5.3 Reliability-Centered Maintenance (RCM) & Maintenance Tactics

Reliability-Centered Maintenance (RCM) & Maintenance Tactics

Quick Answer: Reliability-Centered Maintenance (RCM) determines what must be done so an asset continues to fulfill required functions in its operating context. A conforming process defines functions and failures, examines failure modes and effects, evaluates consequences, and selects technically feasible and worthwhile proactive tasks or appropriate default actions. FMEA scoring can support prioritization, but RPN is not the RCM decision rule.

The Philosophy and Standard of RCM (SAE JA1011)

Following Nowlan and Heap's commercial aviation breakthrough, the methodology was formalized for broader industry by John Moubray in his seminal work Reliability-Centered Maintenance (RCM II). SAE International publishes SAE JA1011: "Evaluation Criteria for Reliability-Centered Maintenance (RCM) Processes."

SAE JA1011 provides criteria for evaluating whether a process qualifies as RCM. In substance, the process determines the actions needed for an asset to continue meeting required functions in its present operating context.

The Fundamental Paradigm Shift

Traditional maintenance focuses on preserving the physical asset (e.g., "keeping the pump running"). RCM focuses on preserving the function of the asset (e.g., "keeping 500 gpm of cooling water flowing to the heat exchanger").

Furthermore, RCM recognizes that all failures do not matter equally. In the past, maintenance tried to prevent every conceivable failure regardless of cost. RCM asserts that the objective of maintenance is not to avoid all failures, but to manage the consequences of failures in the most cost-effective and safe manner possible.

---

The 7 Fundamental RCM Questions (SAE JA1011)

For a process to meet the SAE JA1011 evaluation criteria, its analysis must address the following seven questions strictly in sequence:

1. Functions: What are the functions and associated desired standards of performance of the asset in its present operating context?
   • Primary Functions: The core reason the asset exists (e.g., "To pump crude oil from Storage Tank 101 to Distillation Column 2 at a minimum rate of 1,200 barrels per hour at 120 psig").
   • Secondary Functions: Essential secondary obligations including safety, environmental containment, structural support, instrumentation, hygiene, and aesthetics (e.g., "To contain all process fluid with zero external leakage").
2. Functional Failures: In what ways can the asset fail to fulfill its functions?
   • Total loss of function (zero flow).
   • Partial loss of function (flow below 1,200 bph or pressure below 120 psig).
   • Exceeding functional limits (over-pressurization, excessive vibration, fluid overheating).
3. Failure Modes: What causes each functional failure?
   • Detailed, component-level mechanisms that cause the functional failure (e.g., "Slurry pump impeller vane erosion due to abrasive silicon particulates; mechanical seal O-ring chemical degradation due to solvent incompatibility").
   • A defensible RCM analysis identifies failure modes reasonably likely in that operating context, including relevant history, degradation mechanisms, and human error.
4. Failure Effects: What happens when each failure occurs?
   • A factual narrative describing what occurs locally and globally: What physical evidence emerges (smoke, noise, visual alarms)? What safety or environmental hazards are created? Does the line trip immediately or derate? What physical damage is inflicted on adjacent machinery? How long does repair take?
5. Failure Consequences: In what way does each failure matter?
   • Failures are categorized into four distinct consequence categories:
     • Hidden Failure Consequences: The failure mode gives no direct warning to the operating crew during normal operations (e.g., an emergency pressure relief valve seized shut, or a backup diesel generator with dead starting batteries). Hidden failures carry no direct immediate consequence, but expose the organization to multiple catastrophic failures if a primary event occurs.
     • Safety and Environmental Consequences: The failure mode could injure or kill a person, or violate an environmental permit/standard.
     • Operational Consequences: The failure affects production capacity, product quality, customer delivery, or incurs significant operational repair costs.
     • Non-Operational Consequences: The failure involves only the direct cost of repair, with zero impact on safety, the environment, or operational throughput.
6. Proactive Tasks: What can be done to predict or prevent each failure?
   • Identifying technically feasible and economically worthwhile proactive maintenance tasks (On-Condition / Predictive, Scheduled Restoration, Scheduled Discard).
7. Default Actions: What should be done if a suitable proactive task cannot be found?
   • If a proactive task is not feasible or not cost-effective, default actions must be selected:
     • For hidden failures: Failure-Finding Tasks (periodic functional testing).
     • For safety/environmental failures: Mandatory Engineering Redesign or operational physical reconfiguration.
     • For operational/non-operational failures: Deliberate Run-to-Failure (RTF).

---

Failure Mode, Effects, and Criticality Analysis (FMECA) & RPN Calculation

A core tool deployed during Step 3, 4, and 5 of RCM is Failure Mode, Effects, and Criticality Analysis (FMECA). FMECA extends standard qualitative FMEA by assigning quantitative risk prioritization using the Risk Priority Number (RPN):

$\text{RPN} = \text{Severity (S)} \times \text{Occurrence (O)} \times \text{Detection (D)}$

Standard 1 to 10 Evaluation Scales

• Severity (S): Evaluates the worst-case consequence of the failure effect:
  • 10: Catastrophic hazard involving human fatality, off-site toxic release, or total plant destruction without warning.
  • 7–8: Major operational disruption; severe line stoppage; significant environmental permit violation.
  • 4–6: Moderate operational impact; minor line derate; in-plant containment.
  • 1–3: Minor cosmetic issue; negligible economic repair cost.
• Occurrence (O): Evaluates the statistical likelihood or frequency of the failure mode occurring:
  • 10: Chronic failure; almost certain to occur repeatedly (> once per month).
  • 7–8: High frequency; occurs several times per operating year.
  • 4–6: Moderate frequency; isolated historical occurrences (every 1–3 years).
  • 1–3: Remote or highly unlikely; once in equipment lifetime (> 10 years).
• Detection (D): Evaluates the likelihood that current control mechanisms, inspections, or sensors will detect the failure defect before the functional failure occurs:
  • 10: Undetectable; zero indication; hidden failure mode.
  • 7–8: Low likelihood of detection; requires complex offline teardown or lab metallurgy.
  • 4–6: Moderate likelihood; detectable by routine manual maintenance route (monthly handheld vibration or visual inspection).
  • 1–3: Almost certain detection; automated 24/7 continuous online SCADA vibration/temperature monitoring with automatic trip interlocks.

The Critical "RPN Trap" Warning for CMRP Candidates

A frequent error in maintenance engineering is blindly sorting FMECA results by numerical RPN and addressing only the highest scores. Consider these two failure modes:

• Mode X: Severity = 10, Occurrence = 2, Detection = 2 $\to \text{RPN} = 40$
• Mode Y: Severity = 3, Occurrence = 6, Detection = 5 $\to \text{RPN} = 90$

A simple RPN cutoff can rank a high-severity, low-occurrence event below a minor frequent event. The team should therefore use severity or consequence gates under its approved risk criteria and review high-consequence modes even when the product score is modest. This is a risk-control principle; the exact numeric gate is facility defined.

---

Worked FMEA Table: Critical Process Centrifugal Slurry Pump

---

Maintenance Tactic Selection Logic: The Decision Framework

RCM uses consequence and task-effectiveness logic to select an applicable, technically feasible, and worthwhile policy for each failure mode:

1. On-Condition Maintenance (Condition-Based Maintenance / PdM)

• Applicability: Feasible whenever a failure mode exhibits a measurable P-F Interval (the time elapsed between Point P, the first moment a potential failure is detectable, and Point F, the onset of functional failure).
• Task interval: Set the interval short enough to provide a useful detection and response opportunity, considering variability in the P-F interval, inspection sensitivity, consequence, and time needed to act. One-half of a stable P-F interval is a common heuristic, not a guarantee or universal rule.
• Applications: On-condition work can be suitable when a detectable potential-failure condition exists, the P-F interval is useful, and the task is effective. The historic 89% result should not be generalized to all current assets. Vibration analysis, oil ferrography, thermography, and motor circuit analysis detect impending failure weeks or months before breakdown.

2. Time-Directed Maintenance (Preventive Maintenance — Restoration or Discard)

• Applicability: Scheduled restoration or discard is technically feasible when there is an identifiable age or usage relationship and the task restores resistance to failure or removes the item before an unacceptable probability of failure. A fitted Weibull shape may provide evidence, but $\beta>1$ alone does not establish an effective interval.
• Economic Justification: The cost of the scheduled task must be significantly less than the total cost of an unplanned failure (including secondary collateral damage and downtime).

3. Failure-Finding Tasks (Hidden Failure Management)

• Applicability: Exclusively reserved for hidden failure modes affecting safety, protection, or emergency systems (e.g., high-level alarm switches, emergency fire pumps, safety relief valves, uninterruptible power supply backup batteries).
• The Failure-Finding Interval (FFI): Calculated to ensure that the unrevealed availability of the protective system meets the required safety integrity level: $\text{FFI} = 2 \times M_{\text{Availability Target}} \times \text{MTBF}_{\text{Protective Device}}$

4. Run-to-Failure (RTF / Intentional Corrective Maintenance)

• Applicability: A deliberate, engineered decision for non-critical assets (Class C) where failure consequences are purely economic, and proactive monitoring costs exceed the cost of replacement.
• Pre-requisites: The failure mode must have zero safety consequences, zero environmental impact, zero production bottleneck effect, and must not cause cascading secondary damage to adjacent machinery.