Reading a pressure test on a CPVC hot-water line
Your firm has installed hot and cold water lines in a three-storey hotel. The hot line is CPVC, with 15 mm and 20 mm branches; the 15 mm hot main runs 25 m exposed on the open terrace before dropping to the second floor. Working pressure at the topmost tap is 3 bar, supplied by a 1.5 HP pump from a 500 L tank. On the day after jointing you fill the line and pressurise it with a hand test pump to 10 bar. The gauge falls from 10 bar to 8 bar in the first 15 minutes, then holds steady at 8 bar for the next 40 minutes. It is 2 p.m., ambient temperature 38 C, and the whole 25 m terrace run is in direct sun. The foreman says 'it holds, so sign it'. The client's engineer wants the entire line cut open and re-jointed.
1. Interpret the gauge behaviour. What explains the fall in the first 15 minutes, what does the steady hold prove, and why is 10 bar a questionable test pressure for this line?
Read the worked answer
The two parts of the curve have different causes and must be read separately.
- The fall from 10 to 8 bar is consistent with two effects, not necessarily a leak. First, air trapped at high points - particularly along an exposed 25 m terrace run with small 15 mm branches - compresses and dissolves into the water, so the gauge drops without any water being lost. Second, water and CPVC expand with temperature; the line was filled cold, then sat in direct sun at 38 C, so the pipe and the water inside changed volume. Solvent-cement joints also absorb a little pressure as they cure.
- The steady hold at 8 bar for 40 minutes is the meaningful result. A true leak of any size on a 15 mm line would not stop at a fixed value; it would keep falling, or fall in steps as joints weep. A flat line after the initial settlement indicates the joints are sound, provided no joint shows moisture.
- 10 bar is questionable because the working pressure is 3 bar. Standard practice is to test at about one and a half times the working pressure, and not above the pipe's rated pressure at the test temperature. Ten bar is more than three times the working head, and CPVC derates with temperature - at 38 C in sunlight the pipe wall is warmer still. Testing well above the rated pressure at that temperature can damage good joints and cause the very failure you are trying to detect, and it proves nothing extra about a 3 bar system.
2. List the checks you would carry out before signing the test, and state what would make you reject the line.
Read the worked answer
Before signing I would:
- Verify the gauge. Compare with a second gauge or a known-weight tester; an uncalibrated gauge is the most common cause of a misleading curve.
- Isolate in sections. Close the branch valves and test the terrace main separately from the internal drops, so a fall can be located rather than argued about. If possible, test in the cooler part of the day or shade the exposed run, so thermal effects are removed.
- Run a soap solution test on every joint, elbow and tee on the terrace run and at the accessible drops. Bubbles at a joint are decisive evidence of a leak; a dry, bubble-free joint under pressure is not.
- Inspect for mechanical faults - deformed pipe at solvent-cement sockets, joints made with excess cement, pipe clipped too tightly or resting directly on hot terrace surface, and any joint disturbed before the cement had cured for the full period (typically 24 hours for this size).
- Confirm water was not lost by checking whether the pump needed repeated topping up to hold pressure.
I would reject the line if: the gauge continued to fall after the initial settlement; any joint or socket showed moisture or bubbling; a section could not be isolated and shown to hold; joints had been pressurised before the cement cured; or the pipe showed deformation, which would indicate the test pressure was excessive.
3. Write down the test procedure, pressure and duration you would put in writing, together with the precautions that were missing on this job.
Read the worked answer
Written procedure to be handed to the foreman and the client's engineer:
- Cure before test: solvent-cement joints to remain undisturbed for the manufacturer's full cure period, normally 24 hours for 15 to 20 mm CPVC, and longer in humid weather.
- Test pressure: one and a half times the working pressure, i.e. about 4.5 bar for this 3 bar system, and never above the pipe's rated pressure at the actual temperature. The pump and gauge to be checked before use.
- Fill and bleed: fill slowly from the lowest point and bleed air from the highest outlets until water runs without sputtering. Air removal is what makes the gauge curve readable.
- Duration: hold for one hour, with readings at 15, 30 and 60 minutes. An initial fall followed by a flat line is acceptable only if the flat section lasts at least 30 minutes and there is no visible leakage.
- Environment: shade the exposed terrace run or test in the early morning; record the water and ambient temperature with the readings.
- Record: joint-by-joint visual and soap test, section isolation results, gauge serial number, and signature of both parties.
Missing precautions on this job: air was not bled before testing, the line was tested in direct sun, the gauge was not verified, there was no section isolation, and the line was pressurised the day after jointing without a confirmed cure period. Signing a test on a gauge curve alone, without a soap test and a cure record, is what turns a small weeping joint into a claim after handover.