Bonding a pier-mounted 11kV/400V transformer's HV and LV earths back to the onshore HV earth solves the fault-return problem and creates a cathodic protection problem in the same move: the shore electrode becomes a remote drain for CP current that was meant to stay on the pier steel. For a transformer replacement, the test scope is therefore two surveys, not one — an earthing verification (continuity, conductor adequacy, earth impedance, touch and transfer potential) and a CP baseline taken before the old unit is disconnected so you have something to compare against afterwards. Whether the contractor ends up augmenting the earthing depends on two measured numbers you almost certainly do not have yet: the earth fault current and clearing time at the onshore 11kV switchboard, and the actual end-to-end resistance of the 50mm² earth conductor running 800 m out along the conveyor. Look at the CP rectifier trend first; it will tell you more about the existing bonding than any single-point resistance reading.
What are the symptoms telling you before you touch anything?
The signal chain here runs in two directions. AC fault current wants to leave the transformer tank and get back to the source neutral. DC protection current wants to leave the anodes, enter the pier steel, and stay there. Both use the same metalwork, and every bond you add changes the split. Read the existing values first — a system that is already leaking CP current to shore will show it in the rectifier output long before any corrosion is visible.
| Signal | Where you read it | What a wrong value looks like |
|---|---|---|
| Structure-to-electrolyte potential, pier steel | Portable reference electrode at fixed test points along the full length, including the shore-end trestles | Under-protected readings clustered at the shore end, where current bleeds to the remote earth; readings drift as tide changes |
| Transformer/rectifier (T/R) DC output volts and amps | CP rectifier panel meters and logs | Output cranked up over time to hold the same protection criterion — the classic signature of current draining to a remote earth |
| DC current in the 11kV cable armour | DC clamp on the armour or gland, taken at both the transformer and shore ends | Any standing DC. The armour is carrying CP current to shore, and metal loss concentrates at the remote (source) end |
| Shore-to-transformer earth conductor resistance | Low-resistance ohmmeter or current injection, end to end on the 50mm² core | High or unrepeatable readings — corroded lugs and salt-attacked terminations at the offshore end |
| Tank and earth-bar bonds | Micro-ohmmeter across each bond, plinth steel, gland plates, enclosure doors | Milliohm-level joints that measure in ohms; the tank rises relative to the platform during a fault |
| Earth fault current and clearing time | Onshore protection relay settings and the source earthing arrangement (resistance or solid) | A solidly earthed source feeding a 50mm² conductor that was never checked adiabatically |
| AC current in the earth conductor, load conditions | AC clamp on the 50mm² core, transformer energised | Standing AC circulating current — parallel paths through the structure sharing neutral or screen current |
Why does the shore earth drain the CP system?
A cathodic protection system works by holding the structure negative with respect to the electrolyte, driven from anodes at a modest voltage. Marine systems run at low drive voltages compared with land-based buried systems, which limits how much current can be pushed around an unwanted loop — that is in your favour. It does not eliminate the loop. Bond the pier steel to a low-resistance shore electrode and you offer the protection current a metallic path out of the protected zone. The current leaves the steel where it can, travels back through the earth conductor or the cable armour, and discharges into soil at the shore end.
Two consequences follow. First, protection levels on the pier fall short, so the operator winds the rectifier up to compensate. Overdriving shortens anode and groundbed life and, pushed hard enough, evolves hydrogen at the steel surface — coating disbondment, and a hydrogen embrittlement concern on any high-strength fasteners in the structure. Second, wherever DC leaves metal and enters electrolyte, metal is consumed. That happens at the remote end, onshore, in structures nobody is surveying. The damage does not appear where the bond was made.
What happens on an HV earth fault, and where does the potential go?
Modern 11kV cable construction pushes the fault in a predictable direction. Each phase conductor carries an earthed screen, with overall steel wire armour also earthed. A winding or termination insulation failure therefore hits the screen or armour first, and returns to the source neutral through that path plus the 50mm² conductor. A fault directly to local pier steel, with no prior contact to the HV earth system, is an unlikely sequence. Even if it happened, at 11kV a submerged steel structure in seawater is not a poor electrode — some return via the mass of earth is available.
The hazard that actually matters is transferred potential, not local touch potential. Any rise of earth potential at the onshore feeder substation appears directly on everything galvanically connected to the HV earth — including a transformer tank 800 m out to sea. Whether that is a safety problem is set by the magnitude of the earth fault current and the quality of the substation earth electrode. A resistance-earthed 11kV system limits earth fault current and there is a fair chance the resulting rise of earth potential is too small to worry about. A solidly earthed transmission or primary distribution source is a different calculation, and you need the substation earth grid impedance and the actual fault level to do it.
Separate earths or bonded earths — how do you decide?
Electrical separation is a design intent, not a measured condition. On a 1 km pier with a covered conveyor tunnel, steel trestles landing onshore, cable trays, handrails, instrument cabling and a shore-fed 11kV armoured cable, the realistic starting assumption is that shore and jetty earths are already connected several times over. Prove it or disprove it with an injection test before you spend design effort on maintaining a separation you no longer have.
If separation is genuinely intact and worth keeping, the compromise usually taken is to bond the transformer tank and HV earth to the shore HV earth for fault return while keeping the structural CP zone isolated — and then to live with the transferred potential that this bond carries out to sea. If separation is already lost, quantify what bonding costs the CP system rather than arguing about it. Make a deliberate, bolted metallic connection from the 11kV cable armour or gland to the local steelwork and watch the rectifier and the structure potentials respond. It must be a bolted, cleaned, torqued connection — the CP voltages involved are far too low to drive current through a cable simply laid against steel, and a poor contact gives you a false negative. Log the rectifier output and a set of reference electrode readings before, during and after.
Test procedure for the transformer replacement
- Take the CP baseline first, with the old transformer still connected: rectifier DC volts and amps, structure-to-electrolyte potentials at every established test point along the pier and at the shore transition, and DC clamp readings on the 11kV armour and on the 50mm² earth conductor at both ends. Note the tide state and the time.
- Obtain the earth fault current and the protection clearing time at the onshore 11kV switchboard, and confirm whether the source is solidly or resistance earthed. Everything downstream of this step is arithmetic until you have those numbers.
- Check the earth conductor adiabatically against those numbers:
S ≥ sqrt(I² × t) / k, withktaken from the table in the applicable installation standard for that conductor material, insulation and initial temperature. If the 50mm² core does not pass, that is the augmentation work, and it is better found now than during commissioning. - Measure the 50mm² conductor end to end with a low-resistance instrument, then repeat with the offshore end disconnected from the pier steel to separate the cable from the parallel structural path. The difference tells you how much of the return is really travelling in the conductor.
- Bond and verify the new unit: tank to earth bar, HV and LV earth bars, cable glands and armour clamps, plinth and support steel, enclosure doors and gland plates. Micro-ohmmeter on each joint, recorded individually — not one lump reading across the lot.
- Run a current injection test from the transformer earth back to the source to obtain the real earth return impedance, and measure the resulting rise of potential and touch voltages on the tank, the LV panel, adjacent handrails and the walkway surface. Scale the measured values by the ratio of true fault current to injected current.
- Complete the routine electrical tests on the new unit — insulation resistance, winding resistance, turns ratio, vector group, and the position and integrity of the LV neutral earthing link — before energisation.
- Repeat the full CP survey after energisation and hand the before/after data set to the CP contractor with the bonding changes marked on a single-line diagram.
How do you verify the earthing is actually adequate?
Acceptance is a comparison, not an absolute. Structure potentials at every test point, including the shore-end trestles, meet the criterion in the CP design basis without the rectifier output having been increased from the baseline. DC in the cable armour is stable and no higher than the pre-work reading; a new standing DC where there was none means the replacement introduced a metallic path. Measured touch voltages, scaled to the real fault current, sit under the permissible touch-voltage limit for the measured clearing time — the limit moves with time, so quote both together.
The earth conductor resistance should be repeatable across two separate measurements on different days, and the individual bond readings should be milliohms, not fractions of an ohm. Then set a follow-up: re-read the rectifier output at one month and three months. A slow upward creep in T/R output is the earliest and cheapest indicator that the new installation is bleeding protection current to shore, and it appears long before any corrosion survey would catch it.
Which pitfalls recur on piers and jetties?
Assuming isolation is the first one. Isolating joints and flanges get shorted by a handrail, a cable tray support, a conveyor idler frame or a spare instrument cable, and nobody notices because nothing trips. Test across every isolating joint with the system live before trusting the single line.
The second is treating the CP rectifier as an adjustment. Winding the T/R up to recover protection levels hides a bonding fault and spends anode life doing it. Tuning does not fix wiring — find the metallic path, then reset the rectifier.
The third is the salt environment acting on the earthing system itself. A 50mm² single core terminated 800 m offshore in marine air will show its age at the lugs first. Corroded terminations raise the return path impedance quietly, so more of the fault current diverts into the structure and the transferred potential shifts. Inspect and re-terminate rather than assuming a conductor that was adequate at installation still is.
Finally, remember where DC corrosion damage shows up. Current picked up on the pier and discharged onshore consumes metal at the onshore end — cable armour, gland plates, buried structures near the shore substation. If the DC clamp shows current in the armour, that inspection belongs in the scope even though the transformer work is all offshore.
When to stop and escalate
Stop and bring in the CP contractor the moment you measure DC in the 11kV cable armour or find the rectifier output has been raised to hold protection levels — that is a corrosion engineering decision about anode capacity and bonding strategy, not an electrical one. Escalate to the network operator or the owner of the onshore feeder substation when you need the earth fault current, clearing time and substation earth grid impedance to assess transferred potential; without those three values the touch-voltage assessment cannot be closed out. Take the transformer manufacturer's commissioning requirements and earthing terminal ratings from the official product documentation and their support channel rather than from the old unit's arrangement.
Frequently Asked Questions
Why does bonding a jetty transformer earth to the shore earth cause corrosion?
The shore earth electrode acts as a remote drain: cathodic protection current leaves the pier steel, returns metallically through the earth conductor or cable armour, and discharges into soil onshore. Metal is consumed wherever DC leaves steel and enters electrolyte, so the corrosion appears at the remote onshore end, outside the CP-protected zone.
Why does the CP rectifier output rise after earthing work on a pier?
Because part of the protection current is now going somewhere it was not designed to go, and the operator increases the T/R output to hold the same structure-to-electrolyte potential. Overdriving shortens anode and groundbed life and can evolve hydrogen at the steel surface, causing coating disbondment.
Why is DC current in the 11kV cable armour a problem?
Standing DC in the armour means the armour is carrying cathodic protection current back to shore. That current has to discharge at the remote end, corroding the armour, glands and adjacent buried metalwork there — so a DC clamp reading on the armour at both ends belongs in the baseline survey.
Why does a transformer 800 m offshore see a voltage rise during an onshore fault?
Any rise of earth potential at the feeder substation is transferred along the HV earth conductor and appears on everything bonded to it, including the transformer tank. The magnitude depends on the earth fault current and the substation electrode impedance — a resistance-earthed 11kV system typically produces a rise too small to matter, a solidly earthed source may not.
How do you test pier earthing when fall-of-potential is impractical?
Use current injection between the offshore transformer earth and the onshore source earth to measure the real return impedance and the resulting rise of potential, then measure touch voltages directly on the tank, LV panel and handrails and scale them by the fault-current-to-injection-current ratio. Supplement with individual micro-ohmmeter bond checks and end-to-end resistance on the 50mm² conductor.