The coating gets the spec sheet. The connection gets a torque wrench and a hope.
In the field, current does not enter an MMO titanium anode uniformly through the whole backing plate. It enters through whatever conductive path the installer actually tightened: a bolt, a tab, a welded busbar, a crimped cable, sometimes a clamp that was finger-tight because the glove would not close.
A clean coating on a poorly fed anode ages worse than a modest coating on a well-fed one.
Contact resistance at the interface is not a detail. It is the first thing the current sees. A hydraulic crimp on a cable joint is expected to hold below 0.001 Ω. A loose busbar bolt, an oxidized mating face, an undersized tab, or a paint mark left under the clamp can push that number up by an order of magnitude. At a given operating current, power dissipated at the joint follows P = I²R. The joint heats. The heat accelerates aging of adjacent sealant, micro-cracks form, titanium expands locally, and contact pressure drops further. A positive feedback loop, starting at a connection nobody torqued.
On plate anodes, if a single feed point is used without current-sharing design, longitudinal current gradient appears. Commercially pure titanium runs a typical resistivity of 0.47–0.56 μΩ·m, roughly 25–30 times that of copper. Along the plate, the near end of the coating carries more current than the far end. Multiple feed tabs or a back-mounted copper busbar flatten this curve.
Mesh and tube anodes have the same problem in a different shape. A long tube fed from one end develops a measurable current taper toward the far end. Multiple feed points, or a copper-cored conductor inside the titanium, redistribute the load. Flexible MMO wire anodes are built on exactly this principle: the titanium wire alone cannot carry longitudinal current over long runs, so a copper core takes the current and the titanium only does the electrochemical surface job.
Flange and mounting faces add a second failure mode. The anode must be electrically tied to the source but insulated from the vessel steel. If the insulating washer is missing, cracked, or installed backwards, current bypasses part of the active coating and dumps through the flange. Bolt temperature rises, gasket material degrades, reference potential drifts, and the protection current never reaches the structure.
Commissioning checks that catch this are boring and essential:
four-wire micro-ohmmeter reading across the joint (not a multimeter resistance range), benchmarked against the ≤0.001 Ω internal limit
infrared thermography under rated current for hot spots at bolts, tabs, crimps
insulation resistance between anode body and grounded structure, typically >1 MΩ
confirmation that conductive faces were degreased, derusted, and not painted over
torque record, not torque memory
A constant-potential power supply reading "output normal" does not mean current distributes evenly across the anode body. What decides whether the coating ever gets used is the first ten milliohms after the feed point.
