NIR moves moisture measurement on-line

Over the last twenty years or so, near infrared (NIR) has firmly established itself as a key process monitoring and control technology, measuring a variety of parameters including fat, protein and thickness as well as moisture.

New applications are being developed all the time as industries become aware of its benefits as a safe, hygienic, non-contact technology, and start to realise the clear economic advantages of the close, real-time control it offers. As a result, the technology is revolutionising processing lines in industries from food to chemicals, by replacing time consuming laboratory QC checks.

By far the largest group of NIR applications in use today is concerned with moisture measurement. Any particulate, fibrous or flaky material can be successfully monitored, and since there are three different water absorption bands in the NIR spectrum, of widely differing intensities, moisture levels from 0.01 to 90% moisture can be measured using the same basic technology.

Today's on-line NIR sensors combine a highly accurate optical system with a microprocessor-based control unit. Measurements are unaffected by variations in product particle size, colour or composition, and by using reference measurements and dual detector technology to effect a ratio measurement, sensors are inherently insensitive to changes in ambient or process conditions such as temperature, light intensity, humidity, the height of the product as it passes under the sensor, dust build-up on optical surfaces and so on. Stability of the measurement over time is typically in the region 0.1% or better.

The basic principle of operation of the NIR backscatter gauge is shown in Fig 1. Light from the quartz halogen lamp is focused into a parallel beam and projected onto the product. A filter wheel rotating in the optical path contains NIR optical interference filters which are designed to transmit narrow bands of infrared energy at specially chosen measurement wavelengths, specific to the measurement application. Wheel rotation speeds of 50Hz are typical, effectively providing a continuous measurement.

The pulses of light landing on the product are both scattered and absorbed to some extent. A proportion of the scattered light is focused by the collecting mirror onto the lead sulphide detector which generates signals proportional to the reflected intensities at each wavelength. These signals are manipulated using algorithms specific to each application, to give an output proportional to the parameter being measured, such as moisture content.

In order to develop the algorithm used for calculation of the actual moisture content, raw filter signals are collected from a matrix of samples and used to generate a robust, reliable measurement, linearly proportional to moisture content.

Then finally, since NIR is not an absolute measurement, it needs to be calibrated against a primary reference usually a laboratory technique.

New on-line NIR instrumentation is now just coming on to the market, which is not only insensitive to variations in ambient and process conditions, but, which needs none of the lengthy on-site calibration often associated with such devices.

On this and following pages, a series of case studies demonstrates the wide-ranging capabilities of this remarkably versatile technology.

Dr Ian B Benson is an industry product specalist with Infrared Engineering.