(347h) Simulation-Based Comparison between Three Different Clean-in-Place Configurations Regarding Their Cleaning Efficiency and Water Use

In many process industries, like chemical, food, process and pharmaceutical industries, large tanks are periodically filled, emptied, and refilled with different products. Prior to tank reuse with new liquid, it is essential to ensure any previous residues are completely removed from the tank’s internal walls, as to maintain proper process hygiene. A clean-in-place (or CIP) system is the typical procedure used to clean a vessel after use and before product changeover.

Depending on the product being removed and the equipment size, dismantling of equipment and manual cleaning is usually not attractive economically and a clean-in-place system is desired. A clean-in-place (CIP) process can be automated, standardized, and hence allows for certification. The process involves installing a spray head assembly, figure 1, in which water or cleaning solution jets from the spray head assembly. Different types of spray head assembly exist and are used. The spray head types considered in this study are stationary ball, rotating ball, and orbiting nozzles, figure 2. The stationary ball is the simplest design, but also consumes the most cleaning liquid and takes the most time to cover the same surface. The orbiting nozzles configuration is the most expensive but consumes the least amount of cleaning liquid. The rotating ball is an intermediate design between the stationary ball and the orbiting nozzles.

Typical objectives are obtaining a fully cleaned surface while using the optimal amount of water or solution and within the shortest cycle time. Complete surface coverage with cleaning liquid is essential and “shadowing” effects, i.e., lack of coverage in certain areas, need to be avoided. Due to a wide variety of tank sizes, tank fittings, impeller sizes and their location, among other variables, ensuring the complete coverage and avoiding shadowing can become a lengthy process with each design change. Also, scale up/down rules are nonexistent and engineers rely on previous experiences which changes with time and geographical location making the process difficult to certify and relying on full scale testing which is expensive and time consuming. Computational Fluid Dynamics (CFD) can provide an economical tool to predict film coverage and wetted area. Different tank sizes and accessories (impellers, baffles, dip tubes, internal piping, etc.) can be easily added to the model and their effect on the cleanability can be easily predicted.

For a CIP system, typical input parameters are water pressure (or flow rate), spray head type and its location with respect to the vessel, and finally spray head assembly motion (single-axis rotation for rotating ball, bi-axis or orbiting for orbiting nozzles). In many cases, for larger tanks, there could be multiple spray heads to ensure complete coverage. The engineering challenge is to select and proper place the spray head within the vessel to ensure the complete cleaning, in the shortest amount of time, and using the least amount of cleaning liquid.

Sami et al. [1] studied the effect of inlet water pressure, nozzle design, and distance between the nozzle and target surface on the jet throw (trajectory) and impinging pressure. Metwally et al. [2] studied the effect of nozzle motion on the surface coverage. In this study, different spray heads will be compared, figure 2, in terms of their coverage and time needed to reach that coverage. 100% presented on water jet formation and the impact pressure by a water jet at different nozzle inlet pressure and at various distances between nozzle and surface to be cleaned.

The modeling approach consists of two main ingredients. The first is the water droplet tracking which is done using a lagrangian approach (Discrete Phase Model or DPM for short [3]). This model allows for the accurate tracking of the water droplets and considers its inertia, air drag, and initial droplet conditions from the spray head (location, motion if any). The second ingredient is modeling the water wall film formation [3], spreading, and splashing on the tank surface.

Three different simulations for the same tank were set up, one for each spray head studied. The water flow rate through all the spray heads was identical. Also, the duration for cleaning was identical (one minutes). Figure 3 shows the final film coverage predicted for the three cases studied. It is obvious that the stationary head is suffering from serious shadowing and the top of the tank is not completely cleaned. Both the rotating spray ball and the orbiting nozzles can obtain full coverage. To further quantify the performance of the different spray heads, the % wetted area evolution is plotted in figure 4. The entire cleaning cycle (one minute) % wetted area evolution is shown in figure 4a while the last 40 seconds are shown in figure 4b. Figure 4a shows that the orbiting nozzle starts slower in covering the tank surface, but eventually it overtakes the stationary spray ball. Figure 4b on the other hand further quantifies the shadowing effect when using the stationary spray ball. It also shows that the rotating spray ball is “faster” in cleaning compared to the orbiting nozzle, but the orbiting nozzle is eventually able to catch up.

This modeling methodology is a very powerful tool in expediting the certification and optimization of CIP process. By proper selecting and sizing the system for different equipment configurations, the amount of cleaning liquid used can be significantly reduced and the certification process can be considerably shortened.

[1] M. Sami, H. Metwally, J. Ibrahim, and K. Brown, “Using CFD to Simulate Mixing Tank Clean-In-Place Process”, NAFEMS 2019 World Congress, Quebec City, Montreal, Canada

[2] H. Metwally, M. Sami, K. Brown, “Using CFD to Simulate Tank Clean-in-Place Process”, AICHE 2020

[3] Ansys Fluent User Manual, “Modeling Discrete Phase,” Chapter 24, version 2021 R1