Runaway Exothermic Reaction in a Batch Reactor
Vinit Pandey · Published 29 June 2026
This entry illustrates a representative incident pattern seen across the process industries. It does not describe a specific named company or disclose any client-identifying information.
In short: This lesson illustrates why thermal hazard testing data generated at lab scale must be re-validated whenever batch size or charge sequence changes — a scaled-up batch without updated DSC/calorimetry data lost cooling and ran away within minutes.
- Batch Scale-Up
- Cooling System Fault
- Heat Removal Lost
- Temperature Excursion
- Runaway Reaction
- Emergency Relief Discharge
What Happened
A manufacturing site scaled up a batch chemical process from pilot to full production scale without commissioning new thermal hazard testing for the larger charge.
A cooling water supply fault during the exothermic addition step went unnoticed for several minutes due to an unalarmed temperature trend.
Heat generation outpaced the now-reduced heat removal capacity, and the batch entered a self-accelerating runaway reaction, ultimately discharging through the emergency relief system.
Root Causes
- Adiabatic calorimetry (e.g. ARC/RC1) data from the original lab-scale batch was assumed to extrapolate linearly to the larger scale, when surface-area-to-volume cooling ratios actually changed significantly.
- No independent high-temperature alarm or automatic addition cutoff existed on the reactor — only the cooling water flow itself was monitored, not reactor temperature trend rate.
- Management of Change (MOC) review for the scale-up did not flag thermal hazard testing as a required reassessment trigger.
Lessons
- Thermal hazard testing results are scale- and configuration-specific; a batch size change is a MOC trigger for re-testing, not just a yield or capacity question.
- Rate-of-temperature-rise alarms, not just absolute temperature or cooling-flow alarms, are needed to catch a developing runaway early enough to act.
- Emergency relief systems sized for the original thermal hazard testing data may be undersized once the actual scale-up runaway rate is known — relief system design and thermal testing must stay in lockstep.
Technical Takeaways
- DIERS methodology relief sizing depends directly on validated adiabatic temperature rise and self-heat rate data for the actual production scale.
- MOC procedures should explicitly list batch scale-up as a category requiring thermal hazard re-testing, not leave it to engineering judgement alone.
- A reactor's maximum credible self-heat rate, not just its steady-state heat balance, should size the alarm and trip philosophy.