Pipeline & Gas Journal
October issue 1999:

Technological Solution

A New Flow Meter For Compressor Fuel Gas Measurement

by James E. Gallagher, P.E., and Michael P. Saunders,
Savant Measurement Corporation USA, Houston

Effective natural gas transportation requires the efficient use of compressor technology throughout the system. Obviously, modern compressors are technologically superior to their counterparts of 30 years ago; however, one aspect of compressor technology that has failed to keep pace with the ever-increasing efficiency of engines and controls is the fuel gas measurement device itself. This device is critical to the compressor’s operation and as such provides an area for significant performance improvement within the system.

It is vital that a compressor fuel gas meter offer the following:

• Wide rangeability—to allow for efficient operation throughout the engine’s flow ranges.
• High accuracy—to ensure satisfactory accountability for fuel gas usage.
• Low maintenance—to minimize ongoing costs.
• Minimal calibration—to reduce operational costs.
• Compact design—to allow for smaller and more lightweight installations.

Current Design Challenges
Typical compressor fuel gas measurement installations require significant accuracy and repeatability as a means of assuring operational efficiency and accountability. Maintaining compressor efficiency at all times requires that fuel gas is utilized as efficiently as possible to maintain the engine’s ability to operate within its most effective range. At the same time, accurate accountability for gas usage is critical, not only to ensure contract compliance, but to determine the ongoing efficiency of the compressor itself.

Designers of new compressor facilities and engineers designing modifications to existing installations are faced with numerous flow measurement choices, including:

• Orifice meters.
• Turbine meters.
• Ultrasonic meters.
• Thermal dispersion flow meters.
• Coriolis meters.

With fuel gas measurement for compressors, many of these and other technologies fall short, particularly in the following areas:

• High maintenance.
• High capital cost.
• Unpredictable accuracy.
• Uncertain repeatability.
• Significant operational costs.

New Technology Solution
With the above challenges in mind, a new technology has been developed that, while satisfying the need for optimal performance, does not compromise the application’s requirement for wide rangeability and low maintenance.

This solution is based on a new "head" class flowmeter concept that combines the strengths of a pitot probe and isolating flow conditioner technology to achieve a performance of +/-.75 percent under field piping configurations for Newtonian fluids. Pitot devices measure the velocity of the flowing stream in a circular conduit. Gaussian integration techniques have been incorporated into some product designs to eliminate the sensitivity to piping-induced installation effects. However, field research has shown that such integration techniques’ sensitivities are limited. The flowmeter discussed here does not employ Gaussian integration techniques.

To determine the invention’s validity, experiments were conducted in the natural gas environment. Analysis of the experiments’ results validates the concept and performance of the approach incorporated into the resulting product, named CheckMeter.

Developing A Better Design
Several pitot designs were evaluated during the experiments. The pitot devices employed consisted of both point and averaging pitot designs. All the pitot devices were innovative in their design approach.

Under field piping configurations for Newtonian fluids, it was found that the CheckMeter achieved a flowmeter performance of +/- 0.75 percent. To ensure that the CheckMeter was measuring true flow, an isolating flow conditioner was incorporated into its design. This was essential to establishing an average flow velocity, since pitot probe technology is particularly sensitive to upstream flow disturbances.

The Role Of Flow Conditioning
Flow disturbances in steady-state mass flowrate conditions are deviations of the inlet flow profile, swirl, or turbulence levels from the fully developed flow.

An inferential meter’s sensitivity from "disturbed" flow to "fully developed" flow is dependent on the disturbance, the flowmetering technology, the specific meter design, and the flow field generated by the disturbance (velocity profile, swirl, turbulence intensity, shear stress, etc.).
In general, upstream piping elements may be grouped in either of the following categories:

• Those that distort the mean velocity profile but produce little swirl.
• Those that both distort and generate bulk swirl.

In the industrial environment, multiple piping configurations are assembled in series, generating complex problems for standards-writing organizations and flow-metering engineers. The challenge is to minimize the difference between the actual or "real" flow conditions and the "fully developed" flow conditions in a pipe, in order to maintain a minimum error associated with the selected metering device’s performance. One of the standard error minimization methods is to install a flow conditioner in combination with straight lengths of pipe to "isolate" the meter from upstream piping disturbances.
As a result, the isolating flow conditioner’s role is to lower uncertainty levels associated with "non-ideal" flow conditions.
To truly isolate flowmeters, the flow conditioner should achieve the following design objectives:

• Low permanent pressure loss (low head ratio).
• Low fouling rate.
• Rigorous mechanical design.
• Moderate cost of construction.
• Elimination of swirl (less than 2 degrees).
• Independence of tap sensing location (for orifice meters).
• Pseudo-fully developed flow for both short and long straight lengths of pipe.

When the swirl angle is less than or equal to two degrees as conventionally measured using pitot tube devices, swirl is regarded as substantially eliminated.

A new breed of isolating flow conditioner produces pseudo-fully developed flow conditions for both short and long piping configurations. It has also demonstrated insensitivity to orifice tap sensing location, confirming the presence of pseudo-fully developed flow.

The design has the benefit of being the most recent evolution in the design of flow conditioners. As such, one would expect to have significant performance improvements over existing flow conditioners. The CheckMeter design does exhibit this improved performance as a result of a considerable parametric study for sensitivity (profile device design, settling chamber length, etc.), as well as additional insights gained during the last four years.

The isolating flow conditioner eliminates swirl (less than 2 degrees of swirl) and provides an axisymmetric velocity profile (+/-5 percent between parallel chords) upstream of the pitot device.

Experimental Results
Experiments were conducted at GRI’s Meter Research Facility under the auspices of Southwest Research Institute. Indepen-dent research was conducted extensively on 100-mm (4-inch) and 200-mm (8-inch) meters with both single-point pitot and averaging pitot devices over pipe Reynolds numbers from approximately 500,000 to over 8 million.

To determine the CheckMeter’s optimum design, experiments were conducted in natural gas using proprietary pitot flowmeter designs in combination with an isolating flow conditioner under the following fluid dynamic conditions:

• Fully developed flow.
• Asymmetric, non-swirling flow.
• Swirling flow.

Fully developed flow was established with the use of an isolating flow conditioner, in approximately 40 diameters (40D) of straight pipe prior to the test section.

Non-symmetric, non-swirling flow was established with the use of an isolating flow conditioner, a minimum of 20 diameters (20D) of straight pipe, and a tee mounted in the same plane prior to the test section.

Swirling flow was established with the use of an isolating flow conditioner, using a minimum of 20 diameters (20D) of straight pipe and a 90-degree elbow and tee out of plane prior to the test section. This combination has been known to generate swirl angles of 15 to 20 degrees.

The point pitot probe experiments indicated excellent agreement with the predicted fully developed flow profile of Nikuradse’s Log-Law equation. The preceding results were obtained in fully developed flow conditions in a 200mm artefact over a pipe Reynolds number from approximately 600,000 to 8 million.

The averaging pitot probe experiments indicated excellent agreement with the predicted fully developed profile under "perturbed" flow conditions. For conciseness, the non-symmetric, non-swirling flow scenario will be presented.

The non-symmetric, non-swirling flow case is a robust test for the combination isolating flow conditioner and pitot device. This scenario determines the ability to minimize the sensitivity to non-symmetrical velocity profiles as they approach the pitot device.

The 12 o’clock position is perpendicular to the plane of non-symmetry. The 9 o’clock position is in the plane of non-symmetry. The difference between these two positions represents the worst case performance scenario for the flowmeter.

The preceding averaging pitot probe data was obtained using a 100mm artefact over a pipe Reynolds Number of 500,000 to 7 million.
As indicated in the assembled artefact drawing, the thermowell is located upstream of the profile device. Experiments were performed to determine any sensitivity to the location of the temperature sensing. These experiments indicated that the temperature deviation between the designated location downstream of the pitot device was well within an acceptable tolerance (+/-0.05 degrees C).

The CheckMeter®
The CheckMeter is the result of this careful experimentation period. Its pitot probe measures the differential between the stagnation pressure and the static pressure. This differential pressure is proportional to the square of the velocity. The secondary instrumentation required consists of smart transmitters (differential pressure, static pressure, and temperature) and a flow computer.
As has been stated, advantages of the pitot over other flowmeters are its low cost, rigid mechanical design, and the application of proven highly accurate secondary instrumentation. The disadvantage of the pitot is its ability to measure flowrate only at a single point, its sensitivity to the angle of impact, and its susceptibility to clogging in dirty fluids.

To overcome these disadvantages, the design combines an innovative isolating flow conditioner and a uniquely designed pitot probe. This combination eliminates sensitivity to a single point velocity and the angle of impact.

The unique isolating flow ensures a symmetrical, non-swirling velocity profile approaching the pitot probe. The criteria for this profile is less than 2 degrees of swirl and parallel chordal velocities that differ by +/-5.0 percent.

Since the flowmeter incorporates an innovative pitot design, one can safely assume the relative direction of flow or angle of impact is accurate. In addition, the design minimizes susceptibility to clogging in dirty fluids.

The point velocity, or velocity at any radial position, is calculated from equations which can be used with a high degree of confidence, provided that the Mach number is less than 0.25 for compressible fluids.

Conclusions

• This new "head" class flowmeter concept combines the strengths of the pitot device and isolating flow conditioner technology.
• The flowmeter is an inexpensive, high-accuracy device for compressor, process, and check-metering applications in relatively clean Newtonian fluids.
• The required instrumentation consists of traditional smart transmitters (dP, Pf and Tf) and a flow computer.
• The device can utilize any of the three calibration concepts as demonstrated in the experimental pattern - artefact, central facility, or in situ calibration.
• The mechanical design provides a holistic concept - isolating flow conditioner, pitot probe, upstream and downstream meter runs, and a sensing thermowell. The length of the flowmeter is 10 nominal pipe diameters (10D). The holistic concept does not require additional straight runs of pipe upstream or downstream of the device.
• The CheckMeter’s operating range is independent of operating pressure - very low to very high (5 to 2,000 psig or more).
• For natural gas applications, the design velocity range is from approximately 15 to 300 fps resulting in a turndown of 20:1. For most applications, a smaller flowmeter will satisfy operating requirements, providing the user with additional capital cost reductions.
• The flowmeter has demonstrated a performance of +/-0.75 percent under the following fluid dynamic conditions:
• Fully developed flow,
• Asymmetric, non-swirling flow,
• Swirling flow. P&GJ