The client needed the acceptance of the Bank's Engineer on the NPSH margin to get funding for the project so has accepted the Bank's Engineer's requirements and paid extra for a revised design of boiler feed pump.My recollection is that the Bank's Engineers requirements are based on ANSI/HI 9.6.1 even though it has been withdrawn.
The standard was obviously withdrawn after their lawyers saw it.  Logically its simply a no win situation for HI.  Too little NPSHA = potential liability for any pump damage; too much, and capital cost increases resulting in nobody using it in either case.  Why do we try to dumb these things down all the time by trying to make up some "standard".  Especially where standards don't fit.  Standards only fit standard situations.  NPSHA isn't one of those.  Look at all the variables listed above.  If there ever was a place for a NO standard, NSHA + a "margin" has gotta be one of them.  HI realized exactly that.  My hat's off to them.  Somebody finally let engineers be engineers ... instead of standard writers and standard followers. 
With all due respect, claiming the standard was withdrawn because of the lawyers is pretty speculative.  It is also incorrect IMO.As I mentioned earlier, I attended a number of the HI NPSH standards meetings over a space of about 3 years.The old standard was withdrawn because nobody among the manufacturers and users could agree on what constituted acceptable margin.  The only thing they mostly agreed on was that they didn't like the standard so out it went.Unfortunately last time I checked the new standard was mired in arguments about the basic approach.  Some people wanted simple margins, some wanted a Nss based approach etc.I also disagree that there is no place for a standard on NPSH or if you prefer, some Split case pumps basic guidelines.Look at the example of API 610.  That talks about Suction Specific Speed (Nss) but leaves it open to the engineer to define the maximum allowable.  As a consequence you have a horrible free for all where specifying engineers (who are not usually pump experts remember), require all manner of odd things.  I've seen requirements for Nss <12,000, Nss <11,000, Nss <9500 all with no consideration of the type of pump or the pumped fluid.In some applications this results in a grossly oversized pump.  In others I've seen it result in dangerously small NPSH margins.One can have too much faith in engineers knowing how to do the right thing.

2011-08-16

Back