A new open‑access paper from partners in the SteamDry project presents a dynamic model for filtration of dusty superheated steam, directly supporting the move toward more energy‑efficient paper drying. The study focuses on how dust fouling develops in a dead‑end filter and how this fouling increases pressure drop over time in a closed‑loop superheated steam drying (SSD) system.
The work, titled “Dynamic modeling of fouling development during dead‑end filtration of dusty superheated steam,” was published in Systems & Control Transactions and presented at ESCAPE 36 in Sheffield in June 2026. It offers a compact, physics‑based model that can be used for parameter identification, prediction, and eventually optimization and control of SteamDry lines.
Who Is Behind the Study?
The paper is authored by Felipe de Oliveira, Wijtze Nijhuis, and Edwin Zondervan from the University of Twente (Department of Chemical Engineering), together with Marcel Meinders from Wageningen University & Research (Department of Food Technology).
Within the SteamDry project, these groups play complementary roles:
- The University of Twente team leads process modeling, dynamic simulation, and control concepts for superheated steam drying and related unit operations. Their work focuses on creating models that are simple enough for plant‑wide use, but rich enough to capture key physics such as fouling dynamics and pressure‑drop behavior.
- The Wageningen University & Research team contributes deep expertise in food and biomass processing, particle behavior, and experimental methods, helping to design realistic test conditions and interpret fouling phenomena in terms of cake structure and dust properties.
Together, they provide the modeling and experimental backbone that SteamDry needs to develop robust, validated digital tools for SSD design and operation.
What the Filtration Model Does
The authors develop a “parsimonious” dynamic model for dead‑end filtration of dusty superheated steam containing paper‑like dust. The model is built on Darcy’s law and represents the total resistance as the sum of two parts:
- A constant intrinsic filter resistance.
- A time‑dependent cake resistance that grows as dust accumulates on the filter surface.
Cake thickness is linked directly to the deposited dust mass, effective filter area, and cake density, so that, over time, higher dust loading translates into higher pressure drop. Key parameters, such as filter resistance and specific cake resistance, are obtained from experiments, while gas properties are calculated using thermophysical correlations.
From Experiments to Predictive Power
To calibrate and validate the model, the team built a dedicated superheated steam filtration setup. Dust‑laden steam is produced, routed through a PTFE filter mounted in an oven, and monitored for temperature and pressure drop; total deposited mass is determined by weighing the filter before and after each run.
When the dust dosing rate is assumed to be constant, the model captures the overall trend in pressure drop but shows only limited agreement with experimental data in the dust filtration stage (R² ≈ 0.24). By inverting the problem and estimating the time‑varying dust load that best matches the measured pressure drop, the authors improve the fit dramatically to R² ≈ 0.94, revealing how strongly pressure‑drop predictions depend on realistic solid‑loading profiles.
Why This Matters for the SteamDry Project
For SteamDry, this dynamic filtration model is an important step toward fully model‑based design and operation of superheated steam drying systems for paper and board. It can be:
- Extended to a wider range of temperatures, flows, and dust levels, with statistical analysis of fitted parameters to capture variability.
- Coupled with filter‑cleaning models to simulate long‑term cyclic operation and evaluate cleaning strategies.
- Integrated into dynamic optimization and control frameworks that keep pressure drop, energy use, and cleanliness in balance under realistic disturbances.