“…..uncalibrated elbows will be subject to a tolerance (uncertainty) of about + 4%. With a calibrated elbow, the tolerance should be comparable to that for other types of differential pressure meters. With either calibrated or uncalibrated elbows a high degree of repeatability is attainable."

(Fluid Meters - their theory and application ASME 6th Edition 1971)

GENERAL INFORMATION
Unlike most other flow primaries each Measurell is individually calibrated. Our test facilities allow us to calibrate up to 66" diameter lines. The standard accuracy is + 1.5% of actual flow (with higher accuracies optional ). Repeatability is + 0.2%. Every Measurell is supplied with the calibration flow curve including flow constant ( C ). This “ C ” is also stamped on the side of the Measurell.

Measurells can be constructed with identical specifications as the line in which it will be installed. Non-standard construction to date includes titanium, 90/10 cupro nickel, fiberglass and cast iron. We have also provided special features such as coal tar and epoxy coatings, galvanizing and nickel plating. Note: calibration is performed after any internal modifications.

The Measurell is completely non-invasive and thus permanent pressure losses are minimal and can be easily calculated from readily available tables of permanent pressure loss through elbows. Generally, permanent pressure loss is about 1/3 of pressure loss using an orifice plate with comparable beta ratio and the Measurell signal is up to 4 times greater.( See Fig. 5 )

Much higher differential signals are produced (up to 4X) than those of venturi tubes or orifice plates with comparable beta ratio. This is caused by the effect of the centrifugal force of the fluid as it passes through the elbow. This is especially significant with reduced throat configurations.

INTRODUCTION
The use of a common elbow fitting as a primary flow device is not a recent development. Since early this century, it has been recognized that the differential pressure produced between the inner and outer radii of a 90º fitting was proportional to flow within the line. However, for many years the lack of applied research prevented realization of the many advantages of using elbows as flow elements.

A simple, effective modification to previously conventional tap placement, a unique discharge section design, and individual shop calibration have made the Measurell, i.e. elbow flow element, a non-invasive primary device available as a replacement for the familiar orifice plates and venturiis. All types of line material, weld specifications and end connects are customer specified. Standard accuracy of + 1.5%, excellent repeatability and very low permanent pressure drop are gaining the Measurells a widespread use in the monitoring of gases, liquids and slurries including corrosive medias and other problem fluids.

Operating Principles
The Measurell operates on the momentum principle of requiring a force to turn a moving fluid in another direction. For a fluid flowing through an elbow fitting, this force is provided by the continuous pressure generated on the inner and outer surfaces of the bend (The magnitude of this force is dependent upon fluid density and flow rate). The continuous pressure rises above line static pressure on the outside of the bend, and drops below line static pressure on the inner radius. This pressure variation is gradual along the direction of flow (Fig.1) and also depends on flow rate and fluid density.

FIG. 1: PRESSURE PROFILE

By placing pressure ports or taps in the elbow and sensing the difference in pressure between the inner and outer radii, the flow rate can be determined. Past efforts to use elbow flow meters have usually tried using taps inserted at the 45º position along the bend. However, problems arise since pressure distribution is sensitive to details of tap finishing, upstream surface roughness, internal weldment and manufacturing variances.

Research using flow visualization techniques, has shown that stagnant or separated regions of flow can occur, primarily on the inside bend (Fig.2).

FIG. 2: FLOW SEPARATION

The size of this area is influenced by all the previously mentioned factors. A pressure tap placed within such an area, results in variable readings and generally unsatisfactory flow-metering performance.

Our investigation and development resulted in special tap placements to ensure reliable and repeatable differential pressure signals and thus flow-metering. Measurels are individually calibrated and if necessary, include three diameter lengths of approach piping incorporating anti-swirl vanes. Although standard accuracy is + 1.5%, calibration accuracy of + 0.3% can be achieved for special circumstance with ASME published repeatability of + 1.2%.

Applications for low flow rates, or low-density fluids such as natural gas or steam normally require elbow assemblies (Fig.3) incorporating a size reduction at the elbow, followed by a gently expanded section for excellent pressure recovery.

FIG. 3 REDUCED CONFIGURATION

PERFORMANCE

Signal & Pressure Loss
The overall performance of Measurells vs. conventional venturi tubes and orifice plates is summarized in (Fig. 4).

FIG. 4: SIGNAL GENERATION

From theoretical considerations, an elbow with no restriction develops a pressure difference of almost two velocity heads while an orifice or venturi of course develops no signal at all with no restriction. As the throat (elbow diameter) is reduced below line size, the combined effects of constriction and centrifugal force produce a much larger read-out signal than for either venturi meters or orifice plates at the same restriction ratio. This effect is of great advantage for low flow measurement, as well as for gas, steam, or airflow determinations.

Gradual expansion back to full line size to minimize friction loss results in extremely low permanent pressure loss. A comparison of pressure loss for Measurels, orifices and venturi tubes at different restriction ratios is indicated in (Fig. 5).

FIG. 5: PERMANENT PRESSURE LOSS

Although Measurells in the reduced configuration have a permanent pressure drop it is only a fraction of the pressure loss of an orifice plate measuring the same flow. Venturi tubes also have low frictional loss but are more susceptible than Measurells to flow irregularities and as a result must be protected to a greater degree against inlet swirl by straightening vanes, grids or long lengths of entry pipe. Orifice plates suffer similar disadvantages as the venturi but as mentioned previously, also have a much higher friction loss. The Measurell features of high signal generation and very low permanent pressure loss are particularly important in natural gas pipelines or steam flow applications, where orifice plate losses result in large energy consumption.

Effect of Approach Piping
Specific measurement of the effect of upstream disturbances on the discharge coefficient has been accomplished by testing with a variety of simulated valves and pumps placed immediately upstream of a Measurell. Numerous test runs are summarized in (Fig. 6).

FIG. 6: EFFECT OF UPSTREAM DISTRUBANCE ON DISCHARGE COEFFIEIENT

A gate valve type obstruction (half shut, valve stem in the plane of the elbow) had less than 0.1% effect on discharge coefficient when located only 1.5 diameters upstream. The largest disturbance effect was due to a special in line plate, which produced a swirl velocity approximately equal to the
average pipe velocity. (A condition similar to that at the discharge of a pump or compressor).
The calibration flow constant changed approximately 0.5% in this case. These tests indicate that the Measurell can be designed into systems having minimum approach lengths and still be relied upon to give accurate flow monitoring when other types of flow elements cannot be employed.

Discharge Coefficient
Although the Measurell works on a momentum principle, rather than energy conversion such as anorifice plate or venturi, the differential pressure produced is also of the square root type, making this primary element compatible with most process instrumentation. As with venturiis and orifice plates, the discharge coefficient varies with some fluid conditions. However, if the Reynolds number (an index of these effects) exceeds 100,000 the discharge coefficient remains constant over the complete flow range. (Fig. 7)

FIG 7: ELBOW REDUCTION EFFECT ON DISCHARGE COEFFICIENT

Most flow metering applications have Reynolds numbers well in excess of this range. All Measurels are characterized by their own unique flow constant as a result of individual calibration.

Special Taps for Contaminated/Corrosive Liquids, Gases, Slurries
During the development of the Measurell it was discovered that while tap placement is important, tap size had virtually no effect on discharge coefficients. Extensive series of test indicated tap diameters as large as 30% of line diameter continued to give accurate and reliable differential signals.

This has great practical advantages for use in contaminated fluids, or corrosive gases and liquids in addition to slurries and sewage that are prone to plugging smaller pressure ports present in venturiis or orifice systems. Simple periodic blowout or purging systems can also be devised with excellent results. The Measurell is available with isolating diaphragm mechanisms inserted flush with the inside surface of the elbow, completely isolating corrosive working fluids from a conventional instrument air read-out system.

AREAS OF APPLICATION

To date, thousands of Measurells have been installed in a variety of applications over a wide range of service conditions and line specifications. The advantages of using a Measurell vary with the installation. Measurells with isolating diaphragms have found ready use for sewage and slurry measurement. Furthermore, the very low-pressure drop and excellent repeatability enable economical service for gas pipeline compressor surge control.

Cast Measurells are extensively utilized in HVAC systems as an economical and easily installed means of system balancing. In many cases the fact that the Measurell substitutes for an existing elbow means low costs, while the relatively short approach lengths have resulted in close coupling with pumps or valves and elimination of long lengths of approach piping. To date construction materials have included Cupro-Nickel, stainless steel (cast and forged) cast iron as well as the catalogue standards of PVC, cast bronze and forged steel. The latter has included Sch. 40 to Sch. 120 with any type of pipe connection and weld specifications.

The variety of service conditions and line specification applied to Measurels are continually expanding.

HOW TO SELECT MEASURELLS AND FLO-PROBES
General: For a given line size, flow rate and type of fluid, a simple formula will predict the differential pressure signal produced. For example, for water flows, h = (Q/C)². Where Q = USGPM, C = Measurell or Flo-Probe Constant (tables available on request) and h = readout signal, inches water column.The forged steel forms of Measurell can be made with reduced size elbows at the bend, fitted with appropriate discharge and  approach sections. If the differential signal based on full line size is too low, simply select another Measurell constant for a smaller line size and repeat the above calculation until a suitable differential signal is reached.

ALTHOUGH THE EXAMPLES BELOW USE MEASURELLS, THE SAME FORMULAS ARE USED TO SELECT FLO-PROBES FOR ELBOW OR STRAIGHT LINE INSERTION - WHEN THE APPROPRIATE "C" IS UTILIZED .

For Liquids: For any process liquid, the selection is as indicated above, except the appropriate fluid density for specific gravity must be known. The formula for differential pressure is

h = S(Q/C)², where ;

S = specific gravity of the fluid at line conditions
Q = USGPM,
h = inches water column,
C = the appropriate Measurell or Flo-Probe constant.

For Gases:

Air: The selection procedure is as indicated above, except a constant factor to account for conversion of units and air density must be made using the following formula.

h = (0.896)(d)(Q/C)², where ;

Q = CFM at line conditions,
D = air density at line conditions, pounds/cu ft
h = inches water column,
C = the appropriate Measurell or Flo-Probe constant.

Steam: The selection procedure is as above using the following formula

h = (v)(W/C)²/4015, where;
W = pounds/hour steam flow,
v = specific volume of steam at line conditions, ft²/pound
h = inches water column,
C = the appropriate Measurell or Flo-Probe constant

Natural Gas: The selection procedure is as above using the following formula

h = 32.6(MZ)(T/p)(MMSCFD/C)², where;

MMSCFD = gas flow rate, million standard cu. ft/day
T = line temperature, degrees Rankin,
p = line pressure, p.s.i.a.,
M =gas molecular wt.,
Z = compressibility factor
h = inches water column
C = the appropriate Measurell or Flo-Probe constant.