Water hammer in vortex flowmeters?

Q: We have some vortex flowmeters in various sizes of pipes. They seem to have been installed correctly (upstream and downstream pipe runs are OK), but the ones on hot water service are highly inaccurate. We also have water hammering in these lines. What are your recommendations to resolve this problem?

Rahim Salamat
[email protected]

A1: Water hammer occurs when the flow of a water column is suddenly blocked and the momentum of that column is arrested. It can cause damage as it creates a pressure rise near the blockage. This can be caused by the sudden closure of a check valve, which can be corrected by replacing it with a slower-closing design. When water hammer occurs, you might hear bangs and the pipe might vibrate, but none of these occur because of using a vortex shedding flowmeter. In your case, what's likely happening is cavitation or flashing due to the high vapor pressure of the hot water.

Before addressing how to resolve your problem, let me first tell you how these flowmeters were discovered. A kid named Kármán Tódor was fishing in a spring in the beautiful Hungarian/Transylvanian countryside (today Romania). He noticed that in the water downstream of the rocks, swirls developed and traveled with equal distances between them. That distance remained the same, no matter the flow of water from the spring. So, little Tódor could measure the flow of water from the spring by counting the number of swirls passing by each minute.

Growing up, he became Theodore von Kármán, but this memory stayed with him, and helped him develop the JATO rocket motor as chief of NASA’s Jet Propulsion Laboratory, and later become the father of space travel and the best-known aerodynamic theoretician of the 20th century. To me, his experience proves that we should live with our eyes open when we see complex symptoms—like climate change or the misuse of AI to collect ransom by shutting down pipelines—and always try to figure out "why is this happening?," just as little Tódor did when seeing those vortices.

Now let's look at your vortex flowmeters. In liquid flow, the total energy is constant, so when the kinetic energy (velocity) rises, static energy drops (pressure). In vortex meters, the maximum velocity occurs in the region of the bluff body, where the cross-sectional area is the minimum (the vena contracta or VC). Cavitation occurs when VC pressure drops below the vapor pressure (P_V) because vaporization starts and vapor bubbles form at that point (Figure 1). This doesn't occur when flowing water is cold (green line) because P_V is low, but it does occur with hot water (red line) if P_V is higher than the pressure at the VC.

Downstream of the VC, the pressure recovers, and vapor bubbles violently condense. This collapse creates erosive micro-jets, noise (rumble, rattling or squeal), vibration and loss of measurement accuracy.

As vapor pressure rises, vaporization increases and causes the velocity at the VC to rise, until it reaches sonic velocity (Mach 1). As at that point, the flowing velocity can't increase further and "choking" occurs. The pressure corresponding to this condition is called the choking pressure (P_CH). Under these conditions, if the downstream pressure (P₂) drops further (or if the liquid temperature rises and P_V increases), a region of supersonic flow forms just downstream of the throat of the flowmeter. This condition is called "flashing." The resulting vapor flow with water droplets in it accelerates as it moves away from the throat, and as the area of the diverging section increases, this supersonic acceleration is terminated by a shock wave and drops back to a subsonic condition. This shock wave can be mistaken for water hammering. Whichever it is (cavitation, choking or shock waves), the measurement becomes useless.

As to what to do, you can:

1. Use another type flowmeter.
2. Measure the flow where the temperature is lower.
3. Measure the flow where the pressure is higher by moving the meter upstream.
4. Increase the size of the meter.
5. Select a meter design having a lower pressure recovery factor.

Béla Lipták
[email protected]

A2: I suspect you have cavitation in those meters. Water hammer is another thing. At the point of pressure drop, the water is flashing into steam, and where the pressure rises again, the stream collapses back into water accompanied by high velocity water jets. These are very damaging and meter tube failure is possible. There is no simple or cheap solution. The meter is in the wrong place. Relocation to a position with higher pressure or lower temperature, or replacement with a larger meter with less pressure drop would help. Any water meter where the water is too close to boiling is trouble.

Cullen Langford
[email protected]

Q: How much distance should separate 110 VAC alarm signal wiring (such as for a beacon) and a 4-20mA signal cable (class VI) in a tray with a metallic barrier? Same question for separation distance from a 24 VDC switch signal cable (Class I but with shield)? Finally, how much distance should separate the 110 VAC and 24 VDC signal wires described above in a tray but without a metallic barrier?

Mehdi Mnchri
[email protected]

A1: I assume that the “AC alarm” is a normally-open circuit, and therefore carries no current during normal conditions. If so, there are no minimum separation distances required, especially in a cable tray with a metallic barrier separating AC from DC signal lines.

The cable tray with metallic barriers is designed to allow AC power wiring and DC signal wiring to share a common cable tray. In such wiring, the DC cable is expected to be twisted-pair with a shield that's grounded at one end, or a multi-pair plenum cable with several shielded twisted-pair wires with an overall shield. All shields should be grounded only at one point (typically at the power supply) to prevent ground loops. The presence of the metallic barrier is sufficient to prevent AC interference with the DC signals. Twisted-pair cabling for DC signals is designed to allow a high degree of common-mode noise rejection by the receiving instrumentation.

Rest easy, the cable-tray specifications don't recommend separation distances because it doesn't matter. Noise rejection relies on shielding, the metallic barrier and using twisted-pair wiring. In your case, the AC alarm circuit normally carries no current anyway.

Richard H. Caro
CEO, CMC Associates,
ISA Life Fellow, Certified Automation
Professional (ISA),
[email protected]