How to conduct the selection study of depth filters?

In general, the more complex the filtration  (such as extreme pH, solvents, late-stage development, or viscous stream), the more necessary to conduct small-scale studies for selection and evaluation. The results of selection studies should be validated at production scale. However, for simple single-step filtration processes (such as in the polishing stage), selection studies may also be considered concurrently during production.

If feasible, selection studies should utilize multiple batches of feed streams and filters to identify the most suitable media and area. Additionally, appropriate safety factors should be incorporated during the scaling process.

The selection of filter media and hardware should be determined based on process compatibility and system requirements, with influencing factors including:

• Feed stream to be filtered
• Flow rate
• Particulate content
• Operating cost
• Cleaning/CIP/SIP
• Local regulatory requirements

During the selection study, the following parameters are commonly evaluated:

• Pressure profile
• Filtrate quality
• flow rate
• Temperature / viscosity
• pH
• Loading capacity assessment
• Protection of downstream filtration

Typical Selection Testing

In small-scale studies, the process feed stream is filtered until   it reaches a predetermined maximum differential pressure or until breakthrough of particles or colloidal complexes occurs.Two commonly used test methods are : constant pressure testing or constant flow rate testing. The quality of the filtrate can be monitored by measuring turbidity.

Constant Pressure Testing:

Pressure is applied to the process feed stream to force it through a small-scale depth filter device. The filtrate volume and flux are monitored, and the test is stopped once all the feed stream has been filtered, or when the filtrate flux drops to a certain percentage of the initial flux. The flux, expressed as a function of the filtrate volume, should be plotted on a flow decay graph.

Constant Flow Testing:

This method offers greater comparability to large-scale production. A feed pump is used to filter the feed stream at a specified flow rate, and the pressure drop across the filter is measured. The test is stopped when the maximum differential pressure is reached or when all feed stream has been processed. urbidity, expressed as a function of the filtrate volume, should also be plotted.