Introduction to DCS Architecture: Understanding its typical structure

With the advent of DCSs, many of the problems of their predecessors (DDCs, or Direct Digital Control) were solved. For starters, DCSs distributed functions much more effectively; controllers, I/O, operator stations, data loggers, alarm systems, and configuration stations were now housed in separate, independent units. A key feature of the system is redundancy for the main data network or data highways, controllers, I/O, fieldbuses, and, in some cases, fault-tolerant workstations.

The evolution of communications and hardware has dramatically altered the structure of control systems. Communication technologies such as Ethernet and TCP/UDP/IP, combined with industrial communication standards like OPC and open protocols, allow for the easy integration of third-party applications into control systems. Likewise, object-oriented design, software components, and implementation support tools have facilitated the development of improved user interfaces and the implementation of reusable software.

Today's leading DCS manufacturers offer all these features and can fully integrate I/O fieldbuses such as FieldBus and ProfiBus seamlessly. This means that new controllers can link devices or interface to integrate multiple I/O based on FieldBus, DeviceNet, AS-Interface, HART, ProfiBus, and conventional (point-to-point) I/O into a single system. The following figure illustrates this.

DCSs have dominated industrial process control for years, improving their performance and reliability over time. As hardware costs have steadily decreased, DCS design has become increasingly modular, making them available even in small plants.

Main Components of a Distributed Control System

In the following figure we can see the main components of a typical distributed control system (DCS).

The operating interface is typically a rugged personal computer (PC) with standard keyboards, mice, and LCD monitors. Additionally, operating consoles may include screen arrays (2, 3, or 4 screens) to display the entire process and facilitate screen operation.

Personal computers also provide the speed, sufficient memory, and hard drive capacity to efficiently meet the requirements of engineering workstations used for system configuration and diagnostics. Furthermore, the price-performance ratio of PCs has led major DCS vendors to use them for integrating control system software. Operating systems such as Windows XP (for clients) and Windows Server 2003 (for servers) are currently in use, although industrial control systems now support Windows 7 and Windows Server 2008. The days of the beloved Unix or Linux in the industrial sector have passed, and major manufacturers no longer support them.

The operating equipment is used to monitor process conditions, manipulate setpoints, and receive and send commands. It is typically located in a Central Control Room. From here, the operator can (1) view the information sent by transmitters from the process areas and (2) change the control conditions using an input device. The control units are distributed throughout the process areas, performing two functions at each location: reading or measuring analog and discrete variables (inputs) and generating output signals for actuators to change the process conditions.

Input and output signals can be analog or discrete. It is understood that the signals must be transmitted or communicated from the central control room to the remote locations where the controllers are located. The communication path can be point-to-point (control room-controller) or via a high-speed data bus that interconnects all the controllers and the main control room. In some cases, this communication may be a wireless connection via radio, microwave, or satellite. Obviously, when control is required via these links, some type of redundancy will need to be included to minimize link failures.

To conclude this introduction to DCS architecture, I want to apply these concepts to the Invensys Foxboro IA Series DCS (and the InFusion ECS from the same manufacturer). The following figure shows the simplest architecture but with the main components of this DCS:

In the next article we will detail the functional components of a DCS.

I want to know what you think. Come on, I want you to comment, now!