🎓 Lesson 16
D5
Thermodynamic Modeling of CO₂ Absorption with Blended Amines
It’s like using special chemical 'sponges' (blended amines) dissolved in water to capture carbon dioxide from industrial gases, and thermodynamics helps us predict how well and how much CO₂ they’ll soak up.
🎯 Learning Objectives
- ✓ Calculate CO₂ loading capacity (mol CO₂/mol amine) for a given blended amine system at specified temperature, pressure, and partial pressure using equilibrium thermodynamic models
- ✓ Analyze and compare the impact of amine blend composition (e.g., 30 wt% MEA + 10 wt% PZ) on absorption enthalpy and solvent regeneration energy using thermodynamic property tables
- ✓ Apply the eNRTL equation to compute activity coefficients for CO₂–amine–H₂O electrolyte mixtures and explain deviations from ideality
- ✓ Design a single-stage absorber column by estimating minimum solvent flow rate based on thermodynamic equilibrium limits and pinch point analysis
📖 Why This Matters
In underground and open-pit mining, diesel-powered equipment and ore processing emit CO₂-rich exhaust streams — especially in enclosed ventilation circuits or low-grade ore processing plants. Capturing this CO₂ before release is increasingly mandated (e.g., EU ETS, Canada’s Carbon Tax). Blended amines offer higher capacity, faster kinetics, and lower regeneration energy than single amines — but only if modeled correctly. Without accurate thermodynamic modeling, simulations overpredict capacity or underpredict reboiler duty, leading to undersized absorbers or oversized steam systems — costing millions in capital and OPEX. This lesson bridges theory to plant design decisions.
📘 Core Principles
CO₂ absorption in aqueous amines proceeds via reversible reactions: physical dissolution, carbamate formation (with primary/secondary amines), and bicarbonate equilibrium (dominant with tertiary amines or at high CO₂ loading). Blends leverage synergies — e.g., piperazine (PZ) accelerates kinetics while MDEA provides high capacity and low heat of regeneration. Thermodynamically, the system is a multicomponent, multiphase, reactive electrolyte: CO₂, H₂O, cations (e.g., H⁺, R₃NH⁺), anions (HCO₃⁻, R₂NCOO⁻, OH⁻), and neutral species (RNH₂, RNHCOOH). Activity coefficients deviate strongly from unity due to long-range electrostatic forces (Debye–Hückel) and short-range molecular interactions (hydrogen bonding, solvation). Models like eNRTL combine local composition theory with electrolyte extensions to regress binary interaction parameters from high-quality lab data — enabling predictive simulation across temperatures (25–80°C), pressures (0.1–1 bar), and loadings (0.1–1.2 mol CO₂/mol amine).
📐 eNRTL-Based CO₂ Loading Prediction
The equilibrium CO₂ loading α (mol CO₂ / mol total amine) is solved iteratively by combining material balances, equilibrium constants (K_carb, K_bicarb), and activity coefficient corrections (γ_i) from the eNRTL model. For a binary blend (e.g., MEA + PZ), the effective equilibrium constant accounts for amine speciation and ionic strength effects.
💡 Worked Example
Problem: Given: 2.5 mol/kg aqueous solution containing 1.5 mol/kg MEA and 1.0 mol/kg piperazine (PZ), T = 40°C, P_CO₂ = 15 kPa. Literature K_carb,MEA = 275, K_bicarb = 10^−6.2 at 40°C; eNRTL γ_HCO₃⁻ = 0.42, γ_R₂NCOO⁻ = 0.38 (regressed parameters). Estimate α assuming >95% CO₂ reacts as carbamate with PZ (fast kinetics) and MEA contributes bicarbonate.
1.
Step 1: Determine dominant reactive species — PZ (secondary amine) forms stable carbamate; MEA (primary) forms both carbamate and bicarbonate, but at low P_CO₂ and high [PZ], PZ dominates loading.
2.
Step 2: Use simplified equilibrium: K_eff ≈ K_carb,PZ × (γ_PZ / γ_PZcarb) × (γ_H⁺ / γ_HCO₃⁻) — substitute regressed γ values and known K_carb,PZ = 410 (at 40°C), yielding K_eff ≈ 410 × (1.0 / 0.38) × (1.0 / 0.42) ≈ 2560.
3.
Step 3: Solve α from K_eff = α / [(1 − α) × P_CO₂] → α = K_eff × P_CO₂ / (1 + K_eff × P_CO₂); with P_CO₂ = 0.15 bar → α ≈ (2560 × 0.15) / (1 + 2560 × 0.15) ≈ 0.998 mol CO₂/mol PZ.
4.
Step 4: Adjust for total amine basis: total amine = 2.5 mol/kg → α_total = (0.998 × 1.0 + 0.35 × 1.5) / 2.5 ≈ 0.62 mol CO₂/mol total amine (accounting for MEA’s lower contribution).
Answer:
The predicted CO₂ loading is 0.62 mol CO₂/mol total amine, which falls within the validated range of 0.55–0.68 for 1.0 M PZ + 1.5 M MEA at 40°C and 15 kPa CO₂.
🏗️ Real-World Application
At Newmont’s Boddington Gold Mine (Western Australia), a pilot-scale CO₂ capture unit treats 5,000 Nm³/h of diesel generator exhaust (8–12% CO₂) using a 30 wt% MEA + 5 wt% piperazine blend. Process simulation (Aspen Plus v12 with eNRTL-RK electrolyte property method) predicted 89% capture at 45°C absorber inlet and 115°C stripper reboiler duty of 3.8 GJ/tonne CO₂ — matching field measurements within ±4% after tuning 12 binary eNRTL parameters against lab VLE data from the University of Melbourne’s CO₂ Capture Lab (2021–2023). The model correctly flagged excessive corrosion risk above 0.85 mol CO₂/mol amine — prompting operational cap on loading, avoiding amine degradation observed in prior MEA-only trials.
🔧 Interactive Calculator
🔧 Open Thermodynamics & Process Simulation Calculator📋 Case Connection
📋 CO₂ Capture Solvent Screening for Natural Gas Sweetening
Solvent degradation and excessive reboiler duty with conventional MDEA
📋 Supercritical Fluid Extraction (SFE) Process Design for Caffeine Recovery
Low selectivity and high CO₂ consumption due to poor phase behavior prediction