Read Conversion of a Propeller Turbine text version

Conversion of a Propeller Turbine to Full Kaplan Operation at Michigamme Falls

By Gerard J. Russell, P.E., American Hydro Corp., York, Pennsylvania, USA, Craig Peterson, P.E., American Hydro Corp., York, Pennsylvania, USA, and Douglas T. Eberlein, P.E., We Energies, Milwaukee, Wisconsin, USA ABSTRACT FERC license renewal stipulations for We Energies' Michigamme Falls plant included new flow constraints that were outside the existing fixed-blade propeller turbines' efficiency range. One of the turbines was converted to fully adjustable Kaplan operation to regain the lost generation. Introduction Renewal of the project's FERC license in October of 2001 required that the minimum flow could be no less than 50% of the maximum flow during a given calendar day. The existing generating units were typical fixed-blade propeller turbines with a very narrow range of efficient operation, so they could not effectively meet this new requirement. We Energies evaluated several options for addressing the new operating regime including spilling the required low flow when necessary, installation of a minimum flow turbine-generator unit, and conversion of one of the propeller units to full adjustable blade Kaplan operation. The Kaplan conversion was chosen as the most effective option to improve the operating efficiency and flexibility of the plant while achieving the required low flow operating capability. This paper describes the Owner's planning process that resulted in the decision to convert the unit from fixed to adjustable blade operation. The conversion not only replaced the turbine runner but required the integration of a blade control system to work with the existing governor. In addition, the mechanical and electrical capabilities of the existing generator were evaluated to address the increased turbine output and higher overspeeds. The hydraulic and mechanical design aspects of the conversion are detailed and compared to actual performance. Guidelines for evaluating the feasibility of performing this type of conversion for similar units are also offered. Background and History The Michigamme Falls Hydroelectric Plant is located on the Michigamme River 10 miles northwest of Iron Mountain, Michigan just upstream of the confluence of the Brule and Michigamme rivers. The project develops 65 feet of head and originally featured two vertical propeller turbines rated at 6,500 horsepower driving 4800 kW synchronous generators. The generating station is owned and operated by We Energies and was first placed into commercial operation in 1953. 1

HydroVision 2004 - Copyright HCI Publications, 2004 -

The original operating regime for the plant was peaking mode. At times of the year when river flows were not adequate for full power operation around the clock, the units were shut down at night and operated only during the day for peaking power production. The Owner began a negotiated process with the affected state and federal resource agencies for renewing the project's license in 1998. This process concluded in 2001 with license renewal, which included new flow constraints to enhance the aquatic environment downstream of the plant. The minimum flow downstream of the confluence of the Michigamme and Brule rivers could be no less than 50% of the maximum flow during a given calendar day. While the confluence is downstream of the plant, the contribution of the Brule River to the total flow is not sufficient to allow the existing turbine generators to effectively meet the new requirement at Michigamme Falls. This presented several problems. First, operating flexibility to dispatch the units when needed for power production was severely limited. Secondly, given the narrow range of efficient operation for these turbines, the owner's analysis of the existing units operating under the new flow requirements indicated that at least 7400 megawatt-hours of electrical generation would be lost annually. Also, cavitation was a concern while operating the units outside of their optimum range. We Energies' staff brainstormed possible solutions to these problems with assistance from Black & Veatch Engineering of Kansas City, Missouri. Running the existing units within the new flow constraints was undesirable for the reasons stated above. Spilling the minimum flow through a tainter gate was considered but rejected due to winter icing conditions and the associated lost generation with spilling. The addition of a small turbine-generator set to accommodate peaking or low-flow operation was also considered. However, space limitations and the cost for necessary civil modifications to the plant made this option unfeasible. Converting one of the two existing turbines to fully adjustable Kaplan operation was selected as the best option. While not the lowestcost option, it would regain the lost generation and preserve operating flexibility for the owner. Several challenges still remained, however. The original unit was set fairly deep in relation to tailwater level and the configuration of the draft tube had always been in question. Index testing performed after original unit commissioning found turbine performance less than expected. The draft tube cross section was reduced and reshaped to improve performance, with mixed results. The existing turbines had experienced moderate to severe cavitation in the past; the draft tubes were suspect. However, given a fixed project budget and difficult access to the draft tube, additional modifications were out of the question. The new turbine would have to be designed to perform satisfactorily with the existing draft tube configuration. Generator overspeed during a runaway turbine event (loss of governor control) was another concern. The overspeed of the proposed Kaplan turbine was significantly higher than the existing propeller turbine. While a typical propeller turbine would have a maximum runaway speed of 1.8 to 2.2 times the normal synchronous speed, a Kaplan 2

turbine in this application could experience a maximum runaway speed of 2.5 times the normal speed. Would the existing generator be able to withstand the higher speed, or would modification or even replacement of the generator be necessary? An analysis of the existing generator rotor was performed and, fortunately, showed that the robust construction could satisfactorily accommodate the increased stress resulting from the higher overspeed without modification. Meeting the project schedule was also a concern. The license renewal was received from FERC in January, 2001 and the new minimum flow requirements became effective in November, 2001. Preliminary engineering had been initiated by We Energies in 2000 to allow the solicitation of proposals for the new turbine, but contract award was delayed until June, 2001. Given the abbreviated schedule, the owner sought and received from FERC and the resource agencies a waiver from the minimum flow requirements until June, 2002. With license relief, the new turbine would still be needed to capture generation from high spring runoff flows, so the completion deadline became April, 2002. The Solution American Hydro's turbine engineers faced several design challenges when converting the Michigamme Falls Unit 2 turbine to full Kaplan operation. The existing turbine and generator were designed originally to operate as a fixed blade propeller, and therefore, lacked the design features required for blade angle control. Conversion to full Kaplan operation would require the design of an adjustable blade runner to fit the existing turbine components with a minimum of modifications. The runner diameter, operating speed and setting with respect to tailwater were essentially fixed. The requirement for a very broad range of high efficiency operation made it necessary to design the runner with a large range of blade angle adjustment. Also, to assure that the maximum power of the turbine was not compromised, it was necessary to keep the hub size to a minimum. Given the relatively small size of the hub, the large range of blade adjustment, and the need to have the blade servomotor in the hub, as discussed later, the design of the hub internal components had to be optimized to assure that no components would be overstressed and that the assembled runner would operate as desired. The existing cylindrical throat ring was not compatible with an adjustable blade Kaplan runner, especially with the extremely large range of blade rotation, and therefore a new semi-spherical throat liner was designed to retrofit into the unit. Hydraulic Design and Analysis The selection of the number of blades and the design of the blade shape were critical to assure that the upgraded turbine would operate throughout the desired operating range at high efficiency levels and without damaging cavitation. The design of new blades to fit within an existing wheelcase and operate at predetermined operating conditions is quite different from selecting a standard blade and runner design for operation at a given site where the designer has the freedom to select the optimum runner diameter, operating speed and setting. The blade shape for Michigamme Falls was carefully custom designed and analyzed using American Hydro's runner design system (AHRDS) 3

Figure 1

to suit the existing generator speed of 276.92 RPM and plant net head of 65 feet. Figure 1 shows the AHRDS blade shape and calculated pressure pattern on the low-pressure side of the blade at the most severe operating condition. The blade design was tailored to optimize efficiency across a wide operating range while limiting cavitation to acceptable levels.

The hydraulic profile of the draft tube is as important as the blade shape when determining the overall performance of a low head turbine. As mentioned previously, the draft tubes at Michigamme Falls had long been a source of concern and were suspected of causing certain poor performance characteristics. To confidently establish expected performance levels with the new adjustable blade runner, it was essential to understand the loss characteristics of the existing draft tubes. American Hydro performed an extensive hydraulic analysis of the Michigamme Falls draft tube using a fully threedimensional viscous computational fluid dynamics (CFD) computer program. Figure 2 illustrates the CFD model and calculated flow velocity profiles throughout the draft tube. While some areas of flow separation and reverse flow can be observed, the hydraulic loss calculations indicated that the draft tube, with the new runner, would operate at good efficiency levels throughout the desired operating range. Figure 2 Blade Operating System The main mechanical design hurdle was centered around actuation of the new adjustable blades. Kaplan turbines typically utilize an oil operated servomotor located either in the runner hub, or in an enlarged section of the turbine shaft, to adjust the runner blade angle to suit the operating conditions. This configuration requires the use of an oil piping system through the center of the turbine and generator shafts to deliver high pressure oil to the blade servomotor. The oil piping must also provide a static head of oil to lubricate the blade trunnion bushings and blade operating mechanism components within the hub, as well as providing blade position feedback. An oil head is 4

required to transfer pressurized oil into the rotating oil pipes from the stationary oil lines originating at the hydraulic power unit. As the existing unit was originally designed as a fixed blade propeller, the turbine and generator shafting system did not have the features typically associated with Kaplan turbine oil pipes, oil head, and blade position feedback. The turbine shaft was a solid forging without a center bored hole. The generator shaft had a small central hole in the upper end used for exciter leads only. The generator was also equipped with a brush type exciter and pilot exciter. The possibilities of extending oil piping through this equipment, plus the addition of an oil head on top of the entire assembly would not fit under the powerhouse roof. The original design concept called for replacement of the turbine shaft with a new two piece shaft set that included an integral blade servomotor. This concept would obviate the need to replace the existing generator shaft. The blade mechanism in the hub would be connected to the servomotor via an operating rod extending through the turbine main shaft. Hydraulic oil would be delivered to the hub and servomotor through an oil transfer / blade position feedback assembly surrounding the intermediate shaft between the turbine and generator shafts. The complexity, and questionable reliability of the "over the shaft" oil transfer box, soon ruled out this concept. The remaining alternatives were to: 1) Replace both the turbine and generator shafts and incorporate the blade servomotor into an enlarged upper end of the turbine shaft, or 2) Incorporate the blade servomotor into the runner hub and attempt to reuse the existing turbine shaft and generator shaft if possible. Preliminary runner design indicated that by using a 1000 psi oil pressure system, the blade servomotor could be incorporated into the small hub. The second of the listed alternatives was therefore selected as most desirable and cost effective. Typically, three pipes are utilized to direct the oil to the opening and closing sides of the blade servomotor as well as supplying pressurized oil to the runner hub cavity. Since the diameter of the turbine shaft was rather small, it was determined that the only way to reuse this shaft was to bore through the length of the shaft, and to use this bore as one of the oil pipes. Increasing the diameter of the shaft would have entailed major redesign of the turbine bearing, packing box and other related equipment in close proximity to the shaft. Analysis of the shaft indicated that even with a 4.375" diameter axial bore, stress levels under all operating conditions fell within normally accepted limits. The generator shaft, however, was another story. The diameters at the upper end of the existing generator shaft were extremely small. The diameter at the root of the circumferential groove for the thrust block collar was not much larger than that required for the oil passage bore. It was decided, therefore, to replace the generator shaft. The bore of the thrust block was increased to accommodate a larger shaft. Redesign of the lock collar, along with the larger diameter at the thrust block permitted the addition of the 4.375" diameter oil pipe bore, plus two axial holes, parallel to the center bore, for the exciter leads. The shaft length was also increased to accept the new brushless exciter, and to connect to the outer Kaplan pipe required at the oil head. 5

A new "three pipe" conventional oil head was designed and fabricated. The inner oil pipe serves as the mechanical feedback from the blade servomotor in the runner and also provides the static oil pressure supply to the hub. The intermediate oil pipe transmits the "closing" oil pressure from the oil head to the blade servomotor. The outer pipe, (the bore of the turbine and generator shafts) transmits the "opening" oil pressure to the blade servomotor. The oil head base was designed to mount directly to the new static exciter housing, appropriately insulated for electrical isolation. The existing Woodward PMG mounts directly to the top of the new oil head. A new splined PMG drive shaft was designed to accommodate the axial motion of the inner pipe. Figure 3 illustrates the oil head design. A dedicated hydraulic power unit with a PLC based control system was specified for control of the new adjustable blade runner. The system was purchased from The Hope Group / Sorensen Governor. The blade control system operates as a "slave" to the existing governor that controls the turbine operation and position of the wicket gates. The optimum runner blade angle is set by the blade control system based on the position of the wicket gates for any point of operation. The Kaplan Runner The adjustable blade runner presented its own design challenges. Even though the configuration is more or less typical of vertical Kaplan runners, the small physical dimensions of the hub, combined with the large angular rotation of the five blades, Figure 3 the servomotor located in the hub and the relatively long servo stroke severely taxed the available space. As runner dimensions become smaller, the need for assembly space for a worker's hands and tools does not decrease and had to be kept in mind during the design phase. The blade tip diameter was increased to 84", 2" larger than the original propeller runner. The hub ratio is 42% for a spherical diameter of 35.25". The cylindrical portions of the hub immediately above and below the "sphere" are only 30" in diameter, smaller than the original propeller hub. The centerline of the runner was also lowered 5" to improve 6

the cavitation performance. All of these factors resulted in a rather long runner assembly, considerably longer than the fixed blade runner. The increased length contributed to field assembly challenges as the clearance under the outdoor gantry crane was not adequate for the increased length of runner and shaft. Refer to Figure 3 Typically, American Hydro prefers plate steel fabrications over steel castings, however, the design configuration of the hub and the servomotor/crosshead did not readily lend themselves to fabrications. These two components were made as castings of ASTM A216 grade WCC. All other components within the hub with the exception of the blades were either forgings, or steel or alloy steel plate. The extremely tight space within the hub necessitated the use of higher strength carbon or alloy steels to minimize the physical dimensions of the parts. Self aligning spherical bearings were selected for use in the link ends, both for their high load carrying capacity and to eliminate any bending loads in the mechanism. The stainless steel blades are ASTM A744 CA6NM castings. The blade leaf was cast to near net shape, only requiring grinding to attain the required finish, fairness and edge shape. The trunnion and disc areas were machined on a five-axis CNC horizontal boring machine. The Kaplan runner was completely assembled in the American Hydro facility, pressure and stroke tested using the blade servomotor to verify oil tightness and correct operation of the blade mechanism. The runner was shipped to the field completely assembled with the exception of the inner oil pipe. Turbine Rehabilitation During the early 1990's, both propeller turbines at Michigamme Falls were overhauled and new runners installed, therefore it was anticipated that minimal refurbishment would be required. The main shaft bearing parts, the inner and outer head covers, the wicket gates and mechanism parts, the gate operating ring and the turbine main shaft were shipped to the American Hydro facility for inspection and evaluation. All bearing surfaces on the gates, outer head cover and gate operating ring were re-machined removing minimal material to establish new uniform sizes and finishes. Extensive corrosion on the lower wicket gate stems necessitated the removal of considerable material so the diameters were restored by the addition of stainless steel sleeves. Minor rework of the adjustable water bearing shoes and installation of new rubber staves was performed. New bronze gate stem bushings, thrust washers and new operating ring pads were installed. The turbine main shaft was dimensionally inspected, the center was bored through the entire length to 4.375" diameter, and the runner end face and spigot were re-machined to restore surface finish and flatness tolerances. The existing embedded throat ring was deteriorated due to cavitation and other wear and tear. To assure close running clearances with the blades at all operating blade positions, a new semi-spherical stainless steel throat liner was fabricated and installed as illustrated by Figure 4. Figure 5 illustrates the final configuration of the Michigamme Falls turbine following conversion to full Kaplan operation. 7

Figure 4

Figure 5

Results The design and manufacture of the Kaplan runner and other new components began in June 2001 and field disassembly commenced in August 2001. Given the need to have the unit running in time for the spring runoff in 2002 it was necessary to have the design, manufacture and field work proceed in parallel. Critical dimensions for some of the new components could not be determined prior to disassembly due to lack of detailed generator component drawings. This resulted in several new components including the new runner, new generator shaft and new oil head, appearing as critical path items on the project schedule. Good communication between American Hydro's designers, field service personnel and We Energies was necessary to keep the project moving toward a timely and successful completion. Following disassembly of the turbine and generator, the turbine components were shipped to American Hydro's shop for modification and refurbishment as required. Figure 6 is a photograph of the old propeller runner during removal from the unit. Figure 7 shows the new Kaplan runner following pressure testing on the shop floor. In parallel with the shop work, the new stainless steel throat liner was installed in the field as illustrated by Figure 8. The new and refurbished turbine components were delivered to site and installation began in mid-January 2002. Figure 9 shows the new runner being lowered into position. The unit reassembly and commissioning were completed in time for the spring runoff in mid-April 2002.


Figure 6

Figure 8

Figure 7

Figure 9

An index test was performed in June 2002 to verify performance guarantees and refine the shape of the blade-gate cam curve. Figure 10 shows the results of the index test plotted as relative efficiency versus turbine output. Also shown on this graph are the expected performance curve and the results of an

Michigamme Falls Kaplan Conversion

100 95 90 85 80 75 70 65 60 0 Relative Efficiency (%)

Index Test After Conversion Expected Performance Curve

Index Test Before Conversion







Turbine Ouptut (KW)

Figure 10


index test performed prior to the conversion. Figure 11 shows the pre and post conversion index tests plotted as relative efficiency versus turbine discharge. As these curves clearly indicate, the project goal of improving operating efficiency and flexibility while meeting the FERC mandated low flow operating capability was achieved. Conclusion

100 90 80 70 60 50 40 30 20 0

Michigamme Falls Kaplan Conversion

Index Test After Conversion Index Test Before Conversion

Relative Efficiency (%)







Turbine Discharge (cfs)

Figure 11

For low head hydroelectric sites with propeller turbines, conversion of an existing propeller unit to full Kaplan operation may be the best alternative when considering options for meeting low flow requirements. As the Michigamme Falls project demonstrates, with careful planning by the owner and due consideration by an experienced turbine designer, conversion from propeller to full Kaplan operation is technically feasible and may be economically justified. The following checklist includes some important items to address when evaluating or implementing a Kaplan conversion: Desired Operating Range ­ is capacity or efficiency more important? Site Hydraulic Conditions Generator Overspeed Capability Thrust Bearing Capacity Existing Governor Capacity Turbine and Generator Shafting Features Detailed Drawings of Existing Components ­ are they available? Powerhouse Headroom Above the Generator Available Budget (Ecomomic Justification) 10

Authors Gerard J. Russell, P.E., is Marketing Manager at American Hydro Corporation. He is responsible for sales and marketing of American Hydro's products and services in North America as well as hydraulic engineering for many of their projects including Michigamme Falls. Craig Peterson, P.E., is the Chief Design Engineer at American Hydro Corporation. He is responsible for mechanical design and analysis for American Hydro and performed the detailed engineering for the Michigamme Falls project. Douglas T. Eberlein, P.E., is a Project Manager at We Energies. He is responsible for various improvement projects at the public utility, including the Kaplan Conversion at Michigamme Falls.



Conversion of a Propeller Turbine

11 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

ERCOT's Dynamic Model of Wind Turbine Generators: Preprint
The History and State of the Art of Variable-Speed Wind Turbine Technology