Powder testing: Supplier vs production reality

When two batches pass identical specifications but behave differently in production, the problem is not the specification limit - it is what the specification measures. This article explains how to build a dynamic testing protocol that detects differences that matter.

The specification problem

Most incoming powder specifications are built around measurements that are quick, reproducible, and historically familiar: particle size distribution, moisture content, Carr's Index, and perhaps angle of repose. These parameters are straightforward to measure and straightforward to set limits for. They are also insufficient as predictors of production behaviour - and the gap between specification compliance and production performance is a well-recognised problem across pharmaceutical, food, and chemical manufacturing.

The issue is not that these parameters are irrelevant. Particle size affects flow behaviour. Moisture content affects cohesion and caking tendency. The issue is that they are upstream measurements - they measure the powder's physical attributes, not its functional behaviour. A batch can be within specification on every attribute and still have a Speed Sensitivity Ratio that is 40% higher than the reference material, or a Bridging Factor that predicts hopper discharge failure under conditions the previous batch handled without incident.

Why traditional specifications miss production-relevant differences

Consider two batches of the same excipient from the same supplier, both within specification on particle size, moisture, and Carr's Index. In a controlled study across multiple batches, dynamic testing reveals the following:

Batch B passes every traditional specification criterion. On the production line, it produces erratic hopper discharge, fill weight drift at the higher line speed, and significantly more caking after storage. None of this is predicted by the Carr's Index or particle size data. All of it is clearly visible in the dynamic parameters.

Building a dynamic supplier qualification protocol

A dynamic qualification protocol does not replace existing attribute specifications - it supplements them. The attribute measurements (particle size, moisture, bulk and tapped density) remain as first-line checks. The dynamic parameters are added as functional performance checks that predict production behaviour.

Step 1: Establish reference values from a known-good batch

Select three to five batches that have performed acceptably in production and run the full Minimum Viable Fingerprint on each under identical, controlled conditions:

•     Standard Cohesion (CI + Bridging Factor)

•     PFSD (SSR + Flow Stability + Comp10)

•     Conditioned Bulk Density

•     Caking test (Cake Height Ratio + Mean Cake Strength)

Calculate the mean and standard deviation for each parameter across the reference batches. These become the reference values for qualification.

Step 2: Set limits based on production sensitivity

Not all parameters are equally important for every application. Set tighter limits for the parameters that correspond to your specific production constraints:

•     If fill weight consistency is critical: prioritise SSR limits. A batch with SSR outside the reference range will produce fill weight drift when line speed changes.

•     If hopper reliability is critical: prioritise Bridging Factor limits. A batch with significantly elevated Bridging Factor will produce more frequent discharge failures.

•     If storage stability is critical: prioritise Mean Cake Strength and Cake Height Ratio limits. Batches outside these limits will produce more severe caking in storage.

•     If batch-to-batch consistency is the primary concern: prioritise Conditioned Bulk Density limits, which are the most sensitive single parameter to manufacturing variation.

Step 3: Define the qualification test set for incoming batches

Not every incoming batch needs the full fingerprint. A tiered approach reduces testing burden while maintaining specification integrity:

How to present dynamic data to suppliers

When a batch fails dynamic qualification, the dynamic parameters provide specific, actionable feedback that attribute measurements cannot. Rather than reporting 'batch failed specification', you can report:

•     'Bridging Factor is 80% above the reference range - likely a change in particle shape or size distribution at the coarser end of the PSD.'

•     'Speed Sensitivity Ratio is 1.67 against a reference of 1.12 - the batch will cause fill weight drift above our standard line speed of 80 units per minute.'

•     'Mean Cake Strength is 89g against a reference maximum of 60g - the batch will produce unacceptable caking in our standard storage configuration.'

This level of specificity allows suppliers to investigate the root cause in their own manufacturing process rather than simply being told the material is out of specification. It creates a more productive supplier relationship and accelerates the resolution of quality issues.

The change control application

Dynamic testing is particularly valuable in change control - when a supplier proposes a change to their manufacturing process that they assess as having no impact on powder performance. A like-for-like dynamic comparison between pre-change and post-change material provides objective evidence of whether the change has affected production-relevant behaviour, regardless of whether traditional specifications show any difference.

This is increasingly relevant in pharmaceutical manufacturing, where process analytical technology requirements and ICH guidelines encourage functional characterisation of raw materials. Dynamic powder flow parameters provide exactly the kind of functional performance data that supports robust change control assessment.