Yes it is [illegal] poor design and poor construction, and it is a shame the guy died.
BUT He should not have tested the tank with compressed air.
If He had completely filled the tank with water before putting the air pressure on it
there would have been a big, wet mess and not an explosion.
Had everyone understood that this is a pressure tank and been following all the required codes,
presumably the welder would have been aware that testing a tank with compressed air is a great risk.
The article suggests the proper solution to all the trucks in service, don't pressurize the tank, pump the water.
Here in Brazil there is the so called "NR13 Standard", a safety regulation issued years ago by the Ministry of Labor, that rules the safe operation and maintenance of boiler and pressure vessels installations. It is a standard related to safety, not calculations and fabrication, so it doesn't replace nor interfere with ASME I and VIII Codes, commonly used also in Brazil.
NR13 states that a pneumatic test of a pressure vessel can be done only in cases that a hydrostatic test is impossible or very difficult to perform. A few examples are:
The vessel foundation or support will not withstand the extra weight imposed by the testing water.
The testing water chemical composition will corrode the vessel metal. An example of this is water with high chlorides content that could produce stress corrosion cracking on stainless steel vessels.
Impossibility to drain the vessel after the hydrostatic test.
In these cases, the test may be replaced by the pneumatic one, provided that the testing procedure is written by, or under the resposibility of, a Registered Professional Engineer. "Under the resposibility of" means that the procedure may be written, for example, by a non registered mechanical engineer who works in an engineering company, but the responsibility will remain on the company's Registered Professional Engineer. Brazilian law requires that an engineering company have at least one Registered Professional Engineer for each technical branch in which the company operates: civil, mechanical, electrical, chemical etc.
NR13 requires also that, once the test pressure has been established (and it is stated how to establish it), the pressure into the vessel be raised slowly up to 50% of the test one. At this point, the pressure will be raised at intervals of 10% of the test pressure until the required pressure is reached. At all times when the pressure is raised, all persons will stand at a safe distance, or in a safe location, away from the vessel. As the last step, the pressure will be lowered to 80% of the test one and the inspection for possible leaks can start.
The standard requires also that the pneumatic test be witnessed by the Registered Professional Engineer or his duly authorized representative, to make sure that the approved procedure is followed.
Giovanni S. Crisi
Sao Paulo - Brazil
By qcrobert
Date 07-07-2011 22:20
Edited 07-08-2011 14:24
Giovanni, thank you for bringing the Brazil safety standard to my attention as I was unaware of it.
As with ASME Sec VIII Div 1, UG-100 Pneumatic Test
(a) Subject to the provisions of UG-99(a)(1) and (b)(2), a pneumatic test prescribed in this paragraph may be used in lieu of the standard hydrostatic test prescribed in UG-99 for vessels:
(1) that are so designed and/or supported that they cannot safely be filled with water;
(2) not readily dried, that are to be used in services where traces of the testing liquid cannot be tolerated and the parts of which have, where possible, been previously tested by hydrostatic pressure to the pressure required in UG-99.
{unquote} other provisions go on to state the pressure to be at 1.1 times MAWP; metal temp during test to be maintained at least 30 deg F above min design pressure to minimize risk of brittle fracture; pressure gradually increased not more than 1/2 test pressure and thereafter increased in steps of 1/10th of test pressure.
QCRobert