Abstract
In countercurrent flow absorption columns and trickle beds, if either the liquid or gas flow is too high the column becomes flooded with the liquid.
Flooding is accompanied by a dramatic increase in pressure, resulting in inefficient operation and potential damage to equipment. To optimize
efficiency, however, it is desirable to operate as close to flooding as possible. This article discusses the feasibility of using acoustic emissions for
a non-intrusive, in situ method of monitoring column operations. Six piezoelectric microphones were placed on the outside of a packed column
and acoustic signals between 0 and 20,000 Hz were acquired over conditions ranging from non-flooded to flooded. The raw signals were then
processed to obtain their standard deviation and entropy. Standard deviation and entropy were both found to increase at the onset of flooding.
Entropy however, was more useful because it was not sensitive to the type of packing, air flow rate or water flow rate. Additionally, entropy began
to increase sooner than standard deviation, providing an early warning. The results from both entropy and standard deviation proved acoustic
detection can be used non-intrusively to monitor operations, allowing for better optimization and improved efficiency of absorption.
http://www.4shared.com/file/6779637...g_in_absorption_columns_and_trickle_beds.html
