middot warning cancel success information linkedin google twitter facebook whatsapp user-stroke rss yacht-silhouette library user ship tel email print share lock spyglass arrow--down arrow--up arrow--left arrow--right coins city yacht warranty pin

The importance of a correct predictive maintenance.

I had the experience with a couple of companies trying to perform some works on board. "We can take the oil sample from the oil pan with the stopped engine."

We (Engineers) need to understand that we won't know how the real condition is on our machinery if we do not have the proper analysis result. It is our responsibility.

Oil analysis is a condition monitoring tool designed to evaluate:

1-The properties of the lubricant

2-Lubricant contamination

3-Machinery Wear

However, the analysis of a sample depends mainly on the quality of the sample itself. A high-quality sample must be data-rich and interference-free.

From my perspective, this is the list of what is due and shouldn't be done for oil sampling:

Samples should be taken with the machines operating. Do not take samples in "cold" systems. This practice goes beyond simply putting the machine into operation to take the sample. The philosophy of oil analysis is to take an "instant" picture of the system at the time of sampling. That right moment should be when the system is under the most significant amount of effort. Typically, the best time is when the system is in normal operating and charging conditions. This task is complicated when a sample must be taken on a machine that operates through production cycles (intermittently). Under these conditions, we are going to get a representative sample of that condition that could cause accelerated machine wear.

What should and should not be done when taking oil samples:

1-   Samples should be taken before the filters and the output of the machine components (pumps, etc.). The filters are designed to capture particles of wear and contaminants, so taking samples after these data subtractors doesn't add value. On the contrary, taking a sample before and after a filter to affect a particle count will allow you to see how well it's operating at that time.

2- Specific written procedures should be made for each system. The sampling methods or ports should not be changed. Everything we do in oil analysis and machine lubrication must have a detailed procedure that supports the task.

We need to identify the sampling port, the volume of oil to be purged, the sampling frequency, the right time for sample taking in a cyclic system and indicate which tools and accessories to use at that specific sampling point based on the type of lubricant, pressure and amount of oil required.

3-   A proper purge procedure should be done on the sampling valves and the sampling devices before taking the sample. Do not use dirty sampling equipment or reuse the sample hoses and bottles. Cross-contamination has always been a problem in oil sampling. The reality is that purging a system is an important task that is often ignored. Failure to properly purge the sampling port will produce a sample with a high level of failure. Purging the dead space will also purge your other accessories, such as the sample valve adapter and the hose. A proper purge procedure should be done on the sampling valves.

Predictive maintenance is developed to save resources. It is an investment, not an expense.

Ch Eng Rodrigo Pizzarro Marquez