The Case of the Mysterious False Full-Scale Flow

By John Cargin

In a Parshall flume flow measurement, the flow rate would become full scale when there was no flow.

At a wastewater treatment facility they used an ultrasonic level sensor and a Parshall flume to measure the effluent flow from a sequencing batch reactor. With this system of treatment, the flow from the plant is intermittent when the tanks are decanting the clarified water and no flow during the remainder of the time. After about a year in operation the operator noticed that the effluent flow meter would mysteriously begin to show full scale flow when there was actually no flow at all. The major problem this caused was the totalization of the effluent quantity, a required number for the state regulatory agency, was very wrong.

By checking not the flow, but rather the level reading the ultrasonic sensor, was showing the level sensor was seeing the edges of the Parshall flume and not the actual level in the flume. This indicated that the placement of the transducer was incorrect and allowing too much signal to be “seen” from the edges of the flume.

The question was why it took a year for this problem to start. Surely, if the placement of the sensor transducer was wrong now, it should have been wrong when it was first installed. The edges of the flume were way outside the beam of the transducer. Some of the suspected causes were investigated.

The signal cable connectors and junction boxes were tested for leakage. The transducer was shielded from sunlight. The transducer was replaced. The level sensor transmitter was sent in for repair. Nothing has corrected the problem or changed since installation. I realized that something was causing the reflected ultrasonic echo from the edges of the flume to be greater in proportion to the reflection from the bottom of the flume. I noticed that there was a thin layer of algae growing on the bottom of the flume. This may be attenuating the ultrasonic reflection from the level sensor causing a relative increase in the reflection from the edges of the flume. This was something that had changed from the installation. I used a hose to wash most of it away from beneath the sensor. The problem went away and with continued periodic hosing of the flume, the problem never came back.

John Cargin characterizes his education as an Electrical Engineer. However, his interest has always been broader than that narrow spectrum of Physics. Plus, he has an interest in Computer Science. He was first educated back when the university had one mainframe computer. Calculations were performed with slide rules and drawings came from pencils and paper. His experience started with a company that produced a microcomputer-based order entry and point-of-sale terminals when programs were submitted on punch cards and you had four bits to work with. For the last decade he has been implementing PLC-based automation systems together with SCADA systems in both the municipal and private sectors.