Engineers enjoy a multitude of options for measuring flow. Selecting the right unit requires an understanding of the fluid, the piping layout and required accuracy and range (turndown). Often, the multi-port averaging Pitot tube is an appropriate and economic choice as the primary element for measuring liquids, gases or steam. This widely used flow-sensing device is an adaptation of the instrument invented by the French physicist Henri Pitot in 1732.
As an insertion-type flow meter, it is quickly and easily installed through a small hole drilled into a pipe or duct. As a mechanical device without electronics or moving parts, it doesn’t suffer from electronic drift or physical wear, making it a dependable, repeatable primary element requiring no periodic recalibration. Maintenance may be as simple as inspect and replace as needed.
These averaging Pitot tubes are available for horizontal and vertical pipes and ducts in sizes ranging from 1/2-in. up to 72-in. diameter or more. In contrast, in-line flow meters often are unavailable or significantly more costly in larger diameters. Moreover, insertion-type flow meters frequently have lower unrecoverable pressure losses than in-line meters and create minimal obstruction of the flow area. Wet-tap insertion meters are available for instances where it is impossible to shut down the process for installation or maintenance.
Units are offered in a wide variety of materials. Probes and heads commonly come in stainless steel and other steel alloys, polyvinyl chloride, brass and copper. As with any process hardware, material selection should take into account the nature of the fluid and operating conditions.
Averaging Pitot tubes are most suitable for flow measurements of clean liquids, gases or steam.
The devices are not the best choice, however, for applications such as liquids with high viscosity, gases with low velocity, or where harmonic vibrations in the probe cannot be avoided (don’t be alarmed here — manufacturers screen applications for harmonic vibration). Also, multiphase fluids, such as a gas with significant amounts of entrained liquid or a liquid with entrained gas are not good applications for this technology. Finally, dirty gas or liquid flows can be problematic with the sensing ports on the averaging Pitot tubes. Purging systems can be used to reduce or eliminate blockage in some of these applications.
The measurement mechanism
Pitot tubes produce differential pressure (DP) to determine flow rate. The multi-port averaging Pitot tube incorporates two isolated plenum chambers in a single probe. Numerous ports are precisely drilled into each chamber to sense both the higher flow rate that occurs near the center of a straight pipe and the lower rate that occurs at its wall. The probe is positioned so that the ports in one chamber are facing upstream and the ports in the other are facing downstream (see Figure 1).
The probe’s intrusion into the flow stream creates both dynamic and static pressures. The upstream ports sample and average the dynamic or impact pressure. The downstream ports sample and average the static pressure. A readout device measures the DP between the plenum chambers. The Pitot tube is a “square root device,” meaning the square root of the DP is proportional to the flow rate. Multiplying the velocity by the cross-sectional area of the pipe and a flow coefficient yields the volumetric flow rate.
An empirical flow coefficient, K , is essential to obtain an accurate flow rate from the DP as sensed by the Pitot tube. Among the reasons for this are percent blockage of the probe in the pipe and shedding vortices that cause a suctioning effect at the downstream static pressure ports, producing a larger measured DP than the theoretically predicted value.
The coefficients are determined as a result of water testing of different probe diameters and different pipe diameters via NIST-traceable standards. The manufacturer provides these coefficients for each multi-port averaging Pitot tube.
Flow coefficients for a given probe and pipe diameter fluctuate with changes in operating conditions, including fluid velocity, density and viscosity. By plotting the Reynolds Number versus the flow coefficient (Figure 2), the appropriate coefficient can be determined for any relevant operating condition (velocity profile). However, flow coefficients may vary by ±2% with the Reynolds Number; so, a flow-coefficient correction factor is used to obtain a Reynolds-corrected K . Manufacturers’ published curves of flow versus DP incorporate all applicable corrections. Pressure and temperature corrections are needed only for gases and vapors.
Advertised accuracies of probes vary from ±3/4% to ±1% of reading. Even though this difference may be important in some services, it is relatively insignificant in many process applications. Insufficient straight pipe runs upstream and downstream decrease accuracy. Straight-run recommendations vary with pipe diameter, upstream disturbances such as elbows, bends or valves, and the presence of straightening vanes.
The development of “smart multi-variable transmitters” has, to a degree, turned the discussion of accuracy differences among averaging Pitot tubes into a secondary consideration. These transmitters can incorporate all correction factors for the entire range of flow and so can calculate corrected flow rate at any Reynolds number. This allows all probe shapes to be accurate to their respective specification.
This transmitter can be mounted directly on the averaging Pitot tube. It can drastically reduce installation cost because it can do the job of conventional DP transmitters, pressure transmitters, temperature probes and flow totalizers. Piping and wiring costs also are significantly decreased.
In addition to accuracy, manufacturers publish specifications for repeatability (approximately 0.1%), uncertainty (approximately 1%), and turndown, that is, the ratio of maximum to minimum flow the probe can sense.
How Pitot tubes shape up
All manufacturers’ multi-port averaging Pitot tubes depend upon the measurement of DP, even though devices may have dissimilar mounting arrangements and accessories to integrate them with secondary elements (transmitters) and controllers.
Differences in averaging Pitot tubes frequently are evident in the cross-section of the probes, advertised turndowns and costs. To make a cost-effective selection, you need to know the reasons for those differences.
As the probe separates the fluid stream, vortices are created at a certain point along the side of the probe. This creates a vacuum on the downstream side of the probe. Depending upon the shape of the probe, the velocity of the flow media and, to some extent, its density, these vortices are shed from various points around the perimeter of the probe.
Because the multi-port averaging Pitot tube senses the static pressure on the downstream side of the probe, these shedding vortices can cause variations in the static pressure over the range of velocities the probe is likely to encounter. This could affect the accuracy of the differential pressure sensed by the probe if accurate K versus Reynolds number data are not available.
Manufacturers have introduced several shapes over the years with different advertised turndown ratios. (Note that only shapes symmetrical in the upstream/downstream direction with sensing ports facing directly upstream and downstream allow bi-directional measurements for flow that changes direction.) The turndown ratios range from 4:1 for conventional round probes to 10:1 for diamond and bullet shapes to 17:1 for the elliptical shape. Because there is little difference in the accuracy levels of various shapes, the price premium for a non-round probe essentially is buying greater turndown. For many process streams, a turndown of 4:1 far exceeds what actually will be encountered. For large turndowns, the small available DP at low flow must be considered.
It does cost more to produce probes in non-round shapes, with the premium dependent upon the particular material. If the application can employ a stainless steel probe, the advertised extra price for a non-round shape is modest, typically 8-10% more than a round probe. It is relatively uncomplicated to extrude non-round shapes from stainless steel, including the annular plenum chambers. However, if an application calls for an “exotic” material such as Monel, Hastelloy or Inconel, the price may balloon to 20-40% more than round tubes. While these materials are readily available in round-cross-section tubing, non-round shapes must be machined from bar stock to generate the outer profiles and drilled to create the plenum chambers.
Where intuition may fail
When it comes to flow meters, greater cost and greater flow metering do not necessarily go hand-in-hand. If you instinctively choose a more expensive option than an averaging Pitot tube without knowing precisely what you need, you may get more range and accuracy than you really require and an additional expense you certainly don’t want.
Dave Thomas is manager of pressure and flow products for Meriam Process Technologies, Cleveland, Ohio. E-mail him at [email protected].