Detailed information on energy research

Projektinfo 07/2014

Keepers of livestock including cattle, pigs and poultry expect a consistently high quality of feed for their animals, regard-less of the fluctuating properties of natural raw materials. This customer’s requirement is a top priority for producers. Energy efficiency is becoming an issue due to cost pressure and environmental awareness. Researchers at the University of Bremen are developing an expert system for process control. The system will serve to minimise energy use while at least maintaining product quality.

Producers in the animal feed industry process, depending on market availability, different natural ingredients to form a mixed feed with a constant and controlled quality. Despite fluctuating ratios of the components, the animals receive the exact amount of nutrients such as proteins, fats, trace elements, and vitamins that they require on a daily basis. Not only the ingredients, but also the consistency of the feed has an important effect on usability, and ultimately on animal health.Every delivered batch of raw materials must be individually assessed and analysed for specific product parameters. For best results, many companies go by the principle of “truth is not so much taught as caught”. Depending on the material properties, experienced staff adapt the process and draw upon the intuitive knowledge and instinct gained throughout their long professional experience. For example, they are able to interpret machine noises or literally feel the required fine­tuning by crumbling the feed pellets in their hands. Since pellet quality often correlates with energy expenditure, many plants produce over the required level, and therefore consume more energy than necessary.

Producing animal feed with less electricity and heat Expert system improves energy efficiency and product quality in a compound feed plant

This research project is funded by the

Federal Ministry for Economic Affairs and Energy (BMWi)

Increases in the prices of energy sources increasingly motivate compound feed producers to optimise their energy consumption. An inhibiting effect is that energy consumption is not yet recorded in detail in many stages of the production process. Currently only few compound feed plants in Germany are obliged, according to Section 41 of the German Renewable Energy Sources Act (EEG), to run an energy management system as they consume over 10 gigawatt hours per year.

Energy efficiency and product quality brought together under one roof In the Fu2 expert project, researchers at the University of Bremen aim to increase the energy efficiency of the pro­duction process together with a feed producer in Lower Saxony. They are currently developing a computer­based expert system that draws on the experience of the staff for adaptive process control, and which also accounts for and optimises energy flows using a wide range of measurement instrumentation. The system is to recognise the relations between input products, process control and product quality and offer timely recommen­dations for process control.

From the raw materials to finished productsShredding, mixing and pressing are the basic processes in feed production whose energy flows were analysed by the researchers (Fig. 2): the different feeds of the parti­cipating producer consisted of 12 – 18 components, in­cluding corn meal, soybean meal, and rapeseed meal. Vibrating sieves first separate the fine fractions of the individual raw materials from the coarser components. The latter are uniformly crushed in a hammer mill or a roller mill. Afterwards, a mixer mixes the batch along with liquid ingredients such as fats or acids. With the addition of molasses, the result is a floury feed, as it is mainly used for laying hens. Pellet production has the largest share of specific energy consumption. For a final compaction in the pellet mill, saturated steam heats the middlings to a temperature of about 80 °C. So thermal energy is required in addition to electricity.

Online measurement technology monitors productionSuitable Internet­capable measurement instruments that provide information on energy flows and production pa­rameters are the requirement for optimising the energy efficiency of processes. The researchers installed an ener­

gy monitoring system that registers the electrical energy of each machine as well as the thermal energy in certain areas. Fig. 2 shows the energy use of individual areas. With the combination of energy monitoring and produc­tion control information, it is now possible to relate energy consumption to individual productions and compound feed types.Energy monitoring is complemented by the monitoring of key production parameters. A near­infrared measurement system in the receiving depart­ment directly provides information on the properties of the input materials. It monitors moisture, protein, starch, fibre, and fat content. Additional devices are employed directly in production. In this way, the addition of saturated steam in the pellet press section can be optimised depending on the input moisture content of the milled starting product.The grain size distribution of the milled raw materials is monitored by a photo­optical particle size analyser (CPA). With this information, the opti­mal press values for pelleting can be set. The researchers integrated the CPA device between the batch mixer and the molasses mixer directly downstream to the milling units by adding an automated sampler into the product stream. An additional advantage of real­time grain size analysis is

2 BINE­Projektinfo 07/2014

Fig. 3 The control concept of the expert system optimises energy use.

Fig. 1 Compound feed plant in Damme

2 % · Loading

61 % · Pelleting and cooling

29 % · Milling and mixing 4 % · Compressed air 4 % · Reception

Extended production control

Automated production control Production

Apply

Knowledge management system

Manual input

Current value

Production log

Formulation database

Target value

Sensors

Production unit

No Yes

Fig. 2 Share (in % / tonnes compound feed) of the production steps in overall energy consumption

that inadequate machine settings (e.g., rotational speed of the hammer mill or the gap width of the roller mill) is immediately detected. Wear or defects are also detected by the particle analysis.

What are the potential savings?The researchers examined the entire production chain for potential optimi­sations. Batch­wise operation, for example, leads to idle processes that can be minimised with intelligent planning. Furthermore, the optimal batch size of a given production batch can ensure an optimal plant capacity utilisation. However, the scientists see the greatest potential in the control and regula­tion of the press operation. This step consumes nearly 60 % of the total energy, mostly for steam generation, compaction and subsequent cooling. Both processes are currently manually controlled. The parameters are specified, but their regulation is not yet optimised. Experienced employees regulate steam temperature and pressure as a function of the desired for­mulation, product characteristics and the quality of the raw materials – all on the basis of their previous experience. However, this has two major drawbacks: first, the knowledge of the employee is not recorded anywhere,

3BINE­Projektinfo 07/2014

Fig. 4 Path logistics: the ingredients are taken from large silos.

Fig. 5 Hammer mill and saturated steam production

Fig. 6 Route optimisation also helps to save energy.

Energy in the feed industry

Approximately 22.9 million tonnes of compound feed were produced for the supply of farm and domestic animals in Germany in 2011. The largest share is pig feed with about 44 %, followed by cattle (29 %) and poultry feed (25 %). Following the differing customer requirements, compound feed manufacturers often produce several hundred different feed formulations. Energy consumption for the production of compound feed varies between 15 and 80 kWh/t depending on the refinement. As a rough estimate, the total annual con­sumption of the feed production industry in Germany is about 1.0 billion kWh per year.

so it is lost as soon as the employee leaves the company or calls in sick. Secondly, too much energy is generally expended with this method, just to ensure that the final product complies with the requirements.Energy consumption can be lowered for various feed formulations – as was demonstrated in test series. De­pending on the feed formulation, a few machine para­meters must be adapted to the moisture content and particle size of the raw material. For example, sieve size, rotational speed during the milling process, length of the compaction channels in the pelleting process, and finally the steam temperature all play an important role.

Artificial intelligence makes recommendations A newly developed, adaptive expert system is to suggest the formulation­dependent adaptation of machine pa­rameters to the raw material properties, and also specify the necessary machine conversions. In particular, it op­timises the application of steam, or rather energy, du­ring conditioning and compacting. To achieve this, the expert system incorporates employee experience, and the evaluation of the test series, into a simple control system. This system is based on fuzzy logic and is able to process fuzzy information, such as “a little more” or “slightly reduce”, and to concretise this in learning pro­cesses. At present, the system is in the implementation phase at the compound feed plant of the project part­ner. It is being systematically evaluated and adapted to the requirements of the users.

New formulations – new motivationUsing the results of the test series, the project partners were able to improve the energy efficiency in the areas of milling and pressing. While optimising the product portfolio, the employees contributed their expertise in an intensive exchange of experience, and developed a deeper understanding of the involved energy relation­ships. The staff were highly motivated to make a major contribution to the improvements. With the integration into production control, they had the ability to track energy use in production in detail. Effects of changes to the process flow became more transparent and verifiable.

2 % · Loading

61 % · Pelleting and cooling

29 % · Milling and mixing 4 % · Compressed air 4 % · Reception

Extended production control

Automated production control Production

Apply

Knowledge management system

Manual input

Current value

Production log

Formulation database

Target value

Sensors

Production unit

No Yes

BINE Projektinfo 01/20104

Benefits for the food industry The researchers aim to generalise and modularise the developed concept by syste­matically abstracting the requirements. New requirements can be better and sooner considered, and it is ensured that the concept is not only specialised for a single application. The empirical knowledge base obtained during the project is to be supplemented by the use of deterministic models. The results of the Fu2 expert project can be transferred to production plants in other industries. In particular in the food industry, there are many processes in which the controlled heating and cooling of sensitive raw materials is required. For example, coffee beans are heated to 200 – 220 °C during roasting. Then they must be cooled back down to room temperature as soon as possible. Here, there are again the conflicting targets of quality and minimising energy consumption.An interesting test of the concept would likely be its application in a malt house. The quality of the finished malt product is highly sensitive to changes in input and process parameters. Production is divided into three sub­steps during which the method of the researchers could be used: First, the barley is soaked for a period of two days until the moisture content of the grain reaches 40 – 45 %. The barley then germinates for three to four days at a tempera­ture of 12 – 15 °C. In the final step, the green malt is kilned. At a kilning temperature of 50 – 95 °C, the moisture content of the malt drops to 3.5 – 4.5 % within one or two days. The entire process currently consumes about 750 – 780 kWh of thermal and 73 kWh of electrical energy per tonne of malt.

Project participants >> Project management: University of Bremen, Institute for Integrated Product Development (BIK),

Dr. André Decker, decker@uni­bremen.de, www.bik.uni­bremen.de>> Application partner: Austing Mischfutterwerk GmbH & Co. KG, Damme, Bernhard gr. Austing,

www.austing.de

Links and literature >> www.fu2­experte.de >> Forschungsstelle für Energiewirtschaft e. V., München (Hrsg.): Ermittlung von Energiekennzahlen

für Anlagen, Herstellungsverfahren und Erzeugnisse. Zielsetzung, Durchführung, Methodik, Kennzahlen. 1999. 235 S.

>> Decker, A.; Redecker, M. A.; Thoben, K.­D.: Wissensmanagementsystem zur Verbesserung der Energie­ und Ressourceneffizienz in der Produktion. In: Mühle + Mischfutter. Die Fachzeitschrift für Getreideverarbeitung und Tiernahrungs­Produktion Verfahrenstechnik im Schüttgut­, Lebensmittel­ und Non­Food­Bereich. Jg. 151 (2014), Nr. 4, S. 106 – 109

>> Redecker, M. A.; Thoben, K.­D.: An approach for energy saving in the compound feed production. In: Emmanouilidis, C.; Taisch, M.; Kiritsis, D. (Eds.): Competitive Manufacturing for Innovative Products and Services. Proceedings of the International Conference of Advances in Production Management Systems (APMS 2012). Rhodes Island (Greece), 24 – 26 Sept. 2012. Revised Selected Papers, Part I. Berlin, Heidelberg: Springer, 2013. XXXIV, 748 p., ISBN 978­3­642­40351­4, DOI 10.1007/978­3­642­40352­1_10, p. 73 79.

More from BINE Information Service>> This Projektinfo brochure is available as an online document at

www.bine.info under Publications/Projektinfos.

BINE Information Service reports on energy research projects in its brochure series and newsletter. You can subscribe to these free of charge at www.bine.info/abo

ImprintProject organisation Federal Ministry for Economic Affairs and Energy (BMWi)11019 Berlin Germany

Project Management Jülich Forschungszentrum Jülich GmbHDr Michael Gahr52425 Jülich Germany

Project number 03ET1036

ISSN0937 ­ 8367

PublisherFIZ Karlsruhe · Leibniz Institute for Information Infrastructure GmbH Hermann­von­Helmholtz­Platz 1 76344 Eggenstein­Leopoldshafen Germany

AuthorDr Franz Meyer

Copyright Title image, Fig. 1, 4, 5 and 6:Austing Mischfutterwerk GmbH & Co. KGFig. 2 and 3: University of Bremen

Text and illustrations from this publication can only be used if permission has been granted by the BINE editorial team. We would be delighted to hear from you.

Conc

ept a

nd d

esig

n: is

erun

dsch

mid

t Gm

bH, B

onn

– B

erlin

, Ger

man

y · L

ayou

t: K

ERST

IN C

ON

RAD

I · M

edie

nges

taltu

ng, B

erlin

, Ger

man

y

Contact · InfoQuestions regarding this Projektinfo brochure? We will be pleased to help you:

+49 228 92379-44 kontakt@bine.infoBINE Information ServiceEnergy research for application A service from FIZ Karlsruhe

Kaiserstraße 185­197 53113 Bonn, Germany www.bine.info

4 BINE­Projektinfo 07/2014