Page 207 - Practical Machinery Management for Process Plants Major Process Equipment Maintenance and Repair

Reciprocating Gas Engines and Compressois 189

Therefore, it is just as important to repair and square the guard. The
third, and correct, approach is to repair both seat and guard. The first
and second approaches are admittedly popular; however, there is no es-
caping the fact that regardless of design, materials, etc., valve failures
will multiply after the first normal expected failure if the assembly is not
repaired correctly.
It is worth mentioning here that Bakelite, Nylon, and other nonmetallic
plates cannot withstand high discharge temperatures and pressure ratios,
but since they do not damage seats and guards when they break, they
should be used when possible.
The practice of lapping valve plates and seats to square and remove
dents is highly recommended, but many people fail to realize the impor-
tance of removing the sharp edges formed in the lapping operation. Con-
trary to general belief, a plate does not lift straight off the seat as it opens;
instead, one side lifts first. Thus, if sharp edges are left on either the
valve or the seat, they will dig into each other. For that reason, all sharp
edges should be broken after lapping.
There are many other examples of bad practices but they apply to spe-
cific types of valve designs and cannot be covered here. The point to re-
member is that valve assemblies are subjected to adverse conditions at
start-up and after overhaul.
After the initial start-up pains, the valve troubles usually level off, but
since anythmg that moves is prone to fail ultimately, it is only a matter of
time until problems will develop and will need attention. Perhaps the best
known tool for dealing with valve failures is a complete history on every
valve assembly. The expense of keeping a history on an installation that
is giving good service is not justified, but once anyone issues a valve
complaint on the unit, it will be profitable to go back in with new valves
and start keeping accurate records. Information such as the life of each
plate, its exact location (Le., crank end, head end, suction or discharge),
and whether there have been any changes in capacity, pressure and tem-
perature canditions is valuable. The number of times that such a history
has revealed the trouble to us is almost unbelievable.
An interesting point often brought out in such a history is that there are
more valve assemblies in the plant than realized and that the actual life is
as much as three years. Sometimes the failures will be repeated in a cer-
tain stage, cylinder, or end, or perhaps in the same valve pocket of the
cylinder. A close inspection may reveal that the seat in the cylinder that
receives the valve assembly was damaged. In one specific case of re-
peated failures in one location, the entire cylinder was found to be dis-
torted due to piping strain. The notation on the positions of the unloaders
is important in the case of some variable piston unloaders. The valve
plate may tend to flutter at very high clearance volumes.