The Vibration Analysis Workflow: From a Measurement to a Maintenance Decision

From Mechanical Maintenance to Vibration Analysis - Part 2

In Part 1 - You Are Not Starting from Zero, we established that a mechanical maintenance professional is not starting from zero when moving into vibration analysis. Knowledge of machines, failure modes, operating conditions and maintenance history is already part of the diagnostic process.

Now we move to the central question: What should an analyst actually do when new vibration data arrives?

A reliable diagnosis does not begin with selecting a fault from a chart. It begins before the measurement is collected and continues after the maintenance work is completed.

The objective is not to find a pattern that looks familiar. The objective is to build enough reliable evidence to support the right maintenance decision.

The complete workflow at a glance

Stage	Main question	Expected output
1. Understand	What machine and operating state are we assessing?	Asset and operating context
2. Plan	What measurements can reveal the expected failure modes?	Measurement strategy
3. Collect	Was the data collected safely and repeatably?	Traceable measurements and field notes
4. Validate	Is the change real, or could it be bad data?	Accepted or repeated measurement
5. Analyse	What changed, where and at which frequencies?	Defined vibration symptoms
6. Diagnose	Which causes fit all the available evidence?	Ranked fault hypotheses
7. Decide	What action is justified, and how urgent is it?	Risk-based recommendation
8. Verify	Did the action correct the condition?	Confirmed result and updated history

Each stage protects the next one. Excellent analysis cannot rescue unreliable data, and a technically correct diagnosis has little value if the recommendation is unclear or arrives too late.

Stage 1: Understand the machine and its operating context

Before opening a spectrum, establish what you are looking at. Useful questions include:

• What is the asset ID, function and consequence of failure?
• Which driver and driven components form the machine train?
• What are the running speeds, gear ratios, blade or vane counts and bearing types?
• How are the components coupled and supported?
• Is the machine fixed-speed or variable-speed?
• What load, flow, pressure, temperature or production state is normal?
• What maintenance was recently performed?
• Which failure modes are credible for this design and service?

This information turns frequency peaks into mechanical possibilities. For example, a peak at 25 Hz has little meaning by itself. If the shaft is running at 1,500 rpm, 25 Hz is running speed, or 1X. If the machine is running at 750 rpm, that same peak is 2X. Context changes the interpretation.

Stage 2: Plan measurements around the machine

Do not collect every possible measurement without purpose. Select points, directions, transducers and acquisition settings that can reveal the expected behaviour of the asset.

A basic route on a horizontal motor-pump set often includes measurements near the bearings in horizontal, vertical and axial directions. However, the correct plan depends on the machine, bearing type, casing, speed, accessibility and suspected fault.

The plan should define:

• measurement-point names and exact physical locations;
• sensor type and mounting method;
• measurement direction;
• units and amplitude convention;
• frequency range, resolution and other acquisition settings;
• expected operating state; and
• required process values and field observations.

Route consistency matters because condition monitoring depends on comparing the present with the past. If the point, orientation, mounting or operating state changes, the vibration may change even when the machine condition has not.

Stage 3: Collect safe, repeatable data

Repeatability means reducing unnecessary differences between one measurement and the next. Use the same marked point, direction, sensor, mounting method and suitable operating condition whenever practical.

Before taking a reading:

1. Follow site safety requirements and confirm that the measurement can be taken without exposure to rotating, hot, pressurized or energized hazards.
2. Confirm the correct asset and measurement point.
3. Check that the machine is running in the intended operating state.
4. Inspect the sensor, cable, connector and mounting surface.
5. Mount the sensor firmly and in the correct direction.
6. Allow the signal to stabilize and check whether the reading appears reasonable.
7. Record useful observations before leaving the machine.

Field observations can be diagnostically valuable. Note leaking seals, loose guards, damaged bases, unusual noise, product buildup, oil on the floor, recent maintenance and comments from operators. The person collecting data is not merely carrying an instrument; that person is also gathering context.

Stage 4: Validate the measurement before diagnosing the machine

When a value changes sharply, first ask whether the machine changed or the measurement changed.

Possible data problem	Basic validation action
Loose, tilted or inconsistent sensor mounting	Remount at the marked point and repeat the reading.
Wrong point or direction	Check the route definition, point label and orientation.
Different speed or load	Record the state and compare with data from a similar condition.
Damaged cable or poor connector	Inspect the setup and compare with a known-good sensor or cable where permitted.
Incorrect acquisition setup	Confirm units, frequency range, resolution and sensor configuration.
Transient process event	Check process trends and repeat under a stable condition if appropriate.

A repeated high reading does not prove a specific fault, but it gives more confidence that the condition is real.

Stage 5: Analyse the evidence in a logical order

Many analysts begin with the alarm list or exception report, but an alarm is a screening device, not a diagnosis. Analyse the alerted point in context.

1. Review the trend

Determine when the change began, how quickly it developed and whether it correlates with maintenance or operating changes. A gradual rise over months suggests a different investigation from a step change immediately after overhaul.

2. Compare related points and directions

Identify where vibration is strongest and how it travels through the machine. Compare driver and driven components, inboard and outboard bearings, and radial and axial directions. Spatial distribution is part of the fault pattern.

3. Review the spectrum

Identify the dominant frequencies and relate them to running speed, harmonics and known machine components. Look for changes in amplitude, new peaks, sidebands, broadband energy and high-frequency content.

4. Review the time waveform where useful

The waveform can reveal impacts, modulation, clipping, looseness and non-steady behaviour that may not be obvious from the spectrum alone.

5. Use additional plots and techniques when justified

Phase, enveloping or demodulation, orbits, shaft centerline, run-up or coast-down data and other techniques can help answer specific questions. Use them because the investigation requires them, not simply because the software offers them.

Overall vibration tells you that energy changed. Frequency, time, phase, location and operating context help explain why.

Stage 6: Build and test fault hypotheses

A diagnosis is stronger when the analyst considers competing explanations. Instead of saying, "There is a 1X peak, therefore the rotor is unbalanced," write a short hypothesis table.

Possible cause	Evidence that would support it	Useful confirmation
Mass unbalance	Dominant 1X response, commonly strongest radially, with behaviour consistent with the rotor and support.	Phase and spatial pattern; inspect for buildup, damage or missing material.
Misalignment	Vibration pattern across the coupling, often with axial and harmonic content depending on the case.	Phase relationship, coupling inspection and alignment check.
Mechanical looseness	Harmonics, nonlinearity, impacts or localized response consistent with a loose interface.	Inspect hold-down bolts, base, bearing fits, guards and structural joints.
Hydraulic excitation	Vibration changes with flow or pressure and may include vane-related or broadband components.	Compare process state, listen for cavitation and review pump operation.

No single row is a universal rule. Machine construction and operating behaviour can alter the pattern. The purpose of the table is to make your reasoning visible and testable.

Stage 7: Convert analysis into a maintenance decision

A useful report answers five questions:

1. Where? Identify the asset and measurement location.
2. What changed? Describe the trend and relevant vibration symptoms.
3. What is the probable condition? State the diagnosis with an appropriate confidence level.
4. What should be done? Recommend inspection, further testing, monitoring or corrective work.
5. When? State urgency based on trend, severity, failure consequence and local criteria.

Avoid vague reports such as "high vibration - check machine." A stronger report might say:

Pump P-204 drive-end horizontal velocity increased from 2.1 to 5.8 mm/s RMS over three weekly measurements under comparable speed and load. The spectrum is dominated by 1X running speed, with the highest response on the pump. The pattern is consistent with probable impeller unbalance. Inspect the impeller for buildup or damage and check base tightness during the next planned opportunity. Continue weekly monitoring and escalate sooner if the trend accelerates or operating behaviour changes.

The severity and timing in a real report must follow the organization's approved alarm philosophy, applicable guidance, machine history, operating risk and engineering judgement. A numerical value should not be copied into a universal action rule without context.

Stage 8: Verify the result and preserve the lesson

After maintenance, collect data under a comparable operating condition. Compare before and after values, spectra, waveforms and process conditions. Record what the maintenance team found and what action was actually performed.

If the vibration falls and the suspected defect is physically confirmed, confidence in the diagnosis increases. If the vibration remains high, do not hide the result. Reassess the hypothesis, measurement and repair quality.

Verification converts an isolated diagnosis into organizational knowledge. It improves future alarm decisions, reports, fault recognition and maintenance planning.

Practical case: a pump with rising 1X vibration

Consider a simplified training example involving a motor-driven centrifugal pump operating at approximately 1,480 rpm, or 24.7 Hz.

1. Understand: The pump normally operates at stable speed and similar flow. The impeller handles a product that can accumulate deposits.
2. Collect: Weekly readings are taken at marked bearing locations using the same sensor and mounting method. Speed, flow and pressure are recorded.
3. Detect: Pump drive-end horizontal velocity rises from 2.1 to 5.8 mm/s RMS over three comparable measurements.
4. Validate: The analyst remounts the sensor, repeats the measurement and checks the nearby points. The increase remains.
5. Analyse: The spectrum is dominated by 24.7 Hz, matching running speed. The response is strongest radially on the pump. The time waveform is mainly periodic, and bearing-condition indicators have not changed significantly.
6. Hypothesize: Impeller unbalance is considered probable, but looseness, support problems and hydraulic effects are also reviewed.
7. Confirm: Hold-down bolts and the base show no obvious looseness. Process conditions are stable. The team plans an impeller inspection.
8. Act: Deposits are found and removed from the impeller. Its condition is checked before the pump is returned to service.
9. Verify: Under a comparable operating condition, vibration falls to 2.2 mm/s RMS and the 1X component reduces substantially.

The lesson is not that every 1X peak means impeller buildup. The lesson is that a diagnosis becomes credible when measurement quality, machine context, pattern, inspection and post-maintenance response agree.

A field checklist for new analysts

• Correct asset and point confirmed
• Safe access and site requirements satisfied
• Speed and operating state recorded
• Sensor, mounting and direction verified
• Measurement repeated if the result is unusual
• Field observations and recent work recorded
• Trend reviewed before isolated plots
• Related points and directions compared
• More than one possible cause considered
• Recommendation states action and urgency
• Post-maintenance verification requested

Final takeaway

Vibration analysis is not a contest to name a fault quickly. It is a disciplined process for reducing uncertainty.

Understand the machine. Plan the measurement. Collect repeatable data. Validate the result. Define the symptoms. Test competing explanations. Communicate the decision clearly. Then verify what happened.

When these steps become habitual, software changes from a collection of plots into a tool for making better maintenance decisions.

Coming in Part 3

Part 3: Displacement, Velocity and Acceleration - What Each Measurement Tells You. We will explain the three common vibration quantities, their units, where each is useful and why selecting the wrong measurement can hide an important machine condition.

Discussion question: Which step of the workflow is most often missed in your workplace - measurement validation, diagnosis confirmation, clear reporting or post-maintenance verification?

References and further learning

• Mobius Institute, Vibration Analysis Category I training material, especially Chapter 4 topics on data acquisition, repeatability, field observations, routes and the start of the analysis process.
• Emerson Process Management, Basic Vibration Analysis - Course 2031, especially the introduction to vibration, measurement parameters, spectra and monitoring fundamentals.
• Vibration Analysis Guide, especially the beginner sections on amplitude, frequency, waveforms and spectra.
• Mobius Institute, Vibration Analysis Faults booklet, used as a fault-pattern reference.

Educational note: The numerical values and pump case in this article are simplified examples, not universal alarm limits. Apply site procedures, approved alarm criteria, equipment-manufacturer guidance and qualified engineering judgement to real machinery.

From Mechanical Maintenance to Vibration Analysis: You Are Not Starting from Zero

From Mechanical Maintenance to Vibration Analysis - Part 1

A mechanical maintenance professional recently asked me an honest question:

I am moving from mechanical maintenance into a vibration analyst role. I know the basics, but my knowledge is limited. How difficult is the transition, and can software such as System 1 help me make an exact diagnosis?

If you are facing the same transition, the most important thing to understand is this: you are not starting from zero.

You already know that machines rarely fail without context. A bearing may run hot because of poor lubrication, excessive load, incorrect fit, contamination or misalignment. A pump may vibrate because of unbalance, looseness, pipe strain, cavitation, resonance or an operating condition far from its preferred range. That mechanical understanding is not separate from vibration analysis. It is one of its foundations.

Your next task is to learn how machine behaviour appears in measurements - and how to turn those measurements into a defensible maintenance decision.

What does a vibration analyst actually do?

A vibration analyst monitors the dynamic behaviour of machinery to detect abnormal conditions, assess their severity, investigate probable causes and communicate what should happen next. The work normally includes:

• understanding the machine, its components, speed, load and normal operating range;
• selecting appropriate measurement locations, directions and settings;
• collecting repeatable vibration data safely;
• reviewing overall values, trends, FFT spectra, time waveforms, phase and other relevant plots;
• identifying patterns associated with faults such as unbalance, misalignment, looseness, bearing damage, gear problems, resonance and electrical effects;
• comparing vibration evidence with process conditions, inspections and maintenance history;
• estimating urgency and recommending a proportionate action; and
• following the machine after corrective work to determine whether its condition improved.

The analyst's output is therefore not merely a graph. It is a clear statement such as:

The pump outboard bearing vibration has increased over the last three weekly measurements. The dominant component is at running speed and is strongest in the radial direction. Before balancing, inspect the base and hold-down bolts, confirm the impeller condition and repeat the measurement under the same operating condition.

Notice the discipline in that statement. It describes the change, identifies the evidence, avoids claiming certainty too early and recommends confirmation checks.

Your maintenance experience is a technical advantage

Existing maintenance knowledge	How it helps in vibration analysis
Bearings, shafts, couplings and gears	Helps you connect frequency patterns to real components and possible failure modes.
Alignment, balancing and fits	Helps you evaluate whether a suspected fault is mechanically credible.
Lubrication and contamination control	Adds context when high-frequency vibration or bearing-condition indicators change.
Maintenance and failure history	Shows whether the signal is new, recurring or related to recent work.
Machine sounds, temperature and physical condition	Provides observations that can support or challenge the vibration evidence.
Work execution and plant constraints	Helps you make recommendations that are safe, specific and practical.

The new knowledge you must add includes vibration terminology, transducers, measurement units, sensor orientation, signal processing, FFT spectra, time waveforms, phase, alarm philosophy and fault-pattern recognition. These subjects can appear difficult at first, but they become manageable when learned through real machines rather than memorized as isolated charts.

How difficult is the transition?

The transition is challenging, but it is realistic. The hardest part is usually not learning that unbalance often produces strong vibration at running speed or that bearing faults can generate characteristic frequencies. The harder part is dealing with uncertainty.

Different faults can produce similar symptoms. One fault can produce several symptoms. Speed and load can change the data. A badly mounted sensor can create misleading results. A single overall value can hide important frequency information. A spectrum that resembles a textbook example may still have a different cause on the actual machine.

That is why competent analysis is a process of building and testing a case:

1. Understand the asset. What machine is it, how is it supported, what components does it contain, and what are its normal operating conditions?
2. Verify the measurement. Was the correct point, direction, sensor mounting and acquisition setup used?
3. Describe the symptom. What changed, where is it strongest, at which frequencies, and under what operating condition?
4. List plausible causes. Which faults could realistically create the observed pattern?
5. Seek confirming evidence. Check other measurement points, directions, phase, waveform, temperature, oil, process data and physical condition.
6. Assess severity and risk. Is the condition stable, deteriorating or immediately dangerous?
7. Recommend and verify. State the next action and confirm the result after maintenance.

A pattern suggests a fault. Corroborating evidence strengthens the diagnosis. Verification after maintenance closes the loop.

What System 1 and similar software can do

Condition-monitoring software is extremely valuable when it is supported by good measurements and good analysis. Bently Nevada describes System 1 as a plantwide condition-monitoring platform built around connectivity, analytics and visualization. It can bring vibration, process and control-system information together so users can trend machine condition and investigate changes in context.

Depending on the installed monitoring hardware, configuration and licensed capabilities, software can help you:

• collect and organize machine-condition data;
• display trends, spectra, waveforms, orbits and other plots;
• compare measurements at different times or operating states;
• configure alarms and notify users when defined conditions are exceeded;
• correlate vibration with speed, load, pressure, temperature or other process variables; and
• preserve evidence for diagnosis, reporting and follow-up.

However, software cannot guarantee an exact diagnosis. It does not automatically know that a sensor was loose, that a machine was unloaded, that pipe strain was introduced during recent maintenance or that the wrong number of motor poles was entered in the database. Automated diagnostics and decision-support rules may highlight probable conditions, but their output must still be tested against the machine and its operating context.

Software provides	The analyst provides
Data storage and visualization	Machine context and interpretation
Trends and alarms	Validation of whether the change is real and significant
Analytical plots and calculated indicators	Competing fault hypotheses and confirmation checks
Decision-support suggestions	Risk-based recommendation and accountability

The software helps you see and organize the evidence. The analyst decides what the evidence means.

A practical example: rising vibration on a motor-pump set

Imagine that System 1 or a portable data-collection program shows rising vibration at the pump outboard bearing. The alert is useful, but it is only the start.

A developing analyst might immediately conclude, "The pump is unbalanced." A disciplined analyst asks more questions:

• Was the pump running at the same speed, flow and pressure during each measurement?
• Is the increase present in the horizontal, vertical and axial directions?
• Is the dominant frequency exactly at running speed, or is it another component?
• Did phase relationships change?
• Is there evidence of harmonics, broadband energy or impacts?
• Were the base, hold-down bolts, coupling and pipe supports inspected?
• Has the impeller recently been repaired, cleaned or exposed to product buildup?
• Could the operating point be causing hydraulic excitation or cavitation?

The final recommendation may be balancing, but it may instead be correcting looseness, inspecting the impeller, improving support, changing the operating condition or collecting additional data. The value of analysis lies in narrowing the possibilities responsibly before time and money are committed.

A practical learning path for your first 90 days

Days 1-30: measurement discipline

• Learn the machines, measurement-point naming and safe access requirements.
• Understand displacement, velocity, acceleration, RMS, peak and peak-to-peak.
• Practise sensor positioning and repeatable mounting.
• Record speed, load, process condition and useful field observations.
• Review normal data with an experienced analyst before studying abnormal cases.

Days 31-60: interpreting the basic plots

• Relate frequency in hertz, cycles per minute and orders to machine running speed.
• Learn what the FFT spectrum and time waveform reveal - and what each can hide.
• Compare overall values with trends and frequency content.
• Study unbalance, misalignment and looseness using actual plant examples.
• Write short observations without making unsupported diagnoses.

Days 61-90: supervised diagnosis

• Build fault hypotheses and list the evidence for and against each one.
• Use multiple points, directions and techniques to confirm a suspected condition.
• Correlate vibration with temperature, lubrication, process and maintenance history.
• Draft recommendations that specify the asset, evidence, urgency and next action.
• Compare before-and-after data whenever corrective work is completed.

This is not a promise that someone becomes an independent expert in 90 days. It is a structure for developing safe habits and useful supervised competence. Experience grows through repeated exposure to machines, faults, corrections and verified outcomes.

Seven rules for a new vibration analyst

1. Never diagnose from one number alone.
2. Always confirm machine speed and operating condition.
3. Question the measurement before questioning the machine.
4. Use trends to understand change, not only alarm status.
5. Treat fault charts as guides, not automatic answers.
6. Ask operators and maintainers what recently changed.
7. Verify the result after maintenance and save the lesson.

Final message to anyone making the transition

Feeling uncertain at the beginning does not mean you are unqualified. It means you understand that diagnosis carries responsibility.

Your mechanical-maintenance background gives you knowledge that software cannot supply: how machinery is assembled, how it is operated, how it is repaired and how faults appear in the field. Vibration analysis adds another way of observing that machinery - often early enough to act before a functional failure occurs.

Learn the fundamentals, collect repeatable data, seek confirming evidence and remain willing to revise your conclusion. Confidence should come from a disciplined process, not from pretending to be certain.

You do not need to know every fault on your first day. You need to know how to observe carefully, ask the right questions and build evidence one step at a time.

Coming in Part 2

Part 2: The Vibration Analysis Workflow - From a Measurement to a Maintenance Decision. We will follow the complete process of understanding the asset, collecting reliable data, analysing the evidence, assessing severity and communicating a recommendation.

Discussion question: If you are moving into vibration analysis, which subject feels most difficult - data collection, spectra, fault diagnosis or using the software?

References and further learning

• Vibration Analysis Guide, beginner guide to machine vibration, particularly the sections introducing amplitude, frequency, waveforms and spectra.
• Emerson Process Management, Basic Vibration Analysis - Course 2031, particularly the introduction to vibration, FFT concepts and monitoring fundamentals.
• Mobius Institute, Vibration Analysis Category I training material, particularly condition monitoring, data collection and introductory fault diagnosis.
• Mobius Institute, Vibration Analysis Faults booklet.
• Bently Nevada, System 1 Asset Health Management Software.

Educational note: Vibration patterns are not universally conclusive. Machine design, sensor type, measurement setup and operating condition affect the data. Follow site safety procedures, equipment-manufacturer guidance and your organization's escalation requirements when assessing machinery.

How to Build an Executable Maintenance Job Plan: Scope, Labour, Materials, Tools and Duration

A maintenance work order is not ready merely because it has a description and a target date. Before it enters the weekly schedule or a shutdown programme, the planner must prove that the work can be performed safely, efficiently and without avoidable delay.

That proof is the job plan. At its centre are five questions: What exactly will be done? Who will do it? Which materials are required? Which tools and equipment are needed? How long will the work take?

A schedule assigns time to work. A job plan establishes whether that work is executable.

1. Define a clear and controlled scope

The scope converts an equipment problem into a specific physical action. It should allow a competent supervisor or technician to understand the job without guessing the planner's intention.

A complete scope identifies:

• the asset, functional location and component affected;
• the defect, condition or failure mode that initiated the work;
• the required corrective or preventive action;
• the physical boundaries of the job;
• activities included and specifically excluded;
• required permits, isolations and safety controls;
• inspection, testing and quality hold points; and
• the acceptance criteria for returning the equipment to service.

Descriptions such as repair pump, check bearing or service motor are not adequate scopes. They do not define the work boundary, completion evidence or recommissioning requirement.

A stronger scope would read:

Isolate pump P-101, remove the coupling guard, disconnect the motor, replace the drive-end and non-drive-end bearings, inspect the shaft and bearing housings, reassemble the unit, perform laser alignment, complete a no-load test and confirm vibration is within the approved limit.

The second description creates a measurable finish condition. It also exposes the labour, materials, tools and duration that must be planned.

2. Build the labour plan from the task sequence

Labour should be estimated after the job has been divided into logical tasks. Assign the required trade, competence, crew size and effort to each task instead of applying one general labour estimate to the entire order.

Activity	Required skill	Crew	Duration
Electrical isolation	Authorised electrician	1	1 hour
Disconnect motor and coupling	Mechanical technicians	2	2 hours
Replace bearings	Mechanical technicians	2	4 hours
Laser alignment	Reliability technician	1	2 hours
Test and recommission	Operations and maintenance	3	1 hour

Duration and labour-hours are not the same

If two technicians perform a four-hour task, the activity duration is four hours but the required effort is eight labour-hours.

Labour-hours = crew size × task duration

This distinction is essential for capacity planning, backlog measurement, contractor forecasting and maintenance-cost control. The planner should also identify whether work can proceed in parallel or whether access and safety restrictions require sequential execution.

3. Identify and verify materials

A material list must cover both major spares and the small consumables that frequently stop a job. Each requirement should include the correct description, part number, quantity, storage location and required-on-site date.

Typical checks include:

• equipment bill of materials and manufacturer manuals;
• drawings, specifications and previous work-order history;
• physical stock availability and serviceable condition;
• material reservation or purchase-requisition status;
• interchangeability and approved substitutes;
• shelf-life, preservation and certification requirements; and
• consumables such as gaskets, seals, grease, cleaning agents and fasteners.

A green stock balance in the CMMS is not enough. The item must be the correct specification, physically available, serviceable and reserved against the work order. When the job is critical, the planner should arrange kitting so that the complete material package is checked before the equipment is released.

4. Specify tools, equipment and access

Standard hand tools may be assumed only when site practice allows it. Every special, calibrated, shared or hired item should be listed and reserved.

Depending on the job, this may include:

• calibrated torque wrenches;
• bearing pullers, heaters and hydraulic presses;
• laser-alignment and vibration-analysis equipment;
• lifting beams, cranes, chain blocks and certified slings;
• scaffolding, platforms and confined-space equipment;
• welding, machining or cleaning equipment;
• test instruments with valid calibration certificates; and
• task-specific personal protective equipment.

Tool planning must include the conditions required to use the tool. A crane without a lift plan, certified operator, suitable access and inspected rigging does not make the lifting task ready. Likewise, a laser-alignment unit without a trained user or correct mounting fixtures is only an item on a list.

5. Estimate a defensible duration

Duration should reflect the planned method, quantities, crew and constraints. Copying a duration from an old work order without checking the conditions creates false precision.

A practical estimate considers:

• work quantity and expected production rate;
• crew size, competence and learning curve;
• equipment accessibility and working position;
• permit, isolation and handover time;
• disassembly, cleaning and inspection;
• material handling, lifting and scaffolding;
• quality inspections and hold points;
• reassembly, torqueing and alignment;
• testing, reinstatement and recommissioning;
• shift arrangements and simultaneous-operation restrictions; and
• identified uncertainty and appropriate contingency.

A useful starting relationship is:

Task duration = work quantity ÷ crew production rate + fixed support time

Where inspection may reveal additional damage, do not hide the uncertainty inside an inflated duration. Create an inspection task, a decision point and prepared contingency activities. This makes the schedule easier to control and allows stakeholders to understand the real source of risk.

6. Check the interfaces around the job

The five planning elements are connected to other readiness requirements. Before scheduling, confirm:

• operations can release the equipment at the required time;
• all energy sources and isolation points are identified;
• approved procedures, drawings and permits are available;
• contractor mobilisation and site access are confirmed;
• quality, safety and operations representatives understand their hold points;
• removed components, waste and contaminated materials have a disposal route; and
• the recommissioning and handover sequence is included.

Common job-planning failures

Failure	Better practice
The scope contains only the symptom	Define the physical action, boundary and acceptance criteria
One labour estimate covers every task	Estimate trade, crew and effort task by task
Only the main spare is reserved	Plan the complete kit, including consumables and certificates
Special tools are listed but not reserved	Confirm availability, condition, calibration and competent users
Duration is copied from history	Re-estimate using the current method, quantity, crew and constraints
Recommissioning is omitted	Include testing, reinstatement, handover and operating acceptance

The work-order readiness gate

A work order should enter the committed schedule only when:

1. The scope and completion criteria are clear.
2. The task sequence and interfaces are understood.
3. Required labour skills, crew size and hours are confirmed.
4. Materials are available, serviceable and reserved.
5. Tools, equipment, access and lifting arrangements are secured.
6. Permits, isolations and safety controls are defined.
7. The duration is supported by quantities, productivity and constraints.
8. Testing, recommissioning and closeout requirements are included.

Conclusion

Good maintenance planning removes uncertainty before equipment is stopped. A complete job plan gives the execution team a clear scope, the right people, verified materials, suitable tools and enough time to perform the work safely.

The result is more than a better work order: it is higher schedule compliance, less waiting time, lower emergency purchasing, safer execution and more reliable equipment history. When these five elements are properly defined, maintenance moves from firefighting toward controlled, repeatable performance.

The Planning Engineer: Turning Scope into Safe, Executable Work

Why the planning engineer matters

Projects and plants rarely fail because nobody created a list of activities. They fail because scope was incomplete, interfaces were missed, approvals arrived late, materials were unavailable, production windows changed, risk controls were not ready, or the schedule did not reflect how the work would actually be performed. The planning engineer sits at the point where these constraints must be made visible and resolved before they become delay, cost or downtime.

In construction, the role connects contract requirements, design information, quantities, productivity, resources, procurement and site logistics to a controlled baseline. In maintenance, it converts notifications and backlog items into executable work orders and weekly or shutdown schedules. In both environments, the planner creates a reliable bridge between management intent and field execution.

Core idea: A schedule shows when work is intended to happen. A plan proves that the work can happen.

Planning is not the same as scheduling

The two disciplines are closely linked, but they answer different questions. Planning defines the method: what must be done, in what scope, with which sequence, labor, materials, tools, permits, drawings and quality requirements. Scheduling places ready work into time, balances it against available capacity and access, and communicates commitments.

Planning asks	Scheduling asks
Is the scope clear and broken into controllable work?	When can the work start and finish?
What logic, method, skills, materials and equipment are required?	How will activities fit within milestones, shifts and access windows?
What risks, permits, drawings and hold points apply?	Which path controls completion and where is float available?
Is the package ready for execution?	Which ready work receives committed resources this week or shutdown?

A technically elegant programme built from unready activities is still unreliable. The planning engineer therefore protects the schedule from false certainty by using readiness criteria before work is committed.

The end-to-end planning cycle

1. Define the scope and acceptance criteria. Confirm the contract, asset boundary, drawings, specifications, quantities, exclusions, quality requirements and completion evidence.
2. Break the scope into controllable work. Use a work breakdown structure for projects or a properly coded work order or task list for maintenance. Each package should have a clear owner and measurable finish condition.
3. Develop the execution method and logic. Sequence the work according to physical constraints, interfaces, access, temporary works, isolations, testing and commissioning. Estimate durations from quantities, productivity and crew composition—not guesswork.
4. Load resources and costs. Identify internal labor, contractors, plant, tools, materials and services. Link budget and resource demand to the activities that consume them.
5. Integrate procurement and information. Track submittals, shop drawings, purchase requisitions, long-lead items, fabrication, delivery and material reservations against need dates.
6. Plan risk, safety and quality controls. Include permits, energy isolation, lifting, work at height, confined space, inspections, test plans and contingency actions as real constraints or activities.
7. Baseline and obtain commitment. Validate logic and calendars, identify the critical path, reconcile resource peaks, agree milestones and secure stakeholder sign-off.
8. Control, learn and close out. Update with credible field data, forecast the effect of variance, issue recovery actions, record actual hours and history, and feed lessons into future norms and job plans.

What the planning phase must produce

A planning phase is complete only when it produces usable control documents. Core planning outputs typically include a time schedule, cost loading, resource loading, cash flow, a manpower histogram, drawing and submittal logs, a procurement log, site layouts, temporary-works plans, cost estimates and risk information.

Output	Decision it supports	Readiness test
Baseline schedule	Sequence, milestones and forecast	Logic is complete; calendars and constraints are justified
Cost/resource loading	Budget timing and capacity	Quantities, rates and crew assumptions are traceable
Cash flow / S-curve	Funding and progress control	Time-phased values reconcile to the approved budget
Manpower histogram	Recruitment, shifts and work fronts	Peaks are achievable and match available supervision
Drawing/submittal log	Information release	Required-on-site dates are linked to review and approval lead time
Procurement log	Long-lead material availability	Approval, order, fabrication, shipment and delivery dates are visible
Risk and site-logistics plan	Safe, feasible execution	Access, lifting, temporary facilities and major hazards are resolved

Maintenance planning: from notification to history

In a work-order system, the notification identifies a defect or potential problem; the work order describes how the maintenance activity will be executed; the backlog holds identified work not yet completed; and the weekly/daily plan converts selected ready work into labor and time commitments.

A complete maintenance job plan should contain:

• a precise equipment location, problem statement, scope and expected condition after completion;
• operations or task-list steps in a safe and practical sequence;
• estimated internal labor by craft, external services and planned hours;
• stock and non-stock components, quantities, reservations and need dates;
• special tools, lifting equipment, access requirements and equipment operating condition;
• PPE, permits, isolations, risk controls, inspection points and test requirements; and
• priority, basic start date, revision/shutdown assignment and completion documentation.

This preparation reduces waiting for instructions, tools and materials. It also creates an auditable history that improves future estimating, task lists and reliability decisions. The goal is not to maximize the number of work orders released; it is to increase the proportion of ready work that crews can complete without avoidable delay.

Shutdown planning and scheduling

Shutdowns compress a large, interdependent scope into a short production window. The definition phase should establish the sales and production forecasts, team, dates, duration, budget and shutdown backlog. Planning then turns that definition into a validated activity list, while scheduling integrates the work, identifies the critical path, optimizes interactions, obtains sign-off and communicates the approved plan. Preparation activities continue in parallel.

The shutdown meeting is a decision forum, not a status-reading exercise. Notifications should be transformed into work orders; duration, material, resources, services and safety requirements should be planned; the backlog should be screened for readiness; and purchase requests should be released early. Once the activity list is integrated, the team identifies and optimizes the critical path before sign-off.

Shutdown rule: Freeze the committed scope through governance, not silence. Emergent work must be assessed for safety, reliability, time, resource and critical-path impact before it enters the schedule.

What a planning engineer does every day

• Facilitates short review meetings to validate new issues, execution feedback and the next planning priorities.
• Walks the job or work front with supervisors, operations and discipline engineers to verify scope and constraints.
• Maintains the baseline, detailed schedules, look-aheads, work orders, backlog status and progress cut-off discipline.
• Checks that materials, drawings, purchase orders, contractors, access and permits will be ready when needed.
• Compares actual progress and hours with the plan, identifies slippage and develops time-phased recovery actions.
• Coordinates production, maintenance, engineering, procurement, warehouse, safety, quality and commercial stakeholders.
• Preserves records: approvals, updates, assumptions, change history, completed work and lessons learned.

Measure the health of the planning system

A useful dashboard connects leading indicators of readiness with lagging indicators of delivery. Definitions should be documented and kept consistent across reporting periods.

Measure	What it reveals	Practical interpretation
Schedule compliance	Execution reliability	Completed committed work divided by scheduled work, using a locally agreed rule
Planning accuracy	Quality of estimates	Variance between planned and actual duration or labor hours by work type
Ready backlog	Future schedule resilience	Weeks of executable work by craft, priority and operating window
Backlog age / overdue work	Risk accumulation	Old high-consequence work requires escalation, not automatic reprioritization
Material reservation usage	Supply readiness discipline	Low use can signal late identification or weak CMMS practice
Critical milestone variance	Forecast stability	Track movement and root cause, not only the latest date
SPI / CPI	Earned-value performance	SPI = EV/PV and CPI = EV/AC; use only when the underlying progress and cost data are credible

A mature work-order system tracks schedule compliance, planning accuracy, aging requests, scheduled ratio, overdue and outstanding work, reservation usage and call-outs. Construction-planning experience adds baseline control, critical-path recovery, resource/cost loading, S-curves and earned-value reporting.

Tools and competencies that matter

Software is necessary, but it does not replace engineering judgment. A capable planning engineer combines four layers of competence:

• Engineering literacy: drawings, quantities, construction methods, equipment behavior, temporary works, commissioning and quality requirements.
• Planning and controls: WBS, CPM logic, calendars, constraints, resource and cost loading, baselines, forecasting, recovery planning and change control.
• Systems and data: Primavera P6 or equivalent, SAP-PM/CMMS, spreadsheets, dashboards such as Power BI, document control and disciplined coding structures.
• Leadership and communication: facilitation, concise reporting, challenge without conflict, problem solving, time management and teamwork under pressure.

The most valuable planner is often the person who can explain a complex variance in plain language, show the decision required, name the owner and state the latest date for action.

A realistic career path

Many strong planning engineers begin in execution roles because field experience teaches production rates, constructability, quality and the real causes of delay. A typical path is site or maintenance engineer, planner/scheduler, planning engineer, senior planning engineer, and then project-controls lead, planning manager or maintenance-planning manager. This progression is common because field responsibility builds the judgment needed for senior planning and controls roles.

To progress, build evidence rather than a software-only CV:

• one baseline schedule with a clear WBS, logic narrative and assumptions register;
• one monthly update showing progress, critical-path movement and a defensible forecast;
• one resource/cost-loaded example with S-curves or histograms;
• one recovery plan linked to root causes, owners and achievable production changes;
• one maintenance job plan or shutdown package demonstrating materials, permits, isolations and readiness; and
• one dashboard that turns planning data into management decisions.

Common failure modes

Failure mode	Better practice
Activities are added before scope is mature	Use planning maturity and readiness gates; retain assumptions explicitly
Durations are copied from old schedules	Estimate from quantities, productivity, crew composition and constraints
The programme contains excessive hard constraints	Model real logic; reserve constraints for genuine contractual or external conditions
Procurement is tracked separately from the schedule	Link submittal, approval, order and delivery milestones to consuming activities
Progress is accepted without evidence	Define measurement rules and use field quantities, confirmations or approved milestones
Recovery means compressing dates on paper	Change method, sequence, work fronts, shifts or resources and quantify the consequence
Shutdown scope keeps growing	Apply a controlled emergent-work process with explicit critical-path and risk impact

The planning engineer's 12-point readiness check

1. Scope and completion criteria are clear.
2. The responsible owner and work front are identified.
3. Approved drawings, specifications and quantities are available.
4. The method, sequence, interfaces and hold points are defined.
5. Labor skills, crew size, hours and supervision are estimated.
6. Materials, services, tools and equipment are reserved or ordered against need dates.
7. Access, production windows and equipment condition are confirmed.
8. Safety risks, permits and isolations are planned.
9. Quality inspections, testing and commissioning are included.
10. Logic, duration, calendars and constraints are defensible.
11. Cost, resource and critical-path effects are visible.
12. Execution feedback and completion history will be captured.

Conclusion

Planning engineering is the discipline of making work executable before the clock starts. Its best outputs are not colourful reports; they are fewer surprises, safer handovers, stable commitments, improved resource use, lower downtime and earlier decisions. Whether the environment is a construction project, a preventive-maintenance programme or a major shutdown, the planner earns trust by connecting the schedule to physical reality—and by updating that connection honestly when reality changes.

Centrifugal Pump Selection and Design: Choosing the Right Operating Point

Pump selection is not simply a matter of matching flow and head. A reliable design chooses an operating point that balances hydraulic performance, energy efficiency, cavitation margin, motor capacity and long-term equipment life.

RCM GLOBAL selection example
For this 50 Hz application, the required duty point is 90 m³/h at 74.2 m head.

Selected Centrifugal Pump

• Required duty point: 90 m³/h at 74.2 m
• Selected pump: 4×2–10 centrifugal pump
• Operating speed: 2,950 rpm
• Efficiency at duty point: approximately 71%
• Power requirement: approximately 25.8 kW
• NPSHr: approximately 2.85 m

Pump performance curve used to evaluate the operating point, efficiency, power demand and NPSH requirement.

Why the Operating Point Matters

1. Operate Near the Best Efficiency Point

The Best Efficiency Point (BEP) is the region where a centrifugal pump converts input power into hydraulic energy most efficiently. Operating reasonably close to BEP generally reduces hydraulic imbalance, vibration, recirculation, heat generation and premature wear of bearings, seals and impellers.

Continuous operation too far to the left or right of BEP can shorten equipment life even when the pump still appears to meet the required flow and head.

2. Check the NPSH Margin Carefully

The system must provide sufficient Net Positive Suction Head Available (NPSHa) above the pump’s Net Positive Suction Head Required (NPSHr). For this selection, NPSHr is approximately 2.85 m at the duty point.

Adequate margin helps prevent cavitation, which can cause noise, vibration, loss of performance, pitting damage and accelerated deterioration of seals and bearings. Final verification should use the actual liquid temperature, vapour pressure, site elevation, suction losses and tank operating level.

3. Verify Motor Sizing Across the Curve

The selected motor must safely handle the pump’s power demand across the expected operating range—not only at the nominal duty point. Include the applicable service factor, drive efficiency, fluid specific gravity and any credible maximum-flow condition.

The approximate requirement of 25.8 kW should therefore be checked against the full power curve before the final motor rating is selected.

4. Review the Full Pump Curve

A single point cannot describe the behaviour of a centrifugal pump. Review the head, efficiency, power and NPSHr trends together with the system curve. Also confirm the allowable operating range, minimum continuous stable flow and run-out conditions.

Engineering Checks Before Final Selection

• Confirm the design, normal, minimum and maximum flow cases.
• Build or verify the system resistance curve, including static head and friction losses.
• Check liquid density, viscosity, temperature, solids content and corrosiveness.
• Confirm suction conditions and an appropriate NPSH margin.
• Verify materials of construction, seal arrangement and bearing design.
• Check motor rating, starting method and available electrical supply.
• Assess control philosophy, including throttling, bypass control or variable-speed operation.
• Review maintainability, spare-parts availability and lifecycle cost.

Reliability and Lifecycle Benefits

A well-selected pump improves energy efficiency, reduces maintenance costs and increases system reliability. It also lowers the probability of cavitation, chronic vibration, seal leakage, bearing failure and repeated corrective maintenance.

The most economical pump is therefore not always the unit with the lowest purchase price. The better choice is the pump that operates reliably within the system’s real duty range while delivering acceptable lifecycle cost.

Conclusion

Reliable pump design comes from understanding the complete interaction between the pump curve, system curve and operating conditions. Selecting close to BEP, maintaining adequate NPSH margin, checking motor capacity and reviewing the full operating envelope produce a more robust engineering solution.

RCM GLOBAL shares practical engineering knowledge that supports dependable rotating equipment and better maintenance decisions.

View the original LinkedIn post

Engineering note: This example is for educational discussion. Final pump selection must be verified against complete process data, manufacturer documentation, applicable standards and site-specific operating conditions.

Pump selection is not simply a matter of matching flow and head. A reliable design chooses an operating point that balances hydraulic performance, energy efficiency, cavitation margin, motor capacity and long-term equipment life.

RCM GLOBAL selection example
For this 50 Hz application, the required duty point is 90 m³/h at 74.2 m head.

Selected Centrifugal Pump

• Required duty point: 90 m³/h at 74.2 m
• Selected pump: 4×2–10 centrifugal pump
• Operating speed: 2,950 rpm
• Efficiency at duty point: approximately 71%
• Power requirement: approximately 25.8 kW
• NPSHr: approximately 2.85 m

Pump performance curve used to evaluate the operating point, efficiency, power demand and NPSH requirement.

Why the Operating Point Matters

1. Operate Near the Best Efficiency Point

The Best Efficiency Point (BEP) is the region where a centrifugal pump converts input power into hydraulic energy most efficiently. Operating reasonably close to BEP generally reduces hydraulic imbalance, vibration, recirculation, heat generation and premature wear of bearings, seals and impellers.

Continuous operation too far to the left or right of BEP can shorten equipment life even when the pump still appears to meet the required flow and head.

2. Check the NPSH Margin Carefully

The system must provide sufficient Net Positive Suction Head Available (NPSHa) above the pump’s Net Positive Suction Head Required (NPSHr). For this selection, NPSHr is approximately 2.85 m at the duty point.

Adequate margin helps prevent cavitation, which can cause noise, vibration, loss of performance, pitting damage and accelerated deterioration of seals and bearings. Final verification should use the actual liquid temperature, vapour pressure, site elevation, suction losses and tank operating level.

3. Verify Motor Sizing Across the Curve

The selected motor must safely handle the pump’s power demand across the expected operating range—not only at the nominal duty point. Include the applicable service factor, drive efficiency, fluid specific gravity and any credible maximum-flow condition.

The approximate requirement of 25.8 kW should therefore be checked against the full power curve before the final motor rating is selected.

4. Review the Full Pump Curve

A single point cannot describe the behaviour of a centrifugal pump. Review the head, efficiency, power and NPSHr trends together with the system curve. Also confirm the allowable operating range, minimum continuous stable flow and run-out conditions.

Engineering Checks Before Final Selection

• Confirm the design, normal, minimum and maximum flow cases.
• Build or verify the system resistance curve, including static head and friction losses.
• Check liquid density, viscosity, temperature, solids content and corrosiveness.
• Confirm suction conditions and an appropriate NPSH margin.
• Verify materials of construction, seal arrangement and bearing design.
• Check motor rating, starting method and available electrical supply.
• Assess control philosophy, including throttling, bypass control or variable-speed operation.
• Review maintainability, spare-parts availability and lifecycle cost.

Reliability and Lifecycle Benefits

A well-selected pump improves energy efficiency, reduces maintenance costs and increases system reliability. It also lowers the probability of cavitation, chronic vibration, seal leakage, bearing failure and repeated corrective maintenance.

The most economical pump is therefore not always the unit with the lowest purchase price. The better choice is the pump that operates reliably within the system’s real duty range while delivering acceptable lifecycle cost.

Conclusion

Reliable pump design comes from understanding the complete interaction between the pump curve, system curve and operating conditions. Selecting close to BEP, maintaining adequate NPSH margin, checking motor capacity and reviewing the full operating envelope produce a more robust engineering solution.

RCM GLOBAL shares practical engineering knowledge that supports dependable rotating equipment and better maintenance decisions.

View the original LinkedIn post

Engineering note: This example is for educational discussion. Final pump selection must be verified against complete process data, manufacturer documentation, applicable standards and site-specific operating conditions.