We have discussed the valve manufacturer’s control of these foundries and documentation requirements. The goods-in area inspection and variable interpretation in “casting surface review” to ANSI/ MSS SP55 “Quality Standard for Steel Castings for Valves, Flanges, Fittings, and Other
Piping Components - Visual Method for Evaluation of Surface Irregularities” was identified. The castings are then either released into production or put in a rejection area for further action. In production weld repairing machined areas where internal defects have surfaced during the machining operation is often seen. Gasket surfaces are of main concern though often a “rash” of weld repair can be seen on other surfaces.
Being an Auditor you can stand back and look at the castings to identify ‘common issues’ in defects and repairs. Repair patterns and trends are very obvious. As quality data is not collected, correlated nor subject to improvement, ‘poor practices’ continue unabated. When the valve manufacturer is challenged answers vary. “It’s acceptable to the client”, “It meets the standards”, “It’s always been done like this”, “It’s too costly to correct”.
Assembly and testing
The castings move through to the valve manufacturing assembly and testing. Some NDT or surface crack detection may be done depending on the client’s request or materials used. For the lower end carbon steel materials such testing is very rare indeed.
The auditor will review the test area and witness the procedure being used by the fitters etc. Is the test area well organised? Are the hydrotest machines
well maintained/calibrated? Are gas tests possible? How closely do they try to see leakages? How is the lighting? Are test records clearly kept?
Looking around at the valves being tested the ‘worst nightmare is seen through wall thickness leakage.’ The valve is marked up to be repaired by welding. The leakage occurred at the discontinuity between the body/ the top works branch. I have witnessed such events and it’s amazing. You can guess how the weld repair was done. When the repair procedure was further investigated it did allow such practice. This should have been picked up in the documentation review? Another observation around the test area is containers of water large enough to submerge the complete valve body and pressurize with low pressure air. They are looking for bubble leakage through the casting into the water. I have been quoted up to 7% of the castings leak in this way. So what happens when bubbles are found? Yes you’ve guessed it – they are welded up and retested. At yet another valve manufacturer the container did hold water but it was so dirty that it would be impossible to locate the bubble source.
An auditor from an end user that witnesses either of the above has no option but to reject these castings as they are not safe to use on any operational plant. Any casting failing in this manner should be scrapped. As temperature and high pressure hydrocarbons are introduced to the valve future leakage will be inevitable.
I first came across weeping castings on a plant where I was instructed to fly
immediately to assess the situation. A4 inch class 300 on medium hydrocarbon duty
was clearly weeping product from the casting body. The valve had to be isolated
and the domino effect was that the plant had to shut down. How many other weeping valves were on the plant? What a headache to sort out. The root cause was identified to the valve manufacture having changed foundries without informing the end user. Within 3 months the contract was terminated as many more safety issues began to become apparent. For example flange end bolt holes were off center.
Whatever has happened during the valve manufacture the valve is then covered up with the painting system. There is no evidence of any weld repairs. Don’t tell anyone but the informed eye can still identify the not so good valve castings.
The auditor has now completed his tour of the valve manufacture and returns to
the meeting room for closure. A resume of his findings are given. Where such practices are witnessed approval of the valve manufacturer’s sub-supplier foundry is highly unlikely. It may also jeopardize approval of the valve manufacturer. What is frustrating for an auditor is that when you advise the valve manufacturer of your findings and suggestions for improvements it often falls on deaf ears. Too many clients do accept such practices as the bottom line is a certificate of conformance at the cheapest price. When there is another client willing to accept the valve then the valve manufacturer is not keen to invest. Why should he?
Note; Some end users take the precaution of selecting a completed valve off the shelf for API 591 testing. Though there are discussions on where the test samples can be taken and whether they are representative of the casting API 591 is an excellent
standard to gain confidence in the valve. Though the internal condition of the casting can be seen the more valuable information is that heat treatment has been correctly done giving acceptable mechanical results.
Where no such findings are observed then the end users approval is straight forward.
Discussions will develop into the three other valve manufacturer and foundry models in the upcoming articles.
Thank you for your continued support. Please contact me if you have any questions or different views. They are most welcome.
Barrie Kirkman, BSc CEng MIMechE, writes a regular column for Valve World, bringing his own personal views from inside the valve business.
Barrie can be reached on: firstname.lastname@example.org.