INTRODUCTION

Time duration	December 28, 2017 to Jan 2, 2018
Location	Saudi Arabia
Organization	General Electric (www.gepower.com)
Project	High Vibration on Air Cooled GE 7A6 Generator
Position	Senior Manager (Site Manager)

[CE 2.1]

I joined GE in a Services Leadership Program which was designed to train experienced new hires about GE’s design, operation and financial processes on fast track basis. I had the opportunity to learn and get hands-on experience of maintenance and troubleshooting of several Industrial Gas Turbine ranging from 50MW to 180MW. After completing my one year on job training, I was assigned as Lead Contract Performance Manager to a multi-million dollar long-term maintenance contract with one of prestigious customer i.e. ENGIE. After my hiring I was responsible for preparing computerized maintenance plans, budgeting, coordinating to resolve all technical and contractual issues, executing planned & unplanned outages of assigned power plants. I was also responsible for availability and reliability guarantees of the plants.

BACKGROUND

[CE 2.2]

My assigned Power Plant is having 9 x GE Frame 7FA.03 & 2 x 7EA Gas Turbines with the total capacity of 1800 MW. GT2 installed in an oil refinery of Saudi Arabia. The machine train consists of a GE 7E turbine driving a generator running at a design speed of 3600 RPM. Proximity probes are permanently installed at all fluid film bearings of the train to monitor radial vibration levels. Two proximity probes are installed axially on turbine outboard to monitor rotor axial position. The unit underwent a major inspection (MI) where generator rotor was rewound. Comparison of the vibration data before and after MI outage revealed an increase in vibration amplitudes in most of the locations (specifically on the generator). As vibration was on increasing trend so customer was worried. Customer requested me to perform vibration analysis & evaluation of gas turbine along with generator train balancing.

The unit underwent a major inspection (MI) where generator rotor was rewound. Comparison of the vibration data before and after MI outage revealed an increase in vibration amplitudes in most of the locations (specifically on the generator) but they are still well below the alert limit. I decided to start the analyses by ADRE data collection. Onsite data collection was done because seismic vibration data was not available in System 1 Turbine mid-span probe 39VS-21 gap voltage was outside OK limits since May 2017. Data collection from Mark VI was conducted on 31/12/17 and 01/01/18 to capture steady-state, shutdown and cold start-up events.

[CE 2.3]

Main objective of this project was to reduce the vibration values below 15 mm/sec by insitu balancing of generator rotor. This includes 4 phases:

• Data collection
• Analyses of data
• Calculation of balance weight/location
• Installation of balance weights
• Trial run to see affect and repeat above

As this was hit & trail method so we anticipated that multiple attempts will be made to solve the issue. In this case it took 7 trials to reach up to required vibration levels.

[CE 2.4]

The project Hierarchy is given below. I was responsible to carry out analyses and coordinate meetings with all involved parties. I met field service team who performed last outage on this machine. I came to know that stator rewind was carried also out on generator so I also carried out review meetings with workshop team as well.

PERSONAL ENGINEERING ACTIVITY

[CE 2.5]

My Roles & Responsibilities

I worked as Site Manager on this project. Some of tasks I performed were:

• I discussed in detail with customer to understand their Critical to Quality (CTQs) about this job.
• I clarified the scope, duration and challenges of the project to customer. In this case I explained them that they must shut down & start the machine multiple times since it was hit and trial method.
• I explained the support required from customer during the project execution like logistics, site accesses, work permits, lighting etc.
• I arranged meetings within my company (GE) with multiple linked stakeholders to explain the required resources, customer CTQs and expertise. In this case arrangement of ADRE equipment, qualified technician and craft were required.
• I coordinated with the parts team and provided them the part number of possible balance weigh which will be used.
• While unit was running, I mobilized the technician and asked him to hook up ADRE so that I can capture steady state and shutdown data to perform baseline assessment. On 28/12/17, vibration data was captured for steady-state and shutdown events. On 01/01/18 data was captured for cold startup and steady-state events to identify machine thermal behavior and compare slow roll vectors in hot and cold states.
• With the baseline data I examined different plots like 1X-filtered slow roll orbits, direct orbits, polar plots and average shaft centerline plots. With this examination I ruled out the other unbalance reasons like misalignment, thermal effects or bowed/bent, seal / bearing wear or looseness.
• Then I asked the customer to shut down the machine. After cool-down and making access to the rotor I marked /identified timing mark (key phasor in this case).
• Finally, I made drawing which showed rotor blades, probes placement and angles. This served as starting point of calculating trial weight amount and angle which I performed afterwards.

[CE 2.6]

PROBLEMS FACED & THEIR RECTIFICATIONS

Problem definition:

During balance correction, data gathered while trial runs I used two different ADRE units.The X-Y readings indicated swapped probes which meant that ADRE units were hooked up in wrong orientation.

Effect of problem:

Resulted in incorrect assessment and 1 day of delay

Root Cause of Problem:

The proper labeling cable was not made for ADRE data cables. Technician didn’t install correctly.

Final Solution:

I performed 3 actions to resolve this issue:

1. Contacted the GE vibration team and arranged a configuration drawing of ADRE & BN 3500 circuit.
2. I asked the technician to perform loop check.
3. I made firm Document about configuration of Probe orientation for all GT units and Labeled in PLC Panel where the BN-3500 was installed.

Problem definition:

While installing balance weights I found cracked balance weight stake marks.

Effect of problem:

These cracks have the potential to propagate due to continuous cyclic loading at high temperature. Crack propagation increases the risk of material liberation during operation which can lead to hardware damage downstream in the flow path.

Root Cause of Problem:

Crack initiation in stake marks were due to tensile overload during the staking process residual stresses & stress concentrations.

Final Solution:

All cracked stake marks were removed by the following steps.