Verifying Open Channel Flowmeter Performance

Infrastructure Security Article December 28, 2000
Printer-friendly version

The Need for Field Testing of Flowmeters

Accurate data from open channel flowmeters are essential to the decision process regarding the current adequacy and the need for upgrading municipal wastewater collection systems. Engineering studies that misstate the true system performance can result in costly errors resulting in misspent public funds and potential liabilities.

The measurement of flow in open channel sewers can be complex. Recently portable flowmeters that measure both velocity and area and calculate flow based on the Continuity Equation Q = V x A have become the industry choice. There is current industry controversy over the performance of some of the popular brands. Therefore, potential purchasers are looking for ways to get the facts regarding accuracy.

Flow Lab Tests

Traditionally the most exacting way to determine a flowmeter's accuracy and repeatability over a wide range of flow conditions has been the hydraulic flow laboratory.

Because of the expense of these laboratory tests, most users tend to rely upon either the manufacturers specifications or the tests of others. In some cases, users attempt to test the meters themselves under available field conditions. Furthermore, since meters generally perform worse in the field than under ideal laboratory conditions there is an additional need to determine the performance of instruments in the field.

Traditional Field Testing Methods Dye Dilution Method

Obtaining a flow standard in the field is difficult. One technique utilizes the dye dilution technique where a known concentration of fluorescent dye is injected at a known rate into the upstream portion of the piping system. Analysis of the resulting dye concentration in the downstream samples can give a very high degree of flow rate accuracy. While useful only for spot checks, flow rate accuracies on the order of + 2% can be expected under carefully controlled conditions.

The Salt Velocity Method

If one wishes to determine the average velocity from point A to point B a "slug" of salt water can be injected into an upstream manhole. By knowing the distance between the injection point and detecting the arrival of the relatively higher conductivity salt solutions at the sample point by using a conductivity sensor, the time of transit can be determined and hence the average velocity calculated. Unlike the flow rate measurements made through the dye dilution technique, this technique determines the average velocity, not the flow rate. If the entrance and exit depths are similar, one can reasonably assume that the flow cross-section is relatively the same throughout the measured distance and therefore a good estimate (±4%) of flow could be calculated by the Continuity Equation

Q = V x A.

Other Velocity Timing Techniques

Visual techniques using drogues are similar in nature to the salt injection technique. Such drogues can be placed in an upstream manhole and the transmit time (and distance) measured to a downstream manhole to determine the average velocity.

The Salt Slug Method as well as the drogue technique can yield average velocities of + 3-5%. The user is encouraged to perform the tests several times and to throw out the unusually slow readings (they may have gotten hung up if they were drogues). The remaining transit times are then averaged to obtain the average velocity.

Portable Velocity Meters

Portable velocity meters have been used for decades for the measurement of flow in rivers and streams as well as other open channels such as sewers. Hand-held velocity meters based upon the electromagnetic principal are popular both in the United States and in Europe. These are solid-state sensors with no moving parts and are far more suited to the sewer environment than are mechanical meters.

To measure the average velocity of flow in an open channel such as a sewer, multiple velocity readings are generally taken at defined locations throughout the cross-section and these velocity readings averaged to obtain the average stream velocity. In general, the average velocity can be measured to approximately + 5% by utilizing good field practices.

Lift Station as a Field Hydraulic Laboratory

Many flow experts are beginning to use lift stations as their "flow laboratory" for field verification of flowmeter performance. The lift station uses a volumetric tank that is periodically emptied. This tank can provide a measuring chamber that is similar to those used in hydraulic laboratories for precision flow instrument calibration.

Lift stations that are designed around a wet well with constant speed pumps which cycle periodically are good candidates for a facility that can provide a periodic sampling of the inflow.

As the storage chamber (wet well) fills up, the wet well level will reach a point at which the pump(s) turn on and evacuate the chamber. The pump is generally sized so that the time that it takes to evacuate the chamber is small compared to the time it takes to fill the chamber. By knowing the dimensions of the wet well and measuring the time it takes to fill the chamber one has a direct measure of the average inflow over the fill period.

While the pump is evacuating the wet well the flow continues to enter the wet well. To estimate the inflow during this period of time, a simple technique is to take the average inflow rate during the fill time and then assume that this inflow rate continues unchanged during the pump on time.

To provide an acceptable measurement of the inflow, the lift station should cycle at a high enough rate that any changes in the inflow are not "overly averaged out."

Lift stations that periodically cycle between an upper and lower level (but which also have continuous level devices) are ideal for measuring the inflow to the lift station. An improvement to the "one sample per lift station cycle" technique is the addition of a continuous level monitoring device to the wet well. Unlike the "one sample per lift station cycle," a lift station instrumented with a continuous level sensor can provide a near continuous readout of the inflow rate. There is, however, an interruption in the direct measurement of inflow that occurs during each of the pump on times (evacuation period). Again, inflow must be estimated during this evacuation period.

It is obvious that by knowing the dimensions of the wet well (including any variations in cross-section over the distance from the all pumps off to the lead pump) and by monitoring the rate of rise of the water in the wet well, the inflow rate can be continuously monitored and even rapidly changing inflow can be tracked. By measuring the instantaneous fill characteristic of each cycle one ends up with a piecewise curve that is a very accurate representation of the inflow during the fill time. This piecewise data can be input into a PC and a best-fit curve performed to better estimate the flow rate during the evacuation period. This setup provides a near-perfect laboratory for meter testing and the accuracy (typically ±1%) of this "field standard" is unsurpassed by any of the previously described techniques.

Even if one chooses to be more exact and not use the estimates of inflow during the pump on time, a large amount of accurate "lab quality data" can be obtained during just the fill period.

Points to Consider

The lift station should be such that a single inflow line exists between the location of the test meter and the lift station. Otherwise, not all the flow coming in the lift station would be passing by the test meter.

The wet well should not "back up" into the inflow line. Instead, the flow should free fall into the wet well.


Not all flowmeters are suitable for the engineering studies to which they are applied. Many commercially available flowmeters have not been thoroughly tested by the manufacturer. Fortunately the user can, with some effort, evaluate the flowmeters in the field prior to the purchase decision.
About the author: 
Larry Marsh is the President and CEO of Marsh-McBirney, Inc. in Frederick, MD. For further information call (301) 874-5599 or (800) 368-2723
Overlay Init