Frame 6 balanced long distance by phone, text, e-mail

When a Kuala Lumpur-based overhaul contractor encountered ongoing turbine vibrations at the Sabah Electricity Power Plant in Labuan, Malaysia, its call went 13 time zones away to Paul Tucker in Texas. Tucker’s first call was to Ken Knecht in Alaska. They agreed to work from their respective home offices to fix the problem.

Site data at Labuan was showing continued high vibration on both the cold-end and hot-end bearings of the GE Frame 6B turbine. To complicate matters, a major overhaul of the unit had been completed in late 2014. All efforts to improve the post-rebuild vibration signature were unsuccessful.

When the Malaysia call came in, Tucker, president, First Independent Rotor Services of Texas (FIRST), was just returning from Asia to his Humble (Tex) shop. Knecht, who recently joined FIRST, was away from his Eagle River base. Traveling quickly to Malaysia would have been difficult. They had a better idea: Solve the problem by long distance using modern communications technologies (Fig 1).

F6 balance Fig 1

FIRST immediately asked for all site test data. Vibration spectra indicated a predominant frequency at the gas-turbine (GT) operating speed; the phase angle, a coupling imbalance in the turbine rotor. After reassembly data from the 2014 repair were received, Team FIRST spoke with key personnel at the rotor repair shop in Dubai and concluded balance weights installed during rotor balancing after the overhaul were not arranged properly.

Digging further, Tucker and Knecht found an issue with the way the rotor had been repaired. They had some concerns about how it was final-balanced, and specifically about the rotor’s mid-span weight distribution. If that turned out to be the case, they could fix the issue at hand and pinpoint the original repair problem.

By now, the communication triangle was fully operational among Texas, Alaska, and Malaysia. All commonplace essentials were used: phone, text, and e-mail.

Discussing the data further with the outage contractor in Kuala Lumpur, all agreed to attempt a full-speed trim balance on the GT rotor in its current, assembled condition. All parties knew the alternative—to open the machine and go back in—would be costly in time and money. This particular engine was dispatched as a base-load unit in emergency-need situations. However, it generated income for its owner by being “available.” It was not.

How it worked. Communicating with the contractor’s vibration technician, Knecht developed a detailed procedure and prepared custom drawings plus a list of test weights that could be made or obtained in Malaysia. Specialized weights for the balance shot would be sent from Houston. When Tucker and Knecht concurred, all material was transmitted to the contractor. Back in Kuala Lumpur, the contractor reviewed the procedure and information, confirmed their agreement, and prepared to mobilize to the relatively remote site.

The setup. The Luban Frame 6 is equipped with accelerometers at all bearing locations, but non-contact proximity probes were never installed on this unit. Knecht noted that late-model 6Bs allow for adding external rotor weights in the field, but this an older unit, did not.

Portable balance pickups would be installed on the output shaft of the accessory gearbox, the cold-end (P1) bearing, the hot-end (P2) bearing, the input side of the load gear, and the opposite side of the input high-speed pinion gear (Fig 2). A photocell pickup was installed on the accessory-gear coupling, and a reference point on the accessory coupling would be used for the full balancing procedure.

F6 balance Fig 2

The accessory-coupling-hub turbine-side bolt pattern was used for test and balance-weight location P1. The turbine-output-shaft hub bolt pattern was used for a balancing location relative to the reference on the accessory coupling.

Knecht explained that a test-weight value could be used by removing one of the 12 coupling bolts on the cold end (P1). Since the unit could not accept an external weight, removing a bolt from this end essentially added that weight on the other side, the quickest way to do a test weight. An external balance groove exists just forward on the rotor’s output coupling hub. It was used for test and balance weight location P2 on the hot end.

Monitoring from afar. With all portable pickups and key phase probes installed, site testing began. For each test, all instruments and locations were first photographed and transmitted to Knecht for verification. The original run was completed then all data recorded and transmitted. FIRST next calculated a correction weight at P1, to be installed by onsite personnel. A second run was completed, recorded, and transmitted.

Then FIRST calculated a summed weight which was installed for a third run. From this run, Tucker and Knecht determined that an additional trim balance weight would be needed by additional correction to the hot end (P2). A test weight was installed, and another run was completed, recorded, and transmitted. With the test weight data from both P1 and P2 in hand, FIRST calculated the balance shot. The balance calculations were correct.

Run lengths were determined by the stability of the readings, and most reached stability between one and four hours. All participants put in extra hours as needed. As Knecht stated, “you just can’t rush a balance job.”

Some cultural details added to the time-zone differentials. Most site conversations took place around 1 or 2 a.m. in Alaska. And although all participants communicated in English, there were dialect variations and some participants spoke very quickly. Therefore, requests to slow down or repeat became the norm. Most conversations then were confirmed by back-and-forth texts and e-mails (texting being the most useful). Transmitted photographs of equipment setups also were critical.

The common communication channels became Tucker with the contractor and Knecht with the contractor’s site technician. Interestingly, these four never spoke together as a group. Knecht would also receive an occasional call from the owner/end user in Kuala Lumpur, which he deemed “reassurance” calls.

One significant adjustment was units of measurement. In North America, vibration levels are well known in inches per second. But the site was using millimeters per second, which can be difficult if you are not using that metric every day. As Tucker noted, someone tells you a journey will take 4 gallons, that’s one thing. When they say it could take 15 liters, you might need to stop and think about it.

Moving forward. Long distance did turn out to be a better idea, and the project was completed within 10 days, including logistics delays. The more traditional response of air travel would have added travel time and expense, as well as delays coordinating logistics within Malaysia.

Instead, each group was able to concentrate on its participation in, and contribution to, the solution. Knecht believes the FIRST team would not have been as effective if it had taken the traditional response. Tucker’s summary: “You don’t always need to be on site to help. And with this experience, we could do it again, perhaps even faster.”

Posted in Best Practices |

Comments are closed.