# Sour Gas Amine Treating

> >-

- **Type:** Skill
- **Install:** `agentstack add skill-jskherman-engg-skills-sour-gas-amine-treating`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [jskherman](https://agentstack.voostack.com/s/jskherman)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** Apache-2.0
- **Upstream author:** [jskherman](https://github.com/jskherman)
- **Source:** https://github.com/jskherman/engg-skills/tree/main/skills/sour-gas-amine-treating

## Install

```sh
agentstack add skill-jskherman-engg-skills-sour-gas-amine-treating
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Sour Gas Amine Treating (Screening)

## Overview

Lightweight mass-balance helpers for aqueous amine sweetening:

- Lean/rich loading in mol acid gas per mol amine from measured flows.
- Required lean amine mass flow for a target rich loading.
- Warning if rich loading exceeds widely cited envelopes (corrosion risk).

This skill is for screening, troubleshooting, and operator-discussion
support. It does NOT model the absorber rate-based kinetics, the
regenerator stripping behaviour, or the corrosion-loading interaction.

## Prerequisites

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

## When to Use

- Estimating required amine circulation for a known acid gas load.
- Checking whether observed rich loading is in the safe envelope.
- Cross-checking a vendor sizing for a sweetening unit.

## Don't use for

- Rate-based simulation of the absorber or regenerator (use ProMax,
  ProTreat, Aspen Amines).
- Predicting foaming, corrosion, or heat-stable salt formation.
- Selecting the amine itself (DEA vs MDEA vs blends): consult vendor
  recommendations and operating history.

## Utility Scripts

- `uv run scripts/amine.py loading --acid-gas-mol-s 0.5 --amine-mol-s 5 --output /tmp/loading.json`
- `uv run scripts/amine.py circulation --acid-gas-mol-s 0.5 --rich 0.40 --lean 0.05 --amine DEA --wt-fraction 0.30 --output /tmp/circ.json`

## Procedure

1. From the inlet sour-gas composition and flow, compute the acid-gas
   molar load (H2S + CO2). Use `vle-flash-calculations` if needed.
2. Decide a target rich loading (typical envelopes: DEA 0.35-0.45 mol/mol,
   MDEA 0.40-0.55 mol/mol). Cross-check against vendor envelope.
3. Decide a lean loading (regenerator performance dependent; typical
   0.01-0.10 mol/mol).
4. Run `circulation`; the script returns required amine molar and mass
   flow, and the solution mass flow at the chosen wt%.
5. Verify the result against the existing pump capacity and lean amine
   exchanger duty.

## Pitfalls

- Treating "loading" as wt%; loading is mol acid gas per mol amine. They
  are very different quantities.
- Using a single rich loading for H2S+CO2 mixed streams without thinking
  about the relative absorption rates (DEA absorbs both; MDEA preferentially
  absorbs H2S over CO2).
- Forgetting that the rich loading envelope depends on amine concentration:
  higher wt% generally lower allowable loading due to corrosion.
- Sizing circulation only for the average case and ignoring upsets (e.g.
  feed sulfur swings).
- Assuming the lean amine is at the stripper outlet condition; lean amine
  cooler and pump can change concentration (e.g. cold lean amine may
  contain dissolved gases that re-vapourise in the absorber).
- Using mass balance loading without checking the regenerator's actual
  performance (it may not strip to the assumed lean loading).
- Ignoring heat-stable salt accumulation in long-run operation.

## Fallback Strategies

- For rigorous design, surface to the user that a rate-based simulator is
  required.
- If acid gas load is unknown, use `vle-flash-calculations` to estimate
  H2S/CO2 partial pressures and flow.

## 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/loading_envelopes.md` — published rich loading envelopes and
  caveats.
- GPSA Engineering Data Book (Section 21 — Hydrocarbon Treating).
- Kohl & Nielsen, *Gas Purification* (5th ed).

## Anti-Patterns

- Reporting "amine circulation 100 GPM" without stating lean and rich
  loadings.
- Picking the rich loading from the top of the envelope without an
  operating margin.
- Using this skill output to specify a tower without rigorous review.

## Source & license

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

- **Author:** [jskherman](https://github.com/jskherman)
- **Source:** [jskherman/engg-skills](https://github.com/jskherman/engg-skills)
- **License:** Apache-2.0

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

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-jskherman-engg-skills-sour-gas-amine-treating
- Seller: https://agentstack.voostack.com/s/jskherman
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.