Logo HOSOKAWA ALPINE Aktiengesellschaft
23. May 2016

Automated Cleaning of Complete Process-Technological Systems

Fast product change, safe working conditions as well as high and reproducible hygiene standards: these are the main advantages offered by the automated cleaning of process-technological components. In spite of this, suitable cleaning-in-place systems are still the exception, above all when it comes to complete systems. The peripheral conditions and process parameters are often considered to be too complex, but on closer inspection, they turn out to be easy to manage. Especially in the case of new systems which operate in batch mode, it pays off to include technical cleaning solutions in the design right from the outset and to implement them.

Above all in the pharmaceuticals and cosmetics industry, but also in the food sector, hygiene standards play a pivotal role. This applies especially to production in batch mode, where the systems need to cleaned thoroughly from all residues of the foregoing batch after every run. The European Hygienic Engineering & Design Group (EHEDG) lays down general guidelines for the hygiene-compliant construction of machines, components and surfaces with the aim of avoiding the build-up of dirt deposits and making cleaning easier.

Besides an EHEDG-approved design, it is above all the cleaning of all system components that plays a crucial role in guaranteeing high hygiene standards. In the strictly regulated pharmaceuticals branch, regulatory requirements laid down by the American Food and Drug Administration (FDA) are applied as well as the Good Manufacturing Practice (GMP) guidelines. Annex 15 of the EU-GMP guidelines, for example, dictates that companies who manufacture medicinal products must bring their systems into a condition that is suitable for the production of medicinal products by means of an approved cleaning process. The prevailing guidelines, however, do not unequivocally regulate exactly how the validated cleaning result should be achieved. In other words: exactly how the machine should be cleaned is pretty much left up to the system engineer or operator.

Automated cleaning as an alternative

At the moment, between 85 and 90 percent of all cleaning tasks on this sector are still being carried out manually. This means that under unfavourable conditions, personnel and the environment can be contaminated with highly active medicinal ingredients. Over and above this, manual cleaning is always synonymous with the use of staff capacities as well as the corresponding separate rooms for cleaning. The system also then drops out of the production cycle for the actually avoidable time it takes to clean it.

An alternative is an automated cleaning process by means of integrated systems such as spray nozzles. Generally speaking, a selection can be made between a fully automatic cleaning-in-place (CIP) process which does away completely with all manual cleaning steps or a semi-automatic washing-in-place (WIP) process. In view of the fact that CIP and WIP solutions considerably accelerate the cleaning process, the change-over times can thus be significantly reduced. Over and above this, reproducible hygiene standards are guaranteed.

In spite of all these advantages, it is by no means common today that the component parts of such systems are equipped with CIP-WIP systems. The same applies for complete systems where the advantages are naturally even more salient. One reason for this reluctance is in the complexity of such solutions. However, upon closer inspection, it can be seen that under certain peripheral conditions, an automated cleaning process can be implemented rationally for systems of any size.

Peripheral conditions and parameters

Planning an application-specific optimal CIP system calls for a comprehensive and interdisciplinary approach as well as sound process know-how. The basis is always the question of which individual system components are to be integrated into the solution. The limits should be clearly defined.

The interaction of all affected components in the system including air supplies, inlets and outlets, etc., plays a major role. This applies also to the automated cleaning process. It must not only reflect the individual processing steps but must also be embedded in the higher-order visualisation system of the entire system. Ideally, recipes for different cleaning processes can be stored to memory and can then be started automatically or at the push of a button.

The mechanical construction is based on the general prerequisites which facilitate a cleaning process. Among these, for example, are the use of gradients to accelerate the drainage of cleaning fluids. Here too, a holistic approach is necessary. For instance, in the case of complex systems, it can be expedient to have several drainage points, i.e. in the mill, the filter and the pneumatic system, etc.

Once the peripheral conditions are resolved, the parameters for the cleaning process can be determined. These always depend on the product properties of the coating that needs to be removed, which naturally vary. One can make a distinction between water-soluble (such as lactose) and non-water-soluble substances. In the case of substances that are not water-soluble, cleaning can be performed with the aid of solvents or suitable detergents.

All-important for an optimum wet-cleaning result are the positioning and design of the nozzles via which the cleaning agent is introduced into the machine. The spray angle, spray radius and flow rate all have an influence on the cleaning effect. The number of nozzles, for example, is dependent on the system design: complex geometries with spray shadows call for more nozzles than components with no dead spaces. The final arrangement of nozzles results from practical tests where riboflavin, for example, is sprayed onto the components. Even minimum concentrations of this water-soluble vitamin B2 illuminate under UV light so that residues left over from the cleaning process can be detected very quickly.

 

The actual cleaning process itself constitutes a product-specific complete package made up of medium and cycle. An automated cleaning cycle usually includes several process steps: for example, the substance is solubilised or dissolved during preliminary cleaning or a drying process concludes the cycle. The actual cleaning process can also be performed in several steps. The process can be individually adapted at all points by the selection of cleaning agents, solvents and process temperature.

Example system at a cosmetics manufacturer: cleaning at the flick of a switch

Just how a complex CIP solution can be realised is depicted by the example of several new milling machines installed at a large cosmetics company. They replace the forerunners – meanwhile around 30 years old – and have a throughput rate of around 300 kg/h. This project was realised by Hosokawa Alpine, a globally active specialist for classic mechanical engineering and process technology. The particular challenge here for the fully automatic CIP system was the enormous diversity of some 150 cosmetic ingredients. Because these contain not only a great variety of different colorants but also active ingredients such as UV protection, the hygiene requirements at every product change are extremely high.

One of the products manufactured with the system, for example, is a greasy, non-water-soluble intermediate for face cream. Water at ambient temperature was therefore rejected as the cleaning medium. The implemented solution combines two cleaning processes: the first step is a dry-cleaning step with polycarbonate, a transparent and colourless plastic. The particles are ground in the normal manner in the machine. In this way, stubborn deposits are detached from the walls of the machine. Then follows a wet-cleaning step with hot water (80°C) under the addition of a highly effective cleaning agent that was developed specifically for the task. Additional advantage: the moisture dries on its own as a result of the heat.

The cleaning parameters were determined and optimised over extensive tests. The new system has about three times as many cleaning nozzles as before. In part quite complex, the cleaning programs run completely automatically. The benefit is enormous: instead of the two to three days needed before, the manufacturer now only needs two to three hours for a complete product change.

Conclusion

Automated cleaning systems in complete systems can reduce the time for a product change from days to hours. Moreover, they guarantee reproducible quality standards and thus facilitate a validation. Corresponding CIP and WIP solutions are consistently geared to the product and constitute a complete package comprising cleaning cycle and cleaning medium.