Testing and Validation of Hydro Turbine Shaft Seals

Extending the performance envelope of turbine shaft seals in hydropower applications offers opportunities to reduce leakage and improve reliability. HRW-Hydro Review Worldwide chief editor David Appleyard visited James Walker’s laboratory in the north of England to see research and development work in action.

By David Appleyard

Lying in the heart of the UK’s Lake District, James Walker’s elastomer division is a specialist manufacturer of sealing solutions for static, rotary and reciprocating applications. Offering a bespoke seal design service for a range of industrial applications – including main shaft seals for hydro turbines and valves – the company has invested in the testing and analysis of sealing materials and seal design.

The 130 year-old company prides itself on its abiilty to formulate and compound its own materials, combining high-specification ingredients to create compounds with the ideal properties for specific sealing applications. This capability provides the company with control over both quality and material properties it believes are required when looking to develop high-performance, reliable sealing solutions.

Having produced a potentially suitable raw material, further testing and analysis is conducted in athe on-site, state-of-the-art materials laboratories. The current capability includes spectrometers for detailed chemical analysis, as well as tensometers for exploring a material’s mechanical properties, such as torsional and tensile resistance. Tensometer tests can also be carried out within an environmental chamber, allowing the effect of temperature on mechanical properties to be studied. Andrew Douglas, laboratory manager, says that the company has long invested in analysis and testing equipment for its elastomer division, which regularly undertakes design validation and materials analysis for customers and other external third parties.

The materials testing lab is located adjacent to the applications engineering department led by Ray Clifford, applications engineering manager. Clifford explains that the company, among other tools, uses finite element analysis (FEA) to validate new seal designs. In fact, it is the ability to produce new seal designs, develop appropriate materials and then test and validate those materials and designs, all within a single site, that rests at the core of the James Walker approach, the company says. As well as making an effort to understand customer applications to develop targeted solutions, the company believes its bespoke and high-quality strategy allows it to distinguish itself in the market.

Developing new sealing solutions

With a sophisticated laboratory available to assess the mechanical and physical properties of materials and a range of analytical tools to assess design and operational characteristics, it is the application of dedicated test rig analysis and field assessment that forms a crucial third arm of the design and manufacturing process at this facility.

For example, among the company’s latest seal designs is its HydroSele family of cartridge main shaft seals for Francis and Kaplan turbines.

Developed over a 15 year design, development and operational period and working under a wide variety of conditions, these cartridge seals have seen about 50 installations to date. HydroSele cartridges are custom-manufactured from modular components to suit each specific application, and the company offers full support, including installation and servicing if required.

Indeed, the company claims a HydroSele installation should have paid for itself in terms of maintenance costs, turbine downtime and improved efficiency often in less than two years. (James Walker actually has evidence of an exceptional cost savings of 800,000 euros in just 15 days at the Iren hydro station in Italy.) With a very low and controlled level of water leakage past sealing faces reducing the risk of flood damage to the plant, at four years, total investment in cash and downtime could be just one-quarter of that for an equivalent mechanical seal, James Walker says. It bases these claims on the fact that some prototype HydroSele installations are still operating 15 years after commissioning.

Given that the HydroSele design can be applied to shafts up to more than 1,000 mm and a maximum surface speed of 33 m/sec at greater than 1,000 Kpa pressure, the importance of a thorough testing program cannot be overstated. Each of the designs is assessed on a test rig at Cockermouth, where the company has explored the performance of more than 25 material combinations. It is these investigations that have enabled the company to improve the performance envelope of its products. Technical advances have resulted from the data derived from field and rig trials.

Testing sealing success

Early work to test the HydroSele design was conducted on small-diameter shafts, typically about 80 mm, and was primarily conducted to verify the sealing principle.

However, this rig did not allow sufficient knowledge to be gained regarding how the relationship between pressure and speed affected the performance of the seal due to low motor loads, meaning the tests were restricted to low pressures and speeds. Consequently, early applications for HydroSele cartridge seals were limited to those with a low pressure-velocity factor, typically turbines with a low system pressure and low to medium speed ranges.

Workers at James Walker's Cockermouth facility can manufacture shaft seals with a diameter of up to 2.3 meters. (Photo courtesy David Appleyard)
Workers at James Walker’s Cockermouth facility can manufacture shaft seals with a diameter of up to 2.3 meters. (Photo courtesy David Appleyard)

Subsequently, the testing progressed on to a dedicated rotary test rig, which confirmed the seal’s ability to perform up to the rated pressure of 145 psi (1,000 Kpa) and speed of 3,937 fpm (20 m/sec). In response to a growing requirement for seals in demanding run-of-river and pumped-storage applications with higher speeds and pressures, development work focused on determining the upper performance limits of the seals. In addition, development aimed to deliver a system capable of operating and surviving through periods of spinning in air, overspeed, axial and radial shaft movements and eccentricities, the company says.

In fact, according to James Walker, previous testing produced vital data but provided limited value in terms of pushing the performance envelope. There were also a number of advantages in investing in a purpose-built rig, notably larger shaft diameter capability, higher motor power and increased pump capabilities that are better able to replicate many turbine conditions and configurations.

Seal forms used in the presses are manufacturerd from aluminium at the company's Cockermouth site. (Photo courtesy David Appleyard)
Seal forms used in the presses are manufacturerd from aluminium at the company’s Cockermouth site. (Photo courtesy David Appleyard)

Commissioned in August 2009, the horizontal shaft configuration of the rig acts to replicate a horizontal turbine and was chosen to include the effects of gravity on the seal, as this is a much more arduous test than a vertical configuration, the company says.

With a 1,500 rpm, 326 nm motor, driving a shaft unit consisting of two removable 15.94 inch-diameter (405 mm) shaft sleeves, the rig was designed to be able to assess four seals per test within two cartridge assemblies. Containing two Hydrosele sealing elements either side of a central mounting block, which also acted to contain the system pressure, the design replicates the typical cartridge setup, with a pressure differential replicating the application. The rig also allowed for alterations to the test setup when necessary.

Three pumps were installed, allowing a flush pressure of 348 psi (2,400 Kpa) and a system pressure of 363 psi (2,500 Kpa) to be achieved. A cooling unit maintains a low water inlet temperature.

Tests are monitored throughout, with thermocouples recording head temperatures, pressure transducers monitoring the flush and system pressures, a torque meter, and a shaft speed output. This feeds information into a custom-designed data capture program, allowing the user to quickly analyze seal performance and act accordingly where required.

About one year after the rig was installed, several changes were implemented to further enhance its performance. These modifications included the addition of automatic globe valves to regulate both the flush and system pressures, based on the output from the pressure transducers, to ensure test conditions could be maintained while unmanned.

The data capture setup was also altered to allow the rig to shut down if unwanted changes in torque, temperature and water levels were recorded. This ensured tests were preserved in the event of rig malfunction, a factor that is of particular reassurance when conducting long-term tests.

Development work evolved into three separate test rig phases:

Phase 1: Material validation and benchmarking

The first phase centered on gaining a full insight into the capabilities of the existing seal. It involved analyzing key performance indicators over the established pressure/velocity range, allowing its specific pressure/velocity capabilities to be understood and clearly defined in terms of the pressure-speed limitations of the seal and how the two relate to one another in terms of overall performance. This allowed areas for improvement to be highlighted and targets for new material set, leading into a full seal development program.

Initially, a series of static trials aimed at providing a platform for dynamic trials were conducted, to ensure best “fits” were achieved when dynamic trials were conducted. These involved fitting seals sized with increasing clearances and testing under increasing pressures. The tests were conducted without rotating the shaft, allowing an analysis of how static shaft clearances and fluid film thicknesses relate to leakage rates.

Using information gathered from the static trials, about 35 dynamic trials were conducted under various shaft speeds and pressures. Observations made throughout the dynamic testing also flagged the importance of allowing the seals to “bed in” despite initial “higher” leakage in order to achieve optimum performance. This knowledge provided confidence to fit the seals with a larger initial shaft clearance and therefore larger fluid film, which, once in a balanced energized state, still produced a low level of leakage.

Trends witnessed in high pressure/velocity factor tests showed the sleeve material to become more sensitive to movement as a result of the increased temperature generation under the seal running face combined with the increased pressure behind the seal. As a result of this, shaft contact became inevitable as the seal lost its fluid film, resulting in the seal not operating as desired. However, notes were made that if this effect was eliminated completely, the seal failed to energize and ran simply to throttle the shaft.

Phase 2: Material development

Focusing on testing and developing new materials, this phase allowed specific “trigger points” to be flagged and the performance and condition of the seal to be mapped. Screening trials were conducted across a range of about 20 material variants, with each seal tailored to possess particular characteristics. Testing consisted of many comparable trials to gauge the potential of each material variation, selecting the most likely candidates for further testing with the aim of achieving a greater performance envelope.

Once the new material choice was established, observations made throughout the screening process led to a “fine tuned” seal that subsequently underwent benchmarking trials and proved much more adept under demanding conditions. The seal was seen to be much less sensitive to pressure and temperature than the previous derivative and offered a harder wearing running face, which afforded the seal a greater safety factor during times of dry running or extensive loads, such as start ups and shut downs.

More than 3,100 hours of benchmark testing have been conducted on the improved design, under pressures ranging from 58 to 232 psi (400 to 1,600 Kpa), and speeds of 1,575 fpm to 5,905 fpm (8 to 30 m/sec).

Valuable information was gathered on the fundamental qualities required by the seal to allow the energization mechanism to take place and to perform during all running phases. The stiffness of the sleeve material and optimum seal design were key to both parts of the process, and work was done to enhance and control the behavior and performance range associated with existing materials by using composites.

Phase 3: Customer validation

The third phase applied knowledge gained from the development to specific customer trials. This allowed testing to focus on the finalized design and validate it against performance criteria set by end users.

This was essential, as many tests centered on validation under conditions not previously trialed. This allowed captured data to be produced ahead of installation, as testing under pressures and speeds – as well as conditions such as shaft offset and overspeed cycles – provides a close representation of what customers can expect. For example, in one case, the requirement for customer validation test work arose as a result of unexpected performance during commissioning.

Workers hand finish seals produced at the James Walker elastomer division in the north of the UK. (Photo courtesy David Appleyard)
Workers hand finish seals produced at the James Walker elastomer division in the north of the UK. (Photo courtesy David Appleyard)

The commission was on a vertical Francis turbine with a shaft diameter greater than 27.5 inches (700 mm) running under conditions of 58 psi, 3,346 fpm (400 Kpa, 17 m/sec). After issues with the initial installation, investigations showed that a shaft offset compromised the seal performance.

Initial tests showed that the original material supplied was unsuitable, as a result of the shaft offset combined with high speed. Tests conducted included a 1,000 hour application test that closely replicated the conditions faced on site, such as stop-start intervals and overspeed running phases at 5,315 fpm (27 m/sec). In addition, there were two tests over a total of 480 hours under stop-start conditions, in which the seals were subjected to several 30 minute static periods per day, as well as three tests over a total of 215 hours under shaft offset conditions, during which the seals were tested under three different shaft offset positions, ranging from 0.02 inch to 0.057 inch (0.5 to 1.45 mm).

More than 1,730 hours worth of tests were conducted specifically for these applications, including a 50 hour test conducted on a 6.37 inch (160 mm) diameter test rig, closely replicating the shaft size and conditions of another of the customer’s sites.

Advancing sealing solutions

Advances in turbine design and components have led to more turbines with a high pressure/velocity factor and also increased the requirement to cope with demands from the energy industry, such as running at variable speeds and with intermittent operating cycles.

The relationship between leakage, flow, torque and pressure all contribute to the life span of the seal, and this can be mapped and predicted to provide valuable data for conditioning monitoring systems. Expected leakage rates and power consumption in particular are of value.

Meanwhile, benchmarking and validation testing of new design derivatives is ongoing, and the company has been working with clients to invest in and develop additional test equipment. For example, an investment decision has already been made by James Walker to develop a larger test rig for its rotary seals that will run a larger 1,000 mm shaft diameter at higher speeds in a bid to further extend the applicable range of pressure/velocity values. Work will continue to focus on gaining high levels of understanding of the critical success factors of the tribological conditions and on developing materials to create these successful conditions repeatably and reliably.

David Appleyard is chief editor of HRW-Hydro Review Worldwide.

We welcome your input!

The editors of HRW-Hydro Review Worldwide welcome your suggestions for articles and departments that you feel would be helpful to you and your colleagues.

Please send your ideas to David Appleyard, Chief Editor, HRW-Hydro Review Worldwide, The Water Tower, Gunpowder Mill, Powdermill Lane, Waltham Abbey, Essex EN9 1BN United Kingdom; (44) 1992-656659; E-mail: davida@pennwell.com.


Previous articleCanada’s 824-MW Muskrat Falls jumps in expected cost
Next articleHRW Volume 20 Issue 6

No posts to display