ARTICLE FEED & ADDITIVE MAGAZINE April 2022 51 similar to the variability between reference vs. NIR (Figure 1). PRACTICAL EXAMPLE: USING HHNIR TO MONITOR QUALITY OF INCOMING SOYBEAN MEAL The portability of the HHNIR lends itself to being at locations other than in a laboratory, for example at the intake of the feed mill for raw material analysis. In the example below (Table 2), a HHNIR instrument was used to measure the crude protein (CP) content of incoming soybean meal, which then provided an actionable insight to either accept or reject the incoming load from the supplier. The minimum acceptable crude protein level was 46.5%. All the incoming loads were scanned with the HHSoybean Meal Crude Protein (N = 20) Table 1. Hand-held NIR, Benchtop NIR and Wet Chemistry results for 20 soybean meal samples compared to the reference method *SEPc: standard error of prediction corrected with the bias Average Min Max Standard Deviation SEPc* Reference 47.35 45.22 49.62 1.07 - Benchtop 47.00 45.17 48.76 0.86 0.58 HHNIR 47.10 46.31 48.13 0.51 0.69 Secondary Wet Chemistry 47.10 44.87 49.77 1.21 0.50 Figure 1. Crude protein content as measured by hand-held NIR, benchtop NIR and wet chemistry compared to the reference method
RkJQdWJsaXNoZXIy MTUxNjkxNQ==