Near Infrared and acoustic chemometrics monitoring of volative fatty acids and dry matter during co-digestion of manure and maize silage

In this study, two process analytical technologies, near infrared spectroscopy and acoustic chemometrics,
were investigated as means of monitoring a maize silage spiked biogas process. A reactor recirculation
loop which enables sampling concomitant with on-line near infrared characterisation was applied. Near
infrared models resulted in multivariate models for total and volatile solids with ratio of standard error of
performance to standard deviation (RPD) values of 5 and 5.1, indicating good on-line monitoring prospects.
The volatile fatty acid models had slopes between 0.83 and 0.92 (good accuracy) and RPD between
2.8 and 3.6 (acceptable precision). A second experiment employed at-line monitoring with both near
infrared spectroscopy and acoustic chemometrics. A larger calibration span was obtained for total solids
by spiking. Both process analytical modalities were validated with respect to the total solids prediction.
The near infrared model had an RPD equal to 5.7, while the acoustic chemometrics model resulted in a
RPD of 2.6.