After experiencing two hydraulic pump seal failures in as many months on a pump that normally “runs forever” with no problems, the maintenance staff at Alcoa Mill Products, Lancaster, Pa., decided that something out of the ordinary was going on.
Even though the routine oil analysis results indicated the pump and its lubricant were normal, the oil’s strange color prompted the staff to investigate further. A Fourier transform infrared spectroscopy (FTIR) analysis was conducted, which led the staff to the root cause of the seal failures - cross contamination.
Alcoa Mill Products’ Lancaster Works facility is a 1.3 million square-foot plant that produces common alloy products, specializing in aluminum sheet and cast plate. In late Fall 2003, a hydraulic pump seal failed.
The pump was part of a new hydraulic unit that had been in operation only about six weeks. This particular type of hydraulic pump was typically found to be very reliable, with a long meantime between failure.
However, because the oil that vented off a system filter vessel was an odd color, a cardboard tan rather than the usual red color, the oil analysis results on file were reviewed to see if anything looked out of place.
The oil analysis results indicated that everything was in order and still within limits. It was noted that the seal failure on the hydraulic pump would allow oil to escape on the pressure side, but would not allow contaminants to enter the system against the pressure.
Given that only a small amount of oil was lost, topping off the system was not recommended. In addition, it was decided that another oil sample did not need to be taken at that time. However, the reason the oil had changed color was still not understood.
During early December 2003, the same seal on the same hydraulic pump began leaking again. This was the second seal failure in two months for this unit. Therefore, the decision was made to investigate the discoloration of the oil - a water glycol hydraulic fluid.
Previous oil sample results from Alcoa’s off-site lab, were compared and nothing out of the ordinary was noted. Fresh samples were taken so that the viscosity could be compared to the viscosity of the previous month’s samples. It was then discovered that the old sample from the system filter vent had a layer of an unknown substance floating on the surface. These samples were taken to the plant chemistry laboratory for an FTIR scan. The FTIR scan initially revealed that the oil was a petroleum-based lubricant, not a water glycol hydraulic fluid.
When the sample was run on a centrifuge, the unknown substance separated to the top, allowing the water glycol hydraulic fluid oil to settle on the bottom. The oil on the bottom was a distinctive red color, much different than the off-colored sample that was delivered.
The fluid from the top layer was used for a second FTIR scan. Further testing revealed that the fluid in the top layer was a match for a petroleum-based lubricant oil. The only petroleum-based lubricant used at the facility is used for casting equipment. The petroleum-based product’s data sheet states that it is a clear, dark brown oil.
Figure 1. Water Glycol Hydraulic Fluid Contaminated Sample
Figure 1 is an FTIR scan of the contaminated oil sample, Figure 2 is the FTIR scan of the subtracted sample after centrifuging and Figure 3 is a representative sample of the petroleum-based lubricant.
A comparison of the subtracted scan (Figure 2) and the representative scan (Figure 3) was close enough to confirm that the fluid was in fact the petroleum-based lubricant. At this point, it was realized that the contamination of the water glycol hydraulic fluid by the petroleum-based lubricant was the likely cause of the pump seal degradation.
Because the pumps are ordered with seals that specify a water glycol lubricant, they would have to be a different composition to withstand a petroleum-based oil.
Figure 2. Suspect Fluid Subtracted Sample after Centrifuge Separation
Based on the FTIR results, it was recommended that the whole system be drained, the filters changed and all vessels be wiped down prior to refilling with fresh water glycol hydraulic fluid.
Since then, there have been no problems with the system and no further seal failures. To ensure that these seal failures do not occur again, the facility’s mechanics have been reminded of the importance of using the correct equipment and fluids for system top-ups.
Figure 3. Petroleum-based Oil Representative Sample
Whenever a significant color change is observed in any lubricant, it is often a good idea to determine the root cause of the color change, because, just like in this case, the color change can often be symptomatic of a more serious problem.
While not always the right choice, FTIR is an excellent tool to determine the origin of chromophoric (colored) material in a lubricant. Because FTIR excels at detecting organic molecules (in this case, a petroleum-based fluid), it can often reveal problems other oil analysis tests are unable to detect.
The author would like to thank Harold Matroni for providing the information for this article.