With Oil Analysis "What You Know is What Counts"

Matt Spurlock, Noria Corporation
Tags: oil analysis

For the well-honed reliability program, knowledge equals power and financial success. Without an appreciation of how to derive knowledge from data, it is not difficult for the poorly informed, or misinformed, to conclude that oil analysis is a waste of precious dollars. Let’s look at a case in which site management was convinced that oil analysis was wasting money and simply wanted to rely on vibration analysis with no assistance from oil sampling.

Case Study
A recycle pump had been sampled regularly on 30-day intervals since October 2003. Oil analysis results continuously indicated high levels of dirt and water contamination with high levels of wear expected with continuous and appreciable contamination. The laboratory repeatedly made recommendations to check for the source of abrasives and water contamination, followed by recommendations for disassembly and inspection of components for severe wear.

In September 2004, a replacement pump was used while the original was torn down and inspected. To the personnel’s dismay, mechanical wear appeared minimal with the only notable discrepancy being minor shaft wear at the seal point. The shaft was resurfaced and placed in stock as a spare.

In the meantime, an oil sample was pulled from the new pump placed in the same operating conditions with identical contamination control measures. The results of this sample were frighteningly similar to the samples from the previous unit. The laboratory analyst was not aware of the replacement of equipment; however, the reduction in wear debris and change in lubricant properties did prompt the analyst to assume that a lubricant change had occurred. Figure 1 shows a brief history of the previous pump results as well as samples taken from the new pump in September 2004.

Spectroscopic Analysis
Date
9/22/2004
8/19/2004
7/22/2004
5/13/2004
Lab No.
206227
204170
202511
198382
Iron
267
454
1122
2979
Chrome
3
6
13
33
Nickel
8
41
108
108
Copper
25
118
216
274
Lead
9
4
4
4
Tin
0
0
1
1
Silver
0
0
0
0
Aluminum
8
9
57
102
Silicon
55
216
607
1668
Sodium
7
0
7
18
Boron
24
3
12
17
Molybdenum
1
1
1
2
Magnesium
5
4
15
18
Calcium
268
125
237
224
Barium
0
0
0
1
Phosphorous
246
420
380
475
Zinc
34
428
212
135

Figure 1

Given the consistently high wear levels from this particular machine, from an analyst’s perspective it became evident that it was time to have a long discussion with site personnel in hope of discovering the factors that lay at the root of the symptoms. By the time this discussion occurred, the site management had concluded that oil analysis was not reliable and was not worth the expense. This reasoning grew out of two considerations: First, there was apparently contradictory “evidence” from the vibration program suggesting no sign of a problem. Second, the new pump samples indicated high wear and contamination numbers (just like the old pump). Without an obvious explanation for the results, it was concluded that these results in particular were not dependable, and by default the science was also not trustworthy.

A private conversation with the site vibration analyst revealed several thoughts, making it apparent that there was a lack of appreciation for the value of oil analysis.

The site personnel could not understand why a brand new unit could show the same type of data as a used unit. They did understand the phenomenon of break-in wear and new unit contamination; however, they also understood that the values of both pumps were too similar to be attributed solely to either condition. With regard to contamination, site personnel could not see how “all of this could be getting into the pump.” After thorough prodding of environmental conditions, this key statement was made: “I can see the possibility of silicon getting in there, but nothing else.” It was also mentioned that the desiccant breather consistently became spent within a week.

Silicon is a key element in dirt. When seen on an oil analysis report, silicon in combination with aluminum, is notorious for being a strong dirt indicator. Where dirt can enter, so can other contaminants such as water (remember, water and dirt contamination were present in this component since 2003).

At this point, it is important to draw attention to a group of graphs to show the powerful results of oil analysis (Figure 2).


Figure 2

As can be seen, the wear levels are directly related to the dirt ingression levels. These patterns are textbook lock-step trending patterns and indicate a condition that cannot be detected through vibration analysis, as was the site’s original belief.

Further conversation and photo examination of the unit in current operating conditions indicated that this pump was subjected to constant dirt and water falling from above. It was this dirt and water that consistently contaminated the pump. A cover was proposed to divert the water and debris away from this obvious ingression-sensitive machine. At the time this article went to press, wear debris and contamination were both at their lowest levels since data was first sampled in October 2003.

The lack of knowledge of the actual operating conditions severely limited the original analyst’s ability to provide adequate evaluations and recommendations. In addition, the manager’s lack of knowledge of the limits and abilities of oil analysis, and a poor understanding of the meaning of the data, severely hindered site confidence level in the oil analysis data.

The Lesson for Today
Proper education and training are vital to maintaining high levels of confidence in the technology, to deriving knowledge from the data points, and ultimately to demonstrating rock-solid evidence of a return on the meager investment. This will result in improvement across the board and create an overall positive impression.


About the Author
Create your survey with SurveyMonkey