Turbine Pump – Case Study
Berg Engineering was requested by the client to carry out mechanical works on site at a power station.
This condition assessment details findings and recommendation for the turbine runner and seal rings.
A subsequent condition assessment will be issued for the remainder to the works.
The equipment assessed is a 1.4MW hydro electric Francis turbine
Manufacturer: GE Energy
Details of Assessment
The following inspections were carried out to assess the conditions of the runner and upper and lower seal rings.
Visible cavitation damage on the turbine runner was clearly visible on the leading edge of the high pressure side as well as less extensive cavitation damage to the low pressure side. These damaged areas range from 50mm – 70mm in length, with the damage extending completely through the blades on 5 of the 13 blades.
The upper and lower seal rings showed signs of corrosion and pitting, the pitting was visibly worse on the upper seal ring.
Positive Material Identification
A PMI test was carried out on the turbine runner and we have determined the material to be identified as EN 1.4418 martensitic/austenitic quench and tempered stainless steel.
Hardness checks were done on the runner and values of 315 – 326 HB recorded. This plate is typically used in an annealed or tempered state. The measured hardness indicate the runner material is in the tempered state.
Ultrasonic testing was carried out on the turbine runner to check for any defects in the material. No additional defects were found other than those visible.
Component Dimensional Check
Dimensional checks were carried out on critical dimensions of the turbine runner and upper and lower and seal rings. The records show that all the critical dimensions are in tolerance on the runner except for the shaft seal running surface which measures 0.053mm undersize.
Detailed measurements were also carried out on the running diameter for the upper and lower seal rings.
Image (A) is a view on the low pressure side of the blade, with cavitation damage.
This image is representative of the low pressure damage found on 50% of the blades, the remaining blades have mild damage by comparison.
Lower seal ring
Upper seal ring
Pitting on the seal rings is quite evident, but could be cleaned up with a 0.2mm/0.25mm skim on radius.
If the surfaces were to be machined clear of pitting, the clearances would at roughly 140% of allowable clearance.
Conclusions & Recommendations
Berg has identified an Avesta product 248SV which is the ideal weld consumable for this parent metal. This would be the easiest way of repairing the runner, the Avesta product is tailor made and there is no guess work required on weld parameters and post weld heat treatment. To stabilise the weld metal and minimise the martensite in the HAZ the runner would be annealed at 590 degrees C for 4 hrs, followed by air cooling.
As annealed the hardness would be around 260 HB. The material is then tempered QT900 (590°C – 620°C) for 8 hours followed by oil cooling. This process would involve significant heat input via the overlay weld and the heat treatment, which may lead to distortion. The distortion can be corrected by skimming the OD of the runner, which would necessitate the replacement of the bronze seal rings.
In liaising with welding specialists it has emerged that 2209 duplex wire is widely used for turbine runner repairs. Admittedly these runners are fabricated from CA6NM which is a martensitic material. CA6NM being a martensitic material is ideal for high cavitation resistance, and is consequently widely used in Francis hydro turbine runners. This combination is successful and does not require post weld heat treatment.
Presumably if the damage is more than 3-4mm depth then overlay will be applied in 2 or 3 runs, in which case the surface material is largely undiluted and would have a duplex stainless metallurgy.
Right is a simulation of the parent metal overlayed with 2209 wire on 1.4418 parent metal. The resultant metallurgy is denoted by (R), which is a duplex metallurgy, with a likely hardness of 240 HB.
This is quite a viable option, but the hardness is on the low side and the plate will be less tolerant of cavitation attack going forward.
Given that 5 of the 13 blades have through damage, and most blades also have damage on the low mpressure side of the blade, most blades will require a buttering overlay repair on both sides of the blade.
Since the damage is on the same plate section HP and LP side, there is good motivation for a window plate repair. The damage section can be machined out, with a weld prep cut on the HP side. Plate sections can be dropped in and welded from the HP side. Each repair could then be dressed by hand to achieve a smooth plate profile. This would yield a quality repair with a minimum of work on the inaccessible LP side. The welds can be radiographed for QA purposes. The heat input would also be very much lower for such a repair.
With a window repair on the table it is viable that we opt for a plate with higher hardness like CA6NM. This can then be welded in with the 2209 wire. With a low volume weld prep, the dilution can be maximised to improve the weld metal hardness.
Such a repair with martensitic insert (IN), will yield a weld (WM) of very similar properties (hardness) to the parent metal (PM). Heat treatment would be unlikely.
The recommendation are as follows:
• Purchase 1.4418 plate and overlay with Avesta 248SV to determine hardness achievable without post weld heat treatment, and with post weld heat treatment
• Purchase 1.4418 plate and overlay with 2209 wire to determine hardness achievable without post weld heat treatment
• Purchase 1.4418 plate and CA6NM plate and join with 2209 wire. Determine the hardness achievable without post weld heat treatment
• Purchase materials for seal ring replacement. Final sizes to be determined once the runner repair dimensions are established