π Lesson 11
D5
Modeling Real Reactors with Segregation and Maximum Mixedness
Segregation and maximum mixedness are two ways to model how fluids or particles mix inside real industrial reactors when the flow isnβt perfectly uniform β like imagining how sand and water swirl differently in a stirred tank versus a narrow pipe.
π― Learning Objectives
- β Calculate the exit concentration of a first-order reaction using both segregation and maximum mixedness models given an experimental RTD
- β Analyze and compare conversion predictions from segregation vs. maximum mixedness models for a given E(t) curve
- β Explain why segregation always predicts higher conversion than maximum mixedness for autocatalytic or higher-order reactions
- β Apply the intensity of segregation model to interpolate between the two extremes using mixing intensity parameters
π Why This Matters
In real mining blast holes, leaching columns, or heap bio-oxidation reactors, fluid flow is never perfectly plug or perfectly mixed β itβs messy: channelling, dead zones, and preferential paths distort reaction efficiency. Understanding segregation and maximum mixedness lets engineers bracket performance expectations, avoid overdesigning costly reactors, and diagnose poor leaching or incomplete oxidation before it impacts metal recovery or environmental compliance.
π Core Principles
All non-ideal reactors can be characterized by their residence time distribution (RTD), measured experimentally via tracer tests. Segregation theory treats the reactor as a bundle of independent fluid 'elements' β each with its own residence time β reacting independently as if in a batch reactor. Maximum mixedness assumes perfect micromixing at every location downstream, meaning concentration gradients vanish instantly; this yields the lowest possible conversion for positive-order reactions. The true behavior lies between these bounds β and the gap reveals how much mixing quality matters. For autocatalytic or complex kinetics (e.g., acid-consuming mineral dissolution), the direction of the bound reversal must also be recognized.
π Conversion Prediction for First-Order Reaction
For a first-order irreversible reaction (A β B, rate = kΒ·C_A), the exit concentration can be predicted exactly using either model from the RTD function E(t). The segregation model integrates batch reactor solutions weighted by E(t); the maximum mixedness model solves a differential equation backward in time using the integrated RTD function I(t).
Segregation Model Conversion (First-Order)
X_{seg} = 1 β β«β^β e^{βkt} E(t) dtPredicts conversion assuming fluid elements react independently according to their residence time.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| X_{seg} | Conversion (segregation) | dimensionless | Fraction of limiting reactant consumed |
| k | Reaction rate constant | minβ»ΒΉ | First-order rate constant |
| E(t) | RTD function | timeβ»ΒΉ | Probability density of residence times |
Typical Ranges:
Heap leaching (Cu oxide): k = 0.02 β 0.2 minβ»ΒΉ
Bio-oxidation columns (refractory gold): k = 0.001 β 0.01 minβ»ΒΉ
π‘ Worked Example
Problem: Given experimental RTD data for a copper heap leach column: E(t) = 0.25Β·e^(β0.25t) [minβ»ΒΉ], k = 0.15 minβ»ΒΉ. Calculate X_seg and X_maxmixed.
1.
Step 1: Recognize E(t) is exponential β corresponds to a CSTR (Ο = 4 min). For first-order, segregation gives X_seg = kΟ/(1 + kΟ) = (0.15Γ4)/(1+0.6) = 0.6/1.6 = 0.375.
2.
Step 2: For maximum mixedness, solve dC/dt = βkΒ·C β (1/Ο)Β·(C β Cβ) with boundary condition C(tββ)=Cβ; analytical solution yields same result for exponential E(t) β X_maxmixed = 0.375.
3.
Step 3: Confirm equivalence: For single-parameter exponential RTDs (CSTR-like), both models converge β but for bimodal or skewed E(t), divergence appears immediately.
Answer:
The result is X = 0.375 (37.5% conversion), which matches the CSTR prediction and confirms consistency for this idealized case.
ποΈ Real-World Application
At the Escondida copper heap leach operation (Chile), tracer studies revealed a bimodal RTD β fast channels (Οβ β 2 days) and slow diffusion zones (Οβ β 45 days). Using segregation modeling, predicted Cu recovery was 78%; maximum mixedness predicted only 61%. Field measurements showed 72% recovery β validating that real behavior lies between bounds and prompting installation of liner modifications to reduce channelling, improving uniformity and pushing performance closer to the segregation bound.