FPL Energy experienced multiple failures of the thrust bearing in the single turbine-generating unit at its 6-MW Cataract plant in Maine. To solve the problem, FPL Energy installed a new eight-pad, spring-supported PTFE thrust bearing and a new thrust block. The retrofitted unit began operating in July 2006 and has been failure-free ever since.
By Paul J. Plante, Eric D. Soule, and Mike A. Dupuis
FPL Energy’s 6.65-MW Cataract project is a run-of-the-river hydro facility on the Saco River in Maine. The station has a single Kaplan turbine-generating unit that began operating in 1939. Between 1959 and 2005, the unit’s thrust bearing failed eight times, with half of the failures occurring between 2003 and 2005.
To deal with the situation, FPL Energy installed a low-profile, eight-pad, spring-supported thrust bearing and a new thrust block. This modification solved the problem – the unit has operated since July 2006 with no thrust bearing failures.
Problem with the thrust bearing
The spring-bed babbitt thrust bearing at Cataract is above the rotor in the upper bridge, which also houses the upper guide bearing. There is a lower guide bearing under the generator rotor and a water-lubricated turbine bearing in the head cover. In 1959, FPL Energy repaired the thrust bearing because it had suffered from eccentric wear over the initial 20-year operating period. The eccentric wear was believed to be associated with concrete growth at the station. The repair work included installing a sleeve on the thrust block. Since that repair, the thrust bearing failed eight times, with four of those failures occurring since 2003.
In 2004, FPL Energy took the unit out of service to repair an oil leak in the Kaplan head. When the unit was disassembled, personnel discovered two significant adverse conditions. First, the babbitt shoes on the thrust bearing were cracked. Second, misalignment of the powerhouse as a result of alkali-aggregate reactivity (AAR) had progressed to such a degree that the unit centerline needed to be reestablished. Work to correct these two problems took about ten months.
In June 2005, personnel began to start up the rehabilitated unit. Personnel conducted mechanical runs and then initiated an auto-start sequence. Within 30 minutes, the unit tripped as a result of high thrust bearing temperature. Personnel performed an inspection after the trip and discovered a severely wiped bearing with a babbitt-filled oil reservoir.
FPL Energy personnel then conducted an investigation to determine the cause of failure during start up. During disassembly of the failed bearing, personnel discovered that the round keys that hold the split thrust runner halves to each other were distressed. The two keys are held in place by set screws. Personnel found one ejected key in the thrust bearing oil reservoir; the other key was still in place. Both keys had sheared set screws. And, personnel noticed displacement of about 1/16 of an inch between the thrust runner halves. However, they were not able to target a conclusive root cause for the failure.
To recover from this failure, personnel first reengineered the thrust bearing components. They installed a new split half thrust runner that included a robust key set. Additionally, personnel were concerned that rebabbitting the original backing plate might result in warping. Instead, they decided to install a two-piece babbitt plate. Personnel reassembled the unit and prepared to restart it in September 2005.
During this start up, personnel developed a start-up procedure, intended to address potential issues from the June start-up failure. This included a program of progressive starts and stops consisting of mechanical runs at various speeds, speed-no load runs, and runs of varying duration. Personnel also conducted intermediate inspections and cleaning and scraping to check for damage.
The second start up progressed normally through a run that included flashing the field. The unit was then auto-started and synchronized. Thrust bearing temperatures started to climb dramatically and the unit tripped within three minutes. Upon disassembly of the unit, personnel discovered a preferential wipe in the babbitt that was so severe that the thrust bearing components would have to be either repaired or replaced. Personnel also noted displacement between the two halves of the thrust runner, despite the enhancements made to improve rigidity and stiffness of the keys.
At this time, personnel completely removed the thrust bearing from the unit. FPL Energy then assembled a ten-member multidisciplinary team to determine the root cause of the thrust bearing failures and the appropriate corrective actions.
The team worked on the problem for six months. They performed an exhaustive study, evaluated the failed components, and consulted with several thrust bearing performance experts. Eventually, the team came up with one potential root cause and six contributors that enhanced the likelihood of the root cause. The team determined the likely root cause of both failures was the marginal load capacities of the original bearing (subject of the initial start-up failure) and of the reengineered bearing (subject of the second start-up failure).
This thrust bearing, from the 6.6-MW Cataract plant, failed during start up of the rehabilitated unit in June 2005. The oil reservoir of the bearing was filled with babbitt as a result of severe wiping of the bearing.
It has been widely reported that two-piece babbitt bearings on spring beds in hydro service have lower load-bearing capacity than more modern independent pad bearings.1,2 In the case of the thrust bearing at Cataract, calculations indicate that the design load is within 10 percent of the limit for babbitt, which is generally accepted to be 400 pounds per square inch (psi).3 With such a small margin between the design load and the load limit for babbitt, along with other factors at the station – including the situation with AAR that will progressively increase the amount of misalignment – FPL Energy’s focus moved away from refurbishing the existing two-piece spring-bed bearing to retrofitting the unit by installing a higher-capacity bearing.
The thrust bearing assembly at Cataract consists of two flat annular rings, one fixed to the shaft and one to the housing. These rings turn against each other in a flat-bottomed “pot” that holds a bath of turbine grade oil. The bottom ring (“bearing plate”) is fixed to the pot bottom and has eight radial grooves to promote continuous feed of oil between the surfaces of the two plates. This bottom ring has a babbitted bearing surface. The top rotating ring (“runner plate”) turns with the shaft. As is typical with this style of bearing, the runner plate also has six radial grooves.
The typical flat plate thrust bearing design was used on ships for many years. The original design of this type of bearing came as either one complete base ring or two base ring halves. Both arrangements had babbitted bearing pad sections bonded to the base plate. This arrangement was frequently held by blocks in four locations on the outside diameter of the bearing base plate(s). The thrust runner plate for this bearing was a micro-finish plate with radial grooves on the running surface. These grooves acted as a centrifugal oil circulating pump for the bearing pot. This design was used in many hydroelectric turbine-generators as late as the mid-1940s.
After this time, bearing designers understood that a tilting pad plain bearing has a two- to four-times greater load carrying capacity than the flat plate design. A flat plate babbitt bearing design can safely operate at about 150 psi, compared with the 400 to 450 psi of a babbitt tilting pad design. The polytetrafluoroethylene (PTFE) tilting pad bearing has a design load of 943 psi, which is more than twice that of babbitt. All three styles of bearings have operated successfully at greater than the design loads. However, the design safety factor is reduced.
Flat plate babbitt pads are bonded to a single base plate, keeping them parallel to the runner (rotating) plate. Lubrication is drawn onto a bearing pad’s surface by oil sticking to the runner plate as it rotates. This creates an “oil wedge” at the leading edge of the bearing pad. As oil travels along a pad, some oil leaks from the inner and outer pad edges, thereby leaving less oil pressure at the trailing edge of the bearing segment (see Figure 1). This is an effect of the bearing pad’s parallel position to the running surface, resulting in reduced oil pressure at the back of the pad due to leakage. This leaves only the leading edge of the pad to support the majority of the thrust load. The effective working area of the flat plate bearing pad is only about one-third its surface area.
Figure 1: With a flat plate design for an oil wedge, oil leaks from the inner and outer pad edges, leaving less oil pressure at the trailing edge of the bearing segment. This means only the leading edge of the pad supports the majority of the thrust load.
By contrast, a modern bearing design uses a forward-tilting pad to create a smaller gap between the runner plate and the back of the bearing pad. This allows the trailing edge of the bearing pad to maintain pressure, despite reduced oil (see Figure 2). Thus, a more even distribution of load is created along the entire bearing surface, increasing the thrust carrying capacity.
The thrust runner plate resembles a runner plate of current design, except for the presence of radial grooves on the face of the running surface. Originally designed to circulate oil in the bearing pot and to wash cool oil over the face of the babbitt pads, these grooves actually contribute to the bearing’s weakness. As the radial grooves pass over stationary pads, the oil wedge is broken, allowing pressure developed to support the thrust load to drain from the groove. This results in an unequal distribution of load on the thrust bearing pads by changing pressure across the bearing pads from maximum to zero pressure (six or more times per revolution, depending on the number of radial grooves).
Figure 2: With the tilting pad design for an oil wedge, there is a smaller gap between the runner plate and the back of the bearing pad. This allows the trailing edge of the bearing pad to maintain pressure, despite reduced oil.
Such extreme pressure changes inevitably cause problems. “Oil pressure hammer” may cause fatigue cracking of the babbitted surface. In other cases, pressure drop on a babbitt bearing pad has resulted in cavitation. In both situations, the higher the bearing is loaded, the more prone it is to failure. These failures are sometimes difficult to detect due to the nature of their occurrence. The bearing operates perfectly while slowly deteriorating. Then, after several years of operation, the bearing suddenly self-destructs, leaving little evidence of the root cause.
Choosing a thrust bearing design
FPL Energy chose the PTFE tilting pad for its Cataract plant for several reasons. These include:
– Superior thrust capacity (more than double that of babbitt);
– Need for less auxiliary equipment (a high-pressure system is not required for start up and shutdown), producing a lower capital and maintenance cost;
– The brake can be applied at a lower speed because the PTFE performs better at lower speeds, saving wear on the brakes and brake ring; and
– Higher reliability during start up because fewer operations are required during the start sequence.
In addition, PTFE bearing pads do not fail in the same way as babbitt bearings. PTFE wears over the life of the bearing pad. If a failure begins, the pad temperature will begin to rise, and bearing pads will begin to wear as opposed to failing. This allows an operator the chance to schedule an outage and repair the root cause without replacing the bearing pads. This ability has proven effective in the past during severe misalignments and guide babbitt bearing failures, which are contained in the same bearing pot.
Although some people consider PTFE bearing pads to be a new technology, they have been in commercial operation for 30 years. This technology has been used in plants in Europe and Asia for many years. Some North American companies are turning exclusively to PTFE for the reasons stated above. There are more than 800 PTFE thrust bearings installed around the world.
Installing the new thrust bearing
Conversion of the flat plate bearing to a tilting pad arrangement presents obstacles that are inherently difficult to overcome:
– Finding a way to hold the individual pads in place;
– Frequently, the thrust pad arrangement is very low profile, making it difficult to fit a tilting bearing in the height available; and
– Maintaining the ability to easily adjust the bearing pads for alignment and even pad load distribution.
To have six, eight, or more tilting shoes replace the old bearing requires a design that can withstand torque developed by the turbine on the bearing. Each pad must be held individually to the bottom of the thrust pot.
In many cases, the original design has a very low profile. At Cataract, the combined support system and bearing pad height was only 4.06 inches. Having such a low profile made it difficult for any individually adjustable shoe arrangement. To allow individual shoe adjustment, the bearing pot would likely have to be modified.
FPL Energy contracted with Hydro Tech Inc. to overcome this obstacle by designing a new way in which to support the bearing and hold the pads without modifying the bearing pot. Once the four blocks for the old bearing were removed, two tradesmen installed the new support system in a single morning. The new, rigged support system supplies a greater thrust load capacity than the previous bearing and, because of the greater efficiency of the PTFE bearing (three to five times lower coefficient of friction), results in reduced torque load.
The new Cataract bearing now operates at 50 degrees Celsius (C), about 30 degrees lower than the previous operating temperature. The oil bath temperature was reduced by 25 degrees C. This is especially important during operation in the summer months, when the unit has a history of running hot due to high ambient temperature and increased river temperatures (cooling water to the thrust bearing oil reservoir).
FPL Energy has used this bearing solution successfully on both Kaplan and Francis units. At Cataract, a number of factors caused bearing failures over the past 50 years, and particularly during the start ups from the major overhauls in 1957, 1989, and 2005. This included a bearing design capacity that was marginal for the conditions, increased load on the bearing due to misalignments caused by concrete growth associated with AAR, and decreased bearing capacity due to past modifications implemented to a bearing design that is not fully understood.
The new polytetrafluoroethylene (PTFE) bearing for the unit at the 6.6-MW Cataract plant is made with eight tilting shoes held individually to the bottom of the original thrust bearing. This arrangement supplies a greater thrust load capacity than the old babbitt bearing.
The station was retrofitted with a PTFE design using individual pads and springs in the existing bearing envelope. The goal was to install a bearing design that could tolerate a wide variety of operating and induced loading without significantly affecting bearing design margin capacity. This was achieved, as evidenced by a successful start up and subsequent operating history. Both bearing operating temperatures and oil bath temperatures were reduced by about 30 and 25 degrees C, respectively.
About 18 months after the thrust bearing was installed, the upper guide bearing failed. This guide bearing is in the same bearing pot as the thrust bearings, about 12 inches above the thrust pads. Babbitt from the guide bearing was circulated throughout the bearing pot and over the PTFE thrust bearing pads, but performance of the thrust bearing was not affected. After close inspection of the thrust bearing pads, FPL Energy personnel cleaned and reinstalled the pads. No measurable wear or significant damage was observed on the PTFE surface. The bearing continues to operate as effectively as when it was first installed.
- Fergusen, J., J. Medley, and B. Boueri, “Improving Thrust Bearing Performance When Upgrading Older Hydroelectric Generators,” Waterpower XIV Conference Technical Papers CD, HCI Publications, Kansas City, Mo., 2005.
- Moss, W., and R. Knox, “PTFE Thrust Bearings for Hydro Generators and their Applications to Dinorwig and Ffestiniog Pumped Storage Power Plants, UK,” HydroVision 98 Conference Technical Papers CD, HCI Publications, Kansas City, Mo., 1998.
- Standard Handbook for Mechanical Engineers, Seventh Edition, Desert Sage Books, Moreno Valley, Calif., 1967.
Paul Plante, P.E., and Eric Soule, P.E, engineers with FPL Energy, Maine, were responsible for identifying the root causes of the bearing failures at the 6-MW Cataract plant and identifying and implementing corrective actions to return the unit to service. Mike Dupuis is president and chief executive officer of Hydro Tech Inc., which supplied the new thrust bearing for Cataract.
This article has been evaluated and edited in accordance with reviews conducted by two or more professionals who have relevant expertise. These peer reviewers judge manuscripts for technical accuracy, usefulness, and overall importance within the hydroelectric industry.