Clamp-on, non-intrusive ultrasonic flow measurement provides the process engineer with an opportunity to measure flow in a filled pipe without the need to interfere with the process, simply by clamping a pair of compact transducers to the outside of the pipe.
Easy to install, low maintenance, completely hygienic and reliable. While there are a couple of pitfalls to avoid to make sure you get the best from the technology, clamp-on ultrasonic metering is a great opportunity to measure a flow rate without the hassle of breaking into a pipe, and without the need to stop the process.
When a process engineer is faced with the requirement for flow measurement of liquids or gas in a pipe, there are many technologies that they can reach for, each with its own set of benefits and limitations.
With a very few exceptions, though, they share the same challenge – that you need to interrupt the process or break into the pipe to make the measurement.
The exception is, of course, the clamp-on transit-time ultrasonic meter, a technology that has matured and produces high quality results in often difficult applications, without the installation challenges that other technologies demand.
At its simplest, a compact set of transducers are clamped on to the outside of a filled pipe, connected to an electronic transmitter. Ten minutes setup, and a user has a reliable and repeatable measurement, accurate to within a couple of percent of volumetric flowrate.
In this article, I’m going to look at the principles and technology of this type of retrofit flowmeter, we’ll look at the challenges of correct installation and I’ll also talk about the best ways to get an installation wrong. I’ll also discuss the suite of diagnostics that the top manufacturers provide to help users to get the best possible results.
First things first. The measurement principle depends on the fact that an ultrasonic pulse travelling through a flowing medium will travel more quickly in a downstream direction than it will upstream, and the difference between those two times is proportional to the velocity of the flow.
We fire pulses across the width of the pipe at an angle, so the measurement is often referred to as ‘transit time difference’ measurement.
The transmitter analyses the returning signal, the echo, from each transducer, thus measuring the difference between the upstream and downstream transits to a very high degree of accuracy.
Incidentally, another technology that looks very similar is clamp-on Doppler measurement, which measures a frequency shift as ultrasonic waves reflect from moving particles.
In general, this is not as accurate a measurement approach as transit time measurement but has the benefit that it works well in ‘dirty’ fluids, with more than around 10% solids or bubbles. This article is only concerned with transit time measurement.
The clue is in the name, really, but ‘ultrasonic’ is a sound wave above the frequency of human hearing, around 20kHz, and it’s a very versatile technology, used in cleaning, in ultrasound scans that provide images of the inside of the body and in level and flow measurement.
On a personal note, I’m about to be a grandad for the first time so I’ve become happily familiar with the imaging technology that has allowed my wife and I to see our new grandchild for the first time! In measurement, ultrasonic waves can be used to measure level, velocity, and thickness.
Ultrasonic sensors can be used to measure the distance to an object, or to measure the speed of an object moving through a liquid or gas. Ultrasonic thickness gauges can be used to measure the thickness of materials, such as metal or plastic.
An ultrasonic transducer converts electrical energy into mechanical vibrations at an ultrasonic frequency. Different types of transducers can be used to generate ultrasonic signals, although piezoelectric transducers are most commonly used in measurement.
A piezoelectric material, such as quartz, expands and contracts when a voltage is applied to it, creating the ultrasonic pulses which then travel through the pipe wall and the fluid in the pipe. When an ultrasound pulse reaches the other transducer, it re-excites the crystal, causing it to emit a very much smaller voltage that is then analysed by the transmitter.
The speed at which the wave travels depends on a number of factors, such as density, temperature and elasticity – the wave speed tends to be higher the denser the medium. The intensity (amplitude) of the wave varies with a range of factors: distance from the source as the energy of the wave is dissipated through the medium, and as boundaries or obstacles refract or reflect the wave.
So, the skill of the flowmeter manufacturer comes in part from the ability to generate a packet of signals from a transducer, reliably transmit it through the pipe wall and the medium, and discriminate that data from the mass of competing information, noise and vibration when that signal is received by the other transducer, and do that both upstream and downstream.
The size and construction of the pipe, the pipe wall thickness, the medium and the temperature all have an effect on the speed of sound and, therefore, how the signal wave will propagate, all of which affect the path that the signal follows through the application.
The set-up process when a flowmeter is installed is largely concerned with the accurate positioning of the sensors on the pipe to maximise signal strength (though more of that in a while).
At the end of the process, what the instrument has actually measured is a pair of transit times, and it knows the difference between those times to within a few nanoseconds. A really important point is that this is the only thing that the flowmeter actually measures, and the accuracy of the measurement of flow depends very much on what happens next.
The main mistake people make when setting up a clamp-on flowmeter is to underestimate the importance of accurately measuring the pipe. The flowmeter uses the transit time difference to calculate the velocity of the flow, which it will do to within around 0.5% accuracy, then it uses the dimensions of the pipe to calculate the volumetric flow of the fluid in the pipe.
Any error in the pipe dimension measurement has a significant effect. I’m writing this piece in an airport where, purely for research purposes, I’m sitting in the pizza restaurant. I went for the mushroom pizza, with a choice of 10 inch or, for an extra £1, 13 inch.
The extra 2 inches gave me an extra 54 square inches of pizza, or an increase of around 65%! We can argue about whether there was actually any more topping involved, but you can’t ignore a saving of £75 per square metre…
Now, nobody is going to make a mistake of 2 inches or 50 mm in a pipe diameter, but let’s take a more realistic example: If we have a perfect pipe in a perfect application with an outside diameter of 200 mm and pipe wall thickness of 5mm, then a flow velocity of 1.96 m/s will give a volumetric flow rate of 200 m3/hr.
If the operator makes an error of only 5mm in the outside diameter measurement, 195 mm instead of 200 mm, then the displayed flow rate becomes 190 m3/hr. In other words, the error of 5mm in outside diameter results in an error in the flowrate of 5%. The flowmeter has done its job perfectly, it’s just the calculation that gives the issue.
So, practically, what can the flowmeter do to help?
Performance and signal strength is optimised by getting the sensors in the right spot. The way we at Katronic go about this is with our Audible Positioning Assistant. After the basic pipe parameters have been entered using the Quick Setup wizard, the first screen you see is the positioning screen.
The flowmeter tells the user the recommended sensor separation, measured from the inner faces of the sensor bodies.
Once the user has them in position, two bars across the screen indicate signal strength and signal confidence, with an additional dot in between them to help the user tweak the sensors into the perfect position.
The dot moves as the user moves the sensors slightly closer together or further apart, a central position indicating the best possible signal.
We also have the first opportunity to check whether the flowmeter set up is correct. The flowmeter calculates the sensor separation value from the parameters that were entered, while the fine tuning using the dot is a ‘tweak’ driven by the signal strength being picked up by the flowmeter.
In theory, the maximum signal strength should be within a couple of millimetres of the recommended sensor position. So, if the actual maximum signal is picked up in a significantly different place, that’s a very strong indication that the pipe parameters have been entered wrongly, usually an error in the pipe diameter (back to pizza!).
There is also a whole host of diagnostic data to help the user to check that everything is working properly. Although there is a ‘headline’ signal strength figure displayed, in most instruments that’s more accurately the ‘signal to noise ratio’ (SNR), calculated by subtracting the noise (electrical and acoustic) from the raw signal strength, both measured in decibels.
Ideally, we want to see a high signal strength with low noise, but we can see the same SNR from a strong signal with high noise, or a weaker signal with very low noise.
That data, taken with the ‘Gain’ figure which indicates how hard the flowmeter is having to work to get a signal, tells us whether the transducers are positioned properly, with the correct configuration of reflections and are of the right type to get the best possible result for the measurement.
There are other diagnostic functions available to the operator that helps them to fine-tune applications.
One of the most valuable screens for initial set up is the speed of sound (SoS) screen. Variations in the speed of sound don’t significantly affect the flow measurement accuracy because it’s the transit time difference we’re interested in, not the absolute transit time measurement, but they do affect sensor placement.
For that reason the flowmeter has a set of tables that indicate SoS for different temperatures in each medium we measure. The flowmeter also measures the actual speed of sound measured in the application, a simple calculation of distance the signal travels divided by the time taken.
These should agree, pretty much. Any significant difference between the measured and theoretical speeds of sound is another clear indicator that the parameters have not been correctly entered into the meter. Again – usually the pipe size or the temperature of the medium is significantly out.
The Quick Setup routine deals with 90% of all applications, and as long as the parameters are entered correctly most users will never need to see the diagnostic functions.
There are, however, occasions where it is valuable to understand some of the detail of the application, especially when commissioning a permanently installed flowmeter in a more difficult application.
Clamp-on flowmeters such as Katronic’s well-proven KATflow range provide the user with a powerful series of tools to help the operator to really understand how the pipe, the flow and the process are all taken together to give them the best possible performance and reliable, accurate measurement.
