Oil analysis reduces excessive PM procedures at a southeastern power plant
"We were trying to reduce the cost of our PM program," advised the preventive maintenance technician at a Southeastern power plant. "About a third of our maintenance man hours are spent doing preventive type activities mostly on a calendar basis and we wanted to reduce the time intervals at which we would go inside the equipment. Instead of doing a PM on a piece of equipment every six months, we felt that we could use periodic oil analysis to indicate when to do the PM and in some cases eliminate the need to shut down the equipment altogether."
During the review of the PM program they evaluated several different oil analysis labs and eventually selected Petroleum Analytical Labs, Inc., of Eaton Park, Florida. Important factors in the selection were the convenience of their location, the fast turnaround of samples, and the immediate use of sample data.
"We had two criteria for selecting equipment to be included in the oil analysis program," continued the PM technician. "The first was equipment that was critical in our operation and the second was equipment whose oil was most likely to become contaminated."
"Interestingly, very little oil in a power plant ever runs to its full life we usually change it due to a specific level of contamination, whether it is fly ash, coal dust, sand or water. So for those systems that had a large enough volume of oil we felt, that oil analysis could identify impending or progressing machine problems before they became critical."
"For example, a soot blower lance has a gearbox that turns the lance in the furnace and even though it is an important piece of equipment, the volume of oil in it is very small. Experience has shown that it is more cost-effective to simply change the oil once a year. On the other hand, a bearing on a forced draft fan doesn't have a circulating oil filtering system, and we have only two fans to produce all of the combustion air for the furnace. The equipment has an oil volume of less than 2 gallons, yet it is such a. critical piece of equipment and has such a high probability of contamination that it pays for us to take a monthly sample".
"All of the equipment, in the program receives a spectrographic analysis a water determination, usually by percent and a total acid number. For our turbine oil we do a Karl Fischer water analysis since the specifications require a water determination in ppm and we do a particle count to determine the efficiency of our circulating oil filters. Since we started the program about a year ago we've gradually cut back on the frequency of running total acid numbers since the acidity wasn't changing very much over time."
"The first results were discouraging because we had to go out and pay a lot of attention to the equipment...trying to resolve all of the problems that we found immediately. More than a third of the sample analyses came back with problems for the first several months...and that's good because we were catching the problems early, but it was difficult for the maintenance department to respond to all of them promptly."
"In the long run, however, the program has already started to pay large dividends. For example, after an 8-week outage we sampled the forced draft fans and found a lot of babbit in the oil. Since we had just overhauled the unit, we went back and inspected the bearings and checked for misalignment but still couldn't find the problem. After the fourth oil analysis still indicated a high level of babbitt, we finally found a failure of a shaft thrust spring which had gone undetected, and after replacing it, the babbitt levels went down. Had we not done the oil analysis we would not have known that we had a problem particularly since the system was recently overhauled and the glass on the side of the bearing continued to show clean oil. As it was, the repairs took a few hours during a low-load night shift and we avoided the two or three days of downtime that a more severe bearing failure would have required."
"Another good example is a horizontal air preheater that we had just overhauled and changed the oil on. For the first two months after start-up there was a rapid increase in ferrous metals and I called for a bearing inspection. We eventually found out that the retaining plate bolts were of insufficient strength and had sheared off the inner bearing race had come loose from the shaft, and the bearing had moved out of place. Had the oil analysis not alerted us to pull the unit off of the line, we would probably have had a major bearing failure, and we estimated that it would' have cost us almost $500,000 for repair parts and materials not including the cost of lost generation capacity. As it was the bill came in at about $30,000.
"We had never had this type of situation before, and since we had just replaced, the bearings, it was the type of problem that we never would have suspected. As it turned out, we had been shipped retaining plate bolts that had been incorrectly stamped."
"Once we solve a problem, we often consider reducing the sampling frequency," he continued. "For example, coal is moved around the plant by a series of conveyors, and since the availability of coal is a critical part of our operation, we want to be sure that the gear reduction systems on the conveyors work properly. We were having a problem with contamination from coal dust and water, and eventually corrected it by installing breather filters on each system. As soon as we saw the contamination levels start to drop, we reduced the sampling frequency from monthly to quarterly."
"We had a similar contamination problem with the circulating oil systems on our coal crushers. The design of the system was such that to check the oil level or add oil to the system you had to open the top of the reservoir. Each time this was done coal dust would enter the system, and as a result, we were constantly changing the oil and the filters. Our answer to this problem was to install a bulk oil tank and hard pipe it to each of the reservoirs. Now to refill each system the maintenance man simply opens a valve and pumps clean oil through a meter, directly into the reservoir. It has been almost 12 months since we have had a problem in this area, and the equipment is running longer between repairs."
"In my opinion, there are two different types of operations that still don't use oil analysis," he concluded. "There are those plants that don't care about maintenance and will run their equipment until it fails and there are those who currently have a PM program that could benefit from the additional planning, scheduling and predicting information that oil analysis provides. In our case, I'm sure that we'll be able to reduce our PM activities by over 30% during the next two or three years and we'll be able to take some of the guesswork out of our equipment condition monitoring."