A major pharmaceutical plant contacted Tracerco regarding an operational problem on a separation tower operating on a batch process. The tower consisted of a single bed of structured packing and approximately 5 hours into the batch a high-pressure drop continually developed, tripping the tower out. This resulted in extended batch times and therefore a reduction in the total production capacity of the plant. The company requested a TRU-SCAN®/Tru-Grid™ Scan to examine the bed section for mal-distribution, which the client thought was the most likely reason for their problems. When the Tracerco Engineers arrived on site the tower was at the end of a batch and this gave them the opportunity to perform a virtual, “ off line” scan. The results (see Fig. 1 - black line), confirmed that the bed was correctly located and had a similar density to the engineering specification.
"On-line" scans were then performed during the next batch and it was agreed that a four “grid line” scan would provide the most useful information for a mal-distribution study. However, during the first scan the high pressure drop condition was reached and the tower tripped. The scan continued however, and it showed that the tower was flooding from the bottom of the vessel to an elevation of around 1000mm, actually inside the bed itself. This was very different to what was expected as it clearly showed that the tower was flooding from its base rather than the bottom of the bed, thus indicating that the problem was in the exit line from the tower base rather than the bed itself. Further scans were performed before the tower was re-started and then when it was back on-line. All of the scans indicated a level build-up was developing from the tower base up to elevations within the bed.
After the measurements were performed, our client undertook an initial visual inspection of the vessel and it revealed no restriction within the line and confirmed the control valve was operating satisfactorily and after the inspection the problem disappeared. Three weeks after the inspection the problem reappeared and the client had no idea why it was occurring again. Further analysis of plant data was performed and this revealed that the batches, which had been produced before the problem reappeared, had been allowed to cool. The client concluded that this resulted in the formation of a solids build-up in the exit pipe of the tower, leading to a blockage and causing the flooding.
The trace heating duty on the line was increased by a small percentage and to our knowledge there are no longer any high-pressure drop trip outs on the batch reactor, leading to a much greater production throughput.
Overall, as a result of solving the problem, the batch time has been decreased from 22 to 14 hours together.