The Full P-F Curve: Reliability Is Decided Long Before the First Alarm

Most people meet the P-F curve as a picture of the last stretch of an asset's life: a defect appears, a condition-monitoring tool picks it up, and failure follows. That part matters, but it's only the right-hand end of the story.

An asset's reliability is shaped much earlier, in how it was designed, specified, installed, and maintained in its first months of service. By the time a failure becomes detectable, most of the damage is already locked in. The strongest reliability programs work on the entire curve, not just the end of it.

This article walks the curve from left to right.

Stage 1: Design and specification, where reliability is born

Every asset has an inherent reliability, which is the best performance it can deliver given how it was designed, built, and selected. No maintenance program can lift an asset above that ceiling, and a poor design will cost you for its entire life.

Decisions made here include:

• Selection for the real duty. Was the pump chosen for the duty it actually runs at, or for the lowest purchase price? A pump operating far from its best efficiency point will suffer for years, no matter how well it's looked after.
• Maintainability. Can a technician reach the grease point, the sight glass, and the alignment fixtures? If not, the work will be done badly or not at all.
• Condition-monitoring readiness. Are there sensor mounting points, test ports, and accessible inspection windows, or will you have to retrofit them?
• Specifications and standards. Clear requirements for balance grades, lubrication, sealing, and materials prevent defects from being built in at the factory.

The cheapest time to fix a reliability problem is at the drawing stage. The most expensive time is after it has started failing.

Stage 2: Procurement and receiving

Reliability can be lost between the purchase order and the plant floor. Components get damaged in shipping, stored in humid warehouses, handled roughly, or substituted with "equivalent" parts that aren't. Bearings left sitting on a shelf in vibrating or contaminated conditions can arrive at installation already compromised.

Good practice includes receiving inspection, proper storage, and controlled substitutions, so that what you install is what you intended to install.

Stage 3: Installation and commissioning, where defects get built in

This is where many of the failures you'll see years later actually begin. Installation defects are the ones nobody sees, because the machine starts up and runs. But the curve has already started moving.

Common examples:

• Misalignment between driver and driven equipment, loading bearings and seals before the first hour of service
• Soft foot and poor baseplate or grout condition, distorting casings and shifting alignment
• Pipe strain pulling on pump or compressor casings
• Incorrect fits and torque on bearings, couplings, and fasteners
• Contamination introduced during assembly, such as dirt in bearings, debris in lube systems, or flushing that was skipped
• Skipped commissioning checks, so baseline vibration, temperature, and electrical signatures were never recorded

That last point matters for everything that follows. Without a baseline from a healthy, correctly installed machine, you have nothing to compare future readings against.

Early-life failures, often called infant mortality, are largely a quality-of-installation problem. A machine that is installed well starts its curve from a stronger position and takes longer to reach the point where defects appear.

Stage 4: Precision maintenance, protecting the asset while it runs

Once an asset is running, the job is to keep it away from the conditions that cause defects. This is often called precision maintenance or proactive maintenance, and it works on root causes rather than symptoms.

It includes:

• Precision alignment and balancing to defined tolerances
• Contamination control for lubricants, including clean storage, filtration, breathers, and sealing
• Correct lubricant, quantity, and interval, because both under-lubrication and over-lubrication cause failures
• Proper fastening, fitting, and assembly practices, with torque values and procedures that are actually followed
• Operating within design limits, supported by operators who understand what normal looks like

The key idea is that you're extending the time before a defect ever starts. Everything to the left of Point P is your chance to push that point further out in time, and no later technique can do that for you.

Stage 5: Point P, the first detectable defect

Despite good design, installation, and care, defects will eventually form. Point P is the moment one can first be detected, and it marks the start of the P-F interval.

This is where condition-based and predictive maintenance come in. Techniques such as ultrasound, vibration analysis, oil analysis, motor circuit analysis, and infrared thermography are all ways of looking for P. Each has a different sensitivity, so each sees the defect at a different point along the curve. The earlier a technique can detect the problem, the longer the warning you have.

Predictive maintenance is only valuable if three things are true: the technique matches the failure mode, the inspection frequency is shorter than the P-F interval, and someone acts on what is found.

Inside the P-F interval, your window of control

Between P and F, you have options. The asset still works, but it is telling you it is on its way out. This is the zone for:

• Planned corrective work, scheduling the repair for a convenient window
• Ordering parts at normal lead times rather than expediting
• Planning the job properly, with the right crew, tools, permits, and procedures
• Intervening on the cause, such as fixing the misalignment or contamination source rather than just replacing the part again

As the defect progresses along the interval, your choices narrow. Detection by obvious noise, heat, or smell happens late, and by then the schedule is no longer yours.

Where preventive maintenance fits

Preventive maintenance, meaning time- or usage-based tasks such as inspections, lubrication, cleaning, and scheduled replacements, supports the whole curve rather than sitting at a single point on it.

Used well, PM keeps conditions healthy and catches problems through routine inspection. Used badly, it causes harm:

• Over-maintenance can introduce defects through unnecessary intrusive work, incorrect reassembly, or contamination
• Calendar-based replacement of healthy components wastes parts and labor, and can restart the machine's early-life risk
• Mismatched tasks that don't address how the asset actually fails give a false sense of security

A good strategy uses PM where failure is age-related or where the task prevents the cause, and uses condition-based methods where failure is random but detectable.

Stage 6: Functional failure and reactive maintenance

Point F is where the asset can no longer do its job. When you reach it without having detected the problem, you're in reactive mode: breakdown repairs, emergency orders, overtime, and unplanned downtime.

Reactive work isn't just more expensive per job. It also:

• Often causes secondary damage to connected components
• Creates safety exposure from rushed, unplanned work
• Disrupts the planned schedule, pushing other work further behind
• Pulls skilled people away from proactive tasks, which feeds the same cycle

Plants that stay reactive often do so because the firefighting consumes the time they'd need to fix the underlying causes.

Moving your program to the left

The goal isn't to become better at emergency repair. It's to shift your effort leftward, so that fewer assets reach Point P early, and those that do are caught with plenty of time to act.

A practical way to start:

1. Rank assets by criticality, so effort goes where failure hurts most.
2. Identify the dominant failure modes for those assets and how each one starts.
3. Audit installation and precision-maintenance practices for your critical equipment, because this is where hidden defects often come from.
4. Match detection methods to failure modes and set inspection intervals shorter than the P-F interval.
5. Close the loop by tracking findings, planning the work, and eliminating root causes so that the same defect doesn't return.

How KANE Industrial Consulting can help

The full P-F curve is the philosophy behind KANE, and it's why our logo is built from it. We help plants work across the whole curve rather than only at the end of it, including:

• Criticality and failure-mode assessments to focus effort where it matters most
• Review of installation, commissioning, and precision-maintenance practices to stop defects at the source
• PM and condition-monitoring strategy optimization, aligned to how your equipment actually fails
• Planning and scheduling workflows that turn early detection into planned work

If you'd like to find out where your current program sits on the curve, contact us at adiln@kane-ic.com or (657) 478-8727.