Case Study: Large Synthetic Rubber Producing Company identifies causes of off‑spec batches and limits revenue loss with Aspen Technology's Aspen ProMV

Multivariate Statistical Analysis Finds the Bad Actors in Out-of-Spec Batches

Large Synthetic Rubber Producing Company had been producing out-of-spec batches from a reactor-based polymer process, resulting in revenue loss from rework or downgraded sales and an inability to pinpoint the root cause despite five months of data (55 batches). Aspen Technology was engaged and used its Aspen ProMV desktop batch modeling solution to investigate which process variables were driving the quality issues.

Aspen Technology developed an Aspen ProMV batch model that analyzed roughly 80 variables, pruned low-variance noise, and uncovered three latent variables that separated good and bad batches; the analysis identified three key process drivers (including temperature during Additive 3 addition, initiation temperature, and catalyst level) and showed how reactor phase timing affected quality. The Aspen ProMV results made it easy for engineers to see root causes and actionable changes, providing proof the solution can quickly pinpoint issues to limit losses and improve batch quality.