Absorption Stripping Design

Absorption and Stripping Shortcut Design

Overview

Implements the Kremser equation and helpers for stage counting in dilute absorber/stripper systems. The Kremser formulation is exact for systems with constant absorption factor A = L / (K V) and a linear equilibrium relationship y* = K x.

For concentrated systems, reactive absorbers, or rate-based design, use a rigorous simulator or rate-based correlation; this skill is screening only.

Prerequisites

1. uv available.
2. On first use the script writes LICENSE_NOTIFICATION.txt.

When to Use

• Sizing a dilute absorber (lean oil, glycol dehydrator screening, dilute

acid-gas scrubber).

• Estimating the stripping factor and stage count for a desorber.
• Cross-checking vendor sizing for a tray or packed column.

Don't use for

• Reactive absorption with chemical equilibrium (DEA/MDEA + acid gas):

use sour-gas-amine-treating for screening or a rate-based simulator.

• Concentrated solute (mole fraction > 5-10%): Kremser breaks down.
• Detailed packing-specific design: use vendor correlations or a

rate-based simulator.

Utility Scripts

• uv run scripts/kremser.py --N 8 --A 1.4 --K 0.5 --y-in 0.02 --x-in 0.0 --output /tmp/kremser.json
• uv run scripts/kremser.py --N 8 --A 0.8 --K 0.5 --y-in 0.02 --x-in 0.0 --output /tmp/stripper.json

Procedure

1. Decide on solute, lean-solvent K-value at column conditions (often

evaluated at a representative T).

1. Pick a target removal fraction (e.g. 95% of solute absorbed).
2. Compute the required absorption factor A = L / (K V); aim for

A = 1.2 - 1.5 (industry rule of thumb for non-reactive dilute systems).

1. Solve Kremser for N stages.
2. Convert equilibrium stages to actual stages with a Murphree efficiency

(~0.5 - 0.8 for trays; for packed beds, divide column height by HETP).

1. Sanity-check operating line vs equilibrium line: lines must not cross.

Pitfalls

• Using Kremser for a reactive absorber. The acid-gas + amine system needs

a different framework.

• Picking A = 1.0 (or just slightly above) and reporting it as feasible;

it implies infinite stages at the limit.

• Using a single K for the whole column when K varies strongly with T or

composition. Average it with care.

• Confusing absorption factor A = L/(KV) with stripping factor S = K V/L.
• Reporting equilibrium stages as actual trays without applying efficiency.
• Forgetting to verify that the operating line and equilibrium line do not

cross within the column (a sign that the requested removal is infeasible at the chosen A).

• Picking L/G from the gas mass balance only — heat balance and rich solvent

loading change the effective L.

• Using Kremser for solute mole fractions > 5-10%; non-linear equilibrium

invalidates the constant-A assumption.

Fallback Strategies

• For reactive absorption, switch to sour-gas-amine-treating (screening) or

a rate-based simulator (ProMax, ProTreat, Aspen rate-based).

• For high-concentration systems, do a manual stage-by-stage operating /

equilibrium analysis on a y-x diagram with varying K.

Verification

• Run the listed script with representative inputs and an --output file when a deterministic calculation is available.
• Confirm the JSON result contains ok: true, expected units, and no unhandled warnings.
• Check result magnitudes against the stated assumptions, references, and a hand calculation or known operating range before reporting them.

References

• references/kremser_derivation.md — derivation and assumptions.
• Treybal, Mass-Transfer Operations.
• Sinnott, Coulson & Richardson, Volume 6.

Anti-Patterns

• Reporting "5 actual trays" without specifying Murphree efficiency.
• Picking A from the recommended range without justifying it from

capex/opex trade-off.

• Mixing up "absorption factor" with "absorption efficiency".

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: jskherman
• Source: jskherman/engg-skills
• License: Apache-2.0

Install and usage instructions live in the source repository linked above.