Cr. Hurburgh, IDENTIFICATION AND SEGREGATION OF HIGH-VALUE SOYBEANS AT A COUNTRY ELEVATOR, Journal of the American Oil Chemists' Society, 71(10), 1994, pp. 1073-1078
For three harvest seasons, 1989-1991, a whole-grain near-infrared tran
smission analyzer was used at a large Iowa elevator to measure oil and
protein content of unground soybeans. In the third year, soybeans wer
e physically sorted on the basis of the sum of protein and oil. The su
m is a direct measure of processor revenue potential. The top 23% (app
roximately 150,000 bu or 4500 MT) of soybeans were isolated from the l
ower 77%. The ''high-value'' beans (with sum 0.7 percentage points abo
ve the average) were isolated for future sale to a processor. The anal
yzer performed accurately and required about 1.5 min per test. Compare
d to the normal 1.0 min per test for moisture measurement, this caused
an additional delay of about 30 s per load for farmers, which is a si
gnificant cost over the 100-300 loads per day received by the elevator
. Because this was a new program for elevator employees, errors in dat
a transcription and communication reduced the value difference of the
high-sum beans from its theoretical maximum of 1.2 percentage points,
but the high-value beans still had a theoretical 10 cent/bushel advant
age over average unsegregated beans delivered to that elevator. Cost a
ccounting showed the testing and segregation to cost about 2-3 cents p
er bushel. Testing and segregation is feasible in the U.S. grain marke
t.