cs 35-52 waukesha a solution to years of high … of science, mechanical engineering ... components,...
TRANSCRIPT
A SOLUTION TO YEARS OF HIGH VIBRATION PROBLEMS IN THREE REINJECTION COMPRESSOR TRAINS RUNNING AT 33 MPa DISCHARGE PRESSURE
Jong Kim, PhD/Presenter
Reproduced with permission of the Turbomachinery Laboratory (http://turbolab.tamu.edu). © 2016.
AuthorsJong Kim, PhD Senior Principal Engineer, Waukesha Bearings Corporation Bachelor of Science, Mechanical Engineering, 1985 – Busan National UniversityMaster of Science, Mechanical Engineering, 1987 – KAIST (Korea Advanced Institute of Science and Technology)Doctorate of Philosophy, Mechanical Engineering, 1991 – KAIST (Korea Advanced Institute of Science and Technology)
Marcio Felipe dos SantosSenior Maintenance Engineer, Major South American Oil CompanyBachelor of Science, Mechanical Engineering, 1979 – Universidade Federal of Rio de Janeiro
Barry J. BlairChief Engineer, Waukesha Bearings Corporation Bachelor of Science, Mechanical Engineering, 1990 – University of VirginiaMaster of Science, Mechanical Engineering, 1990 – University of VirginiaMaster of Product, Design and Development, 2015 – Northwestern University
AbstractOver a 13‐year span, a major South American oil company’s maintenance department fought high vibrations in three gas reinjection compressor trains. To reduce the chances of machine trips, technicians field balanced the compressors every year and replaced worn point contact pivot tilt pad journal (TPJ) bearings and O‐ring squeeze film dampers (SFDs) with new ones yearly. The downtime from implementing these preventative measures and from actual trips in the trains resulted in a loss of capacity of 1% a year and additional flaring of the gas.
After a thorough analysis of the compressors and inspection of damaged components, it was determined that the reoccurring problems would be solved by installing optimized Flexure Pivot tilt pad journal bearings with Integral Squeeze Film Damper (ISFD) technology into the compressors.
In 2013 the reinjection compressors were placed back into service with Flexure Pivot TPJ bearings with ISFD technology. Since then the reinjection compressors have exhibited lower vibration levels that do not grow over time, have had ZERO trips, and have not required field balancing for continuous operation. Overall efficiency has increased by approximately 1%.
Contents
• Background• Problem Statement• Modeling• Upgraded Bearing and Damper
– Design– Optimization
• Summary• Lessons Learned
Unit Background Information• 3 gas reinjection compressor trains operating since 2000• Each compressor train has two casings
– First casing (LP) experienced vibration issues– Second casing (HP) did not have vibration issues
• Discharge pressure: 33 MPa• Rated speed: 11,456 rpm • OEM bearing information
– 114.3 mm bore x 50.8 mm long TPJ– 5‐pad, load on pad– 60% offset
Vibration Issue
|18‐Nov‐2012
|24‐Jul‐2012
|6‐Dec‐2011
|01‐Nov‐2011
|24‐Dec‐2011
|10‐Jan‐2012
|19‐Nov‐2011
Problem Statement• Rotor vibration levels increased over time
Vibration trend of LP compressor with OEM bearings over 5‐month span
30.5 ‐
0 ‐
45.7 ‐
61.0 ‐
76.2 ‐
15.2 ‐Vibration pk
‐pk
(microns)
Vibration with new OEM bearings of the original design
Vibration with worn pivots
__ Compressor DE X__ Compressor DE Y__ Compressor NDE X__ Compressor NDE Y
0 ‐
40 ‐
0 ‐
60 ‐
80 ‐
100 ‐
20 ‐
Vibration pk
‐pk
(microns)
__ Compressor DE X__ Compressor DE Y__ Compressor NDE X__ Compressor NDE Y
– Downtime due to high vibrations resulted in about 1% loss in production time and additional flaring gas
– Field balanced every year– Installation of new OEM bearings
every year
Root Cause• Severe pivot wear in TPJ bearings
– Bearing clearance increased by 63+microns
• O‐ring damper performance changed– Damper film eccentricity ratio change
(bottoming out)
Top view Bottom view
Wear mark on pad Wear mark on housing
Pivot wear
Fretting damage on bearing OD
O‐ring grooveOil Inlet
Casing
Bearing Shell
Tilt Pad
Damper Axial Pressure Profile
End Seals
O-ring
Rotordynamic Model
1 5 10 15 60 65 6820 25 30 35 40 45 50 55
69 70
-400
-200
0
200
400
0 400 800 1200 1600 2000 24
Shaf
t Rad
ius,
mm
Axial Location, mm
Divisio
n wall
Baseline Mode Shapes with OEM Bearings• Small log dec for baseline model (no SFD and no aero cross‐coupling
stiffness from seals and impellers)
Cmin (minimum clearance)
-400
-200
0
200
400
0 400 800 1200 1600 2000
Sha
ft R
adiu
s, m
m
Axial Location, mm
f=4400.1 cpmd=.2065 logdN=11456 rpm
Cmin (minimum clearance)
Stability with OEM Bearings• Level I stability predicts that the rotor is unstable without SFD
15000 · 3 · 6319.64 · 0.78 · 12142 · 8 121.932 2.13 07 /
-400
-200
0
200
400
0 400 800 1200 1600 2000 2
Shaf
t Rad
ius,
mm
Axial Location, mm
f=4427.8 cpmd=-.4406 logdN=11456 rpm
Cmin (minimum clearance)
Division Wall Seal Contribution
• Insignificant improvement to stability with division wall hole pattern seal
• For stability, SFD required
daQA
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.0E+00 1.0E+07 2.0E+07 3.0E+07Lo
g D
ec
Applied Cross Coupled Stiffness, Q N/m
Existing Brg (Cmax)
Existing Brg (Cmin)
With DW (sw ratio=0.5)
Without Division Wall (DW) seal
Estimated aero cross coupled stiffness
-0.75
-0.50
-0.25
0.00
0.25
0.50
0.75
1.00
1.25
0.0 0.2 0.4 0.6 0.8 1.0
Log
dec
of 1
st m
ode
O-ring SFD eccentricity ratio
OEM bearing with SFDOEM bearing without SFD
With Applied Cross Coupled K=2.13E+07 N/m
O‐ring Squeeze Film Damper (SFD)• Damper radial clearance (c) = 0.110 mm• Damper radius (R) = 95.25 mm• Effective damper length (L) = 37.85 mm• Stability is very sensitive to damper
eccentricity ratio ()• Added O‐ring stiffness
21 1 /
Damper stiffness and damping coefficients (without O‐ring stiffness):
K=1.94E+07 N/m, C=1.88E+05 Ns/m at =0.25K=6.06E+07 N/m, C=2.62E+05 Ns/m at =0.50K=1.70E+09 N/m, C=2.06E+06 Ns/m at =0.90
Unstable
Pivot Wear Effects on Synchronous Vibration• Pivot wear increased operating bearing clearance
and reduced preload, resulting in increased synchronous vibrations
• A bearing with an SFD can make the rotor less sensitive to pivot wear than a bearing without an SFD
Max
CO
S
Min
CO
SR
ated
CO
S
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 2000 4000 6000 8000 10000 12000 14000R
espo
nse,
mm
p-p
Rotor Speed, rpm
Pivot wear effect (with SFD)
No pivot wear0.013 mm pivot wear0.025 mm pivot wear0.038 mm pivot wear0.051 mm pivot wear
Max
CO
S
Min
CO
SR
ated
CO
S
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0 2000 4000 6000 8000 10000 12000 14000
Res
pons
e, m
m p
-p
Rotor Speed, rpm
Pivot wear effect (without SFD)
No pivot wear0.012 mm pivot wear0.025 mm pivot wear0.038 mm pivot wear0.051 mm pivot wear
SFD:K=4.375E+07 N/m, C=2.625E+05 N‐s/m
Bearing Upgrades (Journal Bearing)1. Flexure Pivot Tilt Pad Journal Bearing
– No pivot wear• Integral pivot• Maintains bearing clearance
– High pivot stiffness• No pivot stiffness effect on bearing
dynamic coefficients
– Tight control of clearance and preload• Electrical Discharge Machining (EDM)
– No pad flutter
Conventional Tilt Pad Journal Bearing (Rocker Back)
Flexure Pivot Tilt Pad Journal Bearing
Bearing Upgrades (Damper)2. Integral Squeeze Film Damper (ISFD)
Conventional SFD ISFD
Damper Film
O‐ring
‘S‐spring’
0.11 mmOriginal Damper
Clearance
0.35 mmISFD Clearance
– Accurate stiffness control by eliminating O‐ring support
– No change in stiffness and damping over time
– Designed to counter static load
– Optimized damping
– Less cavitation
Optimization of ISFD
• Optimized stiffness and damping are 4.375 E+07 N/m (250,000 lb/in) and 2.625E+05 N‐s/m (1500 lb‐s/in)
• Additionally, the ISFD is designed to center the Flexure Pivot TPJ under gravity load by countering the static deflection
Redesigned Bearing
Original Design Optimized DesignConventional TPJ with SFD Flexure Pivot TPJ with ISFD Technology5‐pad, Load On Pad 4‐pad, Load Between PadShaft diameter 114.300 +0/‐0.013 mm Shaft diameter 114.300 +0/‐0.013 mmBearing bore 114.414 +0.025/‐0 mm Bearing bore 114.427 +0.025/‐0 mmClearance range 0.124/0.156 mm Clearance range 0.124/0.156 mmPreload range 0.293/0.501 Preload range 0.230/0.273L/D 0.444 L/D 0.500Pad arc 60° Pad arc 72Pivot Offset 60% Pivot offset 55%
Optimized
1st Mode Shapes: Original and Upgraded Bearings• Applied destabilizing cross‐coupling
stiffness of 2.13E+07 N/m to mid‐span• Without SFD, unstable • Both O‐ring SFD and ISFD can make the
rotor stable• No change in stiffness and damping of
ISFD over time-400
-200
0
200
400
0 400 800 1200 1600 2000 2
Sha
ft R
adiu
s, m
m
Axial Location, mm
f=4427.8 cpmd=-.4406 logdN=11456 rpm
-400
-200
0
200
400
0 400 800 1200 1600 2000
Sha
ft R
adiu
s, m
m
Axial Location, mm
f=3594.5 cpmd=1.0362 logdN=11456 rpm
OEM bearing w/out O‐ring SFD
Upgraded bearing (Flexure Pivot TPJ + ISFD)
-400
-200
0
200
400
0 400 800 1200 1600 2000S
haft
Rad
ius,
mm
Axial Location, mm
f=3624.7 cpmd=.3504 logdN=11456 rpm
OEM bearing w/ O‐ring SFD
No Subsynchronous Vibration (SSV) with Upgrade
With OEM bearing (Acceptance test)
Upgrade bearing
• Small SSV with O‐ring SFD• No SSV with ISFD
40 ‐
0 ‐
60 ‐
80 ‐
100 ‐
20 ‐Vibration pk
‐pk
(microns)
|
2‐Jun‐2013|
12‐Dec‐2013|
15‐Oct‐2014
Compressor B with Upgrade (10‐month span)Compressor B with OEM bearing
C-1252 01B VIBRAÇÕES e DESLOCAMENTO AXIAL DO COMPRESSOR D12
0
0
100
0
100
0
100
27,6019,2050,6058,60
25,1015,5085,4090,70
75,5072,5025,6012,90
21,9022,4014,1025,30
40 ‐
0 ‐
60 ‐
80 ‐100 ‐
20 ‐
Vibration pk
‐pk
(microns)
|
24‐Jul‐2012|
8‐Aug‐2013|
14‐Mar‐2013
Vibration Improvement (Comp A & B)• Compressor A
vibration dropped to less than half and maintained that level over time
• Compressor B vibration also down to below 50 µm from 90 µm (original) and kept the same level over time
Compressor A with OEM bearing
Compressor A with Upgrade (4‐month span)
__ Compressor DE X__ Compressor DE Y__ Compressor NDE X__ Compressor NDE Y
40 ‐
0 ‐
60 ‐
80 ‐
100 ‐
20 ‐
Vibration pk
‐pk
(microns)
|
8‐Mar‐2005|
30‐Dec‐2014
40 ‐
0 ‐
60 ‐
80 ‐
100 ‐
20 ‐Vibration pk
‐pk
(microns)
|
3‐Jun‐2013|
17‐Apr‐2014|
14‐Oct‐2014
Vibration Improvement (Comp C)• Compressor C was also
upgraded with a Flexure Pivot TPJ with ISFD technology
• Again, the vibration level decreased to below 30 µm with the upgrade and maintained that same level due to no change in bearing clearance and SFD performance over time
Compressor C (vibration trend over 10 years)
Compressor C with OEM bearing Compressor C with Upgrade
__ Compressor DE X__ Compressor DE Y__ Compressor NDE X__ Compressor NDE Y
Summary• Three reinjection compressor trains suffered
from excessive vibration over many years– Original configuration: point contact pivot tilt pad journal bearings with an O‐ring SFD
• The root cause was excessive pivot wearand degradation of the O‐ring SFD Pivot Wear mark on housing Fretting Damage
– Bearing bore increased
– Stiffness and damping changed over time
Summary• The compressors were retrofitted with optimized
Flexure Pivot tilt pad journal bearings with ISFD technology
• Operating exceptionally well– Since 2013– Low vibration levels
• 50% drop pk‐pk compared to OEM bearings• Do not grow over time
– No field balancing required so far (2 years)– No trips (continuous production)– No expensive bearing replacements– Overall efficiency increased by 1%
Lessons Learned• Increase in synchronous vibrations may be an indication of bearing clearance
increasing from pivot wear and/or change in O‐ring damper performance
• Pivot wear may accelerate over time from increasing imbalance due to deposits on impellers
• Without eliminating pivot wear, just replacing the worn bearing with new build of the same design is NOT a long‐term solution
• Proper bearing and damper selection and optimization can reduce or eliminate the likelihood of increasing vibrations and pivot wear
• Flexure Pivot technology is a proven design to eliminate pivot wear
• ISFD technology maintains performance over time
Feedback and Questions
Case Study: A Solution to Years of High Vibration Problems in Three Reinjection Compressor Trains Running at 33 MPa Discharge Pressure
Appendix: Damper Design Comparison
Shaft
Oil
Bearing
Oil
Damper Flow
Damper Flow
Anti‐Rotation Pin
CsCg
Squeeze Film (Outer Oil Film) – bearing whirls or orbits (not spins) in a precessional motion due to synchronous (unbalance) or non‐synchronous excitation, squeezing the oil and thus generating an oil film pressure, and subsequently a damping force. Flow can be axial too, depending on sealing.
Conventional SFD Integral Squeeze Film Damper
Shaft
Oil
Oil
Cs
CgNO Circumferential Flow
Damper Flow In/outof Orifices and Axial Gaps
‘S‐spring’
AuthorsJong Kim, PhD, (+1 262.506.3055 [email protected]) is a Senior Principal Engineer at Waukesha Bearings Corporation, headquartered in Pewaukee, Wisconsin (USA) and a Senior Consulting Engineer at Bearings Plus, Houston, Texas (USA), which is a business unit of Waukesha Bearings. Dr. Kim has overall responsibilities for rotordynamic analysis, bearing upgrades and bearing/seal technologies including Flexure Pivot bearings, ISFD technology and brush seals. Dr. Kim joined Waukesha Bearings in 2011. Prior to joining Waukesha Bearings and since 2001, he worked for KMC and Bearings Plus. Dr. Kim received his Bachelor of Science (Mechanical Engineering, 1985) from Busan National University and both his Master of Science (Mechanical Engineering, 1987) and his PhD (Mechanical Engineering, 1991) from KAIST (Korea Advanced Institute of Science and Technology). He has authored several papers and has been granted multiple patents on bearing technologies.
Marcio Felipe dos Santos (+55 092.3616.6614 [email protected]) is a Senior Maintenance Engineer at a major oil company in South America, in the Amazon. Since 2000, Mr. Dos Santos has provided maintenance in LPG plants and on reinjection compressors. Prior to joining his current company, he maintained drilling rigs. He has his Mechanical Engineering Degree from Universidade Federal of Rio de Janeiro (1979). Throughout his career Mr. Dos Santos has had maintenance duties.
Barry J. Blair (+1 262.506.3043 [email protected]) is the Chief Engineer at Waukesha Bearings Corporation, headquartered in Pewaukee, Wisconsin (USA). Mr. Blair has overall responsibilities for research & development activities at Waukesha, including new products and overseeing the refinement of bearing design tools and methods. Mr. Blair joined Waukesha Bearings in 1993 and has served in increasingly responsible engineering and technology roles. Mr. Blair received both his Bachelor of Science (Mechanical Engineering, 1990) and Master of Science (Mechanical Engineering, 1990) from the University of Virginia, completing requirements of both degrees concurrently. He has authored and coauthored several papers on the development of both hydrodynamic and active magnetic bearing technologies.