Cooling System Failure / Root Cause Analysis

Mechanism of Failure

The 2000L still was connected to a cooling reservoir designed for the 300L legacy still. Under sustained load, heat returned to the reservoir faster than it could be dissipated. Temperature increased at approximately 1–2°C/hr, rising from around 30°C → 54°C over 12 hours.

As coolant temperature rose, the system became less effective at condensing vapour back into liquid. More vapour passed through without being properly captured. Distillate output temperature increased, less intended flavour was retained, and the final profile no longer matched the expected standard.

The system behaved the same way in every batch. The difference was duration. During the 7.5 hour Orange Liqueur batches, temperature increased but stayed within a range that did not affect flavour. The gin batch ran for an additional 4.5 hours, giving heat time to keep accumulating until output quality was affected.

Failed Assumptions

The failure followed directly from the system configuration and the conditions it was run under. Three assumptions made it both likely and late to detect.

Assumption 1
Cooling performance was assumed to be independent of water temperature
The manufacturer asked detailed questions about the operating environment and recommended a pre-cooler for Cambodia. That reinforced the view that the thermal constraints had been addressed. What neither the manufacturer nor I modelled was that the pre-cooler depended on city water temperature. In Europe this was typically around 20°C. In Cambodia it was closer to 30 to 35°C. Its effectiveness was assumed rather than validated under local conditions.
Assumption 2
Short runs were treated as representative of sustained operation
The two 7.5-hour Orange Liqueur runs showed the same temperature drift and passed quality control. I took that as a sign that the system was operating normally. When the same pattern appeared in the gin batch, I expected the outcome to be the same. That held true at 7.5 hours, but the gin batch ran for 12. Over that extra time, heat continued to build until flavour was affected.
Assumption 3
System scaling was treated as linear
The move from 300L to 2000L was treated as a scale increase, not a change in the system. In practice, the reservoir was sized for a 300L still and could not dissipate heat during sustained 2000L runs.

Why the Root Cause Was Confirmed Late

The system was not tested under the conditions where the failure occurred. Test runs were shorter, took place in lower ambient temperatures, and were focused on whether the recipe and process worked as expected, not on how the system behaved over longer runs.

Coolant reservoir temperature was not measured during production, and no thresholds were set to show when cooling performance was degrading. The pre-cooler was treated as a fixed solution, and its dependence on water temperature was not validated under local conditions.

Finding
The root cause could only be confirmed later, through controlled testing after shipment, because the conditions that caused the failure were not measured during production.