It is considered to be a valid method for using the waste water of an industrial plant in a wet scrubber to control the air pollution of the exhaust gas of the same plant. A new Gas-Liquid Cross-flow Array (GLCA) wet scrubber system is proposed to reduce the particulate matters of the exhaust gas. Thereby, the plant's waste water is formed into a continuing and regular liquid columnar array to separate the exhaust gas particles. The principle of this method is that many continually falling waste water liquid columns act as dust collectors. These dust collectors are perpendicularly streamed by the dusty gas flow and particles can be captured by the water stream due to inertial, diffusion and interception mechanisms. Generally the GLCA system is applied to exhaust gas purification with hot temperatures, where inertial, interception, diffusion, condensation, and thermophertic mechanisms could occur. However, in the first scientific investigation the gas temperature was fixed close to the water temperature. So the condensation and thermophertic effects can be nearly excluded. A model approach for this "cold version" was developed to optimize the design and operating parameters. Experiments on a lab-scale test rig are shown, by which the pressure drop and the grade penetration efficiency were measured for different parameters (number of unit rows, geometric parameters and gas velocities...). To get the optimal GLCA parameters for a high separation efficiency with lowest pressure drop the quality-factor is used. Further a method is presented to calculate the single unit row grade penetration and pressure drop from measurements of the total grade penetration and total pressure drop of a high number of unit rows. With these method it was possible to calculate for a certain given particle size separation efficiency, the needed number of unit rows, respectively the length of the GLCA filter apparatus and the expected total pressure drop. (C) 2016 Elsevier B.V. All rights reserved.