University of Southern QueenslandFaculty of Engineering and SurveyingInduced Draft (ID) Fan Lubrication SystemDesign Review and Proposed ModificationUpgrade at Callide C Power StationA dissertation submitted byStuart Bakerin the fulfilment of the requirements ofCourses ENG4111/4112 Research Projecttowards the degree ofBachelor of Engineering (Mechanical)28 October, 2010

AbstractAustralia’s coal fired power stations are the most efficient form of providingbulk base load power generation (i.e. electricity) to consumers. This is dueto Australia having an abundance of thermal coal reserves, which is the fuelused in coal fired power stations. Therefore it is extremely important thatthese power stations operate at maximum availability and reliability to ensurethe consumer receives cost effective and uninterrupted electricity.Callide C Power Station in Biloela Queensland is a 900 Megawatt (MW) coalfired power station that was commissioned in 2001. Unfortunately Callide CPower Station has been plagued with continuous operational and reliabilityproblems caused from the induced draft (ID) fans since initial commissioning.The ID fan problems have arisen from the bearing lubrication system whichprovides oil recirculation to the induction motor bearings and fan main shaftbearings. Consequently these issues have caused half-load unit (225 MW)run-backs and full unit (450 MW) trips over the past decade.This project’s aim is to analyse the ID fan lubrication system and thenidentify and define all root causes and their associated failure modes. Onceall root causes are identified through Root Cause Analysis (RCA) process,effective design solutions can be researched and evaluated so a proposedmodification design project can be finalised. This final design proposal willbe used to justify capital expenditure so implementation can occur in thenear future.i

University of Southern QueenslandFaculty of Engineering and SurveyingENG4111 & ENG4112 Research ProjectLimitations of UseThe Council of the University of Southern Queensland, its Faculty ofEngineering and Surveying, and the staff of the University of SouthernQueensland, do not accept any responsibility for the truth, accuracy orcompleteness of material contained within or associated with thisdissertation.Persons using all or any part of this material do so at their own risk, and notat the risk of the Council of the University of Southern Queensland, itsFaculty of Engineering and Surveying or the staff of the University ofSouthern Queensland.This dissertation reports an educational exercise and has no purpose orvalidity beyond this exercise. The sole purpose of the course pair entitled"Research Project" is to contribute to the overall education within thestudent’s chosen degree program. This document, the associated hardware,software, drawings, and other material set out in the associated appendicesshould not be used for any other purpose: if they are so used, it is entirely atthe risk of the user.Prof Frank BullenDeanFaculty of Engineeringii

CertificationI certify that the ideas, designs and experimental work, results, analyses andconclusions set out in this dissertation are entirely my own effort, exceptwhere otherwise indicated and acknowledged.I further certify that the work is original and has not been previouslysubmitted for assessment in any other course or institution, except wherespecifically stated.Stuart D. BakerStudent Number: 0050068776SignatureDateiii

AcknowledgementsThis project was conducted under the principle supervision of Mr BobFulcher, Faculty of Engineering and Surveying, University of SouthernQueensland.I would also like to acknowledge and thank the technical engineering staffand Callide C maintenance staff at Callide Power Station for their continuoussupport, mentoring and assistance.Stuart D. Bakeriv

GlossaryAC:Alternating CurrentAS:Australian StandardsASME:American Society of Mechanical EngineersCNC:Computer Numerical ControlCO2:Carbon DioxidecSt:CentistokesDC:Direct CurrentDE:Drive endDN:Nominal DiameterDNB:Departure from nucleate boilingEP:Extreme PressureFD:Forced DraftHHI:Hyundai Heavy IndustriesHV:High VoltageICMS:Integrated Control Management SystemID:Induced DraftIHI:Ishikawajima – Harima Heavy Industrieskg/s:Kilogram per secondkPa:Kilopascal 1000 PascalskV:Kilovolt 1000 voltskW:Kilowatt 1000 wattsL/min:Litre per minutev

limetre squared per secondm/s:Metre per secondm3/s:Metre cubed per secondMPa:Megapascal 1000000 PascalsMRC:Maximum Continuous RatingMW:Megawatt 1000000 wattsMWh:Megawatt hour 1000000 watts for one hourNDE:Non-drive endOEM:Original Equipment ManufacturerPA:Primary AirP&ID:Piping and Instrumentation DiagramPF:Pulveriser FuelRCA:Root Cause Analysisrpm:Revolution per minuteV:VoltVI:Viscosity Indexµm:Micrometre (micron) 1 x 10-6 mvi

SymbologyACross-sectional area[m2]DPipe inside diameter[mm]D1Orifice inside diameter (DE)[mm]D2Orifice inside diameter (NDE)[mm]fFriction factorFVoltage frequency[Hz]gGravitational acceleration[m/s2]hPotential head[m]h1Major head loss[m]h1mMinor head loss[m]HTotal head loss[m]KLoss coefficient factorLLength of pipe[m]nRevolutions per minute[rpm]pPressure[kPa]PNumber of motor winding polesQVolumetric flow rateReReynolds numberSrpmSynchronous revolutions per minute[Srpm]vKinematic viscosity[m3/s][L/min]vii

VDisplacement per revolution[cm3/rev]V&Fluid flow rate[m3/s]VFluid velocity[m/s]ηVolumetric efficiency[%]viii

Table of ContentsAbstractiLimitations of mbologyviiTable of ContentsixList of FiguresxiiList of TablesxivList of AppendicesxvChapter 1 – Introduction11.1Project Objectives.11.2Background .21.3Reliability Issues with Callide C Power Station.41.3.1ID Fan Lubrication System Reliability Issues.61.3.2Pressure Monitoring versus Flow Monitoring.8Chapter 2 – Literature Review102.1General operation of Callide C Power Station .102.2Characteristics of Supercritical Boilers .112.2.1Supercritical versus Subcritical.11ix

2.3Callide C Boiler Draft System .152.4Boiler Draft Efficiency and Design .172.4.1Fan Margins .172.4.2Fan Types .182.5Tribology of Rotating Machines .212.5.1Bearing Design .212.5.2Callide C ID Fan and Electric Motor Bearings .272.5.3Lubrication Theory.272.5.4Fluid Film Lubrication .292.5.5Callide C ID Fan Lubrication.29Chapter 3 – Methodology3.131Further Research and Appraisal.313.1.1Bearing Selection .313.1.2Lubrication Selection .353.2Root Cause Analysis Methodology .383.2.1Pump Internal Wear.403.2.2Airlock in Stand-by Pump .453.3.3Complex System Design .483.3.4Flow Transmitters Removed.513.3.5Incorrect Pressure Trip Values .53Chapter 4 – Effective Design Solutions4.155Research and Evaluation of Solutions .554.1.1Particle Contamination .554.1.2Internal Tank Design .57x

4.1.3Simplify System Design.604.1.4Suitable Flow Transmitters .614.1.5Pressure Alarm and Trip Values.69Chapter 5 – Modification Design Proposal5.170Design Considerations.705.1.1Filter Element Selection.705.1.2Internal Tank Modifications.705.1.3Fixed-size Orifices .725.1.4Accurate Flow Transmitters.755.1.5Modified Pressure Alarm and Trip Values .785.2Modification Design Overview .82Chapter 6 – Conclusion836.1Project Outcomes .836.2Further Work.84References85xi

List of FiguresFigure 1.1: Callide C ID fan and induction motor .5Figure 1.2: Lubrication system P&ID.6Figure 2.1: Boiler circulations methods .12Figure 2.2: Furnace tube size and temperature comparison.13Figure 2.3: Subcritical versus supercritical steam .14Figure 2.4: Callide C – Balanced draft PF boiler .16Figure 2.5: Typical draft-loss pressure diagram for a PF Boiler .16Figure 2.6: Centrifugal fan flow path .18Figure 2.7: Axial fan flow path.19Figure 2.8: Efficiency comparison of various fan types .20Figure 2.9: Typical ball bearing cross sections .23Figure 2.10: Hydrodynamic journal bearing cross section .25Figure 2.11: Hydrostatic journal bearing cross section .25Figure 3.1: Induction motor bearing selection graph .33Figure 3.2: Causal Tree Analysis .39Figure 3.3: Exploded assembly drawing of gear pump .41Figure 3.4: Abrasion wear on gear pump outer casing .44Figure 3.5: Abrasion wear on gear pump shaft bearing .44Figure 3.6: Lubrication pump P&ID .45Figure 3.7: Individual pump pressure test .46Figure 3.8: Lubrication pump P&ID with local test pressure gauge.47Figure 3.9: ID fan lubrication tank internal arrangement .48xii

Figure 3.10: Lubrication system P&ID.49Figure 3.11: Two throttle valves located inside cabinet.50Figure 3.12: Adjustable orifice.50Figure 3.13: OEM Hedland flow transmitter .51Figure 4.1: Variable area measurement principle .63Figure 4.2: A typical Yokogawa variable area flow transmitter.64Figure 4.3: Operation principle of an oval gear flowmeter.65Figure 4.4: A typical Macnaught positive displacement flow transmitter .66Figure 4.5: Coriolis mass operating principle .67Figure 4.6: A typical Yokogawa coriolis mass transmitter .68Figure 5.1: Proposed de-aeration screen design .71Figure 5.2: Fixed-size orifice 3D model.74Figure 5.3: Fixed-size orifice Matlab script.74Figure 5.4: Modified lubrication system P&ID .77xiii

List of TablesTable 2.1 – Shell Tellus S 46 oil properties.30Table 3.1 – Comparison of gear pump specifications .41Table 3.2 – Comparison of oil particle contamination .43Table 3.3 – Comparison of flow transmitter specification.52Table 3.4 – Comparison of flow rate and pressure protection values .53Table 4.1: Comparison of cleanliness code .56Table 4.2 – Advantages and disadvantages of variable area flowmeters .63Table 4.3 – Advantages and disadvantages of positive displacementflowmeters.65Table 4.4 – Advantages and disadvantages of coriolis mass flowmeters .68Table 5.1 – Comparison of Macnaught flow transmitter specifications .76Table 5.2 – OEM flow rate protection values .76Table 5.3 – Alarm and trip conditions for head loss calculations .78Table 5.4 – Fan main shaft bearing protection values .80Table 5.5 – Induction motor bearing protection values .81Table 5.6 – Comparison of pressure protection values.82xiv

List of AppendicesAppendix A – Project Specification.87Appendix B – Engineering Design Drawings .88Appendix C – Flow Meter Specifications .90Appendix D – Matlab Programs .94xv

Chapter 1 – Introduction1.1Project ObjectivesThe aim of this project is to conduct an engineering appraisal on theoperational and reliability issues in the existing Callide C Power Stationinduced draft (ID) fan lubrication system, and recommend a cost effectiveand reliable design for a future modification project. Project objectives are:1. To describe the general operation of Callide C Power Station withspecial attention to boiler section pertaining to induced draft (ID) fanoperations.2. Research theory on lubrication and bearing systems in rotatingequipment.3. Investigate further information on the different types of bearings andlubrication used in heavy electric drives and axial fans.4. Identify the operational and reliability failure modes in the existing ID fanlubrication system which is causing protection alarms and trips.5. Research and evaluate effective design solutions to prevent reoccurringoperational and reliability issues in the existing ID fan lubrication system.6. Prepare a modification design recommendation that will make certainthe lubrication system is designed for correct functionality and long-termreliability.7. Submit an academic dissertation on the engineering researchconducted and proposed design modification to the ID fan lubricationsystem.1

1.2BackgroundThe rapid technological advances in todays electrical hardware andsoftware has lead to our civilization becoming increasing dependant onelectrical power, also commonly known as electricity.This increasingdependence on electricity is forcing power generation companies allthroughout Australia to produce reliable, cost effective and safe electricityfor their consumers. Electricity in Australia is commonly produced in largescale using a four types of electrical power generation methods, theseinclude: Coal fired power stations Gas fired power stations Hydro driven power stations Large wind turbine farmsOf the four types listed above, coal fired power stations are the mostfavoured in Australia as they are the most efficient method of bulk baseload power generation.Australia is also very fortunate to have anabundance of thermal coal reserves, which is the fuel used in coal firedpower stations. In Australia, base load coal fired power generation unitsrange in size from 30 megawatt (MW) units which were built in 1960’s allthe way to 750 MW units built in recent years.Bulk base load power generation is paramount in sustaining a stableelectrical network (i.e. grid) system as it ensures continuous uninterruptedsupply of electricity to the consumer. This stability is built from base loadpower stations which are designed to run at constant operating output, allday and every day with the exception of planned outages.2

A major disadvantage when building electrical network systems on largebase load units ranging from 350 MW to 750 MW is when a powergeneration unit unexpectedly trips offline to the grid, it leaves a big gap inthe transmission supply which can lead to voltage and frequency instability.Such events can cause forced network load shedding, where electricityload is tripped from the grid.An example of this, take Queensland’s 2009 peak electricity consumption of8699 MW (Department of Mines and Energy 2010) and Kogan CreekPower Station in Chinchilla Queensland, a single 750 MW base load unit.When 750 MW is tripped offline from the state grid of 8699 MW,approximately 9% of available supply electricity is lost and this gap must beinstantaneously supplied from other power stations connected to the grid. Ifsuch an event occurs simultaneous with multiple power stations trippingoffline, the entire grid can become unstable which can lead to regional andstate blackouts. To avoid grid instability and power blackouts it is crucialthat power stations are designed, maintained and operated to run atmaximum availability and reliability.However, all power generation stations from time to time suffer fromreliability issues mainly caused from inadequate design factors, lack ofmaintenance procedures and skills, and operator error. To combat suchfactors that cause reliability issues in large base load power stations. It isgenerally the responsibility of the technical engineering staff to adopt acontinuous improvement culture which investigates and analyses keyperformance and reliability issues. Such investigations, better known asRoot Cause Analysis (RCA) are very popular in today’s industry as thisprocess identifies root causes in machinery failure and poor reliability.Once the root causes are identified and understood, the technicalengineering department is able to proceed with modification concepts anddesigns to prevent reoccurrence of the problem causing poor reliability.3

1.3Reliability Issues with Callide C Power StationCallide C Power Station is situated 18 kilometres east from the township ofBiloela, approximately 120 kilometres west of Gladstone Queensland.Callide C Power Station is quite famous as it was the first supercritical coalfired power station built and commissioned in Australia, in the year 2001(Power Technology 2010). Callide C has a total generation capacity of 900MW which consist of 2 x IHI supercritical power generation units capable of450 MW per unit. This amount of continuous generation is the basis forCallide C being an integral base load power generation station inQueensland’s state electricity grid.Each of the two 450 MW units consists of two Howden Variax Dual StageAxial Flow Induced Draft (ID) Fans. Each fan is powered via an alternatingcurrent (AC) 6600 volt - 3 phase 6 pole synchronous induction motormanufactured by Hyundai Heavy Industries (HHI). The four ID fans havesuffered from reliability and operational issues from commissioning stage in2001, which has caused half load unit run-backs and full load unit trips. Ahalf load 225 MW unit run-back occurs when one ID fan trips and a full unit450 MW trip occurs when both ID fans trip. This amount of generationcapacity tripping off the grid unexpectedly has the potential to causefrequency and voltage instability of the state grid at peak times.4

Figure 1.1: Callide C ID fan and induction motorThe ID fan problems have arisen from the bearing lubrication system whichsupplies oil recirculation to the induction motor bearings at 15 – 16 L/minand fan main shaft bearings at 15 - 18 L/min. The lubrication system has 2x 100% duty cycle lubrication pumps, so one pump is always in standby.Prior to 2004, there were two flow transmitters on each lubrication line(refer to Figure 1.2) that were used to provide alarm and trip values forequipment protection. These values were used by the Integrated ControlManagement System (ICMS) for the following parameters: Low oil flow alarm Oil flow for fan start condition Low oil flow fan trip Low oil flow start stand-by lubrication pump5

Figure 1.2: Lubrication system P&ID1.3.1 ID Fan Lubrication System Reliability IssuesTo balance the required oil flow rates between the induction motor bearingsand fan shaft bearings, two throttle valves and two adjustable orifices areused to obtain required flow rates. Due to equipment and personnel safetyfactors, the oil flow balancing set-up can only be performed when thelubrication system is off-line. This is when the oil is at ambient temperature25 C, not operating temperature between 50 – 60 C. Shell Tellus S46 isthe oil used in the ID fan lubrication system with a kinematic viscosity of100 mm2/s at 25 C and 20 mm2/s at 60 C. This shows the oil viscosity-totemperature relationship is logarithm and oil viscosity is a factor of 5 timesless from ambient temperature to operating temperature.6

The problems became apparent after the oil flow balancing set-up wascompleted and the ID fan returned to service and eventually oil temperaturereached 60 C and even higher on hot summer days. The oil viscosity anddensity would decrease causing the flow transmitters to read a lower valuethen previous for oil flow rate. This would then create a low oil flow alarm inthe ICMS and then start the stand-by pump in an attempt to increase oilflow rate to above alarm value. However, due to poor design factors thestand-by pump was unable to self-prime and make system pressure.Therefore no additional oil flow was created and eventually the ID fan wouldtrip on low oil flow.To combat this issue, the operation and maintenance teams wouldeventually get the stand-by pump to prime and return the ID fan back toservice with the both lubrication pumps running simultaneously. Having twopumps running ensured oil flow rate was much greater than alarm and tripvalues, therefore preventing future alarms and trips. However this solutioncreated another problem, if a lubrication pump dropped performance orcompletely failed, the oil flow rate would again drop to alarm and trip value,eventually tripping the ID fan.The flow transmitters also suffered from other failure modes such as: (1) oilleaking internally into the electrical circuitry causing loss of analog outputsignal to the ICMS, (2) could not be site calibrated to suit different fluidviscosity and density, and (3) contamination of small particles in the oilcaused sticking of the internal components. To rectify this problem, in 2004an engineering modification was implemented which replaced the four flowtransmitters with four pressure transmitters in an attempt to reduce thenumber of ID fan alarms and trips. The alarm and trip pressure protectionvalues were supplied from the fan supplier, Howden Australia.7

This modification unfortunately suffered from the same fate as the flowtransmitters. As the oil viscosity and density decreased when oil reachedoperating temperature, the flow resistance also decreased and thereforereducing line pressure to both the induction motor bearings and fan mainshaft bearings.This drop in line pressure would again create low oilpressure alarms and trips, eventually tripping the ID fan. So unfortunatelythis engineering modification did not prove successful and did not eliminatethe root causes.Some work was recently done to decrease the pressure alarm and tripvalues which did prevent oil pressure alarms and trips.However thissolution still did not address the issue of the stand-by pump not makingsystem pressure when started. This aspect is very important as operationsare required to change-over lubrication pump from duty cycle to stand-bycycle every month to prove pump performance and integrity.So everymonth when this operational activity occurs, the stand-by pump can notmake system pressure which leads to a critical machine operating with nostand-by lubrication pump.1.3.2 Pressure Monitoring versus Flow MonitoringPressure is not the desired online monitoring parameter for the ID fanlubrication system as the ‘Original Equipment Manufacturer’ (OEM) plantmanual gives oil flow rates for operation conditions, not oil pressures. Thereason for oil flow rate as the preferred monitoring parameter is becausethe induction motor bearings and fan main shaft bearings are oil bathconfiguration. Therefore oil flow to the bearing inlet is only slightly aboveatmospheric pressure and the only ‘head’ resistance created is fromdynamic friction of pipes, fittings and orifices. This means pressure willdramatically vary with oil temperature however flow rate will always remainconstant.8

It must also be further emphasised when pressure is increased, thecorresponding flow does not necessary increase.If the cross-sectionalarea of an oil line is decreased upstream of the pressure transmitter byreducing orifice size, then line pressure will increase due to an increase in‘head’ resistance. However, oil flow rate will not increase, instead oil flowrate can actually decrease causing possible bearing overheating and failure.The actual decrease in oil flow rate is dependant on system design, pumptype and pump performance.9

Chapter 2 – Literature Review2.1General operation of Callide C Power StationCallide C Power Station utilises 2 x 450 MW supercritical coal-fired boilers,which achieves higher thermal efficiency than conventional subcriticaldrum-type coal-fired technology. Callide C obtains coal from the adjacentAnglo Coal Callide Mine and receives water from adjacent Callide Damwhich draws water via an overland pipeline from Awoonga Dam nearGladstone. General information for Callide C Power Station is shown below:GENERALCommissioned2001Capacity900 MWUnits2 X 450 MWTransmission275 kVFuelBlack Thermal acturerIHI (Tokyo, Japan)Height42 mOperating Temperature1400 CSteam Pressure25 100 kPa ( 250 bar)Steam Temperature566 CCHIMNEYHeight230 mFlue Gas Temperature135 C10

2.2Characteristics of Supercritical BoilersSupercritical boiler technology was introduced to the power industry in theearly 1960’s.Since this time, there have been many innovative boilerdesign configurations and features introduced to reduce capital andoperating costs, simplify operation and maintenance, and increase reliability.2.2.1 Supercritical versus SubcriticalSupercritical ‘once-through’ boilers are constructed differently from theconventional subcritical ‘drum’ boilers. Primarily in two areas which is theboiler furnace and boiler drum (refer to Figure 2.1). For comparison, in adrum type boiler which operates at subcritical pressures and temperatures,large diameter furnace tubes are used to minimise flow resistance so thatsufficient amount of stream and water can flow through the tubes by naturalcirculation.For that reason, drum type boilers by design permit highcirculation rates so the water passing through the tubes at any stage nevercompletely evaporates to steam, so it remains as saturated steam (steamwater mixture) with latent heat addition. This ensures a liquid film (wet wallflow) is maintained on the tube wall inner diameter so that the departurefrom nucleate boiling (DNB) and/or dryout (dry wall flow) does not occur inall conditions of operations. Nucleate boiling is vital as it provides a highheat transfer coefficient so all furnace tubes remain at essentially thesaturation temperature for the operating temperatures of the boiler,preventing tube overheating and puncture.Whereas a supercritical boiler does not use a head drum to promote anatural circulation circuit required for latent heat addition. Steam flow isforced by the boiler feedwater pump as it does not rely on the densitydifference between steam and water to provide proper circulation andcooling of the furnace tubes. As a result, it can be operated at supercritical[ 220 bar] pressures.11

Figure 2.1: Boiler circulations methodsIn the absence of natural circulation and water progressively converting tosteam without boiling, all the water turns into superheated steam in thefurnace tubes of a supercritical boiler. However, small separating vesselsin stages along the furnace tubes are required to separate the waterfraction and recirculate the water when the unit operates at different loadsand pressures.As illustrated in Figure 2.2, the furnace tubes in asupercritical boiler can also be smaller in diameter as natural circulation isnot needed.This reduces liquid inventory in the furnace tubes whichincreases boiler load dynamic response.12

Figure 2.2: Furnace tube size and temperature comparison

vi m: metre M: Metric ml: Millilitre mm: Millimetre mm 2/s: Millimetre squared per second m/s: Metre per second m3/s: Metre cubed per second MPa: Megapascal 1000000 Pascals MRC: Maximum Continuous Rating MW: Megawatt 1000000 watts MWh: Megawatt hour 1000000 watts for one hour NDE: Non-drive end OEM: Original Equipment Manufacturer PA: Primary Air