5.1 Determining Equipment Reliability Expectations

Determining Equipment Reliability Expectations

Quick Answer: Determining equipment reliability expectations begins with defining what the business requires from an asset within its specific operational context. Rather than treating equipment reliability as an abstract or universal constant, reliability professionals conduct RAM analysis (Reliability, Availability, Maintainability) to quantify performance envelopes, distinguish between Inherent Availability ($A_i$) and Operational Availability ($A_o$), and deploy Asset Criticality Ranking (ACR) to prioritize resources based on safety, environmental, throughput, and financial risks.

Defining Reliability in an Operational Context

For this lesson, use the classical engineering definition of reliability:

$\text{Reliability } R(t) = P(\text{Asset performs its intended function without failure for duration } t \text{ under stated operating conditions})$

This classical definition contains four inseparable components:

1. Probability: A quantified statistical value between 0.0 and 1.0 (or 0% to 100%) indicating the likelihood of surviving without functional failure over time $t$.
2. Intended Function & Performance Standards: The explicit operational duty the asset was procured to perform. A pump is not merely required to "run"; it must deliver 450 gallons per minute of sulfuric acid at 65 psi. If output drops to 300 gpm, the asset has experienced a functional failure, even if the motor continues to rotate.
3. Specified Time Duration ($t$): The operational mission time, run cycle, or campaign duration over which reliability is demanded (e.g., surviving a 4,000-hour continuous operating run between planned outages).
4. Stated Operating Conditions (The Operational Context): The physical, process, and environmental environment in which the asset is operated.

The Primacy of Operational Context

A central premise of modern reliability engineering is that machinery does not possess an inherent reliability in isolation. Two identical ANSI centrifugal pumps manufactured on the same assembly line by the same OEM will exhibit radically different operational life spans depending on their operational context:

• Pump 1 (Mild Context): Pumps clean, ambient chilled water at constant flow near its Best Efficiency Point (BEP) in an air-conditioned indoor facility, running continuously 24/7 with clean electrical power.
• Pump 2 (Severe Context): Pumps abrasive mineral slurry at 195°F with high solids concentration, exposed to coastal saltwater humidity, subject to frequent start-stop cycles, line throttling, and cavitation induced by fluctuating suction head.

Assigning Pump 2 the same MTBF target as Pump 1 without considering duty, environment, design, and care strategy creates an unsupported expectation. CMRP Function 3.1 calls for identifying reliability goals and process expectations; those expectations should be defined for the actual operating context.

---

RAM Analysis: Reliability, Availability, and Maintainability

To establish rigorous baseline expectations, reliability engineers deploy RAM Analysis—a comprehensive modeling framework evaluating three interdependent parameters: Reliability ($R$), Availability ($A$), and Maintainability ($M$).

1. Reliability ($R(t)$)

Reliability measures the probability of zero failures over operating time $t$. Under the classical constant failure rate assumption (governed by the exponential distribution where failure rate $\lambda = 1 / \text{MTBF}$):

$R(t) = e^{-\lambda t} = \exp(-\lambda t) = \exp\left(-\frac{t}{\text{MTBF}}\right)$

Where:

• $t$ = Mission duration or required operating hours.
• $\lambda$ (Lambda) = Failure rate (failures per operating hour).
• MTBF = Mean Time Between Failures (total operating hours divided by number of functional failures during that operating time).

2. Maintainability ($M(t)$)

Maintainability is an inherent design and operational characteristic reflecting how easily and rapidly a failed asset can be restored to full operating capability. Mathematically, it is the probability that a failed system will be restored to specified operating conditions within time $t$ using prescribed maintenance procedures and resources:

$M(t) = 1 - e^{-\mu t} = 1 - \exp(-\mu t) = 1 - \exp\left(-\frac{t}{\text{MTTR}}\right)$

Where:

• $\mu$ (Mu) = Repair rate (repairs completed per hour, $\mu = 1 / \text{MTTR}$).
• MTTR = Mean Time To Repair (total active corrective repair hours divided by the total number of corrective maintenance events).

Maintainability is governed by physical equipment design (e.g., modular components, accessible bolting, quick-disconnect fittings, lifting lugs), technician skills, specialized tooling availability, and clear job plans.

3. Availability ($A$)

Availability represents the percentage of planned operating time during which an asset is mechanically capable of performing its required function. Availability is the mathematical synthesis of Reliability and Maintainability. In industrial asset management, two distinct formulations must be mastered:

Inherent Availability ($A_i$)

Inherent Availability measures pure equipment design capability under ideal support conditions. It reflects what the equipment can achieve when no external supply chain, logistical, administrative, or operational delays exist:

$A_i = \frac{\text{MTBF}}{\text{MTBF} + \text{MTTR}}$

• Inclusions: Mean operating time between failures (MTBF) and active corrective maintenance repair time (MTTR).
• Exclusions: Scheduled preventive maintenance, turnaround outages, parts retrieval waiting times, technician travel time, and administrative lockout/tagout (LOTO) delays.

Operational Availability ($A_o$)

Operational Availability reflects the true, real-world availability experienced by plant operations. It accounts for all downtime sources—both planned and unplanned—including logistical friction:

$A_o = \frac{\text{MTBM}}{\text{MTBM} + \text{MDT}}$

Where:

• MTBM (Mean Time Between Maintenance): Encompasses the mean operating interval between all maintenance interruptions, including both unscheduled corrective repairs and scheduled preventive maintenance shutdowns.
• MDT (Mean Down Time): Total elapsed clock downtime per maintenance event, including active repair time, technician mobilization, diagnostic troubleshooting, parts acquisition delays from the storeroom, administrative permitting (LOTO, hot work permits), post-repair calibration, and operational startup/testing.

---

RAM Metric and Formula Reference Table

The following table summarizes the core RAM parameters, standard mathematical formulations, input variables, and operational boundaries:

---

Step-by-Step Worked RAM Calculation: Inherent vs. Operational Availability

To see how these concepts function in practice, consider a critical boiler feedwater pump operating in a power generation facility over an annual operating calendar of 8,760 hours.

==================================================================================
STEP-BY-STEP CALCULATION: INHERENT VS. OPERATIONAL AVAILABILITY
==================================================================================
FACILITY PROFILE: High-Pressure Boiler Feedwater Pumping Unit
- Planned Annual Operating Demand: 8,760 hours
- Number of Unscheduled Functional Failures: 4 events
- Total Active Corrective Repair Time (Craft Wrench Time): 32 hours
- Logistical Delays (Parts delivery, travel, shift changes): 24 hours
- Administrative Delays (LOTO permitting, cooldown, testing): 16 hours
- Scheduled Preventive Maintenance Outages: 4 planned events totaling 48 hours

----------------------------------------------------------------------------------
CALCULATION 1: INHERENT AVAILABILITY (Ai)

Step 1: Calculate Actual Operating Time
        Actual Operating Time = Planned Demand - Total Unscheduled Downtime
        Total Unscheduled Downtime = Active Repair (32h) + Logistics (24h) + Admin (16h)
                                   = 72 hours
        Actual Operating Time = 8,760 - 72 = 8,688 operating hours

Step 2: Calculate Mean Time Between Failures (MTBF)
        MTBF = Actual Operating Hours / Number of Failures
        MTBF = 8,688 / 4 = 2,172.0 hours

Step 3: Calculate Mean Time To Repair (MTTR)
        MTTR = Total Active Corrective Repair Hours / Number of Failures
        MTTR = 32 / 4 = 8.0 hours per repair

Step 4: Compute Inherent Availability (Ai)
        Ai = MTBF / (MTBF + MTTR)
        Ai = 2,172 / (2,172 + 8) = 2,172 / 2,180
        Ai = 0.9963 or 99.63%
----------------------------------------------------------------------------------
CALCULATION 2: OPERATIONAL AVAILABILITY (Ao)

Step 1: Calculate Total Maintenance Events (Planned + Unplanned)
        Total Events = 4 Unscheduled Failures + 4 Scheduled PMs = 8 maintenance events

Step 2: Calculate Mean Time Between Maintenance (MTBM)
        MTBM = Actual Operating Hours / Total Maintenance Events
        MTBM = 8,688 / 8 = 1,086.0 hours

Step 3: Calculate Mean Down Time (MDT)
        Total Down Time = Unscheduled (72h) + Scheduled PM (48h) = 120 total downtime hours
        MDT = Total Down Time / Total Maintenance Events
        MDT = 120 / 8 = 15.0 hours per event

Step 4: Compute Operational Availability (Ao)
        Ao = MTBM / (MTBM + MDT)
        Ao = 1,086 / (1,086 + 15) = 1,086 / 1,101
        Ao = 0.9864 or 98.64%

DIAGNOSTIC INTERPRETATION:
While the inherent equipment design demonstrates an extraordinary 99.63% availability,
real-world operational friction (logistical waiting times, administrative permit delays,
and planned preventive maintenance) depresses actual availability to 98.64%. The 0.99%
gap represents 86.7 hours of lost generation capacity—highlighting that improvement
efforts must target storeroom staging, LOTO streamlining, and PM optimization rather
than redesigning the physical pump casing.
==================================================================================

---

Asset Criticality Ranking (ACR) Methodologies

A maintenance organization that treats every asset as equally important is fundamentally flawed. If a non-critical effluent sump pump and a primary chemical reactor feed compressor are both broken, which work order receives priority? Without an objective, standardized ranking system, work prioritization degenerates into political infighting or "who screams the loudest."

Asset Criticality Ranking (ACR) is a structured engineering process that assigns an objective risk score to each physical asset based on the classic risk formulation:

$\text{Asset Criticality Risk} = \text{Consequence of Failure (CoF)} \times \text{Likelihood / Probability of Failure (PoF)}$

Multidimensional Consequence Criteria

A cross-functional criticality method can evaluate failure consequences across four dimensions, using a controlled ordinal scale such as 1 to 5 or 1 to 10:

1. Personnel Safety and Occupational Health:
   • Catastrophic: Potential employee or community fatality; off-site toxic exposure (Score = 5).
   • Major: Lost-time injury; irreversible disability; OSHA recordable (Score = 4).
   • Moderate: Medical treatment beyond first aid; temporary restricted duty (Score = 3).
   • Minor: First aid treatment only; ergonomic strain (Score = 2).
   • Negligible: Zero injury potential; fully guarded failsafe system (Score = 1).
2. Environmental and Regulatory Compliance:
   • Catastrophic: Major reportable hazardous release; EPA violation; plant operating permit revocation (Score = 5).
   • Major: Contained on-site hazardous spill requiring external regulatory notification (Score = 4).
   • Moderate: Minor permit exceedance managed internally without environmental dispersion (Score = 3).
   • Minor: Localized leak contained within secondary sump or catch basin (Score = 2).
   • Negligible: Zero environmental impact or discharge (Score = 1).
3. Production Capacity, Throughput, and Customer Fulfillment:
   • Catastrophic: Immediate total plant shutdown; bottleneck process halted; lost revenue > $100,000/hr (Score = 5).
   • Major: Major line derate (> 30%); customer delivery dates missed; no buffer inventory (Score = 4).
   • Moderate: Secondary production derate; buffer inventory permits 12–24 hours of operation (Score = 3).
   • Minor: Minor throughput speed reduction; readily absorbed by existing surge capacity (Score = 2).
   • Negligible: Zero production impact; redundant standby unit auto-starts instantaneously (Score = 1).
4. Direct Maintenance Repair and Asset Replacement Cost:
   • Catastrophic: Repair/replacement cost > $250,000; specialized OEM machining; 8+ weeks lead time (Score = 5).
   • Major: Repair cost $75,000 to $250,000; extensive craft overtime; major component replacement (Score = 4).
   • Moderate: Repair cost $20,000 to $75,000; standard overhaul kits required (Score = 3).
   • Minor: Repair cost $5,000 to $20,000; routine in-house shop rebuild (Score = 2).
   • Negligible: Repair cost < $5,000; consumable wear parts replaced during standard shift (Score = 1).

---

Asset Criticality Matrix (Consequence vs. Likelihood Criteria)

By cross-referencing maximum Consequence against Likelihood of Failure, assets are categorized into clear actionable tiers:

---

Establishing Baseline Performance Expectations from Operating Context & Duty Cycles

Once an asset's criticality and operational context are codified, reliability leaders establish quantitative baseline expectations across several key dimensions:

• Operating Duty Cycles: Assets running continuously 24/7/365 experience steady-state thermal and mechanical equilibria. Conversely, assets running in cyclic, batch, or intermittent duty experience severe thermal expansion/contraction, transient electrical starting currents (5x to 7x full-load amperage), mechanical shock loads, and cyclic fatigue. Baseline MTBF expectations must be adjusted downward for severe cyclic duty, and lubrication intervals must be calibrated accordingly.
• Process Variations and Buffers: The existence of surge tanks, bypass piping, and intermediate inventory buffers decouples equipment failure from immediate production loss. A pump feeding a surge tank with 12 hours of buffer capacity requires different availability expectations than a pump feeding directly into a continuous exothermic reaction.
• Environmental Envelope: Operating in sub-zero Arctic conditions, dusty cement clinker plants, or high-humidity coastal environments requires specific sealing, synthetic lubricants, enclosure ratings (NEMA 4X / IP66), and modified inspection frequencies.
• Setting testable operational expectations: A site may use an asset profile or equivalent controlled record to state required function, duty, demand, performance limits, reliability and maintainability measures, and safety or environmental requirements. Use MTBF, MTTR, OEE, or SIL only where the definition and application are appropriate.