|
Oil Analysis Without Particle Counting?
James Fajcz, Reliability Engineer/Supervisor,
Georgia Pacific Corporation
I would first determine where the samples are being taken in
the lube system. If they are always clean, I would suspect that
the samples are being taken "after" the filters or just prior
to the lube point. "Fewer than one percent reportable" is what
I would expect if the filters were doing their job.
If not already done, I would setup sample points to determine
the condition of the equipment, and oil. Before filters on the
return side to determine the condition of the oil coming back
from the lube points, and then after the filter (sampled when
necessary) to determine the cleanliness of the oil going to the
lube points (bearings, sleeves) and the efficiency/effectiveness
of the filters.
Also setup sample points just after each lube point to determine
the condition of each lube point. Setup the routine sampling as
follows, including particle counts:
1) Determine the cleanliness of the new or makeup oil entering
the system (from the barrel, or tote, or other source of makeup
i.e. particle count).
2) sample on a routine basis only from the return line of the
lube points just prior to entering the filter. (either by hour
or by another set period).
3) based on your trends, if you see an increase in one of the
testing areas (particle counts are good for this), resample, but
sample at each of the lube points upstream. This will indicate
the condition of each of the lube point, and which lube point
is causing the increase in the trend.
4) So, not only will you determine if the oil is dirty, you can
determine which point is causing it, and you can correct the issue
before it causes downtime. 2 birds with one stone.
See other responses
to this Readers Challenge.
|
