Scribd
Upload
35 views

Axial Shift Issue Due To Governor Value

The steam turbine driving a compressor experienced speed fluctuations and excessive axial movement due to the governor actuator failing twice in two weeks. An investigation found the actuator's internal filters were clogged with carbon particles, as was the lube oil system. The carbon was introduced during a "Coke Burn" process prior to an overhaul when fine carbon dust mixed with the compressor gas and contaminated the seal oil. Not fully addressing this during the overhaul allowed the carbon to accumulate over time and foul the actuator filters, causing the failures. Flushing and replacing filters and oil resolved the issue.

Uploaded by

Sasi Nimmakayala
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
35 views

Axial Shift Issue Due To Governor Value

The steam turbine driving a compressor experienced speed fluctuations and excessive axial movement due to the governor actuator failing twice in two weeks. An investigation found the actuator's internal filters were clogged with carbon particles, as was the lube oil system. The carbon was introduced during a "Coke Burn" process prior to an overhaul when fine carbon dust mixed with the compressor gas and contaminated the seal oil. Not fully addressing this during the overhaul allowed the carbon to accumulate over time and foul the actuator filters, causing the failures. Flushing and replacing filters and oil resolved the issue.

Uploaded by

Sasi Nimmakayala
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

Steam Turbine High Axial

Displacement due to Governor


Actuator Filter Clog

KER PIN JUN – EXXONMOBIL


NG WEI SING - EXXONMOBIL

SY M P O S I A : 2 4 – 2 6 M AY 2 0 2 2
SHORT CO U RS E S : 2 3 M AY 2 0 2 2
Author
Ker Pin Jun Ker Pin Jun is a Machinery Engineer in ExxonMobil Singapore. He has 14
Machinery Engineer years of rotating equipment experience in refining and petrochemical
industry. Ker Pin Jun holds a Bachelor’s degree in Mechanical Engineering
from University of Melbourne.

Ng Wei Sing Ng Wei Sing is a Lead Machinery Engineer in ExxonMobil Singapore and
Lead Machinery has 9 years of rotating equipment experience in refining and
petrochemical industry. Ng Wei Sing has a Postgraduate Doctoral degree
Engineer in Mechanical Engineering from University of Leeds.
Abstract
The governor actuator for a steam turbine was observed to be hunting and failed to maintain stable operating speed. The steam
turbine is used to drive a centrifugal compressor for H2 service in a recycle loop. The extent of speed oscillating was exaggerating
over time and caused excessive turbine axial movement. The unit was forced to shut down to prevent secondary damage due to
turbine high axial movement.

The steam turbine was overhauled approximately 9 months prior to the incident. Oil flushing was performed, lube oil was
changed out and main oil filters were replaced as part of machinery overhaul scope. The main oil filter elements were inspected
during troubleshooting of this event and were badly fouled with blackish substance.

The governor actuators were inspected and the internal filters for the oil supply line were clogged with soft, blackish substance.
Foulant analysis was performed using Energy Dispersive X-ray (EDX) which identified the main component to be Carbon. De-
gassing tank oil was drained out and observed layer of blackish substance. EDX identified the main component to be identical to
that found in main oil filter. Water sample from de-gassing tank bottom drain was tested and the pH was 3-4 (i.e. acidic).

Further troubleshooting revealed that fine carbon particles were carried over from upstream and mixing with the process gas
prior to the unit shut down for the compressor train overhaul. The presence of carbon particles had contaminated the seal oil
and control oil circuit which then clogged up the actuator internal fine filters (~40um). This case study presents the unusual
phenomenon, troubleshooting tools, recovery process and key learnings which provide user operating with similar service a
reference.
Agenda
1 Case Study Background

2 Problem Description

3 Findings

4 Analysis

5 Key Learning
Equipment General Info
Steam Turbine Details
Type: Back pressure turbine
Frame: 1U2R
Rating: 1.2MW (3 stages)

Rated Speed: 9,500 RPM


Max. Speed: 11,550 RPM
Min. Gov.: 6,050 RPM
st
1 Critical: 5,000 RPM
Trip Speed: 12,705 RPM
Equipment General Info
Compressor Details
Type: Centrifugal
Model: 4B26 (4 stages)
Suction: 17 barg @43degC
Disch: 23 barg @79degC
Power: 1,050kW
Bearing: Tilting pad
Seal: Mechanical oil contact seal

Lube Oil System


• Common circuit for lube oil, seal oil and control oil
• Lube Oil: 138kpa @145L/min
• Seal Oil: 2,240kpa @76L/min
• Control Oil: 828kpa @19L/min
• Oil Filter: 10um
Equipment General Info
Internal
70um filter

External
40um filter

Model: TM-25LP
External filter: 40um (nominal) @ oil inlet
Internal filter: 70um (nominal) @ coil inlet
Control oil press: 552kpag
Control oil flow: 3.8L/min
Problem Description
• The steam turbine governor valve actuator had failed twice in the space of two weeks.
• First failure on 11th Dec 2019 (5 days outage):
o Turbine sudden loss of speed from 8,800 RPM to 3,800 RPM. Speed gradually picked-up to close to 1st critical speed of 5,000 RPM due to compressor
load reduction.
o Speed was considerably stable prior to the loss of control, but rapid fluctuation in axial displacement was observed 1 – 2 weeks prior to the incident
• Second failure on 24th Dec 2019 (7 days outage):
o Turbine speed started to fluctuate wildly between 8,800 RPM and 11,500 RPM (max. continuous speed)
o Axial vibration swing violently between -0.15mm to +0.10mm. Abnormal sound can be heard at site
• In both events, turbine was manually shut down to prevent any secondary mechanical damage (which can be very costly to repair with long outage)

2nd outage on
1st outage on 11th Dec 2019 24th Dec 2019

Speed and axial displacement rapid fluctuation prior to shutdown


Key Findings
1. Turbine was just overhauled in April 2019, approximately 8 months prior to the failure.
2. As part of overhaul, governor valve actuator was replaced with new. Oil flushing was performed and lube oil and oil filters were replaced.
3. Actuator internal coil filters were inspected and found to be badly clogged with blackish substance.
4. Lube oil filter elements were inspected and found to be badly coated with similar blackish substance.
5. Draining of degassing tank found similar layer of blackish substance which float on oil when left to settle.
6. EDX analysis shows blackish substance was mainly carbon.
7. Degassing tank bottom draining found water and tested to be slightly acidic (suspect carbonic acid)

4 4
3 3
Key Findings
5 6

7
6
Degassing tank – think layer of blackish substance and acidic water sample

EDX analysis of blackish particles shows


main composition of carbon
Analysis
1. Prior to shutting down the turbine for overhaul, the compressor was used for “Coke Burn” operation to avoid catalyst poisoning.
2. By product of “Coke Burn” are soot (fine carbon dust), carbon dioxide and water.
3. During “Coke Burn” process (unit shutting down), the Nitrogen at buffer gas header was only maintained at 700kpag vs. design 1,375kpag for
effective operation of buffer gas system.
4. During “Coke Burn”, the compressor suction pressure was 600kpag. Taking into account the pressure drop from battery limit to compressor, the
buffer gas pressure will not be adequate to prevent “Coke Burn” process gas to come in contact with the seal oil.

Seal oil Circulated


Sour oil Degassing Lube oil
Process gas Buffer gas drainer in oil
system tank reservoir
module system

Carbon Dust Foulant path into oil system


Analysis
Why carbon dust in degassing tank was not identified during overhaul?
➢ Degassing tank inspection and cleaning was not included in the overhaul work scope
➢ During oil flushing, the sour oil return circuit, including the degassing tank was not flushed

Why only after 8 month of operation?


➢ Compressor is equipped with mechanical contact oil seal
➢ Under normal circumstances, the seal oil leakage rate is very low
➢ It takes time for the carbon dust to be flushed to degassing tank and overflow back to oil reservoir

Why carbon dust was not fully trapped by lube oil filter (10um)?
➢ Carbon dust particle size range is between 5 – 15um
➢ Carbon dust smaller than 10um can still pass through the filter and coagulate forming
bigger particle size downstream of the filter
➢ Inspection showed potential by-pass of some of the filter elements
Action Taken & Results
• Performed oil flushing prior to restart – using mesh as well as Swagelok 7um hydraulic filter
• Replaced oil filter elements
• Full replacement of lube oil with fresh oil
• Replace the governor valve actuator with tested and calibrated spare
• Modified TM-25LP governor actuator external filter to be of duplex design with dP monitoring. Online replacement is feasible.
• Steam turbine successfully restarted and governor actuator in stable and healthy operation to date.

NEW external duplex Original external


filter location filter location
Key Learnings
1. Risk of abnormal operation (i.e. “Coke Burn” operation) was not fully recognized and understood.
2. Oil flushing should be planned to cover all sub-circuits / oil paths in the oil system.
3. Both lube oil reservoir as well as degassing tank should be inspected during overhaul, clean if required.
4. Clean and adequate supply of buffer gas is critical not only for mechanical seal reliability, but also to prevent
lube oil contamination due to sour oil contact with dirty process gas.

You might also like