An anode bed fails at the wire, not at the metal
BURIED, permanently wet, packed in backfill chosen for how well it conducts: the lead-wire joint is the deepest thing on this site, and the only seal here asked to hold with direct current running through it.
The bed is many anodes; the continuity is one joint
A groundbed is designed around consumption. The anode metal is meant to go, and the design life is a sum about how much of it there is. The lead wire is not, and neither is the point where it enters the anode. Every ampere that anode delivers passes through that one connection, so its condition rather than the metal still left decides whether the anode is in the circuit.
Three loads act on those couple of centimetres at once, which is why this is the part that goes. Mechanically, backfill consolidates for months after placement and the cable gets pulled on while the hole is made good, so the exit takes tension and bending long after anyone is watching. Electrically, the connection carries the anode's full output continuously, and a joint that has turned resistive heats where it turned resistive. Environmentally, the ground was chosen for conductivity: moisture-retaining, low-resistivity backfill is the point of the installation, and also the most aggressive place to put a junction of two metals.
Nothing at the surface changes when it opens. The rectifier goes on supplying whatever the remaining anodes will accept, and the loss appears as drifted potentials at the next survey, after which somebody works out which anode left and digs for it. The metal is usually still down there. It is the wire that has gone.
The reasoning carries to any joint that is buried as well as wet: a cable end going into a duct that floods takes an adhesive-lined cap, an entry to a below-ground enclosure is closed at the gland, and a pre-insulated pipe end is a different family, written up on heatshrink-sleeve.com. What is particular to the anode is the current.
How the joint opens, in the order it happens
Five stages, none of them dramatic. The gap between the first and the last is counted in surveys, not days.
Moisture reaches the connection
It needs no hole. A cap that only covers, or a lining that never bonded to the wire insulation, leaves a capillary path along the cable. At this depth groundwater is continuous rather than cyclic, so the path has only to exist once.
Corrosion product builds at the wire exit
The connection is now a wetted junction of two metals with current through it. Product forms at the exit, occupies more volume than the metal it replaced, and jacks the insulation away from whatever was gripping it.
The wire works loose under settlement load
Backfill keeps consolidating long after installation. With the exit already opened up, movements the joint would previously have absorbed go straight into the connection.
Circuit resistance rises before anything opens
The anode does not switch off; its share of the output falls first. Where the bed runs through a junction box with a shunt per anode, that fall is visible there before it is visible as a structure potential.
Coverage is lost quietly
The other anodes take up the redistributed current, so averaged readings still look acceptable while the length of structure that anode covered has stopped being polarised. The anode is cheap. The months of unprotected steel are not.
What a cap at the lead wire does not do
It does not make the anode last longer. Consumption is a function of current and mass and sealing the joint changes neither, so a cap sold on anode life is describing a different failure from the one it prevents. It is no substitute for the right backfill, for the connection having been made properly underneath it - a sealed bad crimp is still a bad crimp - or for testing. It removes one failure mode and reports on none of the others.
What the cap changes at the exit, as published figures
The manufacturer's properties, and why a cap specified for this joint is used instead of a general-purpose end cap. Each figure has its test method beside it in the table on the anode cap page; this page states the values bare.
- Mastic lining, not adhesive. The rubber-based mastic is specified to adhere to the anode material and to the wire insulation, two dissimilar surfaces at one point. That is the difference between covering the connection and sealing it, and the sealed construction is rated to IP68.
- Stress relief at the exit. The moulding is sized at both ends, so the recovered wall grips the cable beyond the connection and takes the bending load that would otherwise reach the joint.
- Non-corrosive material. That matters more here than anywhere else on this site: whatever touches a wetted junction carrying current becomes part of the chemistry.
- Operating range -55 to +100 °C (-67 to 212 °F). Colder at the bottom and cooler at the top than the cable end caps described here. Neither figure carries across.
- Dielectric strength 12 kV/mm minimum, volume resistivity 1 x 10^14 ohm-cm minimum. Insulation is a duty at this joint rather than a by-product of the material, because the surrounding ground is a deliberate conductor.
- Shrink temperature 125 °C. Recovery is by torch, so the real constraint at a repair is what else is within reach of the flame: existing coating, the cable run, the sides of the excavation. Technique on the installation page.
- Two diameters select it. Codes run GAC 0420 to GAC 0550 and every row carries an anode-end and a wire-end dimension, so cable size alone will not pick a part. Chart on the size charts page.
Questions this page gets asked
I searched for anode cathode end cap tooling. Do you supply tooling?
Does any of this apply to zinc anodes?
Can I close an anode joint with a cable end cap instead?
Is the cap the whole seal, or does the connection need filling as well?
Where this goes next
- The cap itself, with the full GAC chartBoth-end dimensions, and the three searches it is not for
- Six ways a closed end opens againThis failure beside the five that happen shallower
- No water, no cellWhat sealing an end honestly does for buried metal
- What an end-cap datasheet provesWhere the test method behind each figure above lives
Send the two diameters, not the anode's rating
Sizing a cap for a lead-wire joint needs the outside diameter of the anode at the end being covered, the outside diameter over the wire insulation, and confirmation that the joint finishes buried rather than in air. Weight, alloy and current output do not select the part.