Before The Probe Leaves The Floor: Positive-Pressure Helium Leak Testing
Oct 28, 2026
Report In late October, the deep-processing bay at Dinoer ran positive-pressure helium leak testing on every piece of a batch of ICCP MMO probe anodes. No order details are disclosed. This was a routine quality gate on the production line.
A probe anode is not complicated in shape: a length of Grade 2 titanium tube, one end welded to a flange, the other capped, a copper conductor running through the center, the cable end potted in resin. What is complicated is the seal chain. The flange O-ring, the weld heat-affected zone, the resin interface, the point where the cable exits - any one of these can hide a micron-scale path. Once seawater or soil moisture gets in, it tracks along the copper, oxidizes the conductor, and kills the whole probe two years after burial.
"If the coating fails, you can re-coat. If the seal fails, there is nothing to save." The operator at that station said it without looking up from the gauge.
The test uses positive pressure. The part is filled with high-purity helium and placed in ambient air. A helium mass spectrometer probe scans the outer surface slowly. Helium has a small molecular size and low viscosity, so it finds pores no eye can see. The instrument reads down to the 10⁻¹² Pa·m³/s range - better coverage than vacuum sniffing, especially at corners and the cable transition zone, where negative-pressure methods tend to miss.
On the first piece of this batch, the probe picked up a signal 5 mm inside the L-corner. Reading: 1.2×10⁻⁷ Pa·m³/s, two orders of magnitude over the internal limit. They cut it open. The corner had a hairline lack-of-fusion - not a through crack, just a micro-porosity left at the root of the keyhole closure. Almost invisible on the radiography film.
The fix was plain: back to the welding bench, re-bevel, lower heat input, re-weld, re-test. Second pass read 3×10⁻⁹ Pa·m³/s, one order of magnitude under the limit.
From that point the batch moved from sampling to 100% inspection. Every probe now has its leak-reading, weld-batch number, and coating-load ID bound together in the archive. One chain, traceable both ways.
The workshop engineer put it directly: a cathodic protection system goes underground for ten years or more. Anode failure usually does not start with the coating. It starts with the seal. "The customer never sees this step. The seawater does."
Positive-pressure helium leak testing is now a standard gate in the deep-processing sequence - covering probes, flexible anode joints, tubular anode end-caps, anything with an internal conductive connection.
