ChemProCal
Knowledge Base + New Calculation Login Sign Up
ChemProCal • Separation & Distillation • Membrane Acid Gas Separation Unit Sizer
SEPARATION & DISTILLATION

Membrane Acid Gas Separation Unit Sizer

Size polymeric membrane acid gas separation units (hollow-fiber polyimide / cellulose acetate) for bulk CO2 and H2S removal from natural gas with single-stage and two-stage permeate recycle architectures.



Feed Gas Stream Conditions

MMSCFD
barg
°C
barg

Contaminant Loadings & Sales Gas Target

mol %
mol %
ppmv

Polymer Membrane Specifications

dimensionless
GPU
m² / cartridge
% (for 2-stage)

📐 Engineering Formula & Specification Sheet

Membrane Acid Gas Separation Unit Sizer: Sizing Methodology & Engineering Basis

Size hollow-fiber polyimide and cellulose acetate membrane systems for natural gas CO2 removal, calculating required surface area, stage cut, hydrocarbon loss, and recycle compression.

👨‍🔬 Reviewed by: Process Design Engineer (B.Tech Chemical Engineering, 5+ Yrs Exp)
🗓️ Standards: ASME / API / GPSA / ISA
⏱️ Format: Quick Reference Spec Sheet

1. Fickian Solution-Diffusion Permeation Model

Gas permeation across a dense non-porous polymeric membrane follows the classic solution-diffusion mechanism, wherein gas molecules dissolve into the upstream polymer face, diffuse across the active selective skin, and desorb into the low-pressure permeate channel. The molar flux \(J_i\) of component \(i\) is governed by Fick's first law:

\[ J_i = \frac{P_i}{l} \left( p_{f,i} - p_{p,i} \right) = Q_i \left( x_{f,i} P_{\text{feed}} - y_{p,i} P_{\text{perm}} \right) \]

where \(P_i\) is intrinsic membrane permeability, \(l\) is the effective selective layer thickness (\(0.1\text{--}1.0\ \mu\text{m}\)), \(Q_i = P_i/l\) is permeance expressed in Gas Permeation Units (\(1\ \text{GPU} = 10^{-6}\ \text{cm}^3(\text{STP}) / (\text{cm}^2 \cdot \text{s} \cdot \text{cmHg}) = 3.348 \times 10^{-10}\ \text{mol}/(\text{m}^2 \cdot \text{s} \cdot \text{Pa})\)), and \(p_{f,i}, p_{p,i}\) are feed and permeate partial pressures.

2. Membrane Ideal Selectivity & Separation Factor

The inherent selectivity \(\alpha_{i/j}\) between fast-permeating acid gas (\(\text{CO}_2\)) and slow-permeating methane (\(\text{CH}_4\)) is the ratio of their permeabilities:

\[ \alpha_{\text{CO}_2/\text{CH}_4} = \frac{P_{\text{CO}_2}}{P_{\text{CH}_4}} = \frac{Q_{\text{CO}_2}}{Q_{\text{CH}_4}} \]

Commercial polyimide hollow fibers exhibit ideal selectivities \(\alpha \approx 25\text{--}35\), whereas cellulose acetate modules achieve \(\alpha \approx 20\text{--}28\) under typical natural gas pipeline conditions.

3. Log Mean Partial Pressure Driving Force (LMPD)

Because the feed gas becomes progressively depleted in \(\text{CO}_2\) as it travels from module inlet to retentate outlet while permeate partial pressure builds along the bore side, the effective driving force is calculated via the logarithmic mean partial pressure difference:

\[ \Delta p_{\text{LMPD}} = \frac{\Delta p_{\text{in}} - \Delta p_{\text{out}}}{\ln(\Delta p_{\text{in}} / \Delta p_{\text{out}})} \]

where \(\Delta p_{\text{in}} = (z_{\text{CO}_2} P_{\text{feed}} - y_{\text{CO}_2,\text{perm}} P_{\text{perm}})\) and \(\Delta p_{\text{out}} = (x_{\text{CO}_2,\text{ret}} P_{\text{feed}} - y_{\text{CO}_2,\text{perm,local}} P_{\text{perm}})\).

4. Active Membrane Area Sizing & Commercial Module Packaging

The total net membrane surface area \(A_{\text{mem}}\) required to achieve the target sales gas \(\text{CO}_2\) specification is derived from the molar permeation balance:

\[ A_{\text{mem}} = \frac{\dot{n}_{\text{CO}_2,\text{perm}}}{Q_{\text{CO}_2} \cdot \Delta p_{\text{LMPD}}} \quad [\text{m}^2] \]

Commercial hollow-fiber skid assemblies package between \(250\ \text{m}^2\) and \(500\ \text{m}^2\) of active membrane area into each cylindrical module vessel (\(8\text{--}12\ \text{inches}\) in diameter, \(3\text{--}4\ \text{meters}\) long). The total module count is:

\[ N_{\text{modules}} = \left\lceil \frac{A_{\text{mem}}}{A_{\text{module}}} \right\rceil \]

5. Two-Stage Recycle Recompression Power Balance

In high-recovery two-stage configurations, permeate from Stage 1 is recompressed by an electric-motor-driven reciprocating or centrifugal compressor and fed to Stage 2, whose retentate recycles to the feed inlet. The compression work required is:

\[ W_{\text{comp}} = \frac{\dot{m}_{\text{recycle}}}{\eta_c} \left( \frac{\gamma}{\gamma - 1} \right) Z_{\text{avg}} R_{\text{gas}} T_{\text{suct}} \left[ \left( \frac{P_{\text{feed}}}{P_{\text{perm}}} \right)^{\frac{\gamma - 1}{\gamma}} - 1 \right] \quad [\text{kW}] \]

📚 ChemProCal Knowledge Base

Need Complete Mathematical Derivations & Industrial Case Studies?

Explore our in-depth engineering textbook guide: Acid Gas Removal (AGRU) Technology Selection: Chemical Amines, Physical Solvents, Membranes & Cryogenics Comparison Guide. Features step-by-step manual hand calculations, thermodynamic proofs, and commercial simulator benchmarks.

Read Full Theory Guide →
Peer Review & Engineering Accuracy

This specification sheet and its mathematical formulations adhere to internationally recognized process codes (ASME, API, GPSA, TEMA, and ISA). Sizing parameters and default design allowances are cross-referenced with empirical industry design practices.

About This Tool

The Membrane Acid Gas Separation Unit Sizer delivers comprehensive engineering simulation of polymeric hollow-fiber and spiral-wound membrane systems for bulk \(\text{CO}_2\) and \(\text{H}_2\text{S}\) separation from raw natural gas. Designed for offshore platforms, FPSOs, and remote onshore wellheads where conventional chemical solvent absorption towers are weight- and footprint-prohibitive, this tool evaluates Fickian solution-diffusion transport, log-mean partial pressure driving force (LMPD), stage cut (\(\theta\)), hydrocarbon slip loss, and two-stage permeate recompression power.

Unlike aqueous amine scrubbing systems that require large circulation pumps, high-energy thermal reboilers, and liquid disposal systems, membrane separation is driven entirely by partial pressure differential across a dense selective polymeric skin layer. By modeling both single-stage cross-flow modules and two-stage permeate recycle configurations, process engineers can pinpoint the optimal balance between capital expenditure, active membrane area (\(\text{m}^2\)), module count, and methane recovery efficiency (up to \(98.5\%\)).

1. Governing Principles of Polymeric Gas Separation & Solution-Diffusion Mechanics

Polymeric gas separation membranes operate on the fundamental principle of differential molecular solubility and diffusivity through a dense, non-porous polymer matrix. Unlike microfiltration or ultrafiltration which separate particles based on physical pore sieving, gas separation relies on the solution-diffusion mechanism.

When high-pressure raw natural gas contacts the upstream active surface of the hollow fiber, each gas component dissolves into the polymer matrix according to its solubility coefficient \(S_i\) (governed largely by condensability and critical temperature) and diffuses across the polymer chains driven by its chemical potential gradient, characterized by diffusivity coefficient \(D_i\) (governed by molecular kinetic diameter). The overall permeability coefficient is the product of solubility and diffusivity:

\[ P_i = D_i \cdot S_i \]

Carbon dioxide (\(\text{CO}_2\)) has a kinetic diameter of \(3.30\ \text{Å}\) and high condensability (\(T_c = 31.0^\circ\text{C}\)), making it permeate roughly \(25\) to \(35\) times faster than methane (\(\text{CH}_4\)), which has a larger kinetic diameter of \(3.80\ \text{Å}\) and lower condensability (\(T_c = -82.6^\circ\text{C}\)). Hydrogen sulfide (\(\text{H}_2\text{S}\)), with a kinetic diameter of \(3.60\ \text{Å}\) and very high dipole moment, also permeates rapidly (\(\alpha \approx 20\text{--}30\)), enabling simultaneous bulk \(\text{CO}_2\) and \(\text{H}_2\text{S}\) acid gas removal in a single skid.

2. Single-Stage vs. Two-Stage Recycle Process Architectures & Hydrocarbon Recovery

The choice between single-stage and multi-stage membrane system architectures represents a classic trade-off between capital simplicity and hydrocarbon recovery efficiency:

  • Single-Stage Flowsheet: Raw gas is filtered, heated above its hydrocarbon and water dew points, and passed directly across the membrane tubesheet. Retentate exits at high pressure (\(50\text{--}70\ \text{barg}\)) with \(< 2\ \text{mol}\%\ \text{CO}_2\) as sales gas. Permeate exits at low pressure (\(1.5\text{--}3\ \text{barg}\)) enriched in \(\text{CO}_2\) (\(70\text{--}85\ \text{mol}\%\)). While mechanically simple and requiring zero rotating machinery, single-stage units inevitably slip \(8\%\) to \(15\%\) of the feed methane into the low-pressure permeate, where it must either be flared, used as low-BTU fuel gas, or sequestered.
  • Two-Stage with Permeate Recycle: To eliminate unacceptable methane slippage on large-volume gas fields (\(> 25\ \text{MMSCFD}\)), permeate from the first stage is routed to a recycle compressor and fed into a second membrane stage. The Stage 2 retentate, which is rich in recovered methane, recycles back to mix with the raw feed gas, while Stage 2 permeate vents highly concentrated acid gas (\(90\text{--}95\ \text{mol}\%\ \text{CO}_2\)). This cuts total hydrocarbon losses from \(> 10\%\) down to \(< 2\%\), generating millions of dollars in incremental annual gas revenue at the expense of adding a recycle compressor.

3. Membrane Material Selection: Polyimide, Cellulose Acetate, and Plasticization Resistance

The two dominant polymeric materials utilized in industrial natural gas acid gas separation are asymmetric cellulose acetate (CA) and aromatic polyimide (PI) / thermally rearranged (TR) polymers:

  • Cellulose Acetate (CA): Historically the industry workhorse, CA membranes offer robust resistance to fouling from trace heavy hydrocarbons and moderate plasticization resistance up to \(\sim 10\ \text{bar}\) \(\text{CO}_2\) partial pressure. However, CA is sensitive to liquid water hydrolysis and has lower inherent \(\text{CO}_2/\text{CH}_4\) selectivity (\(20\text{--}25\)), necessitating larger membrane surface areas.
  • Polyimide & Cross-linked Polyamides: Modern high-performance hollow fibers utilize rigid polyimide backbones with covalent cross-linking (e.g., monoesterification or decarboxylation). These polymers exhibit superior \(\text{CO}_2/\text{CH}_4\) selectivities (\(30\text{--}40\)) and high \(\text{CO}_2\) permeance (\(60\text{--}120\ \text{GPU}\)), significantly reducing skid footprint and weight.

A primary operational consideration is \(\text{CO}_2\) plasticization, wherein high concentrations of dissolved \(\text{CO}_2\) swell the polymer matrix, increasing free volume and drastically deteriorating selectivity. Cross-linked polyimide hollow fibers suppress plasticization up to \(\text{CO}_2\) partial pressures exceeding \(25\ \text{bar}\), making them the premier choice for offshore sour gas projects.

Frequently Asked Questions

How do polymeric membranes separate CO2 and H2S from natural gas without chemical solvents?
Polymeric membranes separate acid gases via the solution-diffusion mechanism across a microscopic, non-porous polymeric skin layer. Gas molecules dissolve into the polymer matrix and diffuse across it under the chemical potential gradient established by the partial pressure difference between the high-pressure feed side (50 to 70 barg) and the low-pressure permeate side (1.5 to 3 barg). Because CO2 and H2S possess smaller kinetic diameters and higher quadrupole/dipole interactions than methane, they dissolve and diffuse 25 to 35 times faster, permeating selectively while pipeline-spec methane remains in the high-pressure retentate stream.
Why is plasticization a critical operational risk for high-pressure CO2 membranes?
CO2 plasticization occurs when high partial pressures of CO2 (typically > 10 bar) dissolve extensively into the polymer matrix. The dissolved CO2 acts as a molecular lubricant, swelling the polymer chains and increasing the interstitial free volume. This uncoiling severely compromises the polymer sieving ability, causing a catastrophic drop in CO2/CH4 selectivity from ~30 down to < 12. As selectivity collapses, excessive volumes of valuable pipeline methane leak across into the permeate stream, causing runaway hydrocarbon losses. Cross-linked polyimide fibers mitigate this effect.
When is a two-stage membrane system with permeate recycle economically justified over a single-stage system?
A two-stage permeate recycle architecture is economically justified when feed flow rates exceed 15-20 MMSCFD and raw gas CO2 concentrations exceed 6-8 mol%. In a single-stage unit, 8% to 15% of the total feed methane is lost into the permeate stream. At current gas prices ($3.00 to $4.50/MMBTU), that lost methane represents $2M to $5M+ per year in wasted revenue. A two-stage system captures this permeate methane and recycles it, recovering > 98% of feed methane. The capital and operating cost of the electric recycle compressor is typically paid back within 8 to 18 months.
How does membrane separation compare to amine absorption for offshore and remote gas applications?
Membrane systems provide decisive advantages offshore and in remote desert/Arctic environments: (1) Footprint & Weight: Membranes require 60% to 75% less footprint and weight than amine contactors and regeneration towers; (2) Mechanical Simplicity: No circulating liquid solvents, reboilers, reflux condensers, or foaming issues; (3) Turndown & Modularity: Membrane skids offer near-infinite turndown by valving off individual hollow-fiber modules; (4) Zero Water & Chemical Supply: No water makeup or solvent disposal logistics. However, for extremely stringent specs (e.g., LNG feed requiring < 50 ppmv CO2), amine or hybrid membrane-amine systems are preferred because membranes become uneconomical below 1-2% CO2.
Tool Units

Input / Output Units

Configure global unit system preset or individual parameter dimensions.