Background: Accurate prediction of pressure drop across air filters is a key factor in the development of energy-efficient filtration systems for industrial air cleaning. Pressure-drop models such as the Jump-Channel model and Davies-type correlations are commonly used during filter development and material screening to estimate the future pressure drop. However, these models are often derived or validated at media or laboratory scale. Therefore, further research is needed to assess their applicability to full-scale filters under realistic conditions.

Objective: The objective of this study is to evaluate how well models such as Jump-Channel and Davies-Extended pressure-drop models predict steady-state pressure drop in full-scale filters. The study aims to increase understanding of the models’ performance at full scale and quantify deviations between model predictions and experimental measurements. In addition, determine if deviations are systematic or dependent on key operational parameters such as airflow rate, filter geometry, and filter loading state. The overarching goal is to enable more efficient filter development and to evaluate the value of models in testing filtering solutions.

Method: Experimental tests were performed using a full-scale test rig consisting of three parts: (1) a bottom section where oil mist (under realistic load) is introduced, (2) filter housing accommodating filter cassette, and (3) a top section equipped with a fan to drive the airflow. Pressure drop and filter mass were measured for new filters and when the filters had reached steady state. The experimental pressure-drop predictions were compared with predictions using Jump-Channel model and the Davies-Extended model.

Results and Discussion: Preliminary results indicate that the...