Separator Vessel Sizing

Two-Phase Separator Vessel Sizing

Overview

Implements the Souders-Brown velocity criterion with GPSA-style K-factor ranges (no proprietary table reproduced) for vertical and horizontal two-phase vessels. Liquid holdup time defaults to 5 minutes (typical operator-response horizon); override for your service.

Prerequisites

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

When to Use

• Sizing a knock-out drum, flash drum, suction scrubber, or compressor

inlet drum.

• Screening vendor sizing or rating an existing vessel for new conditions.

Don't use for

• Three-phase separators with explicit oil-water-gas residence time.
• Cyclonic / vane / coalescer-only separators.
• Tower-bottom sumps where downcomer hydraulics dominate.

Utility Scripts

• uv run scripts/sizing.py vertical --vapor-volumetric 0.4 --liquid-volumetric 0.005 --rho-l 600 --rho-g 30 --demister --output /tmp/v.json
• uv run scripts/sizing.py horizontal --vapor-volumetric 0.4 --liquid-volumetric 0.005 --rho-l 600 --rho-g 30 --demister --LD 4 --output /tmp/h.json
• uv run scripts/sizing.py watkins-k --quality 0.6 --rho-l 600 --rho-g 30 --output /tmp/k.json

Procedure

1. Get the vapor volumetric flow at operating P,T (from a flash, or from

measured density and mass flow).

1. Get the liquid volumetric flow.
2. Decide demister or no demister (demister = ~50% smaller diameter for the

same separation).

1. Pick liquid holdup time: 5 min for operator response, 2 min for high

reliability shutdown systems, 10 min for upstream / unattended skids.

1. Pick L/D for horizontal: 3 (compact), 4 (default), 5 (long, high gas

load).

1. Run vertical or horizontal; check that the available liquid holdup

meets requirement.

Pitfalls

• Picking the K-factor at the top of the range without justifying it.
• Sizing for max vapor flow but ignoring the relief contingency flow (which

may govern the design).

• Forgetting to add a demister allowance to tan-tan height (~0.3 m typical).
• Using a vertical vessel for a high-liquid-loading service; horizontal is

often better when liquid-to-vapor flow ratio is large.

• Treating the diameter as exact; round up to standard pipe / fab sizes.
• Reporting K without specifying horizontal vs vertical and with/without

demister.

• Forgetting that high-pressure systems need different K factors than

atmospheric service (the GPSA chart trends with P).

• Sizing on lab/measured density and forgetting that density at operating

conditions can differ by 5-15%.

Fallback Strategies

• If quality x is known, use watkins-k to derive a more realistic K

from fluids.separator.K_separator_Watkins instead of the rule-of-thumb ranges.

• If the holdup time check fails, increase L/D for horizontal or vessel

diameter for vertical and rerun.

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/k_factor_ranges.md — K-factor ranges and citations.
• GPSA Engineering Data Book (procure separately).
• API 12J — Specification for Oil and Gas Separators.

Anti-Patterns

• Treating the vessel as final design without HAZOP, relief, or vendor

review.

• Specifying tan-tan height without showing the holdup calculation.
• Picking K to fit a pre-existing nozzle size rather than to meet the

separation duty.

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.