A Relief System Change During a Debottlenecking Project
An engineering project team proposes to increase a reactor's batch size by 20% to meet demand. The reactor has a rupture disc and relief valve routed to a scrubber. The team's memo states the relief valve 'has margin' because the design case was a utility failure and the new case is only a 'slow overpressure'. The memo also notes the scrubber has a 10% spare capacity and that the change will be implemented under the plant's 'like-for-like replacement' route to avoid a full hazard study. Commissioning is scheduled in six weeks. The process engineer who raised the original relief case has left the company.
1. Critique the engineering reasoning in the memo and explain why the proposed change route is inappropriate.
Read the worked answer
The memo contains several unsafe assumptions:
- 'Slow overpressure' vs design case: Relief sizing depends on the controlling scenario and the required mass flow, not on how fast the pressure rises. A larger batch changes the wetted area, the vapour generation, and potentially the runaway or decomposition case. A slow scenario can still govern if it produces a larger relief load than the original utility-failure case (e.g. blocked outlet with continued heating, or a reaction runaway at higher inventory).
- 'Has margin' without a basis: Margin is a conclusion from a re-run of the relief case, not an assertion. The original author has left, so the design basis may not be fully documented - a documentation gap that must be closed before any change.
- Scrubber spare capacity: The scrubber is part of the relief path. 10% spare may be consumed by the increased relief load, and the downstream back-pressure affects the relief valve's capacity. The whole path (disc, valve, pipework, scrubber, vent) must be checked, not just the valve.
- Wrong change route: A 20% batch increase is not like-for-like. It changes the process envelope and invalidates the basis of safety. It requires a management-of-change review, and likely a hazard study (HAZOP or equivalent) plus revalidation of relief and relief disposal.
Conclusion: the memo's reasoning is not defensible; the change should not proceed on that basis.
2. Outline the technical work you would require before the change could be approved, and how you would handle the six-week schedule pressure.
Read the worked answer
Required technical work:
- Recover the design basis: locate or reconstruct the original relief case, including scenario, relieving rate, set pressure, back-pressure, and disc/valve data. If it cannot be recovered, re-derive it.
- Identify controlling scenarios at the new batch size: utility failure, blocked outlet, thermal expansion, reaction runaway, fire exposure as applicable. Determine which gives the largest relief load.
- Re-rate the relief path: disc, valve, inlet and outlet pipework, and scrubber against the new load and back-pressure; confirm the scrubber can handle the flow and that the vent is safe.
- Hazard study of the change: a HAZOP or focused study on the reactor and relief system, with the new operating envelope, plus review of any new interactions.
- Management-of-change: documented risk assessment, approval by an authorised engineer, updated operating procedures and training, and a pre-startup safety review.
Handling schedule pressure:
- State clearly that the six-week date is not achievable if the relief case cannot be validated; propose a realistic date.
- Offer interim options: debottleneck elsewhere, run at current batch size, or a temporary operating limit with compensating measures - but only if the relief case is shown adequate for that limit.
- Escalate the conflict between schedule and safety to the accountable manager in writing. Do not let the 'like-for-like' label be used to bypass the change process.