CHAPTER 11 - ACOUSTIC FLOW MEASUREMENT
7. Open Channel Acoustic Flowmeters
Open channel acoustic flowmeters are based on the same principles as pipeline flowmeters. However, open channel acoustic flowmeters are more complicated than pipeline flowmeters because the cross-sectional area varies with changing water level or stage. In general, these flowmeters are only economically practical for use where the following conditions exist:
This section will cover any additional considerations associated with open channel acoustic flowmeters not covered in the previous section on closed conduit acoustic flowmeters. Laenen (1985) and Laenen and Curtis (1989) contain more detailed information on open channel acoustic flowmeters.
Design of open channel meters is complicated by the potential errors introduced by a variable water surface and because the open channel environment can cause acoustic signal attenuation and refraction (bending). Another potential problem is signal deflection caused by density gradients or signal reflection from the channel bottom or water surface.
(a) Single-Path Acoustic Velocity Meters
In general, single-path acoustic velocity meters (AVMs) are used as flowmeters by calibrating acoustic path velocities against mean channel velocities computed using standard stream gaging techniques. The discharge rating procedure for an AVM gaging site will involve developing ratings for both cross-sectional area and mean channel velocity. Necessary data required to develop these ratings are a stage-area relationship, acoustic path velocities, and the mean velocities through the discharge measurement cross section for a range of flows and stages. A data set should uniformly cover the expected range of stage and discharge. A velocity rating is developed using linear regression techniques to find the best-fit equation, with the instantaneous mean channel velocity as the dependent variable, and/or stage (acoustic path velocity) as the independent variables. After a calibration is established, discharge is computed by multiplying the instantaneous mean channel velocity, predicted from the best-fit equation, and the channel's cross-sectional area, which is determined using the stage-area rating. This method of flow measurement is only as accurate as the ratings developed during the calibration. Therefore, care must be taken while measuring the discharge and in determining the channel's cross-sectional area for a range of stages. For installations where appreciable changes in stage occur, the transducers will have to be positioned to allow a full range of measurements.
(b) Multipath Flowmeter
This type of flowmeter uses several acoustic paths which are mounted at various elevations throughout the measurement section. The average velocity for each path is used to establish the velocity profile. The velocity profile is then numerically integrated over the channel's cross-sectional area to determine the volumetric flow rate. As a result, flowmeter accuracy is relatively independent of the velocity profile. However, integration errors are unavoidable because the velocities near the channel bottom and water surface cannot be measured because of acoustic interference caused by signal reflection.
Flowmeter accuracy and performance are limited by four factors:
1. Location of acoustic paths with respect to water surface and the channel bottom, which are reflective surfaces that can cause multipath interference at the receiving transducer(s).
2. Density gradients (usually caused by different water temperatures or salinities) cause the acoustic path to bend, which changes the acoustic path length.
3. Acoustic signal attenuation caused by varying concentrations of air bubbles, sediment, organic matter, and aquatic organisms.
4. Streamflow variability, which causes the angle between the acoustic path and the flow to change.
Two types of equipment are available for use in measuring velocity: (1) a simple one- or two-path microprocessor based, preprogrammed system that will measure velocity only, and (2) a more complex, programmable, multipath minicomputer that can calculate discharge. At present (1997), open channel systems use 12 volts direct current or 110/220 volts alternating current.
(e) Site Selection
A thorough review of system limitations and equipment requirements is necessary prior to site selection (Laenen, 1985; Laenen and Curtis, 1989). A good measurement site has a reach where the velocity distribution is uniform and the channel is confined; areas with eddies or a high degree of turbulence should be avoided. It is recommended that the channel be relatively straight for 5 to 10 channel widths upstream and 1 to 2 channel widths downstream from the measurement section. The channel bottom should be stable or easily monitored for variations. A constant cross-sectional area and shape over the upstream and downstream extent of the measurement section is desirable. If this condition cannot be met, an "effective" cross-section shape must be determined. The "effective" cross section is determined by taking the cross-sectional area along the acoustic path multiplied by the cosine of the path angle, . A concrete-lined section with a straight reach located upstream is ideal. During site selection, obtain cross-section survey information and note obstructions which may block the acoustic signal. Obtain temperature, total dissolved solids and sediment concentrations, and possible sources of air entrainment (overfalls, spillways, etc.). Variations in stage should be known in order to determine the number of acoustic paths required to assure the system accuracy.
(f) Transducer Mounting Requirements
When transducers are installed, their position and elevation must be measured and adjusted accurately for each transducer pair. Likewise, path lengths and path angles must be measured accurately. Transducer alignment is critical for establishing a strong acoustic signal and is usually performed by divers. Mountings should be designed so that transducer maintenance can be performed without using divers. Transducers are normally mounted near the banks, so mounting transducers on existing structures simplifies the installation process. Cabling options include submarine, overhead, or a responder link which eliminates the need for a cable crossing the channel. Cabling must be protected from damage from dredging, marine traffic, or vandalism.
(g) Site Analysis
The acoustic path(s) at each site should be checked for multipath interference caused by the water surface or channel bottom. In general, for every 100 feet (ft) of acoustic path length, about 1 ft of clearance is necessary to prevent multipath interference. The transducers should be located at least 20 in below the water surface to prevent signal bending caused by solar warming. Signal bending will affect the flowmeter accuracy. Therefore, avoid conditions where the acoustic signal is bent and is reflected off the water surface and/or channel bottom. Check the acoustic signal for potential attenuation. The normal sediment concentration that can be tolerated by most systems is about 2,000 milligrams per liter. However, tolerable concentrations are a function of transducer frequency, particle size, and acoustic path length.
Flowmeter calibration can be done using current meter measurements, other velocity-area methods, or using computations based on theoretical velocity profiles. The effort expended for calibration will depend upon factors such as number of acoustic paths, flow conditions at all stages, channel stability, and accuracy requirements. However, accuracy can be verified only within the limits of the calibration method used.
For many streamflow conditions, a single-path flowmeter can measure flow within an accuracy of 3 to 5 percent. For multipath systems, accuracies of 2 percent or better can be achieved over a wide range of flow rates and channel conditions if the system design addresses the major sources of errors of acoustic flow measurement. Errors associated with open channel acoustic flowmeters are usually attributed to three sources:
1. Transit-time measurements, where timing errors can be on the order of 0.1 foot per second (ft/s) for systems which employ signal validation routines or 0.3 ft/s when signal validation techniques are not used.
2. Acoustic path angle variation. In general, for every one degree of uncertainty in path angle, about 1 percent uncertainty occurs in velocity measurement. Use of crossed acoustic paths will compensate for variations in streamflow direction.
3. Acoustic signal bending. For path lengths less than 1,000 ft, this error is usually less than 3 percent in velocity.
(j) Operation and Maintenance
Acoustic flowmeters are advanced electronic systems that require specialized maintenance. Properly trained technicians are needed to keep the flowmeter operating. An electronic technician and proper test equipment are needed to troubleshoot the equipment. This requirement is especially true during the initial phases of installation. Likewise, transducer mounts should be designed to allow access for transducer cleaning, alignment, and replacement without using divers.