The Artful Dance
September 09, 2010
by Dan Malovany
Sometimes it just takes a little luck, a wealth of experience and knowing some of the tricks of the trade to figure out how to best ramp up a new product from the 5-lb-capacity McDuffee bowl mixer in the R&D lab to a 1,000-lb-capacity horizontal mixer on a high-speed bun or cracker line. For Brian Strouts, nearly two decades of developing new products taught him a thing or two about going from bench to oven — such as how to adjust the formula from tabletop batch in the lab to a 125-lb dough on a pilot plant line.
“Our experience tells us when we take a bread formula from the lab to the pilot plant we typically need to bump up the water by 0.5 to 1%,” said Mr. Strouts, head of experimental baking at AIB International, Manhattan, KS.
TRICKS OF THE TRADE. One trick of the trade helps, but it takes a whole bag of them to resolve the dozens of variables that often occur during the artful challenge of scaling up. “Even when we have done all of our work at the bench, we’re always tweaking it as we scale up,” Mr. Strouts noted. “With bread, the two primary variables are the effect on absorption because of how long you are handling it, how you are mixing it and how the mixing action affects the absorption. The second is mixing time. Depending on what mixers you use, sometimes it’s more intensive and other times it’s less intensive in the lab compared with when you scale up.
For Ken Zvoncheck, director of the Science and Innovation Center for Reading Bakery Systems in Sinking Spring, PA, years of experience taught him that reducing the bake time is the norm in taking cookies, crackers or other snacks from a rack oven in the lab to the 32-ft long tunnel oven in the company’s pilot plant. “What we see time and time again is that people tell us what bake time they use in the lab, and we’re always able to reduce that bake time considerably because our tunnel oven, with its convection forces, is going to be much more effi cient than what they used in the lab,” he explained.
The pilot plant’s tunnel oven also has several bake zones, allowing Mr. Zvoncheck to customize the baking process from beginning to end by adjusting temperatures and air velocity and manipulating the exhaust along the way. To avoid the blistering on a high-moisture cracker, for instance, he will close the exhaust and trap the humidity in the first baking chamber before releasing the exhaust in the second zone to dry out the cracker. He also can adjust the convective/radiant properties of the heat, from the bottom or top, to balance out the bake and the color of the products as they travel on the oven’s mesh band.
Even with these tools, Mr. Zvoncheck must tweak the process when he is overseeing the commissioning of a full-scale production line with a 300-ft tunnel oven and equipment that runs 10 times the speed of the pilot plant line. “Everything is running faster, but the bake time remains the same,” he noted. In addition to the line’s speed, there are many variables, including the amount of time that it takes to get the dough from the mixer into the oven. “On a production line, it may take half of the time that it takes in our tech center. Small things like that can cause subtle differences in the final product characteristics,” he said.
In the lab, it may take just 30 seconds to make up two prototype loaves of bread while the floor time on a full-scale production line can range from 10 to 20 minutes, significantly altering final product characteristics, Mr. Strouts said. Likewise with a cake formula, beating 90°F ambient air into the batter on the production floor could create quite a different final product than when aerating a test batch in a 72°F air-conditioned lab, he noted.
That’s why many bakers will build a good amount of tolerance into their formulas, said Jesse Weilert, manager of bakery applications, Caravan Ingredients, Lenexa, KS. During the ramp-up process, a handful of successful test loaves now must become a continual flow of flawless products coming down the line. “The complications of large batch sizes, extended processing times, unchangeable automated systems, limitations with production schedules, fluctuations within large production equipment and even the bumps and jarring of conveyors can all make products that were successful in a lab environment become failures,” he said.
In many cases, simulating line production is not possible, added Gary Tull, research manager bakery industry at Puratos USA, Cherry Hill, NJ. “We try to provide enough tolerance to any base or improver to withstand the higher sheer and stress imparted by machinery in most plants,” he said. “It is difficult to advise customers on matters such as mix times. Water absorption can deviate quite a bit due to the size of the dough. Industrial customers often need concentrated product, and there are precautions that must be taken to avoid any potential overscales, especially pertaining to enzymes.”
PLAY DOUGH. Understanding the art of taking a new or reformulated product from bench to oven requires equal comprehension of cereal chemistry, ingredient technology and equipment capabilities, noted Ed Fay, c.e.o. of CMC America, Joliet, IL. “It’s not just the machine. It’s not just the formula. You need to understand the whole process,” he said.
Some enzymes reduce mixing times while others extend it. How the dough is divided, moulded, sheeted, proofed or baked impacts the final product. “You can put the same product through three different ovens and get three different products,” Mr. Fay added. “You need a critical understanding of all products, so it becomes a teamwork exercise in the end. It can be costly. It’s fascinating. It’s an artful dance to achieve these targeted characteristics in the most optimal way.”
He suggested creating bigger batches in the product development stage as a way to reduce the time it takes to scale up and to more accurately predict the variables in the process. Producing test batches on 400-lb-capacity mixers, for instance, can replicate full-scale production better than on machines that make 50-lb batches. “Large companies have dedicated R&D lines almost as big as their full production lines,” Mr. Fay said. “That’s how artfully they want to prove that they can achieve the desired design characteristics on the final product.”
Size can be everything when it comes to mixing, according to Mr. Strouts. “If we’re starting at bench scale with a small spiral mixer, in some cases, we may see a little stronger development in that small spiral,” he said. “When we go to a larger spiral or a horizontal-arm mixer, we will need to play with the dough.”
Alpha Baking, for example, previously tested products on its full production lines, but last year, the Chicago, IL based company installed a 1,000-sq-ft R&D lab with a bench mixer, 1.5-bag spiral mixer, roll divider, moulder, sheeter, box proofer, deck oven, rack oven and a couple pans of every size from each of the company’s bakeries, said George Poulos, vice-president of manufacturing.
In the lab, Mr. Poulos can develop several variations of a formula, then run tests of the small batches that often last no more than a couple of minutes on the actual production line. To determine if a dough needs more mixing time, enzymes or fermentation, many bakers such as Mr. Poulos do it the old-fashioned way. “Some people have that engineering expertise. I just stick my hand in the dough,” he said. “Usually, our biggest challenges [during ramp-up] are in fermentation, because in the lab, it is pretty well controlled. Out in the shop, there are a lot of outside influences.”
HEALTHY CHALLENGES. Controlling time and temperature through the entire R&D process may expedite scaling up to full production, Mr. Strouts said. “We have an advantage. When we’re doing bench top, we’re actually controlling mixing bowl temperature pretty carefully with jacketed bowls even at a small scale just like we can control the final dough temperature on a large scale,” he said.
Prior to conducting nonmainstream testing, Mr. Zvoncheck may ask clients to send him small samples of the dough and a rudimentary formula. To evaluate the dough, he runs it through a couple of gauging rolls to test sheetability. He also can check the dough’s viscosity and stickiness and may adjust the level of water in the formula or add, for example, a combination of lecithin and emulsifiers to make it more machinable.
The sheer mass of a scaled-up dough can create all sorts of problems, according Jim Bohrer, vice-president, technical support services, AB Mauri Fleischmann’s, Wilsonville, OR. “This creates time and temperature challenges that lead to the need to optimize final oxidation and strengthener levels,” he said. “The actual bakery environment is usually better suited to achieve proper fermentation in the sponge, and typically the dough will have better development in the large-scale mixers.” Heat often is an issue when scaling up.
Contrary to popular belief, it’s not the friction from mixing a large batch that generates most of the heat, according to Mr. Fay. Rather, it’s the latent heat of the hydration of the flour. “When you wet flour to more than 30% absorption, you create 6.2 Btu of latent heat per pound of flour,” he noted. Larger batches create more heat chemically. “That heat may be defined as latent, but it produces very real effects during mixing, and must be accounted for,” he said. “But this is only one variable in the process.”
Natural, clean-label and health-and-wellness products can be especially tricky to scale up. Mr. Poulos recently found himself developing breads and rolls for a school lunch program that prohibited use of high-fructose corn syrup, artificial preservatives and chemical additives. The program also required at least 51% whole-wheat flour in the formula, so whole wheat could be listed at the top of the ingredient legend. “The rigors of the production line became a challenge,” Mr. Poulos explained. “We had to beef up the natural oxidation systems, the enzyme replacers for the wheat gluten and the sodium stearoyl lactylate to get product to run properly on the line. In the lab, the pans don’t get bounced around on the conveyors. When we’re working on dough that’s kind of tender, that’s the biggest challenge. We need to consider on which line to run such products.”
Adding nuts, fruit pieces and other inclusions to cookies and baked snacks can cause tearing in the sheet if they are larger than the depth of the sheeted dough, Mr. Zvoncheck said. Likewise, doubling the fiber in a low-pressure extruded snack can also create challenges. “We don’t want to generate more pressure than necessary, because we have the gluten properties developed during the mixing process,” he explained. “As we exert pressure, we can deteriorate the gluten. In turn, that weakens the product, and it may not give us the texture that we’re looking for.”
Oak State Products, Wenona, IL, has seen a huge increase in demand for nutrition bars and other healthy products. During the past year, many customers began adding protein to their formulas, creating an unpredictable process. If the batch mixes even slightly too long, the extra heat can denature the protein, which could alter the final product’s taste and texture. Flavors also can have unintended effects because of their components. “We had to watch the flavors and how they interacted with the protein when it was used,” said Chris Bruch, project leader in the company’s R&D department. “The protein is a lot more sensitive, so it can affect your flavor, and texture becomes a big issue. We had to use a combination of protein blends to balance it out.”
Oak State Products developed a bar that qualifies for school lunch programs and contains a full serving of fruit using raisins and raisin paste, said Dave Van Laar, president and c.e.o. Working with raisins was a natural choice because the company had vast experience making oatmeal raisin cookies. “Schools need a healthy bar that children will eat,” Mr. Van Laar said. “The challenge is finding a way to make a healthful product that tastes good, is all-natural, fits into the school lunch program requirements.”
TWEAK. TWEAK. “Let’s face it, no money is made in R&D,” Mr. Weilert explained. “If the product can’t make it to the store shelf, it doesn’t do anyone any good. With this in mind, R&D labs need to work outside of their bubbles and follow procedures that represent the real world.”
In the lab, mimicking the in-plant process works better than relying on calculations or assumptions, Mr. Weilert added. “If the process time is going to be 20 minutes in the plant, then the lab dough should be evaluated right after mixing and again after 20 minutes of floor time,” he said. “This will help identify potential concerns related to the way the dough ages."
To improve the transfer from lab to production floor, R&D staff must follow production times and temperatures, use drop or shock tests to emulate the jarring of conveyors and run production scenarios such as overmixing and overproofing to see how the product responds, Mr. Weilert advised. Typically, he prefers to see new formulas run on current or standard equipment. Sometimes, however, that’s not feasible when products require new scaling weights, pans or formulations. “When this is the case, it is good to get the product to 90% of target during formulation and tweak processing to pull it up to 100%,” he said.
Establishing precise quality parameters should be the first step in the process of replicating a product on a production line, noted Mr. Bohrer. AB Mauri Fleischmann’s offers a roadmap tool, also known as a spiderweb chart, to assist in creating a methodical step-by-step approach with clear controls and documentation.
“If a bakery’s process is well controlled and systems are in place for quality checks and follow up, then tweaking should only be to adjust those individual bakery differences such as water hardness, fermentation systems and other factors,” Mr. Bohrer said. “All too often, tweaking of formulas, new and old, is done in response to a system or problem that the bakery should be addressing. Band-aiding is an expensive ongoing cost that bakeries just cannot afford.”
Many mistakes can be avoided by implementing standard operating procedures (SOPs) and good manufacturing processes (GMPs). Such steps include frequently monitoring the temperature and amount of metered water and confirming the proper scaling of ingredients such as flour and sugar, Mr. Tull added. “We suggest proper SOPs and check-off sheets to ensure any metered ingredients are scaled within parameters,” he said.
Bringing in the technical experts who understand the overall artful dance of product development can save time and money. “Ingredient knowledge allows them to understand cause and effect within formula changes, and production expertise allows them to envision the manufacturing challenges as they formulate,” Mr. Weilert said. “Anyone who has lost a dough or had to dig bread out of pans by hand because the volume was too low for the depanners knows the obstacles that production faces. If projects are particularly challenging, assistance from an experienced technical services team early in the lab development process can speed up the timeline and improve the chance of success.”
Maintaining the integrity of the final product remains the real bottom line no matter how large the order, according to Rob Burch, c.o.o., Labriola Baking Co., Alsip, IL. “We don’t modify our formula to fit the equipment. We modify the equipment needs to fit the formula,” he said. The artisan baking company won’t take a project that is sizable until it believes it can be made up on a machine first. “There is not much sense in sending a handmade sample to a customer when you know it will be done by a machine when they start ordering,” Mr. Burch explained. •