• Measurement Microphones
• Sound Level Meters
• Measurement of Sound Intensity
• Calibration Sources
• Ear Simulators
• Compatibility Issues
• New Measurement Technique
• Principle Level Measurement
• Level Sensor Selection
• Boiling & Cryogenic Fluids
• Sludge, Foam, & Molten Metals
• Flow Measurement Orientation
• The Flow Pioneers
• Flow Sensor Selection
• Accuracy vs. Repeatability

Technical Guide - Sound Measurements

Measuring Instrumentation

 Although Rayleigh completed the first edition of his 'The Theory of sound' without the benefit of electroacoustical instrumentation (only in the second edition did he mention 'Bell's Telephone' with no thought of using this device for measurement), modern acoustical measurements rely invariably on electroacoustics. Anyone intending to measure any of the descriptors described above would want to turn the acoustical signal into an electrical signal (if not a digital signal) as early in the measurement process as possible, and would use some of the electroacoustical instrumentation shown in Figure 1.

Measurement Microphones

The transducer which converts the acoustical signal to an electrical one is usually a condenser microphone. Given the requirement for a 120 dB dynamic range (12 orders of magnitude) this is one of the few suitable transduction mechanisms. The condenser microphone operates on the principle that the capacitance of two electrically charged plates varies with the separation between them. The charge may be generated by an external polarising voltage, or by the inclusion of an electret material into one of the plates. One of the plates is an extremely light diaphragm which moves in response to acoustic pressure variations and the resulting change in capacitance then produces the output voltage.

Because the introduction of the microphone into an acoustic field will change that field there are three different sensitivities defined for a measurement microphone:

• The pressure sensitivity is defined as the voltage produced for a given sound pressure applied uniformly over the diaphragm of the microphone.

• The free-field sensitivity is defined as the voltage produced for a given sound pressure in a free progressive sound field that existed before the introduction of the microphone.

• The diffuse-field sensitivity is defined as the voltage produced for a given sound pressure in a diffuse field (i.e. one where the sound is equally likely to arrive from any direction) that existed before the introduction of the microphone.

In order for condenser microphones to have sufficient sensitivity they are typically 1 cm in diameter. Because the wavelength of sound is comparable with this at the higher audible frequencies there can be a significant difference (up to 15 dB) between these different sensitivities due to diffraction effects.

Few real acoustic fields will fall into one of these ideal cases but the procedural standards define which should be used for a particular measurement and the uncertainties have to allow for the possible differences.

Sound Level Meters

The most common form of instrumentation to use the signal from the microphone is a sound level meterThere are two types of sound level meter: An exponential-averaging meter is designed for the measurement of continuous sounds. It measures the root mean square value of the sound pressure. The signal from a microphone is fed to amplifier and weighting circuits which limit the frequency content to a prescribed range. The signal is then squared and passed through a single-pole low-pass filter providing a prescribed exponential time constant. This signal is then displayed in decibels as a sound pressure level.

An integrating-averaging sound level meter is designed to measure sounds from specific events. It detects, frequency weights, and squares the signal in a similar way to an exponential-averaging meter. Then the squared signal is integrated. The logarithm of the integrated signal has the logarithm of time subtracted from it. The result is then displayed in decibels as an 'equivalent continuous sound pressure level' (Leq).

Instruments for the Measurement of Sound Intensity

Sound intensity is a measure of the magnitude and direction of the flow of sound energy. The unit is Wm-2. Although acousticians have attempted to measure sound intensity since the 1930s, the first reliable measurements of sound intensity did not occur until the late 1970s when advances in digital signal processing and the availability of digital instrumentation allowed commercial instruments to be produced. Most modern measurements of sound intensity are made using the simultaneous measurement of sound pressure with two closely spaced microphones. The sound intensity can be calculated from the mean pressure, the pressure difference (including phase information) and a knowledge of the specific acoustic impedance of the air. Sound intensity can be expressed in decibels relative to 10^-12 Wm^-2 when it is known as sound intensity level.

The measurement of sound intensity as a way of determining the sound power of a source offers distinct advantages over traditional methods which would require the source to be put in a specialised environment (either anechoic, hemi-anechoic, or reverberant room). But the instrumentation and procedural standards are complicated and the method is not yet in widespread use.

Calibration Sources

Several devices exist which allow the routine calibration of measuring instrumentation.

A sound calibrator contains a small, stable sound source which can be coupled to the microphone of a measuring instrument. The simplest device will generate a sound pressure at a single frequency and level, more complicated devices will generate a variety of levels and frequencies. They allow the calibration of an entire measurement system to be checked.

A reference sound source generates a stable broad-band noise at a known sound power and allows systems that determine the sound power of machines to be checked.

Neither of these devices are absolute. They both require calibration themselves with microphones of known sensitivity.

Ear Simulators

Ear simulators are standardised devices used in the calibration and characterisation of audiometers, earphones, hearing aids, telecommunications and audio equipment. There are a wide variety of designs of ear simulators standardised by different organisations. They range in complexity from a simple hollow cylinder with a microphone diaphragm occupying one end to a series of interconnected cavities, resistances and controlled leaks designed to simulate not only the acoustic impedance of the human ear but also designed to account for the way a user might apply an earphone to the ear.

Ear simulators for audiometric earphones allow audiometers to be calibrated in terms of 'hearing level' i.e. a decibel scale where 0 dB is intended to represent the threshold of hearing for 'otologically normal people'. This was achieved by testing the hearing thresholds of large groups of subjects and relating those thresholds to the sound pressures developed in the ear simulators. Consequently each audiometric ear simulator has a set of Reference Equivalent Threshold Sound Pressure Levels (RETSPLs) which allows the output of an earphone to be related to hearing level.

An ear simulator that simulates the bone conduction path is known as a mechanical coupler and a set of Reference Equivalent Threshold Force Levels (RETFLs) allow the output of bone vibrators to also be related to hearing level.

Compatibility Issues

Measurement microphones, sound calibrators and sound level meters are normally produced to standard patterns, in principle enabling a device from one manufacturer to be used with an associated device from a second manufacturer. However caution should be exercised with such practice. First, devices of a specific type are likely to have a different performance when used with associated devices from different manufacturers, e.g. a sound calibrator may produce different sound pressure levels on two generically similar microphones, but of different types. Second, the control of production tolerances may be different between manufacturers, resulting in subtle incompatibilities that can seriously affect performance or even cause damage to the devices, e.g. a microphone may fit well in its designated sound calibrator, but may be excessively tight in a sound calibrator from another manufacturer.

For the reasons given above, it is recommended that instrumentation from different manufacturers should only be combined where a manufacturer has stated specifically that this is acceptable. Ideally the manufacturer should also state the validity of any calib

New measurement technique will help in fight against cancer

A new product from NPL will ensure an exciting new screening technique can be relied upon by hospitals to identify early signs of cancer.

NPL's Dr Pete Tomlins with a point spread phantom  A point-spread phantom The technique, Optical Coherence Tomography (OCT), is an increasingly popular method for looking beneath the surface of certain materials, notably human tissue. It is higher resolution and much quicker than techniques such as MRI or ultrasound, with no ionising radiation, making it ideal for detecting changes in tissue structure which can indicate the early stages of cancer.

However creating such images requires high precision, and any inaccuracy can lead to incorrect assumptions about cell disruption. This can mean missing opportunities for early, potentially life-saving treatment.

A new NPL product, called a 'point-spread phantom', will eliminate the risk of such errors. The phantoms are translucent cylinders of resin containing specially arranged particles designed to reflect light in a very specific way. By viewing the phantom with an OCT machine and analysing the image with NPL software, users can be certain the machine is producing accurate images, which they can rely on for important medical decisions.

These 'phantoms' will also allow manufacturers of OCT technology to meet the necessary standards to guarantee to hospitals that their machines are sufficiently accurate. This will help speed the route to market of products using this important new technology, and assure hospitals of their ongoing reliability.

Michelson Diagnostics is the first UK company to use NPL's phantoms to validate the accuracy of their machines. CEO Jon Holmes said:

"We developed breakthrough technology for imaging living tissue and for detecting diseases, but we needed to validate our performance claims, to provide customers with greater confidence in them. NPL's phantoms and analysis have enabled us to validate our claims beyond doubt, thereby demonstrating the superiority of our scanners and giving us the edge over our competitors. We expect that this validation will give OCT technology the backing it needs to become standard in hospitals around the world, and thereby make an important progression in the battle against cancer".

NPL recently completed laboratory tests and is now running trials with companies before bringing the product to market. Anyone interested in more information or trialling the new technology should contact Pete Tomlins

Read more technical information in:
P H Tomlins, 'Point-Spread Function Phantoms for Optical Coherence Tomography', NPL Report OP 2 (2009).

The Principles of Level Measurement

With the wide variety of approaches to level measurement and as many as 163 suppliers offering one or more types of level-measuring instrument, identifying the right one for your application can be very difficult. In recent years, technologies that capitalized on microprocessor developments have stood out from the pack. For example, the tried-and-true technique of measuring the head of a liquid has gained new life thanks to “smart” differential pressure (DP) transmitters. Today’s local level-measuring instruments can include diagnostics as well as configuration and process data that can be communicated over a network to remote monitoring and control instrumentation. One model even provides local PID control. Some of the most commonly used liquid-level measurement methods are:

• RF capacitance

• Conductance (conductivity)

• Hydrostatic head/tank gauging

• Radar

• Ultrasonic

Before you can decide which one is right for your application, however, you need to understand how each works and the theory behind it. (Each method has its own abbreviations, so you may find the sidebar, “Abbreviations for Common Flow Sensing Terminology,”, a useful reference during the discussions that follow.)

RF Capacitance

RF (radio frequency) technology uses the electrical characteristics of a capacitor, in several different configurations, for level measurement. Commonly referred to as RF capacitance or simply RF, the method is suited for detecting the level of liquids, slurries, granulars, or interfaces contained in a vessel. Designs are available for measuring process level at a specific point, at multiple points, or continuously over the entire vessel height. Radio frequencies for all types range from 30 kHz to 1 MHz.

Capacitance Measurement Theory. All RF level systems make use of enhancements of the same capacitance-measuring technique, and the same basic theory underlies them all. An electrical capacitance (the ability to store an electrical charge) exists between two conductors separated by a distance, d, as shown in Figure 1. The first conductor can be the vessel wall (plate 1), and the second can be a measurement probe or electrode (plate 2). The two conductors have an effective area, A, normal to each other. Between the conductors is an insulating medium—the nonconducting material involved in the level measurement.

The amount of capacitance here is determined not only by the spacing and area of the conductors, but also by the electrical characteristic (relative dielectric constant, K) of the insulating material. The value of K affects the charge storage capacity of the system: The higher the K, the more charge it can build up. Dry air has a K of 1.0. Liquids and solids have considerably higher values, as shown in Table 1.

The capacitance for the basic capacitor arrangement shown in Figure 1 can be computed from the equation:

C = E (K A/d) (1)

where:

C = capacitance in picofarads (pF)

E = a constant known as the absolute permittivity of free space

K = relative dielectric constant of the insulating material

A = effective area of the conductors

d = distance between the conductors

To apply this formula to a level-measuring system, you must assume that the process material is insulating, which, of course, is not always true. A bare, conductive, sensing electrode (probe) is inserted down into a tank (see Figure 2,) to act as one conductor of the capacitor. The metal wall of the tank acts as the other. If the tank is nonmetallic, a conductive ground reference must be inserted into the tank to act as the other capacitor conductor.

With the tank empty, the insulating medium between the two conductors is air. With the tank full, the insulating material is the process liquid or solid. As the level rises in the tank to start covering the probe, some of the insulating effect from air changes into that from the process material, producing a change in capacitance between the sensing probe and ground. This capacitance is meas ured to provide a direct, linear meas urement of tank level.

As shown in Figure 2, the electrode sensor, or probe, connects directly to an RF level transmitter, which is mounted outside the tank. In one design, with the probe mounted vertically, the system can be used for both continuous level measurement and simultaneous multipoint level control. Alternatively, for point level measurement, one or more probes can be installed horizontally through the side of the tank; Figure 2 shows this type being used as a high-level alarm. Photo 1 shows a typical probe assembly with an enlarged view of the microprocessor-based transmitter that fits in the housing; in use, its digital indicator faces up. Trans mission of the level-measurement signal can take several forms, as can the in strument that receives the signal at either a local or a remote location.

Referring to Figure 2, the transmitter output is 4–20 mA DC plus optional HART Protocol for remote diagnostics, range change, dry calibration, and so on. The instrument receiving the signal can be a distributed control system (DCS), a programmable logic controller (PLC), a Pentium III PC, or a strip or circular chart recorder.

When the process material is conductive, the sensing probe is covered with an insulating sheath such as Teflon or Kynar. The insulated probe acts as one plate of the capacitor, and the conductive process material acts as the other. The latter, being conductive, connects electrically to the grounded metallic tank. The insulating medium or dielectric for this application is the probe’s sheath. As the level of conductive process material changes, a proportional change in capacitance occurs. Note that this measurement is unaffected by changes in the temperature or exact composition of the process material.

RF Impedance or RF Admittance. When another electrical characteristic, impe dance, enters the picture, the result is further refinements in RF level measurement. Offering improved reliability and a wider range of uses, these variations of the basic RF system are called RF admittance or RF impedance. In RF or AC circuits, impe dance, Z, is defined as the total opposition to current flow:

Z = R + 1/ j 2 p f C (2)

where:

R = resistance in ohms

j = square root of minus 1 (–1)

p = the constant 3.1416

f = measurement frequency (radio frequency for RF measurement)

C = capacitance in picofarads

An RF impedance level-sensing instrument measures this total impedance rather than just the capacitance. Some level-meas uring systems are referred to as RF admittance types. Admittance, A, is defined as a measure of how readily RF or AC current will flow in a circuit and is therefore the reciprocal of impedance (A = 1/Z). Thus, there is no basic difference between the RF impedance and RF admittance as a level-measurement technology.

In some cases, the process material tends to build up a coating on the level-sensing probe. In such cases, which are not uncommon in level applications, a significant meas urement error can occur because the instrument measures extra capacitance and resistance from the coating buildup. As a result, the sensor reports a higher, and incorrect, level instead of the actual tank level.

Note that the equation for impedance includes resistance, R. The RF impedance method can be provided with specific circuitry capable of measuring the resistance and capacitance components from the coating and the capacitive component due to the actual process material level. The circuitry is designed to solve a mathematical relationship electronically, thereby producing a 4–20 mA current output that is proportional only to the actual level of the proc ess material. It is virtually unaffected by any buildup of coating on the sensing probe, enabling an RF system to continue functioning reliably and accurately.

Conductance

The conductance method of liquid level measurement is based on the electrical conductance of the measured material, which is usually a liquid that can conduct a current with a low-voltage source (normally <20 V). Hence the method is also referred to as a conductivity system. Conductance is a relatively low-cost, simple method to detect and control level in a vessel.

One common way to set up an electrical circuit is to use a dual-tip probe that eliminates the need for grounding a metal tank. Such probes are generally used for point level detection, and the detected point can be the interface between a conductive and nonconductive liquid.

Figure 3 shows an arrangement with two dual-tip probes that detect maximum and minimum levels. When the level reaches the upper probe, a switch closes to start the discharge pump; when the level reaches the lower probe, the switch opens to stop the pump.

Hydrostatic Head

One of the oldest and most common methods of measuring liquid level is to measure the pressure exerted by a column (or head) of liquid in the vessel. The basic relationships are:

P = mHd

or:

H = mP/d (3)

where, in consistent units:

P = pressure

m = a constant

H = head

d = density

P is commonly expressed in pounds per square inch; H, in feet; and d, in pounds per cubic feet; but any combination of units can be used, so long as the m factor is suitably adjusted.

The density of a liquid varies with temperature. For the highest precision in level measurement, the density must therefore be compensated for or expressed with relation to the actual temperature of the measured liquid. This is the case with hydrostatic tank gauging (HTG) described below.

For decades, DP-type instruments—long before the DP cell—were used to measure liquid level. Orifice meters, originally designed to measure differential pressure across an orifice in a pipeline, readily adapted to level measurement. Today’s smart DP transmitters adapt equally well to level measurements and use the same basic principles as their precursors. With open vessels (those not under pressure or a vacuum), a pipe at or near the bottom of the vessel connects only to the high-pressure side of the meter body and the low-pressure side is open to the atmosphere. If the vessel is pressurized or under vacuum, the low side of the meter has a pipe connection near the top of the vessel, so that the instrument responds only to changes in the head of liquid (see Figure 4).

DP transmitters are used extensively in the process industries today. In fact, newer smart transmitters and conventional 4– 20 mA signals for communications to remote DCSs, PLCs, or other systems have actually resulted in a “revival” of this technology. Problems with dirty liquids and the expense of piping on new installations, however, have opened the door for yet newer, alternative methods.

Hydrostatic Tank Gauging. One growing, specialized application for systems that involve hydrostatic measurements is hydrostatic tank gauging (HTG). It is an emerging standard way to accurately gauge liquid inventory and to monitor transfers in tank farms and similar multiple-tank storage facilities. HTG systems can provide accurate information on tank level, mass, density, and volume of the contents in every tank. These values can also be networked digitally for multiple remote access by computer from a safe area.

Figure 4 shows a simplified system that incorporates only one pressure transmitter (PT) with a temperature transmitter (TT) and makes novel use of a level transmitter (LT) to detect accumulation of water at the bottom of a tank. Mass (weight) of the tank’s contents can be calculated from the hydrostatic head (measured by PT) multiplied by the tank area (obtained from a lookup table). The liquid’s temperature-density relationship can be used to calculate the volume and level, provided the tank is not under pressure. Data fed into a computer system make it possible for all calculations to be automatic, with results continuously available for monitoring and accounting purposes.

The level transmitter, with its probe installed at an angle into the bottom portion of the tank, is an innovative way to detect accumulation of water, separated from oil, and to control withdrawal of product only. Moreover, by measuring the water-oil interface level, the LT provides a means of correcting precisely for the water level, which would incorrectly be measured as product.

Though the DP transmitter is most commonly used to measure hydrostatic pressure for level measurement, other methods should be mentioned. One newer system uses a pressure transmitter in the form of a stainless steel probe that looks much like a thermometer bulb. The probe is simply lowered into the tank toward the bottom, supported by plastic tubing or cable that carries wiring to a meter mounted externally on or near the tank. The meter displays the level data and can transmit the information to another receiver for remote monitoring, recording, and control.

Another newer hydrostatic measuring device is a dry-cell transducer that is said to prevent the pressure cell oils from contaminating the process fluid. It incorporates special ceramic and stainless steel diaphragms and is apparently used in much the same way as a DP transmitter.

Radar or Microwave

Radar methods of level measurement are sometimes referred to as microwave types. Both use electromagnetic waves, typically in the microwave X-band (10 GHz) range. This technology is being adapted and refined for level measurement, so you should check out the latest offerings. Most applications have been designed for continuous level measurement.

Basically, all types operate on the principle of beaming microwaves downward from a sensor located on top of the vessel. The sensor receives back a portion of the energy that is reflected off the surface of the measured medium. Travel time for the signal (called the time of flight) is used to determine level. For continuous level meas urement, there are two main types of noninvasive systems, as well as one invasive type that uses a cable or rod as a wave guide and extends down into the tank’s contents to near its bottom.

One type of noninvasive system uses a technology called frequency-modulated continuous wave (FMCW). From an electronic module on top of the tank, a sensor oscillator sends down a linear frequency sweep, at a fixed bandwidth and sweep time. The reflected radar signal is delayed in proportion to the distance to the level surface. Its frequency is different from that of the transmitted signal, and the two signals blend into a new frequency proportional to distance. That new frequency is converted into a very accurate measure of liquid level.

The sensor outputs a frequency-modulated (FM) signal that varies from 0 to ~200 Hz as the distance ranges from 0 to 200 ft (60 m). An advantage of this technique is that the level-measurement signals are FM rather than AM, affording the same advantages that radio waves offer. Most tank noise is in the AM range and does not affect the FM signals.

The second noninvasive technology, pulsed radar or pulsed time-of-flight, operates on a principle very similar to that of the ultrasonic pulse meth od. The radar pulse is aimed at the liquid’s surface and the transit time of the pulse’s re turn is used to calculate level. Because pulse radar is lower power than FMCW, its performance can be affected by obstructions in the tank as well as foam and low-dielectric materials (K < 2).

Antennas for the noninvasive methods come in two designs: parabolic dish and cone. Sche matically, Figure 5 shows that the parabolic dish antenna tends to direct the signals over a wider area while the cone tends to confine the signals in a narrower downward path. The choice of one or the other, and its diameter, depends on application factors such as tank obstructions that may serve as reflectors, the presence of foam, and turbulence of the measured fluid.

Guided-wave radar (GWR) is an invasive method that uses a rod or cable to guide the micro wave as it passes down from the sensor into the material being measured and all the way to the bottom of the vessel. The basis for GWR is time-domain reflectometry (TDR), which has been used for years to locate breaks in long lengths of cable that are underground or in building walls. A TDR generator develops more than 200,000 pulses of electromagnetic energy that travel down the waveguide and back. The dielectric of the measured fluid causes a change in impedance that in turn develops a wave reflection. Transit time of pulses down and back is used as a measure of level.

The waveguide affords a highly efficient path for pulse travel so that degradation of the signal is minimized. Thus, extremely low dielectric materials (K < 1.7 vs. K = 80 for water) can be effectively measured. Further, because the pulse signals are channeled by the guide, turbulence, foams, or tank obstructions should not affect the meas urement. GWR can handle varying specific gravity and media buildup or coatings. It is an invasive method, though, and the probe or guide may be damaged by the blade of an agitator or the corrosiveness of the material being measured.

Ultrasonic and Sonic

Both ultrasonic and sonic level instruments operate on the basic principle of using sound waves to determine fluid level. The frequency range for ultrasonic methods is ~20–200 kHz, and sonic types use a frequency of 10 kHz. As shown in Figure 6, a top-of-tank mounted transducer directs waves downward in bursts onto the surface of the material whose level is to be measured. Echoes of these waves return to the transducer, which performs calculations to convert the distance of wave travel into a measure of level in the tank. A piezoelectric crystal inside the transducer converts electrical pulses into sound energy that travels in the form of a wave at the established frequency and at a constant speed in a given medium. The medium is normally air over the material’s surface but it could be a blanket of nitrogen or some other vapor. The sound waves are emitted in bursts and received back at the transducer as echoes. The instrument measures the time for the bursts to travel down to the reflecting surface and return. This time will be proportional to the distance from the transducer to the surface and can be used to determine the level of fluid in the tank. For practical applications of this method, you must consider a number of factors. A few key points are:

• The speed of sound through the medium (usually air) varies with the medium’s temperature. The transducer may contain a temperature sensor to compensate for changes in operating temperature that would alter the speed of sound and hence the distance calculation that determines an accurate level measurement.

• The presence of heavy foam on the surface of the material can act as a sound absorbent. In some cases, the absorption may be sufficient to preclude use of the ultrasonic technique.

• Extreme turbulence of the liquid can cause fluctuating readings. Use of a damping adjustment in the instrument or a response delay may help overcome this problem.

To enhance performance where foam or other factors affect the wave travel to and from the liquid surface, some models can have a beam guide attached to the transducer.

Ultrasonic or sonic methods can also be used for point level measurement, although it is a relatively expensive solution. An ultrasonic gap technique is an alternative way to measure point level with low-viscosity liquids. A transmit crystal is activated on one side of a “measurement gap” and a receive crystal listens on the opposite side. The signal from the receive crystal is analyzed for the presence or absence of tank contents in the meas urement gap. These noncontact devices are available in models that can convert readings into 4–20 mA outputs to DCSs, PLCs, or other remote controls.

Selecting the Best Method

Figures 7 and 8 summarize some guidelines that will help you select the right level measurement method for your application. Remember, however, that initial cost is only one consideration—a low initial cost may be far outweighed by high maintenance costs or loss of accuracy over time.

Suppliers often provide recommendations if you specify your needs, usually by filling out a form. Five types of information commonly define the level-measuring instrument or system needed:

• Process material. Give the generic name of the material, such as a 5% sodium hydroxide solution.

• Material characteristics. Specify whether you need to measure a liquid, slurry, solid, interface, granular, or powder. Give values of the material’s dielectric constant, K, conductivity in microsiemens per centimeter (mS/cm), viscosity in centipoise (cP), and density in pounds per cubit foot (lb./ft.³). Also describe consistency in such terms as “watery,” “oily,” “like a batter,” or “like molasses.” If this information is not available, send the supplier a sample for evaluation.

• Process information. Give values of the normal temperature and pressure, as well as the minimum and maximum. If turbulence is present, indicate its degree as light, medium, or heavy. Describe vessel material: Is it metallic, nonmetallic, or lined? Give materials of construction of wetted materials, for example 316 stainless, Kynar, Teflon, or other. Describe area classification: nonhazardous, hazardous (list them), or corrosive (list them too).

• Vessel function. Describe the main function of the vessel, such as sump, reactor, storage, water separation at bottom, and so on. Provide a schematic diagram showing the vessel size and shape, the probe mounting and location, 0% and 100% of level, and the presence of an agitator or other internal obstruction.

• Power requirements. Specify from the following: 115 VAC, 230 VAC, 24 VAC, or loop-

powered (24 VAC, two-wire type).

With a firm grasp of the principles underlying the methods, you should be able to intelligently choose among the options the supplier offers you.

For Further Reading

Bacon, J.M. June 1996. “The changing world of level measurement,” InTech.

Boyes, W. Feb. 1999. “The Changing State of the Art of Level Measurement,” Flow Control.

Carsella, B. Dec. 1998. “Popular level-gauging methods,” Chemical Processing.

Considine, D.M. 1993. “Fluid Level Systems,” Process/Industrial Instruments & Control Handbook. 4th Ed. New York, McGraw-Hill:4.130-4.136.

Gillum, D.R. 1995. “Industrial Pressure, Level, and Density Measurement,” ISA Resources for Measurement and Control Series. Research Triangle Park, NC, Instru ment Society of America.

Johnson, D. Nov. 1998. “Process Instru mentation’s ‘Utility Infielder,’ ” Control Engi neering.

Koeneman, D.W. July 2000. “Evaluate the Options for Measuring Process Levels,” Chemi cal Engineering.

“Level Measurement.” 1995. Instrument Engineer’s Handbook: Process Measure ments and Analysis, B.E. Liptak, Ed., 3rd Ed., Vol. 2. Radnor, PA, Chilton Book Co.:269-397.

“Level Measurement and Control.” Apr. 1999. Measurements & Control:142-161.

“Level Measurement Systems.” 1995. Omega Complete Flow and Level Measure ment Handbook and Encyclopedia. Vol. 29, Stamford, CT, Omega Engineering Inc.

“Level measurement, tank gauging sectors grow, diversify,” Apr. 1999. Control Engi neering:13.

Owen, T. Feb. 1999. “Advanced Elec tronics Overcome Measurement Barriers,” Control.

Parker, S. 1999. “Selecting a level device based on application needs,” Chemical Proc essing, 1999 Fluid Flow Manual:75-80.

Paul, B.O. Feb. 1999. “Seventeen Level Sensing Methods,” Chemical Processing.

Ramirez, R.C. Oct. 1999. “Microwaves calm down black liquor recovery,” InTech:50-53.

RF Level Measurement Handbook. 1999. Princo Instruments Inc.

A Level Measurement Orientation

On the 28th of March, 1979, thousands of people fled from Three Mile Island (near Harrisburg, PA) when the cooling system of a nuclear reactor failed. This dangerous situation

Figure 6-1: Click on figure to enlarge.

developed because the level controls turned off the coolant flow to the reactor when they detected the presence of cooling water near the top of the tank. Unfortunately, the water reached the top of the reactor vessel not because there was too much water in the tank, but because there was so little that it boiled and swelled to the top. From this example, we can see that level measurement is more complex than simply the determination of the presence or absence of a fluid at a particular elevation.

Level Sensor Selection

When determining what type of level sensor should be used for a given application, there are a series of questions that must be answered:

Can the level sensor be inserted into the tank or should it be completely external?

Should the sensor detect the level continuously or will a point sensor be adequate?

Can the sensor come in contact with the process fluid or must it be located in the vapor space?

Is direct measurement of the level needed or is indirect detection of hydrostatic head (which responds to changes in both level and density) acceptable?

Is tank depressurization or process shut-down acceptable when sensor removal or maintenance is required?

By evaluating the above choices, one will substantially shorten the list of sensors to consider. The selection is further narrowed by considering only those designs that can be provided in the required materials of construction and can function at the required accuracy, operating temperature, etc. (Table 4). When the level to be measured is a solid, slurry, foam, or the interface between two liquid layers, it is advisable to consult not only Table 4, but other recommendations, such as Table 5.

If it is found that a number of level detector designs can satisfy the requirements of the application, one should also consider the traditions or preferences of the particular plant or the particular process industry, because of user familiarity and the availability of spare parts. For example, the oil industry generally prefers displacement-type level sensors, while the chemical industry favors differential pressure (d/p) cells. (The petroleum industry will use d/p cells when the span exceeds 60-80 in.)

If the tank is agitated, there is often no space in which to insert probe-type sensors. Plus, because the liquid surface is not flat, sonic, ultrasonic, or radar devices typically cannot be used, either. Even with displacer or d/p sensors, agitation can cause the level signal to cycle. These pulses can be filtered out by first determining the maximum rate at which the level can change (due to filling or discharging) and disregarding any change that occurs faster than that.

Figure 6-2: Click on figure to enlarge.

The relationship between level and tank volume is a function of the cross-sectional shape of the tank. With vertical tanks, this relationship is linear, while with horizontal or spherical vessels, it is a non-linear relationship (Figure 6-1).

Table 4: Click on figure to enlarge.

If the level in a tank is to be inferred using hydrostatic pressure measurement, it is necessary to use multi-transmitter systems when it is desirable to:

Detect the true level, while either the process temperature or density varies;

Measure both level and density; and

Measure the volume and the mass (weight) in the tank.

By measuring one temperature and three pressures, the system shown in Figure 6-2 is capable of simultaneously measuring volume (level), mass (weight), and density, all with an accuracy of 0.3% of full span.

Boiling & Cryogenic Fluids

Figure 6-3: Click on figure to enlarge.

When a d/p cell is used to measure the level in a steam drum, a reverse-acting transmitter is usually installed (Figure 6-3). An uninsulated condensing chamber is used to connect the high pressure (HP) side of the d/p cell to the vapor space on the top of the drum. The steam condenses in this chamber and fills the wet leg with ambient temperature water, while the low pressure (LP) side of the d/p cell detects the hydrostatic head of the boiling water inside the drum. The output of the d/p cell reflects the amount of water in the drum. Output rises as the mass of water in the drum drops (because the steaming rate and the associated swelling increase). It is for this reason that a reverse acting d/p cell is recommended for this application.

Table 5: Click on figure to enlarge.

When the process fluid is liquid nitrogen (or some other cryogenic material), the tank is usually surrounded by a thermally insulated and evacuated cold box. Here, the low pressure (LP) side of a direct acting d/p cell is connected to the vapor space above the cryogenic liquid (Figure 6-4). As the liquid nitrogen approaches the HP side of the d/p cell (which is at ambient temperature outside the cold box), its temperature rises. When the temperature reaches the boiling point of nitrogen, it will boil and,

Figure 6-4: Click on figure to enlarge.

from that point on, the connecting line will be filled with nitrogen vapor. This can cause noise in the level measurement. To protect against this, the liquid filled portion of the connecting line should be sloped back towards the tank. The cross-section of the line should be large (about 1 inch in diameter) to minimize the turbulence caused by the simultaneous boiling and re-condensing occurring at the liquid-vapor interface.

Sludge, Foam, & Molten Metals

Many process fluids are aggressive or difficult to handle and it's best to

Figure 6-5: Click on figure to enlarge.

avoid physical contact with them. This can be accomplished by placing the level sensor outside the tank (weighing, radiation) or locating the sensor in the vapor space (ultrasonic, radar, microwave) above the process fluid. When these options are not available or acceptable, one must aim to minimize maintenance and physical contact with the process fluid.

When the process fluid is a sludge, slurry, or a highly viscous polymer, and the goal is to detect the level at one point, the design shown in Figure 6-5A is commonly considered. The ultrasonic or optical signal source and receiver typically are separated by more than six inches so that the process fluid drains freely from the intervening space. After a high-level episode, an automatic washing spray is activated.

When the sludge or slurry level is detected continuously, one of the goals is to eliminate dead-ended cavities where the sludge might settle. In addition, all surfaces which are exposed to the process fluid should be covered with PFA. Figure 6-5B shows such an installation, employing PFA-coated extended diaphragms to minimize material buildup.

In strippers, where the goal is to drive off the solvent in the shortest period of time, one aims to keep the foam level below a maximum. In other processes, it is desirable to separately control both the liquid level beneath the foam and the thickness of the foam. In the paper industry, beta radiation detectors are used for such applications (Kraft processing), while other industries detect the degree of foaming indirectly (by measuring related variables, such as heat input or vapor flow), or they use capacitance, conductivity, tuning fork, optical, or thermal switches, all provided with automatic washers.

Measuring the level of molten glass or metals is another special application. The most expensive (but also most accurate) technique available is proximity capacitance-based level measurement, which can provide a resolution of 0.1 mm over a range of 6 in. Laser-based systems can provide even better resolution from distances up to 2 ft. If such high resolution is not required and cost is a concern, one can make a float out of refractory material and attach a linear variable differential transformer (LVDT), or make a bubbler tube out of refractory material and bubble argon or nitrogen through it.

A Flow Measurement Orientation

Our interest in the measurement of air and water flow is timeless. Knowledge of the direction and velocity of air flow was essential information for all ancient navigators, and the ability to measure water flow was necessary for the fair distribution of water through the aqueducts of such early communities as

Figure 1-1: Click on figure to enlarge.

the Sumerian cities of Ur, Kish, and Mari near the Tigris and Euphrates Rivers around 5,000 B.C. Even today, the distribution of water among the rice patties of Bali is the sacred duty of authorities designated the "Water Priests."

Our understanding of the behavior of liquids and gases (including hydrodynamics, pneumatics, aerodynamics) is based on the works of the ancient Greek scientists Aristotle and Archimedes. In the Aristotelian view, motion involves a medium that rushes in behind a body to prevent a vacuum. In the sixth century A.D., John Philoponos suggested that a body in motion acquired a property called impetus, and that the body came to rest when its impetus died out.

In 1687, the English mathematician Sir Isaac Newton discovered the law of universal gravitation. The operation of angular momentum-type mass flowmeters is based directly on Newton's second law of angular motion. In 1742, the French mathematician Rond d'Alembert proved that Newton's third law of motion applies not only to stationary bodies, but also to objects in motion.

The Flow Pioneers

A major milestone in the understanding of flow was reached in 1783 when the Swiss physicist Daniel Bernoulli published his Hydrodynamica. In it, he introduced the concept of the conservation of energy for fluid flows. Bernoulli determined that an increase in the velocity of a flowing fluid increases its kinetic energy while decreasing its static energy. It is for this reason that a flow restriction causes an increase in the flowing velocity and also causes a drop in the static pressure of the flowing fluid.

The permanent pressure loss through a flowmeter is expressed either as a percentage of the total pressure drop or in units of velocity heads, calculated as V²/2g, where V is the flowing velocity and g is the gravitational acceleration (32.2 feet/second² or 9.8 meters/second² at 60° latitude). For example, if the velocity of a flowing fluid is 10 ft/s, the velocity head is 100/64.4 = 1.55 ft. If the fluid is water, the velocity head corresponds to 1.55 ft of water (or 0.67 psi). If the fluid is air, then the velocity head corresponds to the weight of a 1.55-ft column of air.

The permanent pressure loss through various flow elements can be expressed as a percentage of the total pressure drop (Figure 1-1), or it can be expressed in terms of velocity heads. The permanent pressure loss through an orifice is four velocity heads; through a vortex shedding sensor, it is two; through positive displacement and turbine meters, about one; and, through flow venturis, less than 0.5 heads. Therefore, if an orifice plate (Figure 1-2) with a beta ratio

Figure 1-2: Click on figure to enlarge.

of 0.3 (diameter of the orifice to that of the pipe) has an unrecovered pressure loss of 100 in H₂O, a venturi flow tube could reduce that pressure loss to about 12 in H₂O for the same measurement.

In 1831, the English scientist Michael Faraday discovered the dynamo when he noted that, if a copper disk is rotated between the poles of a permanent magnet, electric current is generated. Faraday's law of electromagnetic induction is the basis for the operation of the magnetic flowmeter. As shown in Figure 1-3, when a liquid conductor moves in a pipe having a diameter (D) and travels with an average velocity (V) through a magnetic field of B intensity, it will induce a voltage (E) according to the relationship:

where C is the constant for units conversion.

Over the past several years, the performance of magnetic flowmeters has improved significantly. Among the advances are probe and ceramic insert designs and the use of pulsed magnetic fields (Figure 1-4), but the basic operating principle of Faraday's law of electric induction has not changed.

In 1883, the British mechanical engineer Osborne Reynolds proposed a single, dimensionless ratio to describe the velocity profile of flowing fluids:

Figure 1-3: Click on figure to enlarge.

Where D is the pipe diameter, V is the fluid velocity, ? is the fluid density, and µ is the fluid viscosity.

He noted that, at low Reynolds numbers (below 2,000) (Figure 1-5), flow is dominated by viscous forces and the velocity profile is (elongated) parabolic. At high Reynolds numbers (above 20,000), the flow is dominated by inertial forces, resulting in a more uniform axial velocity across the flowing stream and a flat velocity profile.

Until 1970 or so, it was believed that the transition between laminar and turbulent flows is gradual, but increased understanding of turbulence through supercomputer modeling has shown that the onset of turbulence is abrupt.

When flow is turbulent, the pressure drop through a restriction is proportional to the square of the flowrate. Therefore, flow can be measured by taking the square root of a differential pressure cell output. When the flow is laminar, a linear relationship exists between flow and pressure drop. Laminar flowmeters are used at very low flowrates (capillary flowmeters) or when the viscosity of the process fluid is high.

In the case of some flowmeter technologies, more than a century elapsed between the discovery of a

Figure 1-4: Click on figure to enlarge.

scientific principle and its use in building a flowmeter. This is the case with both the Doppler ultrasonic and the Coriolis meter.

In 1842, the Austrian physicist Christian Doppler discovered that, if a sound source is approaching a receiver (such as a train moving toward a stationary listener), the frequency of the sound will appear higher. If the source and the recipient are moving away from each other, the pitch will drop (the wavelength of the sound will appear to decrease). Yet it was more than a century later that the first ultrasonic Doppler flowmeter came on the market. It projected a 0.5-MHz beam into a flowing stream containing reflectors such as bubbles or particles. The shift in the reflected frequency was a function of the average traveling velocity of the reflectors. This speed, in turn, could be used to calculate a flowrate.

The history of the Coriolis flowmeter is similar. The French civil engineer Gaspard Coriolis discovered in 1843 that the wind, the ocean currents, and even airborne artillery shells will all drift sideways because of the earth's rotation. In the northern hemisphere, the deflection is to the right of the motion; in the southern hemisphere, it is to the left. Similarly, a body traveling toward either pole will veer eastward, because it retains the greater eastward rotational speed of the lower altitudes as it passes over the more slowly rotating earth surface near the poles. Again, it was the slow evolution of sensors and electronics that delayed creation of the first commercial Coriolis mass flowmeter until the 1970's.

It was the Hungarian-American aeronautical engineer Theodore von Karman who, as a child growing up in Transylvania (now Romania), noticed that stationary rocks caused vortices in flowing water, and that the distances between these traveling vortices are constant, no matter how fast or slow the water runs. Later in life, he also observed that, when a flag flutters in the wind, the wavelength of the flutter is independent of wind velocity and depends

solely on the diameter of the flag pole. This is the theory behind the vortex flowmeter, which determines flow velocity by counting the number of vortices passing a sensor. Von Karman published his findings in 1954, and because by that time the sensors and electronics required to count vortices were already in existence, the first edition of the Instrument Engineers' Handbook in 1968 was able to report the availability of the first swirlmeter.

The computer has opened new frontiers in all fields of engineering, and flow measurement is no exception. It was only as long ago as 1954 that another Hungarian-American mathematician, John Von Neumann, built Uniac--and even more recently that yet another Hungarian-American, Andy Grove of Intel, developed the integrated circuit. Yet these events are already changing the field of flowmetering. Intelligent differential pressure cells, for example, can automatically switch their range between two calibrated spans (one for 1-10%, the other for 10-100% of D/P), extending orifice accuracy to within 1% over a 10:1 flow range. Furthermore, it is possible to include in this accuracy statement not only hysteresis, rangeability, and linearity effects, but also drift, temperature, humidity, vibration, over-range, and power supply variation effects.

With the development of superchips, the design of the universal flowmeter also has become feasible. It is now possible to replace dye-tagging or chemical-tracing meters (which measured flow velocity by dividing the distance between two points by the transit time of the trace), with traceless cross-correlation flowmeters (Figure 1-6). This is an elegant flowmeter because it requires no physical change in the process--not even penetration of the pipe. The measurement is based on memorizing the noise pattern in any externally detectable process variable, and, as the fluid travels from point A to point B, noting its transit time.

Flow Sensor Selection

The purpose of this section is to provide information to assist the reader in making an informed selection of flowmeter for a particular application. Selection and orientation tables are used to quickly focus on the most likely candidates for measurement. Tables 1-I and 1-II have been prepared to make available a large amount of information for this selection process.

At this point, one should consider such intangible factors as familiarity of plant personnel, their experience with calibration and maintenance, spare parts availability, mean time between failure history, etc., at the particular plant site. It is also recommended that the cost of the installation be computed only after taking these steps. One of the most common flow measurement mistakes is the reversal of this sequence: instead of selecting a sensor which will perform properly, an attempt is made to justify the use of a device because it is less expensive. Those "inexpensive" purchases can be the most costly installations.

The basis of good flowmeter selection is a clear understanding of the requirements of the particular application. Therefore, time should be invested in fully evaluating the nature of the process fluid and of the overall installation. The development of specifications that state the appl

Figure 1-5: Click on figure to enlarge.

ication requirements should be a systematic, step-by-step process.

The first step in the flow sensor selection process is to determine if the flowrate information should be continuous or totalized, and whether this information is needed locally or remotely. If remotely, should the transmission be analog, digital, or shared? And, if shared, what is the required (minimum) data-update frequency? Once these questions are answered, an evaluation of the properties and flow characteristics of the process fluid, and of the piping that will accommodate the flowmeter, should take place (Table 1-I). In order to approach this task in a systematic manner, forms have been developed, requiring that the following types of data be filled in for each application:

Fluid and flow characteristics: In this section of the table, the name of the fluid is given and its pressure, temperature, allowable pressure drop, density (or specific gravity), conductivity, viscosity (Newtonian or not?) and vapor pressure at maximum operating temperature are listed, together with an indication of how these properties might vary or interact. In addition, all safety or toxicity information should be provided, together with detailed data on the fluid's composition, presence of bubbles, solids (abrasive or soft, size of particles, fibers), tendency to coat, and light transmission qualities (opaque, translucent or transparent?).

Expected minimum and maximum pressure and temperature values should be given in addition to the normal operating values. Whether flow can reverse, whether it does not always fill the pipe, whether slug flow can develop (air-solids-liquid), whether aeration or pulsation is likely, whether sudden temperature changes can occur, or whether special precautions are needed during cleaning and maintenance, these facts, too, should be stated.

Concerning the piping and the area where the flowmeter is to be located, the following information

Figure 1-6: Click on figure to enlarge.

should be specified: For the piping, its direction (avoid downward flow in liquid applications), size, material, schedule, flange-pressure rating, accessibility, up or downstream turns, valves, regulators, and available straight-pipe run lengths.

In connection with the area, the specifying engineer must know if vibration or magnetic fields are present or possible, if electric or pneumatic power is available, if the area is classified for explosion hazards, or if there are other special requirements such as compliance with sanitary or clean-in-place (CIP) regulations.

The next step is to determine the required meter range by identifying minimum and maximum flows (mass or volumetric) that will be measured. After that, the required flow measurement accuracy is determined. Typically accuracy is specified in percentage of actual reading (AR), in percentage of calibrated span (CS), or in percentage of full scale (FS) units. The accuracy requirements should be separately stated at minimum, normal, and maximum flowrates. Unless you know these requirements, your meter's performance may not be acceptable over its full range.

Accuracy vs. Repeatability

In applications where products are sold or purchased on the basis of a meter reading, absolute accuracy is critical. In other applications, repeatability may be more important than absolute accuracy. Therefore, it is advisable to establish separately the accuracy and repeatability requirements of each application and to state both in the specifications.

When a flowmeter's accuracy is stated in % CS or % FS units, its absolute error will rise as the measured flow rate drops. If meter error is stated in % AR, the error in absolute terms stays the same at high or low flows. Because full scale (FS) is always a larger quantity than the calibrated span (CS), a sensor with a % FS performance will always have a larger error than one with the same % CS specification. Therefore, in order to compare all bids fairly, it is advisable to convert all quoted error statements into the same % AR units.

It is also recommended that the user compare installations on the basis of the total error of the loop. For example, the inaccuracy of an orifice plate is stated in % AR, while the error of the associated d/p cell is in % CS or % FS. Similarly, the inaccuracy of a Coriolis meter is the sum of two errors, one given in % AR, the other as a % FS value. Total inaccuracy is calculated by taking the root of the sum of the squares of the component inaccuracies at the desired flow rates.

In well-prepared flowmeter specifications, all accuracy statements are converted into uniform % AR units and these % AR requirements are specified separately for minimum, normal, and maximum flows. All flowmeter specifications and bids should clearly state both the accuracy and the repeatability of the meter at minimum, normal, and maximum flows.

Table 1 provides data on the range of Reynolds numbers (Re or R_D) within which the various flowmeter designs can operate. In selecting the right flowmeter, one of the first steps is to determine both the minimum and the maximum Reynolds numbers for the application. Maximum R_D is obtained by making the calculation

Table 1: Click on table to enlarge.

when flow and density are at their maximum and viscosity at its minimum. Conversely, the minimum R_D is obtained by using minimum flow and density and maximum viscosity.

If acceptable metering performance can be obtained from two different flowmeter categories and one has no moving parts, select the one without moving parts. Moving parts are a potential source of problems, not only for the obvious reasons of wear, lubrication, and sensitivity to coating, but also because moving parts require clearance spaces that sometimes introduce "slippage" into the flow being measured. Even with well maintained and calibrated meters, this unmeasured flow varies with changes in fluid viscosity and temperature. Changes in temperature

Table II: Click on Table to enlarge.

also change the internal dimensions of the meter and require compensation.

Furthermore, if one can obtain the same performance from both a full flowmeter and a point sensor, it is generally advisable to use the flowmeter. Because point sensors do not look at the full flow, they read accurately only if they are inserted to a depth where the flow velocity is the average of the velocity profile across the pipe. Even if this point is carefully determined at the time of calibration, it is not likely to remain unaltered, since velocity profiles change with flowrate, viscosity, temperature, and other factors.

If all other considerations are the same, but one design offers less pressure loss, it is advisable to select that design. Part of the reason is that the pressure loss will have to be paid for in higher pump or compressor operating costs over the life of the plant. Another reason is that a pressure drop is caused by any restriction in the flow path, and wherever a pipe is restricted becomes a potential site for material build-up, plugging, or cavitation.

Before specifying a flowmeter, it is also advisable to determine whether the flow information will be more useful if presented in mass or volumetric units. When measuring the flow of compressible materials, volumetric flow is not very meaningful unless density (and sometimes also viscosity) is constant. When the velocity (volumetric flow) of incompressible liquids is measured, the presence of suspended bubbles will cause error; therefore, air and gas must be removed before the fluid reaches the meter. In other velocity sensors, pipe liners can cause problems (ultrasonic), or the meter may stop functioning if the Reynolds number is too low (in vortex shedding meters, R_D > 20,000 is required).