Feed Formulations and Manufacturing

Feed Formulation

Definition and History

Feed formulation is defined as the exercise of using mathematical formulas to determine volumes of ingredients and additives to blend to create compound feeds that meet the known nutrient requirements of targeted species while achieving production goals at an optimized cost.  It is extremely important in the manufacture of animal feed for livestock, poultry, fish, and swine that are raised for their commercial products as well as for domestic pet food. 

The output of a commercially raised animal’s products is directly proportional to the nutritional intake, especially in the case of dairy cows.  Most pet foods are designed to be the sole source of nutrition for the pet, making proper nutrition in the food for both commercially raised animals and domestic pets crucial for good health.  In the case of commercially raised farm animals, feed cost typically makes up 65% to 75% of the total production costs. 

Two basic prerequisites are implied in any feed formulation process.  The first is that the essential nutrients along with their needs for the targeted animals are known to the fullest extent possible.  The second is that the candidate sources and raw material products used to compose compound feed are clearly outlined and suitable for feeding animals. Once these prerequisites are met, feed formulas can be further refined in many ways. 

Effects of processing on the quality of raw materials, functionality of ingredients, nutrient requirements at different life stages, dynamics of production systems, and market expectations are all factors that must be taken into account when determining the exact final composition of an animal feed product.  It requires a thorough understanding of the nutrient requirements of different classes of animal species and how they vary. 

For example, proper nutrition for egg laying, meat, and breeder chickens can be quite different, as can nutritional requirements for meat and dairy cows.  Feed programs must meet nutritional requirements based on the animal’s stage of growth or production, genetic capacity, health, and facilities.  It also requires knowledge of the nutrient composition, constraints in terms of both nutrition and processing, cost, and availability of raw feed ingredients.  Different proportions of ingredients can impact feed flow through the mill, pellet quality of the diet, response of the diet to feed additives, and gut health of the animal. 

In the case of pet food feed formulation especially with niche products, the color, smell, shape, and particle size of the final product are important for marketing purposes and the perception of the consumer. Even though these factors may have little impact on the actual nutritional composition of the feed, they still must be taken into account during the feed formulation process. 

Ultimately, optimizing feed formulation comes down to economics of both manufacturing the feed and satisfying the animal’s nutritional requirements at minimal cost.  This is defined as least cost feed formulation and is the standard method used by most feed manufacturers.  It involves feed production as the most cost-effective level in terms of resources, time, energy, and money by minimizing input while maximizing output.  Depending on the specific animal, manufacturing process, and product output, it can vary based on optimizing the best feed conversion efficiency of the animal or the least cost per unit of product output.

The process of feed formulation has evolved from rudimentary hand calculation methods using the crude nutrient content of raw materials to advanced computer programs that incorporate the nutrient availability values of numerous ingredients and additives. 

The first feed formulation techniques for composing animal feed were used in the early 1900s.  Fish feeds were formulated on the basis of proximate composition, availability, cost, and the palatability of raw materials.  Typical raw materials used were beef heart, beef and pig liver, meat meal, and fishmeal.  This period introduced rapid technological innovations for feed production in the United States and Europe. 

During the 1920s and 1930s, the pelleting process was introduced which created opportunities to use unpalatable, difficult to handle, and varying in density raw materials for the manufacture of uniform and easy to handle pellets.  In 1944, the first nutrient requirement tables were published for farm and laboratory animals after a directive by the Committee on Animal Nutrition of the National Research Council.  The NRC tables are updated regularly and are still used as a global standard for formulation and research for nearly all commercial animals, domestic pets, and laboratory animals.  Formulations became more complex during this time as essential trace minerals were identified, synthetic vitamins were developed, and the addition of antibiotics came into practice. 

There was a massive expansion of new companies in the animal feed business as well as the introduction of extruded pet food in the 1950s.  In 1951, a paper was published by Frederick Waugh which tested the linear programming method for feed formulation – “The Minimum-Cost Dairy Feed – An Application of Linear Programming”.  Linear programming is a technique for the optimization of a linear objective function that is subject to linear equality and linear inequality constraints.  In the case of feed formulation, the constraints are product specifications that are defined as the minimum and maximum levels of nutrients and ingredients that the products can have.  The objective is to find the lowest cost at which various ingredients can be combined to make the product. 

In 1967, the first published book on computer based formulation was published and entitled “Linear Programming and Animal Nutrition”.  In the 1970s, computers became affordable in large industries and became widespread for feed formulation in the 1980s.  Today, advances in animal nutrition, the complexity of compound feeds, and software have made feed formulation a very advanced technique that can be very complicated but also able to incorporate many different variables into the calculations with the use of software. Feed formulators can use different feed formulations that are readily available on the Internet or in textbooks but can also apply intricate and comprehensive approaches that are facilitated by sophisticated mathematical solutions and advanced algorithms.  Methods increasingly incorporate increased nutrition knowledge, diversification of production goals, risk management, uncertainty relative to access safe and high quality raw materials, and commodity price volatility. 

One factor in the use of feed formulation that can reduce its effectiveness in practice is the variability of raw materials.  The nutrient composition of raw materials used to manufacture animal feed often has variation, especially in protein composition.  Current testing methods are insufficient for determining differences in composition.  The use of DDGS (Dried Distiller’s Grains with Solubles) is increasing in animal feed manufacturing and the variability within a batch is much larger than that of corn.  One study showed that the variability in fat content of DDGS can be four times greater than that of corn and that the variability in protein content of DDGS can be sixteen times greater than corn. 

As advances in feed formulation and manufacturing continue, fast, non-invasive, and effective testing methods that can accurately determine the variation in raw materials will be needed.

Composition, Products, Methods, and Software

A standard feed formulation process using software will contain at least three databases: nutrients, ingredients, and products.  Nutrients play the central role in the formulation system because they link both ingredients and products.  The database can be a simple list with corresponding units or can also list the species to which the nutrients can be relevant. 

The term nutrients does not necessarily refer to only macro or micro nutrients in formulation.  Any measurable parameter that can be optimized in the formulation process can be called a nutrient.  Physical parameters such as color, smell, and density and functional parameters such as attractability and palatability can be treated as nutrients as well. 

An ingredient database will store four pieces of information: ingredient names, availability, nutrient composition, and cost.  The nutrient composition table of an ingredient in the database is referred to as the matrix.  It is used to reflect the effects that an ingredient can have on the nutritional value of the feed. 

Matrix values can be constant and published values of ingredients in NRC books and similar resources can be used.  However, it is important to not rely solely on these published values as the nutrient composition of feed ingredients can vary widely all the time.  Values should be updated based on the analysis of ingredients that are available for formulation.  Since it is not practical to conduct an extensive analysis of every ingredient from every batch, manufacturers rely on analyzing a few components from which other components can be predicted. 

For example, crude protein value can be used to estimate amino acid composition.  NIR spectroscopy has enabled fast and accurate analysis of incoming ingredients used for feed manufacturing.  It offers the advantages of being fast, non-invasive, and the ability to determine multiple parameters with a single measurement and is especially valuable as a tool for determining variation within batches of raw materials.  Product specifications define the nutrient levels desired in the formula and the ingredient inclusion levels.  Upper and lower limits for each nutrient and ingredient are set as constraints and drive the final solution for the formulation. 

Potential costs must be considered when setting constraints.  Common least cost formulation constraints include meeting nutrient requirements, minimizing excreted nutrients, meeting required dry matter intake, maximizing carcass fat quality, and simplifying feed manufacturing.

In general, there are six defined steps in the feed formulation process.  The first is to define the animals that will consume the formulated feed.  Species alone is insufficient for this process.  Weight, age, genetics, and end use of the animal product must all be considered. 

The second is to select the right source of nutrient specifications.  Experienced nutritionists will consider the starting base information and then adjust to suit the individual needs of the farm.  For example, a chicken broiler farm that is experiencing early deaths from the fastest growing birds will adjust diets to feed lower nutrient and energy levels than normally recommended. 

The third step is to list all ingredients with characteristics and prices.  Different types of the same ingredient must be carefully considered.  Wheat can vary greatly in crude protein level depending on whether it is hard or soft and even wheat of the same genetic stock can vary based on different climate and growing area.  Similar potential variation exists for many raw materials. 

The fourth is deciding on the minimum and maximum allowances.  Some ingredients rarely have a maximum level, such as corn and soybean meal.  Fish meal is a good example of an ingredient that must be limited. It is expensive, animals can be turned off by excessive fish oil, and products can be tainted by the smell.  Nutritionists must use experience, discretion, and updated product information when deciding allowances. 

The fifth step is using a feed formulation software program to prepare the formula.  Many different programs exist with varying capabilities.  Selection of a suitable software depends on careful analysis of needs and cost benefits.  Updates are continuously required not only for usability, functionality, and data input but to meet changes in organizational functions like purchasing, regulatory compliance, production, quality control, and sales. 

The sixth and last step is to review and adjust the formula to production limits.  Formulas are reviewed continuously after runs and must be examined from the manufacturing point of view.  Loading allowances and the handling of materials at different proportions as the ingredients go through the process must work in practice as well as on paper. 

Ultimately, the feed formulation process is not just about numbers but the use of experience and an understanding of the practical aspects of feed manufacturing to be successful.

In most animal diets, crude protein is the most expensive portion and is usually the first nutrient to be computed in diet formulation.  Nutrients that are required in many types of animal feed include the following: crude protein, crude fat, crude fiber, carbohydrates (often expressed as nitrogen-free extract), ash, microminerals, trace minerals, amino acids, fatty acids & lipids, and vitamins.  Energy is expressed as metabolizable energy or digestible energy.  Moisture is expressed in the nutrient database but dry matter is used in the ingredient list. 

Once the nutrients and products are listed in the software with costs, limits, and constraints the software can calculate the optimal least-cost formulation.  Advances in software and programming algorithms have expanded the scope of what feed formulation can do.  Advanced software can provide means to integrate and manage multiple plants, products, and users.  It can automate several formulation functions such as updating nutrient values of ingredients based on testing and generation of product labels.  Analytical tools can examine the impacts of ingredient and nutrient restrictions on formula cost and suggest solutions for cost optimization. 

Many feed formulation software programs use shadow pricing to monitor how much the price of a given ingredient must fall before being included in the formula.  Other algorithms and programming techniques used in feed formulation include stochastic programming (SP), goal programming (GP), dynamic programming (DP), non-linear programming (MLP), and fuzzy programming method.  SP is designed to help account for nutrient variability.  GP is used along with linear programming to help overcome a supply of certain nutrients and achieve nutritional balance in selected feed mixes.  DP helps solve complex models by breaking them down into smaller models.  NLP assumes a non-linear relationship and helps better measure animal performance for things like milk yield and weight gain. 

Fuzzy programming technique overcomes the basic assumption of using a deterministic coefficient for objective function and constraints.  All techniques have advantages and drawbacks.  Expertise is required to properly implement them for feed formulation.  Feed formulation has been described as both a science and an art, as it requires not only mathematical prowess but also the experience to go beyond the math to make formulations work in the real world. 

One such factor is the variability of raw materials and more advanced testing methods will be required to continue to use feed formulation as an effective technique for animal feed manufacturing.

Manufacturing

Definition

Feed manufacturing is defined as the process of producing compound animal feed from raw agricultural products.  It is a means of converting raw materials of widely ranging physical, chemical, and nutritional composition into a homogenous mixture that is suitable for producing a desired nutritional response in the animal that eats the mixture. 

Generally, feed manufacturing is a physical process with minimal chemical changes.  However, some raw materials can undergo extensive processing prior to inclusion in a compound feed.  According to the American Feed Industry Association (AFIA), there are five basic steps in animal feed manufacturing: receiving of raw ingredients, creation of a formula (feed formulation), mixing of ingredients, packaging, and labeling.  

Over nine hundred agricultural ingredients and coproducts are approved for use in the United States.  Corn makes up over half of the diet composition for both poultry and livestock and is the most important ingredient in most animal feed manufacturing processes. 

The manufacturing process can vary depending on the exact product but in nearly all circumstances, the following steps are included: raw material storage, selection, weighing, and grinding, mixing of dry ingredients and addition of liquids, pelleting of mixed feed, and final blended feed bagging, storage, and dispatch.

Animal Feed

The first step in animal feed manufacturing is storage of the raw materials.  Some products may only be available seasonally and price fluctuations can also necessitate raw material storage. Raw materials are inspected and tests are conducted for quality control and to verify nutrient composition, although current wet chemistry testing methods are insufficient for such testing on a large scale. 

Proper storage is essential to prevent physical losses and is important for quality control.  The weighing of raw materials requires great care, especially in the weighing of small quantities of raw materials that can have a large effect on the growth performance of animals.  For example, omitting 25 kg of bran in a recipe that calls for 400 kg of bran will have a much lower nutritional effect than omitting 1.5 kg of vitamin premix from a recipe requiring 2.5 kg of it. In the case of livestock and ruminant animals, the raw materials will contain both animal fodder and forage, as a certain amount of forage is required for food digestion. 

Raw material grinding can occur at different parts of the manufacturing process.  It can occur before or after weighing.  Some materials can be blended together before grinding to facilitate the grinding of materials that do not do so easily. 

Grain materials are run through a roller or hammer mill to reduce particle size and create uniformity.  Increasing the number of particles and surface area per volume is critical for increasing access to digestive enzymes. 

The desired fineness of particles varies by the type of animal. Poultry raw materials should be ground more finely than those for cattle or swine.  If the raw material is ground for extrusion, it is typically ground finer than if it is to be used for pelleting.

The difference between pelleting and extrusion is primarily the high temperature and short time used for extrusion, resulting in a lower density product with less nutrient loss.  The primary mixers are used to produce a homogenous blend of all the raw materials desired in a formulation.

Proper mixing is crucial to ensure that each animal receives a balanced portion of nutrients at each feeding.  They can be horizontal or vertical, although vertical mixers are seldom used in an industrial setting these days and are usually only found on small farms. 

Horizontal mixers contain a shaft and blades that move material from one end to the other, allowing it to tumble as it goes. Material is discharged from the bottom and most horizontal mixers are capable for blending material that is liquid up to 8%.  Horizontal mixers are typically two to three times faster than vertical mixers and vertical mixers cannot handle liquid material. 

Before pelleting, the mixed feed materials go through a scalper that removes any undesired material.  Material is fed into a conditioning bin and flows by gravity into a pellet mill.  The conditioner is fed with dry saturated steam. 

Steam serves two purposes before pelleting.  The high volume of water vapor condenses on the surface of feed particles, which wets and softens them for better compression.  The high temperature of the steam can help to gelatinize raw starch present in vegetative ingredients, which assists with adhesion in firm pellets. 

The temperature of the feed in the conditioner should be between 80°C and 85°C.  The moisture before conditioning should be between 12% and 13%. Approximately 14% moisture is ideal after conditioning. Pelleting can demand up to 50% of the total energy in the feed manufacturing process. 

There are two standard types of pellet mills: flat die mills and ring die mills.  In flat die mills, the die rotates a roller that presses the powder through the holes in the die.  A cutter then cuts the exposed pellet free from the die. 

In a ring die mill, there are radial slots throughout the die.  Power is fed into the inside of the die and spreaders evenly distribute the powder.  Two rollers compress the powder through the die holes and two cutters cut the pellets free from the outside of the die. 

After pelleting, the soft feed mixture is transformed from a meal-like material at a bulk density around 0.4 g/cc to a pellet with bulk density of 0.5 g/cc to 0.6 g/cc.  It is important to consider and adjust the variables in the mechanical operation. The pellet die thickness as related to the diameter of the hole is a factor in pellet quality. 

The speed of rotation needs to be taken into account for each die thickness/hole diameter combination.  The speed of the feed affects the moisture/temperature relationship, which has a strong effect on pellet quality. 

Atmospheric conditions also affect die selection and operational setting.  Interestingly, smaller die holes require more effort to force the meal through the hole, which means that the smaller the size of the pellet, the greater the cost of manufacturing. 

Pellets are dried and cooled after being discharged from the pellet mill.  They exit at a high temperature and are typically fed on a moving wire belt or sectional belt of perforated metal trays.  The cooler can be a single deck with pellets discharged at the end opposite the intake or a double deck with two belts in the same enclosure where the pellets return to the same end from where they entered. 

Air from a centrifugal fan can flow from the bottom of the layer of pellets which removes fine particles that separate from the pellets.  The air discharges into a collecting system that continuously feeds them into the mill for repelleting. 

If crumbles are desired (mostly for fish feed manufacturing), the cooled pellets can be ground on corrugated rolls and then the product can be sifted into the desired size of crumbles or granules.  A final sifting occurs using a sieve before the pellets are sent off for packaging. 

The final product is usually distributed in sacks that are filled directly from mixers or from holding bins.  Representative samples of batches are taken for final quality control tests before product distribution.

Pet Food

Pet food can be manufactured in multiple forms.  It can be made dry by a baked or extruded process, wet in a can, pouch, or tray, or in snack and treat form. 

Dry pet food manufacturing is very similar to the process for making traditional animal feed.  Ingredient mixing can differ in that many of the ingredients are animal or fish derivatives and by-products. 

Typically, pet food is made from food that has been deemed as unfit for consumption by humans although it still must meet safety and quality requirements.  Most recipes call for a meal form where meat and animal derivatives are cooked to remove the fat and then dried to make a meal.  Other ingredients come from cereals, grains, and vegetables. 

As with animal feed, feed formulation is used to carefully construct recipes that meet the target animal’s nutritional needs.  It is also important to ensure that the texture, smell, and taste are palatable to the animal. 

Pet food for some animals (mostly small mammals) is manufactured in pellet feed using extrusion just like animal feed pellets.  Typical dry dog and cat food only differ in the shape at which the material is extruded out of the mill into kibble form.  It is then dried and cooled like animal feed.

Unlike animal feed pellets, dry food often enters a revolving drum for coating with various materials before packing.  Applied coatings are a mix of flavors to enhance taste and preservatives to prevent spoilage and extend shelf life.  If dry food is baked, the dough is pulled out after steam conditioning, cut into the desired shape, and baked to form the final product. 

Manufacture of treats and snacks can be done using extrusion, baking into biscuit form, or injection molding, where ingredients are mixed together before injection into a shaped mold.  After cooling, the treats are released from the mold and this process is typically used to make chew treats. 

For wet food, recipes are formed and made in the same fashion but after ingredient selection, weighing, and mixing, the material is then placed into a hermetically sealed can, tray, or pouch.  The food is cooked in the package and meant to stay sterile and fresh for the shelf life of the product or until it is opened. 

For many forms of wet food, a mix of water, thickening agents, and flavors are added to form a gravy or jelly during the cooking process.  The temperature and cooking time are carefully controlled and regulated during the process to prevent spoilage, optimize taste, and protect the required nutritional content.  The final product is labeled, packed, and dispatched to customers. 

While the animal feed and pet food manufacturing processes are carefully regulated, the variability in the raw materials and inability of current testing methods to be implemented on a large scale has made it difficult to ensure that nutritional content of finished products is consistent.  One method that has been implemented for fast, non-invasive, and large-scale testing of animal feed and pet food products is NIR spectroscopy. 

Feed Formulation, Manufacturing, and NIR Spectroscopy

NIR spectroscopy has emerged as a fast, non-invasive testing method for parameters of interest in animal food quality control.  It offers the advantages of little to no sample preparation, the ability to be used for large-scale testing, and is able to determine multiple parameters with a single measurement. 

The use of NIR spectroscopy has evolved from research and development studies in the animal food industry to being used extensively by many large companies in the animal feed, animal forage, and pet food industries. 

Examples of macronutrients and energy components that can be determined using NIR spectroscopy include dry matter, crude protein, fiber, crude fat, carbohydrates, ash, metabolizable energy, and gross energy.  NIR spectroscopy can be used for analysis from the beginning to the end of the animal food manufacturing process.  It is used for raw material ID and component analysis as well as a tool for checking variation and shipment within batches.  

NIR spectroscopy is an aid for feed formulation by both determining variation in materials and giving manufacturers the ability to make mixing adjustments on the fly.  It is used as a real-time process control tool in the feed mill. When products are finished, NIR spectroscopy can be used for the final quality control check.  

While the principles of NIR spectroscopy have been well-known for many years, recent technological advances have enabled its advancement as a practical tool in industry.  Handheld and portable instruments have enabled analysis in the field.  NIR spectroscopy requires the creation of calibration models to correlate NIR spectra to parameters of interest. Companies have created pre-built calibration models for parameters of interest, reducing both the labor and costs required to implement NIR spectroscopy. 

The largest pet food company in the United States has successfully adapted NIR spectroscopy for determining quality assurance.  Material arriving on trucks is scanned and tested inside the original package before it even enters the facility and subsequent tests are performed throughout the manufacturing process.  This company has had so much success with NIR spectroscopy that they even share the technology with competitors in the interest of creating safer pet food products for the entire industry, ultimately benefitting all manufacturers, customers, and pets themselves. 

Third-party programs have enabled the use of web-based technology that supports database management, quality control, and trend analysis to optimize processes and protocols for animal food analysis. 

Research and development for new technologies and products is ongoing.  As the worldwide demand for food continues to increase in coming years, NIR spectroscopy will play an essential role in many segments of the food industry, including animal food.