From Batch to Flow

A response to volume variation of both the external and internal type has been a transition away from a batch to a flow model. As with many other trends, this one has been conhned mostly to single companies — sometimes with immediate suppliers included. An industry in the United States where it has advanced is the automotive industry. The Big 3 have copied Japanese approaches to what has become known as lean manufacturing. Operations are closely linked with components fabricated in the morning finding their way into a vehicle in the afternoon. 

The analogy for the flow vision ideal is a river and is a foundation for lean manufacturing. This river flows evenly at a constant pace from its source to its customers there are no dams and lakes, rapids, or waves 

For many, this view must remain a remote hope. For these companies, their supply chains are too long most of their sales are at Christmas or other peak periods there are too many players to make coordination feasible and, in addition, their crystal ball forecasts are not accurate enough to be successful. 

However, most can do better. Those who are successful in managing their supply chains will make this transition. The following sections trace the evolution of this transition from batch to flow and some of the enabling methodologies that will work in many supply chains. We also believe that companies with uncontrollable external barriers such as seasonality can tailor these methods to make life at least a bit easier. 

The flow vision ideal has a river as an analogy and is a foimdation for lean manufacturing. This river flows evenly at a constant pace from its somce to its customers there are no dams and lakes or rapids to alter the flow rate. The river delivers only what customers want when they want it — not too much, not too little. There is little waste in terms of imwanted product and inventory. The river banks adjusts easily to changes in the water level — at least within the boimds established for it. 

---

Dies (SMED) in a single digit. The SMED philosophy is important in moving from batch to flow oriented supply chains. 

Free phenol is a major concern in the manufacture of novolac resins. This is true for several reasons. The strongest drivers are probably EPA classification of phenol as a Hazardous Air Pollutant and worker safety concerns. However, free phenol also has significant technical effects on such parameters as melt flow characteristics. In this role, free phenol may undermine the desired effects of a molecular weight design by increasing flow beyond the desired point. Since free phenol is often variable, the effects on flow may also cause variation in product performance from batch to batch. Fig. 18 shows the effects of free phenol on the flow across a series of molecular weights. Free phenol contents between 1 and 10% are commonly seen. In recent years, much work has been aimed at reducing the free phenol. 

A careful analysis of the current portfolio of one major pharmaceutical company indicates that about 60% of the chemistry is suitable for continuous processing. About 50% of this chemistry is homogeneous and therefore readily transferable to existing continuous processing technology. The remaining 50% is heterogeneous and will therefore require implementation of some of the current advances in continuous flow equipment such as oscillatory flow reactors [13]. Technically, the transfer of these processes from batch to continuous could happen within... 

A second flow measurement to be considered during scale-up is the ability of the granulate/powder to fill the dies. This can most efficiently be monitored by punch force variability and individual core weight measurements. Acceptable weight control (<3% RSD) and force (<5% RSD) may be masked at slower compression speeds typically used for development or when a tablet press is not fully tooled. Production operations will tend to run at the high end of any validated range, so flow must be consistent from batch to batch. 

Most batch reactors today employ, at best, only partially automatic control. The operator must keep track of numerous valves, motors, and flow and temperature gages — just to control temperature, and just for one reactor. Failure to continuously coordinate these devices leads to varying temperature profiles from batch to batch, and, thus, inconsistent or unacceptable product quality. 

The size reduction step can be the source of problems. Firstly, it is often difficult to replicate the milling and sizing processes from batch to batch which, in turn, leads to poor reproducibility. We have found that a combination of wet ball-milling, using a ceramic system, and laser particle sizing allows the production of higher quality material [67]. However, since the MIP particles are not uniform in shape, MIP columns still tend to demonstrate relatively poor flow dynamics, which result in poor chromatographic performance. In spite of this problem, the exceptional selectivity of the polymers has still made them useful. 

Although the rapid growth of micro-reactor technology has led to the transfer of many common synthetic reactions from batch to chip little attention has been paid to the problems associated with the continuous purification of these reaction products. To tackle this, the incorporation of solid-supported catalysts within miniaturised flow reactors, leading to the synthesis of analytically pure compounds and aiding the development of efficient multi-step processes. 

As already noted, this stems from the fact that the process is almost always developed from a batch-operated beaker or flask. However, it is worth observing that an output of 500 tonnes/year of active substance corresponds to a continuous process flow rate of around only 70ml/sec. This allows various items of intensified equipment to be assembled and operated continuously, literally on a desktop , to meet the production demand. The decision to switch from batch to continuous processing immediately confers a useful intensification benefit because the peak batch process loads (e.g. heat output, liquid removal, etc.) are distributed in time, so equipment size can be reduced. Thus, with a new process, the laboratory scale becomes the full scale when allowed to run continuously and the scale-up delays described earlier are largely avoided. This strategy is generating considerable industrial interest as the commercial pressure to bring new molecules to market rapidly continues to increase. 

In the time or temporal domain, periodicity in operation is incorporated to realize all four principles of PI. A combination of adsorption-reaction-desorption on catalyst surface by periodic forcing of temperatures and pressures demonstrates the application of first principle. Oscillatory baffled flow reactor enhances uniformity, and illustrates the second PI principle. The application examples for third and fourth PI principles are pulsation of feed in trickle bed reactors enhancing the mass transfer rates, and flow reversal in reversed flow reactors shifting the equilibrium beyond limitations respectively. Switching from batch to continuous processing also result in realization of second and third PI principles. 

The Bucher-Guyer horizontal rotary press is a highly automated batch process machine that requires no press aid. The press consists of a horizontal hydrauhc ram inside a rotating cylinder containing many flexible rods covered with a knitted synthetic fabric. The rods have serrated surfaces to allow juice which passes through the fabric to flow to the discharge ends. Hydrauhc pressure is apphed for a preset time, the ram is retracted, and the cylinder is rotated to break up the press cake. This cycle is repeated several times before the press cake is removed from the cylinder and the press is cleaned (16). Juice yield for this horizontal rotary press is 84% with secondary water addition it is increased to 92% (15). 

---