John R. Pugh

The loose particulate nature of bulk solids has characteristics that challenge the design of instrumentation systems for solids handling. Examples include: particulate solids do not distribute themselves evenly inside storage vessels; they travel in complex forms, whether flowing along pipes or being carried by conveyor belts; low-density solids will often result in ill-defined surfaces, making level measurement difficult; and dusty solids will often gather in measurement ports, such as those leading to pressure transducers.

While manufacturers have developed excellent transducers and associated systems that tackle these difficulties, a full appreciation of the complexity of solids handling systems is required, because the ‘wider system’ can often destroy the inherent performance of the transducers.

It is arguable, at least by me, that a major aspect of successful measurement is the communication between instrumentation supplier/developer and the end user of the instrumentation, both of whom may of course be in the same company.

In my view, the vital first step is to consider the measurements in broad terms, including topics such as the surrounding environment, the material and transport characteristics, the purpose of the measurements, what needs to be measured, and how well each quantity needs to be measured. This initial information should help identify the range of measurement options. For example, is a weighing system required or would a level measurement system be more appropriate? Do we need an in-line measurement of mass flow rate or would a gain-in-weight system be sufficient? Is the instantaneous mass flow of material fairly constant or, if not, how well will the measurement technique (such as a belt weigher) accommodate the fluctuations? There are, of course, many other factors to consider.

The maximum measurement uncertainty that can be tolerated by the user for each of the measurements also has to be established. All measurements have uncertainty and consequently, to some degree, they will all be wrong. The user is the person who is best placed to evaluate the consequences of a measurement that is ‘too wrong’. The evaluation of the maximum tolerable uncertainty has to be a balance and should be based on what is needed rather than what ‘would be nice’.

As the requirement for a low uncertainty rises, in addition to the cost of the instrumentation rising, the demands on the calibration technique are also likely to grow. There is no point in buying or developing a measurement system that operates at a level that cannot be verified.

The need to discuss calibration is particularly important where the overall system can have an impact on the measurement system. For example, load cells can be supplied fully calibrated. However once they are attached to a silo, the connected pipes, the support frame, and the silo itself will deflect when loaded and will also expand with temperature increase. Connected pipes may partially support the load (force shunting) and thermal expansion will add additional forces. Stirrers and agitators will cause vibration. Through calibration, the deflection and force shunting effects can at least be accommodated, and other techniques can be employed to reduce effects such as thermal expansion and vibration.

In determining what and how much needs to be measured, the user should be aware that it is not always cost effective to reduce the number of measurements. In one troubleshooting project we specified several pressure measurement points. The client questioned the number of measurement points as ‘they knew’ that the pressure-drop across a particular element was negligible. As you would expect (or I wouldn’t be telling the story), the problem in the pneumatic conveying system turned out to be a surprisingly large pressure drop across that ‘negligible’ component. The other side of the story is that on occasion we have specified measurement points that simply confirmed what everybody knew. The only additional information gained being that we now knew as well.

John Pugh is Head of Centre for Industrial Bulk Solids Handling Glasgow University, Glasgow, Scotland. A professor in the School of Engineering and Computing, Professor Pugh has pursued an interest in instrumentation and measurement for over 30 years, with more than 15 years in the area of instrumentation for bulk solids handling. Pugh is vice president (Learned Society matters) of the Institute of Measurement and Control in the United Kingdom, and chair of the Institute’s Learned Society Board and Measurement Science and Technology Panel.