Skip to content

Ontario Grain Farmer Magazine is the flagship publication of Grain Farmers of Ontario and a source of information for our province’s grain farmers. 

Nothing but good grain


taking the good with the bad is just a part of life, never mind farming. Except in the new world of small grains, researchers are finding a way to take only the good – thanks to a seed sorting prototype currently being tested.


In early 2012, the University of Saskatchewan was the first place in North America to install a BoMill TriQ 30 seed sorter. According to Swedish manufacturers, the machine is able to separate small cereal grains damaged by disease based on crude protein content using near infrared reflectance (NIR) technology. Dr. Tom Scott of the University of Saskatchewan says his research shows the unit can sort approximately 30,000 seeds per second (3MT/hr) of wheat, barley or durum seed. He also believes the system can sort out very low crude protein seeds that can indicate fusarium damage. He thinks this is a particularly important challenge to address in Canada where roughly 20% of the grain crop is downgraded to feed because of mycotoxins which isn’t good for animals. Farmers are potentially losing $800 million annually.

PHOTO: Bomill triq 30 seed sorter.

“Crude protein is one of the most important traits in cereals not only from a nutritional value but also there’s many important relationships to functional values such as bread or pasta quality,” said Scott. “The benefits are primarily for high value products but, as a feed researcher, I saw the opportunity to be able to separate out more unique feed from a sample of grain because typically that sample can be quite variable.”

sorting the grain
In removing the worst 10% of kernels based on crude protein, Scott found he could upgrade a feed wheat to the number 1 class, recovering approximately $80/MT. He also wants to turn the negative impact of protein content variability in grains, which is as high as 6% among kernels in one head of wheat alone depending on weather, into a potential positive. Scott’s team at the Canadian Feed Research Centre in North Battleford, Saskatchewan hopes they can use the sorter to take advantage of that variability, using it to positively impact grain pelletizing processes and improving animal performance through higher quality feed.

“Now I can take fractions of wheat or barley that are different in crude protein or other characteristics, process them, and see which one optimizes the feed value for animals,” said Scott. “You could remove mycotoxins, other damaged seeds, you can guarantee much higher quality food or malt, and provide more consistent production, so it’ll not only be good for agriculture, it will be extremely good for research.”

Although Scott is a livestock feed expert, he isn’t limiting his research with the seed sorter just to improving livestock feed production. He says he’s extremely interested in whether the sorter can determine amylaze activity in barley seed, which could help malters run more efficiently by grouping slow sprouting kernels in separate batches from quick ones rather than altogether like they do today. He also wants to investigate whether the machine’s NIR technology can be calibrated to sort barley based on beta-glucan, which is a nutraceutical able to lower cholesterol.

Archie Wilson, General Manager of C&M Seeds and Palmerston Grain, was also attracted to the technology as a means of potentially creating niche products to target specialty food markets.

“There’s possibly going to be more segmentation of grain deliveries based on what end-users want long term,” said Wilson. “This could be a tool to help you get there and get market share with certain end-users.”

Wilson believes there could be a lot of value in the future in selling grains based on their crude protein values, in the same way livestock rations are often marketed today. Dr. Rex Newkirk with the Canadian International Grains Institute (CIGI), sees a similar opportunity to use seed sorting technology to offer a wider range of products to buyers. He is particularly excited about opportunities in the pasta and bread flour market, where high amylase activity inevitably leads to terrible dough production.

“What captured my imagination was the opportunity to do things like sort wheat based on protein content,” said Newkirk. “It intrigues me that technology has come this far, that you can actually look at individual seeds, pick them for what you want, and select a product.”

potential for change
Newkirk says that if the system works as well as they think it might, this technology could be a game changer. Premiums are already offered for wheat that has crude protein values exceeding 13.5% as determined by NIR technology. With the right marketing and enough time to gain full confidence in these systems, there’s no telling how much of an impact it will have on the industry. However, as both Newkirk and Scott witnessed in 2012, it will be the bad years that advance the technology the fastest. Scott could only test the sorter on two fusarium infested wheat samples  because disease pressure was so low.

“It’s one of those things where it will be extremely valuable in those challenging years but there will be other years where it won’t be quite as active,” said Newkirk. “But I hate seeing people make an investment in something before we really know how well it will work in the industry and so this is a great intermediate step.”

Wilson is a little more reserved about the full potential impact the seed sorter might have on the industry, especially in the near future. He travelled out to see the sorter in action and says he was generally impressed with the technology but not the $300,000-400,000 price tag each unit comes with. In a facility operating at economies of scale, 10 units would have the capacity to sort 250,000 MT/yr and the estimated ownership and operation cost would be $10/MT. But both Scott and Wilson point out that there’s a considerable amount of additional infrastructure required to handle the grain post-separation. Scott estimates the increased cost of grain handling brought his final total up to $580,000 for one unit and they’re still not fully automated. He suspects the unit is best suited to operations that already move a lot of grain, like seed cleaning companies. At its current operating capacity, Wilson feels the purchase cost is too high to apply within his own commercial operations right now.

“I think the technology is neat, I think it can lead to something, I’m just not sure how soon that will be,” said Wilson, “but I think once you get it out there, people do learn different ways to use and adapt it.”

Newkirk agrees that the machine is well placed in a research environment right now. The potential value that could be captured by further adapting the current set up for a wider variety of applications, such as use in separating corn or soybeans, is obvious. Though some progress is being made sorting flax seed, and the next logical target is pulse crops, there is still a lot of ground to cover before larger kernels like corn can be processed.

Producers who have downgraded wheat, barley or durum they’d like to have tested can contact Dr.Scott at Project funding was provided by the Alberta Crop Industry Development Fund (ACIDF), Saskatchewan Ministry of Agriculture, Canadian International Grains Institute, Western Economic Development and BoMill.

First, a grain sample is analysed in the machine by software that estimates the range of protein values available. The operator identifies the desired ranges and the full load of kernels are fed into the centre of a spinning cylinder. Once inside, centrifugal force pushes each grain into kernel shaped pockets which line the interior surface of the cylinder. These pockets then pass through the near infrared detector, which shines a beam of red light on each kernel and determines the seed quality through a quick analysis of reflected light patterns.

The result of the scan is read by the ejection unit which triggers one of three high-pressure valves. Each of these valves blows the kernel out of its pocket and into either the high, mid-range, or low quality collection channel, according to the protein limit ranges set by the operator.

Currently the machine can sort 30,000 kernels per second for commercial use, but the design allows for modules to be stacked and connected so multiple units can sort in tandem to increase output capacity.  •


In this issue: