1

40th

TURBOMACHINERY SYMPOSIUM

CASE STUDY

‘BALANCE INSTABILITY AND VIBRATION ON A 6 MW

INDUCTION MOTOR ROTOR’Authors

Gunnar Porsby (ABB, Sweden) Sameer S. Patwardhan (Bechtel OG&C Corp., Houston, Texas)

Siddharth (Sid) Shinde (Chevron, Houston, Texas)Barry Wood (Chevron, Richmond, California)

338

2

Induction Motor at Test Stand

339

3

Stabilizer Overhead Compressor• Centrifugal compressor (with ‘side stream’)• Two stage compression used for mixing

vapor/gas from ‘stabilizer’ column into feed gas stream

• Speed increasing gearbox• 6 MW induction motor driver with variable

frequency drive• Large variation in inlet flow• Complicated control system• No spare compressor

340

4

Induction Motor Driver• 6 MW induction motor

• 6.6 kV 3 Ph 60 Hz

• Coupled with variable frequency drive (VFD)

• Motor designed per API 541 and IEC (hybrid standard)

• Routine and complete run tests including heat run test, unbalance response test, mechanical run test, over-speed test etc.

341

5

Motor Design: Squirrel Cage Induction Rotor

Rotor core(Laminated steel)

Spider shaft(Homogenous forged steel)

Squirrel cage(Copper)

Unbalance weights

Complete rotor

The sheetlaminations are

shrunk on tothe rotor shaft

and compressedin axial direction

Rotor assembly AND

342

6

Motor Design: Sleeve Bearings and Stiffness Chain

Stiff shaft design with unbalance force pathway

Sleeve bearings with oil film for damping and stiffness

343

7

Separation Margin and Balancing

Normally 4 pole motors are run below first critical speed with sufficient separation margin(sub-critical operation)

Motor speed range:1350-1890 rpm

2-plane balancing at reducedspeed is often sufficientfor sub-critical operation

344

8

API 541 Vibration Requirement

Shaft vibration limits(Relative bearing housing using non-contact probes)

Bearing housing vibration limits(Using bearing-mounted velocimeters)

345

9

Factory Acceptance Test (FAT)

• High vibration levels during over-speedtest

• Vibrations above normal for the machine type at running speed (1800 rpm)

• High vibrations during over-speed test led to bearing failure

346

10

Factory Acceptance Test Results

API Criteria is 1.5 mills = 38µm

Bearing Vibrations during Overspeed Test

Failed NDE Bearing during 4 hour Mechanical Run Test

347

11

Investigation After FAT• Assumption:

– The vibrations are caused by unbalance in the rotor due to initial settlement during the first heat run and/or over speed test

• Decisions– Residual Unbalance Check � Re-Balance at

Low Speed (1000 rpm) � New Test

• Result– Vibrations still not meeting the requirements

348

12

FAT Results After Re-Balancing

Bearing Vibration Results After Re-Balancing

Balancing p lane

Re sidual unbala nce

[kg mm ]

Residual unbalance ma ss [g ]

Re sidual unbala nc e

[kg mm ]

Residual unbalance ma ss [g ]

AP I 541 requirem ent re sidual unbalance

m ass [g ]DE 33 117 2 6 .2 7.8

ND E 47 169 1 2 .1 7.8

After R e-balancingFAT

Residual Un-Balance After Re-Balancing

349

13

More Investigations…• Assumptions:

– Other effects than only settling.

• Decisions:

– Check balancing state again, this time at different speeds

• Findings

– The balancing state at 1000 rpm had changed again

– Balancing state was also changing with speed

• Conclusions

– Unbalance of the rotor was not caused by only settlings in the rotor

Speed [rpm]

Residual unbalance[kg mm]

Residual unbalance mass [g]

Residual unbalance[kg mm]

Residual unbalance mass [g]

249 7.4 26.3 18.8 67.1500 3.8 13.7 20.2 72751 1.9 6.7 23.2 82.7

NDE DE

350

14

A Theory Was Formulated…• Measurements showed that individual sheets in the

lamination were skewed/buckled• Centrifugal force acts to ‘straighten’ the sheet metal plates

which could lead to changed balancing state

Rotor Core

Centrifugal Force

Sheet metal plateat ‘low’ speed

CentrifugalForce

Axis of rotation

Sheet metal plateat ‘high’ speed

Rotor core measurements

351

15

Corrective Actions• Attempt to recompress the rotor core lamination to make it

perpendicular to the rotor centre line

• Result:

– Improved vibration but still not meeting requirement, balancing state still changes with speed

• Conclusion:

– Recompression not working due to high friction betweenrotor core and spider shaft

Rotor Core Compression

‘Press-rings’ ‘Pull-rods’Speed [rpm]

Residual unbalance[kg mm]

Residual unbalance mass [g]

Residual unbalance[kg mm]

Residual unbalance mass [g]

249 20.2 72.1 17.1 61.2500 19.7 70.5 20.1 71.7750 17.4 62.1 22.3 79.8

1000 14.4 51.3 25.1 89.7

NDE DE

Balancing plane

Residual unbalance[kg mm]

Residual unbalance mass [g]

API 541requirement residual unbalance

mass [g]DE 1 4.8 7.8

NDE 2 7.6 7.8

Balancing @ 1000 rpm

After recompression:

352

16

Test After Operating SpeedBalancing

• Solution for the problem:

– Balancing of rotor at full speed (settling effects within balancing)

• Results

– After full speed balancing vibration levels are within required limits

• Forced unbalance tests for final verification

NDE Y-direction

API Criteria is 1.5 mills = 38µm

NDE X-direction

API Criteria is 1.5 mills = 38µm

353

17

FAT: Residual Unbalance TestPer API 541

• The residual unbalance in the rotor was found to be above maximum allowable residual unbalance

• It was now decided that a new rotor should be manufactured

354

18

Root Cause Analysis Findings

High vibrationsbearing failure

Balancing statechanges with speed

Skewed/buckledsheet metal plates

Insufficient compression of rotor core due to malfunction in thecooling process of rotor core during manufacturing process

Problem:

Analysis:

Manufacturing of new rotor with changed manufacturing process for the core shrinking

Resolution:

355

19

Resolution: Manufacture New Rotor• New process for shrinking the core onto the

shaft:

– Cooling from the top down to get axial pressure on the entire length of the rotor core

Fans to cool ‘TopPart’ of rotor core

‘Insulation’ around bottom part of rotor core

Rotor core cooling

356

20

New Rotor: Residual Unbalance& Vibration Tests

• Residual unbalance test was successfully passed

• Vibration levels were lowercompared to old rotor, normal for the machine type and within API 541 prescribed limits

Old rotor

New rotor

357

21

CONCLUSION• It is very difficult to correct a rotor after a

distorted cooling or skewed lamination fit on the rotor core

• Trial and error attempts to diagnose and repair this type of rotor problem can be very time consuming and without guarantee of success.

• In a schedule oriented environment, it is important to have all the necessary resources involved to quickly determine if the problem can be corrected and a quality machine assured. Sometimes it may be necessary to move in parallel in attempting to repair the rotor and preparing to manufacture a new rotor.

358

T40LTCDSLecturesT40LECT1LECT2LECT3LECT4LECT5LECT6LECT7LECT8LECT9/

LECT10LECT11LECT12

Turbo40TutorialsTUTT1TUTT2TUTT3TUTT4TUTT5TUTT6TUTT7TUTT8CONSEQUENCES OF POOR INLET FILTRATIONErosionFoulingCorrosion

FILTRATION CHARACTERISTICSFiltration MechanismsFilter Efficiency and ClassificationFilter Pressure LossFilter Loading (Surface or Depth)Face VelocityHigh Velocity SystemsLow Velocity Systems

Water and Salt Effects

COMPONENTS OF A FILTRATION SYSTEMWeather Protection and Trash ScreensAnti-icing ProtectionInertial SeparatorsMoisture CoalescersPrefiltersHigh Efficiency FiltersSelf-Cleaning FiltersStaged Filtration

OPERATING ENVIRONMENTCoastal, Marine, or OffshoreLand Based EnvironmentDesertArcticTropicalRuralLarge CityIndustrial Area

Temporary and Seasonal Contaminant SourcesSite LayoutSite Evaluation

LIFE CYCLE COST ANALYSISLife Cycle Cost BasicsConsiderations for an Inlet Filtration SystemPurchase Price/Initial CostMaintenance CostAvailability/Reliability of Gas TurbineGas Turbine Degradation and Compressor WashingPressure LossFailure/Event Cost

SUMMARY

TUTT9

Turbo40CaseStudiesCaseT1CaseT2CaseT3CaseT4In-house Engineering for Resolution of Chronic 4th Stage High Discharge Pressure Limitation on Carbon-dioxide Reciprocating Compressor Problem StatementAnalysisSlide Number 4AnalysisAnalysisAnalysisAnalysisAnalysisAnalysisSlide Number 11Conclusion

CaseT5CaseT6Advanced Vibration Analysis on Gear Box Train and vibration elimination ���by ��Udayashankar P. Eng., MBA,CMRP�Suncor EnergyMachine Train configuration��Brief History��Problem description��Analysis��Rectification��ConclusionsMachine Train ConfigurationSlide Number 4Unit DetailsBrief History-7K-20 Gear box Brief History ( Continued)Problem DescriptionSlide Number 9AnalysisAnalysisAnalysisAnalysisRectification- Action PlanEfforts MadeSlide Number 16Animation- ODS of the Gear BoxSlide Number 18Slide Number 19Slide Number 20Slide Number 21Slide Number 22Slide Number 23Gear Box Vibration Before/ After correctionConclusions

CaseT7CaseT8Investigation of Engine Vibration for Natural Gas �Gathering and TransmissionBackground16-Cylinder Natural Gas Engine Mounted on Foundation at Another LocationCompare to 12-Cylinder Natural Gas Engine Mounted on SkidEquipment for Case HistoryEngine – Compressor SystemVibration Measurement on Damper End of EngineVibration Measurement at�Middle of Engine FrameSummaryModel for Operating Deflection Shape (ODS) MeasurementODS – Iso ViewODS – Top ViewODS – End ViewODS – End ViewRecommendationsFinite Element Analysis (FEA)Example of Kick Brace With �Insufficient Stiffness (Different Unit)Proposed Modifications – Add Gussets to Skid Under EngineEngine Skid ModificationsAfter Modifications Performed

CaseT9CaseT10Slide Number 1ObjectivesContentsTurbo-Expander - ApplicationTurbo-Expander – ComponentsThe Beginning of ProblemsFailure Modes ExperiencedFailure Mode 1 – Axial Shuttling (Surge Failure Z12)Failure Mode 1 – Axial Shuttling (Surge Failure Z12)RCA Work/ CA CompletedFailure Mode 1 - Current Status Unit #1 TECFailure Mode 2 – Transfer Function ChangedAMB/Rotor Dyn TF Measuement In FieldFailure Mode 2 – Transfer Function ChangedA1 - Typical Transfer Function PlotsA2 – Unit #1 TF Change at High FrequencyA3 – Unit #1 Controller Modified to Counter TF ChangeB1 – Unit #2 TEC Unstable Vibration following TF ChangeB2 – Unit #2 TF Changed at Low FrequencyB3 – Unit #2 Machine Center Section Root CauseSummarySlide Number 22

CaseT11Slide Number 1Speed Signal Deterioration at High Speeds in Electronic Governor and Trip Systems �40th Turbomachinery Symposium Case Study��BackgroundDrawing of Bracket, Speed Gear and Probe ConfigurationAxial View of Speed Gear, Probes and BracketPlan View of BracketBackgroundSystem CharacteristicsSystem Characteristics (cont’d)Problems Appear. . .And Disappear . . .Slide Number 12Speed Signal IssuesSpeed Signal IssuesBracket Deformation Due to Thermal StressSpeed Probe Voltage OutputSpeed Signal IssuesSignal Voltage ReductionSignal Strength Reduction SummaryLessons LearnedLessons Learned (cont’d)Lessons Learned (cont’d)Lessons Learned (cont’d)DisclaimerBackup SlidesCalculation of Approximate Signal LossProbe Test ResultsShop Test DataSignal Before and After Patch

CaseT12Slide Number 1OutlineIntroductionProcess OverviewProcess OverviewCompressor Design & ConstructionReverse Rotation EventsCause of Reverse RotationCause of Reverse Rotation Cause of Reverse RotationPossible Impacts of Reverse RotationMitigating Actions. Phase 1 CompressorsMitigating Actions. Phase 2 CompressorsSlide Number 14Lessons LearnedConclusionsSlide Number 17Slide Number 18Slide Number 19

CaseT13“Beating” Effect Caused by Two Closely Spaced Mechanical Frequencies Observed on Two-Shaft, Gas Turbine Drive �Two Shaft Gas Turbine ConfigurationCross section of a similar two-shaft gas turbinePhoto of Gas Turbine/Compressor EnclosurePhoto of Gas Turbine-CompressorBackground InformationVibration Response AnalysisBeating IssuesResulting Graphs: Set 1Resulting RMS Graphs: Set 2Investigation of System and AnalysisZoom Analysis Results from PT Speed SweepFrequency AnalysisThree (3) similar gas turbines at the site were running at the time of this comparative analysis. Here is a plot showing GP vibration at various locations along the three engines analyzed. Note: Turbine B below is the engine described in the case study. Notice that for some unknown reason it transmits the highest level of GP vibration to the power turbine end.SolutionConclusions and Lessons LearnedQuestions?

CaseT1440th Turbomachniery Symposium�Abstract Slide Number 2Slide Number 3Slide Number 4Slide Number 5Slide Number 6Slide Number 7Slide Number 8Hammering Test for PedestalSlide Number 10Slide Number 11Slide Number 12Slide Number 13The eigenvalue problem is solved.Slide Number 15Slide Number 16

CaseT15 (2)Thermoplastic Labyrinth Seals in Centrifugal Compressors -�15 years of experiencesThermoplastic Labyrinth SealsThermoplastic Labyrinth SealsIntroductionLabyrinth SealsLabyrinth Seals – Compressor SealsLabyrinth SealsLabyrinth SealsMetallic Seal Rubs Polymer Seal RubsThermoplasticsThermoplasticsThermoplasticsThermoplasticsThermoplasticsDMA Plot for Various ThermoplasticsThermoplasticsCLTE PlotRelative Thermal PropertiesThermoplasticsPEEK Chemical ResistanceTorlon Chemical ResistanceEngineeringEngineeringEngineeringTensile Strength vs. TemperatureTensile Modulus vs. TemperatureCompressive Strength vs. TemperatureUpgrade Payback CalculationsUpgrade Payback CalculationsCase HistoriesCanadian Ethylene PlantCanadian Ethylene PlantCanadian Ethylene PlantCanadian Ethylene Plant 1996 outageCanadian Ethylene Plant 1996 outageCanadian Ethylene PlantTorlon Tooth Scrapping ToolShaft Seal as removed after 4 and 5 year runs – 9 year total run time – reinstalled will run another 6 yearsEye Seal as removed after 4 and 5 year runs – 9 year total run timeCanadian Ethylene PlantCanadian Ethylene PlantTexas Ethylene PlantDuPont Cracked Gas TrainDuPont Cracked Gas TrainTexas Ethylene PlantTexas Ethylene PlantNew Polymer Seal–Newly Coated RotorNew Polymer Seal–Newly Coated RotorTexas Ethylene Plant �Used Booster Compressor Shaft SealTexas Ethylene Plant �Used Charge Gas Compressor eye SealTexas Ethylene PlantTexas Ethylene Plant �ConclusionsConclusionThermoplastic Seals

TurboDG_TurboDGTSC_Layout 1Turbo40TACProfessional Staff_Layout 1Turbomachinery Laboratory