501D5-D5A USERS: Engine vibration continues as a major concern of users
The 501D5-D5A Users can be characterized as an intimate group, historically attracting about two-score attendees to its annual conferences and vendor fairs, fewer at the “mid-year” meetings in January. Most attendees have been repeat participants, several from far-away places like Korea and Indonesia, with the typical user knowing perhaps half of the owner/operators in the room.
The chairman, Gabe Fleck, recently promoted to manager of gas plant operations at Associated Electric Cooperative Inc, has been at the front of the room for a decade, supported by Vice Chair Barry Mayhew, maintenance manager at Cardinal Power, and Director Lonnie Grote, compliance coordinator for NAES Corp.
Fleck’s low-key demeanor and unflinching dedication to the group is critical to the organization’s success. He works diligently at technical program development, and with sponsors to provide users new experiences in terms of meeting locations and social events. This June’s conference was in Louisville, affording visits to Churchill Downs, the Louisville Slugger Museum, and a local brewing company—plus bourbon tasting on a private yacht. The Seelbach Hilton Hotel was an apropos conference location for a mature group, dating back to 1905 and listed in the National Register of Historic Places.
That time always seems to stand still at D5-D5A meetings is not surprising because of the continuity of attendees and discussion threads. Technical topics are linked from meeting to meeting by interactive messaging via the group’s robust online forum. Over time, conclusions are drawn and solutions developed—with help from the OEM, Siemens Energy Inc, and participating third-party equipment and services providers, of course.
Fleet characteristics contributing to the group’s stability include the following:
- Both engines served are no longer in production (although there are three new D5As in storage), the two fleets holding steady with a total of 150 units worldwide.
- The service factors (operating hours divided by period hours) for engines not in cogeneration service are relatively low, based on information shared by asset owners. One reason for this undoubtedly is the increased use of more-efficient F-class engines in peaking service.
- Given the foregoing perspective, why these “surprises” at the 2013 annual meeting:
- Record attendance, breaking the 50 mark for the first time in the 17-year history of the group.
- Most first-timers ever, 22; plus a dozen who had participated in only one other meeting previously.
The only logical answer may be that the retirements of many senior plant professionals, predicted for several years, is now occurring and that their replacements have little or no experience in operating and maintaining D5s or D5As. Participating in this and other user groups is the only practical way new O&M personnel can come up to speed on what they have to know, and whom they should know to get the help needed going forward.
The technical program began after breakfast on Tuesday, June 4, with presentations by Mitsubishi Power Systems Americas Inc and a few suppliers that would participate in the vendor fair later that afternoon. Four hours of closed user discussion rounded out the day. The second and third days of the meeting were divided between Siemens technical presentations and closed user discussions.
Engine vibration
Engine vibration occupied a considerable amount of discussion time. Identifying and correcting sources of vibration has been a major initiative between the users group and the OEM for the last two years. The collaborative Vibration Improvement Program is guided by an executive committee consisting of user representatives and Siemens management.
Work packages were developed and teams formed to perform root-cause investigations, testing, rotor-dynamics modeling, and data analysis, and to make operational recommendations. Creative Power Solutions, an independent turbomachinery consultant based in the Phoenix area, was retained by Siemens to assist in the effort.
A user started the discussion by saying his D5A trips on high vibration about once every 10 starts, with the trip setting at 10 mils. It takes two hours for the vibration to build to that point. He thought the turbine blades might be locking-up on startup, so root springs were installed in the gaps between the R4 blade roots and the discs to make rotor vibration levels more repeatable and to facilitate balancing.
The low-cost consumable part was developed to minimize disc wear by preventing blade rock during turning-gear operation. However, springs were not the answer to this user’s issue. Where springs do provide benefit, a recommendation was made to replace them at each hot-gas-path (HGP) inspection. More information on blade-root springs is available in the presentation Mitsubishi engineers made to the group. It is posted on the D5-D5A website in an area accessible only by registered users. The steering committee urges qualified owner/operators to register today.
Another attendee said vibration on his machine was traced to coke buildup on the bearing exhaust line; when coke broke off (Fig 1), the balance would shift. He learned that running unbalanced—example: x=2, y=5—contributed to more stable operation. Each time technicians tried to do a proper balance, vibration increased and the unit tripped. Someone suggested that a successful vibration analysis demands that you know what you’re looking for.
Yet another user got into the passionate exchange, suggesting that overtightening of bolts on the marriage coupling to adjust run-out can cause vibration when the bolts relax. He said this happened on his unit.
Someone else asked if everyone with a vibration issue is sure their bearing probes are properly located. Here’s why: Installing exhaust-bearing vibration pickups on the bearing housing rather than outboard can reduce readings by up to one-half during startup, full speed/no load, and at base load. This location, which is consistent with ISO standards, is said to minimize the contribution of bearing-housing motion to the vibration signal. Redundant probes were suggested when making this modification, to improve serviceability.
The discussion continued with an attendee who was puzzled as to why vibration on his machine was at 4 to 5 mils—until plant personnel really dug into the matter. Turns out the stack was pushing down on the end of the machine and causing the unbalance. The solution was to lift the stack by about an inch.
The air separator also was mentioned as a possible source of vibration. One of the Mitsubishi speakers had addressed that at the beginning of the meeting. He said the original air separator uses spring force to maintain contact with the R1 disc and that relative movement at the R1 disc face can cause fretting. Retrofit with a bolted air separator was said to reduce R1 disc fretting potential and increase the through-bolt compressive load on turbine discs—effectively decoupling the air separator. This mod, developed for the M501F gas turbine, is applicable for the D5 and D5A turbines as well. A drawing is provided in the Mitsubishi presentation cited earlier.
You could sense the frustration of a user who told that group that the vibration gremlin has resided in his D5A since 2001, tripping the unit every dozen starts or so. It seems that rust/dirt accumulated and then released, upsetting the engine’s balance. His experience suggested that balance weights aren’t necessarily the solution. The unit had so many weights it was ridiculous, he said. Every service technician believed he had the solution, the user continued, and added weights. All were wrong.
Maintenance outages
The closed user discussion shifted to engine maintenance outages, experience with conventional and extended-life parts, repairs and coatings, etc—a cornucopia of subject matter. One attendee with a pending major inspection polled the group for experiences in using contractors for this work.
Standard practice at his facility has been to perform maintenance outages with plant personnel under the direction of an OEM technical advisor. This has worked well over the years, but a multitude of projects would prevent staff from doing the work on the upcoming major. He was leaning toward hiring the OEM or Mitsubishi to perform the outage. What was the group’s experience?
Hands shot up in all directions. The majority of attendees had opinions, as you might imagine. Difficult to draw objective conclusions from such an exchange: The experience base of the commenters varied widely, plus the crew from Company X that did a good (or bad) job for you likely was not the same crew that did the bad (or good) job for the colleague disagreeing with your assessment.
Also obvious from the discussion: Performance may vary widely among repair shops, depending on the owner and key personnel at the time the work is scheduled to be done. Thorough due diligence of candidate shops was strongly suggested. A couple of participants stressed very objective evaluation of low-cost providers.
They said some service providers may not have the specific experience, or even qualified in-house procedures, for your work. Those shops may look to you to provide repair procedure. In sum, the users were in general agreement that there is no “right” answer on which way to go regarding repairs. You have to make decisions based on what “fits” at the time.
Good discussion. However, the value of the exchange definitely was positive. For those in the room who had experiences they wanted to forget, it was a welcome opportunity to vent; for those with good experiences, an opportunity to pass on to colleagues their good fortune with the hope that others might benefit as well. There seemed to be unanimity of opinion on two points:
- Leading aftermarket repair shops typically will scrap fewer parts, cost less, and have better turn times than an OEM.
- Owner/operators must have a shop presence to assure proper conduct of repair jobs, assess repair quality, evaluate parts the contractor wants to scrap, etc. If there’s no direct employee with the capability to perform these functions, use of third-party expert was highly recommended. Two oversights mentioned: Repaired transition pieces provided without cooling holes; a new vane row that included a refurbished vane.
As the discussion of overhaul experience continued, one user mentioned that his plant had essentially completed its major when the rotor being lifted for reinstallation in the casing was dropped. The lengthy exchange following this revelation left one with the impression that the owner must remain involved in transportation (Fig 2), rigging, lifting (Fig 3), and other such tasks contracted to others. It’s your unit that won’t be producing power when the unthinkable happens.
There have been several reported “drops” in the last year or so. This particularly disturbing fact might well point in the direction of inexperienced riggers. Rigging is both engineering and art and one must assume that some of the more talented people have retired—riggers get old, too—and that their replacements may not have been properly trained.
Another thread related to annual maintenance outages was started when the subject of contingency was brought to the floor. The question that stimulated the discussion went something like this: For a base-load D5A, how much contingency would you allow for a combustion inspection? First reply: “We aim for 8000 hours but will go to 10,000 or 10,500 if necessary.”
Next question: How far can you stretch out an inspection without incurring problems? A different user replied: You should not put off an HGP or major, although some believe 10% in hours is acceptable. It’s all about risk assessment, he said. The ductility of R1 and R2 turbine blades is the variable that should drive decision-making. For a base-load unit, material degradation is not as much of a concern as it would be for a peaking machine.
The user continued, saying it’s important to get critical hot parts into the shop for thermal rejuvenation. The life cycle of components comes into play here, he added. Is this the first cycle, the second, the third? The younger the component, and the fewer the thermal cycles, the more flexibility you have in scheduling—generally speaking.
 |
 |
 |
Service agreements had a brief run in the open discussion session. Renegotiation and what should be included in the new arrangement were key talking points. A parts agreement was suggested in place of a service agreement. If you go that route, the group was told by a colleague that they have to consider who handles and stores the parts. In addition to tax implications, how will you confirm that all parts are of the quality level you believe they are? Recall the earlier comment on the reconditioned vane segment that was included in the new vane row.
One more thing the experienced user suggested: Specify in the contract that the customer has the right to have the service provider’s program manager replaced at its discretion. This is important in situations where the original manager approved by the customer is changed by the contractor or leaves its employ and the replacement is not to the customer’s liking.
Finned-tube heat exchangers used for rotor air cooling (RAC) were introduced by a user who wanted to know how to clean fins to improve heat transfer. Pressure washing is not effective he was told and there’s the possibility that a high-pressure water jet will tear fins off the tubes. Flooding the housing with a suitable solution was suggested as the best method. But if you go this route, don’t forget to put plastic over the motor underneath the unit.
Several attendees were interested in how you might fix flattened-over fins on this and other types of air-cooled heat exchangers (lube-oil coolers, for example). A hail storm really can squash down the fins, one person said (Fig 4). But he was told that this condition had relatively minimal effect on the unit’s cooling performance compared to the separation of the fin strips from the tubes. The only practical corrective action for fin/tube separation is to replace the tube bundle, which can cost about a quarter of a million dollars for D5A air-to-air RAC. An infrared scan of the unit can provide an indication of heat-transfer efficiency, someone said.
Two discussion topics linked up when someone asked what type of rotor balancing should be done during a major—high- or low-speed? One attendee said a low-speed balance was fine but another disagreed considering the rotor vibration issues many owner/operators are experiencing with these engines. A user said R1 and R2 blades were changed in the field; the new sets were moment-weighed and properly installed. No balancing was done. This rotor suffered balance problems. Another attendee said they did a high-speed balance during a major with a fully bladed compressor and no blades in the turbine. There no balance issues following the installation of moment-weighed blade rows with no additional balancing.
Transition pieces always get air time at GT user-group meetings. In Louisville, the owner of two D5A peakers commissioned in mid-1999 with water-injected DF42 combustion systems reported transition failures, which occurred in the upper panel near the exit mouth (Fig 5). Temperatures in this region were said to be higher for DF42 systems than they are for DLN. Longer run times and higher outputs probably contributed to the cracking, the user said.
 |
 |
 |
Discussion threads that didn’t extend very far included generator issues such as corona damage, partial discharge, and vibration; plus, wet compression (Fig 6) and shaft grounding (Fig 7). Disc cavity No. 3 temperature excursion on startup received brief mention. It is conducive to a turbine trip, which can be avoided by slower ramping. It was said that if operators don’t catch the increasing temperature and slow down the ramp, the engine will run back. A user said he avoided run-backs by restricting the ramp to 3.5 MW/min. He added that, in his case, the issue was seasonal, but didn’t elaborate.
Air filters. No user-group meeting would be complete without discussion of the filter house. One attendee questioned something he had heard regarding the placement of an evaporative cooler in front of the filters. He mentioned Donaldson Company Inc as the supplier. There were lots of shoulder shrugs with some users thinking they might have heard of this but not sure. An email from the meeting room to Donaldson’s GT aftermarket manager, Barry Link, got this response within a couple of minutes:
Donaldson manufactured several two-stage static inlet filter systems for Siemens V-series machines that were located behind the evaporative cooler. These filtration systems used conical/cylindrical element pairs with a panel-style prefilter. He thought the arrangement probably was specified by the OEM and believed there probably were fewer than a dozen such inlet systems operating in North America.
Generally speaking, Link continued, Donaldson prefers not to put evaporative coolers upfront in first-fit applications based on the following factors, among others:
- Shorter filter life.
- Shorter evap-media life.
- Increased water consumption because additional blowdown is required.
- Additional inlet-system maintenance.
However, in some retrofit cases, he thought upfront evap coolers could be a viable option based on the lower cost of the turnkey project to procure and install.
Bearings. A user asked the group how often others change out journal bearings. One person said “each major.” Several attendees suggested simply rebabbitting pads and resetting clearances. This was discussed in detail last year during the tour of the Pioneer Motor Bearing shop in Charlotte. By show of hands, owner/operators experience more babbitt cracking than loss, probably because of vibration.
User presentations are always instructive, enabling other attendees to see what their colleagues are doing to improve performance and safety. One utility engineer reviewed the capital/O&M projects recently implemented at his plant. Given today’s focus on physical and electronic security, and the plant’s location in a tornado-prone area, control-room “hardening” was given top priority. This involved adding reinforced cinder-block walls, a hurricane-rated door, and bullet-proof windows. Electronic security upgrades also were implemented. Ethernet outlets were installed and electrical outlets increased in number.
Generator upgrades included installation of a bus coupler on generator phases to monitor stator-winding partial discharge and the generator pole crossover mod. Safety enhancements: Installation of access ladders, platforms, and steps, and addition of a safety fence around the turbine roof. The handwheels for onsite gas supply valves were painted red so first responders know what valves must be closed.
Lube-oil coolers were modified to provide an additional 5 deg F of cooling and a dam was built around the coolers to catch any leakage. Building HVAC equipment was replaced and upgraded to reduce the headaches and maintenance costs associated with overheating of the electrical package. Turbine insulation was replaced as were station batteries. A turning-gear economizer was installed and the control system upgraded.
Siemens plans its eight hours of presentation/Q&A time at D5-D5A meetings based in large part on topics the users would like the OEM to address. The group’s steering committee maintains a collaborative and productive relationship with the Siemens legacy team headed by Mark Kamphaus and Jeff Kain, both of whom have worked closely with Fleck and Mayhew for years. Topics for this year’s presentations included the following:
- Axial rubbing program.
- Blade-ring face rope seals.
- Exhaust-bearing extension cover.
- Bently update.
- Exhaust baffle seal.
- Vibration.
- Controls.
- Exhaust cylinder.
- Generators.
If you are an owner/operator of a D5 or D5A and were unable to attend the Louisville meeting, you can access the presentations on these subjects on Siemens’ Customer Extranet Portal (CEP). Don’t know how to do this? Either solicit help from fellow users via the group’s forum or call your Siemens representative.
In their introduction to the specific technical presentations, the Siemens engineers noted that the top reliability detractors in the D5 fleet in the last year were compressor blade rubs, heavy tip rubs in the turbine section, and impact damage on turbine blades. Performance of the D5A fleet has been a solid “A” for the last decade, with reliability, availability, and starting reliability all above 98%.
Technical advisories have been issued in the last year on the following:
- Tip shroud wear on R4 turbine blades. Users also are seeing wear on the new material, IN738.
- Turbine R2, R3, and R4 potential rubbing issue and recommendations.
- Relocation of vibration probes and addition of additional probes, as noted earlier.
- Inspection of compressor-blade closing and locking keys.
- Urgent product bulletin: Modify the exhaust-bearing-cover removal tool before next use.
Looking ahead, work is ongoing on developing an upgrade for the exhaust baffle seal. The prototype for the 501F is being adapted for the D5.
The Mitsubishi presentation covered a lot of territory in only two hours. It began with a company update that discussed shipment of the company’s first air-cooled M501GAC, the acquisition of Pratt & Whitney Power Systems, status manufacturing and repair facilities, etc. Safety performance came next, followed by a review of the various service offerings and engineering solutions for the D5 and D5A fleets. Perhaps of greatest immediate value to attendees was the presentation on planning generator inspections—simplified, robotic, and major. The checklists presented on components to inspect when and by what means are a worthwhile addition to your outage planning library. Access the paper on the user group’s website; registered 501D5 and D5A owner/operators only. CCJ
