(402d) Monitoring Cleaning-in-Place (CIP) of Multiphase Formulated Products By Electrical Resistance Tomography

Manufacturing competitiveness in fast-moving consumer goods (FMCG) through novel processing technologies and control engineering is an important research area with significant industrial interest. This paper presents the results of a study targeting faster cleaning protocols which minimizes the volume of the cleaning effluent required.  Inline  Electrical Resistance Tomography (ERT) has been demonstrated for monitoring cleaning-in-place (CIP) strategies.  A water–based cleaning protocol via continuous or semi–continuous process flow, to improve product recovery and/or CIP efficiency cycles, was selected as an experimental evaluation method.

A double screen MS type Silverson rotor –stator inline processor was used as the inline test equipment with ancillary pipework attached, as before and after the high-shear inline processor. The pilot scale processor is a Silverson MS 150/250 model (with outer and inner rotor diameters of 63.5 and 38.1 mm and rotor-stator gap of 0.229 mm ).  The 34.78 mm internal diameter stainless steel pipework was filled with various non-Newtonian fluids, including a model-type of structured fluids comprising of SCMC  in commercial haircare products.  The pipework was then cleaned by pumping water through it and monitoring the removal by ERT and visual inspection.  A variety of flow rates were used as the main process parameter.  Two inline ERT test sections with different geometries were selected to assess the CIP processes.  Firstly, a straight in-line Perspex pipe section containing 6 planes of 16 electrodes around the circumference of the pipewall enabled imaging of the fluid displacement from this test section.  Secondly, a Perspex T-piece pipe section containing 2 planes of 16 electrodes around the circumference of the pipe.  The T-piece was intended to mimic a cross-vertical to inline flow sample or instrument port where products could build up.  Consequently, this arranges the ERT sensor planes to be not in the main flow but rather in a short section of dead end pipework.  This enabled the fluid displacement from a model ‘difficult to clean’-type section to be monitored during CIP processes.

The pre- and post–cleaning CIP phases were detectable under the ERT resolution limits over the range of flow rates from ~ 1000 – 3000 kghr⁻¹.  The ERT global conductivity, in terms of mean spatial distribution and variance, were obtained by reconstructive image and transient analysis. Tomography profiles and process normalized dimensionless cleaning time constants have been correlated to illustrate the localized/delocalized heterogeneous nature of inline product-to-water dilution, displacement and cleaning time indexes .