Fundamentals of Liquid Flow Measurement

This course covers the basic methods used to measure the flow rate of liquids in commercial, municipal and industrial applications. It is intended to be an introductory course and somewhat broad in discussing a range of current methods of measurement such as differential pressure flow meters, vortex shedding flow meters, ultrasonic flow meters, magnetic flow meters, Coriolis flow meters, positive displacement meters, and turbine flow meters; as well as open-channel flow meters, such as weirs and flumes. In addition, we will discuss applications using these meters, the advantages and disadvantages of each; and conditions that can influence flow meter accuracy, repeatability and reliability. All of the flow meters that we will be discussing in this course are commercially available, calibrated instruments that use various methods to indirectly measure the velocity of a flowing liquid within a closed system such as pipe, or open channel flow, such as in a culvert or trough. Flow Characteristics in Filled Pipe Systems One cannot discuss flow measurement, select or specify the proper flow meter without an understanding of the characteristics of liquid flow within a filled pipe system, and its related terminology. The primary objective of a stationary (fixed-position) flow measurement instrument is to determine the velocity of the liquid that is passing by it. Once the velocity of the liquid and the inside diameter (I.D.) or cross-sectional “free area” of the pipe or conduit through which it is flowing are known, one can easily calculate the volumetric flow rate of the liquid. For example, if the velocity of water flowing through a 1-inch I.D. hose is measured to be 30 feet per minute, then we can determine the volumetric flow rate of the water that is passing through the hose in gallons-per-minute using the following calculation:

Volumetric Flow Rate (GPM) = cross-sectional area x fluid velocity where: GPM = gallons per minute diameter (d) = 1 inch one gallon of water = 231 cubic inches of water velocity (v) = 60 feet/min. x 12 inches/foot = 720 inches per minute

the equation becomes: GPM = 720 inches/min. x (πr2) x 1 gallon/231 cubic inches = 720 x (3.14 x 0.52) x 0.004 = 2.26 gallons per minute

The volumetric rate of water flowing out of the hose will be 2.26 gallons per minute if the measured flow velocity of the water is 720 feet per minute. Friction Loss in Pipe Friction loss of fluid flow within a filled pipe refers to the friction that occurs between the flowing liquid that contacts the stationary inside surfaces of the pipe as it flows.

Figure No. 1 – Liquid Flow Profile in Pipe The most accurate method of estimating frictional head loss in steady pipe flow due to friction is generally considered to be the Darcy-Weisbach Equation (developed by Henry Darcy and later refined by Julius Weisbach in 1845) used in conjunction with the Moody Diagram (developed by L.F. Moody in 1944) . The Darcy-Weisbach Equation is:

where: hf = friction head Note: the term “head” is an engineering unit of hydraulic pressure used commonly in discussing liquid statics and dynamics. “Head” is the pressure equivalent produced by a stationary column of water, typically measured in inches or feet of head. One foot of head is equal to 0.4335 psi.

f = coefficient of friction (obtained from the Moody Chart below) L = length of pipe V = liquid velocity D = inside pipe diameter (I.D.) g = gravity (32 ft/sec2)

Another formula often used to calculate head (pressure) loss due to friction is the Hazen-Williams Equation that was developed in

where:

Pd = pressure drop (psi per foot)

Q = volumetric flow rate (gallons per minute)

C = friction loss coefficient (the greater the C factor, the smoother the pipe)

d = inside diameter of pipe (inches)

Chart No. 1 – Moody Chart

Examples of the relative roughness of the interior wall of some of the more common pipe materials are as follows:

• Concrete 0.001 to 0.01 • Cast Iron 0.00085 • Commercial Steel 0.00015 • Drawn Copper Tubing 0.0000005

(Source: Pump Handbook, 2nd Ed., 1986, Igor Karassik, McGraw-Hill Book Co.)

Liquid Turbulence in Pipes and Reynolds Number The accuracy of many of the flow meters that we will be discussing in this course is directly affected by turbulence of the liquid. Therefore, it is important that we determine whether or not turbulence will occur at the point of measurement in the pipe system. A basis for determining at what point liquid turbulence occurs in a pipe and why, resulted in the establishment of The Reynolds Number (Re), named after Osborne Reynolds, who experimented with flow variations approximately one hundred years ago. Dr. Reynolds determined that if the liquid velocity or pipe diameter is relatively small and the viscosity of the liquid is relatively large, the Re will be small and the flow will tend to be laminar (non-turbulent). Increasing the pipe diameter or liquid velocity, or decreasing the liquid’s viscosity, will result in a higher Re. Reynolds concluded that, for a given flow rate in a pipe, regardless of which of the three particular parameters he varied, as long as the Re value was less than approximately 2300, the flow in the pipe remained laminar. Above that value, turbulence would consistently occur. The formula for calculating the Reynolds Number (Re) for flow within a pipe is:

Re = u d / ν

where:

u = kinematic viscosity of the liquid

v = liquid velocity in feet per minute

d = diameter in feet Therefore, when selecting the proper meter for an application, the Reynolds number must be calculated in order to determine if a turbulent flow condition will exist at the point of measurement; and that the meter is suitable for that condition.

Chart No.2 – Pipe Friction Identified on Moody Chart

Revisiting the Moody Chart in Chart No. 2 above, please note the relationship between friction factor, Reynolds Number and relative roughness and at what point the flow condition transitions from laminar flow to turbulent flow. Flow Straighteners A flow straightener, such as the one appearing in Figure No.1, is, as its name implies: a device that straightens the flow pattern within a pipe. They introduce further restriction in flow; and a greater potential for fouling (accumulation of particulate or debris at the tube inlets). However, they provide an effective means of eliminating turbulence when there is insufficient space available ahead of the flow meter location for a sufficiently long run of straight pipe. In other words, you may have no choice but to specify such a device if there are space constraints. When the pump system producing the flow that is being measured is initially sized, or the flow straighteners are added to an existing pipe system, the additional system pressure drop produced by the flow straighteners will need to be considered.

Figure No. 1 – Fluid Flow Straightener

To calculate the pressure loss introduced by the pipe itself, one would use the Fanning Equation.

The Fanning Equation is expressed as:

Where:

D = inside diameter of pipe in inches

f = Fanning friction factor - which is a function of the Reynolds number

g = gravity (32.2 ft/s2)

L = length of the pipe section in feet

V = average velocity of the liquid in feet per second

= pressure loss due to friction effects in pounds per square inch

= density in pounds per cubic foot

Evolution of Liquid Flow Measurement In 1783, Swiss physicist Daniel Bernoulli introduced the concept of the conservation of energy for fluid flows. He determined that an increase in the velocity of a flowing fluid will increase its kinetic energy while at the same time decreasing its static energy. That explains why a flow restriction within a pipe causes an increase in the liquid’s flow velocity at the restriction while also causing a decrease in the static pressure of the flowing liquid downstream of the restriction. If a an orifice plate (a plate with a small diameter hole in the center) is installed in a pipe line that has water flowing through it, the velocity of the liquid as it passes through the orifice will increase substantially while the pressure downstream of the orifice will be

measurably less than the pressure on the upstream side of the orifice. In essence, the restriction in flow caused by the orifice plate produces a backpressure on the upstream side of the orifice. Since Bernoulli’s time, there have been other methods of indirect measurement of liquid flow developed that are based on other technology besides the pressure velocity relation. The development of those other technologies has been driven by technological advances in electronics, metallurgy, and the unique problems encountered when attempting to measure flow. There is a very wide range of liquid flow meters that are commercially available. Each has distinct advantages and disadvantages, and is selected based upon the instrument that best fits the particular application. Flow instruments tend to be rather expensive; particularly when a high level of accuracy and reliability are needed. It can become quite costly to misapply a flow meter and have to replace it when it either fails to provide an accurate measurement; or it fails to operate properly a short time after being installed. Initial consideration must be given to the characteristics of the fluid to be measured, including it’s:

• pressure • temperature • density (or specific gravity) • conductivity • viscosity, and • vapor pressure

Note: For most of our examples in this course we will be using water as the measured liquid. In addition to the characteristics of the fluid being measured, we must consider the environment that the rest of the flow meter will be exposed to, including such things as heat, vibration, moisture, and corrosive chemicals. All of these considerations will be discussed in more detail later. However, it is important to consider them before reviewing the various types of liquid flow instrumentation and its associated hardware, as it will make it a bit easier to understand why there are so many different styles of flow meters being manufactured and on the market today. Differential Pressure Flow Meters The oldest and perhaps one of the most reliable forms of flow measurement is based on differential pressure. Bernoulli’s equation

V = k(h/d)0.5

where:

V = velocity of the liquid k = discharge coefficient of the element h = pressure differential d = pipe inside diameter

allows us to calculate the velocity of a liquid passing through a pipe by measuring its pressure drop across a known flow restriction. The variable k is influenced by the liquid’s Reynolds number and by the "beta ratio," which is the ratio between the bore diameter of the flow restriction and the inside diameter of the pipe.

Figure No. 2 – Standard Orifice Plate

Figure No. 3 – Differential Pressure Measured Across Orifice Plate

An advantage of a differential pressure style flow meter is that the flow element typically has no moving parts, which makes it a highly reliable instrument. That is not to say that a device such as the orifice plate appearing in Figure No. 2 is completely reliable. Periodically, at least once or twice a year, the orifice plate should be removed and examined to insure that nothing has fouled the opening and that there is no significant wear that has increased the size of the opening. The plates are installed between two welded orifice pipe flanges which are designed so that the bolts of the flange can be loosened and the orifice plate removed. Most such plates are constructed of stainless steel and are good for many years of reliable service. For further information regarding the established engineering standards for orifice meters refer to ANSI/ASME Standard MFC-14M-2003, “Measurement of Fluid Flow Using Small Bore Precision Orifice Meters” and International Standards Organization (ISO) Standard ISO 5167-1:2003. For over 50 years orifice plates were considered to be the most reliable style of flow meter for measuring water and steam flow in steam generation plants. However, a downside to using these differential style pressure instruments is that they each introduced a restriction, and therefore induced a pressure drop, in the pipe system - which meant that the pumps producing the water flow had to work harder, or be sized larger, to handle that additional pressure. The harder working or larger pumps would consume more electrical energy which over time increased the cost of operating the equipment, not to mention the utility rate increases over time. This prompted instrument manufacturers to begin to develop flow meters that would not introduce a significant pressure drop in the piping system. Further prompting the development of other fluid metering technology was the need, in some applications, to measure flow rate in liquids that might contain particulate or debris that would foul or plug an orifice, such as in certain parts of a wastewater treatment facility. What was desired in those types of applications was a “non-intrusive” style of flow meter. An instrument that could measure flow rate without having a flow element protrude into the flows stream that particulate or debris could collect on or foul. Another style of differential pressure flow element is the venture nozzle. A venture nozzle produces a pressure drop similar to an orifice plate and has the advantages over an orifice plate of a flared or tapered constriction in a pipe that is less likely to become fouled by debris or particulate.

Figure No. 4 – Venturi Nozzle and Diagram

As with orifice plates, there are no moving parts to the flow element, giving it good reliability and virtually no wear, allowing for many years of service. A disadvantage shared by any type of differential flow element is the limitation in its accuracy during low flow conditions. A flow meter’s ability to accurately measure flow over a range is commonly referred to as the meter’s rangeability or its turn-down ratio.

Turn-Down Ratio (rangeability) = Max. Flow Rate / Min. Flow Rate The typical turn-down ratio (i.e. rangeability) for an orifice plate is approximately 4:1. There are many free programs readily available on the Internet for calculating the size and pressure drop across an orifice plate or venturi nozzle, such as http://www.pipeflowcalculations.com/orifice/index.htm and http://www.efunda.com/formulae/fluids/calc_orifice_flowmeter.cfm Target Flow Meters Target flow meters are essentially meters that have a “target” object that is centered in the liquid flow stream. The rate of flow is determined by the force imposed on the target by the flowing fluid striking it.

Figure No. 5 – Target Meter and Diagram

This style of meter produces a pressure drop (restriction in flow); but much less pressure drop than an orifice plate or Venturi. Turbine Style Flow Meters Another relatively common style of flow meter is the axial vane turbine-style flow meter that was reportedly first developed by Reinhard Woltman in 1790. The turbine flow meter employs a spinning turbine blade or impellor in the fluid stream that spins faster or slower depending upon the velocity of the fluid impacting it and causing its rotation. The rotational speed of the impellor is detected by installing a small magnet in the tip of an impellor blade and then detecting it each time that it passes by a small solid-state device referred to by meter manufacturers as a “pickups” that uses magnetic reluctance to create an electrical pulse that is detected and converted from a pulse frequency into fluid velocity. There are a very large number of design variations in this style of flow meter that includes axial turbines, pin wheels, and paddle wheels. Axial turbines tend to perform best when measuring clean, conditioned, steady flows of liquids that have a low kinematic viscosity. Available sizes range from ¼-inch to over 30 inches I.D. A small diameter meter with a signal transmitter attached is shown in Figure No. 6.

Figure No. 6 – Axial Turbine Flow Meter w/ Signal Transmitter In addition to the axial style turbine meters there are insertion-style turbine meters, such as that appearing in Figure No. 7. The insertion-style turbine flow meters consist of a small axial rotor that is mounted on a retractable stem that is inserted through a shut-off valve and the pipe wall so that the impellor is located near or at the center of the pipe through with the liquid is flowing. The purpose of the shut-off valve is to allow quick retraction and removal of the meter from the pipe without having to drain the entire pipe. The shut-off valve is quickly closed off once the meter has been removed, limiting the

amount of liquid that escapes during meter removal. This type of meter installation is commonly referred to as a “hot-tap” or “wet-tap” configuration. Design considerations of the rotor component of a turbine flow meter include its blade shape and pitch, the blade count, balance, rigidity, stress, drag and inertia. Bearings must allow low-friction rotation of the impellor. The breakaway torque required to initially induce rotation establishes the lowest velocity that the meter is capable of measuring. Most turbine meters employ tungsten carbide, fluid bearings, or ceramic journal and thrust bearings. Some of the smaller meters use jeweled pivot bearings that will allow sensitive measurement even at very low flow conditions.

Figure No. 7 – Insertion-style Turbine Meter The delicate nature of the impellor or wheel of turbine flow meters makes them unsuitable for liquids that contain a lot of particulate or solids. They can foul or impact the impellor or wheel in such a way as to potentially damage or destroy it. Turbine meters that have been installed in steam lines have been known to be damaged by a slug of condensate that is rapidly pushed through a steam line and then impacting the turbine. Turbine flow meters have very good rangeability or turn-down (as high as 20:1). These meters provide very good reliability if properly applied in clean, low viscosity liquids, and are installed in a straight run of pipe or with upstream flow straighteners (so that there is relatively low turbulence upstream and downstream of the point of measurement). Turbulence can cause erratic and inaccurate measurement, and the turbine wheel and bearings will wear out more rapidly when continuously subjected to a turbulent flow pattern.

The axial and insertion style flow meters introduce a pressure drop in the pipe system. However, it is considerably lower than the pressure drop introduced by a differential style instrument such as an orifice plate or Venturi. A variation of the turbine meter that was mentioned earlier is the paddle wheel style meter such as that appearing in Figure No. 8. Note the small paddles at the base of this insertion style meter.

Figure No. 8 – Insertion Style Paddle Meter and Diagram

As with other types of insertion style flow meters, paddle meters can be used in pipe diameters as small as one inch and as large as 30 inches depending upon whether or not the wheel can reach the middle of the flow pattern in the pipe. Paddle wheel meters are available with full-scale flow ranging from 5 to 99,000 GPM. They are a bit more forgiving when it comes to particulate in the liquid stream being measured. However, as with a turbine meter, the paddle wheels are most reliable when used to measure the flow rate of clean, low turbulence and low viscosity liquids. Vortex Shedding Flow Meters Another style of insertion flow meter is the vortex shedding flow meter. This meter has the benefits of an insertion style meter such as low pressure drop and accessibility; and, like the turbine flow meter, exhibits very good rangeability. The principle advantage of the vortex shedding meter over the turbine meter is its ability to function well in fluids having some particulate, and there are virtually no moving parts, which tend to give it a higher degree of reliability than that of the turbine meter. Vortex shedding meters, as their name implies, utilize a blunt-shaped object in the fluid stream that creates eddies (turbulence) immediately behind it.

Figure No. 9 – Vortex Shedding Meter and Diagram These eddy currents, which vary with the velocity of the liquid, cause a slight oscillation in the wedge (or other style of blunt object sometimes referred to by manufacturers as the

“bluff body”) that is detected by sensitive electronic circuitry within the bluff body utilizing an ultrasonic, electronic, or fiber optic sensor that monitors the changes in the vortex pattern and converts the impulses into a industrial standard output signal such as a square wave pulse, or a 4-20 mAdc analog signal. The average velocity of the fluid is proportional to the frequency of the vortex shedding and the width of the bluff body. The typical style of vortex shedding flow meter appearing on the left in Figure No. 9 has a transmitter mounted on it that produces the industry standard output signal. Theodor Von Karman is credited with developing early bluff body designs for vortex metering in 1912. His work in that area led to awareness that sharp-edged bluff bodies provide greater consistency of vortex shedding. Early vortex shedding meters developed a rather poor reputation, largely due to their inability to distinguish (filter out) pipe vibration from that produced by the eddy currents. Advanced designs, utilizing better electronic filtering of the vortex shedding sensor output resolved that problem and eventually these meters made a strong comeback in industrial and municipal flow measurement applications. Positive Displacement Meters Another type of flow measurement meter is the positive displacement meter. These meters measure the volumetric flow of fluid passing through them by repeatedly counting the filling and discharging of known fixed volumes (batches) of the liquid. A typical positive displacement meter contains a chamber that obstructs the flow. Within the chamber, a rotating or reciprocating mechanical unit is configured so that it creates fixed-volume discrete volumes (batches) of the passing liquid. The volume of the fluid that passes through the chamber can be obtained by counting the number of passing batches, or equivalently the number of cycles of the rotating or reciprocating mechanical device. The volumetric flow rate is derived from the cycles of the device.

Figure No. 10 – Positive Displacement Meter and Diagram

This type of meter is used in small water meters, such as that appearing in Figure No.10, and more frequently in measuring liquids that have relatively high viscosities, such as fuel oil. They can accept small particulate (suspended solids) in the measured liquid. However, as with other meter types that have components directly in the path of the fluid stream, the particulate produces more erosion wear due to the particulate impacting the components and in some cases increasing the amount of friction between moving components, such as meshed gears. If the various types of meters that have been discussed thus far are not particularly effective when used with fluids containing suspended solids or debris, what types of meters can be used in those applications? The solution is to use a “non-intrusive” type of meter, which has virtually no objects directly in the path of the flowing liquid that is being measured to impede its flow. Coriolis Meters The first such “non-intrusive meter” that we will discuss is called a Coriolis style meter. I chose this style of meter first because it may not be considered by all to be a true non-intrusive flow meter due to its unique method of sensing flow rate. However, other than curvature of the pipe through with the sensed fluid travels there are no other obstructions on the flow path. Therefore, I would categorize it as a relatively non-intrusive instrument. The Coriolis meter, introduced in 1977 by a firm called Micro Motion Inc., is a flow measurement instrument that indirectly measures the mass flow rate and density of a fluid flowing through a helical tube. The Coriolis meter (refer to Figure No. 11 below) functions by inducing a vibration of the helical tube through which the fluid passes. The vibration, though not entirely circular in motion, provides a “rotating reference frame” that gives rise to the “Coriolis Effect”, after which the meter has been named. Sensors monitor and analyze changes in frequency, phase shift, and amplitude of the vibrating flow tubes to determine the mass flow rate and density of the fluid. The Coriolis Effect is the same phenomenon that causes the winds of a tornado or cyclone to always swirl in a counter-clockwise direction in the Northern Hemisphere, and swirl in a clockwise direction in the Southern Hemisphere. This condition is directly influenced by the Earth’s rotation.

Figure No. 11 – Coriolis Style Flow Meter

These meters are quite accurate and as mentioned earlier they are capable of measuring variations in the density of the measured liquid, unlike all of the previously discussed meter types. A potential disadvantage of these types of meters is their size, which can be substantially larger than the pipe diameter that the measured fluid is passing through, and the need to rigidly support such an instrument. Those issues aside, these instruments, although relatively expensive, provide a very accurate measurement of variable density fluids. Magnetic Flow Meters Used largely in the pulp and paper and wastewater industries is a type of non-intrusive flow meter referred to as a magnetic flow meter. As its name suggests, a magnetic field is used to determine flow velocity. Faraday's Law states that “the voltage induced across any conductor as it moves at right angles through a magnetic field is proportional to the velocity of that conductor”. This phenomenon is referred to as “electromagnetic induction”. In the case of a magmeter, the aqueous fluid flowing through the meter’s magnetic field serves as the conductor. A magnetic field is applied to the metering tube, which results in a potential difference proportional to the velocity of the liquid flowing perpendicular to the magnetic field’s flux lines, as illustrated in Figure No. 12 below.

Figure No. 12 – Magnet Flow Meter Diagram

In order to function properly, a magnetic flow meter requires a conductive fluid (such as water) and an electrically insulated pipe surface, such as a section of plastic-lined steel tube. Conductivity of the liquid in a typical magnetic flow meter application must be maintained above about 5 micro-Siemen per centimeter. Some magmeter designs require more than 50 micro-Siemen per centimeter in order to function properly.

Figure No. 13 – Magnetic Flow Meter

In-line magnetic flow meters, also referred to as “magmeters”, such as that appearing in Figure No. 13, can be as accurate as 0.5% of the rate of flow. Insertion style magmeters typically have an accuracy of 1 to 2% of the rate of flow. There are no protruding objects in the flow path, making this type of meter very useful in the measurement of fluids that contain suspended particulate or large solids. Such meters are available in sizes ranging up to 30 inches in diameter. Straight pipe run requirements are relatively short, as these meters are less affected by turbulence than most, making magnetic flow meters attractive where limited straight run is available. Clamp-On Ultrasonic Flow Meters A “non-intrusive” type of flow meter design to be discussed in this course is referred to by several names including clamp-on, ultrasonic, Doppler and transit time flow meter. Each of these descriptions partially captures its method of measurement. The device is very portable; and does get clamped onto the exterior surface of a pipe. It then uses an ultrasonic signal to penetrate the pipe wall, pass through the flowing liquid within until it strikes a suspended particle or bubble, is reflected back off of the suspended particulate or entrained bubbles moving with the flow of the liquid; and echo back to a receiving device using the Doppler principle - much like sonar. The transit time (which is the distance the bubble or particle travels from the time that it is detected by the previous pulse echo) is determined and converted into a liquid flow velocity. Doppler flow meters will normally emit a continuous frequency. Dual piezo-ceramic crystal transducers are clamped onto a pipe a fixed position apart from one another (refer to Figure No. 14 below). One of those crystals (the one further “upstream”) acts as the emitter; and the other as the receiver. If there are small particles or bubbles in the liquid they act as reflectors and reflect back a frequency that is different than the initially transmitted frequency, depending on their average velocity in the pipe. This condition is referred to as the “Doppler shift”, which is the same principle used in radar and sonar systems.

Figure No. 14 – Clamp-on Flow Meter The obvious advantages of this type of instrument are that it is completely portable and can be clamped on a pipe wherever there is adequate clearance to do so. Just as a magmeter requires the measured flowing liquid to be conductive, the fluid measured by the clamp on, transit time meter must have entrained air bubbles (typically caused by turbulence) or suspended solids from which the Doppler signal can bounce back. This is the opposite requirement of some of the earlier meters that we considered, such as the orifice plate, Venturi nozzle, vortex shedding flow meter, and the turbine meters, which all require a straight run of pipe upstream and downstream to minimize the turbulence. A clamp-on flow meter does not work well without the turbulence or some suspended solids. High Temperature and Other Environmental Considerations Of ultimate importance is knowing the hazard classification of the location in which the meter will be installed. If the area in which the meter will be installed is rated as a Class 1, then it is vital that the meter being selected and specified, much like any other instrumentation within the hazardous area, must be of an intrinsically safe design. This usually means providing an instrument with an explosion rated termination head, and intrinsically safe (current limiting) barriers between the flow transmitter’s electronics and any electrical power and signal sources. Class 1 areas have vapor or dust, or the potential for their presence that can result in a catastrophic and potentially deadly explosion if the vapor or dust cloud is ignited by an electrical spark. It is important, for protection of the meter itself, when selecting and specifying a flow meter to consider the maximum temperature that the flow element, transducer, and transmitter (electronics) will be subjected to while installed. Referring back to the illustrations of some of the flow meters we have discussed, you will see examples of the electronics being an integral part of the meter installation. In other words, the meter is attached directly to the flow element. In that type of meter design you have good thermal conductivity between the flow element and the electronics due to the properties of the

metal components connecting them together (refer to Figure No. 6 as an example). Some, if not most, manufacturers have the option of allowing you to specify a meter that has a thermal isolation option (a thermal insulator between the metallic portion of the flow element and the metallic portion of the electronics housing) for high temperature applications; or what is referred to as a “remote head”. With remote head meters the electronics are remotely located in a cooler location so that the electronics are not subjected to high temperatures. Most solid-state electronic components will fail when subjected for any length of time to ambient temperatures in excess of 180 degrees Fahrenheit. This is because in order for many of the components to function properly, they must be able to dissipate heat. If the air around them is as high, or higher, than the temperature that should be dissipated, then the air no longer acts as a suitable heat sink and the electronic component will literally cook itself and soon fail. Whenever the meter to be used is not position-sensitive (check with the manufacturer) they should be mounted in a horizontal orientation so that the electronics are to the side of a hot pipe, rather than directly above it. Therefore, it is important to take note of the maximum ambient temperature that the instrument can accommodate and take that into account when selecting a suitable meter configuration and location. I recall being asked years ago to investigate a metering problem involving multiple flow meters that had been installed on medium pressure steam lines in a power plant. All of the meters had failed shortly after installation and start-up of the plant. The electronics were cooked because the transducer/termination heads of the meters were connected directly to the flow elements and only a couple of inches directly above the steam pipes. With no thermal isolation the electronics had been cooked. Specifications for the instrument listed a maximum operating temperature of 140 degrees F. The destroyed instruments were replaced with the same meters having remote heads located on a nearby wall and the problem was solved. The humidity (can condensation occur?) and corrosiveness of the atmosphere where the meter is to be located must also be given consideration. The housing in which the meter’s electronics are contained must be of a weather-resistant design if it will be installed where the instrument can be sprayed with water or condensation can occur. Most industrial grade meters are inherently designed to be weather and corrosion resistant. Accessibility is an important consideration, given the fact that any of the flow meters described should be examined for proper operation, and their calibration verified, on an annual or semi-annual basis. Calibration service organizations that are used for this purpose should have fully equipped calibration laboratories and offer NIST [National Institute of Standards and Technology] traceable flow calibration for flow meters. There is no other reliable way, short of setting up your own certified lab, of reliably confirming the accuracy and performance of the flow meters. This requires temporary removal of the instrument. Therefore, another consideration must be whether or not the pipe system can be isolated in such a manner as to allow the instrument to be removed without draining the entire pipe system. This can be done by installing manually operated

isolation valves up and/or down stream of the meter using valves that allow unrestricted flow when fully open, such as a full-port ball valves. Vibration transmitted through the pipe from a pump or other source is of less concern than the previously mentioned conditions. However, severe vibration can eventually result in metal fatigue and wear that should be avoided when possible. Additional information on flow meter technology for filled (pipe) systems can be found at these links: http://www.omega.com/techref/flowcontrol.html; http://www.isa.org/ Closed System Flow Meter Selection Determining the best flow meter to use in a particular application will depend on a variety of factors and conditions. The comparison table below has been created to make it a bit easier to determine the best meter type for various conditions.

METER TYPE High Viscosity Particulate Turbulance Large Flow Turn-DownMagmeter Very Good Very Good Very Good GoodTurbine Poor Poor Fair Very GoodVortex Shedding Poor Fair Poor Very GoodPos. Displacement Very Good Fair Good FairOrifice Plate Poor Poor Poor PoorVenturi Nozzle Poor Fair Poor PoorClamp-on Very Good Very Good Very Good Fair

Table No. 1 – Simple Closed System Flow Meter Comparison The chart is not to be used for any purpose other than as a simple comparison of some of the more common meter types and how they compare to other styles of flow meter for the purpose of this course. Substantially more data will be needed in order to make a proper meter selection, as noted at the end of this course material. A more detailed flow meter selection guide can be found at either of the following links: http://www.apengr.com/Flowmeter%20Selection/Flowmeter%20Selection%20Guide.xls and http://www.tagteam.com/tagteam/client/detail.asp?dbid=539&siteid=836048&dataid=95930 Open Channel Flow Measurement Up to this point in the course we have discussed the types of flow meters that may be used in closed pipe systems. We will now discuss the instruments used to measure open-channel flow, such as in an open culvert or trough. The means by which open channel flow is measured is commonly referred to as the hydraulic structures method.

Two of the most common types of flow measurement utilize either a flume or a notched weir. Both devices work on a similar principal, which involves flowing water through a calibrated restriction, such as a narrow channel or notch, where the free area geometry (height and width of the opening) are known and indirectly determining the rate of flow by measuring the level of the water at the point where it is passing through that narrow channel or notch restriction. A weir (pronounced "wēr") typically refers to a dam placed across a stream to either raise or regulate the water level upstream; or on a smaller scale as a means of measuring the rate of open channel flow that is moving downstream. The discharge of the stream flows through a v-shaped opening in the weir, whose lowest level is referred to as the “sill height”. The most common weir constructions include the v-notch weir, the rectangular notch weir and the broad-crested weir. Refer to Figure No. 15.

Figure No. 15 – V-Notch and Rectangular Notch Weirs

The formula for calculating flow rate of water passing through a v-notched weir is:

Q = 8/15 cd b [2 g]1/2 tan(θ/2) h5/2 (2)

The flow rate measurement in a rectangular weir is:

Q = 2/3 cd b [2 g]1/2 h3/2 (1)

For the broad-crested weir the flow rate is expressed as:

Q = cd h2 b [2 g (h1 - h2)]1/2

where (in all three equations):

Q = volumetric flow rate

h = head on the weir

b = width of the weir

g = gravity

cd= discharge constant for the weir (which must be determined by calibration tests)

Non-contact ultrasonic level transmitters are typically positioned directly above the flow cascading over the weir notch without any direct contact with the flow. Digital flow computers simplify and automate the calculations needed to determine the flow rate utilizing the equations above. Palmer-Bowlus and Parshall Flumes Palmer-Bowlus and Parshall flumes have a narrow shaped open channel flow section which may be installed in a trench, canal, ditch, or lateral to measure water flow rate (refer to Figure No. 16 below). The Parshall flume is a particular type of Venturi style flume and was named after its principal developer, Ralph L. Parshall. Distinct advantages of a Venturi style flume such as the Parshall flume include 1.) a larger constriction point - allowing it to be used to measure greater flow rates; and 2.) there is less risk of debris in the flowing water fouling the point of measurement and thereby rendering it inaccurate. These more reliable, non-contact level transmitters have replaced the older, mechanical float level meters that could potentially become fouled or fail due to wear over time.

Figure No. 16 – Parshall Flume

Palmer-Bowlus flumes, which have been widely used since 1937, constrict flow vertically and horizontally, and are generally installed in U-shaped channels; whereas Parshall flumes constrict primarily horizontally, and are typically installed in square or rectangular channels.

Figure No. 17 – Palmer-Bowlus Flume

During flume or weir installation, they must be positioned vertically in order to accurately determine the water surface variation. Further, if they are installed too high, the upstream channel may tend to overflow. If they are installed too low, they can end up submerged, rendering them useless as a measuring device. Note: For additional information on selecting and specifying weirs and flumes, refer to ASTM Standard D5640 – “Standard Guide for Selection of Weirs and Flumes for Open Channel Flow Measurement of Water”. Summary As you can see, there are a considerable number of conditions that must be considered when fully specifying the proper flow meter for a given measurement application. At a minimum, you must be aware of the:

• fact that it is open channel or closed system (filled pipe) • maximum and minimum design flow rates (range) that you will be measuring • pH of the liquid • density of the liquid • viscosity of the liquid • temperature of the liquid • amount of suspended solids in the liquid • system design pressure • available length of straight pipe run • flow turbulence at the metering location (calculated) • meter accuracy requirement • turn-down (rangeability) of the meter • pressure drop introduced by the meter

• accessibility of the meter location • level of mechanically-induced vibration in the pipe and meter • ambient (atmospheric) condition at the meter location (wet, corrosive, etc..) • signal compatibility with any remote display, control, or recording instruments • calibration requirements • service requirements • replacement parts availability • warranty provisions