Modifying Generator Shafts at the Lookout Shoals Project

A problem with bearing cooling at the Lookout Shoals plant helped Duke Energy uncover several other issues that needed to be resolved. Through creative approaches and significant machining work, the units are now operating dependably.

By Tyler Barrett and Russell Martin

The use of water-cooled turbine bearings at hydroelectric plants eliminates environmental concerns with regard to getting oil in the water, but these bearings do not always work as desired. The Lookout Shoals plant on the Catawba River near Statesville, N.C., was built in 1915 and has three 8.5-MW units (Units 1 through 3) and two 300-kW units.

Unit 1 had low cooling water flow to the turbine bearing, resulting in overheating failure. Duke Energy needed to replace the bearing and turbine shaft bushing. While replacing this bushing in place, additional concerns arose because the turbine wear rings were galling to the point of needing to completely disassemble the unit for replacement. Plant personnel took advantage of the disassembly and machining of the parts to modify the generator shaft to eliminate the threaded nut arrangement that supported the unit. The original thrust assembly was creating static shaft runout seven times greater than specification.

This article covers the machining work performed during the turbine bearing and bushing replacement, along with the innovative method used to remove the generator shaft for redesign.


Lookout Shoals is the fourth of 13 hydroelectric stations on the Catawba River between Marion, N.C., and Camden, S.C., owned by Duke Energy. The Lookout Shoals units have about 76 feet of normal gross head and discharge 1,500 cubic feet per second of water at 11,700 horsepower at a best efficiency point of 90.5%.

In the spring of 2009, Unit 1 showed signs of decreased water flow in the lignum vitae bearing. The bearing had been installed since the early 1990s, with no issues. The bearing was pulled up once the issue was discovered to clean out the slots, as freshwater clams and silt make their way into the bearing and slow the water flow over time. During the bearing removal, a crack was found on the lower third section of the bushing. The naval bronze bushing was original to the unit and had exceeded its life expectancy. To plan for a repair, the seal of the bearing was moved up and the lower third of the bushing was removed. The bearing is 52 inches long, so plenty of material was left to endure the load.

Planning the repair

The plan was to replace the bushing material in place and machine the bearing with new material to match. The old bushing was removed and the shaft was field machined by Duke Energy personnel. An electric-operated slide allowed the bushing to be machined with a single-point tool, and a hydraulic wheel attachment was used to polish the sleeve. The shaft was held in place with rollers that bolted to the head cover extension. The unit was rotated using hydraulic drives attached to the bottom of two bridge tree legs with a rubber wheel on top of the rotor.

To replace the Unit 1 bushing, the old bushing was removed and the shaft was machined in the field. This photo shows the field machining slide and support bracket with rollers.

During machining of the turbine shaft bushing, the runner seals showed signs of smaller clearance compared to previous measurements. The wear rings were a rotating nitronic 60 material and a stationary 304L stainless steel. The two materials were suspected to be galling, which was closing up the clearance in various areas of the wear rings. The design clearance of the upgraded 1990s vintage runners was 0.086 inch to 0.096 inch diametrically. The most clearance found diametrically at this time was about 0.035 inch.

The original design of the threaded generator shaft for the units at the Lookout Shoals plant is shown, with the thrust runner and cradle in place.

An attempt to cut clearance between the wear rings, using a carbine saw blade and diamond-coated blade, was made without success, so the unit was set up out of plumb slightly due to the settling of the Lookout Shoals powerhouse. While Duke Energy personnel were setting the unit out of plumb, another issue arose, with the static runout of the unit being 0.1 inch or more. The turbine shaft sleeve was finish machined and the turbine bearing was machined skewed to match the shaft angle and head cover extension. After the turbine bearing installation, the runner seals were still tight due to the galling. Too many variables were in play to allow the unit to run smoothly, so a complete disassembly was required.

Performing further repairs

Disassembly of the unit went smoothly and other repairs were discussed. Lookout Shoals was the first plant to go through Duke Energy’s hydrovision program, in which the units were updated to modern controls and other equipment. Because these were the first units, they did not receive field machining to level up the embedded flanges where the head cover and bottom ring bolt down. Duke Energy’s equipment and operator were used to machine the 148-inch-diameter head cover flange and 139-inch-diameter bottom ring flange.

The field machining results on Units 1 and 2 were both within 0.004 inch of level across the diameters. The laser tracker was used to set up the boring bar to plumb. Different methods were used for setting up the bar. The first method was measuring a level plane with the tracker and then measuring the bar as a cylinder. The bar could be adjusted to plumb by use of the jacks that secured it in place on the lower bracket. The second method used seemed to be easier. The tracker was used once again and the level plane was measured. A coordinate system was created and a live readout was given. The SMR (spherically mounted retroreflector or target for the laser tracker) was attached to the end of the arm where cutting would take place and the live readout was recorded at various positions of the arm. With the live readout, the position of the bar with respect to plumb was easily detected.

The disassembly of Unit 1 showed signs of wicket gate bushing wear. Measurements of the stems and bushings showed excessive clearance, so the bushings were replaced by an outside shop during the other machining. It was learned from the bushing replacement, using a local shop, that the bores need to be machined on a vertical boring mill with a live head. Unit 1 had the bores machined on a horizontal mill, and the sag of the bar threw the holes off by about 0.005 inch from the intermediate bushing to the top bushing of the head cover.

The wicket gates on Unit 1 were in poor condition, but due to the lead time required and budgetary constraints, new gates were not obtained. The wicket gates on Unit 2 were replaced with ASTM A487 grade CA6NM class A.

Another issue was the static runout on Unit 1. Being a vertical Francis unit, static runout develops from the thrust bearing. The original 1915 thrust bearing design consisted of a threaded nut supporting the unit and distributing force through the thrust block to the thrust runner and then to the babbitted shoes. The threaded nut was creating all the static runout during initial field machining of the turbine shaft sleeve. The thrust nut wasn’t giving consistent static runout results as it would take up the clearance in the 0.5-inch-thick threads differently each time the rotating weight was lifted on jacks. With the complete disassembly of the unit, the opportunity for a new designed thrust bearing was utilized.

The thrust bearing supports the rotating weight and hydraulic thrust of the unit. The rotating weight — the rotor, generator shaft, turbine shaft, runner and nose cone — is a total of 160,980 lb. The hydraulic thrust was calculated to be 85,578 lb and is derived from the discharge diameter of the runner, head on the runner and thrust coefficient. Based on this information, the thrust bearing was redesigned to remove the threaded nut arraignment. Duke Energy engineers felt the best results would be achieved with the generator shaft in a lathe at the shop. The generator shaft had been removed at other locations with relatively minimal effort, but for the Lookout Shoals plant it proved to be challenging.

The new thrust assembly for Unit 1 features a new split ring, block and runner.

The design of the Lookout Shoals rotor is different from Duke Energy’s other conventional hydro plants. The original drawings were lost over the years, so the exact interference of the shaft to the rotor hub was unknown and the shaft had never been removed. The interference was estimated to be about 0.01 inch based on other units of that vintage but resulted in 0.09 inch average. The rotor is designed with an outer cylinder the rotor poles attach to as one piece and is shrunk onto the inner cylinder and pinned. The inner cylinder has six spokes that attach to a two-piece rotor hub. The rotor hub is split in half, with three legs connecting each. The rotor poles were in good shape and were not removed, which in turn left the outer cylinder in place. The rotor hub has six split bolts that hold the hub together and a 3-inch gap on each side of the two halves of the hub. The generator shaft has a shoulder the rotor hub rests on and a retaining ring on top to keep it in place.

The first attempt at removing the generator shaft involved trying to get the rotor hub to open up with force and pressure on the shaft. Wedges and jacks were placed in the split of the rotor hub and heat was applied to the outside of the outer rim. The split bolts were loosened, split ring was removed from the top, and hub was soaked with penetrating oil. While the load was applied to the jacks on the hub, heat was applied directly out from the rotor hub splits to get the outer rim to grow.

Galling on the rotating wear ring of the runner occurred due to the plant settling and the wear ring material selection. These rings were replaced with 304L stainless steel.

No movement of the generator shaft was observed, so force was applied to the bottom and top of the shaft. Four 1-inch B7 all-thread rods were anchored into the concrete floor of the basement. The rods went through the coupling bolt holes and 60-ton hollow ram jacks were placed on top of the coupling. Along with the jacks on the bottom, a strongback arrangement was placed on top that was connected with all-thread rod to the top of the rotor hub. On top of the strongbacks were placed four 60-ton hollow ram jacks. The rods on the bottom were loaded up to 25 tons total and the jacks on top were loaded to 50 tons. This approach did not result in any movement of the shaft.

The next idea was to get the spokes of the rotor to shrink and draw the hub with it toward the outside diameter of the rotor. To achieve this, 4,000 lb of dry ice was applied. A box was constructed around the spokes to the outer rim and up to the hub. The ice was applied and the hub drew apart 0.12 inch. Precision 123 blocks were placed on the hub split line to get a precise measurement of the growth on the line. The shaft was able to be lowered to the basement after two hours of application of the dry ice.

The shaft was then sent to the shop for machining. The new design of the generator shaft involved machining the diameter down to remove the threads and put a groove in the shaft for a split ring. To complete this, a good measurement was needed to know where to place the groove. The threaded nut was put back on the shaft to the scribe mark to obtain a precise measurement with the laser tracker from the bottom of the nut to the coupling face. After the distance was measured, a 1-inch-thick disc was cut off the end of the shaft to test the tensile strength in three samples. The yield strength and allowable stress was obtained from an average of the three samples.

At this point, the final design modification could be completed. The threads on the shaft were 0.5-inch male threads and the shaft diameter was 12.5 inches. The final diameter the shaft was machined to was 12.49 inches on Unit 1. The total force on the thrust bearing was 246,558 lb using the hydraulic force and rotating equipment weight. Using the total force, a radial keyway depth of 1 inch, and a safety factor, the shaft stress was calculated to be under allowable limits. The split ring stress was also looked at, with the ring being 1.5 inches thick. The ring was designed so the actual stress was less than the allowable shear stress.

A new thrust bearing was supplied for each unit. The diameter of this bearing remained at 34 inches, but the split ring was added and a new block was required. The split ring was 1.5 inches thick and made of ASTM-516 grade 70 steel. The ring had 0.56-inch through holes drilled in eight spots for 0.5-inch bolts to retain the ring while being loaded with the weight of the unit. A new thrust block made of ASTM-A668 grade D steel forging was supplied with an interference fit to the generator shaft of 0.001 inch to 0.0015 inch. This fit allowed the block to be heated slightly for installation and prevents fretting of the generator shaft during operation.

A new runner for the thrust bearing was also supplied for the units because the running surface of the old one could not be repaired. The thrust shoes required re-babbitting after they went through the condition assessment. The inspection of the thrust disc and screws showed signs of cracks along with galling, so a new disc and screws were supplied.

The runner on the Lookout Shoals units had galling attributed to the plant settling and the wear ring material selection. The rotating wear rings were replaced with 304L stainless steel and the stationary rings were replaced with nickel aluminum bronze.

Completion and results

During re-assembly, the unit responded well to the modifications. The generator shaft was installed again using dry ice around the rotor spokes. The key to the installation was getting the shaft plumb and freezing it with dry ice as well. The thrust block was heated with heat lamps and slid on with no issues. The thrust shoes were loaded within +/-3% of the average. The shoes were loaded using the slug arc measurement. The arc length from loaded to 50-pound pull was measured and adjustments were made to get to the +/-3%. With the new thrust bearing design, the static orbit on Units 1 and 2 were well within the specification of 0.012 inch, over the distance the measurements were taken.

The original problem with the Unit 1 turbine shaft sleeve was actually the victim of other issues on the unit. The static runout of the thrust bearing was instigating increased wear on the bearings and triggered galling on the wear rings. The modification to the generator shaft greatly reduced the static runout and increased unit reliability. The other work on the unit improved dependability as well.

Unit 3 at Lookout Shoals is currently undergoing the same modifications performed on Units 1 and 2. The Unit 3 work is under the direction of Duke Energy personnel, with the use of outside resources.

Tyler Barrett is engineering technologist III and Russell Martin is senior engineering technologist with Duke Energy.

More HR Current Issue Articles
More HR Archives Issue Articles
Previous articleCanada Les Cedres Celebrates a Century of Innovation
Next articleModeling Potential Turbine Upgrades for the Brazeau Plant
Renewable Energy World's content team members help deliver the most comprehensive news coverage of the renewable energy industries. Based in the U.S., the UK, and South Africa, the team is comprised of editors from Clarion Energy's myriad of publications that cover the global energy industry.

No posts to display