The Crw-y-Cawl hydroelectric power station supplies 300MW to a metal reprocessing plant, Penrhys Special Steels. The steel plant operates a maximum of 16 hours/day, six days/week, consuming 250 MW. Excess capacity is sold for grid distribution. Apart from late summer, the power station can run 24 hours/day. In late summer the lag from lighter precipitation means that only the steel plant's demand can be satisfied.
The majority of operation is fully automatic. There are few staff on site, and when breakdown occur, Penrhys Steel's support team can react, but only if time allows, because steel production is the first priority.
Background
A major breakdown last year led to severe damage to shafts and one of the alternators. The problem was traced to inadequate lubrication of an alternator bearing, which was fitted with an automatic greasing can. The cost of repair was around £120,000 - it was debatable whether the plant was worth repairing at all. The records, which are quite good paper based records, show that the majority of failures are related to controls but are usually rectified at little expense in less than four hours.
Objective
The main objective of this report is to develop a CBM system; using relevant CBM techniques such that as to optimise the availability improve the reliability and prevent a reoccurrence of a breakdown of the Crw-y-Cawl hydroelectric power station.
Hydroelectric Power Stations
The principle behind hydroelectric power stations is simple, it seeks to convert the energy (as it flows or falls) in water to convert it to electrical energy.
(source: howstuffworks.com)
The main components of a conventional hydroelectric power plant are;
The Dam
This serves as a large water reservoir for holding water
Intake
The water held in the dam is drawn in under gravity through a penstock; this is a pipe that carries water from the dam to the turbine.
Turbine
The water in the penstock builds up pressure. The water which is now under pressure strikes the blades of the turbine. The blades of the turbine turn as the water strikes it drives a shaft which is attached to the generator above it
Generators
The generator produces the electricity via the turbine. This electricity which is AC is transmitted to a transformer
Transformer
The transformer simple converts the AC in to current with a much higher voltage.
Power lines
The power lines are used to transmit the power to where it is needed.
Outflow
This is to carry the used water back to the river.
Used water is carried through pipelines, called tailraces, and re-enters the river downstream.
The water in the reservoir is considered stored energy. When the gates open, the water flowing through the penstock becomes kinetic energy because it's in motion. The amount of electricity that is generated is determined by several factors. Two of those factors are the volume of water flow and the amount of hydraulic head. The head refers to the distance between the water surface and the turbines. As the head and flow increase, so does the electricity generated. The head is usually dependent upon the amount of water in the reservoir.
The Generator
A generator is made up of the following maintainable items
Shaft
Excitor
Rotor
Stator
The turbine turns, the excitor responds by sending electric current to the rotor. The rotor is a series of electromagnets which spin inside tightly-wound copper wire. This coil of wire is what is referred to as the stator. A magnetic field is created between the rotor and the stator which produces electricity.
Theory of operation
As stated earlier water flowing is directed such that it strikes the blade, the energy in the water is transferred to the turbine which in generates electrical energy via the generator attached to the turbine.
There are two types of water turbines namely;
Reaction turbines
Impulse turbines
The precise shape of water turbine blades is a function of the supply pressure of water, and the type of impeller selected.
Reaction turbines
Reaction turbines are acted on by water. The pressure of the water changes as it moves through the turbine and transfers its energy. These types of turbines should be such as to contain water via being encased (or suction), or be completely submerged in flowing water.
Impulse turbines
Impulse turbines operate on the principle of changes in momentum (impulse). The change in momentum results in the rotation of the turbine and the kinetic energy in the water is removed.
These type of turbines do not required a housing as no pressure change takes place at the blades. It takes place at the nozzle which is focused on the turbine, where potential energy is converted to kinetic energy
Most turbines operate on the principles used in both impulse turbines and reaction turbines.
Power
The power which is obtained from the water can be expressed as;
P = η ρ g h q
where:
P = power (J/s or watts)
η = turbine efficiency
ρ = density of water (kg/m³)
g = acceleration of gravity (9.81 m/s²)
h = head (m)
q= flow rate (m³/s)
Pumped storage
Some water turbines are designed for pumped storage hydroelectricity. They can reverse flow and operate as a pump to fill a high reservoir during off-peak electrical hours, and then revert to a turbine for power generation during peak electrical demand. Efficiency
Specifications and Diagrams
Turbine selection is based mostly on the available water head. The following is a list
typical head ranges measured in metres;
Hydraulic wheel turbine 0.2 < H < 4 (H = head in m)
Archimedes' screw turbine 1 < H < 10
Kaplan 2 < H < 40
Francis 10 < H < 350
Pelton 50 < H < 1300
Turgo 50 < H < 250
The Cwm-y-Cawl Hydroelectric Power Station , supplies 300MW of electricity to Penrhys Special Steels. It can be deduced from the chart above that the Hydroelectric Station uses a Kaplan turbine. This means that it requires between 2m to 40m of water head.
(Source:http://upload.wikimedia.org/wikipedia/commons/1/16/S_vs_kaplan_schnitt_1_zoom.jpg)
Literature Review
Condition Monitoring
Introduction
Condition monitoring is a systematic approach to achieving optimum system availability. It is a planned maintenance task.
CBM involves using monitoring the state of equipment and using its current state to prompt the appropriate maintenance response.
This type of maintenance can be formally defined as 'The preventive maintenance initiated as a result of knowledge of the condition of an item from routine or continuous monitoring' (BS 3811:1984; Williams et al 1984).
In summary maintenance is only carried out when needed and the timing and actual task to be performed is based on the actual state of the item.
The advantages of CBM include;
Early knowledge of impending failure: this makes for easier planning
Actual condition knowledge and accurate failure prediction
Diagnosis potential
Flexibility
Generally speaking the fundamental principle in the application of CBM is that maintenance can be carried out based on the ability to predict failure in a time sufficient enough.
Condition Monitoring Methods
In order to be able to predict the failure of equipment accurately, the selection of the right parameter to monitor the equipment condition is very important.
Vibration Analysis
Lubricant Oil Analysis
Thermography
Corrosion Monitoring
Vibration Analysis
Vibration can be said to be motion which repeats itself after a time interval it is one of the primary tools used in condition based monitoring.
The reason behind this is that all structures are subject to vibration. Equipment used in industry produce a vibration profile which can be monitored.
A vibration profile is made up of the various components of the equipment each vibrating and generating its own individual profile. When all these individual profiles are added up they produce the overall vibration of the equipment. This profile gives a description of the overall health of the equipment.
Vibration monitoring is not limited to just limited to just the determination of the current state of equipment. It can be used in fault diagnosis.
Basic Vibration Theory
Simple Harmonic motion is the simplest form of vibration. This can be represented as a mass spring system. If the an external force is applied to the system which causes the mass to oscillate the equation of motion can be represented as
External force = Inertia force + Damping force + stiffness force
Ma(t) + cv(t) + ky(t) = f(t)
This can be represented graphically:
Vibration is measured in terms of displacement, velocity or acceleration. It is quantified as root mean square (rms), 0-peak and peak to peak.
Vibration data Interpretation and Analysis
The time domain is amplitude of vibration against time. It is useful for linear and reciprocating motion. It can be used to identify change conditions in a machine train. The limitation in the use of time domain in the analysis of vibration data is that the determination of individual contributions to the overall vibration profile.
To overcome this limitation the time domain is converted to frequency domain using the Fast Fourier Transformation (FFT). The frequency domain is amplitude of vibration against frequency. Under the frequency domain each vibration source is seen a unique peak which is a multiple of the basic running speed if there is a fault.
The usefulness of vibration analysis is the ability to distinguish between when the machine is a healthy state and when it is in an unhealthy state.
Mobley (2002) states the following assumptions in the use of vibration data;
All common machinery problems have a unique vibration frequency component which can be isolated and identified.
The frequency domain vibration signature (profile) is used in analysis because discrete peaks represent a specific vibration source.
By comparing the vibration profile of a piece of equipment overtime, changes in the profile can be identified as the profile repeats itself overtime.
Data acquisition
There are three main devices used to acquire vibration data, they include;
Proximity Probes
Velocity Transducers
Accelerometers
It must be said that they are all transducers, which measure different vibration parameters.
Proximity Probes
These measure the actual movement (displacement) of a machine shaft relative to the probe. To obtain accurate and repeatable data, it must be mounted rigidly to a stationary structure.
Displacement transducers (proximity probes are displacement transducers) are used for measuring low frequency vibrations usually between 10 - 1000Hz.
These types of transducers are very expensive to uses
Velocity Transducers
These are electromechanical sensors which measure the rate relative displacements. They are usually very bulky and expensive. They are used to measure low frequency vibrations also between 10 - 1000Hz.
They are sensitive to mechanical and thermal damage and required a very rigid recalibration programme.
Accelerometers
They are best type of transducers especially piezo-electric accelerometers. They have a very wide frequency range between 1 - 10,000 Hz and have good linearity for a wide range of high frequencies. It is simple in design and has no mechanical moving parts and as such they are reliable and not subject to wear.
The following should be considered in the selection of accelerometers;
Sensitivity
Frequency range of measurement
Resonance Frequency
Amplitude Linearity
Phase distortion
Maximum Acceleration Measurement Limit
It must be noted that linearity is usually accurate for about 20% of the frequency range.
Measurement Points
Measurements are most often taken from the rigid part of the machine usually the machine housing or bearing cap. Direct measurements of the shaft displacement are taken only to acquire a true picture of the machine dynamics.
Measurements should be taken in both radial and axial directions to aid the detection of root cause deviations. This is because of unbalanced forces which occur when the machine is either a healthy or unhealthy state.
Transducer mountings
Permanent mounting
This method gives the best results as accuracy and repeatability are guaranteed. It is a very expensive method of mounting transducers.
Stud mounting
This method of mounting is much cheaper than permanent mountings and gives the same level of results as permanent mounting methods in terms of accuracy and repeatability.
Magnet mounting
The resonant frequency of the transducer is reduced distorted and as such give inaccurate readings. Repeatability is also an issue as it might be difficult to obtain the exact location it was initially placed before it was removed.
Adhesive mountings
These present the same challenges as the magnet mounting. The only difference is that it is less reliable than magnet mounted transducers.
The reason for this is the change is stiffness value of the system, by either the magnetic or the adhesive.
Vibration analysis techniques
Trending
This involves the use of historical vibration data as analysis tool. It makes use of the overall vibration which is taken at regular intervals over time to identify when the machine is in an unhealthy state. Overall vibration is usually recorded as RMS. Changes to the amplitude of the overall vibration, indicates a change in the operating conditions and as such can be used as a diagnostic tool. In order for trending to be effective it has to be filtered to remove noise and so as to obtain a true picture of the machine conditions.
Broadband
Broadband analysis has been used to monitor the overall condition of equipment. It uses the overall vibration from frequency range of zero to the maximum desired frequency. This is usually in the RMS. Broadband analysis is only useful for long term trending of the overall vibration in setting warning/alert and alarm limits.
Narrowband
Narrowband analysis also monitors the overall vibration like broadband analysis. The difference is that ability to select a specific set of frequencies. The advantage of this is that it can be used to monitor specific components in a machine. Alert/warning and alarm limits can be set for various equipment components.
Warning limits
These give information that a there has been a change in the overall vibration. The change is such that it does not require immediate action but that more attention needs to be paid to the monitoring parameter.
Alarm limits
These give information that a significant change has occurred, a fault has developed and action needs to be taken to prevent complete failure.
Diagram
The peaks seen in the frequency domain increase as faults develop. This might not be captured immediately using RMS to record the overall vibration. The crest factor or kurtosis might be used as they are much quicker to respond to changes in the overall vibration compared to RMS.
Comparative Analysis
Comparative analysis entails comparing data sets in order to detect changes in the operating conditions. Data needed to carry out this analysis includes baseline data, present machine condition or industrial reference data.
Baseline Data
This is the data set which is acquired for either the various components of a piece of equipment or the overall machine. This data is usually collected when condition monitoring programme is being established if it was not collected after installation. The usefulness of this data set is since it represents the machine operating under normal conditions; it can be used to compare data sets collected in the future.
Baseline data should be update after any major maintenance work is carried out.
Signature Analysis
This provides information that uniquely identifies all components in a machine. It gives specific data on every vibration component within the overall frequency range. This makes the form of analysis very useful in fault diagnosis.
Lubricant Monitoring
Solid particles in lubrication oils and contamination of the oil can result in equipment failure. Lubricant monitoring is used to determine the health of the oil and of the machine.
Lubricant oil analysis is a method used to determine the condition of the oil in a piece of equipment. On the other hand wear debris analysis is the method used to monitor the health of equipment and in fault diagnosis.
Lubricating oil is used for the following;
To reduce wear and friction
To cool and protect surfaces
To remove contaminants
The results of lubricant oil analysis are used to determine when to schedule oil changes. This is cost effective and helps to reduce inventory as oil is changed only when it needs to be done.
Wear debris analysis provides information in relation to the particle content in lubricating oil. This information provides insight in relation to the actual condition of the equipment. Information on the quantity, size and shape of the particles is useful in failure and root cause analysis.
The challenges in the use of lubricant monitoring are the cost and sampling.
Oil sampling
Sinha et al (2010) lists three things are very important in oil sampling these are;
Positioning
Preparation
Time
Positioning
The position of where the sample is to be collected is very important. It needs to be such that it gives a picture of the true condition of the oil in the system. There are three techniques of taking samples they are online, offline and inline.
With online sample taking the tap off method is the best way of taking a sample. If this is not possible syringes or vacuum pumps can be used.
The sample point should be downstream of the component being under observation and upstream of filtration.
Preparation
There should be a well defined procedure for collecting oil samples as cleanliness is very important in ensuring that the accuracy of the test results. Sample should be compare with pre-cleaned fluids.
Time
It is recommended that samples be taken when operating temperatures have been attained, with the machine still running with the oil in turbulent flow. The time, date and the conditions of operation should be recorded.
It should be said that the sampling frequency is a function of the mean-time-to-failure (MTTF). The period starting from the time of abnormal wear to catastrophic failure (Mobley 2002)
The following is a list of parameters measured while carrying out lubricant monitoring;
Viscosity
Microbiological growth
Oxidation
Acidity
Additive concentration
Dilution
Debris content
Viscosity
Viscosity is an important property of lubricating oil. Very low viscosity results in oil film strength being reduced, and this leads to an increased metal-to-metal contact. Excessively high viscosity on the other hand indicates a reduction in lubricating ability as oil to cannot flow to where it is needed.
Fuel Dilution
Oil contamination results in oil dilution. The strength, the sealing ability and detergency of the oil is affected. It can be attributed to fuel leaks, improper operation and timing.
Solids Content
The solid content in oil can led to increased wear of parts. Abnormal levels or sudden rise in the solid content calls for attention.
Fuel Soot
This is an important factor for oil used in diesel engines. It serves as a measure of how efficiently fuel is being burnt.
Oxidation
The presence of oxygen in lubricants (oil) leads to oil thickening and corrosion. Even though additives are present in most lubricant, when they are depleted oxidation commences.
Wear Debris Analysis
As stated earlier this is used to monitor the condition of the equipment in which the oil (fluid) is placed.
It considers the following
What particles might give indicate trend toward failure
What is the most appropriate sampling procedure is
The analysis technique which should be used ( Sinha et al 2010)
Wear debris is classified using the following features
Size
Quantity/Concentration
Distribution
Shape
Composition
There are number of lubricant monitoring methods these include;
Total Acid Number (TAN)
The acidity in lubricating oil sample is simply a measure of the acid content in the sample. It is important to compare used oil samples with never before used oil of the same type. The reason behind this is that new oil contains additives which influence the TAN.
Total Base Number (TBN)
The base number is simply a measure of the ability of oil to neutralise acidity. A high TBN indicates a greater ability to neutralise acidity. TBN is affected by the use of wrong grade of oil, the interval between oil changes and sulphur levels in fuel.
Particle Count
This is useful in anticipating problems with equipment. It is quite different from wear debris analysis. The particle count can give indications to abnormally wearing. The number of particles is what is observed here, not wear patterns or other factors which can give insight to failure modes.
Spectrographic Analysis
Spectrographic analysis measures the following parameters concentration of additives, soot, debris content composition and oxidation accurately and rapidly. A disadvantage of using this is the inability to detect large particles (it only detects particles in size of 5 -8 microns).
(Williams et al 1984)
Ferrography
This is similar to spectrography. It separates the particle content by using a magnetic field and as such it is limited ferrous particles.
It can however detect particles of sizes greater than 10 microns, the range it can detect is between 1 - 100 microns. In this regard it can be said to give a better picture of oil contamination when compared with spectrography.
Depending on the system and the parameters being measured alert and alarm limits can be set up.
Thermography
Introduction
Thermography is a technique used to determine the actual temperature of a body by measuring the emission of infrared energy from its surface. Infrared emissions are outside the visible spectrum and as such they require special instrumentation. The intensity of infrared radiation is a function of its surface temperature. Measuring temperature using infrared methods is complicated because a number of the sources of thermal energy can be detected from any object. These energy sources are the energy emitted from the object, energy reflected from the object, and energy transmitted by the object. The energy which is either reflected or transmitted distorts raw infrared data and as such need to be filter out so as to obtain more accurate readings.
The emissivity of an object is affected by a number of factors such as the surface condition, and protective coatings. Also the atmosphere between the object and the instrument of measurement has to be taken into consideration. Gases in the atmosphere absorb radiation and as consequence distort the infrared radiation.
BASIC INFRARED THEORY
All objects emit energy when heat is applied. The amount of energy is emitted has a direct relationship with the temperature. The electromagnetic spectrum made of various forms of radiated energy, of which infrared energy is a part. Infrared energy covers the spectrum
of 0.7 micron to 100 microns. Electromagnetic spectrum is a continuum of all electromagnetic waves arranged according to frequency and wavelength.
Infrared energy is not within the visible spectrum, devices have been developed to enable the human eye to see it and measure the heat.
Data from the infrared imagers is converted into digital data and processed in video images known as thermograms. The pixels of a thermogram have individual temperature values
and the image's contrast is because different surface types have different surface temperatures.
Heat Transfer Concepts
All objects emit infrared energy through via three different ways: conduction, convection, and radiation.
Conduction
This is simply the transfer of energy through or between solid objects.
Convection
This simply the transfer of energy through or between fluids or gases.
RADIOSIBLE
RAVIOLET
Radiation
This is simply the transfer of heat by wavelengths of electromagnetic energy.
Blackbody
A blackbody is a perfect thermal emitter. Its emissivity value is 1, as it neither reflects nor transmits energy.
Black boy radiation
There are three physical laws which are relevant to the generation of radiation they include (Sinha et al 2010);
Stefan-Boltzmann Law which defines the relationship between total energy flux density from a body to the temperature of the body
Plank's Law which defines the relationship between the energy flux density of a body at each wavelength to the temperature of the same body
Wein's Displacement Law this defines the relationship between the energy flux density of a body at each temperature with that at another temperature
Real Body radiation
Emissivity
Emissivity is the percentage of energy emitted by an object. Infrared energy hits an
object; the energy is then transmitted, reflected, or absorbed. Different surfaces have different emissivity values between 0 and 1.
Sinha et al 2010 states a number of challenges in using thermography in condition monitoring they are as follows,
Materials actually which are actually at the same temperature can appear to be different. To overcome this challenge the emissivity value must be correct, thermal cameras have built in emissivity tables.
Difficulty in matching thermal images to the real life situation. Thermal cameras also now that the ability of image blending, GPS position labelling of IR images and voice over recording
Types of thermal Problems
There are three basic types of thermal problems:
• Mechanical looseness
• Load problems
• Component failure
These problems can be seen in different areas like civil, mechanical and electrical engineering.
Measurement Instruments
Three general types of instruments used for condition monitoring using this technique:
Infrared thermometers
Line scanner
Imaging systems (Thermal cameras)
Application of Condition Based Maintenance
CBM is very expensive to implement, so it is very important that is applied to appropriate equipment. This usually the most critical equipment, equipment was a severe impact on production, safety or the environment.
This knowledge is very important before CBM can be implemented.
Plant selection procedure for the application of CBM
Surveying plant criticality
Maintenance Audit
Select units of the plant
Match CBM techniques to failure mode
Routine Monitoring
Review
Cost effectiveness
Diagram
Criticality Analysis
The most important thing in doing a criticality analysis is to answer questions related to reliability issues, availability issues and financial issues (where was the most money spent). Another issue that must be dealt with is the selection of the pilot plant or unit.
The Hydroelectric plant - Criticality Analysis
The information given the major breakdown which occurred cost £ 120,000. The root cause is attributed to inadequate lubrication of an alternator bearing and this resulted in severe damage to the shafts and of the alternators. Other failures which occur are from the records attributed to controls.
It can be concluded that the most money has been spent on the repair of the damaged shafts and an alternator and that the controls are what fails the most.
The focuses to implement CBM on the power plant are on the controls, the shafts and alternator and the alternator bearings which were damaged as a result of inadequate lubrication. These items are found in the generator. The generator would be the main focus.
Failure mode analysis
The fault tree gives us an indication as to what components need to be monitored in order to prevent the top level event of total loss of power. The following is a list of the items and their respective failure modes
Item/Component Failure mode
Runner blades Erosion or cavitations
Bearings Inadequate lubrication
Water leakage faulty inlet valves
Shaft Crack, misalignment or unbalance
Guide Vanes Inadequate lubrication
Alternator
Rotor Inadequate lubrication, voltage overload ,misalignment /imbalance
Stator Insulation failure, overload, misalignment/imbalance
It is worth noting that intrinsic faults are a common failure mode of both the rotor and stator.
The root cause was inadequate lubrication. The shaft damage could have been foreseen with the use of vibration monitoring. The failure of the alternator and the alternator bearing could have been prevented using either vibration monitoring or thermography.
The failure mode analysis fault tree analysis and the root cause of the major break down reveal that
Vibration monitoring, lubrication analysis and thermography can be applied, as summarised in the table below
Item/Component Failure mode Condition monitoring Technique
Runner blades Erosion or cavitations Vibration analysis
Bearings Inadequate lubrication Vibration/Lubricant analysis
Thermography
Water leakage faulty inlet valves Performance monitoring
Shaft Crack, misalignment or unbalance Vibration analysis
Guide Vanes Inadequate lubrication Lube lubricant analysis
Alternator Vibration/lubricant analysis
Thermography
The generator can be regarded as a critical unit.
The bearings are lubricated using grease. Most of lubricant analysis is most suited for oil analysis and not grease analysis.
The recommended CBM techniques are vibration analysis and thermography. Even though information from the above table suggest that most items in question can be monitored just using vibration analysis, thermography can be used to supplement vibration analysis to ensure accuracy and reliability of results. If a fault is missed by one technique it might be picked up by the other also some faults are easier to pick up with thermography than vibration analysis if trending is just used.
An example is a faulty bearing. The overall vibration will pick up a bearing failure, but themography will not only pick it up it will also allow the faulty bearing be identified.
It must be said that inadequate lubrication was the root cause of failure. In a number of cases failure due to lubrication is due to the use of improper grade of lubricant.
Suggested maintenance monitoring programme
Vibration Analysis
The sensors can be mounted permanently or smart studs can be used instead of the sensors. These techniques ensure that repeatability of measurements. The instrument of choice would be an accelerometer. This is because of the wide frequency range, it will ensure that bearing faults are capture. Velocity transducers and proximity probes are not are versatile and as such are not recommended to be used here.
This is because of the bearings. Bearing are associated with high frequencies when they fail. The position of the sensor (accelerometers in this case) should be on the bearing house. The sensors should be in both the vertical and horizontal direction, to take care of any unbalance force which could be along any axis.
If the sensor mounting is to be adopted, a cable can be run from the sensor so that a portable instrument can be used to collect the data which can later be downloaded into the PC for trending or analysis.
It is recommended that a hardwired system be used as the system is critical.
There is not history of vibration data in the form of a database. A baseline needs to be established which identifies when the machine (in this case the generator) is in a health state.
In order to establish this baseline data would need to be taken more often, in this case it is recommended that data be collected weekly. This can then be reduced to every two weeks.
The absence of any vibration data makes the establishment of warning and alert limits challenging. The ISO code 10816 and manufacturer's recommendation can serve as initial guidelines. These limits can be changed over time as required.
Thermography
Thermal cameras are the instrument of choice here is. There are a number of things to consider when selecting thermal cameras.
Portability
Ease of use
Qualitative or quantitative. Does it measure temperatures? If yes, what temperature range will be measured? Will you need more than one range?
Ambient or quantitative measurements. What are the maximum upper and minimum lower
ambient temperatures in which you will be scanning?
Short or Long Wavelengths. Long-wavelength systems offer less solar reflection and operate in the 8- to 14-_m bandwidth. Short-wavelength systems offer smaller temperature errors when an incorrect emissivity value is entered. The operating bandwidth for a short-wavelength unit is 2 to 5.6 _m.
Batteries.
Interchangeable lenses
Monitor, eyepiece, or both
Analog or digital. How will you process the images? Does the imager have analog, digital, or
both capabilities?
Software. Can the software package produce quality reports and store and retrieve images? Do you require colonization and temperature editing?
The unit under consideration is regarded as critical and as such continuous monitoring is recommended. The decision on if to do this will depend on the following;
Time between fault initiation and failure is short
There is a very high cost of failure
The equipment required is very expensive when compared with interval/scheduled monitoring
Scheduled/Periodic monitoring is more appropriate if;
The equipment is expensive
The equipment is not critical or doesn't have any critical parts
There is a long time interval between fault initiation and failure
Placing the camera permanently is not feasible for either safety or logistic reasons
Costing
A cost/benefit analysis would need to be done, in order to select the most cost effective approach for each condition monitoring technique identified.
Discussion and Conclusion
Discussion
No condition monitoring technique is superior to the other. A very good condition monitoring programme will adopt a number of techniques, which have the ability to monitor the selected parameters in a cost effective manner.
Visual inspection should not be over looked and should be included in the condition monitoring programme. This is because man is the best all round sensor. It can be carried out along side preventive maintenance tasks or other condition monitoring tasks.
It includes identification of leaks, abnormal sounds or high noise levels and cracks.
The monitoring of process parameters should also not be ignored and should also be included in the condition monitoring programme. The monitoring of parameters such as flow rate, operating temperature and pressure can give indications of impending failure. The recent advances in technology have made it easier to monitor these parameters as some of these parameters can be read off built in gauges.
The implementation of a condition monitoring programme is not enough to prevent failure. There are a number of other factors which have to be considered. These factors include;
Maintenance Strategy
Preventive Maintenance Schedule
Skilled and Motivated Personnel
Computer Maintenance Management System (CMMS)
Maintenance Awareness in Design
Maintenance Strategy
The formulation and implementation of the appropriate maintenance strategy cannot be over emphasised. Maintenance is not activity carried out in isolation; it is part of a larger system the business. It is should be carried out in such a way as to aid in the achievement of the overall business objectives.
Whatever strategy is implemented is it important that all stakeholders are involved. The stakeholders include; senior management, production, maintenance and engineering. There involvement and support is vital if the maintenance strategy of which condition monitoring is a part of is to be successful.
Formulating and implementing the appropriate maintenance strategy, also the identification of critical equipment easier. The various stakeholders have varying ideas on what the critical equipment is. The classification of critical equipment is usually based on impact to the environment, safety and production. In some cases the price of the machinery and maintainability of equipment is used as criteria. Condition monitoring is usually reserved for critical equipment on the basis of cost associated with it. In light of this it is very important that critical equipment is identified correctly so that condition monitoring is carried out where the best results can be obtained; which is plant optimisation.
Preventive Maintenance Schedule
This is very important in ensuring equipment reliability and availability and as such should not be over looked. Preventive maintenance tasks include; lubrication, calibration, inspections, cleaning of equipment, making adjustments to equipment and parts replacement.
Preventive maintenance tasks are carried out at regular intervals. This requires effective planning and scheduling of activities. Planning is answering the questions 'what task is to be done' and 'how the said task is to be carried out'. Scheduling on the other hand answers the questions 'who is responsible for carrying out the task' and 'when the said task is to be carried out'.
Skilled and Motivated personnel
For any maintenance programme to be a success the people are very important. It does not matter how well the task has been planned or scheduled or if the right condition monitoring techniques are selected. If the personnel do not possess the skills and are not motivated, the programme will not be a success.
The skill requirements need to be address when implementing a condition monitoring programme and appropriate training carried out if it is to be done in house. Personnel need to buy into the programme; a sense of ownership should be developed, as this helps in keeping them motivated.
Computer Maintenance Management System (CMMS)
A good CMMS is important in a CBM programme. It aids in not just planning and scheduling maintenance activities and coordinating resources but also in the documentation (of trends and other feedback information) and communication of information.
The following information is very important in a condition monitoring programme, monitoring parameters, data collection (methods and plans) and communication.
The communication is a very important part of condition monitoring. If the information is not communicated it defeats the objectives of the programme.
Maintenance Awareness in Design
This is considered with the ease or difficulty in associated with performing maintenance task. An item which is good in terms of maintainability, the lower is associated downtime. The issue of maintainability also affects condition monitoring. An example is including tap points in equipment which is oil lubricated in order to make it easier in collecting oil samples, and the location of bearing in order to aid vibration monitoring via locating sensors.
Conclusion
A condition monitoring programme has been suggested to optimise the availability of the Crm-y-Cawl Hydroelectric Power Station.
The suggested condition monitoring techniques are vibration monitoring and thermography.
