Since our Getting Started with Kernza® post, we’ve gone even further into understanding just how and why Kernza behaves the way it does. We’ve done rigorous testing: milling, sifting, blending, baking, comparisons, you name it.

Our aim has been to get a solid idea of how Kernza truly compares to conventional and commercial flours, and, ultimately, aid bakers in understanding what to expect when working with Kernza.

We’ve even gone as far as to send flour samples to a grain analysis lab for rheological testing (Rheology: The study of the flow of matter; used to determine the many aspects of dough quality). Working with this grain is becoming more intuitive to us by the day. The text that follows is a full description of everything we’ve been up to in relation to baking with Kernza. I make a few educated guesses and predictions that I have not had an opportunity to confirm, and there is plenty of new information here for you to generate your own hypotheses.

Short on time? Here's our summary:

• Kernza inclusion percentage with Hard Wheat: 15-25%
• Kernza inclusion percentage with Soft Wheat: Up to 50%. 75% in some cases
• Optimal hydration percentage (100% Kernza): 63-67.5%
• Optimal hydration percentage (Kernza/Turkey Blend): 66-70%
• Optimal kernza mixing time: 6 minutes
• Kernza has starch issues: not enough. Will not set loaves very well without inclusions
• Kernza's gluten structure: Very viscous and somewhat extensible. Not very strong or elastic. 
• Shares many similarities with soft wheat
• Kernza struggles to keep fats emulsified. May benefit from cutting back in recipe. 
• Sifting, oxidization, vital wheat gluten, ascorbic acid, or additional starch may help boost Kernza's structure formation
• Kernza's flavor profile: Honey, cinnamon, maple syrup, nuts, vanilla, butterscotch, almond extract, brown butter 
• Kernza's best applications: scones, pancakes/waffles, quick breads, some cakes, muffins, whole grain preparations

Given the amount of data in this report, it is all broken down into sections making it easier to sift through should you deem only a few sections useful. We feel that this information will give you a substantial jumping off point regarding working with Kernza, and will additionally provide a great reference for many other culinary experiments you may wish to conduct. 

Table of Contents

1. Milling/Sifting
2. Perennial Test Kitchen Bake Tests
   
1. Kernza and Turkey Red Wheat
   
2. Commercial Flour Comparisons
   
3. Experiments in Versatility
   
4. Kernza and Soft Wheat
   
5. Commercial Flour/Versatility Bake Trials
   
6. Final Thoughts on Baking
   
3. Rheological Flour Analysis
   
1. Moisture, Ash, Protein, & Falling Number
   
2. Extensograph
   
3. Farinograph
4. Rapid Visco Analyzer
   
4. Final Thoughts

Milling/Sifting

What is the look and feel of Kernza: coarse, fine, soft, tough? How do these observational properties compare to similar products from commercial brands? How does this information correlate with finished baked goods? These questions formed the basis of our first round of testing. We needed to know precisely in what ways - before baking -  Kernza flour differs from other grain flours, and how we could use this information to optimize our flour.  

Right away, it is easy to notice Kernza’s size. It is much smaller than wheat and most other related grains. Being smaller in size means there is much more bran in relation to the starch/gluten protein containing center called the “endosperm.” Each individual particle of bran in our flour also tends to appear a bit larger than in other whole grain flours. A whole grain flour from, say, King Arthur, may contain a much more constant and generally smaller particle size. Though these differences do not diminish flavor, larger bran size does have an impact on ease of baking, as it has the potential to disrupt starch/gluten formation in a variety of different ways. This, in turn, could yield flatter loaves of bread. In quickbreads, cakes, and other “low-structure” doughs, however, larger bran is typically not as much of an issue, if at all. 

Through milling and sifting, we were able to incrementally manipulate and test particle size in an attempt to determine what our optimum flour may look like on paper, not just through observation. Details of this experiment are shown below. Each chart contains sifting data for each flour used, including Kernza dried to 10% moisture (our pick when compared other Kernza samples milled at different moisture levels). The mesh numbers (#5 mesh, #10 mesh, etc.) relate to the particle size each lets through; #5 contains wide gaps, and lets through large particles, #230 has extremely small gaps, and lets through only the smallest particles. The way we determined (and continue to determine) flour quality, was by measuring the weight of flour that passed through the #60 mesh sieve. The flour resting in the #160, #230, and Bottom Tray had ideal characteristics (to us and in relative accordance to a commercial flour standard for all purpose flour). As you can see, The Kernza we milled revealed that just over 61% of its flour passed through the #60 mesh, while large scale whole wheat flours sat around 85% or more. Most All Purpose flours will be around 99%! Keep in mind, these are large scale producers with a greater number of resources at hand. Smaller stone millers, such as ourselves, generally have particle sizes slightly less than 85%. It should also be noted that since these tests have been carried out, we’ve had a chance to “dress” (sharpen) our millstones and produce a much finer whole grain flour, but it is still not quite 85%. 

Kernza sift analysis

Commercial flour sift analysis

Though Kernza will generally be more coarse than flours from these large scale operations, its post milled texture is also light, fluffy, and airy, whereas wheat’s flour particles are somewhat hard and dense. This is an interesting attribute, as it seems to mitigate the coarseness of the flour, making it seem like it’s slightly larger particle size is less of an issue. More milling tests have been conducted since, where we determined exact mill gap sizes and flow rates as a way of standardizing our product. Given the nature of stone mills, however, the process is ever changing. It really is something that requires daily observation in order to maintain control of standards. Getting a set of small sifting drums is a tremendous help for this. Our process is simply to add 150 grams of flour to the sieve stack, and vigorously agitate it for five minutes. After agitation, flour remaining in each sieve is weighed, and the data is used to define our standard.

Back to top

Perennial Test Kitchen Bake Tests

Our intent in these applied tests was to gain a better understanding of the effect kernza had when blended with other flours, and make any comparisons between the baking characteristics of these blends and those of other stone milled and commercial flours. Though more testing is yet to come, these initial tests cover a range of applications amongst different varieties of wheat flours (soft wheat, hard wheat, stone milled, roller milled, fresh, aged), and the results provide a great foundational piece to the puzzle when it comes to intuitive baking with Kernza. 

Kernza and Turkey Red Wheat

We wanted to see exactly what effect different inclusion rates of Kernza had on end products, and just how well one, if any, compared to pre-purchased, well known flours available for retail. 

The first round involved baking a simple whole wheat sandwich loaf with four different inclusion percentages of Kernza (0-75%) blended with Turkey Red wheat purchased from a local farmer and milled in house. Right away, some differences became apparent. It *seemed* as though the 75% blend needed a little more water, because it was having trouble adhering to itself. In addition to this, it was much more sticky than the other doughs. Upon proofing, it also expanded much slower and considerably than the other blends. In fact, there was an obvious positive correlation during this first round of baking that revealed the more Kernza present in the dough, the less it wanted to “pop;” it behaved as though it was less active, which contributed to a lesser increase in volume. This correlation held true throughout the actual baking of the bread as well. Each loaf became substantially smaller as the blend percentage approached 75% Kernza. 

You should also take note of the color differences. Yes, Kernza seed is darker than Turkey Red seed, but not by much. A darker loaf can indicate that more may be at play than just a darker color of the raw seed: one of them being an issue of too much starch being broken down too early. Broken down starches readily convert to sugars in the presence of enzymes, and sugars caramelize very easily. Evidence of damaged starch and early enzymatic activity is present in falling number tests, rapid visco analyses, and other dough tests that will be covered in the next section. It is also fairly common knowledge that stone milled flour generally produces more damaged starch than roller-milled flour. Again, whether or not this drastic color change was due to seed color or broken down starch, or both, is up in the air, but the section revealing our lab test results will provide much more insight on why this may be the case. 

Ultimately, it was decided that an inclusion rate somewhere between 1% and 25% Kernza would align with the goal of our outcome for this blend. Subsequent testing was carried out (Photos of 0-25% inclusion tests are below as well) in a similar fashion to land on a final blend ratio: 15% Kernza, 85% Turkey Red wheat. Photos of these tests are shown below, and they provide great visual examples of general expectations when it comes to baking with Kernza.

% Kernza blended with Turkey Red Wheat

% Kernza blended with Turkey Red Wheat

Back to top

Commercial Flour Comparisons

We landed on an acceptable blend ratio; it behaved somewhat normally, proofed well, baked well, and was incredibly delicious. The next question then became “How well does this blend stack up against well established whole wheat flours?” To answer this question, it was decided that we would compare King Arthur, Bob’s Red Mill - both whole wheat - and our own stone milled Turkey Red, each with 15% Kernza. In addition to these three blends, a control loaf made from each unblended flour (excluding Kenza) would also be baked.
Without getting too far into the weeds as far as the results go, here is an overview. In considering flavor, the decision was tough. At some point between the three of us, the Kernza/Turkey blend tied with either the King Arthur blend or the Bob's Red Mill blend. They each contributed their own character that ultimately coerced the decisions to come down to personal preference. It was ultimately agreed upon that the Turkey blend had the most flavor over all: more complex, more potent, strongest aroma, etc. The other two blends were much more mild; some said one of each was bland, others said they were smooth and buttery. Broadly speaking, amongst all six samples, Turkey Red never came in last. Bob’s Red Mill, and King Arthur (blends and controls) both received the lowest score from at least one of us, meaning when it came to flavor, we were on track. 

There were variations in the look and feel of each loaf as well. The commercial flours yielded extremely smooth doughs, while Turkey Red produced a fairly coarse dough. Adding Kernza to each, as you might expect, added more coarseness and texture. In terms of volume, each was fairly similar, but no final height was exactly the same. The Kernza/turkey blend was the smallest, but not by a substantial amount. 

The goal here was to determine whether or not our blend could produce comparable results (better, equal, or slightly less desirable) to popular commercial flours in order to deem it an acceptable substitution. Though the commercial flours we tested had a smoother texture, and produced bread slightly larger in size, comparative differences were slim. It was even said that the crumb of the pure Bob’s Red Mill loaf was too tight, or toothy. To top it all off, Kernza and Turkey Red together created a flavor that was difficult to beat. 

Experiments in Versatility

One final consideration to act upon was how our blend held up in a range of different baking applications. Our conventional pan loaf passed the test, but we were curious to try a focaccia, which required different baking parameters (ingredients, fermentation, times and temps, etc.). This bake test was slightly less involved; only our blend, pure Turkey Red, and King Arthur whole wheat would make up the samples. As a side note: King Arthur’s whole wheat was chosen to be our preferred commercial flour representation over Bob’s Red Mill. 

Slight differences between these samples were, again, revealed throughout the baking process, but over all, there were far fewer than there were in the pan loaf trials. The King Arthur dough did come together slightly faster and easier than the other two samples, but a minute or two of extra kneading brought each to a more level playing field. They all proofed wonderfully, and in about roughly the same amount of time. Upon baking, the King Arthur dough was a little firmer than the other two. It also smelled like a conventional whole wheat bread: not too spectacular, but somewhat pleasant. Our blend yielded the softest crumb, and had an aroma that really stood out; it was noted to be very floral and lively. The crumb of each was fairly tight for a focaccia, but this comes as no surprise, because we pulled a very simple and quick-to-ferment recipe from the internet. With some work, we have few doubts this would yield a top notch focaccia. It also seemed as though the Kerza blend focaccia lost more oil after baking than the other two breads. More on that later. 

15% Kernza blended with different flours and baked into Focaccia

Back to top

Kernza and Soft Wheat

It’s been stated before that there is speculation Kernza would make a good stand in for a soft wheat flour. Drawing from a small amount of testing we’ve done in the past, we agree, but we had to continue vetting our new flour specs (particle size, moisture content), to further explain how this conclusion arrived. This round of testing followed the same procedure as the “Turkey Red Trials” mentioned above: start with a broad range of inclusion levels; hone in on a tighter range; pit it against large scale/commercial brands’ and carry out a final test exploring the versatility of our blend. Rather than a pan loaf for our first round of tests, however, we decided to go with scones, as soft wheat characteristics generally lend themselves very well to this style of pastry. 

We used a soft white wheat called White Sonora for these trials. It is considered the oldest heritage wheat in the Americas, and has a lovely subtle flavor. As before, the trials began with Kernza inclusion percentages of 0 through 75%, allowing us to later hone in on a tighter range. Upon mixing, the pure sonora dough remained fairly wet and increased in dryness as the amount of Kernza increased. This makes sense, because our Kernza flour had a more prominent bran inclusion, causing it to soak up more water. While baking, the pure sonora scone seemed to hold its shape the best and 75% Kernza, the least. It spread out a noticeable amount as well. None of the scones fell flat, though, and only slight differences in height were observed. Each sample developed a very nice top crust, too. 

The taste test was quite an experience. Each sample was absolutely delicious and soft, and the recipe itself was so easy and produced such amazing results, that we had to share it on our website. Go check it out. The most notable change in flavor occurred in the jump from 0% to 25% Kernza. Sonora is delicious on its own. It is also very subtle, however, and our Kernza flour added a great component to the flavor. There was no sacrifice in texture at this point either. At 50%, we noticed a slightly denser crumb, and again in the 75%, but nowhere near something we would consider unpleasant or even “plain.” Each sample was classified as “soft.” It was another tough call selecting a tighter range, here, but ultimately we decided the next round of testing would involve inclusion percentages of 20, 30, 40, and 50% along with a pure soft white wheat control flour. We also used a soft white wheat from Meadowlark Organics instead of the white Sonora from the previous test. 

Round two revealed similar results in comparison to the first round: hard to distinguish substantially, but the biggest jump in flavor happening between the control and the scone with 20% Kernza. It was noted here that “Kernza turns sugar into honey.” Make of that what you will, but we consider that a positive trait. At 30% we noticed a “brany-ness” coming through in the flavor. It was not unpleasant, just intriguing. In fact, we deemed the 30% inclusion to be our favorite, because its flavor was the most balanced between the two flours; both were recognizable and seemingly boosted the flavor of their counterpart. Given that final conclusion, it would be a blend of 30% Kernza, 70% Meadowlark Organics’ soft white wheat used in the next round of bake tests. 

% Kernza blended with soft wheat

% Kernza blended with soft wheat, showing heights

Back to top

Commercial Flour/Versatility Bake Trials

For our final experiment, we lumped two trials together, testing a different style of pastry and a comparison involving commercial flour. Perhaps this is cheating, but we have a business to run here, people. Along with our predetermined blend, the other flours in this trial were, Bob’s Red Mill Whole Wheat Pastry Flour, and pure Meadowlark Organics’ soft white flour. The recipe: Shortbread. We went with Bob’s Red Mill pastry flour, because, well, we had it on hand. But, the package also has a shortbread recipe on the back. This scenario was perfect for a comparison; a recipe developed for a specific flour would represent an ideal standard to emulate. 

The final takeaway from this test was, at this point, a fairly standard result when it comes to baking with Kernza. The blend’s cookies fell a touch more flat than those using the other flours. They were also darker. However, and most importantly, they tasted amazing. We even went as far as to say, while the Bob’s Red Mill cookies held their shape the best, they sort of tasted awful. Perhaps we came to this conclusion solely because the other two cookies were so amazing. It’s hard to say. What we were confident in saying, though, is that a blend of 30% Kernza and 70% soft white flour produces a comparable stand-in for a commercial whole wheat soft white flour. 

One additional thing worth mentioning is I baked a few scones using our Turkey blend as a side test. They still turned out excellent, and the results suggest an even broader range of applications for this blend than previously thought. 

Kernza blended with soft wheat in shortbread

Back to top

Final Thoughts on Baking

Within this swath of testing, we were able to extrapolate on nearly every baking test we’ve carried out in the past, and really hone in on trends that arise during utilization of Kernza in the kitchen. Some of these trends are well known, while others are relatively new. As each trend is revealed and further testing is conducted, recorded, and repeated, predicting baking outcomes becomes more and more tangible. A condensed iteration that represents everything we know and predict about baking with Kernza follows. 

When it comes to pan loaves (or anything of the sort) requiring structure, volume, oven spring and strong bread flour, your Kernza will need help. In fact, I feel it is more appropriate to think of Kernza flour in these applications as taking strength away from bread flour rather than bread flour giving strength to Kernza. Don’t try to strengthen Kernza flour; use it in a way that least diminishes the structure of your strong flour. As we’ve seen: more Kernza in this type of baking will trend positively with less rise, a slightly coarser texture (due to it being whole grain), a slight diminish in its ability to “set” at the end of baking, darker color, slightly stickier crumb (and dough), and higher affinity for water absorption (this is not to say Kernza flour needs more water just because it seems to absorb more, however). To put it generally: it slightly hampers performance, but it tremendously boosts flavor.

Given this knowledge, there are plenty of tricks you can implement when baking breads requiring strength and stability to ensure more satisfaction with your end product. Try starting with an inclusion percentage of about 15-25%. This typically works for us, and it is just the right range to allow plenty of Kernza flavor to still come through. You can also try dropping hydration to around 65%. This is somewhat counter intuitive, as whole grains are typically very thirsty. However, we’ve seen positive results when applying it, and the lab test results in the next section seem to agree. Sifting the flour to rid it of some or all bran content will also contribute to maintaining the structure during mixing and baking, because bran has a tendency to break apart the gluten network when over mixed. It also competes with gluten and starch for water, affecting the development of each. I should note, though, there is evidence stating the minerals and nutrients present in the bran aid in extending shelf life. A few things I have not personally tested yet, but suspect may also help are the following: Allow the flour to oxidize for a week or two (just keep it in a cool dry place with a semi loose fitting lid). Believe it or not, this is a fairly standard practice. Fresh flours continue to undergo changes weeks after milling, and it is at about the two week mark, where these changes begin to flatline. Oxidizing the flour also promotes gluten bonding ability, giving the limited gluten available more strength. Mixing less than one might expect (more on this below) will also benefit Kernza flour ensuring it reaches optimum strength development before beginning to break down. There is also promising evidence that adding extra starch may help the dough’s ability to set during the final minutes of the baking process. The type of starch does matter, however, and it seems like wheat, barley, or something related to taro may be good candidates. Take this with a grain of salt. Again, I have not had an opportunity to test this, but, as you will, again, see below, I suspect Kernza’s starch content (or lack thereof) is a huge factor in why it bakes the way it does. In addition, other dough conditioners like ascorbic acid, vital wheat gluten, and wheat protein isolate have each been shown to improve loaf volume and crumb texture. Ascorbic acid contributes to flour oxidization, while vital wheat gluten and wheat protein isolate can reduce stickiness and improve gluten structure forming ability. I haven’t found a study that combines these conditioners, but it is a common belief that this practice may yet further improve loaf characteristics. Ascorbic acid and vital wheat gluten are also both inexpensive and very easy to find. 

When it comes to baked goods similar to scones, pancakes, and quickbreads, it seems Kernza presents fewer issues. This makes sense, because a strong, elastic, stable structure is less important in these applications. There is still a noticeable decrease in these doughs’ structure holding capacity, however. In other words, more Kernza will, across the board, cause a slightly higher spreading out effect even if no rise is required. But, the effect is much less detrimental in these scenarios. Our inclusion range of success is also substantially greater in this category of baking. In a lot of cases, 25, 50, even 75% inclusions yielded great results, and I’ve taken it as far as making banana bread with 100% Kernza flour in the past that turned out quite well. 

One final characteristic that I’ve noticed tends to reoccur regardless of what you are baking is that if your dough recipe contains a fair amount of additional fat, it will have the tendency to, for lack of a better term, leak out of the bread during baking. To me, it seems like the dough loses its emulsion, and I first noticed this when I was testing pie crusts. It happened again in both the focaccia and shortbread tests mentioned above. As I briefly explained earlier, I think this may be directly related to the starch content of Kernza. Think is the key word, here. I don’t actually know, but what I do know is that lipids (in the form of fat) have the ability to bind to starches in flours forming an amylose-lipid complex that contributes to some of their network forming properties. If there isn’t enough starch in a flour sample (Kernza) or possibly isn’t the right variety (not enough amylose, too much amylopectin), this additional fat, having nowhere to go, may just seep out of the bread. This observation leads me to believe that where wheat flours can generally handle a higher fat inclusion, a recipe involving Kernza may benefit from a reduction in fat. The only fat present in the scone recipe used in testing came from cream. There was no additional butter or yogurt. In turn, the scones themselves were incredibly soft with a minimal “greasy” feeling. The focaccia and shortbreads, however: noticeably greasy. 

Given these findings, it would be great to carry out additional testing. Further exploring Kernza flour’s relationship with fat and oil would be an appropriate next step. I would also like to experiment more with dough conditioners, hydrocolloids, and aging. A research paper I pulled information from regarding similar testing showed promising results. Tests were also conducted on tempered and refined Kernza flour that led to additional benefits. However, I would like to keep my testing in a realm that is useful to home cooks, restaurants, and small scale milling operations. Therefore, if I can’t easily source products necessary for a specific test, I will not prioritize its implementation. 

Baking experiment ingredients prepared!

Back to top

Rheological Flour Analysis

At this point, we were somewhat familiar with Kernza in comparison to other flours. We were even getting accustomed to its behavior in the kitchen: an applied setting. Most of our data, however, was strictly observational and speculative. In order to get a fundamental understanding of a wide range of Kernza’s baking properties, we had to outsource. Great Plains Analytical Laboratory, located in Kansas City, Missouri, ran our flour through a litany of tests, and the results proved to be extremely helpful in understanding Kernza. Below, I’ll do my best to explain each test and what it means in relation to working with Kernza flour. Much of the data aligned with the results of the baking tests, providing helpful additional context. Some of the data, however, was a little abstract and proves difficult for one to draw any solid conclusions from. 

Moisture, Ash, Protein, & Falling Number

Moisture is self explanatory here: The sample of flour was 9.6% water. Our intent was to get as close to 10% as possible. We’ll take 9.6.

Ash content refers to the mineral content of a sample. It is obtained by incinerating a small portion of flour and weighing any remaining ash. Minerals do not fully combust and are, therefore, left over after burning. Everything else gets vaporized. The higher the ash content, the more minerals are present in a sample. In grains, the majority of minerals (nutrients) are stored in the bran and small germ. Very few minerals are present in the starchy part (endosperm) of a grain seed, making it, among other things, less nutritious. For example, all purpose flour contains almost no bran or germ. It’s ash content is usually between 0.5-0.6%, whereas Kernza’s is 2.218%. More ash equals a higher mineral content, which equals more bran/germ presence, which ultimately means more nutrients. 

Protein is typically an excellent indicator of bread quality. Two proteins, glutenin and gliadin, when allowed to combine in the presence of water, form gluten. It is gluten that provides structural capacity in bread. In most cases, a higher protein content will yield a stronger dough that still has the capacity to stretch and expand. This is revealed in a loaf of bread that has an excellent rise and a high final volume. I say “typically” and “in most cases,” because these guidelines predominantly apply to wheat. Grains other than wheat, even wheat grass, play by different rules. 

As you can see, our Kernza sample has a protein content of 15.39%. If a wheat variety were to contain that level of protein, its dough would be among the strongest you could find anywhere. High gluten bread flour (wheat) usually contains around 13% protein. Unfortunately, the protein listed needs the right combination of glutenin and gliadin proteins to form high quality gluten. While Kernza has both, along with somewhat acceptable amounts of each, there are functional and less-functional forms of both proteins. Any glutenin present needs to be “high molecular weight glutenin.” Yes, the weight of each individual molecule determines whether or not that molecule will carry out the desired function. Kernza’s glutenin is predominantly “low molecular weight glutenin,” so it does not perform with the rigor of glutenin present in a high-protein wheat. Instead, its structure is a little less stable. It is partially due to this that we recommend using it in structure deficient baked goods such as quickbreads. Another contributor to the high protein content is the relatively high number of “albumins” and “globulins.” These proteins are not stored where glutenin and gliadin are. They reside in a thin layer of cells just underneath the bran, while gluten and gliadin are contained in the starchy endosperm. Again, due to the high area of bran in comparison to endosperm, the ratio of albumins and globulins will go up, contributing more to the total protein percentage; contributing more of the “wrong” proteins, as they do not serve a strong function to desirable baking properties. A wheat seed is larger, because it contains more endosperm. Making it safe to conclude that an overwhelming amount of its protein comes from (high molecular weight) glutenin and gliadin. 

Falling Number provides an indication of the starch quality in a grain. The method used to produce this number utilizes a machine that operates by heating a mixture of flour and water in a small vial to make a thick slurry. Once heated and allowed to thicken, a small plunger is placed just on top of the slurry, and allowed to fall through it until it reaches the bottom of the vial. The falling number itself is literally the amount of time, in seconds, it takes the plunger to reach the bottom of the vial. Every falling number is greater than 60, because the mixing and heating time is taken into account for each test. As you may have guessed, thicker slurries are more viscous and produce more resistance to the plunger. So, the thicker the slurry, the higher the falling number. 

A general sweet spot for wheat and a few other grains is a falling number around 200-250 seconds. I have seen a few cases where a falling number at or just slightly over 300 is acceptable as well. Too weak of a slurry (low falling number) typically means there is a significant amount of damaged starch or enzyme activity already present in the grain. Damaged starch usually occurs during milling, and a literal breaking of starch granules causes a decline in their stability during hydration and heating. A high enzyme activity can also involve damaged starch, but it usually begins in the field before, during, or immediately after harvest. If a grain begins to germinate too soon, before it has had a chance to dry down to a storage moisture content, enzymes present in the seed will begin breaking down the starch, again causing a loss in stability. 

On the other end of the spectrum, too high a falling number can also be detrimental. From what I’ve gathered, this is usually above 350, and this typically means there is not enough damaged starch. A certain percentage of damaged starch (around 8-10%) is incredibly helpful during baking; the starch chains are broken down enough to become food for the yeast when it is added, increasing its vigor. The falling number is usually a very reliable predictor of potential flour performance. 

Take a look at the two samples: 100% Kernza, and our Kernza/Turkey blend. The blend has a falling number of 340, while 100% Kernza flour’s is 152. A drastic difference. And, as you will see upon continued reading, Kernza does, indeed, have a starch deficiency. The blend may also seem fine for practical uses, but other factors need to be taken into account - whole grain flours in particular, because there is so much variability from product to product - before one can definitively say “this is a completely acceptable flour to use in most applications.”

One final point to keep in mind: The falling number provides data on a specific sample, and falling numbers produced by grain taken from the same plot can vary year to year due to the heavy influence of environmental factors. I suspect the falling number taken from this specific sample may represent the low end of Kernza’s range, and I am confident that subsequent generations of Kernza will reveal an increase. 

100% Kernza

Kernza and Wheat blend

Wet Gluten is a quick and easy indicator of pure gluten quality and quantity in a given sample of flour. It is expressed as a percentage, by weight, of the gluten network remaining after a dough has been formed, and all of the water soluble components of it have been washed off. The idea is: higher quality gluten is better able to form a structure around water, thus holding more. Therefore, the higher the percentage of wet gluten, the higher the quality and quantity of gluten forming potential is present in your dough. 

The results we obtained from each of our samples were <10% wet gluten content from pure whole grain Kernza, and 28.83% from our blend. As a benchmark, most strong hard wheat flours have a wet gluten content ranging from 30-45%. This range of percentages represents the highest quality/quantity, however, and does not mean anything below is unacceptable. I have seen results from other hard wheats weighing in at 27%. The soft wheats that I have seen are around 19-22%. Judging by Kernza’s wet gluten percentage, it is pretty clear that it needs a little help should one need a boost in baking performance. It also seems that the addition of our Turkey Red does just that. It looks like we’re on to something here. 

A final note: Once wet gluten is isolated, there are a few additional tests that can be carried out to hone in on a more precise relationship between quality and quantity of gluten. They involve spinning the gluten ball in a centrifuge to separate the strong and weak glutens and determine the ratio between the two; and drying the gluten to determine its water holding capacity. For now, we are sticking with the initial test considering the extent of additional testing conducted on our flours. 

Bullet point? *High nutrition and fiber, high protein, lacks structure forming proteins. 

Back to top

Extensograph

The extensograph is a great tool that provides insight on dough characteristics and behavior before baking. This information can, in turn, be used to predict what will happen when baking actually does occur. Generally, the shape of the graph reveals how positive the interplay is between glutenin and gliadin (the two proteins that form gluten). It also reveals a potential hydration to consider when baking. 

Take a look at the graph below depicting an extensogram (the graph made by an extensograph) of 100% Kernza flour. It is a tall, thin spike. Its height is around 1400 BU, or “Brabender Units.” Just know they are a unit of force: in this case, resistance to extension. The horizontal axis depicts length in centimeters. Seeing as the extensograph literally stretches a piece of dough, imagine what would need to happen to produce a graph like this: A lot of force would be applied to stretch the dough a very small amount before it snapped. In other words, the dough cannot stretch without breaking. This is not ideal for a loaf of bread. Instead, an extensogram depicting more of an upside down “U” or rainbow shaped curve is a sign of a dough that resists extension just enough to stretch a little farther. 

Take a look at the extensogram produced by our Kernza/Turkey Red blend. According to it, the dough formed here looks much more like what one would desire when making a loaf of bread. Although, when compared to the extesogram of a high quality, high protein dough (bread flour), it is still lacking. Imagine the graph being pulled to the right and slightly upwards, and you have just pictured a more ideal extensograph. 

100% Kernza

Kernza and Wheat Blend

Back to top

Farinograph

A farinograph reveals many dough quality properties that arise specifically during the process of mixing flour and water. There is no stretching, proofing, or baking. Essentially, as two paddles mix the dough, they are measuring the force of resistance. Continued mixing increases the dough’s strength until a peak viscosity has been reached. After this point, continued mixing will slowly begin to break down the structure that had formed. The results from this process give bakers an idea of how much water to add to a given flour, the amount of time it takes for optimal consistency to form, the extent to which it breaks down after peak consistency is reached, and how quickly it will do so. 

Though all of this information is very helpful, it does not always paint a full picture of how your flour will actually perform. As it was explained to me, it is just a strength reading; it determines how strong isolated components in a flour can become, but does not correlate to the total quantity of said components in the flour. It is for this reason that farinographs and extensographs should be used in tandem to give a broader picture of a flour’s performance. 

With that being said, the 100% Kernza farinograph reveals some important information to, at the very least, get you in the ballpark of how to treat your dough.

Water absorption may be the most straightforward data point: Start with 62.7 - 67.5% water to flour by weight. Each of the two numbers represents an optimal percentage corrected for different variables, so creating a range between the two should be alright. 

Development time is the amount of time it takes for a given dough sample to reach peak resistance i.e. maximum strength. The good news here is that there is usually a direct correlation between the development time on this test, and while mixing your own loaf of bread. Our Kernza sample’s time was six minutes. Generally one looks for longer development times, because it can be an indicator of a more stable dough. A dough that peaks too quickly is more likely to break down faster, as well. Stronger flours usually sit around 8-11 minutes of mixing before reaching maximum strength, so at six minutes, kernza flour is not far off. What this should really tell you, however, is that while mixing your own dough, be very aware of what it is looking like at the six minute mark. Once this point is reached, overmixing is approaching very quickly if it hasn’t happened already. 

Stability indicates the “endurance” of your dough in minutes. Longer stability time is equivalent to a longer period of time in which your dough holds near maximum strength. A longer period of time is, again, more desirable in this case. A stable dough will be more versatile in ever changing circumstances: humidity, room temperature, differing fermentation times, inconsistent mixing, etc. It gives the processor more “wiggle room.” To, once again, compare Kernza to a relatively strong wheat, notice its stability is 6.7 minutes. On the graph, this is the range where the curve breaches 500 FU (or BU in some cases), and subsequently falls back below it. FU and BU are units of force or strength. 500 BU is used as a standard for characteristics of a desirable dough. In stronger wheat flours, I have seen stability times of up to 23 minutes, and suspect an even greater number of minutes may be possible. 

Though there are more data points on this graph, the three mentioned above are typically the most helpful. They will tell you how to treat your dough, and how it will behave in an unpredictable environment. I should reiterate, however, that this graph should not be a sole judgment on the quality of a flour, and a comparison between the results produced by each flour provides a great example as to why. The Kernza/Turkey blend results show a very short development time, an even shorter stability time, and a short breakdown time. Short farinograph times are usually undesirable; longer development, stability and breakdown are traits of strong, resilient doughs. The results produced by 100% Kernza flour look much better, but as we know, the performance of Kernza flour cannot compete with wheat in a baking application. This is precisely the reason why a farinograph needs to be considered amongst a range of other tests in order to get a complete picture of flour quality and performance. 

As to why the blend produces such short time intervals, I have a feeling it may have a lot to do with the introduction of more (and potentially stronger) bran contributed by the wheat. Because this dough is being mixed throughout the entire procedure, the bran’s effect may take hold sooner, thus causing a more rapid dismantling of the gluten/starch network. This is a simple explanation, and there is always more at play, but I think it is a significant factor. 

100% Kernza

Rapid Visco Analyzer

The final test on our flour samples was carried out by the Rapid Visco Analyser (RVA). This machine, much like the two mentioned previously, measures viscoelastic properties of a dough. However, the RVA is designed to bring out behavioral characteristics of starches present in grains. It works by mimicking a standard baking process, and records the viscosity of hydrated starch along the way. There are three temperature phases: heating, holding, and cooling, and the starch’s behavior throughout each phase has the potential to reveal an aspect of the final bread quality.

 Two important roles of starch in baking are: providing food for yeast, and preventing the structure (formed by gluten) from collapsing. In other words, starch is responsible for “setting” the dough in the final moments of a bake. The type of starch also determines how rapidly the bread will harden as it sits in storage. Starch works hand in hand with gluten. It helps trap water and CO2 and eventually sets the structure after a certain amount of cooking. A dough without starch (just gluten, and assuming yeast had something else to eat) might still rise while baking, but without the starch to keep it in place, would fall flat. Therefore, starch quality/content is in direct correlation with loaf volume. 

The red line on the RVA graph represents temperature changes, while the blue line represents the real time viscosity of the hydrated flour. There are six variables along the blue line to take into account: Pasting Temperature, Peak 1 (peak viscosity), Trough 1 (hold viscosity), Breakdown (viscosity), Setback, and Final Viscosity. 

The pasting temperature represents the minimum temperature required to cook the starch in a given flour. It is at this point that the viscosity of the mixture begins to increase. This is caused by the swelling of individual starch granules due to increasing temperature. And should be noted when considering heating/cooking options for your dough. In many cases, it can be used to predict final loaf volume. A delayed pasting temperature is generally preferred when a high final bread volume is sought after, as it is a sign of higher amylose content. More on this in the following paragraph. 

Peak viscosity represents the point during heating at which a dough has the highest viscosity, or thickness. It may seem counterintuitive, but a high peak viscosity does not necessarily mean a higher quality of starch. Starch is composed of two molecules: amylose, and amylopectin. We don’t need to get into too much structural detail here, but due to their difference in size, weight, and shape, they behave differently when hydrated and exposed to heat. As was briefly mentioned above, in this hot wet environment, starch granules begin to swell, and amylopectin is the main contributor to this attribute. It is this increase in size that causes the increase in viscosity. During this phase, the amylose molecules begin to “leach” out of the granule and into solution. This event counteracts and reduces the rate of granular swelling. This is an important step, because a delay in starch swelling increases its stability. If starch granules swell too much too quickly, they will rupture and degrade earlier in the baking process, losing a portion of their contribution to the bread. 

A comparison between hard and soft wheat provides a good example of amylose’s role in peak viscosity. Hard wheat has a higher percentage of amylose than does soft wheat, and in many studies I’ve come across, soft wheat has a significantly higher peak viscosity, meaning less amylose is leeched, and amylopectin is able to swell freely (gliadin is also partially responsible for an increase in viscosity). With the quick swelling of soft wheat flour comes an early structural failure (revealing a low starch integrity) and, ultimately, poor bread volume. This is a major reason soft wheat is rarely used when a large loaf of bread is a desired outcome. 

 Breakdown viscosity is the value given to the difference in “peak” and “hold” viscosity. At a certain point (with enough heat), all starch will rupture, and its structure will be compromised. After this rupture, viscosity will drop to a certain point and remain there for a certain period of time. This post-rupture constant viscosity is the “hold viscosity,” and it occurs at some point after the RVA machine reaches a temperature of 95 degrees celsius. The degree of loss in viscosity is yet another indication of starch content and stability, and, again, it is the relationship between amylose and amylopectin that determines the level of breakdown. A more substantial drop in viscosity could mean a high peak viscosity which is an indication of a lack in amylose. In addition to preventing granular swell, amylose is also responsible for setting the dough during baking. Not enough amylose and your bread will collapse regardless of how strong the gluten network is. So, the more the viscosity drops (higher breakdown value) during this phase, the less stable the starch content is said to be. This is an interesting point, because our 100% Kernza flour saw a significantly smaller drop in viscosity (proportionately) than the Kernza/Turkey Red blend. However, digging into research publications studying similar topics, I’ve found this to be a common occurrence. Kernza’s starch may be more stable than wheat’s. As to why, there could be a number of reasons: higher minerality provided by the bran, high lipid (fat) content, or the ratio of amylose to amylopectin all contribute dough conditioning like properties to the starch. I have yet to find an exact explanation.

 Unfortunately, this is not a smoking gun for a complete explanation for how Kernza will behave in the kitchen. As mentioned earlier, the bran may also be detrimental to the gluten structure. What’s more, the overall viscosities of each RVA parameter (from 100% Kernza flour) are very low, implying, again, that there is not enough starch for it to really contribute anything substantial to the flour’s integrity. Something this information may allude to, however, is the potential of a starch isolate from Kernza: being that it seems to be stable and resilient, it may be a practical starch to use in cooked products that require thick texture after prolonged cooking (sauces, soups, pudding, etc.) I may be completely full of it, here, but I did find a paper that mentioned this as a possibility. 

The final viscosity and setback viscosity are parameters that typically correlate with the way a flour will behave while it is cooling after cooking. Where final viscosity is, aptly named, the final viscosity reading taken during the analysis-after cooling and time have expired, the setback viscosity is the difference in viscosity between peak and final viscosity. Depending on the type of starch, the viscosity during cooling will increase substantially or nominally.  A substantial increase in viscosity corresponds to a high amylose content relative to amylopectin. Amylose is the portion of starch that has the ability to quickly form thick gels upon cooling. It is  this quick action that is responsible for setting the dough. Though it contributes positive attributes during baking, a high concentration can lead to products becoming too firm as they cool. However, a higher concentration of amylose along with fat and minerals derived from the bran will contribute to the bread maintaining its quality in the long term as opposed to drying and solidifying even more as the amylopectin retrogrades. Where amylose sets quickly, amylopectin “recrystallizes” days after cooling, and it is the main contributor to bread staling. Again, this is all about the balance between amylose and amylopectin. Too much or too little of each can have a positive and negative effect. Not only that, (to get even more complicated here, sorry) the average shape and size of each of these molecules will cause even more changes in dough characteristics. For example: Amylopectin consisting of longer than average branch chains can mimic the effect of amylose. 

We don’t need to worry about long chain amylopectin in our experimentation, but it ties in to an important point: There are a seemingly endless number of factors contributing to any given doughs behavior during processing: gluten is formed by gliadins and globulins which behave in specific ways; the type of gliadin and globulin again contribute another layer of behavior changes; the many variations of starch contribute another set of factors interacting with each other, each gluten component, the fat in the bran, and the bran itself, and each of these interactions interact with every other combination of interactions. It really seems as though predicting anything about flour with one hundred percent certainty is impossible, and getting to a point of just-applicable-certainty can be quite frustrating. To add to this frustration, the amount of research done on Kerza rheology pales in comparison to that of wheat. However, more studies on Kernza are being published each year, the majority of them coming from the University of Minnesota, and I have no doubt that culinary predictions, standards, behaviors and traits will all be determined and compiled in no time. 

Back to top

Final Thoughts

 With all of this information, it is still difficult to draw many solid conclusions when it comes to cooking and baking with Kernza. The graphs and data points may allude to something, but carrying out actual baking may tell a different story. One thing, I feel, that may help in experimenting with kernza at home or in the kitchen is the fact that the individual seeds are so small. This fact is really the foundation to what throws everything off. It increases the bran to endosperm ratio, which means it has a high protein content; it’s just the wrong type of protein. If Kernza were a larger seed, the starch characteristics may show some promise, but there just isn't enough per gram to contribute much at all. No number on the RVA graph got very high when compared to wheat, even though the shape of the pasting curves look similar (perhaps even better) at a glance. More starch per seed may also help mitigate the undesirable gluten characteristics to a certain degree. For a while, the seed was even too small to remove the bran for a refined flour, though a study published in 2019 determined the size in some strains was finally to a point where tempering and bran removal were possible. Refined Kernza flour does show more promise across the board when it comes to ease of cooking and usability. That being said, whole grain Kernza is what's available for now, so remember: it's small. Small things need a little help sometimes. The help, in this case, will probably come from a little (or a lot of) extra wheat flour, although certain dough conditioners, hydrocolloids, perhaps some refined starch might help out as well. It is also important to keep in mind that a lot of the properties mentioned above are what is required for a higher quality loaf of bread or something of similar nature. Quick breads, cookies, and no-structure-neccessary baked goods are usually a great fit for Kernza (in case you jumped around, read the “Final Thoughts on Baking” section for more on this), but I now know that they have the capacity to firm after cooking a little more than one might expect (amylose). 

It is my hope that anyone will be able to pull at least one useful bit of information from this post to make an accurate prediction much like the one mentioned in the previous sentence. I also hope this article motivates continued fearless experimentation with Kernza. This grain deserves all of the attention right now, because of its tremendous potential to help stave off climate change and heal soil. It is also incredibly delicious, and we have yet to find a flavor profile that it does not work well with. It is truly a powerful and versatile grain that is becoming more so each year, so help save the planet, and eat more Kernza, people!

Back to top