Onsite

Consider the impact of new operating regimes on your SCR

By Team-CCJ | April 19, 2022 | 0 Comments

It’s easy to forget about the “big box” selective catalytic reduction (SCR) unit sandwiched in your HRSG modules, even though it stands between you and compliance with your air permit, and your ability to operate.

At the 2017 conference of the Combined Cycle Users Group (CCUG), held in Phoenix the last week of August, Andy Toback, Environex Inc, did his best to remind users that SCR process parameters have to be re-evaluated when gas turbines are upgraded, plant operating tempos change, duct-burner operation is more prevalent or variable, and/or the latest G and H technology machines are being deployed. Otherwise, you may be leaving money on the table or setting yourself up for unexpected costs and performance issues down the road.

Toback’s main message is that you need to adjust the original design expectations for the SCR based on real-world operating data, the key to optimizing the process for new conditions.

For an actual 7FB.01 to 7FB.04 upgrade, for example, the new combustors added 7 MW of output and lowered turbine NOx levels entering the SCR (Fig 1). Because the catalyst going forward typically will convert 9 ppm NOx levels from the turbine, compared to the original design of 25 ppm, the relative catalyst activity level, representing the expected end of life, can now be projected out beyond 10 years (Fig 2), compared to 5.5 years with the original combustors.

Ammonia consumption and ammonia slip at the stack also are reduced significantly because of the lower NOx conversion requirement for the SCR system. Due to the improved combustor dynamics, gas-turbine CO levels are not expected to change even though the NOx levels have decreased.

Toback’s second example is a 7FA.03 to 7FA.04 + DLN 2.6+ upgrade (Fig 3). The good news is that the megawatt output gain was higher than expected, another reason why actual measurements are important. However, peak-load fuel and exhaust flows increased accordingly, the turbine exit NOx levels did not change appreciably, but the SCR temperature increased by about 20 deg F after the upgrade.

The original Dot 03-machine catalyst design life projection was about 21 years. The upgrade design expectation reduced it to around 13 years. Based on the actual operating data, however, the catalyst life should be more like 18 years (Fig 4). Both an expected increase in engine NOx from 9 to 10.4 ppm and higher exhaust flows impair catalyst life, but because the actual operating data showed no change in gas-turbine NOx levels, the minimum catalyst-activity requirement did not increase as much as anticipated.

For advanced G and H machines, the news isn’t so good. Toback states, “We’re being asked to achieve the same 2 ppm NOx and ammonia-slip levels (typical of the toughest permits) even though these machines have five times the turbine-exit NOx levels.” Plus, they likely will be required to cycle and operate at less than design output for significant operating hours over their lifetimes.

Fig 5 indicates that for these performance specifications, you’ll either have to accept high operating and compliance risk at the ragged edge of the capabilities of traditional single-bed SCRs or resort to a more complicated and more expensive SCR design.

When you operate advanced-technology machines at low loads, you tap out the capabilities of the design (Fig 6). “The ammonia injection grid can’t handle both the NOx levels at the maximum design output and what would be typical at 30-50% load, because of the corresponding changes in mass flow, temperature, and mixing.”

Environex specialists believe owner/operators of G- and H-class machines will have problems because vendors are supplying SCRs with inadequate catalyst volumes. It’s also important to consider adding a permanent grid made of stainless-steel tubing within the HRSG housing which allows you to periodically take 2-D distribution-grid measurements for NOx and NH3 and more accurately tune the AIG distribution valves.

Such capability nominally adds about $50,000 to the budget, a sum that looks paltry compared to the penalties of non-compliance.  The high NOx conversion requirements for these systems coupled with low ammonia-slip limits decrease the tolerance for non-ideal ammonia-to-NOx distribution. You should expect this to increase the required frequency for ammonia-grid tuning.

Duct burners, of course, are another source of NOx and CO which must be accounted for through real operating data. SCR inlet NOx can more than double at design GT output and full duct firing and the SCR operating temperature can climb by 50 to 100 deg F. Inlet CO, meanwhile can decrease. During interim periods as duct burners come up to full capacity, or remain at part load, NOx and CO emissions can be quite variable. These impacts can be quite striking if your unit was designed for baseload operation.

Generally, Toback concludes, increases in exhaust flow to the SCR impairs catalyst life, while decreases in SCR inlet NOx and CO emissions and increasing SCR operating temperature extend it. If your machines are no longer operating the way the SCR was designed, it’s time to consider a program to acquire the operating data needed for optimizing the SCR process for new conditions.

How natural-gas fuel variability impacts GT operation

By Team-CCJ | April 19, 2022 | 0 Comments

Natural-gas fuel composition has considerable influence on gas-turbine emissions and operability. This might not have been a concern to you previously, but the possible impacts on your engines of increasing reliance on non-traditional pipeline gas—in particular, shale gas, but also LNG, and byproduct fuels from industrial processes—suggests you might want to read on.

Even if you have “good gas,” be aware that the negative impacts of gas-constituent variability can be magnified by the high firing temperatures of today’s most advanced gas turbines, Scott Sheppard, Ben Emerson, and Tim Lieuwen of Turbine Logic told the editors. Reason is that these machines must “work harder” to achieve low-NOx emissions and thus have fewer “knobs” to mitigate fuel-composition impacts, the trio added.

The Wobbe and Modified Wobbe indices are two standard properties used in practice to determine the effects of fuel variability. But they only indicate the changes in heating value and usually cannot describe how changing fuel composition will affect emissions and some potentially damaging operability issues.

Aside from NOx and CO emissions, there are four operability issues of importance to plant personnel. They are:

      • Blowout, or when the flame physically exits the combustor; also referred to as blowoff.
      • Flashback, or when the flame moves upstream into premixing sections not designed to withstand high temperatures.
      • Combustion dynamics, or the damaging pressure oscillations associated with oscillations in the combustion heat release.
      • Autoignition, or the spontaneous ignition of a reactive mixture of fuel and air in the premixing section of the combustion chamber. The effects of autoignition and flashback are the same, but the underlying reasons and fuel sensitivities are completely different.

These four issues, which the Turbine Logic team routinely observes when performing root cause analyses, pose risks to gas turbines and are closely tied to operating conditions and fuel composition.

Before discussing the impacts of fuel variability on gas turbines, Sheppard, Emerson, and Lieuwen said it is important to understand the composition of natural gas, which in North America typically contains 85% or more of methane. The next most common constituent is ethane at about 7% or less, followed by propane at about 1.5% or less. Other components include other higher hydrocarbons, nitrogen, carbon dioxide, and hydrogen. From an operability point of view, it is convenient to divide the fuel constituents into three buckets: higher hydrocarbons, hydrogen, and diluents.

As previously mentioned, the Wobbe Index and Modified Wobbe Index both describe the effect of fuel composition on fuel heating value, with the Modified Wobbe Index also accounting for fuel temperature. In addition to Wobbe Index recommendations, OEMs typically include a fuel specification for their gas turbines on levels of higher hydrocarbons and hydrogen. By operating within these recommendations, operators generally can avoid some of the operational issues mentioned earlier. However, the combustion-dynamics issue is particularly sensitive and difficult and is an issue that must be actively monitored and managed by operators, the trio stressed.

There is a relationship between some of these operability issues and Wobbe Index—for example, higher ethane and propane contents in fuel leads to a higher fuel heating value, which also increases autoignition risks. Nonetheless, two fuels with the same Wobbe can have substantially different blowout, flashback, combustion-dynamics, and autoignition tendencies. Thus, Wobbe Index alone cannot be used to ensure safe, reliable operation.

Effects of fuel composition on operability

The operability concerns described below by Sheppard, Emerson, and Lieuwen apply mostly to dry-low-NOx/dry-low-emissions/ultra-low-emissions (DLN/DLE/ULN) systems. They are much more prone to these issues than diffusion systems—most notably combustion dynamics, which are a product of combustor design and the effort to keep NOx emissions as low as possible. Diffusion systems do experience some dynamics issues near blowout, but they can be avoided by operating with a sufficient blowout margin.

Blowout. One of the biggest factors impacting flame blowout—also known as blowoff or lean blowout (LBO)—is the flame speed of the different fuel mixtures. Flame speed is the speed at which the flame moves through the reactants during combustion. Hydrogen addition has particularly significant impacts on increasing blowout margin because of its much higher flame speed than methane.

Propane and ethane also have higher flame speeds, but the effect is much less significant than hydrogen. By contrast, diluents (particularly CO2) reduce blowout margins as they act to reduce the fuel’s flame speed.

Takeaway 1: Increases in diluents pose the greatest blowout risk.

Flashback. As previously mentioned, the presence of hydrogen and higher hydrocarbons can move a flame upstream into premixing passages. Swirlers, fuel lances, and premixing passages are not designed to withstand the high temperatures that would occur with a flame in the premixing section and can quickly sustain damage.

In some events, this hardware may liberate and travel downstream through the power turbine, causing catastrophic damage. Loosely speaking, flashback can be thought of as the “opposite” of blowout; for this reason, the H2 level limits in DLN system fuel specs are set by flashback margin.

Takeaway 2: Increases in hydrogen content pose the greatest flashback risk. 

Combustion dynamics. Fuel composition directly affects combustion dynamics. However, in contrast to other operability concerns, the effect is non-monotonic with operational parameters which make it particularly difficult to predict. For example, while increasing H2 levels always will decrease flashback margin, combustion-dynamics amplitude can either increase or decrease, depending upon other operating conditions.

Takeaway 3: Changes in fuel composition may decrease combustion dynamics, increase combustion dynamics, or have no effect at all. 

Combustion dynamics pose a direct threat to combustion liners, transition pieces, and cross-fire tubes. If the combustion-dynamics amplitudes are severe, liberation of these parts may occur, with hardware travelling downstream through the power turbine. Combustion instabilities can also lead to blowout and flashback.

Because of the sensitivity of combustion dynamics to the details of the flame configuration inside the combustor, dynamics are sensitive to all changes in fuel composition—that is, changes in higher-hydrocarbon, hydrogen, or diluent content. Note: In this case, the risk is changes  rather than increase. The Turbine Logic combustion dynamics monitoring group has observed increased challenges with dynamics, and the need for more frequent tuning because of gas composition issues.

Autoignition. Hydrogen and higher hydrocarbons also significantly decrease autoignition temperature and times, leading to increased autoignition risks. While symptomatically similar to flashback, the physics of autoignition is different. In this case, if the autoignition time of the fuel mixture falls below its residence time in the premixing tubes, the fuel mixture may spontaneously ignite in the premixing zone.

Similar to flashback, autoignition can damage premixing passages, swirlers, and fuel lances.  As with any of these other operability issues, liberation of any of these parts poses a threat to the power turbine. Fig 1 shows the how quickly autoignition times decrease with increasing amounts of ethane and propane in natural-gas fuel.

Takeaway 4: Higher hydrocarbons (especially when coupled with the higher compressor discharge temperatures of aeroderivatives) pose the greatest autoignition risk.

Nitrogen oxides. There are two main NOx formation pathways:

      • NOx produced in the flame, which generally is a few ppm and heavily influenced by fuel composition.
      • NOx produced post-flame, which really only depends on flame temperature.

Thus fuel composition influences on NOx emissions are somewhat dependent on the nominal NOx emissions of the gas turbine. Assuming the firing temperature stays fixed as fuel composition varies, for units with relatively high NOx emissions (nominally 15 ppm and higher), NOx production is dictated mostly by the firing temperature and is largely insensitive to fuel composition.

On the other hand, units that have fairly low NOx emissions (less than 5 ppm) will see a significant effect from changes in fuel composition. Data suggest that NOx concentrations may even double with changes in fuel composition for such low-NOx systems. Fig 2 shows how much NOx production can increase, in the low-NOx case, with increasing amounts of ethane and propane in natural gas.

Takeaway 5: The lower a gas turbine’s NOx emissions, the greater the impact of fuel composition on those emissions.

Carbon monoxide. The effects of fuel composition on CO production are largely controlled by whether it moves the system closer to, or farther from, lean blowout. If the unit is near LBO, CO emissions typically already are high, but can still be raised significantly by the presence of higher hydrocarbons and hydrogen in the fuel. Away from LBO, CO emissions may still rise with the addition of higher hydrocarbons and hydrogen, but the effect is quite small.

Takeaway 6: Similar to blowout, increased diluent content poses the greatest CO risk.

Recommendations

A fuel treatment system will mitigate the fuel-variability effects on each of the operability issues discussed above, Sheppard, Emerson, and Lieuwen said. These systems are equipped for removing solid particulates and higher hydrocarbons. The latter is accomplished by condensation and liquid removal.

Most OEMs also recommend some degree of fuel superheating before use in a gas turbine to ensure that no fuel condenses before reaching the fuel nozzles. Condensed fuel can cause significant autoignition issues, the trio warned. Even with a good fuel treatment system, small concentrations of liquid hydrocarbons from the fuel gas remain a risk when liquids are entrained from the liquid knockouts. This may occur with higher hydrocarbons, because the knockouts fill more quickly and require more frequent draining than experience might dictate.

Along with Wobbe Index recommendations, OEMs issue fuel specifications for their units. They are designed to accommodate blowout and flashback margins, and those phenomena should not occur when operating within the specs. However, even when operating within the recommended specs, combustion dynamics and emissions may be impacted by changes in fuel composition.

The greatest operational risk that cannot be managed by staying within OEM fuel guidelines is combustion dynamics. The Turbine Logic experts strongly recommend users have a combustion dynamics monitoring system (CDMS) to warn of impending issues, and also have protocols in place for ensuring sensors are healthy; plus, appropriate notifications if levels exceed alarms.

In their experience, every engine monitored by Turbine Logic has had multiple dynamics excursions exceeding thresholds as the ambient temperature and load varied. The need for shorter tuning intervals only can be confirmed by CDMS.

Several options are available to control loud combustion instabilities, and all are used by manual- or auto-tuning systems. These include varying the fuel temperature within the OEM spec, altering the fuel staging, increasing the non-premix pilot fuel, varying the amount of inlet air chilling, and varying the amount of steam/water injection. De-rating is the last resort.

To complicate things, because of the non-monotonic nature of combustion dynamics, a change in fuel composition can alter the unit’s response to each of these operational changes. Since combustion dynamics don’t have a one-to-one relationship for different operating conditions, installation of a CDMS can provide great insight into how your units respond to different operational changes. CDM systems provide nearly instantaneous feedback and will facilitate the building of a knowledge base around how your units behave regarding combustion dynamics.

If you continue to have combustion-dynamics issues while keeping NOx concentrations within regulatory limits, consider installing an auto-tuning system. It monitors dynamics and emissions and has control logic to minimize emissions while keeping combustion dynamics within acceptable limits to prevent hardware damage.

Auto-tuning systems typically have the ability to regulate several operational parameters— including firing temperature, fuel splits, and even fuel-split schedules. When either dynamics or an emissions excursion occurs, auto-tuning systems step in and nudge these parameters to return the gas turbine to safe and compliant operation. While seasonal tunes remain a good idea, and sometimes a necessity, the auto-tuning system will handle the day-to-day changes that gas composition variability may introduce.

Questions? Contact Sheppard, Emerson, or Lieuwen at connect@turbinelogic.com or 678-841-8420.

 

Paul White’s legacy

By Team-CCJ | April 19, 2022 | 0 Comments

When Paul White crossed over to the spiritual world from the physical Nov 16, 2017, he was well prepared for the journey and certainly was welcomed by those who had preceded him. White was a deeply religious man who taught Sunday school and gave church activities his full support—all while excelling in his engineering day job, and contributing to gas-turbine user-group initiatives with his deep technical knowledge and considerable people skills.

The challenge in writing an obituary for an exemplary human being and professional is that it never really meets your expectations, and likely those of the many readers who knew White well—there’s so much to say. Perhaps the best way to honor such an individual is to keep him in your conscious mind to help guide your thinking. You might ask yourself when dealing with a knotty problem: What might Paul have done in this situation?

The editors knew White best from his involvement in gas-turbine user groups. His highlight reel includes leadership roles in the following organizations:

      • 7F User Group steering committee member from just prior to the millennium until retirement as a full-time Dominion Resources Inc employee in the first quarter of 2013. He was a past chairman of that august organization.
      • Combustion Turbine Operations Technical Forum (CTOTF™), leadership committee member for many years—including several as chairman of the GE Roundtable and the Siemens Roundtable.
      • Combustion Turbine & Combined Cycle User’s Organization (CTC2), past chairman of the steering committee.

White was among the industry’s best discussion leaders and a great asset in any Q&A session having to do with gas turbines. He was successful in getting his point of view across without raising his voice or denigrating anyone’s opinion. His approach was debate, yes; argue, no. White’s technical knowledge and calm demeanor earned him industry-wide respect. One example: He was recognized by the Combined Cycle Users Group in 2015 with its Individual Achievement Award.

White took great pride in his ability to recognize talent from afar, recruit those individuals, and mentor them to the point where they could advance on their own and expand the capabilities of the gas-turbine technical support organization he led. White made it a point to bring new members of his team to user-group meetings and introduce them to as many participants as he could. He knew well that building a proper network to aid in decision-making was vital to success for both the employee and employer.

The registered professional engineer (North Carolina) also knew that he could not properly motivate and lead without keeping up with generation technology. White participated in many technical symposia over the years, several focusing on new materials and cooling schemes for gas turbines.

Consider that from the time he graduated from North Carolina State University in 1974 with a BS in Mechanical Engineering—two years after the nominal 52-MW GE Frame 7B was introduced—until he retired from Dominion as a part-time employee in December 2016—shortly after the 384-MW GE 7HA.02 became available—turbine inlet temperatures increased from about 1800F to 2900F.

Career profile

1974—Bechtel Power Corp, nuclear focus, design and field engineer.

1978—Duke Power Co, senior engineer, gas turbine (and steam turbine) technical support.

1997—Duke Energy North America, director of engineering, responsible for strategic turbine expertise in both current and developing technologies.

2000—Dominion Resources Inc, manager of O&M, provide technical support and strategic management for a fleet of about 75 gas turbines, spanning legacy to advanced technologies.

2016—GT & ST Consulting Corp, shop build surveillance on complete component and unit assembly tasks for Dominion’s first J machine.

Remembrances from. . .

Bob Kirn. Amidst the early 7F Users Group meetings that were more of a fruitless slugfest between the OEM and a handful of owners suffering from bucket munching turbines,  there emerged an individual who, while already known as a steady hand and cool head, would become one of the stalwarts of the gas-turbine industry.

Paul White, with his steadfast, oft-repeated belief that problem resolution could be best accomplished through cooperative efforts punctuated by full disclosure and free discussion, was quickly recognized by the user community as one who could be trusted to not only “find the solution” but to—and of such greater importance—“share the solution.”

His successful method of cooperative effort was so widely recognized that even GE invited him to speak to their services group in the company’s efforts to foster a more cooperative approach to customer service. Paul had made a lifetime of sharing his talent and he agreed to participate, and with the same level of honestly and enthusiasm that he approached everything.  No doubt the subsequent and on-going success of the 7F Users Group is a legacy to the open-forum structure he promoted and to the personal qualities that he so strongly displayed.

For more than two decades, Paul and I exchanged information, sought and offered advice for never-ending calamities and frequently crossed paths on the user-group circuit.  His commitment to finding solutions and providing the best information possible to anyone who asked never wavered; nor did my personal pleasure in seeing him at meetings and being able to swap stories, bounce new ideas, or just share a genuine handshake.

Christa Warren. Paul was the best manager and mentor I ever had, and will always remember what I learned from him. Thanks to Paul, I had to opportunity to join the gas-turbine industry and work with him. He created a legacy that will live on through all of his mentees. Wherever and whenever his name comes up it always results in a positive comment; no matter whom you talk to. Even I catch myself thinking, “What would Paul have done?”

He presented himself and his team with such class, humility, and respect he made everyone around him feel valued and respected. I will never forget how he would refer to all colleagues as his friends, and how when he described your role on our team he made you feel like a million dollars. It is sentiments like this that make people remember Paul for who he was. Although he will be deeply missed, the impact he made on those around him will continue to resonate through the years.

Sam Graham. Of all the good things I can say about Paul White, the one that stands out the most is his wonderful character. Paul was a true gentleman in every sense of the word, which was obvious to everyone around him; he was a genuine pleasure to be with. Paul worked constantly to mentor younger engineers and to make a positive impact on the 7F Users Group. He was a great contributor to the user community and a long-serving member of our steering committee.

You could always rely on Paul to provide a composed and thoughtful voice to any situation. His wisdom and self-control were vital during the hard conversations required with the OEM when these machines were in their infancy. Paul could be counted on when times were rough, always with a warm heart and a smile. This industry, and this world, could certainly use more people like Paul. He will be greatly missed.

John Gundy. We remember Paul White as a loving and devoted family man, mentor, and friend. Not only was he caring and compassionate, Paul was humble, adventurous, and enjoyed life, sharing his faith with those around him—which was something special. His contributions to the industry and willingness to teach others from his life’s lessons were priceless. I was blessed to be hired by Paul, and he mentored me before I took on the role of engineering manager for Dominion Energy’s Combustion Turbine Operations. Paul will be missed at Dominion Energy and by the many friends he made throughout his career.

Ontelaunee details experience with GT uprate, exhaust cylinder fix

By Team-CCJ | April 19, 2022 | 0 Comments

The 501F Users Group’s 2018 conference and exhibition is only a few months away—Feb 25 – Mar 2, at the Hyatt Regency Grand Cypress, Orlando, Fla. It is your only chance in the coming year to visit with the major service providers and parts suppliers for the 501F fleet. Exclusive closed sessions for owner/operators with Siemens, Mitsubishi, PSM, and GE are a highlight of the program.

Add to that package the following:

      • User presentations on recent experiences (both good and bad).
      • Direct participation in roundtables dedicated to the compressor, turbine, combustion system, rotor, generator, inlet and exhaust sections, auxiliaries, safety, and hot-gas section.
      • The opportunity to select from more than three-dozen vendor presentations on O&M topics selected and vetted by the steering committee.
      • The opportunity to spend quality time with nearly 100 vendors in the exhibition hall.

No budget, no time is no excuse for not attending this meeting. Participation will save significant staff time and O&M costs in the coming year because you’ll be better prepared to anticipate problems already experienced in the fleet, know whom to call about them, and solutions to implement.

Register today!

Following up on key topics from the 2017 meeting a couple of weeks ago, the editors caught up with Adam Sensenig, who made two presentations at last February’s conference in Reno, Nev. During a visit to Dynegy’s Ontelaunee Energy Facility in Reading, Pa, Sensenig, Ontelaunee’s plant engineer, provided additional details on the recent uprate of the facility’s two gas turbines and a “fix” for cracking of strut shields in the two-piece exhaust cylinder, a 501F fleet-wide issue.

As background, Ontelaunee is a 2 x 1 combined cycle which went commercial in 2002, and features 501FD2 machines. In recent years, the plant has been operating primarily at base load, with an average capacity factor of 87% for 2016. In 2014, the plant contracted with PSM for a long-term service agreement (LTSA) and the company’s GTOP 6 upgrade package.

GTOP (Gas Turbine Optimization Package) is PSM’s non-OEM performance-enhancement offering for the 501F market. The uprate increases mass flow to 501FD3 levels.  By signing with PSM, the plant avoided the exhaust cylinder and R4 blade-ring replacements characteristic of the OEM’s uprates.

The plant was first out of the gate with GTOP for the 501F. About 18 months transpired between project kickoff and completion.  Overall, Ontelaunee gained 7% in net plant output with a 1.7% improvement in net plant heat rate. It was able to maintain a large portion of the increase in output by way of an aggressive online water-wash program.

Five upgrade elements. Sensenig breaks down the uprate modifications into three areas: new blade-path components that take advantage of prevailing metallurgical, coating, and design improvements; modified inlet-guide-vane (IGV) actuators to boost air flow by extending stroke length from minus 2 to minus 6 deg; and by adding auto-tune capability.

Overall goals were to increase the air/mass flow through the turbine, reduce the amount of air required for blade path cooling, and achieve a higher total exhaust temperature.

Regarding the blades, the R16 compressor blades feature a different airfoil shape, and were a “drop-in” replacement. The R1 blades, vanes, and seals were modified to decrease cooling-air flow, as were the R2 blades and vanes. The last-stage R4 GT blades are about a ¼-in. taller to accommodate the higher mass flow and additionally reduce exhaust swirl. R3 blades and R3 and R4 vanes remain the same. Of these, Sensenig credits opening up of the IGVs and the taller R4 blades as having the biggest impact on the results.

The IGV actuator mods could be accomplished by replacing the actuator or modifying existing ones. Ontelaunee had one spare actuator which was modified for one unit and the plant purchased a new actuator for the other GT.

Auto-tuning, through the combustion dynamics monitoring system (CDMS) and PSM’s Autotune Version 2, assures flame stability—and emissions stability—under all operating conditions. One consequential saving with the auto-tune, says Sensenig, is that the plant no longer has to call someone out to adjust the controls for seasonal ambient conditions.

Attention to BOP. The upgrade did require the plant’s (and its third-party engineers) careful attention to balance-of-plant (BOP) impacts. PSM conducted the plant assessment up to through generator output but only guaranteed simple-cycle GT performance. Some BOP impact examples:

      • The four-way joint where the turbine casing and the combustor shell meet has significant fleet-level issues. The uprate leads to higher shell pressure, and the impacts, such as greater potential for leakage, has to be monitored.
      • More air flow through the turbine, of course, means more air flow through the HRSG; plant personnel need to keep up with HRSG maintenance, and be cognizant of the higher HRSG backpressure.
      • The HRSG high-pressure steam drum safety valve had to be resized and replaced.
      • The SCR’s ammonia vaporizer is running at near capacity at maximum output; while ammonia consumption per megawatt dropped, the absolute level of ammonia feed increased because unit throughput increased.
      • Water-treatment chemical consumption increased because of higher demineralizer demand and cooling-tower load.

In general, Sensenig notes, “We’re anxious to see what the parts will look like after 24 months of operation.”

Because Ontelaunee was Rev 0 for the PSM GTOP, a substantial effort was required to qualify the R4 blade design, which added three to four days to the outage. Blade monitoring had to be conducted under a variety of operating modes, including startups, shutdowns, speed sweeps, IGV sweeps, and inlet fogging and steam power augmentation.

Other issues Sensenig describes as run-of-the-mill for outages and significant equipment modifications. The plant experienced a trip during over-speed/under-speed testing because of the trip-limit settings. Capability to modify the controls within the TXP DCS “was limited” and the plant had to “clean up” unused counters and other items to free up processing space, but Sensenig notes this should not be an issue with the newer control systems.

Exhaust-cylinder fix solves “most” problems. Ontelaunee was not first for the exhaust-cylinder fix, performed by Texas-based Braunflex LLC. Important to note is that this fix doesn’t solve all the problems experienced fleet-wide with the support struts, but for a fraction of the cost of a new exhaust cylinder, it solves the most important ones.

At the 501F Users Group meeting, Sensenig reported that, upon inspecting one modified unit after 5000 operating hours and 10 starts, only minor stress-relief type cracking on the inner load plates had been observed. At the time of the CCJ visit, both GTs had been modified. With 25 starts and 11,000 hours on the first modified unit, they both are “looking good.” No additional severe cracking has been observed.

The two-piece cylinder design issues stem from differential thermal expansion between the inner diffuser piece and the outer case. The issues range from common strut shield cracking to complete liberation of the load plates. Other repair options offered included new flanged load collars and replacing load plates with a new material, but neither adequately addresses the fundamental thermal expansion issue.

The modification at Ontelaunee is essentially a Hastelloy X collar overlay onto the original strut shield. The collar allows for controlled growth while still supporting the outer diffuser. Said another way, the outer diffuser can expand independently of the inner diffuser and outer casing.

Get the slides! Readers interested in knowing more are urged to visit the 501F Users Group website at http://501f.users-groups.com. The slides include diagrams with rich detail of blade comparisons, 3-D graphics of the modified components, before and after performance graphs and tables, and much more.

Turbine blade, vane cooling—a primer

By Team-CCJ | April 19, 2022 | 0 Comments

The most efficient powerplants ever produced are now entering service. These latest and largest combined cycles—powered by GE’s 7HA, Siemens’ SGT6-8000H, and Mitsubishi Hitachi’s M501J engines—are all clocking in at 62% – 63% thermal efficiency. This makes them the most efficient heat engines yet perfected by engineers, with all three OEMs striving to reach 65% in coming years.

The gas turbines are state-of-the-art, says Lee S Langston, professor emeritus, UConn, an ASME Life Fellow who joined Pratt & Whitney Aircraft as a research engineer after receiving his PhD from Stanford University in 1964. They make extensive use of turbine blade and vane cooling, thereby allowing the high turbine inlet temperatures required to achieve record-breaking efficiencies. Cooling air drawn from the compressor gas path (as much as 20%) is used to protect hot section parts in both combustors and turbines.

Turbine cooling details. Thermal efficiency increases with the temperature of the gas exiting the combustor and entering the turbine—the work-producing component. Turbine inlet temperatures for modern high-performance commercial jet engines can reach 3000F, while gas turbines in electric-power service typically operate at 2700F or lower and military jets in the neighborhood of 3600F. (The turbine designer must accommodate for excursions above these nominal temperatures, because of combustor hot streaks, etc.)

In the highest-temperature regions of the turbine, special high-melting-point nickel-base-alloy cast blades and vanes are used because of their ability to retain strength and resist hot corrosion at extreme temperatures. These so-called superalloys, when conventionally vacuum cast, soften and melt at temperatures between about 2200F and 2500F.

This means blades and vanes closest to the combustor can be operating at gas-path temperatures far exceeding their melting points. To endure temperatures of 500F to 1400F above melting, they must be cooled to acceptable service temperature (typically eight- to nine-tenths of their lower melting point) to maintain integrity (Fig 1).

Thus, turbine airfoils subjected to the hottest gas flows take the form of elaborate superalloy investment castings to accommodate the intricate internal passages and surface hole patterns necessary to channel and direct cooling air within and over exterior surfaces of the superalloy airfoil structure. By turbine design conventions, internal airfoil cooling is usually termed convective cooling, while the protective effect of cooling air over external airfoil surfaces is called film cooling.

Turbine cooling guide. The cooling of turbine blades and vanes goes back to the origin of the gas turbine in early 1940s. During WWII, Germany, faced with increasingly difficult times importing strategic materials such as nickel, used turbine blade and vane cooling extensively in its jet engines. Since then, turbine cooling technology and practice has become a very large segment of the gas turbine world.

In June 2017 at ASME’s Turbo Expo in Charlotte, the biennial International Gas Turbine Institute Scholar Lecture was given by Dr Ronald Bunker. A past IGTI chair, the recently retired GE gas turbine heat-transfer expert, presented his scholar paper, “Evolution of Turbine Cooling.”

Bunker’s paper can now serve as an up-to-date overview of turbine cooling, complete with a listing of 123 references (to order “Evolution of Turbine Cooling” online, visit www.asme.org/events/turbo-expo). His 26-page paper treats the evolution of turbine cooling in three broad aspects, including background development, the current state-of-the-art, and prospects for the future. This is indeed a seminal work by an expert, reflecting his direct research and OEM design experience over a period of several decades.

The author posits that the fundamental aim of a turbine heat-transfer designer is to obtain the highest overall cooling effectiveness for a blade or vane, with the lowest possible penalty on thermodynamic performance. In Fig 2, this is shown in the form of notional (that is, expressing a notion) cooling-technology curves.

On the Fig 2 ordinate, the cooling effectiveness of a turbine blade or vane is made up of its bulk metal temperature (Tm), the hot-gas-path temperature (Tg), and the coolant fluid temperature (Tc). Note that a value of 1.0 would represent “perfect” cooling.

The Fig 2 abscissa is the heat-load parameter, WcCp/UAg, where U is an overall hot-gas-path convective and radiation heat-transfer coefficient, Ag the external surface area, Wc the coolant flow rate, and Cp the thermal-capacity coefficient of the coolant.

Bunker points out that in the last 50 years, advances have led to an overall increase in turbine and vane cooling effectiveness from 0.1 to 0.7, as shown in Fig 2. It started with convection only (for example, the convectively cooled turbine airfoils of the German jet engines of WWII), and has progressed with film cooling, thermal barrier coatings (TBCs), and new materials and architectures—for example, the directionally solidified and single crystal turbine blades, which entered service in the 1970s – 1990s.

Fig 3 shows five conventional investment-casting cooling geometries in use today. They range from convection only (that is, internal-passage heat transfer only), to film cooling, and the combination of both. Bunker’s paper offers a detailed discussion of each. Typical state-of-the-art cooling schemes for high-pressure-turbine vanes and blades are illustrated in Figs 4 and 5, respectively.

Outlook. Turbine blade and vane cooling play a key role in making gas turbines for electric-power service more efficient and reliable. These cooling technologies, especially film cooling, will continue to advance gas turbine efficiency and life in the future.

GENERATORS: Lessons learned, best practices shared at GUG 2017

By Team-CCJ | April 19, 2022 | 0 Comments

There are nearly as many electric generators at fossil-fired and nuclear powerplants as there are gas and steam turbines combined. So, you might think that with thousands of utility/large industrial-grade generators rated more than 25 MW operating in the US alone, the annual meeting of the Generator Users Group (GUG) would draw hundreds of attendees. Not true: There were just 50 owner/operators at the 2017 conference in Phoenix, August 27-30.

This is understandable. Relatively few electrical engineers have permanent offices at generating stations in this country today. Many of the leading generator experts employed by utilities and independent power producers are located at headquarters and have responsibility for multiple machines fleet-wide—some scores, at least one the editors know has more than 200!

Given the state-of-the-art in generator diagnostic instrumentation and data analysis, which has not yet come up to the level of that available for gas and steam turbines, owner/operators might consider it worthwhile assigning someone on-staff to monitor these electrical machines and report suspicious behaviors to the corporate generator engineer. It’s hard to believe this wouldn’t improve unit availability/reliability and reduce maintenance costs over the long term.

But assigning someone at your plant to keep tabs on the generators is only a small first step. That person must motivated to allocate company and personal time and become sufficiently knowledgeable to contribute to the owner/operator’s goal of continuous improvement. Attending the 2018 meeting of the Generator Users Group, in Louisville, August 27-30, is a good starting point.

That generators are “taken for granted” by the majority of plant personnel should not surprise. One of the reasons for this attitude is that staff often is not aware of the many things that can go wrong with electrical machines, how to identify problems, and what solutions are available to mitigate/correct issues.

CCJ ONsite’s coverage of GUG 2017 highlights, which begins below and will continue in at least one more issue of this electronic publication, offers an opportunity to see the many problems generator engineers face regularly and why the opportunity to meet with other experts (users, OEMs, third-party solutions providers, and consultants) regularly is so important. If you have had relatively little exposure to generators over the years you may get “lost” in the technology. But don’t get frustrated if you don’t understand; just look at the many photos provided and begin to build a vocabulary of new words/phrases you might want to better understand as time allows.

Coverage of the 2017 presentations and discussions is divided into these five sections:

    • Stator frames and magnetic cores.
    • Stator windings and bus systems.
    • Fields and excitation systems.
    • Operation and monitoring.
    • General topics.

Summaries of the 10 presentations contained in the first two subject areas are below. They were prepared by IEEE Fellow Clyde V Maughan, president, Maughan Generator Consultants, who supplied the muscle to get the GUG off the ground in late 2015. Users wanting to dig deeper into any presentation can access the PowerPoint in the Power Users library. Note that Power Users Group is the umbrella organization serving the generator, steam turbine, combined cycle7F, and controls users groups.

Members of the GUG steering committee, all of whom have been with the volunteer organization since startup, are the following:

2018 Chair: Ryan Harrison PEng, lead generator engineer, ATCO Power (Canada).
2018 Vice Chair: Dave Fischli, generator program manager, Duke Energy.
Immediate Past Chair: Kent N Smith, manager of generator engineering, Duke Energy.
John Demcko, lead excitation engineer, Arizona Public Service Co.
Joe Riebau, senior manager of electrical engineering and NERC, Exelon Power.
Jagadeesh Srirama, generator engineer, NV Energy.

Stator frames and magnetic cores

    • ELCID trending, analysis. Ryan Harrison, PEng, ATCO Power
    • Electromagnetic signal analysis. Kent Smith, Duke Energy
    • Fiberoptic temperature measurement for continuous monitoring. Craig Spencer, Calpine Corp

ELCID trending, analysis

Electromagnetic Core Imperfection Detection (ELCID) is a low-excitation test with wide industry acceptance for assessing core health. Insulation breakdown causes fault currents to be set up as illustrated in Fig A1. A Chattock potentiometer (Fig A2) is used to measure the magnetic potential difference resulting from this current, with the somewhat complex equipment and circuit illustrated in Fig A3.

Bear in mind that fault currents create hot spots which can cause further deterioration to the core. If left unchecked, they can lead to damage of the stator core, windings, and the machine as a whole.

There are several setup challenges important for you to consider during analysis and trending of ELCID test results, ATCO Power’s Ryan Harrison, the 2018 chairman of the Generator Users Group, told attendees—including the following:

Core length. Depending on the operator and OEM versus non-OEM, different core lengths often are used. This leads to scaling issues in the traces and makes exact positioning a challenge.

Polarity. The orientation of the Chattock coil, and the orientation of reference transducer can lead to inversion of the quadrature signal.

Slot numbering. Decide whether to number the slots clockwise or counter-clockwise, and which slot you select as Slot No. 1.

Trending areas of interest. Results are often standalone and on various scaling in the report. This makes assessment of areas of interest more difficult and, in some cases, more judgement-based. In addition, the digital files, which have valuable information such as phase current, often are not retained by the site/tester.

Software. The owner doesn’t necessarily have the software to read the digital file. Furthermore the software is needed to export the values to a usable format. But only software “*.csv” is available; it is free to download.

Filtering. The raw data files have noise, and filtering is applied to the final reported results. While not necessarily a problem it can be valuable to look at the raw data for which the original data file is needed.

Duration between tests. Depending on the machine, the duration can be quite long. Results can be lost over time which is important for establishing baseline values and fault tracking.

Several trending challenges also were discussed and illustrated by Harrison—for example, filtered versus unfiltered, noise, inversion of signal, alignment, vertical scaling, and vertical offset. Digital data can alleviate some of the challenges associated with trending; however, there is not a commercial solution available to help with this review. Harrison presented a scripted Matlab-based solution with alternative graphic representations.

Electromagnetic signal analysis

The signal capture equipment employed by Duke Energy until 2011, Kent Smith, Duke Energy’s manager of generator engineering, told the group, was the EMC30-MKIV; today the Agilent E7402A. National Instruments’ Labview 8.5 program was used to control (standardize) data capture. Frequency of data capture on 74 generators in the fleet was annually on large steam units and so-called Tier 1 gas turbines; “when available” on smaller GTs.

The Labview program data-capture process was in four ranges of 2000 points each, as recommended by AEP /User’s Group: Range 1, 30 – 300 kHz; Range 2, 300 kHz to 3 MHz; Range 3, 3 – 30 MHz; Range 4, 30 – 100 MHz. At the end of each range the program pauses to allow manual capture of peak signatures. The program consolidates all four ranges and displays signature. After signature is saved, the peaks of interest are demodulated, viewed on-screen, and saved to file. The data then are manually imported into Excel for reporting.

Smith, the GUG’s immediate past chairmen, then discussed two examples of data taken, aided by several slides showing signal analyses:

    • Crystal River Unit 4. It was taken offline for testing in 2005. Results: Passed the Hydraulic Integrity Test (HIT); B phase megger was low (<500 Meg). Online testing in 2005 revealed a small amount of higher-frequency noise. In 2010, the electromagnetic signal analysis (EMSA) signature revealed a “hump” in mid-frequency range and had more of the higher-frequency noise. Work performed in 2010 included rewedging, the rotor mod recommended in TIL 1292, “Generator Rotor Dovetail Cracking,” and repair of five major clip leaks. In addition, the isophase ground flex link at the generator bushings was found pitted and overheated; the link was replaced. A stator rewind was scheduled for 2011.
    • Crystal River Unit 5. Online testing in 2010 found high EMSA signatures in the low and mid frequencies. Shaft voltage readings were extremely high (30 V DC) and EMI sniffer was screaming at the exciter end. Later in the year, offline testing found hydrogen seal grounded, the HIT Skid test passed, and B phase resistance was 700 Meg with PI < 2.

The hydrogen seal was coked and pitted (electrolysis). A new hydrogen seal was installed but personnel could not get the new hydrogen seal insulation package to have a very good megger; resistance was acceptable, but suspect. Sniffer readings were still high. A stator rewind was scheduled in 2012.

Future plans include having baselines on all generators (even those for small GTs), developing database on failure mechanisms with signature data, expanding the program to include large motors, developing a continuous online monitor ported to PI for Tier 1 generators.

Fiberoptic temperature measurement for continuous monitoring

While generator core failures aren’t common, their potential impact is up to the catastrophic level. Most generator cores are only indirectly monitored online through embedded RTDs situated between top and bottom stator bars at specific locations in particular core slots. These point sensors offer little protection to the large volume of the core.

Offline core testing, such as ELCID or ring/loop testing, can catch developing core issues, but both tests offer challenges in correlating measured values to actual online temperatures, and neither one offers protection from emergent issues online.

Fiberoptic temperature monitoring shows great promise in advancing core protection by permitting measurement of distributed temperatures along a length of fiber line, Calpine Corp’s Director of Outage Services Craig Spencer told the group. Working with Fiber Optic Sensors LLC and Oz Optics Ltd, Calpine installed a proof-of-concept application into one of its Hermiston (Ore) combined-cycle generators. The sensing fiber was installed on top of the stator wedges (Fig A4), though an ideal installation would be under the wedges in the base shim stock.

Once online, temperature readings from the fiberoptic line compared well against the existing RTD readings. More importantly, excellent data curves emerged which clearly demonstrated the stator zone-cooling temperature affects along the length of the fiber (Fig A5). There were some small anomalies in the data, but personnel suspected these were installation-driven variances, to be proved out in the next test case.

Overall, the results were very encouraging for developing advanced online core thermal protection, as well as for additional applications of distributed temperature sensing.

 Stator windings and bus systems

    • Stator wedge design. Edward Winegard, GE
    • Aeropac rewind. Derek Hooper, BPHASE Inc
    • Monitoring of endwinding vibration. Mladen Sasic, Iris Power
    • Connecting-ring maintenance. Keith Campbell, MD&A
    • Hot-spot detection. Jagadeesh Srirama, NV Energy
    • Importance of flex-link maintenance. Dave Fischli, Duke Energy
    • Preventive maintenance of bus-duct systems. Jesus Davila, RMS Energy

Stator design

Ed Winegard, GE’s principal engineer for armatures, opened his presentation by noting the high radial slot forces that must be contained in the stators of modern power generators—ranging from 10 to 110 pounds/inch of slot. Fortunately, he said, these forces are predominately downward, adding that about half the slots in a given stator retain bars for different phases, about half the same phase.

For slots with both bars in the same phase the force will be downward on both bars, he said. When the bars are for different phases, the force on the top bar will be slightly upward. Some type of compression system—top ripple springs, for example—is required to minimize bar movement and ensure it remains seated in the slot (Fig B1).

Bar lateral forces are minimal, Winegard continued. However, he pointed out the inherent tangential motion of the slot teeth caused by radial deflection of the core (Fig B2). This motion and its effect on the stator bar must be minimized. A similar method of ensuring the bar remains in contact with the slot wall is required and the side ripple spring (refer back to Fig B1) is ideally suited to meet this requirement.

There are also important stator wedging considerations which must be met, Winegard said: material properties (stiffness, creep, thermal aging, abrasion, etc), dimensions and tolerances (design clearances, tolerance stack up, component machining quality), assembly process (standard methods and sequence, compensation of assembly variation), and inspection (qualitative and quantitative validation).

Aeropac rewind 

The Siemens Aeropac generator discussed by Derek Hooper, president of BPHASE Inc, a small repair, inspection, and consulting company specializing in gas-turbine generators, was rewound by Alstom in 2014 because of moderate-to-severe spark erosion (Fig B3). Numerous concerns were experienced with this rewind, including the following: injection of clear resin into the dry tie material used made it difficult to determine if the cord was fully saturated (Fig B4), difficulty in obtaining proper series connection alignment (Fig B5), and use of semiconductive packing in the phase-break gaps to attenuate partial-discharge damage.

Two years later, BPHASE performed a minor inspection of the Alstom rewind. Focus was on visual inspection of the winding and evaluation of the core keybars. The keybars were intact and within torque specifications (Fig B6). While there was no evidence of keybar fracture in this unit, sister machines had suffered such fractures and plant personnel elected to reduce the keybar torque from 300 to 200 ft-lb.

There was visible evidence of discharge oxidation at the phase splits (Fig B7). There was also significant evidence of movement at the series-connector interface with the outboard ring (Fig B8).

Because of vibration concerns, it was felt that blocking should be installed between the series connections for additional support. However, this would require bump testing. The outage was too short to allow necessary disassembly and the decision was made to install series blocking in short groups to limit any effect on the global modes of the baskets.

Monitoring of endwinding vibration

Iris Power’s Mladen Sasic discussed monitoring of endwinding vibration. Although the problems associated with movement of endwindings are not new, because of changes in the design and operation of generators these issues have become a greater concern in recent years.

The endwinding region of large turbine/generator stator windings is one of the most complex parts of a generator relative to design, manufacturing, and maintenance. During normal operation, the endwindings are subject to high mechanical forces at twice power frequency because of currents in the stator bars, as well as mechanical forces transmitted via the core and bearings at rotational speed. Furthermore, during power system transients, the forces in the endwinding can reach 100 times higher than that of normal operation.

The design of the endwinding also must account for thermally induced axial expansion and contraction as the generator is loaded and unloaded. Metallic components to restrain the movement of stator bars caused by these forces normally are avoided because of the presence of high magnetic and electric fields.

Sasic shared his knowledge on the installation of vibration sensors, offline test results, and online monitoring data from a 288-MVA, 21-kV, air-cooled generator. Offline impact test data led to installation of fibreoptic endwinding vibration sensors. Continuous online monitoring of these sensors revealed an increase in vibration level, encouraging a visual inspection and bump test of the endwinding. The inspection/test confirmed loosening of endwinding support structure. Timely corrective maintenance was then possible to prevent a costly in-service failure.

Connection-ring inspection and repairs

Inspections of connection rings typically are focused on the physical support structure-to-ring interface. While there are other factors to consider—such as the connections to bars or coils and inner cooling circuity—they aren’t as common an issue and should become apparent through other testing, MD&A Generator Specialist Keith Campbell told GUG attendees.

Thorough visual inspection is vital to an accurate assessment of the overall condition of the rings. The 10 photos here illustrate typical problems associated with undesirable movement. To begin, the ties in Fig B9 offer indications of two previous repairs using a weeping epoxy. While oil intrusion was a contributing factor, the contamination (greasing) was removed well enough to allow for an adequate repair.

Fig B10 is of a phase-connection setback that had at least one prior repair attempt using the same epoxy. In this case, the contamination was under the ties and blocking, and could not be removed by cleaning alone. Fig B11 reflects an overall looseness in the endwinding structure as indicated by the large amount of greasing throughout. This was conducive to the possibility of a catastrophic failure. Fig B12 is of an original blocking and tie arrangement that does not meet quality standards.

Fig B13 shows a continuation of the previous repairs by additional application of epoxy. Fig B14 reveals ties removed for a better cleaning and application of new ties. Fig B15 is the result of excessive movement that dictated reinsulating and securing of components with a different material and by different methods than used by the OEM.

Repairs complete, Fig B16 shows the new tie after the old tie and blocking had been removed, cleaned, and new conforming material had been added. Fig B17 illustrates the areas where epoxy was applied; the Fig B18 photo was taken after repairs to the endwinding structure were completed.

Hot-spot detection

NV Energy’s Jagadeesh Srirama, a member of the GUG steering committee, profiled for attendees the recent inspection of a 391-MVA Alstom steam turbine/generator for an F-class combined cycle. This unit was put in service in 2004 and high vibrations had been recorded since installation. No issues were identified during a MAGIC (Miniature Air Gap Inspection Crawler) inspection done in 2013 and all the electrical test results were acceptable during this outage.

The unit was inspected again during a 2017 outage for simultaneous gas turbine, exhaust structure, and generator work. MAGIC identified four hot spots in the core iron and ELCID testing confirmed damage at those locations with exceptionally high readings of 1998, 1363, 674, and 976 mA. In addition to the hot spots, foreign material was found in the air gap. Management decided on immediate corrective action. To address the hot spots, it was necessary to remove the field—a challenge at this outdoor plant with major plant repairs already underway.

After rotor was removed, visual inspection identified several spots with obvious overheating similar to that in Fig B19; the debris was identified as core lamination material (Fig B20). Near the endwindings, some of the side packing had come loose and was migrating upwards into the air gap (Fig B21).

The lamination pieces came from a grossly loose tooth package, photographed in Fig B22. This tooth area was cleaned, inspected, and trimmed to make sure no more of the punchings would liberate. Mica then was placed in the shorted area and a tapered wedge inserted into the tooth to tighten the package. This wedge was epoxied in place. Note that the core step iron will have to be replaced when the stator is rewound in the future.

The side packing that came up from the top of the bar (refer back to Fig B21) was removed, air dry varnish applied, and new side packing installed. All damaged areas had a coat of red dye applied to weep into the laminations before coating with buff paint (Fig B23).

Importance of flex link maintenance

Duke Energy’s Dave Fischli, manager of generator engineering, and vice chairman of the GUG steering committee, reviewed for attendees the case history of a Westinghouse 818-MVA, 20-kV generator (COD 1981) that tripped on an A-phase neutral ground only a couple of months before the meeting. The machine’s field and stator had been rewound by Alstom in 2005.

Subsequent to the trip, a fire was reported at the lead box on the generator; site emergency responders used ABC dry chemical to extinguish the fire as the unit coasted down. External visual review showed significant damage to the A-phase lead area, with heat damage to the B-phase bushing area. Post-event data review showed some electrical anomalies starting eight minutes before trip.

Inspection revealed significant damage to the A-phase links; none of the 32 links remained intact (Fig B24). B- and C-phase links all were connected and appeared fine (Fig B25). There was a heavy layer of soot on the CTs for both the A and B phases, plus contamination at the bottom of lead box from fire-damaged components (Fig B26).

Investigators concluded that loose connections on one or more flex links caused a high-resistance contact which allowed current to flow through the bolt rather than the link contact surface area, and the bolt melted. Thus, loss of one flex link shifted its current to the remaining flex links, adding heat to them and amplifying the loose-connection problems and degraded condition of other flex links.

The A-phase links were too heavily damaged to record torques, but torque checks performed when removing links from the B and C phases were satisfactory. Some Belleville washers removed from the B and C phases had been flattened out from repeated use, others had been installed upside down (Fig B27). Flex links from the B and C phases also revealed fraying and degradation (Fig B28). It was evident that previous visual inspections had not been sufficiently rigorous to identify degraded links for replacement.

Consequential damage. The ABC dry powder extinguishing agents include chemicals such as sodium bicarbonate, potassium bicarbonate, ammonium sulfate, and ammonium phosphate. These chemicals act as a desiccant, absorbing moisture, and under humid conditions become conductive. They are alkaline in nature and corrosive to electrical insulation and metal components within the generator.

There was extensive contamination of the lead box and exciter internals by soot and smoke particles. Basic cleaning was performed of all accessible areas without complete disassembly. Follow-up inspection and full cleaning is planned for a 2018 outage.

Lessons learned:

    • Ensure work-order instructions are written correctly.
    • Ensure craft technicians are trained on the importance of assembling high-current connections properly.
    • Ensure flex links are completely removed for electrical isolation—not just unbolted on one end and bent back out of the way.
    • Ensure fire extinguishers staged around generator and other electrical equipment are CO2 or Halon (not ABC chemical).

Preventive maintenance of bus duct systems important

RMS Energy’s Jesus Davila reviewed for attendees the several types of bus systems and components: cable bus, non-segregated and isolated-phase bus, terminations and disconnect links, insulating materials, expansion joints, seal-off bushings, etc. Each of these requires special maintenance. Critical items on the bus duct include flex/bolted connections (current carrying), expansion bellows/joints, insulators and mountings, seal-off bushings, groundings, and insulated joints. Examples of some of the issues discussed by Davila were the following:

    • Electrical connections. Arc damage to bolted joint bus face (Fig B29).
    • Flexible connectors. Cracked laminations caused by vibration or air flow (Fig B30); flex braids damaged by rubbing and/or abuse (Fig B31).
    • Improper bolting or lack of maintenance at connection points. Damaged contact surfaces and gross heating issues (Fig B32).

    • Expansion-bellows damage attributed to excessive movement often resulting in cracks (Fig B33).

    • Bus failures. Overheating of non-segregated bus attributed to a lack of maintenance (Fig B34); line-to-ground fault causing melting of the bus enclosure (Fig B35).

Most of the deterioration conditions listed above can be detected, particularly in advanced stages, by visual examination and/or temperature monitoring. All require immediate attention to prevent major equipment failure. If the condition is found before failure, refurbishment usually can be accomplished by obvious and/or well-established procedures.

Checklist for success: NV Energy commissions seven turbine DCSs in seven weeks

By Team-CCJ | April 19, 2022 | 0 Comments

Commissioning seven gas-turbine (GT) control systems in seven weeks would probably be taxing under the best of conditions. In NV Energy’s case, there were some extenuating circumstances.

First, the seven systems are located at two generating facilities acquired by the utility in 2014. Second, it is challenging to test a retrofit, due to air permit emission limits. Third, NV Energy, like most utilities today, has limited central engineering staff for projects like this. Finally, the original estimates for the projects were established by the due-diligence team, but had to be carried out by engineering and operations. No field instrumentation was included in the scopes.

As background, three of the GTs, gas-fired (no backup fuel) GE 7EA peakers located at Sun Peak Generating Station (SPGS), have water injection for NOx control and minimum balance-of-plant (BOP) systems. The other four engines are at Las Vegas Generating Station (LVGS), which has three combined cycles powered by LM6000s equipped with water injection for NOx control and SPRINT compressor inlet water spray for power augmentation. Two of the combined cycles are in a 2 × 1 arrangement (installed in 2003), one is a 1 × 1 (installed in 1994). The latter was not included in the controls upgrade project.

The controls retrofit at SPGS was justified on the basis of adding remote operation; the LVGS retrofit was justified to improve and automate transitions in and out of SPRINT mode. Both projects were commissioned in seven weeks during October and November 2016, with no third-party owner’s engineer in the mix.

An interesting wrinkle, important to understanding how the projects unfolded, is that SPGS went first in the schedule. Because it required a Class A Nevada contractor’s license by NV Energy’s procurement group, the electrical contractor was lead, with Emerson, the DCS supplier, as sub. Due to a less restrictive project scope for LVGS, this requirement was relaxed such that a Nevada Class C contractor could lead, which allowed the DCS vendor to be lead, and the electrical contractor as sub. Nevertheless, this led to two separate engineering specifications for each plant.

As reported by Clint Vanderford at the Ovation Users Group Conference in July 2017 and during a follow-up call with CCJ editors, the experience offers valuable lessons—from the acquisitions phase through commissioning new equipment—for others undertaking such projects.

In reviewing the challenges explained below, it’s important to note that each plant had a different owner and both were IPPs, with design features that may not be considered “utility-grade.” Also, some of these are typical “gotchas” which occur with every project of this type and magnitude, but still suggest wise cautions for others contemplating similar work.

Carefully review all instrumentation relevant to the retrofit. Many project issues stemmed from a lack of instrumentation expertise on the acquisition due-diligence team which led to the controls retrofit project team being unaware of important instrumentation issues. All of the gas turbines here feature wet NOx control (water injection). Some of the instrumentation at SPGS is 1980s-vintage. LVGS had LM6000s equipped with newer instrumentation, but valve positioners in the Woodward control system posed problems. Plants often “live” with marginal instrumentation. The larger point here is that existing instruments will exhibit varying degrees of compatibility with a state-of-the-art DCS.

Clarify each participating group’s capabilities and experience. Each facility’s contract was managed differently. At SPGS, a local electrical contractor licensed in Nevada acted as the lead EPC contractor with Emerson as a subcontractor. At LVGS, Emerson was lead contractor with Dynalectric Nevada as sub. Lesson learned here is that the DCS supplier is a better EPC than an electrical contractor, in part because of the inherent understanding of the instrumentation. On the other hand, Emerson had little experience retrofitting LM6000 machines with Ovation; this inexperience surfaced in the areas of cabinet layouts and construction and diagnosing wiring circuits. Substantial new wiring had to be pulled to replace cabinets located at the turbine housing (the gas-turbine OEM’s design basis) to cabinets located outside the control room (the DCS vendor’s design basis).

Pay attention to the soft-hard interface. Vanderford noted that Emerson did a really good job delivering the “soft” product—that is, writing the logic and building the graphics. Wiring the existing plant components to the new DCS equipment was not so straightforward. At LVGS, there was no one who could actually install the new gear, so it had to be hired out. A third-party contractor had to make sure everything was properly wired. The project team found numerous instances of “duct tape” solutions with the existing equipment which had to be remedied to hook up the new controls. At SPGS, the ABB Bailey BOP controls and the GE gas-turbine controls were hard-wired—that is, no bus or data highway. “It was like peeling an onion,” Vanderford said, “it took a few iterations to achieve our objective of having minimum hard-wired stuff.” However, the protective circuits are still hard-wired. The project team also found numerous instances of electrical changes not properly documented.

Focus on the graphics. What the operators see on their screens helps determine how well they can run the equipment. Emerson’s Ovation team includes GT specialists, but they are typically not plant operators, and they build and deliver “standard” graphics packages. Those who specify and bid projects like these have to incorporate flexibility to modify the graphics so they work for GT plant operators, not GT engineers. Vanderford said the Ovation “standard” graphics were crowded and not well organized, and that boilers “don’t exist in the GT controls world.” “Many discussions were required with Emerson’s Clifton Park GT specialists,” he added.

Know your air permits before starting the project. It is challenging to test a retrofit like this because air quality permit limits cannot be violated. Not all of the post-installation testing that was required at SPGS would fit within the air permit emission limitations. As the regulatory agency does not issue variances, there was no way around this limitation. This made it difficult to commission the new turbine governor. This also necessitated postponing the voltage regulator replacement, because it requires one hour of operation at full speed and no load, even though the existing one is old and not easily maintained (though it is still deemed reliable). NV Energy is pursuing the permitting of a limited amount of operation at higher permit limits for the purposes of testing and tuning, which will allow the new voltage regulator to be commissioned.

Check out motor and other critical component specifications prior to testing. NV Energy experienced a starting-motor failure during the commissioning. An underrated motor had been installed before the project as a replacement, and it ended up running longer than it was designed for as part of the commissioning. The relays in the DCS were programmed to protect the load of the old motor, not the new one. An auxiliary lube-oil-pump motor also failed but Vanderford chalked this up to age: “Its time had come.”

Trust, but be in a position to verify. Because of the lack of central engineering resources and minimum plant staff, NV Energy had to place a great deal of trust in its contractors. For example, Vanderford notes, “we lacked experience in governor logic, and had to trust the DCS vendor, while holding some money back until verification that the controls would work as designed.” The electrical contractor was more of a “generalist,” though with abundant industrial facility experience, and required “coaching” by NV experts. Few on the plant staff had the requisite knowledge of the equipment to carry out a project like this. In the end, says Vanderford, success in construction relied heavily on good will and good working relationships among the team members. Each project was run by the plant, although up until 2016, such projects would have been handled by a central project management group.

Overall, Vanderford reports, there were no major failures and no major delays and the transition to normal operations went well. That’s quite an outcome when retrofitting seven gas turbines in seven weeks across two facilities with substantially different systems, equipment vintages, and former owner/operators.

GENERATORS: Lessons learned, best practices shared at GUG 2017 (Part 2)

By Team-CCJ | April 19, 2022 | 0 Comments

CCJ ONsite’s coverage of technical highlights from the Generator Users Group’s Third Annual Conference, held in Phoenix, Aug 27-30, 2017, continues below with Part 2 of the lessons learned and best practices shared among attendees. The three-part series concludes in the next issue of this electronic publication.

Presentations and discussions are arranged in these five sections:

    • Stator frames and magnetic cores.
    • Stator windings and bus systems.
    • Fields and excitation systems.
    • Operation and monitoring.
    • General topics.

Links are provided to Sections 1 and 2 in case you missed last week’s edition. Summaries of the presentations in Sections 3 and 4, which appear below, were prepared by IEEE Fellow Clyde V Maughan, president, Maughan Generator Consultants, who supplied the muscle to get the GUG off the ground in late 2015. Users wanting to dig deeper into any presentation can access the PowerPoint in the Power Users library. Note that Power Users Group is the umbrella organization serving the generator, steam turbinecombined cycle7F, and controls users groups.

Members of the GUG steering committee, all of whom have been with the volunteer organization since startup, are the following:

2018 Chair: Ryan Harrison PEng, lead generator engineer, ATCO Power (Canada).
2018 Vice Chair: Dave Fischli, generator program manager, Duke Energy.
Immediate Past Chair: Kent N Smith, manager of generator engineering, Duke Energy.
John Demcko, lead excitation engineer, Arizona Public Service Co.
Joe Riebau, senior manager of electrical engineering and NERC, Exelon Power.
Jagadeesh Srirama, generator engineer, NV Energy.

Fields and excitation systems

    • Rotor arcing and repair. Chris Keathley, MD&A
    • Collector rings: Inspection and repair. Keith Campbell, MD&A
    • An unusual generator field ground. John Demcko, PE, Arizona Public Service Co
    • Brush-holder experience. John DiSanto, GE
    • Digital excitation replacing ageing technologies. Richard Schaefer, Basler Electric Co
    • Shaft earthing monitoring. Andre Tetreault and Bernard Lemay, PEng, VibroSystM

Rotor arcing and repair

A common and destructive phenomenon in generators is negative sequence currents (I2), MD&A Project Engineer Chris Keathley told GUG 2017 attendees. These can be caused by unbalanced three-phase currents, unbalanced loads, unbalanced system faults, open phases, and asynchronous operation. The result of I2 currents may be rotor body currents that can damage the rotor forging (Fig C1), retaining rings (Fig C2), slot wedges (Fig C3), and to a lesser degree, the field winding.

Three case studies were reviewed by the speaker, who brought to MD&A 16 years of experience with a major utility as a turbine/generator engineer, and another five years with the OEM.

Case Study No. 1. A 500-MVA, 22-kV, 3600-rpm generator had been involved in a motoring event of unknown size and duration five years before the inspection described was conducted. A visual assessment revealed relatively minor problems—such as slight burning between wedges, significant burning on wedge ends, and burning of the forging at wedge junctions. Eddy-current testing revealed 290 indications. Affected areas were cleaned and blended. A TIL 1292 (“Generator Rotor Dovetail Inspection”) repair on the Coil No. 1 Slot was done and steel wedges were replaced with aluminum (except end wedges). A high-speed balance and heat run were conducted after rewind.

Case Study No. 2. A two-pole field suffered a double ground fault that caused severe arcing damage to the rotor, including melted material and cracking on a tooth (Figs C1 and C4). NDE of the area revealed a through-wall crack and an engineering evaluation determined the rotor unacceptable for operation. A temporary weld-repair solution was proposed to get the unit back in service until a replacement rotor could be obtained. The damaged portion of the tooth was removed (Fig C5) and the tooth reconfigured with weld build-up, rotor heat treatment, and re-machining of the tooth (Fig C6).

Finite-element models of the rotor and slot tooth were created to obtain the various stresses along the tooth height, the weld fusion line/HAZ, and the highly stressed wedge groove fillet radius. These stresses and mechanical properties were used in fracture-mechanics calculations; favorable results supported acceptance of the repair. A fatigue analysis of the repaired tooth suggested the reworked rotor was good for 150 start/stop cycles or 10 years of operation, whichever came first.

Case Study No. 3. A generator experienced a collector failure and ground to the main shaft that caused major arcing and heat damage to the end of the generator rotor forging (Fig C7). The amount of damage and heat-affected material made the shaft end forging unacceptable for continued operation. Stub-shaft replacement was proposed and accepted by the owner.

This was a major undertaking. The shaft was severed just outboard of the journal and bolt holes drilled and tapped for the new stub shaft (Fig C8). The assembled new shaft extension is shown in Fig C9, and the final assembly with the fan and collector rings in Fig C10.

To conclude, negative sequence events and ground faults are preventable by good operation and monitoring equipment. When those defenses fail, considerable damage can occur. However, it does not always mean that the damage cannot be repaired and the unit returned to service. These case studies show that even when there is damage, advanced welding and machining processes can restore the unit to service relatively quickly.

Collector rings: Inspection and repair

MD&A’s Keith Campbell, an industry veteran with four decades of experience, showed and discussed numerous photographs to illustrate the inspection, failure modes, and repair of collector rings.

Inspection. Typical conditions found during inspection are shown in photos C11-C15. Threading and grooving (Figs C11 and C12) naturally result from the wear and tear of long service. Photography (Fig C13) is not common nor well understood; it describes the phenomenon in which the pattern of the brush holders is replicated on the rings. There is inevitable contamination associated with the collector (Fig C14) from sources that include brush wear, induction of foreign material with the cooling-air flow, and uncorrected arcing. Massive contamination is caused by flashover failure (Fig C15).

Failure. Typical problems resulting from failure are seen in the photos C16-C19. In the first, the right ring was severely burned by a flashover to ground or severe arcing from poor maintenance. The adjacent image shows a similar condition but with the brush holder removed. Fig C18 is of damage to the main shaft from severe arcing to ground, C19 shows corresponding arc damage to the ring inside diameter.

Repairs. Figs C20-C22 illustrate steps in replacing an old collector with a new collector, Fig C23 is of a grinding operation for truing a worn, or new, ring.

An unusual generator field ground

The incident profiled here by John Demcko, PE, a senior consulting engineer at Arizona Public Service Co and member of the GUG steering committee, occurred at an APS plant equipped with three single-shaft combined cycles installed in the mid-1970s.

The generators serving these units are rated 146.7 MVA/13.8 kV; they were retrofitted with static excitation systems and redundant digital regulators several years ago. This update included a modern 64F field ground detection system, which experienced a continuous field ground alarm over a year ago. The incident was treated routinely—that is, management was informed.

Management was made aware of the risks associated in running with a known ground. The decision was made to remain online until an outage could be scheduled to evaluate the situation.  When that happened, a Megger test of all components in the field winding circuits showed no ground.

The 64F relay is only operative when there is excitation on the machine since it is powered from the excitation power potential transformer. With the unit offline and not spinning the 64F was powered up with a “cheater” cord. No field ground was detected by the relay in this configuration. The 64F relay was switched out for an identical spare which also indicated no ground with the unit offline but did indicate one with it spinning and the field energized.

This implied that the ground was on the rotor and was caused by the centrifugal loading of the field winding. The unit was put back in service and was run almost daily with the apparent field ground while a new 64F relay was ordered from another manufacturer. The new 64F was installed and behaved exactly as the previous two relays.

A last-ditch effort was made to spin the unit up to synchronous speed with excitation off. Meggering of the field via the carbon brushes and slip rings found the resistance to be in the megaohms range. There appeared to be an impasse when a ground was indicated in the field circuit but could not be localized.

After re-verifying the accuracy of all data taken, an investigation was conducted to look for any other elements that can fault to ground when the exciter is running. Several manufacturers of field ground detection equipment were asked if their equipment could detect faults on the AC side of the static excitation system. Responses were mixed, but one vendor said AC-side grounds can definitely be detected, although their occurrence is highly unusual.

Additional voltage-to-ground measurements were made on the AC side of the static excitation system breaker for all three units. A fundamental difference in readings was noted on Unit 2 as compared to Units 1 and 3 which did not have field grounds. This inferred an AC side excitation system ground was causing the 64F relay to indicate a field ground. A precise physical model of the static excitation system, including the same 64F, was constructed. It confirmed that an AC ground is easily detected by the 64F and the lab measurements very closely matched the data taken of the actual machine.

The unit will be operated until a scheduled outage allows for AC fault confirmation and repairs.

Brush-holder experience

John DiSanto’s presentation focused on collector incidents and avoidance. The GE senior engineer, who is responsible for generator controls/excitation and protection fleet-wide with the goal of improving equipment reliability, reviewed nine recent root-cause-analysis investigations.

They involved collector flashovers, collector-ring overheating, improper collector-ring assembly, and a damaged insulating sleeve.

Two of the case studies illustrated unnecessary forced outages. One involved overheating of the inner collector-ring surfaces, molten brush holders, and arcing of collector housing. The second case revealed a heavily worn brush and brush-holder damage. The collector surface was found to have salt deposits and corrosion.

Daily, weekly, monthly, and outage inspection and maintenance were reviewed. DiSanto also offered a few comments on maintenance—for example, the importance of cleaning, collector-ring wear rates (1 mil per 1000 hours of operation is typical); permissible wear (the ring diameter can be reduced but must always be larger than the diameter of the bottom of the spiral groove); maintaining proper cooling on collector rings (design temperature is 40C); and recommended brush grade (National 634).

The presentation closed with a discussion of brush-rigging upgrades. GE was said to have a well-defined set of brush-holder design criteria, including the following:

    • Allow for safe installation and removal of brush holders with the generator online.
    • Improve ease of brush and spring replacement—no tools required.
    • Eliminate brush “hang-ups” within the holder by having a fully supported brush and a smooth/slick brush pocket.
    • Decrease the risk of flashover.
    • Decrease susceptibility to brush current selectivity—that is, uneven current distribution among brushes.
    • Allow easy integration for both servicing and replacement across all GE generator models.

The company’s most recent brush-holder design for generators of moderate size (Fig C24) was said to meet these criteria.

Digital excitation replacing ageing technologies

If you were relatively new to the industry, listening to the presentation by Basler Electric Co’s Richard Schaefer at GUG 2017 provided valuable perspective on excitation systems. A former chair of the IEEE PES Working Group, he’s “seen it all” in more than four decades of service to power producers. His CliffsNotes on the evolution of excitation systems:

    • Before the 1950s, rheostats (Fig C25).
    • 1950s and ‘60s, amplidyne.
    • 1970s and ‘80s, magnetic/analog.
    • Late 1980s through today, digital.

Several factors affect how quickly systems become obsolete, he said—including, available materials, present technology, and available software. Examples of materials availability issues: carbon composition can no longer be manufactured, warlords have taken over the mines, material determined to be hazardous. Examples of hardware availability issues: primitive computers, present-day powerful computers.

Schaefer note actions that have contributed to the slowing of obsolescence—for example, purchase of components from manufactures serving major long-survival industries (such as automotive), redesign of product with new components where practical, purchase large volumes of obsolete components.

Particularly in recent years, there has been upward evolution in software. For example multilingual capability, built-in powerful testing tools, enhancements to speed commissioning. Also there are options to facilitate retrofit—for example, replacement product fits into same location, provide replacement excitation cabinets but keep the power potential transformers, keep SCR bridges and provide new front-end electronics.

Shaft earthing monitoring

Andre Tetreault, director of tests and diagnostics at VibroSystM, and co-author Bernard Lemay, PEng, the company’s Zoom analytical software expert, opened their presentation by reminding the GUG 2017 audience that, in the ideal, the generator shaft should be electrically and magnetically neutral during operation. The use of ferromagnetic materials in the construction of the shaft makes this component extremely conductive and subject to current flow and induced voltages. A typical turbine/generator configuration is shown in Fig C26; the exciter provides DC current for the field, the turbine provides mechanical power.

At least one generator bearing is insulated from ground. On other bearings, a thin oil film is the only barrier separating the shaft from the ground and this film may not act as insulation to current flow.

Both AC and DC voltages can be induced in the shaft, causing potential damage to the unit— especially the bearings. Shaft earthing monitoring systems can record and identify the various sources of voltage and current, allowing for analysis and damage prevention. One brush is installed to monitor voltage (diagnostics), usually on the generator exciter end, and one to monitor current (protection), usually on the turbine end.

Results from the voltage brush are shown in Fig C27. Time, waveform, spectral content, in addition to DC and RMS values, are available for analysis. Alarm levels are set after results are analyzed.

Protection schemes using trends of shaft currents are common, but do not provide diagnostics. In many cases, alarms triggered will confirm existing damage, instead of detecting ongoing issues and/or preventing future damage. Voltage frequency profiles should not evolve over time.

High-speed data acquisition, and comprehensive diagnostics software, allow for in-depth analysis of harmonics to identify various anomalies. Patterns such as high even harmonics may indicate rotor alignment or stator bar problems. If deviations are present, other symptoms may include stator vibration and/or bearing temperature variations.

High DC levels should trigger a ground connection investigation and if the shaft is magnetized, bearing temperatures should be verified.

Operation and monitoring

    • Mini turbine/generator model for training. John Demcko, PE, Arizona Public Service Co
    • Generator abnormal operation. Ron Halpern, Generator Consulting Services
    • Effects of negative-sequence and off-frequency currents. Izzy Kerszenbaum, IzzyTech
    • Impact of cycling duty. Jamie Clark, AGT Services Inc
    • Generator cyclic duty. Ed Winegard, GE

Mini turbine/generator model for training

John Demcko, PE, a senior consulting engineer at Arizona Public Service Co and member of the GUG steering committee, spoke to the industry’s knowledge gap and work his company was doing to bridge that gap. Many new powerplant engineers and technicians, he said, have no formal training on how to bring generating units online and take them offline. Nor are they usually familiar with how generators interact with the grid.

The 40-plus-year industry veteran added that plant engineers and technicians are good at addressing mechanical issues but typically are not well versed in the electrical characteristics of synchronous machines. A PowerPoint-only class had been offered for many years at APS but personnel were not being prepared adequately for the challenges they faced daily.

Demcko’s department of four engineers functions as a “consulting firm” internal to the utility. Over the years, electrical training responsibilities defaulted to his group because it was intimately involved with problems on the generator, exciter, automatic voltage regulator, and power-system stabilizer. Since operating staff tend to be “visual learners,” he continued, a physical model-based training approach was proposed to management. It was accepted and a wish list of model functionality was developed. The search began for a commercially available training model.

The “wish list” of features desired for the model included the following:

    • Three-phase synchronous generator.
    • Power level compatible with a nominal 5-amp CT output and 120-Vac PT output found in utility generation.
    • Utility-style control switches and metering.
    • Digital generation protection relay.
    • Auto-synchronizer and emulation of generator breaker function.
    • Drive motor with the ability to mimic turbine characteristics.
    • Synch to grid at either 208 or 480 Vac.
    • Have the look and feel of a typical physical generator control panel.
    • Portability (movability).

Search results. North American universities were canvased and educational suppliers queried. Nothing commercially available met all of APS’s requirements at any price. Conclusion: The utility would have to build a custom model.

Custom build. Team Demcko received approval to build its own model with a modest hardware budget. Engineering and labor was to be done by staff on a time-available basis—such as between regular assignments. Cost of the model would be contained by purchasing/scrounging as many components as possible off-the-shelf. A target of one year was established for completion of the model.

Results/observations. The model, highly modular in construction, took two years to complete because of limited availability of labor (Figs D1 and D2). It was good for troubleshooting but had many terminal blocks for interconnection that could loosen. Classes started in April 2017; 102 APS employees were trained. The first four weeks of onsite plant training was completed near the end of August with 84 attendees. Remainder of the fossil fleet (several hundred “students”) was scheduled for onsite training before yearend. Results: Almost all the design targets were achieved, but some refinements to the model were suggested.

One use of the model that Team Demcko did not fully anticipate was for training plant operators in manually closing the generator breaker to synch the unit. While an auto-synchronizer normally is relied on for this task, APS expects its control operators to have the ability to do it manually. Trainees appreciated being able to practice synchronizing the generator with the electric system using the model, without risking damage to critical equipment.

Generator abnormal operation 

Ron Halpern of Generator Consulting Services opened his presentation by defining “abnormal operation” as any operation outside normal operating parameters that could damage the generator—such as operation outside of the generator capability curve.

The speaker, who has been involved with generators for well over 40 years, 25 of those at GE, focused his presentation on the following:

    • Stator—including core, oil, hydrogen leaks, grounds, stator cooling, water leaks and restrictions, bushings, and frame.
    • Rotor—including grounds, shorted turns, thermal sensitivity, shaft voltage, and collectors.
    • Auxiliaries—including loss of hydrogen seals, coolers out of service, and on moisture corrosion and contamination.
    • Electrical and grid—including over fluxing, off-frequency operation, loss of synchronization, motoring.

Typical abnormal-operation events discussed included the following:

    • Core failures. They may be caused by foreign object damage, lamination insulation failure (Fig D3), damage from repair work (Fig D4), loose core (Fig D5), etc.

    • Core over-flux, a complex phenomenon. Protection is via volts-per-hertz relay. Minor over-fluxing (105%-110%) increases core losses and elevates core temperature but should cause no damage. Over-fluxing above 110% saturates portions of the core to the point that flux flows out into adjacent structures and, if sufficient and sustained, may cause total core destruction (Fig D6).

    • Rotor ground. The excitation system is ungrounded and a single ground will not cause damage (unless the cause is a broken conductor or coil short). However, a second ground can be disastrous. There are many possible causes—including ground-wall insulation breakdown, contamination, electrical arcing, displaced insulation, and water intrusion into the exciter. Collectors are the most significant contributor to operations-caused forced outages on generators; the results can be dramatic and dangerous (Fig D7).
    • Rotor turn/coil shorts. Shorts may not be a problem unless excessive and you run out of current, or if they result in high thermal-sensitivity vibration. But they can be destructive (Figs D8-D10).

      • Thermal sensitivity can be problematic. It causes rotor vibration to change as the field current is increased and can cause rotor bowing when (1) the temperature distribution is uneven circumferentially around the rotor and/or (2) axial forces are not distributed uniformly in the circumferential direction. The phenomenon, characterized by a once-per-revolution frequency response, may limit operation at high field currents or VAR loads because of excessive rotor vibration.
      • Shorted-turn detection. The most reliable method for detecting shorts is by use of a flux probe. The technology is well understood and reliable.

    Other items briefly discussed included oil in the generator, stator-bar slot support systems, high- voltage bushing, seal leaks, noise causes and investigations, and damage prevention in general. 

    Effects of negative-sequence and off-frequency currents

    From early on, AC synchronous generators were designed to produce three-phase balanced voltages at their terminals, began Dr Izzy Kerszenbaum, PE, of IzzyTech. Over time, the design also incorporated features to reduce the harmonic content of the generated voltage. In the case of generators, the problem was (and still is) mainly related to unbalance in the load currents, while in the case of AC motors, the problem was (and still is) related to unbalanced supply voltage.

    The negative-sequence current component circulating in the stator windings creates a magnetic flux in the airgap of the machine, continued Kerszenbaum, a well-respected teacher of things electrical and prolific author with more than 40 years of service to the industry. This flux rotates at synchronous speed, but in the direction opposite to the positive flux (the “normal” flux), he explained.

    The rotor, also rotating in synchronous speed in tandem with the positive magnetic flux, is subject to a 2× synchronous frequency magnetic flux, by the negative flux. Then, by the law of electromagnetic induction (Faraday), 2× synchronous frequency voltages and eddy currents are induced in the rotor body. Given that these induced currents have a periodicity of 120 Hz in 60-Hz systems or 100 Hz in 50-Hz systems, they tend to flow mainly in the outer regions of the rotor, because of the “skin effect.”

    Net result: If large enough, the induced currents will spark and arc between wedges, wedges and forging, wedges and retaining rings, forging and retaining rings, and any component on the periphery of the rotor. Such sparking/arcing can cause hardening of the metal in critical areas, followed by the generation of cracks.

    From the foregoing, it is obvious that negative-sequence current carries with it the potential to cause significant damage to the generator; thus, protection against these currents must be provided. In the event a large negative-sequence event occurs, (as with a major short-circuit between phases in the vicinity of the machine), it behooves the operators to carry out an assessment of the possible damage incurred by the machine, followed by a proper inspection, if warranted.

    Impacts of cycling duty

    Generators built during the gas-turbine order/installation “bubble” in the late 1990s and early 2000s, look very much like their predecessors built in the 1950s, 1960s, and 1970s. However, unlike their predecessors, the newer machines are not giving the 20 to 30, or more, years of reliable service expected.

    OEMs have designed similarly sized machines for MVA ratings 40% to 50% higher than their predecessors, while pushing material capabilities to their maximum. Plus, demands on equipment have been exacerbated by the need to cycle these generators hundreds of times annually to accommodate must-take renewables.

    Generators were designed to run at base load or, worst case, for minimal annual start/stop counts—perhaps 50 to 75. However, as the charts in Figs D11 and D12 for two case histories show, they are seeing 250+ starts per year.  Units in renewables-heavy markets are exceeding 350-400 annual starts. This takes low-cycle stresses from thermal expansion/contraction, and moves it into a high-cycle realm.  The end result is that units are either suffering in-service failure or, at a minimum, are requiring very costly repairs or maintenance/upgrades at their first major outages, within 10 to 15 years.

    In his presentation, AGT Services Inc’s Jamie Clark pointed to common weaknesses exacerbated by these high cycling operations—including loose stator wedge systems (Fig D13), axially loose core iron, loose endwindings, global endwinding dusting or broken ties, loose belly bands, bar movement in the stator slots, high partial discharge and resulting corona damage, and increased opportunities for seal-oil problems resulting in oil entering the unit, which further accelerates the previous issues.

    In the field, the impact is found in cracked or failed main leads, pole/pole and coil/coil crossover jumpers, migration of slot armor, deformation of field endwindings, loose/missing/broken distance blocking, migrating coils, insulation, or amortisseur springs resulting in blocked cooling, thermally sensitive fields, rapid turn short development, and myriad other issues (Fig D14).  

    Generator cyclic duty

    In recent years there has been a changing of the generator lifecycle. These machines originally were intended for baseload operation and 30 years of service. There has been an industry shift to frequent starting/stopping, load cycling (described as more than two 20% changes in megawatt output in a 24-hr period with two primary load cycles (50% – 100%) in a typical day), VAR cycling, and seasonal influences.

    Frequent starting/stopping imposes additional stress, with faster degradation of insulation and components, negative impact on generator life, higher risk of in-service operating incidents—all likely contributing to increased maintenance.

    Cyclic duty involves start/stop operation, load cycles, and power-factor changes. Impact on stators includes vibration transients, thermal and mechanical stresses, and core-end heating. Some of the effects on stator windings and core are high- and low-cycle fatigue, insulation abrasion, strain, shorts or grounds, localized overheating, and core-iron melting. Typical failures are strand cracking and fracture (Fig D15), lead fracture and extensive arc damage (Fig D16), and insulation abrasion (Fig D17).

    Cyclic-duty impact on rotors includes copper distortion, insulation breakdown, shorted turns, connector failures, grounds and forging damage. Typical resulting failures are shown in Figs D18-D21: slot liner abrasion, insulation fracture, copper distortion, and blocked vent (left to right).

    Twenty-five-year GE veteran Ed Winegard, currently principal engineer for armatures, described for attendees several design features developed to accommodate cyclic operation. You can access a copy of Winegard’s PowerPoint on the Power Users website.

    Maintenance and inspection suggestions for cyclic duty, also covered in the presentation, include the following:

      • Maintain equipment in accordance with GEK 103566.
      • Conduct additional testing during scheduled outages.
      • Perform regular borescope and robotic inspections.
      • Do modal testing of endwindings.
      • Provide for additional monitoring during operation.

Group-sponsored HRSG development projects share results, reduce costs

By Team-CCJ | April 19, 2022 | 0 Comments

European Technology Development Ltd conducts multi-client projects to address concerns and challenges within the international electric-power industry. Known as ETD Consulting, the 20-year-old UK company’s projects are aimed mainly at developing new tools, technologies, and methodologies for powerplant inspection, integrity, and life assessment/extension; and improvements to materials, welding, and design.

Collaboration with the international power industry is key to developing new tools and cutting costs while achieving useful scientific results. This approach pools both financial and technical resources, technical know-how, and lessons-learned databases. Among the company’s current group-sponsored projects (GSPs) of interest to CCJ ONsite readers are these:

    • Aberrant P91 long-term creep rupture data collection, inspection, and repair.
    • Crack assessments in boilers/HRSGs and turbines.

Closer to home for many readers, note that these are among the topics included in the robust presentation/discussion agenda for the HRSG Forum with Bob Anderson, March 5 – 7 at the Hyatt Regency Houston (see sidebar below).

ETD concentrates heavily on the integrity and life assessment of plant systems and components based on disciplines of mechanical engineering, materials science and metallurgy, and inspection and maintenance in high-temperature applications—somewhat similar to what US-based EPRI and Structural Integrity Associates Inc do as well. It specializes in probabilistic crack and life assessments, which require large databases.

According to Dr Ahmed Shibli, CEO, “ETD has expertise in plant services and technical consulting backed by R&D conducted with manufacturers, plant operators, service providers, and researchers from around the world. We are in an ideal position to conduct such projects for industry sponsors, and bring together plant experience from best-run plants in Europe, Japan, US, Canada, Australia, and elsewhere.” A few examples of GSPs follow.

Creep rupture strength of P91 materials and welds, a six-year project, is generating long-term creep rupture data to help plant operators establish safe operating lives of aberrant or abnormal P91 base- and weld-metal microstructures often found in powerplants and HRSGs. The aberrant materials and weldments arise when heat treatment during either steel production or component fabrication is not conducted to the precise requirements for P91—a high-chrome martensitic steel. With incorrect microstructure, creep rupture strength is reduced, leading to early failure.

Given the absence today of long-term rupture strength data for aberrant microstructures, some plant owners and operators are not sure how to deal with such components and often treat them as either P22 or P9 steels, sometimes unnecessarily condemning critical and costly components too early, at great expense to owner/operators (Fig 1).

Within this project, tests are being conducted on 15 aberrant microstructures at durations of up to 30,000 hours. Current sponsors include organizations from Europe and Japan.

As background, ETD plant experience includes NDT of boiler P91 welds at the 460-MW Quezon Power Plant in the Philippines; P91 work for EdF in France, Inpex in Australia, Malakoff and TNB in Malaysia; plus work at Uch Power Plant in Pakistan and other facilities in the Middle East, Europe, and Far East. Also significant is an ETD Grade 91 Users Group initiative for the supercritical Keephills Unit 3 in Canada.

Although not part of the original Abnormal P91 GSP scope, the work on this project also is helping to establish new inspection regimes for both P91 and P92 components.

Inspection techniques and life assessment for P91 and P92 welded in-service components is a closely aligned GSP with a three-year duration. According to Shibli, “Welded-pipe creep tests are being performed and stopped at 30%, 50%, and 70% of estimated life to enable an assessment of new inspection techniques and study of damage development with lifetime consumption.”

Inspection techniques include (1) replication, (2) hardness, (3) innovative UT, (4) boat sampling using a newly developed portable spark erosion machine, (5) portable scanning force microscopy (SFM) to detect early-life damage and the creation of 3-D creep-cavity images (Figs 2 and 3) using cavity volume for life assessment, (6) electromagnetic property measurements, and (7) AC- and DC-potential-drop techniques.

The aim of this GSP is to develop and experiment with innovative techniques that can detect and quantify creep cavitation damage in P91 and P92 steels at an early-life stage. At present this is difficult because traditional NDE techniques are inadequate for the task.

Three large P91 pressure vessels and one P92 vessel, with both butt and seam welds, are being tested under pressure and high temperature—and with end loads to obtain in-service-like Type IV failures in butt-welded pipes at a planned duration of about 10,000 to 15,000 hours.

States Shibli, “Tests have now been completed along with supporting cross-weld rupture data and finite-element analysis. Relationships between damage in the Type IV or fine-grained region of the heat-affected zone (HAZ) and the remaining life of the welded components are developed to help plant operators detect and quantify early-stage damage. This will give them time to develop repair or replacement strategies before potential and possibly abrupt failure takes place.”

Participants are from the UK, US, Japan, Germany, Italy, and Belgium.

A similar project has been planned using P91 pipes containing aberrant microstructures to further demonstrate the above NDE techniques together with a few new and promising methods. This GSP will help in developing inspection strategies for components with aberrant microstructures. In addition, it will demonstrate a Japanese-developed technique for strengthening creep-damaged pipe for which ETD has the rights for further development. Existing sponsors are from the UK, US, Europe. and Japan; new sponsors are being pursued.

CrackFit. This new group-sponsored project uses ETD’s CrackFit procedures and software to assess cracks in both low- and high-temperature high-pressure equipment. The technology presently covers 17 geometries of pressure vessels and piping components typically found in power and process plants. Steam and gas turbine geometries also are being investigated. CrackFit software significantly reduces the time required to complete assessments, in some cases reducing weeks to days.

Initial work by ETD was supported by the European Commission and leading European universities and industry; data were supplied by operating plants.

Says Shibli, “CrackFit comes with an optional crack initiation and growth database for base metal, HAZ, and welds for many materials of interest to the power and process industries. It deals with defects such as lack of fusion/penetration in welds (defects at weld toes), internal- and external-surface emerging or embedded defects in straight pipes and pipe bends, defects at stress concentrations such as sharp corners (T-pieces, nozzles), and defects in plates, among others.”

Shibli further explains that the software is “user friendly for industry engineers who would like to carry out defect assessments without having to go through the different established codes or country/in-house defect assessment practices.” Procedures including BS7910 (a BSI-British Standards Code of Practice for the assessment of flaws), Nuclear Electric (UK) R5, the French nuclear code RCC-MR A16, and the European HIDA (high-temperature defect assessment procedure) are available as software options.

“The software allows for failure analysis (fast fracture, plastic collapse, and ligament rupture) and the evaluation of damage mechanisms such as creep, fatigue, and creep-fatigue interaction,” he explains.

A fitness-for-service program incorporated into the software performs calculations to determine a defect’s criticality in line with the assessment codes. The software carries out both deterministic and probabilistic crack assessments. Modules within the study are:

    • Materials.
    • Cyclic loading of powerplants.
    • Sensitivity and probabilistic analysis.

Weldlife. This five-year GSP launches in early 2018 and will investigate changing the heat-treatment cycle of P91 welding with some changes to the welding processes. The initial basis is specific research in both Japan and the UK that shows weldment rupture life improvement by a factor of two to three. Japanese and UK companies have committed support; additional sponsors are invited. Fig 4 shows some of the planned pipe tests.

Drones. ETD is also beginning a group-sponsored project to study drones, robots, and other automated devices available in the market, and to develop recommendations for appropriate and cost-effective devices that may even be used as consumables. This includes 3-D printed drones with light plastic frames and safety cages. Ultimate uses include inspection and possibly repair capabilities.

Shibli ended the interview, conducted by Consulting Editor Steve Stultz, by referring readers interested in learning more about these and other projects to the organization’s website.

HRSG Forum with Bob Anderson

March 5 – 7, Hyatt Regency Houston

Several presenters and discussion leaders at the second annual HRSG Forum with Bob Anderson will be speaking on the topics described in the accompanying article on research in Europe and Asia regarding failure mechanisms, inspection, and welding of advanced materials for high-temperature/high-pressure steam systems.

A few examples:

    • Jeff Henry, respected internationally for his work associated with creep-strength-enhanced steels, will present on the evolving issues with these materials. The ASME Fellow is intimately familiar with the subject, serving as chair of that august professional organization’s Working Group on CSE Ferritic Steels. Henry, who was director of Alstom Power’s Materials Technology Center in Chattanooga, now is president of ATC Inc, a materials engineering services company.
    • Barry Dooley, perhaps the world’s best-known powerplant water chemist, will speak on flow-accelerated corrosion, the leading cause of HRSG tube/pipe failures. His presentation will cover the causes of FAC, methods to detect and monitor its presence in HRSGs, and chemistry actions to prevent its occurrence. Dooley, who works for Structural Integrity Associates Inc, volunteers most of his personal time to industry organizations focused on helping plant owner/operators improve safety, availability, and profitability. To illustrate: He is executive secretary of IAPWS, which provides Technical Guidance Documents free-of-charge to improve operating practices relating to water chemistry, metallurgy, etc. Dooley also is an active participant on the steering committees for several user-focused organizations—including Air-Cooled Condensers UG, Canadian HRSG Forum, Australasian HRSG UG, HRSG Forum with Bob Anderson, European HRSG Forum, and International Conference on Film-Forming Amines and Products.
    • HRST Inc’s Director of Engineering Bryan Craig will discuss factors affecting superheater and reheater metal temperatures, perhaps the variable with greatest impact on the lives of these components. Long-term overheating of superheaters and reheaters is conducive to repairs and component replacements costing millions of dollars. Many HRSGs installed during the “bubble” are at significant risk today. Craig will help attendees better understand and manage remaining life.

COMMENTARY: What the digital revolution means for electricity

By Team-CCJ | April 19, 2022 | 0 Comments

What’s happening inside the fast-growing, hugely profitable, digital behemoths—one characterization is the “Big Five,” namely, Amazon, Apple, Facebook, Google, and Microsoft—may be more important than what the electricity industry sees from the outside.

We know they are driving the industry towards renewable energy through direct investments in solar and wind facilities and green electricity purchases. We know they are pushing products and services for electricity consumers, such as smart thermostats and home automation devices, to manage demand to lower levels.

One of these firms, however, has an array of internal programs and policies which may be astounding to many. For example, the company burdens (think of it as a tax or penalty) every business unit with a price for carbon that is 40% higher than the prevailing market price.

Corporations charge business units for a variety of “services”; in some cases, these are called “indirects” on the unit accounting statement. But an internal carbon tax is something different.

This becomes a powerful incentive for executives and directors to drive towards a progressively lower carbon footprint. The taxes collected are then directed into a clean energy fund, used to support innovation and technology which reduce carbon and leads the company to its carbon neutral goal.

If that’s not enough, this company, according to a presentation delivered at the Ten West Innovation Conference in Tucson, October 2017, also is beginning to take on the carbon burden associated with customers for its cloud-based services.

The internal cost of energy at this company includes not only, for example the cost of utilities or the cost of an airline ticket, but also the cost to offset the carbon associated with both. Carbon impact becomes a line item in the budget for every business unit. Electricity consumption, however, is the primary source of carbon emissions for its internal operations. Another metric being considered is “emissions per housed employee (direct and contractors) per square foot.”

It doesn’t take a genius to see that such a program can drive behavioral changes far beyond the company itself as well as drive demand for energy-efficiency and demand-management related products and services. According to a company white paper, moving customer services to the “cloud” in itself can reduce energy use and carbon footprint by at least 30%.

Employees also are encouraged to find ways to reduce carbon impact for the company and in their personal lives, including use of mass transit, energy conservation, organic farming, and local resources. It would seem straightforward that the other digital Big Five firms, and the second and third tier firms, are implementing similar programs internally, since they would benefit in similar ways.

Traditional electricity industry players need to take note: Sustainability and carbon neutral programs are more than slogans, suggestion boxes, and thick documents written by consultants sitting on shelves in the C-Suite. For these digital firms, they are driving productivity and profitability. And their profitability is undoubtedly going to be revenue losses for traditional players.

Scroll to Top