pyDAC: Open-Source two-dimensional modeling of s-TVSA for direct air capture: experimental assessment and validity of one-dimensional approximations.
- 1Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, P.O. Box 513, Eindhoven, Netherlands, 5600MB
Abstract
Open and experimentally validated process models remain scarce in direct air capture (DAC), limiting reproducibility and making it difficult to separate physical conclusions from implementation choices. We present pyDAC, an open-source Python framework for packed-bed DAC simulation with one dimensional (1D) and two dimensional (2D) axisymmetric formulations built on the same governing framework. The 2D model is validated against independent zeolite 13X breakthrough experiments, including temperature measurements at multiple radial positions. We then use pyDAC to assess the validity of the 1D approximation for a steam-assisted temperature–vacuum swing adsorption cycle using Lewatit® VP OC 1065. At fixed operating conditions and bed volume, aspect ratios τ = L/D (bed-lenght/bed- diameter) from 1/8 to 8 are examined. The 1D model systematically predicts more favorable performance, overestimating productivity by up to 14 % and underestimating equivalent work by up to 13 %; discrepancies remain around 4 % even at τ = 8. The residual difference originates from the early, conduction-dominated wall-heating transient, which a constant wall heat-transfer coefficient cannot reproduce. During multi-objective optimization, however, the bias reverses because the optimizer selects shorter pre-heating steps within this transient. At τ = 1/2, the 1D Pareto front requires a median 30 % more equivalent work at equal productivity, and the sampled 2D-optimal designs become infeasible when re-evaluated in 1D. At τ = 8, the fronts nearly coincide, but the corresponding operating decisions differ. These results show that 1D models remain useful for screening, but selected designs should be verified in 2D when radial heat transfer controls cycle performance or constraint margins.
Related
- Green synthesis of Ag/MgO nanocomposites from red pepper seed waste for efficient sunlight-driven diclofenac degradation: experimental investigation and EEFO-optimized DT_LSBoost predictive modeling
- Ab initio modeling of morphology with experimental comparison in a nanoscale heterostructure
- Structural Modeling and Theoretical Study of Low-Dimensional Materials
- Bio-based cellulose-filled vitrimers for 3D printing via liquid deposition modeling: Rheological tuning and environmental assessment
- Deciphering pH Behavior in Deep Eutectic Solvents: Synergizing Experimental Data and Computational Modeling
BibTeX
@misc{galanti2026pydac,
title = {{pyDAC: Open-Source two-dimensional modeling of s-TVSA for direct air capture: experimental assessment and validity of one-dimensional approximations.}},
author = {Mattia Galanti and Ivo Roghair and Martin van Sint Annaland},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010382/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010382/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - pyDAC: Open-Source two-dimensional modeling of s-TVSA for direct air capture: experimental assessment and validity of one-dimensional approximations.
AU - Mattia Galanti
AU - Ivo Roghair
AU - Martin van Sint Annaland
PY - 2026
DA - 2026/10/13
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010382/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010382/v1
AB - Open and experimentally validated process models remain scarce in direct air capture (DAC), limiting reproducibility and making it difficult to separate physical conclusions from implementation choices. We present pyDAC, an open-source Python framework for packed-bed DAC simulation with one dimensional (1D) and two dimensional (2D) axisymmetric formulations built on the same governing framework. The 2D model is validated against independent zeolite 13X breakthrough experiments, including temperature measurements at multiple radial positions. We then use pyDAC to assess the validity of the 1D approximation for a steam-assisted temperature–vacuum swing adsorption cycle using Lewatit® VP OC 1065. At fixed operating conditions and bed volume, aspect ratios τ = L/D (bed-lenght/bed- diameter) from 1/8 to 8 are examined. The 1D model systematically predicts more favorable performance, overestimating productivity by up to 14 % and underestimating equivalent work by up to 13 %; discrepancies remain around 4 % even at τ = 8. The residual difference originates from the early, conduction-dominated wall-heating transient, which a constant wall heat-transfer coefficient cannot reproduce. During multi-objective optimization, however, the bias reverses because the optimizer selects shorter pre-heating steps within this transient. At τ = 1/2, the 1D Pareto front requires a median 30 % more equivalent work at equal productivity, and the sampled 2D-optimal designs become infeasible when re-evaluated in 1D. At τ = 8, the fronts nearly coincide, but the corresponding operating decisions differ. These results show that 1D models remain useful for screening, but selected designs should be verified in 2D when radial heat transfer controls cycle performance or constraint margins.
ER -