Plant managers diligently pursue O&M strategies that have a high probability of success for keeping their electric generating assets in service throughout the “must-run” season—this to assure a profitable year. Deciding on what inspections are necessary to access the information needed for reliable decision-making is part of the challenge. Today’s meager budgets do not allow for inspecting everything you’d like to check with the latest diagnostic tools available. You have to pick and choose based on a review of operational data, equipment idiosyncrasies, and your instincts honed over the years.
Consider your heat-recovery steam generator. Many combined-cycle plants do not have a boiler expert on staff—one who is relatively knowledgeable about HRSG water treatment and damage mechanisms, welding, and metallurgy. Given the stationary nature of this equipment, damage, such as cracking at tube-to-header welds, may not be identified until it has reached the point when an outage is necessary.
Thus, given access to industry experience with HRSGs of a design the same as, or similar to, yours, perhaps you can develop a meaningful inspection plan that involves frequent visual checks by staff and detailed inspections by third-party experts every couple of years or so. The experience of others, beyond that provided by colleagues, can be gained through participation in industry meetings—such as the monthly virtual sessions conducted by the HRSG Forum—where you have the opportunity to discuss your concerns with attendees.
Tube-to-header welds. Consider inspections of tube-to-header welds for a moment. You can do a crawl-through and look for signs of cracking when the HRSG has been offline long enough to cool. But you could miss cracks because they tend to “heal” as the unit cools down and it’s unlikely you would see cracking on the back side of the tube in any event. But there could be other visual signs of leaks that you would see—possibly steam scoring on a tube’s external surface.
Magnetic particle inspection (MT) is better than a visual-only check, but surface preparation is necessary and you probably won’t be able to see more than about 200 deg around the tube surface. Phased-array also is a possibility, but it too requires surface preparation and space restrictions might not allow complete access around the tube.
A better choice might be TesTex Inc’s “Claw.” It examines tube-to-header welds for cracking at both the toe of the weld and in the weld (Figs 1 and 2). Typically used for inspections in the HP superheater and reheater sections of the HRSG, it relies on the so-called Balanced-Field Electromagnetic Technique (BFET) for reliable identification of indications. The device is used for header diameters of 4 in. and larger. The claws are able to examine the tube-to-header welds for tube diameters of 1.5 to 3 in.
Tube-wall thinning. Using a proprietary Low-Frequency Electromagnetic Technique (LFET), TesTex technicians can examine HRSG finned tubes from the outside to detect and quantify internal pitting, wall thinning attributed to flow-accelerated corrosion, and under-deposit corrosion (Fig 3). Benefit of LFET is that access holes don’t have to be cut in the tubes to conduct an inspection.
TesTex’s Shawn Gowatski says LFET is very reliable for locating pits of ¼ in. diam and larger, and wall loss in excess of about 20% of the tube wall thickness. This type of damage typically is found in the economizer and back-pass sections of the HRSG, warning of impending issues that can be corrected before a forced outage might be required.
A more exacting inspection of tube internal condition is possible with TesTex’s Remote Field Electromagnetic Technique (RFET) using the internal access tool. The crawler (Figs 4 and 5) is inserted in either the upper or lower header after the end cap is removed and the RFET probe, equipped with a camera, is moved through individual tubes to detect wall thinning and make a video recording of the 360-deg internal surface. Probe travels at 2 to 3 in./sec (Figs 6, 7, and video). Current header-to-header tube-length limit is 70 ft. A benefit of using this tool is that it is able to examine the full lengths of all the tubes in a header.
In a webinar August 12, experts from Mitsubishi Power Americas (MPA) shared a first-order long-term modeling exercise which showed that a centralized hydrogen storage facility and hydrogen-fueled prime movers can economically avoid having to curtail wind and solar energy in the Western Electricity Coordinating Council (WECC) and improve carbon footprint even further than a baseline case.
The modelers acknowledged this type of modeling is a challenge: “WECC has 6000 generating units,” one said. Also, as with all models and forecasts, there are numerous assumptions, including resource costs and electricity demand far into the future, policy frameworks to support decarbonization, and how much renewable energy would actually have to be curtailed.
The so-called “Green H2 Hub” includes massive underground salt caverns (Fig 1), capable of storing up to 150 GWh of hydrogen, produced by 1000 MW of electrolysis units during the winter/spring shoulder months, and available to burn in gas turbine/generators during summer peaks.
In 2050, the need for firm “green” capacity is evident (Fig 2), said the experts. The basic concept is a double bang for the buck—don’t waste carbon-free renewable megawatt-hours because of curtailments, and reduce emissions by burning renewables-derived hydrogen (so-called green hydrogen) instead of natural gas. This long-term storage concept was distinguished from short-term storage using battery-based systems, although MPA also supplies battery systems.
The model also showed that overbuilding wind to solar was optimum.
The Q&A session illuminated these points:
- The well-known (among utilities) Plexis cost model was used, not an in-house model.
- In addition to underground caverns, pipeline packing (increasing pipeline pressure) and “bullet” storage (pressurized vessels) can also be considered, but were not here.
- Only powerplants were considered as off-takers, not transportation or industrial facilities as shown in Fig 1.
- Mitsubishi is targeting 2040 when its turbine/generators are capable of 100% hydrogen firing.
- The capacity factors of the turbine units varied between 28% and 45% in WECC but would not necessarily apply to other regions.
- Though there is no agreed-upon definition, long-term storage has typically been considered as 12 hours or more but is evolving to 100 hours or more.
MPA also took the opportunity to highlight recent accomplishments and projects, including the Advanced Clean Energy Project in Utah, in collaboration with Magnum Development, based on the green hydrogen hub concept; adding 200 MW of battery storage to the Texas grid; J-Power’s 1200-MW Jackson Generating Station in Illinois, a 2 × 1 J-class combined cycle said to be among the world’s most efficient power producers and lowest carbon emitters; and the addition of J-class gas turbines at the coal-fired Intermountain Power Plant in Delta, Utah, which will be an off-taker of hydrogen from the Magnum/Mitsubishi storage facility.
The first thing to note about the GE presentations made at the 2020 virtual conferences of Power Users’ Combined Cycle, Power Plant Controls, Steam Turbine, and Generator Users Groups is that you’re probably going to want to listen to the recordings, if you’re approved to access them at the OEM’s MyDashboard website.
For the Power Plant Controls Users Group (PPCUG), the GE Day content was heavily oriented around elaborating on existing official documents—including numerous Technical Information Letters (TILs), GEH documents for Mark VI control systems, and maintenance documents for specific turbines and subsystems.
Product Services Consulting Engineer Will McEntaggart covered smarter pre-start checks to improve start reliability. Leveraging experience from the aviation industry, he said, “lots of tests can be automated.” Generally, pre-start check philosophy has evolved to better reflect the service duty of the machine.
So, for example, automation startups can proceed with a failed test as long as it doesn’t cause an unsafe condition or risk damage to equipment. Operators are instead given a warning and the system “facilitates testing between runs.”
Included in his remarks were tests for power to DC lube-oil and seal-oil pump motors, manual tests for the large number of fuel and air valves for dual-fuel units (about 70 air/motor/hydraulic valves) and gas-only units (about 16 valves), leakage from small valves in the water-injection purge system, fan motors, and DC batteries.
Some tests are more critical, such as the gas-valve bottle test and the DLN valve tests. Regarding the latter, “all four valves have to be tested simultaneously, which is a pain in the butt,” McEntaggart conceded.
Product Services Engineer Randy Ortiz covered common issues with static-starter systems, specifically switches. He mentioned that TIL-1755 Rev 3, is a “complicated but very important TIL.” It addresses replacing nylon T connectors with stainless-steel flow restrictors in the source bridge. Other topics covered include confusion in troubleshooting exciter trip lockout events in the generator protection panel and generator DC ground faults while at speed.
Product Services Controls Manager Dave Boehmer focused on rationalizing turbine protection. He noted that new software is available for B- and E- class machines that reduce protective actions by 24%. Goal is to retain only those trips, runbacks, and permissives necessary for safe turbine operation, as well as single points of failure. One example given is a turbine trip on oil low pressure delays, which are not required on most new units.
Boehmer also covered stuck bleed-valve trips and wiring issues (such as limit switches which share a common wire and thus constitute a single point of failure) with compressor bleed valves, as well as overspeed testing executed by electrical overspeed protection circuits (in lieu of the mechanical bolt), which reduces subsequent stresses on the rotor from the test.
To conclude, McEntaggart returned to the virtual stage to review fleet lessons learned, which included DC oil-pump systems, battery system run-down tests, inlet-guide-vane and gas-control-valve actuator calibrations, hydraulic-actuator checks, “part substitution risks” when transmitters are replaced by smart transducers, especially when third-party thermocouples are employed (in these cases, you need to pay attention to device time constants in some turbine applications); and general improvements in CIMPLICITY HMI/networking software.
A 480-MW 1 × 1 combined cycle was experiencing NOx emissions above its 2-ppm stack permit limit. The problem was more pronounced with supplementary firing, so the plant was restricted both in its level of duct firing and overall output. This HRSG, which incorporates a heat-recovery section directly upstream of the SCR (the “box” in Fig 1), is of the triple-wide panel design with baffles between the panels.
SCR Solutions LLC was engaged to help the plant determine the cause of the emissions excursion. Principal Consultant Bill Gretta, who specializes in SCR testing and optimization, and AIG (ammonia injection grid) tuning, began his investigation by gathering gas-path data at the catalyst inlet and outlet using the company’s proven top traverse method (Fig 2). Temperature and NOx were measured at 12 elevations via each of the HRSG’s seven roof ports (total of 84 data points), as shown in Fig 3.
Initial results revealed very high NOx (more than 10 ppm) at the SCR outlet in locations directly in line with the baffles separating the tube banks. Temperatures in those same locations exceeded 1000F when duct firing. By contrast, SCR outlet NOx readings in regions away from the baffles, where temperatures typically were about 740F, were generally less than 1 ppm.
Thus, ammonia had to be over-injected to reduce overall stack NOx to permit limits. This led to high ammonia slip and high ammonia consumption (524 lb/hr), and the plant could not maintain its stack NOx limit when duct firing.
SCR Solutions determined that the baffles were not designed properly and failed. Thermal expansion was a contributing factor. The resultant openings in the failed baffles allowed a significant amount of high-temperature gas to sneak through the box and get to the SCR, damaging some of the catalyst.
The baffles were repaired and testing was repeated in the same locations. As shown in Fig 3, the downstream temperatures in line with the baffles were reduced to about 800F. Important to note is that no changes were made to the system beyond baffle repairs.
NOx values at sampling locations in line with the baffles were closer to the stack average post repairs and typical of NOx values in regions of the catalyst away from the baffles.
Ammonia flow was reduced to 417 lb/hr—a 20% reduction—allowing the plant to operate with duct burners at 100% capacity.
A presentation entitled “Does Your Backup System Work—Really?” during Week Three of the Combined Cycle Users Group (CCUG) 2021 virtual conference reminds why user groups can be worth their weight in gold. The presenter, asset manager for a 600-MW, 2 × 1 combined cycle with Siemens 501FD gas turbines and a KN steam turbine/generator, reviewed the root-cause analysis for a catastrophic ST/G failure.
Bottom line: The failure of a $400 relay “wiped out the ST/G” causing “a seven-month outage and tens of millions of dollars in lost margin and equipment repair costs.” Many cliches are represented in this analysis—sweat the small stuff, trust but verify, only the paranoid survive, etc—but perhaps the most important one is “we are all stronger [as an industry] when we learn from each other.”
Some background first: Plant is equipped with 230-kV main transformers. It was designed for auto transfer of ac power to dc, and therefore has limited ac backup from the alternative feed breaker. A 5-kV, 850-amp backup supply comes from an adjacent facility within an eight-second delay. The dc lube-oil pump is supplied with 125-V backup power and a test of the backup supply is included as a permissive in the start sequence.
Sequence of events. After accidental closure of the wrong breaker in the switchyard and loss of station power, the entire plant tripped while operating at baseload. The gas-turbine lube-oil pumps rolled over to dc power and shut down normally, the BOP equipment also came to controlled stops, but the ST/G slowed from 3600 to 0 rpm in about five minutes without the start of its dc lube-oil pump.
During the subsequent restart, the ac pumps—lube oil, boiler feed, circulating water, and aux cooling water—did not turn on, and the dc lube-oil pump did not receive a start signal. Backup from a second alternative source of power did start the ST/G lift-oil pumps, but resulted in a loss of primary containment and a minor lube-oil release in the ST/G building.
Digging deeper. Troubleshooting, including a hand-over-hand wiring check, revealed that an “on” relay was not burned up or damaged, but just failed to operate (a forensic analysis is being conducted to determine why). BOP ac pumps with hands-off auto switches set to “auto” attempted to restart after the trip, but overloaded the feeder, opening the breaker for the backup from the adjacent facility, and tripping the pumps. The second alternative power source, called “construction power,” was found “not fully utilized.”
A deeper review of critical systems revealed several single-point vulnerabilities (SPV) in the ST/G dc logic, but none in the gas-turbine logic. A review of pump load data showed that the total was higher than what the backup ac system was designed for. Learn more about single-point failure vulnerabilities from GTC Control Solutions’ recorded presentation from the 7F Users Group on the subject.
“Circuit heads” will want to take a close look at the two slides comparing the original emergency dc lube-oil pump logic and the revisions, addressing the SPVs. Briefly: A redundant pressure switch was added, coming off a separate existing heater tap (three-way isolation valve). The hot leg is split between two starting contactors with the neutral directly wired, and industrial control relays were replaced with solid-state relays and placed in parallel.
These additional changes also were implemented:
- Changed ac logic so only the ac lube-oil pumps and vapor extractors restart, and eliminated the auto restart of the boiler-feed, circulating water, and aux cooling-water pumps.
- Reviewed all facility schematics and modified for redundancy where appropriate.
- Added construction power backup to all lube-oil and lift-oil pumps, vapor extractors, and battery chargers on the ST/Gs and GT/Gs.
- Arranged to test both circuits automatically during each start sequence; a failure of either relay or pressure switch results in a failed start.
- Emphasize pressure-switch calibration and testing as a PM during outages.
- Determine whether dc lube-oil system can be tested online (in process at the time of the meeting).
Best practices. The most important recommendation to others is to validate control and automation logic, especially backup systems, in operation. Also, if your ac logic has auto-start, calculate total loads to make sure an overload won’t cause a trip, and test the load offline, Finally, review your dc lube-oil systems for SPVs and create, think-through scenarios where primary and secondary losses of power occur and identify contingencies.
Non-chemical alternatives control algae in River Road’s cooling tower
Five years after COD, Clark Public Utilities’ River Road Generating Plant, operated by GE, was compelled by the Washington Dept of Ecology to eliminate the use of chlorine for biological control in its cooling tower. Instead, the plant was instructed to use bromine. Several years after the switch to bromine, River Road began to experience a resilient and chemically resistant form of filamentous blue-green algae. Its growth in spring and summer got out of control.
Expensive and toxic algaecide treatment was required to limit algae growth. The plant discharges tower blowdown to the Columbia River under a NPDES permit with strict discharge limits. One consequence: Blowdown was not permitted for up to 12 hours following this treatment process to assure dissipation of residual algaecide and to maintain permit compliance. That led to high cycles of silica.
During the peak of algae growth, accumulation at the forebay trash screens upstream of the circulating-water pump suction became so great the screens had to be pulled and manually cleaned every three days to prevent collapse. Removal of the trash screens required a crane and crane operator at a cost of $4500 per event (Fig 1).
These factors prompted staff to seek out new technologies to control algae. In 2013, plant personnel located a company that provided algae control using floating ultrasonic devices (Fig 2). They were procured and installed in various locations and orientations in the tower basin. Algae growth was reduced substantially, but not eliminated during the summer.
The search continued for an effective non-chemical technology; reverse osmosis, flocculation, ozone, UV treatment, and electrodeionization were among the technologies investigated.
Because sunlight is required for algae to grow, nursery shade screens were installed around the tower perimeter. They reduced the light required for photosynthesis and did help to reduce the algae more than the ultrasonic devices did on their own. However, the algae problem was not eliminated entirely during the summer.
During the search for a solution, staff located a company, Flow-Tech Industrial Water Treatment Systems, Milwaukee, Wis, that had developed a product used in small-to-medium size HVAC applications for many years, with well-documented success.
The vendor proposed creating and using its patented non-chemical devices on an industrial scale at River Road. The relatively small components are mounted on circulating- and service-water system pipe flanges in strategic locations (Figs 4, 5). Each unit transmits AM radio frequency into the water systems and cooling tower basin.
The theory of design is that the radio frequency disrupts the lifecycle of the algae in its single-cell haploidic form and kills it, before it forms viable colonies. Plant personnel worked with the vendor to develop and implement an industrial-scale test plan for the devices at River Road.
Use of the new devices has measurably reduced the plant’s biological and algae count within the cooling tower (Fig 6). Chemical treatment has been reduced to an algaecide injection once every three weeks rather than every other day. Plant Manager Robert Mash identified the following benefits of the new program:
- A nearly complete elimination of algae in the cooling tower.
- Lower micro-bio counts on weekly dip-slides.
- A significant reduction in chemical consumption for biological control.
- Elimination of crane costs for in-service screen cleaning.
- Reduction in chemical discharges to the river.
- Reduction in the handling of toxic chemicals by plant personnel.
Hydrogen-detection flange tape promotes safety, facilitates leak detection at River Road
Following each generator maintenance activity at Clark Public Utilities’ River Road Generating Plant, the system supplying the machine with hydrogen for cooling was restored and leak-checked using a soapy water product that bubbled or foamed if a leak existed.
A plant employee who attended the 2019 conference of the 7F Users Group saw a demonstration by Nitto Inc, Teaneck, NJ, promoting a pipe flange tape that changes color in the presence of hydrogen. Thus, it provides a visual indication of a leak because it will not return to its original color once hydrogen gas flow is stopped. Even though a soap test indicated no leaks, the staff thought it would be a good practice to use the color-changing tape.
The team member presented this idea as a proactive safety practice to the safety committee. The tape was purchased and installed on every hydrogen flange (Fig 6). Leaks were found, notes Plant Manager Robert Mash.
Thirty-eight flanges were taped. All had been soap-checked prior to using the tape. Once installed, four flanges changed color within two days. Where leaks were indicated, tape was removed from the flanges, and the joints were soap-checked again. Only one of the leaks was large enough to cause bubbling visible to the eye.
The hydrogen-detection flange tape worked well as a pre-emptive safety check and has stimulated proactive thinking about previously unrecognized problems. This fits in well with the long-established plant safety culture at this VPP Star site. Use of the tape is now part of the hydrogen-system restoration procedure.
Real-time monitoring of lube-oil condition at River Road
River Road Generating Plant, owned by Clark Public Utilities and operated by GE, relies on one lube-oil tank to supply the bearings of the single-shaft plant’s gas turbine, steam turbine, and generator. The original oil conditioning unit supplied by the OEM was grossly undersized for the 10,000-gal sump and was removed from service shortly after commissioning.
In 2000, the plant purchased a portable conditioning unit to filter the lube oil. It was functional and cleaned the oil fairly well. As that unit reached the end of its useful life, and the sophistication of the plant team grew, a replacement was sought.
Plant personnel worked with a small, local oil-filtration company to design, construct, and install a new conditioning system—one that would fit in the same physical space as the original unit, have a higher throughput, and include online particulate and moisture analyzers.
Following a water-intrusion experience, plant personnel learned the importance of obtaining oil-quality test results quickly. Relying on sample results from a third party to determine moisture content and particle count had put the staff in a reactive mode instead of a proactive one. The new conditioning system provides real-time monitoring of 4-, 6-, and 14-micron particles. The dew-point monitor and associated instrumentation assure instantaneous detection of moisture in the oil.
The new unit increased the oil conditioning rate from 13 to 38 gpm, reports Plant Manager Robert Mash. Third-party analysis for 4/6/14-micron particulates prior to installation of the new system was 22/20/16; two weeks after installation it was 18/16/12.
Moisture in oil was not tracked regularly before the new system was installed. The data available showed the Karl Fisher number for entrained water averaged around 600 ppm. That number has been reduced to 25 with the new system.
The new conditioning/monitoring system has dramatically improved lube-oil quality. The expected benefit is less bearing wear between maintenance outages.
Oil boom protects against leakage from River Road’s cooling-tower gearboxes
River Road Generating Plant’s cooling tower has five gearboxes, each containing about 20 gal of lube oil. The catastrophic failure of a gearbox poses the risk of releasing oil into the cooling-tower basin. From there it could migrate into the natural environment—more specifically, into the Columbia River—and result in a water-permit discharge violation.
Solution was to install a boom upstream of the cooling-tower circulating-water-pump trash screens to collect any oil that might be spilled, preventing it from reaching the effluent discharge piping. Oil collected can be removed by absorbent pads or skimming equipment, says Plant Manager Robert Mash.
The oil boom installed by owner Clark Public Utilities and operator GE continues to provide a reliable way of preemptively capturing any oil that might leak into the cooling-tower basin. The boom is easy to handle and is cleaned during annual outages.
River Road benefits from combining JSAs and operating procedures
At River Road Generating Plant, owned by Clark Public Utilities and operated by GE, work instructions historically included two separate documents—a job safety analysis (JSA) and an operating procedure. Personnel using the work instructions had to review these documents side-by-side to understand the risks associated with each step of the process.
Moreover, since the facility maintains work instructions for many its processes, reviewing documents and keeping them updated required a considerable time commitment for the onsite team on a regular basis.
Solution supported by Plant Manager Robert Mash was to incorporate the JSAs into the operating procedures, providing a job safety analysis for the task at the beginning of the document. New format also includes more pictures pertaining to the process steps.
- Having the job safety analysis for the given task at the beginning of the document simplifies work instructions. Additional pictures are included for more clarity regarding steps involved.
- Alleviates the need to review multiple documents while performing a task and increasing awareness of the hazards associated with the task.
- Reduces number of documents to be reviewed and updated by half, reducing the time required to keep the documents current and updated.
A combination of ultrasonic energy, AM radio-frequency waves, and sun shades managed to tame severe algae growth in the cooling tower at River Road Generating Station, according to Operations Manager Justin Hartsoch, GE Gas Power O&M. He told participants in the 2021 conference of the Combined Cycle Users Group (virtual) that algae blooms became a problem after the plant was compelled to shift from chloride to a bromine-based treatment to meet EPA NPDES regulations.
In a moment of levity, Hartsoch called the collected algae a “non-cash crop” (Fig 1), but also noted that the plant’s “unique algae bloom is seasonal.” Well water is the source for the plant.
The biocide formula that was being used also fed on silica, creating resistant algae which would coat the forebay trash screens, and require cleaning every two to three days. So, the plant undertook a campaign to identify a non-chemical approach to algae control. The solution to date (“the story is not over,” Hartsoch said) has proved to be the following:
- Adding floating-head ultrasonic transducers at key locations throughout the cooling-tower basin.
- Installing nursery sun shades on three faces of the tower to deprive algae of the sunlight necessary for growth.
- Installing radio-frequency devices on each of the circulating- and service-water lines.
Details of this unique cooling water treatment approach, the theory of design, the circumstances which led to it, and the subsequent results and benefits were reported last year in CCJ. In the CCUG presentation, Hartsoch did note that the payback was a year and a half.
E.C.O. filming technology. Another “cash crop” tale emerged a few presentations later. Greg Boileau, Suez Water Technologies and Solutions SA, described a new proprietary, non-phosphate, circulating-water treatment based on E.C.O., an “engineered carboxylate oxide” filming technology (Fig 2). Phosphates are being phased out because of deposition challenges and formation of algae in receiving waters. Boileau said that one pound of phosphate can lead to 500 lb (wet) of algae.
Boileau reviewed the key drivers for reducing phosphates, sources of phosphates in the plant water balance, and several case studies. Much of the return on investment comes from avoiding the capital cost of a phosphorus removal system. Other Suez information online suggests the proprietary technology reduced phosphates by up to 80% in the pilot study at a large gas-fired plant while maintaining acceptable mild-steel corrosion rates.
You’ll want to watch the video below from Bill Kitterman’s “Tube Repair Clinic: The Good, the Bad, and the Ugly,” even if just to see the pictures of the “uglier” and the “ugliest” tube-to-header welds (photos). In true photojournalism fashion, Kitterman, head of Bremco Inc, now part of SVI Industrial, described six different styles of such welds and the four methods for accessing leaking tubes, including the one fit for an action movie title, “cut your way in, weld your way out.”
Kitterman encouraged industry attendees of the HRSG Forum to “do more to determine the root causes of tube failures.” He also asked users in the audience to understand that, for repairs of creep-strength-enhanced ferritic tubes (such as P91), the downtime required to do quality work could be longer than they might expect. Welding is the fastest part of the procedure, he noted. Stress relief, code requirements (national, state, and local), official inspections, and insurance-company compliance factors take most of the time.
Example: “Bremco has modified its Alloy 91 weld procedure four times since initial qualification—for pre-heat and post-weld heat treatment and weld-wire requirements.” Proper wrapping to maintain the heat during heat-treat is critical.
Kitterman discussed Weld Method 6, a repair that avoids post-weld heat treatment (PWHT) but is only good for butt welds and on tubes with a wall thickness of less than 0.5 in. “This reduces downtime considerably since PWHT can take up to 14 hours,” he added. He also mentioned Supplement 8 for thicker pressure parts, which also avoids PWHT. It’s good for attemperator piping, although Kitterman conceded that Bremco isn’t yet comfortable with the procedure.
One attendee asked about tube plugging, but Kitterman cautioned that plugging a tube can change the flow patterns. It’s no longer being steam-cooled so the hotter gas can impinge on adjacent tubes, and failures could cascade. Another asked if Bremco undertakes turnkey scope; Kitterman answered yes, but prefers to add third-party specialist heat-treat and inspection companies to the team.
Other questions and responses addressed sonic leak detection methods—all captured in the video recording a couple of clicks away.
With pressure parts, so much depends on high-quality welds. Even if welding “isn’t your thing” at the plant you are responsible for, it’s worth watching this presentation to gain a cursory understanding of what’s involved.
Since virtually everyone at a cycling plant faces issues with their attemperators (Fig 1), you’ll probably want to know about a unique design retrofitted to several Duke Energy combined cycles, one that uses the existing spray ring.
Key to the design is use of HP steam to provide the energy to atomize the spray water (Fig 2). This creates a much finer distribution of droplets, regardless of water flow, pressure drop, or steam velocity in the steam pipe, noted Justin Goodwin, director, Steam Conditioning Group, Emerson Automation Solutions.
Steam-atomized nozzles are not new, but are considered unsuitable for high temperature applications like HRSG attemperators. Emerson reached out to corporate colleagues at Fisher™ to design new steam-atomized nozzles that fit into the existing radial-spray, spring-loaded nozzle ring. 3-D printing the nozzles (patented method) of a hardened cobalt chrome alloy (similar to Alloy 6) eliminated the many weld joints, and failure points inside the nozzles of a conventional steam-atomized unit.
“There are no droplets [of water] falling because they are so fine, which avoids the common attemperator failure mode of water impingement leading to damage on internal pipe surfaces. Plus, the design is highly resistant to plugging and corrosion,” Goodwin stressed. A tap at the h-p drum serves as the source of steam.
Note that the design is not applicable to the HP attemperator, only the reheater units. But good news for designers of new HRSGs: Smaller droplets can lead to a 30-40% reduction in piping lengths.
Lessons learned during the field trial are that a 1 in. to 2 in. connection in the atomizing steam supply piping was a choke point, as was use of a Y-pattern valve instead of a full-bore ball valve (a pressure transmitter was added to troubleshoot these issues). Modified control logic design is critical to a successful retrofit. In response to a question, the presenters noted that they replaced the water temperature control valve, but not the block valve.
Eugene Eagle, HRSG engineer, Duke Energy, and Goodwin’s co-author, said that Duke was pleased enough with the initial field trial on one unit at the utility’s Buck Combined Cycle Plant that they installed the new design at Dan River Generating Station on four additional units. The attemperator with the longest service life had 18 months of operating experience at around 85-90% capacity factor at the time of the presentation.
Duke has eliminated several failure modes, as well as the two-year inspection and test schedule for the previous spring-loaded nozzles, and is in the process of determining the cycle life for the new design. Current thinking is that the nozzles could warrant replacement every three years and that the internal piping liner should be borescope-inspected every two years. Thermal fatigue is the expected nozzle failure mode.
Many of the other questions addressed aspects of the control system (such as the operation of the block valve with the control valve), leakage at the block valve (you need a good block valve and trust it to be tight), the potential need for a second block valve, and possible issues with wet steam in the atomizing steam piping.
Presentations made by MD&A, Shell Lubricant Solutions, ARNOLD Group, and EthosEnergy Group to owner/operators participating in Weeks Three and Four of the virtual STUG2020 conference are summarized below. You can access the recordings and PowerPoints on the Power Users website.
This presentation by James G Miller, PE, manager of performance services for MD&A, is a valuable primer for plant personnel participating in their first steam-turbine outage and equally valuable as a refresher for more experienced engineers and technicians. Miller’s message: Use the results of (1) recent performance tests conducted with the unit in service, and (2) steam-path audits made in the early stages of the outage, to reduce both outage cost and duration.
Miller reminded attendees that performance losses are a sign of degrading conditions that adversely impact the plant’s bottom line. The outage affords the opportunity to use this information for pursuing repairs and upgrades of greatest economic value.
The speaker covered the basics of performance testing, and how to conduct the all-important steam-path audit, in his presentation, which is available on the Power Users website complete with formulas, calculation examples, a comparison of as-tested performance to reference data, etc.
Best practices in performance testing—such a making sure there’s at least 25 deg F of superheat when calculating turbine efficiency—are included in the PowerPoint, together with a list of diagnostic parameters or additional tests that can be used to further characterize the sources of loss—such as solid particle erosion (SPE), casing leakage, and valve leakage. Thermal scans are particularly valuable for identifying the locations of leakage (Fig 1).
Examples of typical sources of performance loss identified during the steam-path audit include the following:
- Seal leakage (Fig 2).
- Surface roughness.
- Change in trailing-edge blade profile.
Case studies identifying the reasons for performance loss in a reheat turbine at a combined-cycle plant, in a reheat turbine for a conventional steam plant, and in an industrial double-extraction condensing turbine are highly informative. For the first unit, performance testing revealed gross output had decreased by 2.3%. Excessive surface roughness, worn end packing, rubbed tip spill strips, and leakage by startup vents and HRSG drains were among the primary contributors to the loss.
A checklist of information to review in overhaul planning concluded the presentation.
Shell Lubricants: Choosing your lubricant not a one-size-fits-all
Lubricant selection is one of those subjects you might not think about for years, but when necessary it’s good to have a backgrounder like this at your fingertips—or only a couple of mouse clicks away on the Power Users website.
Key discussion topics include these:
- Base-stock evolution (Groups I through V).
- Mitigation methods for varnish—including top-off fluids, filtration units, and fluid solutions (polyalkylene glycol, gas to liquid—lubricants made from natural gas).
- Field experience.
Varnish elimination with polyalkylene glycol (PAG) was a focal point of the Shell Lubricant Solutions presentation, which included a review of experience since 2001 at two units that switched to PAG to eliminate servo valve issues caused by varnish. Since then there have been no servo failures or trips while on PAG.
Recall that the stress experienced by a turbine lubricant contributes significantly to the ageing of petroleum oil, causing the non-polar fluid to oxidize. However, the resulting byproducts of decomposition are polar and insoluble in the base oil; they come out of the solution as varnish. Polyalkylene glycol, by contrast, is a polar fluid and, while it too oxidizes, the byproducts of decomposition are polar and infinitely soluble in the base stock. No varnish is produced.
ARNOLD Group: Advanced steam-turbine warming for increased startup flexibility
Pierre Ansmann opened his presentation on “the most advanced turbine insulation combined with a high-performance heating system to improve startup flexibility,” by summarizing its value proposition thusly:
- Increased in-market availability.
- Lower startup costs.
- Reduced thermal fatigue and longer mean time to repair for critical components.
- Increased operating flexibility.
He reviewed alternative warming-system arrangements, rejecting those integrating the heating circuits in insulation blankets, installing the heater on a thin mattress below the blanket, and using glass-fiber-insulated heating cable. The optimal system for the upper casing, they said, is heater on metal mesh baffle, for the lower casing, permanent mounting of heating cable below the split line.
The ARNOLD system features interlocking high-performance blankets which conform perfectly to the turbine surface (Fig 3). High-quality materials and manufacturing, and long-term high-temperature resistance, allow the company to guarantee reuse of its insulation system for 15 outages without a decrease in efficiency.
Dozens of thermocouples, strategically located on the turbine, ensure proper heating. Each of the 18 or so heating zones has t/cs installed on the heating wires to double check if the zone is responding correctly and at the specified temperature. Below every heating zone, multiple t/cs are mounted on the casing to confirm even heating of the turbine.
Ansmann said a properly maintained ARNOLD insulation system can maintain your turbine in a hot-start condition for at least four or five days after shutdown. No preheating of the turbine is required prior to a start within this time period, reducing startup fuel consumption and auxiliary power.
Combining high-quality insulation and warming systems enables tight control of casing-to-casing and rotor-to-casing expansion during shutdowns. A goal for operations personnel to aim for, Ansmann said, is a homogeneous cooldown to maintain the temperature difference between the upper and lower casings to less than about 100 deg F. Access the recording and slides here.
EthosEnergy Group: Multiple upgrades improve D11 reliability
Owner/operators of the popular D11 steam turbine are sure to benefit from a review of this illustration-rich, 50+ slide presentation, easy to access the recording and hard copy in the Power Users archives. The presenters from EthosEnergy cover the repair of 40-in. L-0 blades, and upgrades of Smart seals and the N2 packing box, among other things. The subject plant was a 4 × 2, 1240-MW combined cycle. COD for the unit upgraded was 2011; first major inspection in 2020.
Two rows of the damage-prone L-0 blades were weld-repaired prior to the outage to correct excessive leading-edge erosion (Fig 4). Cracking in the blade pin-finger dovetail roots also was addressed. Presentation provides details likely of value to anyone facing the same issues. Photos illustrate key steps in the process, including re-blading of the L-0 row.
The Smart seal upgrade was done to address rotor vibration caused by seal rubs. Experienced users know the HP/IP rotor is very flexible and sensitive to mid-span rubs. Detailed measurements of packing and tip-seal wear (average horizontal, top, and bottom) are presented. Heaviest rubs are identified with the lower-half horizontal joint. The speaker noted that although clearances generally are larger on the bottom, wear is substantial at all locations.
A seal developed to upgrade OEM seals to avoid rubs and wear during startup and shutdown, by way of additional clearance, is illustrated. Reduced vibration during startup is one benefit. Another is increased revenue, said to be upwards of $17-million for a typical 300-MW steamer over an eight-year run time. Information on estimated savings in fuel and carbon emissions also are presented.
N2 packing heads, which contain shaft seals between the HP and IP steam paths, have a history of horizontal joint leakage. This impacts performance because HP inlet steam leaks into the IP section. Plus, steam cutting occurs across the horizontal joint. The presenter highlights what his company’s experts say are design issues that prevent maintaining a closed joint. Described modifications are said to mitigate the problem.