Enabling continuous manufacturing in the pharmaceutical industry through numerical simulation

In the pharmaceutical industry, traditional batch manufacturing processes are proving to be too inefficient for today’s world of economic pressures and increased global competition. There is no question that continuous manufacturing is the path forward towards leaner processes.

The batch-based systems currently in place in the pharmaceutical industry are inefficient due to segmented steps involving multiple facilities and requiring start and stop of the batch, site-to-site transfer and warehouse storage. Product quality assessment is also cumbersome, because it is performed through sampling and in post-production, causing long lead times and waste.

Moving away from batch processing to a continuous, non-stop, end-to-end manufacturing process, however, could modernise the industry and solve its crisis of productivity. Josef Jimenez, CEO of Novartis, recently stated that changing production from batch to continuous will transform the way medicines are made around the world. He said it could cut the time from development to market-entry in half.¹ Implementation of these processes will result in smaller production plants, lower inventory costs, reduction in carbon footprint and higher quality products.²

Figure 1: Possible ultra-lean manufacturing, from chemical synthesis to final dosage form. (Source: Novartis-MIT Center for Continuous Manufacturing)

New moves by regulatory agencies are encouraging the development of new manufacturing technologies by building quality into the process and using a science-based quantified risk approach, by starting to lay the groundwork for continuous manufacturing with several initiatives,^3,4 and with regulatory frameworks such as process analytical technology (PAT) and quality by design (QbD).

Both the chemical and food processing industries have been improving their productivity by successfully integrating continuous manufacturing into their plants. It is clear that regulatory hurdles and conservative thinking by the pharmaceutical industry can no longer be used as an excuse to avoid taking pharmaceutical manufacturing into the 21st century.

Predicting process behaviour through numerical simulation

For continuous processing to become possible, potentially suitable processes must be identified and designed, and the associated risks need to be analysed and mitigated, in order to make a business case for implementation, and to manage regulatory compliance. Multiphysics computational fluid dynamics (CFD), a numerical method for predicting the coupled behaviour of fluid, gas and particulate flows including heat and mass transport, offers a solution for the enhanced understanding and design of these processes. It is emerging as game-changing technology to help continuous manufacturing for active pharmaceutical ingredients (APIs) become a reality through virtual prototyping, optimisation and modelling of the complete system.

Virtual prototyping

Numerical simulations enable the engineer to build a virtual laboratory, providing insight into the performance of a product before tests are carried out. This means that the uncertainty resulting from major process and equipment changes can be evaluated up front, leading to a significant risk reduction and cost savings. Contrast this with traditional manufacturing processes that are based on the ‘design-build-test’ principle - in which the effects of design changes are quantified by experimental tests on physical prototypes. Physical prototyping in the pharmaceutical manufacturing context is anticipated to be very costly.

Multiphysics CFD and state-of-the-art visualisation tools also offer a wealth of detailed information, not always readily available from laboratory or experimental tests. This not only results in an increased level of insight into the details of what is going on inside processes, it also enables innovation. For example, multiphysics CFD can help explore new reactions and molecules for drugs manufactured with a continuous process.

Design exploration and optimisation

The maturing of robust simulation tools and the increase in computing power over the years have made it possible to use numerical design optimisation in production environments. Parameter studies and optimisation will be vitally important for designing and tuning of the new (often smaller) equipment required for continuous manufacturing while ensuring that the operation can efficiently handle fast reactions and remains flexible. In addition, the CFD-generated responses - obtained through design of experiments over a range of operating conditions and equipment design parameters - can be combined with statistical models to identify risk and implement robust real-time process control. This will ultimately result in reduced variability and consistent, repeatable processes.

There are now tools that enable intelligent design exploration and to easily consider ‘what if’ scenarios so as to identify the critical manufacturing points that define quality. For example, feeding devices for continuous manufacturing influence all downstream operations, and design exploration of parameters such as feed rate will help identify their impact on final blend uniformity.

Simulating systems

Solving complex real-world problems demands an accurate, easy-to-use, multidisciplinary approach to simulating complete systems. CFD-focused multiphysics engineering simulation tools such as STAR-CCM+ can accurately deliver full spectrum engineering results and the pharmaceutical industry should fully leverage these tools in support of the development of continuous manufacturing processes.

Up until now, integration of numerical simulations in a production environment has required a great deal of specialised knowledge, but this is no longer a showstopper. Automation and ease of use are enabling the deployment of CFD for complex multiphysics applications. For example, STAR-CCM+ offers state-of-the-art meshing, seamless integration with CAD and easy modelling of complex moving parts, all in a single integrated environment. The net result is more time for an engineer to analyse data instead of preparing and setting up the simulations, resulting in engineering success.

Seeing the ‘big picture’ for continuous manufacturing will require a multiphysics approach to solving problems. Be it mixing, coating or drying, multiphase flows lie at the core of the pharmaceutical processing industry. Capabilities such as discrete element modelling (DEM), a numerical method for computing the interaction of a large number of small particles, and Eulerian multiphase modelling (EMP), a numerical method for simulating several phases in a system, will be invaluable for implementing continuous manufacturing of APIs. Two case studies are presented next to demonstrate these capabilities.

Figure 2: Simulation showing tablet velocity as they tumble in a coating pan.

Case study 1: Direct element modelling (DEM) for pill coating

DEM simulates the motion of a large number of interacting particles and tracks them in a numerically efficient manner, modelling contact forces and energy transfer due to collision and heat transfer between particles. DEM will be particularly important in the design and optimisation of continuous coating processes to help identify the important factors for equipment design (such as the number of spray guns) and to determine optimal equipment operation conditions (such as inlet temperature).

Figure 3: DEM simulation showing pill coating thickness in a fluidised bed.

Figures 2 and 3 show STAR-CCM+ generated solutions for two types of equipment currently used for real-world tablet coating: coating pan (rotating drum) and fluidised bed. In these simulations, DEM is used to analyse the random movement of the particles as layers of coating are applied. Parameters such as particle velocities, residence time and coating thickness are tracked to assess and improve tablet coating uniformity. In addition to tablet coating, DEM can also be used to simulate other steps in manufacturing such as filling, filtering and conveyor processes.

Case study 2: Eulerian multiphase (EMP) modelling for mixing

EMP modelling provides an effective means for studying the interacting streams and randomly dispersed phases in multiphase flows. The EMP model in STAR-CCM+ includes an extensive range of submodels including break-up and coalescence models for bubbles and droplets and a granular flow model for particles. Figure 4 demonstrates an EMP simulation of a gas-liquid mixer with three rotating impellers. Shown are the effects of increasing gas injection rates on gas. The ability to predict gas hold-up, a parameter that governs mass transfer across the phases and consequently rates of reaction, is a key enabler in the design of such reactors. This approach adds valuable scientific insight into the decision-making criteria to develop practical solutions for mixing and other processes in continuous manufacturing.

Figure 4 : Mixer model showing the effects of increasing gas injection rate.

Conclusion

In today’s competitive climate, manufacturing must become leaner with a focus on building quality into the process. Continuous manufacturing for the pharmaceutical industry will change the way drugs are made and multiphysics CFD simulations offer a cost-effective way to perform rapid prototyping for design of new equipment and processes. In particular, design optimisation tools and powerful multiphase models such as DEM and EMP will play an important role, and the pharmaceutical industry should fully leverage these state-of-the-art technologies for the design and implementation of continuous manufacturing processes.

References 1. Jimenez, J 2012, ‘A Defining Moment : The Future of Manufacturing in the US’, conference presentation, The Future of Manufacturing in the US, MIT, 8-9 May. 2. Plumb, K 2005, ‘Continuous Processing in the Pharmaceutical Industry : Changing the Mindset’, Chemical Engineering Research and Design, vol 83, pp 730-738. 3. Woodcock, J & Woosley, R 2008, ‘The FDA Critical Path Initiative and its Influence on New Drug Development’, Annual Review of Medicine, vol 59, pp 1-12. 4. US Food and Drug Administration, Pharmaceutical cGMPS for the 21st century: A Risk Based Approach, <http://www.fda.gov/drugs/developmentapprovalprocess/manufacturing/questionsandanswersoncurrentgoodmanufacturingpracticescgmpfordrugs/ucm137175.htm>.

*Sabine Goodwin is a Senior Engineer, Technical Marketing, and Dr Kristian Debus is Director, Life Sciences, at CD-adapco.