It was going to be a long night — and not because of the 6 in. of ice on the 100-ft. stairway to the top of the roaster. No, this disaster was man-made. Instead of using compressed air to test the buried fiberglass pipe from the tankfarm, which was my suggestion, the dayshift used water.
The water froze and broke the fiberglass pipe about an hour before my shift. Now, in addition to the inclement weather of January along Lake Michigan, we had an environmental release. Losing the large pipeline meant that I was forced to use a smaller spare line. This highlights one problem with pressure tests: fluid choice. After the spools have been tested, they must be tested again in the field. This is where neat plans in quiet comfortable conference rooms fall apart.
There are three choices for the test fluid: air (or nitrogen), process fluid (gas or liquid), or water. Each choice poses its own problems.
Testing with compressed air is limited to about 80 psig, although nitrogen is often available from a utility provider at pressures of about 150 psig. A danger with compressed air is that it may be contaminated with condensates, either oil, water or dust. Oil or dust can be an explosion hazard in some processes, but then, so, too, can air, for some operations, if it isn’t purged prior to commissioning.
Water also may be a contamination problem. Nitrogen, typically delivered via a utility provider, is often available at up to 150 psig. The main peril with nitrogen is the danger of suffocation. Even a single lung full can be fatal. The leading reason for not choosing a compressed gas is the difficulty in regulating its pressure. However, careful design of your test skid can minimize this threat.
Another drawback to testing with gas is that leaks are slightly more difficult to detect. I’ve had success using liquid soap and a water bottle with a drinking nozzle. Make sure the soap is at least 10% by volume; water alone won’t work. Spray the solution completely on the flange. Perfectionists like to use masking tape over the flange so that the leak shows up more prominently.
Choosing the operating fluid offers one advantage for pressure testing: material evaluation. — It can identify the risk of acute failure— if seals and piping may leak and equipment may fail. Usually, this is hardly worth the trouble of fluid disposal or the risk to personnel from a potential exposure to harsh chemicals.
Besides, most chemical corrosion is chronic. Even for a service requiring a tantalum, a Type 316 stainless steel thermowell survived three days before failure — well beyond the time it would take for a pressure test. Using the operating fluid raises one more challenge to consider: the emergency response plan (ERP). If you choose this fluid, your safety manager will certainly insist that all the interlocks be confirmed safe, the operators are trained for any new procedures and that all aspects of your ERP are in place. No doubt, you will have this manager breathing down your neck for the duration of testing. Given all its drawbacks, pressure testing with the process fluid isn’t appealing.
Hydrotesting is often promoted as the easiest and safest method. However, I usually prefer testing with a compressed gas. First, draining the pipe assembly after hydrotesting can be tough. Indeed, nearly every time I’ve done hydrotesting, there’ is a corner that can’t be drained. This Achilles’ heel was exposed during a major expansion at a chemical plant. It happened before my time, but drying was a problem when the chlorine system was hydrotested. It took several weeks for the water dew point to sufficiently drop for the start-up to continue. What a disaster! If you must do a hydrotest in the field, make sure your piping system is designed to be easily drained.
In fact, design your piping for the test. My personal preference would be, for most pipe, to hydrotest the spools in the shop with a foot pump and use a compressed gas or the system pump in the field.
One advantage of pressure testing is that it may allow testing of pressure interlocks, including the instrument and the controller. This type of testing is preferable to a communication check, which often passes for an instrument check. Instrument tests could be performed in the shop, too, but you risk damaging instrument (copper) leads.
One problem to avoid after the test is reused gaskets. Unless your process contains water — don’t reuse the gaskets. Water-soaked gaskets may even be dangerous in your process.
With thoughtful consideration, beginning during the design phase, you can make field pressure testing painless.