El Paso Electric Samples Cooling Tower Gearboxes

One of the first tasks when setting up an oil analysis program is selecting the appropriate sample location for each critical component. Ideally the sample point should be from a turbulent zone, downstream from critical components, but upstream from filters, and most importantly, when the equipment is operating. While satisfying these criteria is a matter of installing a sample valve in an appropriate location in most cases, some components offer more of a challenge. This article outlines the steps taken by El Paso Electric to enable the company to accurately sample cooling tower fan gearboxes, and to reduce the significant cost associated with sampling each gearbox.

Staff Reductions Demand Change
El Paso Electric turned 100 this year and is one of the oldest utilities in the U.S., serving customers in southwest Texas, southeast New Mexico and Mexico. The company has gone through deregulation recently resulting in significantly fewer employees. According to Les Harrington, El Paso’s tribology analyst, “getting an oil sample taken was getting harder and harder, particularly from our cooling tower fan gearboxes, which requires confined space entry.” Yet Harrington and his supervisor, Jaime Viramontes, knew that oil analysis was a vital part of El Paso’s predictive maintenance program, so they set out to find a way around the labor-intensive process of sampling these gearboxes.

The Problem
Sampling cooling tower fan gearboxes is one of the most difficult challenges. At El Paso, prior to the initiatives the maintenance team put in place, this normally simple task was a complex procedure. It involved maintenance and operations personnel, supervisors and planners, in addition to the tribology analyst who actually took the sample. The procedure for obtaining oil samples under these conditions involved:

• Operator to tag and lock out the chlorinator,
• A one-hour wait for residual chlorine to dissipate,
• An operator to turn off each gearbox motor and tag it out,
• Walkways inside stack are slippery and wet, and potentially dangerous,
• The help of an attendant (required for confined space entry),
• Taking of the sample by an tribology analyst and
• An operator to remove all tags and locks and then return the
  gearbox motor to service.

The problem with this scenario was two-fold. First, with so many people involved, it was often difficult to coordinate everyone to be in the same place at the same time. According to Harrington, he would often wait half a day to obtain all the necessary paperwork and personnel to complete the sampling. Second, by the time the sample was collected, the gearbox had been idle for at least an hour. This allowed any contaminants or wear debris to settle, making the sample less than representative. What surprised the team the most was the associated cost of this procedure; the sidebar shows Harrington’s estimate of the cost associated with the old method of sampling. The cost of sampling amounted to more than $20,000 per year, without assessing the cost of analyzing each oil sample, which is done in-house.

The Solution
In researching ways around this problem, Harrington found a procedure already in use by Nevada Power and Light. Its solution was to install an off-line pipe/hose arrangement that took a small side stream sample of oil from the gearbox, and directed it outside the cooling tower. This enabled an oil sample to be taken without tagging out the chlorinator or gearbox motor and without requiring operations assistance or confined space entry. Under normal circumstances, the oil in this side stream is static. However, when an oil sample is required, a small perastaltic pump is attached to a tee on the side stream pipe and the oil circulated for 20 minutes, before the oil sample is taken.

This solution offers two advantages. First, the sample is truly representative of the condition of each gearbox, because the sample is taken while the gearbox is running. Second, because the sample can be taken from the outside, no equipment needs to be shut down or tagged out, meaning that only the tribology analyst is required to take the sample and the fan can continue to run under full load.

Harrington and Viramontes agreed that this procedure would be beneficial to El Paso Electric and implemented it at both plants with the rest of the maintenance team. To date, three cooling towers with a total of 17 gearboxes at El Paso’s Newman Plant, and one cooling tower with eight gearboxes at the Rio Grande Plant have been setup with this new sampling arrangement during routine, scheduled maintenance downtime. At an average cost of $200 per gearbox, plus $1,700 for the perastaltic pump, the initial outlay was fairly high for just setting up oil sampling points. However, based on Harrington’s calculations in the sidebar, the payback in reduced manpower was immediate. Also, once this program was implemented, it eliminated the need for the tribology analyst to enter the dangerous, confined space inside the cooling tower stack.

The Real Payback
One of the first samples taken with this new arrangement was from #4 fan gearbox on the #3 cooling tower at the Newman Plant. The sample showed high levels of large ferrous debris, indicating that the gearbox was approaching failure. These findings were confirmed by subsequent vibration readings by the technician, and the gearbox was taken out of service and found to be close to failure and in need of rebuild. Previous samples from the same gearbox taken with the old method did not find any signs of this problem, presumably because the oil samples were taken with the gearbox out of service, when the oil was static.

The example illustrates the true payback of the initiatives of the El Paso team. Without representative sampling, all other oil analysis benefits can be negated simply by the inability of the sample to reflect the true condition of the equipment. El Paso can now have the confidence that its oil analysis program is capable of finding problems early, avoiding catastrophic failures and unplanned outages.