega
corrected clearing time	t_{clear}=0.32\ \text{s}
corrected touch transfer factor	k_{touch}=0.22
corrected step transfer factor	k_{step}=0.12

Corrected ground potential rise is:

So:

Corrected touch voltage is:

Corrected step voltage is:

The corrected touch-voltage limit at 0.32\ \text{s} is:

The corrected step-voltage limit is:

The corrected utilizations are:

The touch-voltage margin is now substantial, provided the corrected relay settings, bonding, surface layer, and measurements are maintained.

Validation Evidence

The repair is not accepted just because conductors were added. It must be validated as a system.

Useful acceptance evidence includes:

Evidence	Acceptance purpose
continuity test from gate metal to ground grid	proves exposed metal is bonded
low-current touch-potential test scaled to fault current	confirms local touch transfer factor
fall-of-potential or approved grid-resistance test	confirms grid resistance assumption
relay secondary injection and trip record	confirms clearing time used in safety calculation
surface-layer thickness and material record	confirms the assumed contact condition
as-built grounding drawing with test locations	preserves future inspection baseline
commissioning deviation closeout	proves the access restriction was removed by engineering approval

The relay test and grounding test must agree. A grounding calculation that assumes 0.32\ \text{s} clearing is not valid if the installed relay file clears in 0.72\ \text{s}, and a relay test alone does not prove that touch points are bonded correctly.

Measurement Uncertainty

The failed touch-voltage conclusion should be checked against uncertainty. Assume:

• fault current uncertainty: \pm10\%;
• grid-resistance uncertainty: \pm0.04\ \Omega around 0.48\ \Omega;
• touch transfer factor uncertainty: \pm0.03 around 0.36.

The relative uncertainty is approximated as:

For the original touch-voltage estimate:

Even the lower bound:

is above the adjusted allowable touch voltage of 514\ \text{V}. The failed conclusion is therefore robust. The exact margin may change, but the decision to correct and revalidate does not depend on a marginal measurement.

Risk Screen

A risk-priority-number screen communicates the release decision:

Before correction:

Severity is high because the failure mode can expose a person to dangerous touch voltage. Occurrence is moderate because the fault is credible even if infrequent. Detection is weak because the hazard is invisible during normal load operation.

After bonding correction, surface repair, relay timing correction, and scaled touch-potential validation:

The severity remains high because electrical shock consequence remains serious. The risk is reduced by lowering occurrence and improving detection through maintained bonding, relay testing, and repeatable measurement locations.

Lessons for Electrical Engineers

The transferable lessons are:

1. Ground-grid resistance is necessary evidence, but it is not sufficient safety evidence.
2. Touch voltage can govern even when step voltage is acceptable.
3. Clearing time is part of the grounding design, not only a protection setting detail.
4. Fence gates, cable trays, transformer tanks, control cabinets, and remote metalwork must be treated as exposure points.
5. A grounding correction must close with as-built bonding records, relay evidence, surface-layer records, and scaled touch-potential measurements.

The engineering decision is to release the substation only after the installed grounding system, protection clearing time, and measured exposure voltages all match the safety basis.