Packed-column flooding: correlation against one measured operating path

A frozen correlation meets one measured air-water flooding path. No constant was fitted to the pressure or flooding observations on this apparatus.

This is one apparatus, checked against one frozen correlation. One air-water path, one fixed liquid load, and one polypropylene Mellapak 250.Y apparatus.

Verdict

UNDERPREDICTS OBSERVED FLOODING

The frozen correlation predicts flooding before the experiment's observed flooding bracket. The gap remains after the recorded voidage discrepancy is tested as a fixed sensitivity case.

The experiment brackets flooding between 6.74 m/s and 6.89 m/s.

First observed flood 6.89 m/s superficial gas velocity

Frozen model flood 4.77 m/s superficial gas velocity

Model predicts early 2.12 m/s, or 30.7 percent

Does the correlation find the same operating limit?

Primary and sensitivity flooding predictions occur below the observed bracket — **The input discrepancy does not remove the result.** The fixed 0.88 voidage check predicts 5.20 m/s. It remains 24.5 percent below the first observed flood.

The raw differential contains more than packing friction

Measured pressure spans the apparatus. The correlation predicts packing friction. No independent fixed-loss measurement was available.

The model is defined for 23 of 35 measured operating steps.

Its mean model-minus-measurement bias is -1.80 kPa across those steps.

The pressure residual mixes packing friction with unmeasured apparatus losses. It is diagnostic and is not a clean validation statistic.

Measured apparatus pressure drop rises sharply after loading while the model reaches flooding earlier — The model curve stops at its predicted flooding limit. The measured series continues to the visually observed apparatus flood.

One path, one model, no refit

No adjustable control is shipped. A voidage slider could tune the result and soften the fixed negative finding.

Basis	Value	Note
Experimental unit	35 operating steps	Each mean uses 300 source readings from one stabilized window.
Velocity conversion	9 printed pairs	The largest reconstruction residual is 0.0062 m/s.
Packing basis	polypropylene Mellapak 250.Y	Voidage 0.85, area 250 m2/m3, and height 0.945 m.
Operating basis	Air and water	The liquid load is 21.17 m3/m2h at the reported ambient condition.

Every input has a named source or a labelled scenario basis

Measured data	Dataset DOI 10.7488/ds/3151, CC BY 4.0, accessed 2026-06-22.
Observed regimes	Companion paper DOI 10.1109/TIM.2022.3184363.
Exact packing inputs	Author thesis DOI 10.7488/era/2884.
Equations	Stichlmair, Bravo and Fair, 1989, through fluids 1.3.1 under MIT.
Frozen source checksum	`8dc81185d14923547d2503fa3da9ea19dfdc5c639384b13c32407aa13b5b9f30`

What this means

The model is conservative here. That is not a universal safety margin.

The study has one packing, one liquid load, and one apparatus.

It cannot establish a general correlation boundary.