🎓 Lesson 18 D5

Integrating Monitoring Data into Risk Matrices

Using real-time data from sensors and surveys to update how likely and how bad a slope failure could be.

🎯 Learning Objectives

  • Analyze displacement rate trends from inclinometer or GNSS data to assign updated likelihood scores in a 5×5 risk matrix
  • Apply threshold-based trigger logic (e.g., >2 mm/day displacement) to reclassify risk levels in accordance with industry alert protocols
  • Explain how integrating pore pressure ratios (ru) modifies consequence severity ratings for potential landslide runout zones
  • Design a simplified dynamic risk matrix update protocol aligned with AS/NZS ISO 31000:2018 and SME Best Practices

📖 Why This Matters

In modern mining, a static 'one-time' risk assessment is dangerously obsolete. Slopes evolve—rainfall saturates, blasts induce micro-fracturing, and creep accelerates without warning. In 2022, a major Australian iron ore operation avoided a catastrophic failure by detecting accelerating displacement (4.7 mm/day over 72 hours) and triggering immediate evacuation—because their risk matrix was dynamically updated using live monitoring feeds. This lesson teaches you how to turn raw sensor data into actionable risk intelligence—not just reports, but decisions.

📘 Core Principles

Risk matrices classify hazards by Likelihood (L) and Consequence (C) to produce a Risk Level (R = L × C). Static matrices use historical or expert-judgment inputs. Dynamic integration replaces fixed L/C values with data-driven tiers: e.g., displacement velocity thresholds map to Likelihood categories (Low: <0.1 mm/day; Medium: 0.1–1.0 mm/day; High: >1.0 mm/day), while real-time pore pressure ratios (ru) and proximity to infrastructure refine Consequence severity. Crucially, integration requires traceability—each data point must be validated, time-stamped, calibrated, and linked to a defined trigger protocol per ISO 31000’s 'monitor and review' principle. The goal is not automation, but *augmented judgment*: data informs, engineers decide.

📐 Likelihood Tier Assignment from Displacement Rate

Displacement rate (v) measured via GNSS or extensometers is converted to a Likelihood tier using empirically calibrated thresholds. This is not a predictive model—but a standardized, auditable classification rule used in operational alert systems (e.g., Rio Tinto’s Slope Monitoring Framework).

Displacement-Based Likelihood Tier

L_tier = f(v_{24h})

Assigns a discrete likelihood category (1–5) based on maximum 24-hour displacement rate measured at critical monitoring points.

Variables:
SymbolNameUnitDescription
v_{24h} Maximum 24-hour displacement rate mm/day Highest validated displacement increment over any consecutive 24-hour window, derived from GNSS, inclinometers, or GB-InSAR.
Typical Ranges:
Stable highwall: 0.01 – 0.05 mm/day
Accelerating creep phase: 0.5 – 5.0 mm/day
Imminent failure precursor: >5.0 mm/day

💡 Worked Example

Problem: A highwall inclinometer array records cumulative displacement of 18.6 mm over 3 days. The latest 24-hour interval shows 3.2 mm movement. Assign the appropriate Likelihood tier per SME Guidelines (2021) Table 4.2.