Getting the most out of a plant
The nature of dairy operations has changed over the past few decades. Small, local dairies with manual operations have become outdated and been replaced by larger units with factory-style production.
This trend has caused many and far-reaching consequences. Processes in small dairies were supervised and controlled by a few skilled people, who carried out most operations manually and also cleaned the equipment by hand at the end of each run. As dairies expanded, both the number and size of the machines grew, as did the number of manual operations required. Cleaning, in particular, was a laborious business – every machine that had been in contact with the product had to be disassembled and cleaned by hand at least once a day.
Cleaning-In-Place (CIP), introduced in the mid-1950s, is used at most of today’s dairies. CIP means that equipment no longer needs to be disassembled for cleaning. Machines are designed to be cleaned with detergent solutions, which are circulated through the production lines according to a set cleaning program.
Extensive mechanization of dairy operations gradually became a reality, with the result that more and more of the heavy manual labour was taken over by machines. Mechanization, together with the rapid expansion of production capacity, also led to a substantial increase in the number of operations that had to be executed. More valves had to be operated, more motors had to be started and stopped. The timing of individual operations also became critical. Operating a valve too soon or too late, for example, could lead to product losses. Every malfunction in the process, and every operator error, could have serious economic and qualitative consequences. Automation was the solution to handle these problems.
Automation is a fast-moving field. Only a few decades ago, process control systems were based on electro-mechanical relays, wired together in a logical pattern. They were replaced by hardwired electronic control systems, which were faster and more reliable, as they contained no moving parts.
The next improvement was programmable control systems with the logic expressed in data bits stored in an electronic memory, not in the physical arrangement of the wiring. This not only made it easier to modify the program whenever necessary, but also reduced the cost of the hardware.
In modern control systems, the growing capability and reduced cost of computers and microprocessors has been utilized to distribute control functions to local units. This gives the system as a whole more flexibility and a very high potential. The new processors can be used to control a single machine, or build up a total control and management system to make an entire plant more productive.
Totally integrated plant control
Nowadays, the next step in the evolution of automated processes is taken towards the totally integrated plant control system.
A plant consists of more than one process area, eg. reception, cheese and liquid milk production. Each area has a its own configuration of one or more Process Controllers and they will often have a User Interface for operators, handling product transfer from one process area to the other.
It is essential to keep track of production and economy in a plant. The Process Controllers generate a substantial amount of data from the process at all times, day and night, week and month. Knowing what is happening is a key to be able to run the plant more efficiently and economically.
The Process Controllers themselves provide all the raw data for the Manufacturing Execution System (MES), where the data can be further processed and stored in a database. This is handled by a separate computer.
A modern MES system is dedicated to handle large volumes of data. It computes and processes the data to transform it to useful information. Visualizing this information in various types of reports helps the users to analyse production economy, etc. and to assist in planning and making preventive maintenance forecasts.
Why do we need automation?
Several aspects must be considered when designing a dairy. Therefore, the final production solution of a plant is always a compromise between product-related, process-related and economic aspects, in which external demands on the plant must be satisfied. These external requirements relate to factors such as legislation, type and amount of product, product quality, hygiene, production availability, flexibility, labour and economy.
The product-related aspects include raw materials, product treatment and quality of the end product, while the process-related aspects include selection of process equipment to satisfy external demands. Even if the processing units in a plant are chosen primarily to achieve the stated product quality, various compromises must be made, particularly if many different products are to be manufactured.
Such considerations apply, for example, to the cleaning requirements of the equipment and its suitability for connection to the proposed cleaning system. Compromises must also be made on other matters, such as the consumption of energy and service media, and the suitability of the equipment to be controlled. When selecting process equipment, it is important to remember that the process control solution should also be considered.
Correctly applied process control, in which a thorough knowledge of products, processes and process equipment guides the design, has many advantages.
The most important are:
- Food safety
- Consistent product quality
- Production economy
- Flexible production
- Production control
Food safety is secured by the control system through the continuous supervision of equipment and processes. A malfunctioning machine will be brought to a safe status if a serious fault occurs, and a process fault will stop the related process. This system ensures the prevention of unwanted mixing of products, overfilling of tanks and other faults, which might cause product losses and production disruptions.
The process is monitored in exactly the same way during each production run, which means that the finished product will always have the same high quality after fine-tuning of all processing variables for an optimum outcome.
Precise control of the process means that product losses and consumption of service media, cleaning solutions and energy are kept to a minimum. As a result, the production economy of a well-designed and adapted control system is very good.
Flexible production can be achieved by programming the control system with various production alternatives and production recipes. Changes in production can be implemented simply by altering a recipe, instead of modifying the actual program.
The control system can also provide relevant production data and information in the form of reports, statistics, analyses, etc. The data becomes a tool for more precise management decisions.
- External factors
- Food safety
- Consistent product quality
- Production economy
- Flexible production
- Production control
The following definitions have been adopted to describe the level of control in the system:
- Manual control
- Unit control and supervision
- Line control and supervision
- Production management
All operations in the plant are carried out manually. Control modules are manually operated, but normally they are started or stopped from panels with push buttons, with no interlocking function. Some single valves, such as the diversion valve in a manual pasteurizer, may be automatically controlled, but the plant or line is still considered to be manual.
Unit control and supervision
Each process unit is operated from its specific operator panel. Each unit has a standardized way of communicating with other units and supervisory systems. The units either communicate with a limited number of I/O signals or with a communication link. The complexity of the control systems is low, so the demands on the local service organization are limited.
Line control and supervision
The operator supervises the plant or line from one or more User Interfaces. Process units, with their own specific operator panel, are normally supervised from central User Interfaces. Coordination of routings and operation of units is done from one or more plant PLCs.
Line control and supervision gives an excellent plant overview and facilitates increased plant functionality, i.e. operations can be carried out in a sequence and losses can be minimized by optimization of the process sequences. Changes in the process will require modification in the control program, and therefore demands on the local service organization are high.
Production and cleaning can be executed in jobs or batches, using recipes. The Production Manager can schedule batches from an operator station, which can be situated in an office. The operator of the process supervises the execution of scheduled batches from one or more operator stations. In a bigger plant, each operator station should encompass a dedicated production area.
Control of process units that have their own specific operator panels should be included in the execution of batches. One or more plant PLCs control the routings, and the plant server coordinates all activities in the plant. The history of the batches is stored in a database. The use of advanced technology means the control system is highly complex. Changes in the process will result in modifications of the plant models, recipes and programs, and therefore the demands on the local service organization are high.
Operations can be carried out in sequences, and product losses can be minimized by sequence optimization. The performance of the plant can be analysed, and the way a specific end product was produced can be traced back through production.
Requirements for a control system
Reliability, flexibility and economy are the most important requirements for a modern process control system.
This means that the control system should:
- Be reliable and easy to maintain
- Have a user interface that is logical, self-instructing and efficient
- Be based on off-the-shelf hardware and software
- Include software for diagnostic testing and modification
- Be easy to extend
Extending a control system
One of the most important requirements for a control system is the possibility to extend the system when required. It should be possible to build a system of any size, step by step, by adding standard components. A small process controller installed to control a reception line could be extended later with more controllers of the same brand that control milk treatment, filling, etc. At the same time, management routines could be added to existing controllers to feed data into management computers.
When extending a control system, it is very important that all control system components, from the remote sensor to the user interface, are easy to connect to each other in order to create a smooth functioning control system platform. Using products from a sole supplier will normally guarantee this.
How does the control system work?
Automation = Process Control and Production Management.
Automation means that all actions needed to control a process with
optimal efficiency are handled by a control system on the basis of
instructions that have been programmed into it.
Process Control System = The system executing Process Control.
It normally incorporates:
- User Interfaces, which are used by the process operator to communicate with the control system and the process.
- Process Control, normally a PLC (Programmable Logic Controller), which executes actual control of the process.
- I/O-system interfaces with control modules and transmitters in the process.
Management Execution System (MES) = The system executing
- It can also form a link to other company systems such as Enterprise Resource Planning (ERP) systems.
Logic is a fundamental concept in Process Control. It denotes the decision-making mechanism, making it possible to perform a given task according to a given model. The human mind is programmed by education and experience to perform a task in a certain way.
Figure 6.10.7 shows in a manual system, how an operator uses logic to solve a control problem, which involves supplying a process line with milk from a battery of tanks. He receives information from the process, e.g. that tank T1 will soon be empty, tank T2 is currently being cleaned, tank T3 is full of product, etc. This information is processed logically by the operator. The figure illustrates his train of thought – the questions and decisions he has to formulate. Finally, he implements his decisions by pushing the correct buttons on his panel to actuate the right valves, pumps and other control modules.
The operator has no great difficulty in solving this particular control problem. Even so, the potential for errors is always present. Detergent and milk could be mixed by mistake. The process line may run out of milk, resulting in burning-on at the heat transfer surfaces. Milk in the tanks may be wasted when the tank is cleaned. The risk of such errors increases if the operator is responsible for several similar sections of the process at the same time. He may be rushed and under stress, which heightens the risk of him making a mistake.
At first glance it is easy to assume that the operator is constantly faced with choices between many alternative solutions to control problems. A closer look reveals that this is not the case. After many hours of operation the dairy has verified the control sequences, which results in optimum product quality, safety and economy. In other words, the operator has acquired a more or less permanent control logic. He selects tanks according to established routines, uses a stopwatch to time milk drainage from a tank, so that he knows exactly when to switch to a full tank in order to minimize product losses, and so on. Each process can be analysed in this way and it is then possible, on the basis of the analysis, to determine the control logic that produces optimum results.
The control logic is stored in the form of a program in the specific process controller, which is normally a PLC.
All the transmitters and control modules in the process (4) are connected to the logic by the Input/Output (I/O) system (3). In this way, all the necessary information regarding temperatures, flows, pressures, etc. is transmitted to the logic of the control system. After processing of I/O signals and operator commands, the logic sets the correct output signals to actuate the control modules involved in the process. This is done in a certain order to comply with the logical conditions that apply to the process. The control modules send back feedback signals confirming that the commands have been carried out. These feedback signals are used by the logic as conditions, permitting the next step in the sequence to be actuated. The principal layout of a control system is shown in Figure 6.10.8.
If the output signal and the feedback signal do not match, an alarm signal is generated, trying to bring the related process to a safe state. This assumes, of course, that the fault in question can be predicted. As a process becomes more complicated, and demands on operational security and economy become stricter, the required control program (logic) has to be extended accordingly.
All user interfaces (1) are connected to the logic as well as local operator panels.
Efficient process control requires first-class electronic solutions in the process. The operation of the entire automatic process control system will be jeopardized if transmitters and sensors do not work properly.
The valve control system shown in Figure 6.10.9 is an example of distributed intelligence. Running a dairy of any size involves keeping track of hundreds or thousands of valves and operating them in different combinations and sequences. PLCs are dedicated systems to solve these control tasks in the shortest possible time. To do this, the PLC needs a channel for instant communication with all the valves. This makes the installation costly, but new valve control systems have been developed to provide an economical solution.
A modern system consists of a number of valve tops (1), one for each valve. The valve tops are connected to a common fieldbus cable and a common compressed-air line. The fieldbus cable is connected to a gateway communicating with the control system (2) and the power supply serving the valve tops. Several fieldbuses can be connected to the process controller to control the required number of valves.
Another important advantage of the system is that the valve top unit reports the valve status back to the control system. The modem scans the status of all valves continuously and instantly informs the process controller if a malfunction arises. This facilitates fault tracing and maintenance, especially since it is possible to disconnect individual valve units without disrupting the operation of other parts of the control system.
The fieldbus concept is also starting to be applied for transmitters and instrumentation as a whole – distributed temperature control and flow-metering are just two examples.
For the producer, the advantage is not only a significant reduction in installation and commissioning costs, but also the increased amount of useful information, which makes the total investment in a control system lower than for a traditional system.
Production in liquid food plants is becoming more complex as new and more complicated recipes are introduced. Strict recipe procedures must be followed to manage production and guarantee product quality.
The increased number of products demanded by producers means shorter production runs. In order to stay competitive in this situation, the efficient planning and running of production is a necessity.
The manufacture of 50 tons of strawberry yoghurt, for example, is called a batch. Instead of only executing conventional process operations, such as transfers to and from process units, the batch control system takes total control of production, from milk reception until the yoghurt cups are stored for distribution. The major benefit of batch control is that the system helps with all the necessary actions.
Using recipe management, a producer will have full control when introducing new products. If no new process equipment is needed, there is no need to call in external assistance to reprogram the control system. All procedures are edited on site using easy-to-understand tools.
All previous recipes are automatically stored and ready to use whenever needed in the future. Any existing recipe can be easily modified on line and stored as a new version or a completely new recipe.
Flexibility is maximized, as all recipes are scalable.
Control of production
The batch control system gives comprehensive on-line information about what is happening in production: production figures and totals to date, data on products scheduled for runs later in the day, and current problems related to production and lines. All this information can be displayed on any user interface connected to the network.
How does the data management system work?
Logging production data
Everything that occurs in the control system can be logged automatically in a database and tagged with a specific identity. This means it is possible to automatically compare parameters between production runs by producing a report, which will probably reveal any quality problem that has occurred in a specific period. In this way, it is possible to solve problems concerning inconsistent quality or difficulties in running a particular product.
In addition, it is possible to automatically produce a report defining all target and actual values during production – all events and any errors that occurred during a particular production run. Laboratory data can be added and connected directly to the tagged output.
There is a strong relation between the way the production is run and the resulting traceability. E.g. if a tank has been cleaned between two batches they are completely separated. If however the next batch is coming directly after the previous without any cleaning in between there will be traces of the first batch in the second batch. The first method will give a more precise separation but also a significantly higher cost of operation. The producer must define the desired target level of product separation depending on the type of production and need for traceability. The tracing system will record what has happened and the user will know if the batches are separated or not. There are systems and methods available to provide the required level of separation and traceability enabling a choice based on the balance between cost and need.
Increasing cost pressure on food producers is resulting in demands for tools that can help to decrease the cost of production. New forms of information solutions are opening opportunities to obtain information that can be used to identify improvement areas. There are many parameters that influence the cost of production. OEE (Overall equipment effectiveness) is becoming one of the most important measures and evaluates how effectively a manufacturing operation is utilized. It measures:
The top line figures are important to see the trends in a plant but to be useful they need to be broken down into actionable information. For this purpose different reports are made and "drill-down" functions are used to go deeper into different details, depending on where the possible problem is.
For day-to-day production, a report can be produced based on the optimal running scenarios decided during plant dimensioning, optimization or at later stages. Certain figures are shown for each unit. These figures represent specific set values (taken from the optimal scenario) compared with the actual figures.
The figures/unit could be:
Lines, pasteurizers, filling machines
- Ratio of production hours/idle hours
- Ratio of start/emptying/production run hours
- Ratio of circulation (or, for lines or machines: transfer selected, but pump idle) time/production time
- Amount and type of cleaning
- Ratio of product in tank period/24 hr
- Amount and type of cleaning
The figures for optimal and actual running are compared. If the figures differ by more than a certain value, they are highlighted. The reason could be operator error, less than optimal planning or that the plant is not dimensioned for that type of production. The deviation could also be caused by equipment faults (temporary problems). The findings and causes can be scrutinized later by the planning manager.
Planning and scheduling
Development of planning and scheduling systems within the industry has only just begun. The basic idea is to integrate the whole information structure of the plant or the entire company. Planning tools normally work on the ERP level, integrating customer orders with raw material and normal capacity availability, with a week-to-month time horizon. Scheduling tools are at the MES level ,working on breaking down the plans to detailed schedules for individual machines with a minute-by-minute horizon. To be efficient the scheduling tools need to work with real-time information and have enough flexibility to adjust schedules as the production progresses.