22.3 Life-Cycle Costing, Maintainability, and System Availability

22.3 Life-Cycle Costing, Maintainability, and System Availability

Engineers frequently make the mistake of evaluating competing system designs solely on initial acquisition (capital) cost. In practice, initial procurement represents only the tip of the economic iceberg. Over decades of operation, recurring maintenance labor, spare parts logistics, downtime losses, energy consumption, and final environmental disposal dominate total expenditures. Maximizing enterprise profitability requires optimizing Life-Cycle Cost (LCC), engineering high Maintainability, and achieving targeted System Availability. This section covers economic lifecycle curves, maintainability mathematics, maintenance paradigms, and the rigorous formulation of Inherent Availability ($A_i$) versus Operational Availability ($A_o$).

---

1. Life-Cycle Cost (LCC) Analysis & The Cost Commitment Paradox

Life-Cycle Cost (LCC) is the total cumulative expenditure incurred to conceive, design, develop, test, manufacture, operate, maintain, and decommission a system over its entire operational lifetime.

The Four Primary Life-Cycle Cost Categories

1. Research & Development (R&D) / Design: Feasibility studies, systems architecture, detailed CAD design, prototyping, environmental qualification, and software development.
2. Production & Acquisition (CapEx): Tooling fabrication, facilities preparation, raw materials, manufacturing labor, assembly, quality control testing, and initial spare parts procurement.
3. Operations & Support (O&S / OpEx): Consumable energy/fuel, operator salaries, scheduled preventive maintenance (PM), unscheduled corrective maintenance (CM), supply chain inventory holding, depot overhauls, training, and technical documentation updates.
   • Critical FE Exam Rule: In commercial production machinery, industrial facilities, and aerospace/defense systems, O&S costs typically represent 60% to 80% of total Life-Cycle Cost!
4. Retirement & Disposal: Decommissioning, disassembly, hazardous material remediation, site decontamination, material recycling, and regulatory disposal.

The Life-Cycle Cost Iceberg

                  ▲   Acquisition / Purchase Price (15% - 25% of Total LCC)
  Waterline ══════┼════════════════════════════════════════════════════════
                  │   Operations & Support (O&S) Costs (60% - 80% of LCC)
                  │   • Operating Personnel & Consumable Power
                  │   • Scheduled Preventive Maintenance (PM)
                  ▼   • Unscheduled Corrective Repairs & Spare Parts Inventory
                      • Depot Overhauls & Logistics Delay Losses
                      Retirement & Environmental Disposal (5% - 10% of LCC)

Cost Commitment Curve vs. Cost Incurrence Curve

A central principle of systems engineering economics is the cost commitment paradox:

• Cost Commitment Curve: Rises extremely steeply during the initial phases of the lifecycle. By the conclusion of the Conceptual Design phase, roughly $70%$ of total life-cycle cost is committed (locked in) by high-level architectural choices (e.g., selecting hydraulic vs. electric drives, choosing stainless steel vs. composites, or adopting modular vs. integral architectures). By the Preliminary Design Review (PDR), approximately $80%$ to $85%$ of total LCC is permanently committed.
• Cost Incurrence Curve: Represents actual cash expenditures paid out over time. During Conceptual and Preliminary design, actual expenditures are negligible—typically less than $5%$ to $10%$ of the total project budget. The vast majority of cash is spent years later during high-rate production and decades of operations.

Cost Commitment vs. Cost Incurrence Across System Life

100% ┌─────────────────────────────────────────────────────────────┐
     │                                    Cost Commitment Curve    │
 80% │             /───────────────────────────────────────────────┤ (80-85% committed at PDR)
     │            /                                                │
 60% │           /                                                 │
     │          /                                                  │
 40% │         /                              Cost Incurrence      │
     │        /                              /─────────────────────┤
 20% │       /                              /                      │
     │      /                  /───────────/                       │
  0% └─────┴──────────────────┴───────────┴────────────────────────┘
       Concept Definition    Detailed    Production   Operations &
             & PDR            Design                   Retirement

The Cost-to-Change Multiplier (Rule of Ten)

Because early decisions lock in future costs, correcting an error becomes exponentially more expensive as development proceeds:

$\text{Requirements Phase (\$1)} \to \text{Design (\$10)} \to \text{Production (\$100)} \to \text{Field Operations (\$1,000 - \$10,000)}$

A requirement modification that costs $$1,000$ to resolve on paper during concept definition costs upwards of $$1,000,000$ if resolved via a recall and structural rebuild after deployment.

---

2. Maintainability Engineering & Maintenance Philosophies

Maintainability is an intrinsic design characteristic that defines the ease, speed, accuracy, and safety with which a system can be restored to operational status following failure, or retained in operational status through preventive servicing.

Mathematically, maintainability is modeled as the probability that a failed system will be repaired within a specified downtime $t$:

$M(t) = P(\text{Repair Time} \le t) = 1 - e^{-\mu t} = 1 - e^{-t / \text{MTTR}}$

Where:

• $\text{MTTR}$ is the Mean Time to Repair.
• $\mu$ is the Repair Rate, defined as $\mu = \frac{1}{\text{MTTR}}$.

Mean Time to Repair (MTTR) Formulation

For a system composed of $n$ independent components, each possessing a failure rate $\lambda_i$ and a component mean corrective repair time $M_i$:

$\text{MTTR} = \bar{M}_{\text{ct}} = \frac{\sum_{i=1}^n \lambda_i M_i}{\sum_{i=1}^n \lambda_i} = \frac{\sum_{i=1}^n \lambda_i M_i}{\lambda_{\text{sys}}}$

NCEES Formula Trap: Never calculate MTTR as the simple arithmetic average of component repair times! Component repair times must be weighted by their failure rates ($\lambda_i$). Components that fail frequently exert a much larger influence on system MTTR than components that rarely fail.

Maintenance Paradigms

Industrial facilities implement three primary maintenance strategies:

1. Corrective Maintenance (CM): Reactive, run-to-failure maintenance performed strictly after a functional failure occurs. Objectives: diagnose root cause, isolate failed component, replace or repair, and test. Disadvantages: Unplanned production halts, secondary collateral equipment damage, and high overtime costs.
2. Preventive Maintenance (PM): Proactive, scheduled maintenance executed at predetermined calendar intervals or run-time thresholds (e.g., changing oil and replacing seals every $1,000$ operating hours). Objective: replace aging components prior to the onset of the wear-out phase of the bathtub curve. Disadvantage: Incurs labor and replacement part costs on components that may still possess substantial useful life.
3. Condition-Based Maintenance (CBM) / Predictive Maintenance (PdM): Proactive maintenance triggered by real-time sensor monitoring of physical health indicators (e.g., vibration spectral analysis, infrared thermography, ultrasonic leak detection, lubricant spectrochemical debris analysis). Maintenance is scheduled only when physical degradation crosses a quantitative statistical threshold. Advantages: Eliminates unexpected failures while maximizing component service life.

---

3. System Availability Metrics ($A_i$ vs. $A_a$ vs. $A_o$)

Availability ($A$) is the probability that a repairable system is operating satisfactorily when called upon at an arbitrary calendar point in time. While reliability measures the duration of continuous failure-free operation ($P(\text{no failure up to time } t)$), availability accounts for both reliability (uptime) and maintainability/supportability (downtime). The FE exam tests three distinct levels of availability:

1. Inherent Availability ($A_i$)

Inherent availability measures system readiness from a pure design perspective under ideal maintenance conditions. It considers only corrective maintenance downtime and assumes:

• Tools, technicians, and spare parts are immediately on hand (zero delay).
• Scheduled preventive maintenance is excluded.
• Zero logistics or administrative delay times.

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

Where:

• $\text{MTBF}$ = Mean Time Between Failures = $\frac{1}{\lambda_{\text{sys}}}$.
• $\text{MTTR}$ = Mean Time to Repair (active corrective repair wrench time).

2. Achieved Availability ($A_a$)

Achieved availability incorporates both corrective maintenance and scheduled preventive maintenance, but still assumes ideal support conditions (zero logistics or administrative delay):

$A_a = \frac{\text{MTBM}}{\text{MTBM} + \bar{M}}$

Where:

• $\text{MTBM}$ = Mean Time Between Maintenance (accounting for both failure events and scheduled PM events): $\frac{1}{\text{MTBM}} = \frac{1}{\text{MTBF}} + \frac{1}{\text{MTBPM}} = \lambda_{\text{CM}} + f_{\text{PM}}$
• $\bar{M}$ = Mean Active Maintenance Time (the weighted average of active corrective wrench time $\text{MTTR}$ and scheduled preventive maintenance wrench time $\bar{M}_{\text{pt}}$): $\bar{M} = \frac{\lambda_{\text{CM}}(\text{MTTR}) + f_{\text{PM}}(\bar{M}_{\text{pt}})}{\lambda_{\text{CM}} + f_{\text{PM}}}$

3. Operational Availability ($A_o$)

Operational availability reflects the real-world availability experienced by end-users in actual operating environments. It accounts for all causes of downtime, including active corrective repairs, scheduled preventive maintenance, technician travel time, administrative processing, and supply chain delays:

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

Where:

• $\text{MUT}$ = Mean Up Time.
• $\text{MDT}$ = Mean Down Time. MDT is composed of three additive elements: $\text{MDT} = \bar{M} + \text{MLDT} + \text{ADT}$
  • $\bar{M}$ = Mean Active Maintenance Time (direct wrench time for CM and PM).
  • $\text{MLDT}$ = Mean Logistics Delay Time (waiting for spare parts procurement, shipping, customs clearance, and specialized test equipment delivery).
  • $\text{ADT}$ = Administrative Delay Time (work order authorization, shift changes, technician dispatch).

Availability Metric Hierarchy

Inherent Availability (A_i)          Achieved Availability (A_a)         Operational Availability (A_o)
┌────────────────────────────┐       ┌────────────────────────────┐       ┌────────────────────────────┐
│         Uptime             │       │         Uptime             │       │         Uptime             │
│         (MTBF)             │       │         (MTBM)             │       │      (MUT or MTBM)         │
├────────────────────────────┤       ├────────────────────────────┤       ├────────────────────────────┤
│       Downtime:            │       │       Downtime:            │       │       Downtime:            │
│ • Corrective Repair        │       │ • Corrective Repair        │       │ • Corrective Repair        │
│   Wrench Time (MTTR)       │       │ • Scheduled PM Wrench Time │       │ • Scheduled PM Wrench Time │
│                            │       │                            │       │ • Logistics Delay (MLDT)   │
│                            │       │                            │       │ • Admin Delay (ADT)        │
└────────────────────────────┘       └────────────────────────────┘       └────────────────────────────┘
         A_i (Highest)           >               A_a              >           A_o (Realistic / Lowest)

---

4. Step-by-Step Worked Engineering Calculations

Worked Example 22.3.1: Life-Cycle Cost (LCC) Present Worth Analysis

Problem: A manufacturing plant evaluates two competing automated CNC machine cells for an 8-year production contract. The corporate discount rate (cost of capital) is $i = 8%$ compounded annually.

• Machine Alpha (Low CapEx):
  • Initial Acquisition Cost: $$160,000$
  • Annual Operations & Support Cost: $$48,000/\text{year}$
  • Net Salvage Value at Year 8: $$20,000$
• Machine Beta (High Reliability / Quick-Change Modular Design):
  • Initial Acquisition Cost: $$260,000$
  • Annual Operations & Support Cost: $$26,000/\text{year}$
  • Net Salvage Value at Year 8: $$35,000$

1. Calculate the uniform series present worth factor $(P/A, 8%, 8)$ and single-payment present worth factor $(P/F, 8%, 8)$.
2. Compute the total Present Worth Life-Cycle Cost ($PW_{\text{LCC}}$) for Machine Alpha and Machine Beta.
3. Determine the net life-cycle savings achieved by selecting the superior machine.

Solution:

Step 1: Compute Engineering Economics Interest Factors

• Uniform Series Present Worth Factor: $(P/A, i, N) = \frac{(1 + i)^N - 1}{i(1 + i)^N} = \frac{(1.08)^8 - 1}{0.08(1.08)^8} = \frac{1.850930 - 1}{0.08(1.850930)} = \frac{0.850930}{0.148074} = 5.74664$
• Single-Payment Present Worth Factor: $(P/F, i, N) = (1 + i)^{-N} = (1.08)^{-8} = \frac{1}{1.850930} = 0.540269$

Step 2: Compute Total Present Worth Life-Cycle Cost

$PW_{\text{LCC}} = \text{CapEx} + \text{O\&S}(P/A, i, N) - \text{Salvage}(P/F, i, N)$

• For Machine Alpha: $PW_{\text{LCC, Alpha}} = 160,000 + 48,000(5.74664) - 20,000(0.540269)$ $PW_{\text{LCC, Alpha}} = 160,000 + 275,838.72 - 10,805.38 = \$425,033.34$

• For Machine Beta: $PW_{\text{LCC, Beta}} = 260,000 + 26,000(5.74664) - 35,000(0.540269)$ $PW_{\text{LCC, Beta}} = 260,000 + 149,412.64 - 18,909.42 = \$390,503.22$

Step 3: Life-Cycle Savings Comparison

$\Delta PW_{\text{LCC}} = PW_{\text{LCC, Alpha}} - PW_{\text{LCC, Beta}} = 425,033.34 - 390,503.22 = \$34,530.12$

Engineering Conclusion: Despite Machine Beta requiring a $$100,000$ ($+62.5%$) higher initial acquisition cost, its superior reliability and maintainability reduce annual O&S costs sufficiently to generate $$34,530.12 in net life-cycle savings in present value terms.

Worked Example 22.3.2: Multi-Component MTTR and Inherent Availability ($A_i$)

Problem: An automated high-speed parcel induction conveyor consists of three critical subsystems in a reliability series configuration:

1. Conveyor Drive Roller Assembly: $\lambda_1 = 0.00035\text{ failures/hr}$, Mean corrective repair time $M_1 = 4.0\text{ hours}$

2. Optical Line Scanner: $\lambda_2 = 0.00045\text{ failures/hr}$, Mean corrective repair time $M_2 = 1.2\text{ hours}$

3. Pneumatic Diverter Gate: $\lambda_3 = 0.00120\text{ failures/hr}$, Mean corrective repair time $M_3 = 0.5\text{ hours}$

4. Calculate the overall system failure rate ($\lambda_{\text{sys}}$) and Mean Time Between Failures ($\text{MTBF}$).

5. Calculate the system-level weighted Mean Time to Repair ($\text{MTTR}$) and repair rate ($\mu$).

6. Determine the system Inherent Availability ($A_i$).

Solution:

Step 1: Compute System Failure Rate and MTBF

For a series reliability system: $\lambda_{\text{sys}} = \sum_{i=1}^3 \lambda_i = 0.00035 + 0.00045 + 0.00120 = 0.00200\text{ failures/hour}$ $\text{MTBF} = \frac{1}{\lambda_{\text{sys}}} = \frac{1}{0.00200} = 500.0\text{ operating hours}$

Step 2: Compute Weighted MTTR and Repair Rate ($\mu$)

$\text{MTTR} = \frac{\sum_{i=1}^3 \lambda_i M_i}{\sum_{i=1}^3 \lambda_i}$ $\sum_{i=1}^3 \lambda_i M_i = (0.00035)(4.0) + (0.00045)(1.2) + (0.00120)(0.5)$ $\sum_{i=1}^3 \lambda_i M_i = 0.00140 + 0.00054 + 0.00060 = 0.00254\text{ hours/hr}$ $\text{MTTR} = \frac{0.00254}{0.00200} = 1.27\text{ hours}$ $\mu = \frac{1}{\text{MTTR}} = \frac{1}{1.27} = 0.7874\text{ repairs/hour}$

Step 3: Compute Inherent Availability ($A_i$)

$A_i = \frac{\text{MTBF}}{\text{MTBF} + \text{MTTR}} = \frac{500.0}{500.0 + 1.27} = \frac{500.0}{501.27} = 0.997467 \approx 99.75\%$

Worked Example 22.3.3: Inherent vs. Operational Availability with Supply Logistics Delays

Problem: Consider the parcel induction conveyor from Worked Example 22.3.2 ($\text{MTBF} = 500\text{ hours}$, active repair wrench time $\text{MTTR} = 1.27\text{ hours}$).

The facility operates under actual field conditions:

• Scheduled preventive maintenance (PM) is performed every $250\text{ hours}$ ($f_{\text{PM}} = 0.004\text{ events/hr}$) requiring $\bar{M}_{\text{pt}} = 1.5\text{ hours}$ of active wrench time.
• Due to vendor central warehousing, the Mean Logistics Delay Time (MLDT) for corrective repair spare parts delivery is $14.5\text{ hours}$.
• Administrative Delay Time (ADT) for maintenance ticketing and technician dispatch averages $1.5\text{ hours}$.
• PM activities are planned in advance, incurring zero logistics delay.

1. Calculate the total Mean Down Time for corrective maintenance ($\text{MDT}{\text{CM}}$) and preventive maintenance ($\text{MDT}{\text{PM}}$).
2. Compute the overall Mean Down Time ($\text{MDT}$) across all maintenance events.
3. Compute the Mean Time Between Maintenance ($\text{MTBM}$).
4. Calculate the Operational Availability ($A_o$) and compare it directly to Inherent Availability ($A_i = 99.75%$).

Solution:

Step 1: Compute Down Times per Maintenance Category

• Corrective Maintenance Down Time: $\text{MDT}_{\text{CM}} = \text{MTTR} + \text{MLDT} + \text{ADT} = 1.27 + 14.5 + 1.5 = 17.27\text{ hours}$
• Preventive Maintenance Down Time: $\text{MDT}_{\text{PM}} = \bar{M}_{\text{pt}} = 1.50\text{ hours}$

Step 2: Compute Overall System Mean Down Time ($\text{MDT}$)

Total maintenance event frequency: $f_{\text{total}} = \lambda_{\text{CM}} + f_{\text{PM}} = 0.00200 + 0.00400 = 0.00600\text{ maintenance events/hour}$ $\text{MDT} = \frac{\lambda_{\text{CM}}(\text{MDT}_{\text{CM}}) + f_{\text{PM}}(\text{MDT}_{\text{PM}})}{f_{\text{total}}} = \frac{0.00200(17.27) + 0.00400(1.50)}{0.00600}$ $\text{MDT} = \frac{0.03454 + 0.00600}{0.00600} = \frac{0.04054}{0.00600} = 6.7567\text{ hours}$

Step 3: Compute Mean Time Between Maintenance ($\text{MTBM}$)

$\text{MTBM} = \frac{1}{f_{\text{total}}} = \frac{1}{0.00600} = 166.67\text{ operating hours}$

Step 4: Compute Operational Availability ($A_o$)

$A_o = \frac{\text{MTBM}}{\text{MTBM} + \text{MDT}} = \frac{166.67}{166.67 + 6.7567} = \frac{166.67}{173.4267} = 0.96104 \approx 96.10\%$

• Comparison and Engineering Insight: $A_i = 99.75\% \quad \text{vs.} \quad A_o = 96.10\%$ Total annual downtime under $A_i$ ($8,760\text{ hrs/yr}$) is $8,760(1 - 0.9975) = 21.9\text{ hours/year}$. Total annual downtime under $A_o$ is $8,760(1 - 0.9610) = 341.6\text{ hours/year}$! Takeaway: In the real world, logistics delay (waiting for spare parts) causes over $90%$ of all system downtime, proving that stocking critical on-site spares has a far greater impact on operational availability than simply training technicians to turn wrenches faster.

---

5. NCEES Reference Handbook Tips & Realistic Exam Traps

• The Cost Commitment Trap: Exam questions frequently ask: "In which lifecycle phase is the largest portion of total system life-cycle cost committed?" The answer is always Conceptual Design / Preliminary Design (where $70%$ to $80%$ is locked in), NOT during manufacturing or operations, even though actual money is spent during operations.
• Availability Denominator Confusion:
  • $A_i$ uses purely active repair time: $\text{MTBF} / (\text{MTBF} + \text{MTTR})$.
  • $A_o$ uses total downtime: $\text{MUT} / (\text{MUT} + \text{MDT})$, where $\text{MDT} = \text{Active Wrench Time} + \text{Logistics Delay} + \text{Admin Delay}$.
  • Never substitute Mean Down Time (MDT) into the Inherent Availability formula!
• Unweighted MTTR Error: When given multiple components with different failure rates and repair times, never take the simple average $(M_1 + M_2 + \dots + M_n)/n$. Always compute the failure-rate-weighted average: $\text{MTTR} = (\sum \lambda_i M_i) / (\sum \lambda_i)$.