The accuracy of acoustic velocity meters depends on equipment limitations, the accuracy of acoustic-path length and angle determination, and the stability of the mean velocity to acoustic-path velocity relation.

Equipment limitations depend on the path length and angle, the transducer frequency, the timing oscillator frequency, and the signal-detection scheme. Typically, an acoustic velocity meter using a multiple-voltage threshold signal-detection scheme, 200-kilohertz transducers, an 80-megahertz timing oscillator, and averaging 100 interrogations per minute can have a velocity error of about ± 10 millimeters per second for a 20-meter path length and a velocity error of about ± 1 millimeter per second for a 200-meter path length.

Error in the measurement of acoustic-path angle or length can result in a proportional measurement bias. Typically, an angle error of 1 degree can result in a velocity error of 2 percent, and a path-length error of 1 meter in 100 meters can result in an error of 1 percent. In many situations, these measurement errors can be adjusted with check current-meter measurements. For very low velocity flow where check measurements are impractical, special care must be taken to make the best possible path angle and length determinations.

Ray bending (signal refraction) depends on path length and density gradients present in the stream. Any deviation from a straight acoustic path between transducers can change the unique relation between path velocity and mean velocity. These deviations can then introduce error in the mean velocity computation. In many short-path situations (less than 200 meters), this error can be avoided or is of minimal importance. Typically, for a 200-meter path length, the resultant error is less than 1 percent, but for a 1,000-meter path length, the error can be greater than 10 percent.

Recent laboratory and field costs tests at the U.S. Geological Survey hydraulic laboratory facility at the Stennis Space Center in Mississippi have substantiated assumptions of equipment limitations. An acoustic velocity meter was tested in both tow-tank and field installations. Tow-tank tests, based on the use of an acoustic velocity meter with a 4.69-meter path, had a maximum velocity error of 27 millimeters per second and an average standard deviation of 9 millimeters per second; and the field tests, also based on the use of an acoustic velocity meter with a 20.5-meter path, had a maximum velocity error of 8 millimeters per second and an average standard deviation of 4 millimeters per second.

(The entire report is available below)

---

Information about on-line reading and printing of historic documents

These reports and documents have been scanned from the original hard-copy materials and are made available on the internet in both HTML and PDF formats. Because these are scanned documents, we are unable to provide fully-accessible versions of these reports. If you cannot fully access the information in these documents, please contact Heather S. Henkel at hhenkel@usgs.gov.

The HTML versions of these documents have been created to provide the information in a format that is quickly and easily readable over the Internet. Selective pages and images can be printed from this HTML version by placing the cursor inside the right-hand frame and selecting the print option from the browser. The PDF version of the reports are also provided, and are the recommended format to use for the best printer format and resolution.

Please note - some PDF files are very large (over 2 MB) and may take some time to download, depending on your system.

View the HTML version of the report (downloads faster for on-line reading)

View the PDF version of this report (3.65 MB) (best for printing)