CORN PROCESSING
Probe-Sampling of Grain Corn and the Code of Practice
By Brian Doidge, Ridgetown College, University of Guelph

• Grade No. 1 CW/CE 2.0% CCFM
• Grade No. 2 CW/CE 3.0% CCFM
• Grade No. 3 CW/CE 5.0% CCFM
• Grade No. 4 CW/CE 7.0% CCFM
• Grade No. 5 CW/CE 12.0% CCFM

Therefore, for the 90 loads of corn sampled (30 at each of three processor locations), because of the CCFM result:

• The processor probe procedure had 20 (22.2 per cent) that were lower grades than for hand-sampling.
• The CGC probe procedure had 14 (15.5 per cent) that were lower grades than for hand-sampling.
• Hand-sampling had three (3.3 per cent) that were lower grades than for the processor probe procedure.
• Hand-sampling had two (2.2 per cent) that were lower grades than for the CGC probe procedure.

Because none of these processors can utilize (and therefore, do not accept) corn lower than Grade No. 3, and because of the large discount applied against Grade No. 3 corn versus Grade No. 2 corn at processor locations (i.e., $3 - $5/mt in many instances), down-grading from No. 2 to No. 3 on account of CCFM content is of most concern.

Of the 20 samples (22.2 per cent of the 90 samples) that were lower graded by processor probe procedure versus hand-sampling:

• 13 moved from a No. 2 down to a No. 3.
• Two moved from a No. 2 down to a No. 4.
• Four moved from a No. 3 down to a No. 4.
• One moved from a No. 3 down to a No. 5.

Of the 14 samples (15.5 per cent of the 90 samples) that were lower graded by CGC probe procedure versus hand-sampling:

• 11 moved from a No. 2 down to a No. 3.
• Two moved from a No. 3 down to a No. 4.
• One moved from a No. 3 down to a No. 5.

A project steering committee reviewing these and other CGC results suggests the following:

1. Spreaders on loading spouts at the farm would spread the corn and fines more evenly throughout the load, rather than concentrating fines in the centre of the load when loaded without benefit of a spreader. Even distribution of fines would improve samples and make them more representative of the load.
2. The study suggests hopper-bottom trucks and trailers provide better quality and more representative samples compared to flat-bottom trucks and trailers.
3. Probe design, age and set-up affects breakage, depending on distance from probe to scale-house, the number of bends and material of the sample transport piping.
4. Detailed inspection of probe equipment should be conducted several times during the season with attention on efficiency of probe door sealment, to prevent vacuum leakage when the sample is pneumatically transported from the load to the scale-house.
5. The more probe samples taken per load the better, and the closer the result will be to hand-sampling.
6. The higher the ratio of the number of probes on the outside of the load compared to the number of probes in the centre of the load, the more representative will be the sample. For example, four probes taken from the outside (i.e., four corners) to every one probe from the centre will provide a more representative sample than only twoprobes from the outside to every one probe from the centre (i.e., a diagonal across the truck). Recommended procedure in the U.S. is a ratio of 6:1 or 8:1.
7. The probe must go straight down into the load, not on an angle toward the centre of the load.
8. Hand sampling to calibrate probe-sample results perhaps should be conducted randomly (daily?) during the entire season to verify probe sample results and to signal to shippers that loads are indeed being inspected closely.
9. Processor grading personnel should be trained by CGC personnel on proper techniques and equipment for hand-sampling.
10. Probe-sample CCFM results might be adjusted frequently throughout the season based on randomized hand-sampling calibration results to compensate for changes in the corn profile being received.
11. Probe sampling CCFM ratings might better be adjusted as a percentage of the probe CCFM result rather than an absolute number. For example, 0.2 per cent lower at a probe CCFM reading of 2.2 per cent; 0.7 per cent lower at a probe CCFM reading of 3.4 per cent, etc.

Processors have responded to the CGC project and steering committee suggestions with these actions:

• King Milling, even before these results were in, had conducted its own comparison testing of probe sampling versus hand-sampling, had compared results and created its own “line of best fit” (i.e., adjustments necessary at various probe CCFM readings in order to align with hand-sampling readings), and then applied this adjustment as a standard “correction factor” to CCFM results.
• Casco “has taken the approach that we are willing to take additional probes from the outside of the truck in order to parallel hand-sampling. Casco feels that if they take one sample from the centre of the truck and three to four from the side of the truck this should help eliminate the discrepancies for CCFM.”
• Commercial Alcohols’ corn receiving and grading criteria are similar to Casco’s, as Casco purchases all corn for the ethanol plant in Chatham. Results of this study put both producers and processors in an awkward situation concerning sampling of grain corn at processor delivery locations. Producers should know that these processors have agreed to abide by the Code of Practice for Trade in Corn & Canola in Ontario, which governs deliveries at primary elevator locations in the province. The code states that “the elevator’s representative receiving the corn or canola, or the producer, or the producer’s agent, may require a sample of corn or canola to be taken. Such sample shall be accepted by all parties as a fair and representative sample of the corn or canola from which it was taken.” The code makes it clear that the producer and the processor must agree that the sample taken from the load is a fair and representative sample. So, when grain corn grading depends on samples gathered by probes, before accepting a probe sample as representative of a load suitable for testing (especially if the grading result is or might be in dispute) producers should note carefully:
• that the probe went into the load straight down, not angled into the centre
• that there were four probe samples taken from the outside versus 1 from the center
• that probe sampling results in a CCFM percentage about 0.75 per cent higher than hand-sampling, even if both the first two criteria are present. At the request of the producer or the producer’s representative (this would allow a trucker to request), or the representative of the grain receiver, a representative sample of the grain corn shall be taken by probe-sampling procedure as outlined, or by hand-sampling using accepted CGC practice.

It is suggested that prior to resorting to arbitration of disputes concerning grading on account of CCFM content, producers and processors negotiate acceptable resolutions, considering the results of the March 1999 CGC Probe-Sampling project.

1