Log in
en | fr | nl
  • Factsheets
  • In-line moisture monitoring in drying operations

In-line moisture monitoring in drying operations

  • Public
In-line moisture monitoring in drying operations

In-line monitoring of moisture content: an overview

NWGD Symposium
 ‘Numbers tell the tale'
30 Nov 2015

Author: Erik Esveld, Food & Biobased Research, Wageningen UR.

The product final moisture content is a key parameter in any drying process and determines the product quality, its technical performance and storage stability. Detrimental fluctuation or heterogeneity in product moisture content, due to inherent variability in input stream and air conditions, can be corrected via process control only if relevant in-line measures are available. Next to temperature the moisture content of the processing air must be measured, and even better, the product moisture content. An in-line moisture measurement allows for fast feed-back control and continuous quality monitoring, unlike the laborious gravimetric or Karl-Fisher determinations. There are many different measurement principles applied in commercially available moisture sensors, each with different operational characteristics.  On behalf of a Dutch working party on Drying (NWGD), an overview is compiled that discusses main features of the different sensor technologies. The overview can be assessed at the NWGD website and can be sent at request.

There is a huge variation in operational principle and complexity for air humidity measurement. Some are based on an equilibration temperature, such as the classic Psychrometer (based on the wet and dry bulb temperature) the accurate Dew point meter (based on  cold surface condensation) or the (cheap) LiCl sensor (based on the temperature for which the RH = 11%). The change of material property in response to moisture sorption is exploited in hygroscopic polymer sensors. Their applicable temperature range can be as high as 180°C, but condensation and contamination is a concern for these sensors. For even higher temperatures and moisture levels near to saturation, the electrochemical zirconium oxide sensor is applicable (allthough stricty an oxygen sensor). On the other hand, for very low dew points, the aluminium oxide sensor with nanopores condensation is the method of choice. A good way to circumvent the hot and contaminating measurement conditions is a dryer, is to sample the process air via low pressure sampling lines. The absolute moisture at ow pressure can be very accurately be detected with diode IR laser absorption, which is also applied to detect smouldering via carbon monoxide traces. 

For inline product moisture content, there is a choice between near infrared monitoring of the surface moisture and the high-frequency / microwave sensors for the whole product moisture content. Water has a distinct absorption in the NIR spectrum (1940nm)  and its ratio with respect to other reference wavelength can be used as a measure. NIR moisture sensors can measure at large distance (40cm) and even trough sapphire windows. However, the NIR reflectance is quite sensitive to variations is composition, colour and surface texture, and does not penetrate more than 50 microns in a dry product. High-frequency and microwave penetrate the whole product and the electromagnetic transmission is notably influence by water, which has a very high dielectric constant. But also here calibration is a necessity and only valid for that product range for which it is calibrated. Many different sensor geometries and high-frequency measurement technologies are applied, adapted to the form of the material flow (powder, granular, chips, board; free flowing, transported in a duct or on a belt, side stream sample batch, or continuous in main stream etc.)   In all cases is the signal response proportional to the amount of water per volume and not per mass. For true moisture content determination, the density (mass and volume) should be measured independently. A few application however, a certain amount of density variation of the material is acceptable. Particularly for microwave resonance sensors one can correct for density variation by including information stemming from the dielectric loss.


Near Infrared moisture sensor above a fried product measuring the surface reflectance of the water resonance band at 1940nm with regard to two reference bands (Process Sensors Ltd.)


Microwave resonance sensors (tubular and planar versions shown) measure the internal product moisture as function dielectric constant and dielectric loss at GHz frequencies. Correction for bulk density variation is possible when correctly calibrated. (Inventech, TEWS)

More information can be found in the handouts of the presentation given by Erik Esveld.

Commenting only available for logged in users