article

How Mercury rehabilitated its turbine shafts at 104.4-MW Whakamaru Power Station

Save It Saved

When deciding to rehabilitate machinery at its Whakamaru plant, New Zealand utility Mercury evaluated several options and developed a method to machine existing shafts instead of replacing them.

By Glen Twining

The 104.4-MW Whakamaru Power Station on the Waikato River on the North Island of New Zealand is one of Mercury NZ Limited’s nine hydro stations. The project was at a stage in its life cycle where many mechanical components were in need of rehabilitation or replacement to prolong its operation and possibly increase its capacity. As part of the rehab process, Mercury evaluated and developed processes for machining turbine shafts, and this work will be discussed below.

About Whakamaru

Whakamaru houses four 26.1-MW vertical Francis turbines that were commissioned in 1956. These units were designed and built by Dominion Engineering, with 27.8 MVA generators provided by Metropolitan Vickers. Only one unit has had a major overhaul, with Unit No. 1 undergoing service in 2010.

Whakamaru’s generators began to suffer from end-of-life failures, mainly stator winding failures and rotor earth faults. Dissection of a failed winding revealed the extremely poor state of the strand insulation and confirmed that replacement was required to reduce breakdown risk. The turbine runners still had some life left, but they were generally well-worn and required at least an overhaul like Unit 1 had received.

At a minimum, the project’s scope would include generator replacement work and refurbishment of the existing turbine equipment. This led to the consideration of varying options, and an opportunity to reconfigure the flow and output from the power station.

Evaluating the options

With major work required, many options were available for consideration. This led to the analysis of two main options given the turbine equipment still had some remaining life: risk reduction or an uprate.

Option 1: Risk reduction

The risk reduction scope of work included generator replacement to reduce the plant breakdown risk, and disassembling and refurbishing the turbine and governor equipment to bring it back to near-new condition. This option didn’t include any performance uprate but instead focused on reducing plant breakdown risks and hence was assessed as a Net Present Value (NPV) negative project.

This option reduced the overall outage time required over the long-term and removed the need for a second round of disassembly and reassembly at the end of turbine life - plus the costs associated with that work. An earlier replacement of turbine equipment also brought forward the additional generation and revenue benefit for Mercury.

Additionally, but not always as obvious, by replacing turbine and governor equipment, a greater percentage of the original equipment could be disposed of rather than spending large amounts of time and money to refurbish it - thus removing or significantly reducing some large risks during the outage works.

Mercury’s experience in the previous decade refurbishing equipment back to as near-to-new condition as possible was often met with challenges, and in some cases complete replacement would have been only incrementally more expensive.

Option 2: Uprate

The uprate option took advantage of the units already being disassembled to complete the generator replacement and other refurbishment work and added in replacement of turbine and governor equipment to uprate unit performance.

Analysis projected an uprate that would add 28 GWh per annum of generation through about a 20-MW station capacity increase, plus an increase in efficiency. This uprate opportunity was NPV positive and would cost about US$900,000 per MW of additional capacity, but was sensitive to the price of electricity moving forward.

Demand for power had remained flat for several years in New Zealand and at the time project approval was being sought, there was uncertainty over possible reform to the electricity market. That risk did not materialize, and even with a flat electricity price, this option was still considered a valid investment and selected by Mercury.

Existing constraints and project benefits

By selecting the uprate option, there was some opportunity to reconfigure Whakamaru. The plant has a lower flow capacity than the stations upstream and downstream of it, meaning it is one of the least flexible stations for Mercury to operate and maintain. The units have a high utilization and typically run between their peak efficiency point and maximum continuous rating.

A generator shaft is lifted out of its pit at Mercury’s Whakamaru hydro plant after its removal from the rotor.
A generator shaft is lifted out of its pit at Mercury’s Whakamaru hydro plant after its removal from the rotor.

During the early feasibility and planning stages, a registration of interest (ROI) process was run where expected uprated performance figures were submitted by potential suppliers. Using in-house resources, the expected performance curves were run through Mercury’s Waikato River hydro system model to determine the best configuration for the Whakamaru units.

The selected option sees the turbine flow at peak efficiency increased by 14 m3/s, the peak turbine flow increased by 8 m3/s and turbine capacity increased by about 5 MW.

Planning and procurement

As the Whakamaru proposal was being evaluated and developed, Mercury was fortunate to have some major rehabilitation work already in progress, providing experience to help shape its delivery. Four generators were being replaced using a design, build and install model at the 204-MW Arapuni plant, while a turbine rehab project was also in-progress at 112-MW Ohakuri using a design, build and supervise model.

The addition of governor replacements at Whakamaru created a project larger in scale than the other two, causing it to be split into three manageable phases:

  • Phase 1: Feasibility and planning;
  • Phase 2: Design, testing, manufacture, delivery and site works planning; and
  • Phase 3: Installation and overhaul work.

Due to the knowledge and experience Mercury had and the need to evaluate all project scope options, Phase 1 was created specifically to work through this with an appropriate allocation of time. This also allowed planning of the remainder of the project to take place with the benefit of building on the lessons learned from recent projects.

This phase had dedicated funding for two years and included activities such as feasibility studies on the uprating of the units, an ROI process to select a shortlist of suppliers, development of the main equipment supply contracts and tendering.

With the generator plant breakdown risk being high, the replacement generators needed to meet all minimum technical and performance requirements set by Mercury and the relevant regulations in New Zealand. They would also need to be operated and maintained throughout their life cycle by Mercury and its maintenance contractor, rather than having to use the original equipment manufacturer.

To ensure that the project benefits were realized, the turbine that offered the best increase in annual generation was preferred. Meanwhile for governor equipment, all minimum technical and performance requirements set by Mercury and relevant regulators needed to be met.

In parallel with the evaluation of equipment supply tenders, a business case for the second and third phases of the project was being prepared with actual performance and cost data from suppliers. This enabled Mercury to present a business case with guaranteed performance data and certainty of price for the equipment supply. Full approval of the generator equipment supply and installation part of the project was requested during this stage.

To satisfy the uprate requirements, approval only up to the completion of turbine model testing was requested. This was done to include a final decision point prior to fully committing to the full uprate scope of the project. A termination window was written into the turbine equipment supply contract so that there was the option to cancel following completion of the turbine model testing should there be a failure to meet the contracted performance guarantees.

Phases 1 and 2 were run in series, with Phase 2 beginning with the signing of supply contracts for turbine and generator equipment with Alstom (now General Electric) and Andritz Hydro in August 2013. Phase 3 overlapped with Phase 2 as the new equipment arrive on-site and installation work began.

Summary of the project scope

The generator work included the complete replacement of the stators and rotor poles, relocation and rebuilding of the slip rings, and replacement of the rotor leads system. Interfacing a new through-shaft turbine air admission system into the machine resulted in a rotor lead system not seen before by the Mercury team. This presented additional challenges for implementing the refurbishment work on the generator to accept the new parts.

The turbine work included the replacement of the head cover, bottom ring, turbine runner, wicket gates and shaft seals and installation of the air admission system. All original bearings were retained.

Mercury’s Whakamaru plant has been in service since 1956 and is located on the Waikato River.
Mercury’s Whakamaru plant has been in service since 1956 and is located on the Waikato River.

The governor work included a completed replacement of the old Woodward mechanical governor and hydraulics with the installation of a modern high-pressure digital governor.

Various refurbishment and upgrade work was also completed on the remaining parts of the units, from water-to-wire. The top intake screen panels were strengthened, the gate lifting gear was refurbished, and the gates had seals replaced and other minor repairs performed. The embedded intake gate frames were cleaned and painted where required, along with patch painting on upper and lower sections of the penstock, scroll case and draft tube. Three new phase transformers with increased capacity to suit the uprated machines were also installed.

Machining a replacement generator shaft

Before developing a detailed machining procedure for the bore machining, the overall method for modifying the complete generator shaft had to be selected. The choices were to either complete the machining onsite with the shaft still in the rotor or to remove the shaft and complete the machining in a workshop. Mercury’s installation contractor assessed the two options and selected the latter as it was given the highest probability of achieving the design requirements. Next, the details removing the shaft and boring procedure were developed.

Removing the generator shaft

The rotor spider and generator shaft were originally assembled by placing the spider over the shaft and lowering it down to the correct position for bolting. This was possible during the original construction of Whakamaru’s machines as there was no rotor rim on the spider. Now, however, the shaft needed to be lowered down and out of the spider so as not to disturb the shrink fit of the rotor rim.

To achieve this, the installation contractor developed a methodology making use of the empty space once the turbine and generator was fully disassembled. The contractor built a frame to sit on the turbine’s base ring to lower the shaft onto after lowering it through the rotor spider.

The generator’s lower bracket and rotor were then placed back into the machine, the shaft lowered down and placed on the frame, and a dummy hub used to lift the rotor out of the machine. The generator shaft was then lifted out of the machine and transported to the subcontractor’s workshop for the machining work.

The rotor poles were removed before removing the generator shaft to make it easier to lift the rotor once the shaft was removed. Rotor poles were all replaced as part of the rehab work.

Machining the generator shaft bore

A suitable piece of round bar was used to replicate the shaft and perform the machining trial. A boring bar system was used in a 12-meter horizontal lathe and different cutting tools, guides and feed rates were tested to develop a methodology for the machining.

Following these trials, the round bar was cut at the section where the H9 tolerance was required. The result was within the design tolerance, confirming the machining setup would work.

This piece of risk management work demonstrated its value as the work on the actual shaft was completed in eight days - two fewer than scheduled - and many weeks quicker than if the trial had not been done.

Once the actual shaft was delivered, it went through a multi-step process:

  1. A conventional boring bar setup was used to bore a 130-mm-diameter hole deep enough to produce a new guide hole. This hole had to have enough space to accommodate the tooling and guide system for the rest of the boring process as the original hole through the shaft was not used.
  2. Using a customized boring bar and flathead drill from Showa Tool, the shaft was machined from the existing 75 mm diameter to 130 mm for the full 2.4 m length required.
  3. A second Showa drill and guide set at 140 mm diameter was then used to bore the finished bore. For the last 150 mm of the bore, the feed rate was reduced and then honed to achieve the required surface finish.
  4. The new shaft lead holes were drilled perpendicular to the new center bore.

Once the modified shaft was returned to the site, the reverse process was completed to put it back into the rotor, allowing the remaining work to be finished.

Project successes and lessons learned

The work performed at Whakamaru has provided many lessons and led to successes in a number of key areas.

Manufacturing locations and inspections

Mercury has had mixed experiences with manufacturing locations around the world, so significant effort was put into evaluating the options presented in order to protect the long-term value of the project. Locations were evaluated specifically for the equipment that was proposed to be manufactured at that site, considering the risks involved with each piece of equipment.

For example, a high level of importance was placed on the welding process used in manufacturing the turbine runners to provide confidence the new runners would last at least as long as the equipment they are replacing, and also considering the high cost to perform in situ repairs or replace the item earlier than expected.

The generator shaft is shown post-machining (left) and with its new shaft connection ring.
The generator shaft is shown post-machining (left) and with its new shaft connection ring.

Following a desktop analysis and reference checking process, the project team visited manufacturing locations in China and Europe - generally selecting the higher cost or premium manufacturing option. This experience again reinforced previous lessons learned that every manufacturing facility - including subcontractors - needs to be considered as there are no guarantees of good quality solely based on where the equipment is manufactured.

It is also worth noting that the capability, skill and experience of manufacturing locations can change rapidly, therefore historical knowledge cannot always be relied on.

Once a location is selected, an appropriate quality inspection program is required to check the key steps along the way, ensuring any identified risks are well-managed in order for the right level of product quality to be achieved and the project benefits protected.

Design review process

After the turbine and generator supply contracts were signed in August 2013, work on the design continued through mid-2014. A key lesson from Mercury’s previous projects and from other companies was to implement a robust review process.

So often in a project, the early stages of planning and procurement take longer than expected for various reasons. But, due to the desire to maintain overall completion dates, the additional time is usually cut from the design and manufacturing schedule. This forces the project into a reduced design duration and overlaps the design and manufacture stages of the procurement, significantly increasing the risk.

The Whakamaru project was planned with significantly more time for the design period, and due to the way the project approvals were set up, the manufacturing was placed in series with design. While it is common to use many different standards or the technical aspects of a project and even standards for its management, standards for running a design process appear to be less common. To ensure a process was in place - and to provide evidence of good practices to the project governance groups - the team chose IEC 61160 Design Review as a basis for its review process.

Allowing adequate time, using an international standard and being disciplined in adhering to the process were key items in ensuring the equipment would meet the requirements, and that equipment from different suppliers would interface together successfully during installation.

Risk management

Having a pragmatic and robust risk management plan in place was critical for the success of the project given the scale, complexity and amount of interfacing to existing equipment required. The most obvious risk at a high level was not meeting the project benefits, so a multi-staged approach was taken.

There were also some technical challenges identified in the site works where a failure would have had a large impact on the project’s time and cost, and significant business interruptions.

Uprated machine performance

Commissioning and performance testing of the unit and new equipment was completed in May 2017. The generator met its performance guarantees, and the turbine met its power output guarantee and exceeded its efficiency guarantee.


Glen Twining is a mechanical engineer for Mercury NZ Limited.

Suggestions For You: