Hybrid Physics–Diagnostic Estimation of Liquid Loading for Wet-Gas Flowmeter Correction: A Model-Based Uncertainty Framework
- 1Independent Researcher
Abstract
The established wet-gas corrections express the indicated gas rate as the true rate multiplied by an over-reading factor that depends on the Lockhart–Martinelli parameter. At most production metering stations that parameter is not measured, so the correction cannot be closed from the meter signal alone. This paper treats liquid loading as a latent variable to be estimated rather than treating the over-reading as a quantity to be learned. A thermodynamic prediction from composition, pressure and temperature is combined with a structurally distinct estimate from flowmeter diagnostic signals already transmitted at the station, and their disagreement is quantified by an estimator consistency index governing whether the correction is applied. The study is computational: no wet-gas loop data were used, the diagnostic estimator is specified but not fitted, and every figure below is a modeled uncertainty rather than a measured accuracy. In a Monte Carlo study of 300,000 simulated realizations over a rich-gas envelope (60–85 kg/cm2, 30–60 °C), three results follow. Inverse-variance fusion degrades the estimate unless the thermodynamic estimator declares an allowance for upstream carryover it cannot observe; with that allowance and an asymmetric weighting, the simulated 95% band falls from ±5.45% to ±4.28%. The consistency index is a specific detector of that failure mode: carryover occurs in 25% of realizations but in 94% of those rejected. Because the two estimates are not statistically independent, the covariance term must be retained in the index; omitting it progressively disables the acceptance test. A falsifiable experimental protocol is specified.
Related
- PCAM: proteochemometric assay-aware modeling for robust bioactivity estimation and uncertainty quantification
- NephroTox AI: an explainable machine learning framework with uncertainty quantification for drug-induced nephrotoxicity prediction
- Beyond Go/No-Go Decisions: A Regional Selection Framework for Uncertainty-Aware Molecule Screening
- A Physics-Informed Fourier Neural Operator Framework for Zero-Plating Fast Charging of Lithium-Ion Batteries via Real-Time Electrochemical State Estimation and Barrier-Riding Optimal Control
- Addition and Correction to “Numerical Investigation of LPG Leakage Dispersion and Explosion Consequences in a Spherical Tank with Safety Walls”
BibTeX
@misc{sandler2026hybrid,
title = {{Hybrid Physics–Diagnostic Estimation of Liquid Loading for Wet-Gas Flowmeter Correction: A Model-Based Uncertainty Framework}},
author = {Leon Sandler},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15009070/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009070/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Hybrid Physics–Diagnostic Estimation of Liquid Loading for Wet-Gas Flowmeter Correction: A Model-Based Uncertainty Framework
AU - Leon Sandler
PY - 2026
DA - 2026/09/18
PB - ChemRxiv
DO - 10.26434/chemrxiv.15009070/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009070/v1
AB - The established wet-gas corrections express the indicated gas rate as the true rate multiplied by an over-reading factor that depends on the Lockhart–Martinelli parameter. At most production metering stations that parameter is not measured, so the correction cannot be closed from the meter signal alone. This paper treats liquid loading as a latent variable to be estimated rather than treating the over-reading as a quantity to be learned. A thermodynamic prediction from composition, pressure and temperature is combined with a structurally distinct estimate from flowmeter diagnostic signals already transmitted at the station, and their disagreement is quantified by an estimator consistency index governing whether the correction is applied. The study is computational: no wet-gas loop data were used, the diagnostic estimator is specified but not fitted, and every figure below is a modeled uncertainty rather than a measured accuracy. In a Monte Carlo study of 300,000 simulated realizations over a rich-gas envelope (60–85 kg/cm2, 30–60 °C), three results follow. Inverse-variance fusion degrades the estimate unless the thermodynamic estimator declares an allowance for upstream carryover it cannot observe; with that allowance and an asymmetric weighting, the simulated 95% band falls from ±5.45% to ±4.28%. The consistency index is a specific detector of that failure mode: carryover occurs in 25% of realizations but in 94% of those rejected. Because the two estimates are not statistically independent, the covariance term must be retained in the index; omitting it progressively disables the acceptance test. A falsifiable experimental protocol is specified.
ER -