Input from five plants suggests that leading- and trailing-edge erosion of L-0 blades is driven by operating hours and exhaust temperature versus starts. More specifically: The fewer the hours and the higher the temperature the less erosion. Inspections indicate that Stellite shielding noticeably reduces the erosion rate, moisture-removal troughs in the last-stage diaphragm do as well.
Other findings are the following:
- Erosion rate for combined-cycle turbines is higher than for steamers at coal-fired plants, in most cases. Plus, trailing-edge erosion is more prevalent in the combined-cycle fleet.
- Erosion proceeds quickly after COD, then slows. Thus, if you blend out surface irregularities, the airfoils actually erode faster than they would have had you left them alone.
- The highest level of erosion typically occurs from the top of the blade to 3 in. below it.
- Blade damage is related to the moisture content of the steam more so than water chemistry.
Experience with 40-in. titanium L-0 blades—Case History One
Owner/operators of Toshiba steam turbine/generators received an alert in fall 2018 regarding the failure of 40-in. titanium L-0 blades at a Japanese powerplant. The OEM recommended inspections every 24k operating hours or 600 starts.
Personnel from a US combined-cycle plant with two steam turbine/generators affected by the notification shared with STUG2020 attendees their inspection and repair experiences from spring 2018 through spring 2020. The two steam turbines (called ST10 and ST20) typically were starting 125 to 225 times annually prior to the alert.
Readers are reminded that the PowerPoint for this presentation is posted on the Power Users website and accessible only by approved owner/operators. With two steamers and two OEMs involved (GE, the gas-turbine supplier, provided some ST outage support), a review of the author’s slides may facilitate your understanding of this ongoing project.
A spring 2018 L-0 inspection done before the alert was issued revealed no crack indications. However, some trailing-edge wear was in evidence near the dovetails, as was some so-called “wormholing.”
A follow-up inspection of ST10 was conducted the following spring when the unit was out of service to correct diaphragm dishing. Results were dramatically different than those recorded a year earlier. This time, every L-0 blade had cracks emanating from wormholes—from two to three indications per blade to between 20 and 30. Plus, one airfoil had a 0.75-in.-long through- crack.
The titanium blades on the generator end of the machine were removed and were replaced with a pressure plate because a replacement row of blades was not immediately available. The trailing edges of the L-0 blades at the opposite end of the cylinder were checked by fluorescent penetration inspection (FPI), blended, and reinstalled. The emergency pressure-plate repair was effective with no vibration or thrust issues after restart. However, the loss of output was significant at about 17 MW.
Another advisory was released by Toshiba in mid-June 2019. It reported that low-load operation of the turbine increased the rate of trailing-edge erosion because of flow reversal. It also indicated the lower portion of the trailing edge, extending from the platform to about one- quarter of the way up the blade, was most susceptible to damage. Cracks initiated by fretting of the midspan snubber sleeve were identified as well. The fix there would be upgraded material.
The pressure plate was removed from ST10 in fall 2019, replaced with non-OEM refurbished L-0 blades except for one new GE L-0 blade. The unit achieved full capacity on its return to service following a low-speed balance.
Both turbines passed their trailing-edge FPI inspections in spring 2020, leading personnel to assume it would be smooth sailing until the major. But that did not happen as stress corrosion cracking of dovetails was found in ST20. Temporary pressure plates replaced ST20’s L-0 rows and the unit was returned to service with a reduction in output of about 40 MW.
ST10 was restarted with new L-1 blades, upgraded materials, and new locking hardware. Attend STUG2021 in St. Louis to get an update on this project and others as well.
Experience with 40-in. titanium L-0 blades—Case History Two
Another combined cycle equipped with a Toshiba steam turbine then shared its experience with 40-in. titanium last-stage blades in a double-flow low-pressure section. This plant, which began commercial operation in 2005, was designed as a 2 × 1 facility but constructed as a 1 × 1, reducing the output of the 330-MW steamer to about 100 MW.
The owner/operator making the presentation bought this combined cycle in 2016 with 20k to 25k service hours and lots of time on turning gear. There was no record of the plant having had an inspection to that point. The new owner inspected the ST’s last-stage blades in-situ during a valve outage shortly after purchase.
A major inspection was conducted in fall 2017. Plan was to visually check the last-stage blades, but during the outage the OEM recommended adding an eddy-current inspection. This was the owner’s first and last experience (to date) with 40-in. titanium blades. It sold the facility in 2019.
Most blades had worse-than-expected erosion of their trailing edges. The distressed area extended from the platform upwards about 20 in. and from the tailing edge inwards about ¼ in. Some blades also had leading-edge erosion to the same degree. Plus, one airfoil was found with a 150-mil-deep indication about 4 in. up from the platform.
Plant personnel attributed the erosion to multiple years of low-load operation—that is, low steam flow. What happens, the speaker said, is that the root section of the L-0 blades acts like a pump, pulling droplets of moisture back into the trailing edges of the airfoils. This phenomenon is described in the article referenced previously.
New titanium blades, including supporting hardware, were ordered for both rows of last-stage blades. Today there’s a two-year delivery time for these parts.
The following points were mentioned during the presentation (recorded) and follow-on discussion:
- Data provided by pressure transducers installed in the condenser hotwell, said one user, shows promise in helping to determine the degree of wear suffered by last-stage blades.
- The general feeling of the group was that the leading edges of L-0 blades should not be blended to address erosion. Thinking on whether to blend the trailing edges, or not, was split 50/50.
- Stress corrosion cracking is a concern and water chemists should be involved in decisions concerning the low-pressure section.
- Do crawl-throughs and borescope inspections of the LP section at every opportunity.
- Decisions on the lowering of backpressure during winter should be made carefully. Low pressure can exacerbate moisture formation and erosion. As one user noted, it’s a tradeoff between maintenance and additional power.
The speaker ended his presentation by noting that his company sold the plant a year after the major so he did not know when the blades would be inspected again and if they would be replaced with the ones on order.