Shale Spoil, a Holiday at Twelve O'Clock, and Rain

The repair fails on at least four independent grounds, each sufficient by itself.

1. Dew point. At 36°F and 90% RH the dew point is roughly 33°F. Steel at 34°F is barely 1°F above it, and the coating manufacturer's data sheet typically requires the surface to be at least 3°F above dew point, with the pipe dry. Condensation forming under a fresh patch is a guaranteed disbondment.
2. Surface preparation. Hand wire brushing is not a surface preparation for FBE. The specification will call for power-tool cleaning to bare metal at a defined cleanliness and anchor profile - commonly SSPC-SP11 or better, plus abrasive if the patch is large. Hand brushing leaves a surface the patch cannot bond to.
3. Wet substrate. No liquid coating goes onto a wet pipe.
4. Pot life / induction. Four minutes after mixing is plausible for some products and not for others; I check the data sheet's pot life and induction time rather than assume, and check that the mix ratio was measured, not eyeballed.

Acceptance conditions: substrate temperature sustained at least 3°F above dew point and rising; dry, clean, correctly profiled bare metal; product within its shelf life and mixed and applied within pot life; cure verified - including a holiday check at the specified test voltage once cured, and an adhesion check if the specification requires it; and a record of product lot, batch, and cure conditions.

If the weather will not allow it, the answer is to heat and shelter the work area or to defer the repair - not to install a patch that will be buried and never seen again.

Backfill: reject the shale as-is. 2-4 in. fragments are not padding and not rock-free backfill. Under a 0.500-in. wall coated pipe, a shale fragment becomes a point load under several feet of cover and eventually a coating holiday and a corrosion cell; that is exactly the mechanism that produces external corrosion failures years later. The specification is explicit and the crew's temptation is to use what the trench produced.

My requirements before lowering-in:

• Trench bottom free of rock, hard spots, and debris; padded or sand-slid where the native bottom is rock.
• 6 in. of imported sand around the pipe, verified by depth checks, not by eye.
• Rock-free material to 12 in. over the top. That means an imported or screened fill, a padded spoil, or a specified rock shield if the project allows one - the choice is the operator's, not the crew's, but the 12 in. is not negotiable.
• No frozen material in the backfill; air is 36°F and the spoil pile will have frozen lumps that leave voids.

Sequencing. The dry trench is the trap: the crew will lower in and backfill because the trench is open and the weather is closing. But once backfilled, the coating cannot be inspected, holiday-detected, or repaired. I hold lowering-in until the completed coating is inspected at 100% and holiday-detected at the specified voltage, until the backfill source is approved by physical test (not by its name on the ticket), and until padding depth can be verified in the ditch.

I document the hold in the daily report and notify the operator's construction manager the same day. A one-day schedule slip is cheap; a shale-damaged line is not.

I evaluate it as a substitution, not a field decision, and I say so in writing.

Questions the proposal has to answer:

• Is mechanical protection equivalent to padding here? Rock shield protects against a point load in one direction; it does not address the rock bedding under the pipe, the 12-in. cover requirement, or the movement of angular spoil during settlement. On an FBEP-coated line, shield over an uneven rock bottom still produces the same failure.
• Does the operator's specification permit substitutions at all, and who holds the authority to approve one? Usually it requires engineering approval and, in some cases, a documented variance. A verbal "it's fine, everyone does it" is not approval.
• What does the coating manufacturer say about the coating's tolerance for mechanical damage and the required padding?
• Is it verifiable? If the trench bottom is padded, depth can be checked before lowering-in. If not, I have no way to verify anything after backfill.

My recommendation to the operator is the substantive one: keep the padding under the pipe and around the first several inches, approve rock shield only for the upper portion if the specification allows it, and require the imported fill to 12 in. over the top. If the operator grants a variance, I want it in writing with the engineering rationale, I record the affected stationing, and I note the variance in the as-built documentation so the integrity group knows where to look in ten years.