The magnet factory where I cut my teeth in “process control” had precious few process operations, and they were almost all done in batches. We alloyed and cast metals, milled and blended metal powders, and sintered pressed metal shapes in vacuum (batch) and push (sort of continuous) furnaces, but most of our manufacturing was discrete.
Batch processing looks deceptively easy to control: you think you can know what you put in it, know what you did to it, measure the quality in-process and verify it on completion. And it works fine for beer and shampoo, but when you’re mixing solids like metal powders, it’s strictly impossible for every bit of the batch to experience the same process. For example, blending operations are especially tricky—by the time the mix is homogenous, it’s easy to change its properties by grinding effects, introduce contaminants by blender wear, and cause segregation by particle size or density.
In our case, exposure to the atmosphere or heat during milling or blending caused additional changes that were impossible for us to analyze and quantify, so “quality control” involved making test magnets from each batch before we could put it in production. By the time we discovered a systemic problem, we might have a million dollars’ worth of bad powder—chump change to many of you, but a significant percentage of our annual sales.
Toward the end of my tenure, we shifted much of our powder production from batches in wet ball mills to continuous processing in jet mills, but since we were milling batches of crushed ingots, the output varied and we still had to blend and requalify the milled powder. So, 30 years ago, we started developing processes to produce powders directly from melted alloy. It was worth overcoming the significant technical challenges of melt atomization to get away from the vicissitudes of milling and blending.
In my factory, we sought every opportunity to gain the quality control potential of continuous processing, but magnets aren’t easy to pump through pipes. Apparently, neither are pharmaceuticals, as it’s taken the FDA until this year to recognize the benefits of “continuous manufacturing,” as detailed in a Sept. 11 blog post by FDA Director Michael Kopcha.
“Today, a new and exciting technology—continuous manufacturing (CM)—can truly transform the drug manufacturing process, so it's more reliable and efficient,” Kopcho writes. “As with any new technology, implementing CM presents challenges, such as the initial cost of investing in new equipment. However, the CM production method offers clear benefits for both patients and industry. CM can shorten production times and improve the efficiency of the manufacturing process. CM also allows for more nimble testing and control that can help reduce the likelihood of manufacturing failures. These control strategies could potentially contribute to the prevention of drug shortages.”
The 21st Century Cures Act, enacted in December 2016, authorized grants to support studying CM and recommending improvements to the process of continuous manufacturing of drugs and biological products, and the FDA is encouraging pharmaceutical companies to consider CM for new and existing drugs.
It’s heartening that the principal regulator of this highly regulated industry is encouraging innovation. It’s also an opportunity for experts at continuous processing to share their knowledge, experience, instruments and equipment with formerly batch-constrained facilities and fellow professionals.
As for me, I once wanted to help make the world’s finest magnets. Now I’ll be perfectly pleased to see more and cheaper drugs.