|
Selecting Machines for Oil Analysis
Chris Gillard, Industrial Services Development
Team Leader, Shell Global Solutions
1) Establish what is currently happening on site regarding a)
condition monitoring (all techniques); b) failure modes and rates;
c) any PM/CM procedures; d) any machine criticality info; e) what
CMMS (MMS) usage (failure codes, w/o nos & ratios i.e. reactive
to preventative); f) any plant/machine availability & reliability
info; g) plant annual output & costs (maint. & raw mat.); h) output
revenues (if poss.). To gather this info check-out the CMMS (MMS)
reports (if any); speak to operations and maintenance; check for
any control charts; any purchasing, sales & customer data info.
Also, establish what the business drivers are regarding plant
operations, is it cost reduction, increased quality output, transform
to a lean environment, etc.
2) Depending on the outcome of 1) perform (or check and/or repeat)
a quick criticality assessment (this can always be repeated later
in more detail i.e. streamline or full RCM). Establish which machines
are causing problems, which machines are critical to operations,
and which machines, if they fail, could cause a serious QHSE problem.
This could also be done at the same time as 1) if time is short.
3) Having established high priority machines check a) duty cycles;
b) operating conditions/environment; c) lube specs & maintenance
procedures; d) actual & possible failure modes. This will help
establish sample tests & frequency. Work with the laboratory at
this point so that the lab and diagnosticians can help interpret
data results and set-up tests & frequencies. This may also reveal
the opportunity to do inhouse tests as well as lab tests. Most
labs have standard test suites by machine type, but check that
these tests cover the failure modes. For instance, hydraulic systems
should have particle counts; high temp operations may require
RBOT; if you have a high moisture environment you might want a
crackle (go/nogo test) -> if go check the ppm, etc. The lube oil
supplier should also be able to help establish test suites and
frequencies. If no OA is currently performed this may require
a review of OA labs and a lab contracting phase, make sure the
lab is competent to do what is required and the diagnostician
has experience.
4) Train maintenance & operations on why OA is important to machine
operations, how this fits with maintenance activities, and then
how to take a sample, what the reports will tell them, and what
diagnostics will come out. In particular if the OA is new to site,
appoint OA 'specialists' to get the program running (adds to competency
& status as well as process implementation). These will also be
good contacts for the diagnosticians.
5) Ensure OA procedures (plus relevant KPIs) are added to the
CMMS (MMS) systems and that maint. planners know what these new
tasks are. It might be possible to establish OA routes for your
specialists. Alternatively, operators could take samples. Also,
ensure that the planner gets copies of reports and knows what
to do if a diagnosis suggest a maintenance action. Ensure all
feedback into CMMS. when maint. action complete ensure the diagnostician
is informed.
6) When the high priority machines have been completed establish
OA activities for the lower priority machines. This may also be
the time to complete a more rigorous RCM criticality assessment
now that high priority machines are covered under OA procedures.
See other responses
to this Readers Challenge.
|
