Batch Reactor Scale-Up: Fire-Case Relief Is Not the Governing Scenario
A batch reactor of 6.0 m³ working volume is charged with 4,500 kg of reaction mass of specific heat 3.2 kJ/(kg·K). Normal operating temperature is 70 °C on jacket steam at 6 barg. Calorimetry gives a reaction onset at 95 °C and a total heat of reaction of 620 kJ per kg of charge; accelerating rate calorimetry shows the onset of a secondary decomposition at 200 °C. A spring-loaded pressure relief valve set at 4.5 barg, sized strictly for the external fire case to API 521, discharges to atmosphere through a 4-inch vent terminating 3 m above the roof and 12 m from the site boundary. The vapour is methanol-rich and flammable, and the plot adjoins a public road. The client wants your written endorsement to increase the batch charge to 5,400 kg using the same vessel and the same relief system.
1. Evaluate the design basis of the relief system and state whether the fire case governs.
Read the worked answer
Adiabatic temperature rise:
ΔT_ad = ΔH_r / cp = 620 / 3.2 = 194 K
Maximum temperature of a synthesis reaction (MTSR):
MTSR ≈ T_process + ΔT_ad = 70 + 194 = 264 °C
The secondary decomposition onset is 200 °C. MTSR exceeds it by 64 K, so a runaway is not just credible - it is expected to enter the decomposition regime before any relief device can control it. The fire case is therefore not the governing scenario; it was the wrong basis for sizing.
The scale-up compounds it: the heat of reaction per kg is unchanged, so ΔT_ad stays at 194 K and the MTSR argument is identical - but the peak heat-release rate scales with the charge mass. A 20 % increase in charge means roughly a 20 % increase in the required mass relief rate, and therefore at least a 20 % larger required relief area, before any credit is taken for two-phase or foam behaviour. The existing 4.5 barg valve has neither been sized for the runaway nor re-sized for the larger batch.
Conclusion: the relief system is not demonstrated to be adequate for the current process, let alone the proposed one. Endorsement cannot be given.
2. Set out what you would require technically before any charge increase, including the vent-disposal question.
Read the worked answer
Process side:
- Re-run the calorimetry on the proposed 5,400 kg recipe (heat-release rate versus temperature, not just total heat) and derive the design basis case from the MTSR/decomposition pair, per DIERS methodology for the vapour, gassy or hybrid system as the data dictate.
- Size an emergency relief system on that basis: likely a rupture disc upstream of the relief valve (to survive two-phase/gassy flow), with the discharge routed to a catch tank, scrubber or flare - not to atmosphere.
- Reduce the hazard at source before adding hardware: semi-batch operation with accumulation control, reduced charge, lower jacket temperature, addition-rate limitation, a high-temperature trip on the jacket, and independent emergency cooling.
Vent disposal - separate and serious: An atmospheric vent 12 m from the site boundary with a flammable methanol-rich plume is a siting and dispersion problem that exists today. A relief scenario is precisely the moment when the largest release occurs. Routing to a safe location (flare/absorber) or providing containment is a requirement, not an option. This should be assessed against the site's QRA/land-use planning envelope and the Major Hazard Installations regime if the inventory crosses the threshold.
Documentation: a HAZOP covering loss of cooling, mischarge, agitation failure and utility failure; a Management of Change record; and re-examination of the vessel and the relief system by the Authorised Examiner.
3. The client argues the plant has run safely for eight years and asks you to sign 'as the process engineer who knows this plant'. How do you respond?
Read the worked answer
Eight years without an incident is evidence about frequency, not about consequence. The calculation shows a credible event path - loss of cooling or a mischarge at 95 °C leads to an MTSR of 264 °C against a 200 °C decomposition onset - whose outcome is not bounded by the existing relief system. Prior good service does not discharge my obligation.
My response:
- Decline to sign, in writing, stating precisely what is missing: a runaway design basis, an adequately sized relief path, and a safe discharge location.
- Escalate the safety concern to the client's most senior process authority, not just the operations manager. If the response is to proceed anyway, record the advice and the refusal so that the decision is documented as the client's.
- Do not trade scope for approval. I will not sign a narrower document (e.g. 'vessel mechanically fit') knowing it will be used to imply process safety endorsement.
- Offer a route forward: the increased charge can proceed in a smaller pilot quantity, or with the inherent-safety changes and a properly sized relief system, on a stated timescale. A refusal without a path is poor engineering advice.
- Competence and honesty: work in the reactive-hazard area should be signed by someone competent to do the calorimetric interpretation. If that is not me, I say so and bring in the specialist rather than affix my stamp.
The governing principles are that public and worker safety prevail over commercial pressure, and that a professional opinion is given only within one's competence and only when it is honest.