How to tell if a large
BSP leak is size,
thickness, face
damage or load
Wrong gasket size, wrong thickness, damaged face, and insufficient seating load all produce a leak at a large BSP flat face connection. They do not all look the same. The timing of the leak, its location around the joint, the compression mark on the removed gasket, and the face condition together point to which cause is most likely. Reading these signals before ordering the replacement prevents the most common and most expensive mistake: fixing the wrong thing.
Why the cause matters before the replacement
At ¾" BSP, a wrong diagnosis is inconvenient. At 1½" or 2" BSP in a commercial installation, a wrong diagnosis means a second full disassembly — access cleared again, system isolated again, fluid managed again — at a connection that was not straightforward to reach the first time. The cost of the second event often exceeds the cost of the first.
The four causes are distinct enough that the evidence available before and during disassembly can usually indicate which is most likely — if that evidence is read before the removed gasket is discarded and before the face is cleaned without inspection.
Read the removed gasket and inspect the face before cleaning or reassembling anything. The compression mark on the removed gasket and the condition of the face at the time of disassembly are the primary diagnostic inputs. Cleaning the face before inspecting it, or discarding the gasket before measuring and reading it, removes the evidence that identifies the cause.
The four causes — what each one signals
Quick diagnostic table
| Signal observed | Most likely cause | First check |
|---|---|---|
| Leak at first pressurisation, mark narrower than recess, face clean | SIZING | Measure gasket OD, ID vs recess dimensions |
| Union would not fully close, or gasket extruded visibly | THICKNESS — TOO THICK | Measure gasket thickness vs recess depth |
| Uniformly light mark, seals initially then weeps progressively | LOAD | Review tightening procedure and gasket grade creep resistance |
| Uniformly light mark, leaks after thermal cycle, no visible damage | LOAD or THICKNESS | Measure recess depth and gasket thickness; review seating load |
| Interrupted mark at one arc, visible damage or residue at that location | FACE DAMAGE | Inspect and address face condition before re-gasketing |
| Radial gap in compression mark, radial scoring on face | FACE DAMAGE — SCORING | Assess score depth and length; determine if face can be re-sealed or needs replacement |
| Mark light at bore edge only, outer zones adequate | FACE DAMAGE or LOAD | Inspect bore edge of face for damage; check bore-edge seating stress adequacy |
| Joint re-tightened and leak temporarily stopped; gasket intact, face clean | LOAD | Review assembly load and gasket grade for creep resistance |
Signals overlap — the table gives the most likely cause, not a guaranteed diagnosis. Low seating load and face damage can produce similar patterns, and both can be present at the same time. Where the evidence points to two possible causes, address both before reassembling. At 1¼" to 2½" BSP, the cost of a third disassembly because both causes were present and only one was addressed is higher than the cost of treating both systematically the first time.
When the leak comes back after re-gasketing
A leak that returns at the same location after the first repair attempt is the strongest signal that the root cause was not correctly identified — or was identified but not fully addressed. Common patterns:
Same location, same pattern. The new gasket leaked from the same arc or the same radial position as the original. Face damage was present and was not addressed. The new gasket sat on the same damage as the old one.
Outer edge, bore edge, or full circumference unchanged. The same leak geometry after replacement. The cause was sizing or load — the gasket specification or the bolt load was not corrected, and the same underlying condition produced the same failure.
Sealed for a few weeks then failed. Initial seal followed by progressive failure suggests load loss — either insufficient initial assembly load, or a gasket grade with insufficient creep resistance for the service conditions at this connection size.
The diagnosis is in the evidence. The evidence is available before it is discarded.
The removed gasket and the face condition at the time of disassembly identify the most likely cause. Sizing produces a mark that does not cover the full recess. Thickness produces a mark that is uniformly light or shows closure problems. Face damage produces an interrupted mark. Load loss produces a uniformly light mark on a clean face. Read the gasket, inspect the face, measure what needs measuring — then specify the response. At 1¼" to 2½" BSP, doing this before the first repair is completed costs minutes. Skipping it and needing a second repair costs significantly more.