🎓 Lesson 15
D5
Translating Model Outputs into Closure Risk Statements for Regulators
Turning computer model results about mine waste behavior into clear, trustworthy statements that regulators can use to decide if a site is safe to close.
🎯 Learning Objectives
- ✓ Explain how model uncertainty propagates into closure risk classifications using confidence intervals and exceedance probabilities
- ✓ Apply regulatory threshold logic (e.g., 95th percentile metal concentration vs. aquatic guideline) to convert simulated time-series outputs into binary or graded risk statements
- ✓ Design a transparent risk statement template that includes model scope, key assumptions, dominant uncertainty drivers, and lines of evidence supporting closure readiness
- ✓ Analyze a geochemical model output file (e.g., PHREEQC or MIN3P result) to identify critical failure modes and assign qualitative/quantitative confidence levels per exposure pathway
📖 Why This Matters
Regulators don’t approve models—they approve closure decisions. A technically sound geochemical model means little if its outputs aren’t translated into unambiguous, auditable risk statements that answer: ‘Will this site protect water quality for 100+ years?’ In high-profile mine closures (e.g., Mount Polley, British Columbia), ambiguous or overconfident risk language has delayed approvals, triggered litigation, and eroded public trust. This lesson equips you to speak the regulator’s language—converting stochastic outputs, parameter sensitivities, and scenario envelopes into precise, defensible, and legally robust statements.
📘 Core Principles
Risk translation rests on three interlocking pillars: (1) Uncertainty framing—distinguishing epistemic (knowledge-limited) from aleatory (inherent stochastic) uncertainty and selecting appropriate representation methods (e.g., Monte Carlo vs. bounding scenarios); (2) Regulatory alignment—mapping model outputs (e.g., predicted Zn concentration in seepage) directly to compliance metrics (e.g., BC Water Quality Guideline of 5 µg/L for chronic exposure); and (3) Statement architecture—structuring conclusions using standardized logic: ‘Under Scenario X, with 90% confidence, pathway Y will not exceed threshold Z for ≥100 years, supported by lines of evidence A, B, and C.’ Students progress from deterministic interpretation (‘model says pH stays >6’) to probabilistic risk framing (‘probability of pH <4.5 at receptor exceeds 1×10⁻⁴/yr’) and finally to regulatory-grade statements that declare closure readiness—or identify required mitigation.
📐 Exceedance Probability Threshold Mapping
This formula quantifies the probability that modeled contaminant concentrations exceed regulatory thresholds over time—forming the basis for quantitative risk statements. It integrates time-series model output with regulatory limits and confidence bounds.
Exceedance Probability (EP)
EP(t) = P[C₉₅(t) > C_reg]Probability that the 95th percentile of predicted contaminant concentration at time t exceeds the regulatory concentration limit.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| C₉₅(t) | 95th percentile concentration | µg/L | Upper bound of the central 90% of concentration predictions at time t, derived from stochastic modeling. |
| C_reg | Regulatory concentration limit | µg/L | Legally enforceable water quality guideline or standard applicable to the exposure pathway. |
Typical Ranges:
Low-risk closure statement: EP ≤ 0.01
Conditional approval: 0.01 < EP ≤ 0.05
High-risk designation: EP > 0.10
💡 Worked Example
Problem: A MIN3P simulation of acid rock drainage from a waste rock pile yields 1,000 realizations of dissolved Cu concentration at the downgradient aquifer receptor over 200 years. The BC Aquatic Life Guideline for Cu is 3.8 µg/L. The 95th percentile of annual maximum Cu concentrations across all realizations is 4.2 µg/L at year 87; at year 200, it drops to 2.9 µg/L. Calculate EP at year 200 using the 90% confidence interval of the 95th percentile.
1.
Step 1: Extract the 95th percentile concentration time series: C₉₅(t=200) = 2.9 µg/L
2.
Step 2: Determine the lower bound of the 90% CI for C₉₅(t=200): given standard error = 0.3 µg/L, lower bound = 2.9 − 1.645×0.3 = 2.41 µg/L
3.
Step 3: Compare lower bound to regulatory threshold (3.8 µg/L): since 2.41 < 3.8, the probability that the true 95th percentile exceeds 3.8 is <5% — thus EP < 0.05 at t=200
Answer:
The exceedance probability is <5% at year 200, satisfying BC’s ‘low risk’ criterion (EP ≤ 0.05) for closure certification.
🏗️ Real-World Application
At the Antamina Mine (Peru), closure risk statements were developed using PHREEQC reactive transport modeling coupled with Monte Carlo parameter sampling. For the tailings storage facility, the final regulatory submission stated: ‘With ≥95% confidence, dissolved As concentrations at the perimeter groundwater monitoring well will remain below the WHO drinking water guideline (10 µg/L) for ≥1,000 years, contingent on the implementation of the engineered clay cap (Scenario B). This conclusion is robust to ±30% variation in pyrite oxidation rate and saturated hydraulic conductivity, as confirmed by global sensitivity analysis (Sobol indices).’ This statement directly linked model output, uncertainty treatment, engineering control, and regulatory benchmark—enabling rapid regulatory sign-off.
🔧 Interactive Calculator
🔧 Open Mine Waste Characterization & Geochemical Modeling Calculator📋 Case Connection
📋 Copper Mine Waste Rock Stockpile ARD Mitigation at Escondida Extension
High-pyrite waste rock (>3.2% S) stockpiled without cover; predicted ARD onset within 5 years
📋 Gold Tailings Geochemical Stabilization at Granny Smith Mine (WA)
Arsenic-rich tailings (up to 120 mg/kg As) exhibiting elevated As leaching under oxidizing conditions
📋 Iron Ore Mine Waste Rock Long-Term Stability at Brockman 4 (Pilbara)
Massive hematite-goethite waste rock (low sulfide but high Mn/Al) showing delayed acidity and Al leaching post-construct...
📋 Coal Mine Spoil Geochemical Capping at Hunter Valley Reclamation Project
Spoil with pyritic shale interbeds generating ARD despite initial alkaline overburden; inconsistent capping led to local...