The Test Chart That Never Settles
You are inspecting the hydrostatic strength test of a 5.8-mile, 12.75-in. OD, 0.250-in. wall, X52 test section with an MAOP of 1,200 psig and a required test pressure of 1.25 × MAOP. The section rises 240 ft from the low point to the high point. The chart recorder and the deadweight-checked gauge are both installed at the low point. Test water comes from a stock pond at 48°F; ambient is 85°F. Air was displaced by pig during filling, and a 1-in. vent at the high point was closed when water appeared. The chart shows pressure reaching 1,500 psig, then falling to 1,478 psig over the first 100 minutes, then holding essentially flat for the remaining 6 hours of the 8-hour hold. Water temperature in the section rose 6°F during the test.
1. Is the test pressure correct? Do the arithmetic on elevation before you look at anything else.
Read the worked answer
No. The number on the gauge at the low point is not the number that matters.
- Head = 240 ft × 0.433 psi/ft = 103.9 psi
- Pressure at the high point = 1,500 − 104 = 1,396 psig
- Required minimum at the high point = 1.25 × 1,200 = 1,500 psig
The high point is 104 psi short. The test as run does not demonstrate a 1.25 × MAOP strength test at the controlling (highest) elevation, even though the recorder shows 1,500.
What the low-point gauge must read: 1,500 + 104 = 1,604 psig.
Check the hoop stress before you raise it:
- S = P × D / (2 t) = 1,604 × 12.75 / (2 × 0.250) = 20,451 / 0.500 = 40,902 psi
- 40,902 / 52,000 = 78.7% of SMYS
That is within the strength-test hoop-stress limits typically permitted for this class of pipe, so raising the low-point pressure to 1,604 psig is feasible with this wall - but I confirm the exact ceiling in B31.8 §841.3.2 and the operator's test plan before the pump starts, and I confirm the flanges, valves, and temporary test heads at the low point are rated for 1,604 psig plus the gauge tolerance.
2. Interpret the 22 psi drop followed by a flat 6 hours. Is this a leak? What do you do?
Read the worked answer
The flat section of the chart is not the same thing as a passing test, and the drop is not automatically a leak - but it is not automatically air either.
The physics:
- Trapped air. Air compresses and dissolves into the water over time. That produces exactly this signature: an early drop as air pockets compress and the packer settles, then a plateau. Filling by pig and venting at the high point reduces but does not eliminate air; the vent closing "when water appeared" tells you nothing about whether air was still coming through.
- Temperature. The section was filled with 48°F pond water into soil at a different temperature. A 6°F rise over the test is a large volume change; in a restrained steel line, water warming typically raises pressure noticeably. Here pressure is flat while temperature rose 6°F. That means something else is absorbing the volume that temperature is adding - air is the benign explanation, a leak is the other one, and the two are indistinguishable on the chart alone.
- The plateau masks both. A flat 6-hour trace can be a 22 psi drop early plus 22 psi of thermal expansion later.
What I require before accepting anything:
- Stabilize, then test. Do not start the hold clock until the water temperature in the section is within a few degrees of ground temperature and stable - commonly a stabilization period of many hours to a day. The current test began far too early after fill.
- Prove the air is out. Flow and pressure history during fill, pig travel records, and vent behavior at the high point. If air is suspected, a controlled depressurization and refill with a pig or additional venting at the high point.
- Pressure-volume analysis. If a leak is suspected, measure the volume required to hold pressure over a defined interval; a small leak consumes water continuously and will show up as a slope, not a step.
- Isolate if needed. If the trend persists after stabilization, isolate the section into shorter segments and re-test to localize.
3. The contractor wants the test signed off because the chart is flat and the weather is closing. What do you do, and what do you require in the record?
Read the worked answer
I do not sign it. A flat chart at the wrong pressure with unstable temperature and unresolved air is not a strength test; it is a recording. The operator's integrity group will later rely on this document as the basis for the MAOP, and signing it would put a number on paper that the line cannot support.
My disposition:
- Reject the test as run: pressure insufficient at the high point, temperature not stabilized, air not demonstrated to be removed.
- Require a re-test with the low-point pressure computed from elevation, gauges calibrated and deadweight-checked, temperature stabilized, and the hold clock started only after stabilization.
- Require the test manifold, temporary heads, and the low-point piping rated for the elevated pressure, and a written test procedure reviewed before pumping.
- Confirm the section's caliper or pig run and the tie-in welds are complete - you do not test a section and then weld on it.
Records I require before acceptance:
- The calibrated chart and gauge certification, with the gauge elevation and the computed high-point pressure stated on the record.
- Fill records: volume, fill rate, pig runs, venting performed, air-removal method.
- Temperature log at multiple points and the stabilization evidence.
- Hold duration, any pressure adjustments, and who authorized them.
- A signed acceptance that states the pressure at the high point, not just at the gauge.
If the contractor elects to wait out the weather rather than re-test, that is the correct outcome and my daily report says so.