What Changes at Scale, and How Do You Know?

Parag Borgaonkar, Senior Business Development Specialist, Chemical Analysis Division, Thermo Fisher Scientific

Dr. Margarethe Richter, Application Specialist Pharma, Thermo Fisher Scientific

When pharmaceutical processes scale, familiar settings may no longer produce familiar results. Changes in geometry, material behavior, and operating conditions raise a second question: how will teams recognize what matters? This article examines the engineering and analytical perspectives needed to interpret change and carry useful process knowledge into commercial manufacturing.

Introduction

A pharmaceutical process that performs reliably on a few grams of material can behave very differently at pilot or commercial scale. The temperature setpoint that worked on the bench, the mixing time that delivered a uniform blend, the feed rate that kept a reaction on track: none of these is guaranteed to mean the same thing in larger equipment. When that difference goes unnoticed, the consequences are familiar to every development leader: reformulation, delayed technology transfer, rejected batches, and the loss of costly active pharmaceutical ingredients (APIs).

Scale-up therefore raises two questions that need to be answered together: what changes as a process moves into different equipment or operating conditions, and what evidence will show whether those changes matter? Engineering helps explain the physical relationships involved, while analytical information helps investigate how the process is actually behaving. Neither perspective tells the whole story independently.

When Familiar Settings Stop Telling the Whole Story

As pharmaceutical processes move into larger equipment, changes in geometry, mixing, mass transfer, and heat removal can alter the conditions experienced by the material. A temperature setpoint or mixing speed that worked during development may not represent the same physical environment at another scale.

Continuous twin-screw extrusion offers a useful example. As explored in Dr. Margarethe Richter’s previously published interview with The Medicine Maker, geometric similarity, energy input, and thermal behavior are important considerations when transferring formulations between laboratory and larger-scale extruders. Changes in screw configuration, residence time, and heat transfer can influence the material’s processing history. Specific mechanical energy consumption (SMEC) provides one useful point of comparison, but matching a single parameter does not establish equivalent material behavior.

Measurements help investigate these differences. In extrusion, trends in torque, temperature, and pressure can reveal changes in operating behavior, while analytical characterization can help evaluate material attributes relevant to product quality. Neither the operating data nor the analytical results should be interpreted without considering the conditions under which they were obtained.

Bioprocessing presents a different set of physical challenges. Increasing bioreactor scale can influence mixing, mass transfer, nutrient availability, and cell metabolism. Real-time metabolite monitoring in bioprocessing can help scientists investigate how cultures respond as operating conditions change. The measurements and underlying mechanisms differ from extrusion, but the underlying question is similar: what is happening to the process, and what does the available evidence tell us about its behavior?

The question isn’t simply whether the settings match. It’s whether the process still behaves as expected. That requires understanding which relationships matter, how they may change, and what evidence is needed to evaluate the process at a new scale.

“The question isn’t simply whether the settings match. It’s whether the process still behaves as expected.”

From Learning to Decisions: How Analytics Must Evolve

The need to interpret changing process conditions also changes what teams require from their analytical methods. During development, experiments help scientists learn what changes, what drives those changes, and which parameters influence product quality. As Parag Borgaonkar discusses in his Pharma Focus interview, the emphasis shifts as a process approaches manufacturing. Measurements must remain reliable under routine operating conditions, and the information they produce must support decisions about process consistency.

Teams usually recognize when an existing approach has reached its limits. The signs are practical: waiting hours for an offline result, or finding that sampling itself has become difficult or disruptive at scale. Depending on the process and the decision being made, Process Analytical Technology (PAT), including spectroscopic methods, may provide more timely information about what is happening during processing.

More timely measurements can make changes in process behavior visible sooner, but observing a trend is not the same as understanding its cause or significance. The information must still be interpreted alongside engineering knowledge, relevant product attributes, and other analytical evidence. The goal is not simply to detect more changes. It is to recognize which changes matter and determine what decisions the evidence can support.

Watch: How analytical strategies evolve during scale-up

Parag Borgaonkar discusses how analytical priorities change as processes move toward manufacturing, what makes PAT useful in routine production, and why process understanding depends on collaboration across teams.

Making PAT Stick in Routine Pharmaceutical Production

Many organizations have piloted PAT successfully in development, only to find that sustaining it in production is a very different challenge. Demonstrating that something can be measured is relatively easy. Making that measurement useful and dependable in routine manufacturing is far harder.

“Demonstrating that something can be measured is relatively easy. Making that measurement useful and dependable in routine manufacturing is far harder.”

The deciding factor is whether the application solves a real manufacturing need, such as reducing manual sampling, monitoring a critical metabolite, identifying a process endpoint or catching a change earlier. Purpose alone is not enough, however. The entire system must hold up on the manufacturing floor: the sensor, the sampling interface, the software, the data, and the model behind it. Validation and model maintenance need to be planned early, not after installation. When these elements are addressed together, PAT becomes part of the process rather than an experiment running beside it.

Analytical models illustrate why this matters. A model developed under one set of conditions may require reassessment when equipment or measurement conditions change, as highlighted by considerations surrounding Raman model transfer between instruments. A dependable analytical strategy therefore involves understanding not only what a model predicts, but the conditions under which its results can be trusted.

Looking ahead, connected sensors and predictive models could bring separate measurements together into more useful process intelligence. But even as these capabilities advance, the underlying challenge remains: ensuring the information is reliable, meaningful, and appropriate for the decisions it is intended to support.

When Process Understanding Changes Hands

Ultimately, a measurement strategy succeeds only if its information can be understood and used by the people responsible for the next stage of the process. That becomes especially important when development knowledge moves between functions, facilities, or manufacturing partners.

Data only goes so far without the right people to act on it. Analytical scientists understand the measurement; process engineers understand what is happening inside the unit operation; manufacturing teams know what is practical on the floor; and quality functions know what will be required to use the information in a regulated environment. When these groups meet only at the end, the result can be an excellent analytical method that is difficult to use in production.

Early alignment around three simple questions helps: what are we measuring, why does it matter, and what decisions will we make with that information? Documenting the reasoning behind key development decisions, including relevant unsuccessful runs and unexpected results, can also help downstream teams understand the boundaries of existing process knowledge.

The value of that documentation becomes clear when a process moves beyond the team that developed it. A manufacturing team or external production partner may be working with different equipment, different sampling arrangements, or different operational constraints. What they need is not simply a record of successful settings, but enough context to understand why those settings worked, which conditions were important and when further investigation may be necessary.

Conclusion: Process Understanding Must Scale with the Process

Moving from development to commercial manufacturing requires more than reproducing operating settings at a larger scale. Teams need to understand how process conditions may change, which differences matter and how the available evidence should be interpreted.

That understanding develops through the interaction of engineering knowledge, analytical measurements, and experience with the process itself. Its value grows when it can be carried forward, rather than remaining with the people who originally generated it.

The goal is not simply to measure more or control more. It is to preserve enough understanding to recognize meaningful changes, evaluate what they mean and help the next team make informed decisions.

Expert Perspectives

This article brings together insights from Parag Borgaonkar’s Pharma Focus Europe interview, ‘Scaling Process Understanding: How Analytical Strategies Evolve from Development to Manufacturing,’ and Dr. Margarethe Richter’s interview, ‘Scaling Smarter: How Continuous Manufacturing Is Reshaping Drug Development,’ originally published in The Medicine Maker.

Author Bio

Parag Borgaonkar

Parag Borgaonkar is an enterprise-focused commercial and product leader with more than 16 years of experience in analytical instrumentation. He is recognized for building and leading global teams, driving strategic growth initiatives, and translating market insights into profitable, customer-centric solutions.

Dr. Margarethe Richter

Dr. Margarethe Richter is currently working as an Application Specialist at Thermo Fisher Scientific, bringing a decade of professional experience in pharmaceutical production. She specializes in extrusion and granulation processes, with a focus on pharmaceutical manufacturing applications.