π Lesson 3
D2
Resilience vs. Reliability vs. Availability: Key Distinctions
Resilience is how well a system bounces back after disruption; reliability is how long it runs without failing; availability is how often itβs ready to work when needed.
π― Learning Objectives
- β Explain the conceptual and mathematical distinctions among resilience, reliability, and availability using system-level examples
- β Calculate steady-state availability given MTBF and MTTR values
- β Analyze a mine power substationβs resilience posture by mapping threats, recovery actions, and functional dependencies
- β Apply ISO 55000 and IEEE 1366 standards to assess and compare reliability metrics for diesel vs. grid-connected mine power systems
- β Design a minimal redundancy strategy to improve resilience without compromising cost-efficiency
π Why This Matters
In remote mining operations, a single 4-hour power outage can halt ore production, cost $2M+, and trigger safety-critical ventilation failures. Yet many engineers conflate 'the generator started' (availability), 'it ran 99.8% of last year' (reliability), and 'it restored full ventilation within 8 minutes after a lightning strike' (resilience). Confusing these leads to under-investing in adaptive capacity β e.g., specifying ultra-reliable but non-reconfigurable switchgear β leaving mines vulnerable to cascading failures. Mastering their distinctions enables precise risk allocation, smarter CAPEX decisions, and compliance with evolving regulatory expectations (e.g., ICMMβs Climate Resilience Framework).
π Core Principles
Resilience is inherently *dynamic* and *contextual*: it depends on threat type (natural, technical, human), system architecture (centralized vs. microgrid), and organizational response capability (training, spare parts, automation). Reliability is *probabilistic and time-based*, rooted in failure rate modeling (e.g., exponential or Weibull distributions) and governed by design maturity, component quality, and operating stress. Availability is *operational and observable*, expressed as a ratio β it conflates failure frequency (MTBF) and recovery speed (MTTR). Crucially, high reliability does not guarantee high resilience: a highly reliable SCADA system may lack failover logic or manual override capability, making it brittle under novel cyber-physical attacks. Likewise, high availability masks latent vulnerabilities β a diesel plant with 99.2% availability may still collapse during simultaneous fuel delivery delay + cooling fan failure if no contingency exists.
π Steady-State Availability Calculation
Steady-state availability (A) is the foundational metric linking reliability and maintainability. It assumes constant failure and repair rates and is used for long-term planning of mission-critical infrastructure like mine substations or compressed air networks.
Steady-State Availability (A)
A = MTBF / (MTBF + MTTR)Proportion of time a repairable system is operational under steady-state conditions.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| A | Availability | dimensionless (decimal or %) | Fraction of time system is operational |
| MTBF | Mean Time Between Failures | hours | Average operational time between failures for repairable systems |
| MTTR | Mean Time To Repair | hours | Average time required to restore system to operation after failure |
Typical Ranges:
Critical mine power substation: 99.5% β 99.9%
Mine compressed air system: 98.0% β 99.2%
Remote site solar-diesel hybrid: 97.0% β 99.0%
π‘ Worked Example
Problem: A mineβs primary 33 kV substation has a mean time between failures (MTBF) of 2,400 hours and a mean time to repair (MTTR) of 8 hours. Calculate its steady-state availability and interpret the result against industry benchmarks.
1.
Step 1: Recall the formula A = MTBF / (MTBF + MTTR)
2.
Step 2: Substitute values: A = 2400 / (2400 + 8) = 2400 / 2408
3.
Step 3: Compute: A β 0.9967 β 99.67%
4.
Step 4: Compare to typical range: 99.5β99.9% for critical mine electrical infrastructure per IEEE 1366 Annex D
Answer:
The result is 99.67%, which falls within the safe and target range for Tier-1 mine power infrastructure (99.5β99.9%). However, this metric alone does not reflect resilience β e.g., if MTTR assumes ideal conditions (spare parts on-site, trained crew available), real-world disruptions may extend repair to >24 hrs, collapsing effective availability during extreme events.
ποΈ Real-World Application
At Rio Tintoβs Yandi Mine (Pilbara, WA), a 2022 Category 4 cyclone severed the sole 220 kV transmission line and flooded the main substation. Though the siteβs diesel generators had 99.8% annual availability (per maintenance logs) and high reliability (MTBF > 3,000 hrs), resilience was low due to single-point dependency and lack of automated islanding. Recovery took 36 hours. Post-event, Rio implemented a hybrid microgrid with solar-battery-diesel coordination, real-time load shedding logic, and pre-positioned mobile substations β increasing *resilience* (recovery time reduced to <15 mins for critical loads) without changing generator reliability or availability specs. This shift aligned with ICMMβs 2023 Resilience Assessment Protocol and reduced annualized outage risk by 73% (per internal RAMS study).