From Seed to Sugar: A Vertically-Integrated Model for Small-Scale Turbinado Sugar Production from Organic GMO-Free Beets

The first phase of our work was to obtain seed for and then crop and evaluate some non-gmo varieties of sugar beets. Despite difficulty obtaining non-gmo varieties, we were able to obtain four varieties, including three sugar beet strains and one fodder beet strain. Fodder beets are cousins of sugar beets, rarely grown in the U.S. and have somewhat lower sugar content by weight but larger size and greater ease of harvest. In 2011 nearly all of our seedbed was drowned in heavy spring rains. By the time acceptable planting conditions recurred, the season was so far advanced as to greatly limit yield. Therefore the project was extended an additional year. Though heavy rains did recur somewhat in 2012, an acceptable performance of all 4 non-gmo varieties was observed. The crops were carried through to storage and processing.

In 2012, our trial plot took place on one acre of Vergennes B soils with gently sloping terrain.  Initially we had planned to spread the trial over two soil types but after 2011 we lost access to the plot with lighter soils as that was rented land.  Greg Roth at Penn State advised that the Vergennes B soils should be adequate to support beet production.  The soils in this plot tested for optimum to high levels of Phosphorus, Potassium and Magnesium in 2011 and 5 tons of beef manure to the acre were recommended per our crop plan, which was spread the fall of 2011 prior to the commencement of the trial in 2012.
In April 2012, we laid out a trial of four varieties of field beets (all beta vulgaris):  A Scottish Fodder Beet sent to us by a farmer there, a non-GMO sugar beet sold as a deer bait crop by the company MonsterBuck, "Giant Half-Sugar" beets sold by R.H. Shumways, and an experimental non-GMO sugarbeet shared with us by the company Betaseed USA.  All in all obtaining verifiable non-GMO varieties was very challenging.  In terms of reliable, affordable access Shumways stands out and has actually increased its offerings of seed since the beginning of this project.  The field was roughly 120' x 350' and had three rows of one variety at 3' row spacing followed by 3 rows of the next variety, with the sequence repeating every 12 rows.  The field was plowed in the fall and harrowed in the spring and worked down to a fine seedbed.  The rows were laid out with strings and planted with an earthway push seeder.
Following planting on April 10th the weather trended cold and wet which hurt the emergence of beets on the lower, soggier portion of the field but not the upper part.  Where beets succesfully emerged in the more poorly driained portion they still exhibited poor growth and size, especially in areas where water pooled during heavy rains.  Thinning was accomplished with hand hoes, and weeding with a horse-drawn cultivator, hoes, and a walking rototiller.  By July the beets were very well established and they continued to grow new leaves and larger roots until November.  We delayed harvest until mid-november to maximize sugar content as this is a recommended practice.  We harvested the beets by loosening the roots with a "beet lifter," a sort of walking plow that has an offset sole that helps loosen soil around the roots and makes them easier to pull by hand.  A middlebreaker plow would accomplish much the same effect.  After loosening we removed tops in the field and carted the beets to the farmyard.  Without mechanization hand harvesting beets is a fairly big job.  We estimated our progress to be about one ton harvested per laborer per day.  In our experience, with Mangels this rate can go up to about two tons, fodder beets about 1.5 tons, and with true sugar beets it is 1 ton at best.  Often, even after the beet lifter is used, stubborn beets need to be loosened with a fork which greatly slows harvest.
Our data from this trial follows in a table.  We were pleased to see that the varieties performed within an acceptable range and had high sugar content.  We measured the sugar content by extracting the juice and using a the balling scale on a hydrometer.

After harvest the beets were stored in a "clamp," which is a low-cost root storage solution consisting of a shallow pit covered with straw, wood chips, or other insulative material. The beets are piled in a heap within the clamp and are well covered. In December 2012 we dug up the beets one variety at a time in order to begin processing. A few beets were damaged by freezing and thawing but further into the pile the vast majority had held quality well.

Our processing system was set up based on conversations with distillers and other experimenters who had worked with beets for sugar on a small to moderate scale.  Our setup included a washing step involving soaking batches of beets in 55-gallon drums full of water (which had to be changed intermittently) , followed scrubbing and rinsing in a large restaurant-type sink with a sprayer.  Following washing the beets were chopped into very large pieces and run through a commercial juicer (a Breville Juice Fountain Elite, 1000 watts) which separates the beet juice from the pulp.  This method, essentially a centrifuge method, was recommended by my sister Alexis Andrus who was working as a lab tech at the dairy science division at Cornell, who consulted with a senior chemist colleaugue on my behalf about scale-appropriate sucrose-extraction methods.  The initial results were encouraging.  The juice could be produced at a rate of about one gallon per minute with a crew of two at task.  Furthermore, the resulting juice tested very high in sucrose, over 20% in some cases.  But although beet juice is very sweet but we also noted it to contain a powerful off-flavor with a strong lingering aftertaste, which can be best described as over-ripe swiss chard.  The aftertaste had a chemical-like burn quality to it.  This wasn't anticipated.  
Nevertheless, we carried this juice through the process of boiling and crystallization.  At this point our sugarmaking approach resembles the making of sugar from maple sap, with a shallow pan and a heat source.  In our case we used a commercial restaurant stove and shallow "hotel pans" as boiling pans.  This was intended as a simple boiling system to refine our approach, with the possibility of moving the operation to a true outdoor maple syrup arch when we were ready.  We boiled the beet juice solution and continued boiling to concentrate the solution.  Being already very sugary it did not take very long.  Once the boiling point reached 260 degrees F we began the cooling phase, adding a handful of table sugar to each pan and stirring once the solution reached about 250 degrees F in order to obtain large crystals.  We did achieve crystallization but the crystals never came out very large.  Also, the flavor became more concentrated as the syrup became more condensed.   Once the crystallized mass was cooled the appearance and consistency was much like roofing tar, with many crystals of various sizes ranging from 1/16" to 1/8" embedded in the mass.  We removed the sugary mass from the hotel pans put them through the centrifugal juicer and in so doing rid them of some of the off-flavors in the molasses, but even so the crystals were too small and the molasses content too high, rendering the finished product bad-tasting.  If we had been able to obtain very large sugar crystals the molasses effect might have been less, but after several attempts we were not able to achieve this and began to reevaluate our approach.
During processing we noted that the fodder beet variety had less pronounced off-flavors than the other three sugar beet varieties. However sugar crystals made with the above method from the fodder beet were still unacceptable.  We processed approximately 3 tons of beets yielding about 400 gallons of beet juice and about one ton of damp beet pulp before giving up completely on this approach. 
During processing we had noted that the fodder beet variety had less pronounced off-flavors than the other three sugar beet varieties. However sugar crystals made with the above method from the fodder beet were still unacceptable. 
In light of discouraging results with the juicer, we changed strategy decided to experiment with a diffusion method, which involves drawing off the sugar in the beets by steeping in hot water, rather than by centrifuge. We used a 60 quart hobart kettle and filled it with large beet slices, sliced to about 3/4" to 1/2" thick.  We filled the kettle with slices and added hot water, then brought the entire kettle to a boil and then removed heat.  After the kettle cooled we strained the now syrup-like liquid off the slices.  This liquid tested at about 10% balling scale.   We were very pleased that this method resulted in a nearly clear syrup with only mild beet flavor. However when we boiled this syrup down it became noticable that the syrup was laden with what we presume is pectin, making the solution slimy and difficult to work with and also apparently resistant to crystallization, like a jelly. The off flavors were also still present with this method though not nearly as pronounced as in previous batches. Consultation with the UVM Chemistry Department and to our technical advisor Ginger Nickerson at the UVM Center for Sustainable Agriculture did not provide a solution to the problems encountered in our process. Having encountered several barriers to producing the 1000 lbs of sugar we originally set out to make, we decided to conclude processing after this single trial of the steeping method.

Our processing efforts did produce some turbinado crystal sugar. However the presence of strong off-flavors indicated we were still very far from having a product that would appeal to the public, even a committed buy-local public. We had also produced beet molasses as a by-product but the off flavors were even stronger in the molasses. In addition our production method was also very slow, and it appeared that more impurities were being carried through into the beet syrup than is ideal, to general detriment to the process. Therefore we elected to draw our own conclusions as to the marketability of the products in lieu of attending markets and collecting surveys.

• Seed to Sugar Crop Data
• Seed to Sugar Powerpoint Presentation