On February 13-14, 2001 a Near Infra-Red (NIR) sensor was tested at an OSB mill to measure liquid PF resin level of furnish on two forming lines. This was the final step of the Forintek project titled “Online Measurement of Resin Distribution in OSB”. The NIR sensor was a commercially available NIR moisture gauge modified to measure liquid PF. An earlier pilot plant study [1] showed that this modified gauge could work in online situations to show trends in resin level using either liquid PF, powder PF or MDI (isocyanate). In this study, the sensor was not calibrated and gave measurements in signal strength units only. This data was later calibrated using a trend equation relating signal strength values to resin level based on pilot plant trials.
In the mill, the sensor was set-up at two separate locations including the top and bottom face forming lines. Once set up and logging data, the mill made several changes to the resin level from 4-5%. The results clearly showed a good correlation between sensor data and the corresponding resin level changes.
To further test the sensor, the mill changed the solids content of the resin from 55% to 45%, while maintaining the same resin application level, to determine whether the resulting moisture content change on the furnish would affect sensor readings. This caused the sensor signal to significantly drop in value, roughly equivalent to a change of one percentage level, and showed that the sensor is vulnerable to changes in moisture content. In addition, the earlier pilot plant study showed that changes in furnish species and fines content could affect readings as well. As discussed in the earlier pilot plant evaluation report, it may be possible to eliminate the influence of non-resin furnish conditions if these factors could be singled out and subtracted from the measurement data. One suggestion is to mount a second sensor before the blender (no resin condition) to obtain base-line data. Many mills already have NIR moisture gauges mounted on the dry bins which could possibly be modified for this purpose.
This study has shown that NIR sensor technology has the capability to measure the level of adhesive resin in furnish under online mill conditions. Further work is recommended to eliminate the sensor’s sensitivity to other (non-resin) variables in the furnish.