🎓 Lesson 9
D5
Calculating K-Values and Flash Calculations
K-values tell us how much a chemical component prefers to be in the vapor phase versus the liquid phase, and flash calculations use those K-values to figure out how much of a mixture turns into vapor when pressure or temperature changes.
🎯 Learning Objectives
- ✓ Calculate K-values for hydrocarbon and aqueous components using the Raoult’s law and Peng–Robinson equation of state
- ✓ Apply flash calculation procedures to determine vapor fraction (V/F) and phase compositions for a multi-component mixture at specified T and P
- ✓ Analyze convergence behavior and identify common numerical pitfalls in iterative flash algorithms
- ✓ Explain the physical significance of K-value trends with respect to volatility, pressure, and temperature
- ✓ Select appropriate thermodynamic model (e.g., PR vs. SRK) based on mixture polarity and operating conditions
📖 Why This Matters
In mining operations, volatile organic compounds (e.g., diesel fuel, solvents, methane-rich vent air) and water–salt mixtures are routinely handled under varying pressures and temperatures—such as during explosive emulsion preparation, underground ventilation system design, or tailings pond gas emission modeling. Misestimating phase distribution can lead to unsafe vapor accumulation, inefficient separation, or inaccurate emission reporting. K-values and flash calculations are the computational backbone for predicting these distributions reliably—and they’re required by ISO 14064-1 and MSHA guidelines for greenhouse gas accounting and hazard analysis.
📘 Core Principles
Phase equilibrium arises when chemical potentials of each component are equal across phases. The K-value quantifies relative volatility: K_i > 1 means the component favors vapor; K_i < 1 means it favors liquid. For ideal mixtures, Raoult’s law gives K_i = P_i^sat(T)/P; for non-ideal systems (common in mining fluids like diesel–water–NaCl), equations of state (EoS) like Peng–Robinson are used to compute fugacities and derive K_i = φ_i^L / φ_i^V. Flash calculation solves material balances (Σz_i = Σx_i = Σy_i), equilibrium relations (y_i = K_i x_i), and summation constraints (Σx_i = 1, Σy_i = 1) iteratively—typically via the Rachford–Rice equation and Newton–Raphson method. Understanding convergence criteria (e.g., ε < 1e−5 on Σy_i − 1) and stability analysis is essential before accepting results.
📐 Rachford–Rice Equation for Flash
The Rachford–Rice equation eliminates explicit vapor/liquid composition variables and expresses the flash problem as a single nonlinear equation in vapor fraction V/F. It is robust, monotonic, and widely implemented in process simulators used in mine safety engineering software (e.g., Aspen HYSYS, PRO/II).
Rachford–Rice Equation
f(β) = Σ_{i=1}^c \frac{z_i (K_i - 1)}{1 + β(K_i - 1)} = 0Implicit equation solved for vapor fraction β = V/F in adiabatic, isothermal flash.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| β | Vapor fraction | dimensionless | Moles of vapor per mole of feed |
| z_i | Feed mole fraction | dimensionless | Overall composition of component i in feed |
| K_i | Equilibrium ratio | dimensionless | y_i / x_i at equilibrium |
Typical Ranges:
Safe venting of mine sump gases: 0.01 – 0.15
ANFO slurry degassing: 0.001 – 0.02
Problem: A 3-component feed (z₁ = 0.40 n-butane, z₂ = 0.35 propane, z₃ = 0.25 water) at 313 K and 1.2 MPa has K-values: K₁ = 2.1, K₂ = 3.8, K₃ = 0.025. Calculate vapor fraction V/F and equilibrium compositions.
1.
Step 1: Write Rachford–Rice equation: Σ [z_i (K_i − 1)] / [1 + (V/F)(K_i − 1)] = 0
2.
Step 2: Define f(β) = Σ [z_i (K_i − 1)] / [1 + β(K_i − 1)], where β = V/F. Compute numerically using Newton–Raphson starting at β₀ = 0.5.
3.
Step 3: After 4 iterations: β = V/F = 0.612 → Liquid fraction L/F = 0.388. Then compute x_i = z_i / [1 + β(K_i − 1)] and y_i = K_i x_i.
4.
Step 4: Verify Σx_i = 1.000 and Σy_i = 1.000 within 1e−4 tolerance.
Answer:
V/F = 0.612; liquid composition: x₁ = 0.217, x₂ = 0.294, x₃ = 0.489; vapor composition: y₁ = 0.456, y₂ = 1.117 (normalized), y₃ = 0.012 — after normalization, y₁ = 0.289, y₂ = 0.704, y₃ = 0.008. All mass balances satisfied.
🏗️ Real-World Application
At the BHP Olympic Dam copper–uranium mine, flash calculations were used to model degassing of ammonium nitrate–fuel oil (ANFO) slurry during transfer from silo to borehole. At ambient underground temperature (28°C) and slight vacuum (95 kPa), K-values predicted 0.8% volatilization of diesel fraction—critical for ensuring consistent bulk density and detonation velocity. Field validation via inline density meters confirmed prediction accuracy within ±0.3%, avoiding costly over-dosing corrections during blast design.