Application Corner: Measuring under vacuum, part 2

Last month, I mentioned three situations that one should carefully consider: operation under vacuum, measuring downward liquid flow and gravity flow. Let’s continue to discuss operating under vacuum.

Liquids exhibit a pressure loss and subsequent recovery as they flow through an obstruction in the piping, including many flowmeters. Given that operation is often at or near the vapor pressure, it is possible for the liquid to form bubbles when pressure is reduced, only to implode during pressure recovery. This cavitation can catastrophically damage flowmeters and other equipment, so flowmeter selection should be performed carefully. Selection, sizing and specification of control valves (and their internals) should likewise be carefully performed. 

In general, flowmeters should be selected and sized to have a low pressure drop. However, selecting, for example, a full-bore magnetic flowmeter with no piping obstruction would comply with this preference, but particular models may not be acceptable because their liners may not be suitable for vacuum service. Therefore, it is common to specify flowmeters and other instruments for vacuum service to operate between -1 to 6 bar — even when the process does not operate under pressure — to ensure that they have the mechanical integrity to operate in a full vacuum. 

Reverse flow is another phenomenon that can occur in some processes operating under vacuum. Normally, liquid will flow downward by gravity. However, when a vacuum is applied to a vessel, liquid may flow upward from below — like drinking through a straw. This can cause strange occurrences, such as when a flowmeter measures reverse flow while a process is not operating, and one would think that there should be no flow. Careful attention should be given to whether a flowmeter in a given application should measure reverse flow or not.

There are many other idiosyncrasies associated with measuring under vacuum, but measuring downward flow will be discussed next month.