Colloid filtration theories for fibrous bed coalescer design

Fibrous bed coalescers are used in a wide range of applications, from food and soap processing to crude oil and fuel processing, yet there is no formal methodology to size coalescers. The closest form of a design equation comes from Lee and Liu’s (LL) 1980 colloid filtration theory, which considers inertial impaction as a method of drop capture. This work compares the LL framework to two additional colloid filtration theories. The first one is Rajagopalan and Tien’s (RT) 1976 theory, which considers inertial impaction and the detachment of a captured drop due to shear conditions. The second theory is a modified form of the RT framework that includes the coalescence probability (RTCP) between captured drops and drops in the incoming flow. For pad coalescers, the LL, RT, and RTCP theories were compared in terms of the predicted bed length for various combinations of superficial velocities (q) and bed porosities (ε_o). In the typical design range for “q” (0.6 - 3 cm/s), the three frameworks produce similar packed bed lengths. However, only RTCP can be used for high and low “q” values. The RTCP framework was also the only one capable of reproducing the recommended industrial practice for the design of cartridge coalescers. The RTCP framework was also used in a case study for the dewatering of diluted bitumen froth. A fibrous bed coalescer was sized to remove emulsified water from diluted bitumen extracted from the oil sands in the Canadian Athabasca region. The manufacturer-recommended “q” range of 0.6 - 3 cm/s was the most reasonable because higher “q” values produced larger pressure drops in the coalescer and a substantial increase in the size of the gravity separator.