In many industrial plants, equipment receives regular maintenance and still experiences recurring failures, premature wear, or unexpected losses of availability.
When these problems occur, the response often consists of increasing inspections, shortening maintenance intervals, or replacing components more frequently. However, performing more maintenance does not necessarily make equipment more reliable.
Reliability engineering proposes a different approach. Rather than focusing only on repairing equipment, it examines how assets can fail, which factors affect their performance, and which design, operating, and maintenance decisions can preserve their functions for a longer period.
ASQ describes a reliability engineer as a professional who applies performance evaluation and prediction principles to improve the reliability, maintainability, and safety of products, systems, and processes.
Applied to industrial operations, this discipline helps reduce failures, improve equipment availability, and extend service life through decisions based on risk, evidence, and actual operating conditions.
WHAT IS RELIABILITY ENGINEERING?
Reliability engineering studies the ability of an asset or system to perform a required function under defined conditions for a specified period.
This definition includes four fundamental elements:
- A specific function
- Defined operating conditions
- A period of time
- A probability of satisfactory performance
For this reason, saying that a piece of equipment is “reliable” does not simply mean that it is still running.
It is also necessary to determine:
- Whether it provides the required capacity
- Whether it maintains the expected quality
- Whether it operates within safe limits
- Whether it preserves its efficiency
- Whether it can do so for the required period
- Whether it experiences interruptions or degradation
A pump may continue rotating while still losing a significant amount of reliability if it develops leakage, excessive vibration, reduced flow, or increased energy consumption.
From this perspective, industrial reliability is not evaluated only when total failure occurs. It also considers progressive degradation that prevents an asset from performing its function correctly.
RELIABILITY AND AVAILABILITY ARE NOT THE SAME
Reliability and availability are often used as if they were equivalent, but they describe different aspects of asset performance.
Reliability refers to the equipment’s ability to operate without failure or interruption for a defined period.
Availability indicates the proportion of time the equipment is capable of operating when it is required.
IBM explains that availability measures an asset’s operational readiness over time, while reliability describes its ability to operate without interruption under specified conditions.
Equipment may have low reliability but apparently acceptable availability if failures can be repaired quickly.
It may also have high reliability and low availability if an infrequent failure requires excessive time to correct.
Equipment availability depends primarily on two factors:
- How frequently failures occur
- How long it takes to restore operation
Improving reliability therefore involves both reducing failure frequency and making diagnosis, repair, and return to service more efficient.
RELIABILITY BEGINS IN THE DESIGN STAGE
Many maintenance problems originate in decisions made before the equipment begins operating.
These may include:
- Incorrect material selection
- Insufficient capacity
- Improper tolerances
- Limited inspection access
- Lubrication systems that are difficult to maintain
- Insufficient instrumentation
- Designs that are sensitive to contamination
- Components exposed to temperatures that were not considered
- Lack of redundancy in critical functions
NASA notes that the factors influencing reliability include design, manufacturing, final application, and human involvement at every stage.
This means that maintenance cannot indefinitely compensate for a design that does not reflect actual operating conditions.
Designing for reliability requires anticipating:
- Potential failure modes
- Actual loads and temperatures
- Process variability
- Environmental conditions
- Material degradation
- Inspection requirements
- Repair methods
- Spare-parts availability
- The consequences of failure
It also requires designing for maintainability.
Reliable equipment should not only fail infrequently. When intervention is necessary, the work should be performed safely, quickly, and consistently.
IDENTIFY FUNCTIONS BEFORE ANALYZING FAILURES
Reliability engineering begins by understanding what the asset is expected to do.
This step may appear simple, but it is frequently overlooked.
For example, the function of a pump is not merely “to pump.” Engineers must define:
- The fluid it must transport
- The required flow rate
- The expected pressure
- The operating temperature
- The service duration
- The minimum efficiency
- Acceptable leakage limits
- Starting and stopping conditions
- Safety requirements
A functional failure occurs when the asset can no longer meet one of these criteria, even if it continues to operate.
Clearly defining functions makes it possible to distinguish among:
- Total failure
- Partial performance loss
- Acceptable degradation
- Unsafe conditions
- Quality nonconformance
- Operational inefficiency
Without a clearly defined function, it is impossible to determine what reliability actually means for the equipment.
ANALYZE FAILURE MODES, NOT ONLY COMPONENTS
Once the functions have been defined, the next step is to identify how the equipment may fail to perform them.
A failure mode describes the specific way in which a function is lost or degraded.
Examples include:
- Bearing wear
- Loss of lubrication
- Seal leakage
- Line blockage
- Misalignment
- Shaft fatigue
- Motor overheating
- Surface corrosion
- Thermal deformation
- Sensor calibration loss
Failure Modes and Effects Analysis, or FMEA, evaluates components and subsystems to identify potential failures, causes, and consequences. The U.S. Department of Energy notes that this tool is often one of the first steps in a reliability study.
Not every failure mode requires the same strategy.
Some can be prevented through scheduled maintenance. Others require condition monitoring, design changes, operating controls, or different materials.
Reliability engineering seeks to select the most appropriate response for each risk rather than applying the same maintenance approach to every asset.
RELIABILITY-CENTERED MAINTENANCE
Reliability-centered maintenance, commonly known as RCM, uses asset functions, failure modes, and consequences to determine which maintenance tasks are technically and economically justified.
It does not begin with the assumption that every component should be serviced at the same interval.
Instead, it asks:
- Which function must be preserved?
- How can that function be lost?
- What causes each failure mode?
- What consequences would result?
- Can the failure be detected before it occurs?
- Is there an effective preventive task?
- Is redesign necessary?
- Can the risk be accepted?
NASA indicates that an RCM strategy can reduce failures, increase operational availability, and lower maintenance costs.
This approach helps distinguish tasks that genuinely control risk from those performed only because they have always been part of the maintenance routine.
A maintenance strategy may combine:
- Reactive maintenance
- Preventive maintenance
- Predictive maintenance
- Functional inspections
- Protective-device testing
- Condition monitoring
- Redesign
- Scheduled replacement
- Run-to-failure operation for noncritical assets
The correct choice depends on the consequences of the failure mode and whether it can realistically be anticipated.
PREVENTIVE MAINTENANCE DOES NOT ALWAYS PREVENT FAILURE
Preventive maintenance is performed at intervals defined by time, cycles, or use.
It is effective when there is a clear relationship between component age and failure probability.
Examples include:
- Periodic filter replacement
- Lubricant changes
- Drive-system adjustment
- Replacement of limited-life components
- Inspection of parts exposed to known degradation
However, not every failure is age-related.
A component may fail randomly because of contamination, installation errors, process variation, or environmental conditions.
Servicing it more frequently may not reduce that risk. It may even introduce new defects through disassembly, incorrect adjustment, or contamination during maintenance.
The U.S. Department of Energy’s operations and maintenance guidance distinguishes reactive, preventive, predictive, and reliability-centered maintenance as different strategies for helping equipment reach or exceed its design life.
The objective should not be to maximize the number of preventive tasks, but to select those that genuinely reduce the probability or consequence of failure.
THE ROLE OF PREDICTIVE MAINTENANCE
Predictive maintenance uses data about the actual condition of equipment to identify degradation and estimate when intervention will be required.
It may rely on technologies such as:
- Vibration analysis
- Thermography
- Oil analysis
- Ultrasonic testing
- Current monitoring
- Thickness measurement
- Particle analysis
- Pressure and temperature sensors
- Performance tracking
IBM explains that predictive maintenance uses operating and environmental information to identify patterns related to failure and evaluate equipment health.
This makes it possible to replace calendar-based interventions with evidence-based decisions.
However, installing sensors does not automatically create a predictive strategy.
To make the data useful, organizations must:
- Establish normal operating conditions
- Define relevant limits and trends
- Connect signals to failure modes
- Ensure measurement quality
- Assign responsibility for interpreting results
- Convert alerts into work orders
- Verify whether the intervention corrected the problem
The value lies not only in collecting data, but in using it to make decisions before degradation becomes a functional failure.
MEASURE PERFORMANCE TO MAKE BETTER DECISIONS
Reliability engineering uses indicators to understand asset behavior and determine whether implemented strategies are producing results.
Common indicators include:
Mean Time Between Failures
MTBF represents the average operating time between failures for repairable equipment.
An increase may indicate that implemented actions are reducing failure frequency.
Mean Time to Repair
MTTR measures the average time required to diagnose, repair, and return an asset to service.
Reducing MTTR can improve availability even when failure frequency remains unchanged.
Availability
Availability relates the time during which an asset is capable of operating to the total time it is required.
Failure Rate
Failure rate indicates how frequently failures occur within a defined period or asset population.
Remaining Useful Life
Remaining useful life estimates how much time or how many operating cycles remain before a component reaches a failure limit.
Maintenance Compliance
This indicator measures whether scheduled maintenance tasks are completed within the required period.
These metrics must be interpreted in context.
A higher MTBF does not necessarily represent improvement if failures are no longer being recorded correctly. A low MTTR is not necessarily positive if the same component is repeatedly repaired without eliminating the root cause.
Reliability cannot be managed through a single metric. It requires a combination of technical, operational, and economic indicators.
LEARNING FROM RECURRING FAILURES
When a failure repeats, industrial reliability requires investigating why previous actions did not eliminate the problem.
Root cause analysis may reveal factors such as:
- Incorrect material selection
- Inconsistent procedures
- Lubrication deficiencies
- Contamination
- Process changes
- Conditions outside the original design
- Assembly errors
- Insufficient training
- Inadequate controls
- Recommendations that were never implemented
A repair returns the equipment to operation.
A reliability action seeks to prevent the same mechanism from producing another failure.
This requires preserving evidence, analyzing patterns, validating hypotheses, and implementing measurable corrective actions.
The learning should also extend to similar assets. When a problem is identified in one pump, bearing, or seal, engineers should determine whether other equipment shares the same design, condition, or risk.
OPERATION ALSO DETERMINES RELIABILITY
Operational reliability does not depend exclusively on the maintenance department.
Production decisions can significantly alter equipment service life.
Operating conditions that affect reliability include:
- Overloading
- Frequent starts and stops
- Operation outside the design point
- Temperatures above expected levels
- Cavitation
- Fluid changes
- Process contamination
- Inadequate heating or cooling
- Speed modifications
- Extended dry-running periods
Properly designed and maintained equipment can still fail early if it continuously operates outside its intended limits.
For this reason, a reliability strategy requires collaboration among:
- Engineering
- Operations
- Maintenance
- Production
- Quality
- Purchasing
- Safety
- Technical suppliers
Reliability must be managed as a system-wide responsibility, not as an isolated task assigned to the people who repair assets.
MATERIAL SELECTION AS A RELIABILITY DECISION
Material selection directly affects maintenance frequency, wear rate, and equipment service life.
A component may initially meet dimensional and mechanical requirements but degrade rapidly when exposed to:
- Extreme temperatures
- Corrosive fluids
- Abrasive particles
- Insufficient lubrication
- Low-viscosity fluids
- Sanitary environments
- Continuous operation
- Variable loads
When an application experiences recurring failures, the analysis should not be limited to determining whether the part was installed correctly.
It should also ask whether the material is compatible with actual operating conditions.
Proper material selection can help:
- Reduce friction and wear
- Maintain tolerances
- Dissipate heat
- Prevent corrosion
- Reduce dependence on lubrication
- Protect more expensive components
- Extend maintenance intervals
- Eliminate recurring failure modes
In this sense, materials engineering is part of reliability engineering.
WHEN CARBON GRAPHITE CAN SUPPORT RELIABILITY
In some systems, recurring failures are related to the limitations of oils, greases, metals, or conventional polymers.
Carbon graphite components may be considered when the application requires:
- Self-lubrication
- Dry-running capability
- A low coefficient of friction
- Wear resistance
- Dimensional stability
- Thermal conductivity
- Chemical resistance
- Operation at extreme temperatures
- Reduced risk of galling against metals
These properties can help eliminate lubrication-related maintenance points and reduce certain failure mechanisms.
However, reliability is not achieved by automatically replacing one material with another.
Engineers must evaluate:
- Load
- Speed
- Temperature
- Counterface
- Environment
- Available lubrication
- Tolerances
- Geometry
- Operating cycles
- Failure consequences
Metcar develops different carbon graphite grades to adapt properties such as strength, lubricity, impermeability, thermal behavior, and chemical compatibility to specific operating conditions.
Proper grade selection can help simplify the system and reduce failures when it is based on a complete application analysis.
FROM REPAIRING EQUIPMENT TO MANAGING ITS LIFE CYCLE
Reliability engineering changes the central question of industrial maintenance.
Instead of asking only:
When should this equipment be serviced?
It also asks:
Which function must be preserved, how can it be lost, and what is the most effective strategy for controlling that risk?
This shift makes it possible to move from a reactive culture toward management based on:
- Functions
- Risks
- Data
- Failure modes
- Actual condition
- Consequences
- Continuous improvement
Effective maintenance is one of the most important tools for ensuring reliability, safety, and efficiency. The U.S. Department of Energy also warns that inadequate maintenance can produce significant operational and energy losses.
Extending equipment service life does not mean keeping assets in operation indefinitely.
It means ensuring that they perform their functions safely, efficiently, and predictably for as long as continued operation remains technically and economically practical.
When design, operation, materials, and maintenance are managed as parts of the same strategy, reliability no longer depends on emergency repairs. It becomes an integrated capability of the system.
