Caustic Merox Extraction

Caustic / Merox Mercaptan Extraction Screening

Overview

Thin mass-balance helpers built around the chemistry of mercaptan extraction with aqueous NaOH:

RSH + NaOH → RS-Na + H2O

and the subsequent regeneration (oxidation):

2 RS-Na + 0.5 O2 + H2O → RSSR + 2 NaOH

Provides:

• Stoichiometric RS-Na loading at steady extraction.
• Kremser-style stage performance with a user-supplied distribution

coefficient K = C_LPG / C_caustic.

• Risk-score heuristic for disulfide carryback to product LPG (caustic

disulfide loading + caustic age + separator dP).

These are screening and troubleshooting tools, not licensor design calculations.

Prerequisites

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

When to Use

• Troubleshooting a Merox unit (rising RSH slip, caustic age question,

carryback suspicion).

• Sanity-checking caustic circulation rate vs RS-Na uptake.
• Order-of-magnitude operator discussion.

Don't use for

• Licensor design (UOP Merox package and successors are proprietary; use

the licensor's calculations).

• Amine treating: use sour-gas-amine-treating.
• Exact prediction of product disulfide concentration — the carryback

score is directional, not quantitative.

Utility Scripts

• uv run scripts/merox.py loading --rsh-ppmw 500 --lpg-kg-s 10 --caustic-kg-s 0.5 --naoh-wt 0.12 --output /tmp/load.json
• uv run scripts/merox.py kremser --K 0.02 --lpg-vol 0.014 --caustic-vol 0.0006 --stages 3 --rsh-in 500 --output /tmp/k.json
• uv run scripts/merox.py carryback --caustic-disulfide-ppmw 300 --caustic-age-days 60 --separator-dp-kpa 8 --output /tmp/risk.json

Procedure

1. Compute RS-Na loading from inlet RSH and current caustic circulation.
2. Compare with the bound-alkalinity available in lean caustic (total

alkalinity − free alkalinity).

1. If loading is approaching available bound alkalinity, regeneration is

not keeping pace — investigate the oxidizer / separator side.

1. Use the Kremser estimate to bracket required stages for a given RSH

removal.

1. Use the carryback risk score to triage process data when product

disulfide is rising.

Pitfalls

• Treating K as a constant when it depends on caustic concentration, T,

and RSH chain length.

• Confusing "free alkalinity" with "total alkalinity" — RS-Na consumption

reduces free alkalinity.

• Ignoring water carryover between the LPG and caustic phases (and the

reverse): the coalescer / sand filter is part of the design envelope.

• Treating the carryback heuristic score as quantitative.
• Treating disulfide rise in product solely as a Merox issue when it could

be feed (DMS / DEDS / DMDS passthrough).

• Forgetting that caustic age increases mercaptide loading even at constant

RSH inlet; bound alkalinity climbs over time.

• Sizing without considering the disulfide-separator (Stelter/Hsu-style)

performance.

Fallback Strategies

• If the Kremser estimate predicts unachievable removal at any reasonable

stage count, the dominant problem is likely K (poor distribution) rather than stages; revisit caustic strength and contactor design.

• If disulfide carryback score is "high" but product disulfide is low,

there may be a strong DMS feed passthrough; surface to vle-flash-calculations or product-sulfur speciation skill.

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/merox_chemistry.md — chemistry, regeneration, and the

practical role of free alkalinity.

• Kohl & Nielsen, Gas Purification.
• UOP Merox descriptive literature (public).

Anti-Patterns

• Reporting "Merox is the cause" without considering feed sulfur

speciation.

• Treating the carryback score as a value to optimise on rather than a

triage indicator.

• Using these helpers for licensor design or rerate.

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.