As of now, there is no strict boundary between DCS and PLC. In the eyes of most people, large systems are referred to as DCS, while smaller systems are called PLC. Of course, this statement is not entirely incorrect, but it is also not entirely accurate. Let us redefine this concept.
First, what are the differences between DCS and PLC?
1● From the perspective of development direction:
DCS has evolved from traditional instrumentation monitoring systems. Therefore, DCS inherently focuses more on instrument control, such as YOKOGAWA CS3000. DCS systems even have no limit on the number of PIDs (PID, Proportional-Integral-Derivative algorithm, is the standard algorithm for closed-loop control of control valves and frequency converters; typically, the number of PIDs determines the number of control valves that can be used).PLC has evolved from traditional relay circuits, and the early PLCs even lacked the capability to process analog signals. Thus, PLC has emphasized logical operation capabilities from the beginning.
2● From the perspective of system scalability and compatibility:
There are many control products on the market, and both DCS and PLC have many manufacturers producing and selling them. For PLC systems, there is generally little or no demand for expansion, as PLC systems are typically used for specific equipment. Generally speaking, PLCs also have few compatibility requirements; for example, it is quite difficult for two or more systems to share resources. Moreover, PLCs usually adopt dedicated network structures, such as Siemens’ MPI bus network, making it challenging or costly to add an operator station.During the development of DCS, each manufacturer has formed its own system. However, most DCS systems, such as YOKOGAWA, Honeywell, ABB, etc., although the internal (process level) communication protocols differ, have commonly chosen Ethernet for the operational network platform, using standard or modified TCP/IP protocols. This provides convenient scalability. In such networks, controllers and computers exist as nodes, and as long as the network reaches a location, nodes can be added or removed freely. Additionally, based on open protocols like OPC and DDE, various systems can communicate easily to achieve resource sharing.
3● From the perspective of databases:
DCS generally provides a unified database. In other words, once a piece of data exists in the DCS database, it can be referenced in any situation, such as in configuration software, monitoring software, trend charts, reports, etc. In contrast, the database of PLC systems is usually not unified; configuration software, monitoring software, and even archiving software each have their own databases. This is why it is often said that Siemens’ S7-400 is only considered DCS when it reaches model 414 or higher, as the PCS7 system uses a unified database, which requires the controller to be at least S7414-3 or higher.
4● From the perspective of time scheduling:
PLC programs generally cannot run on a pre-set cycle. The PLC program executes from start to finish and then starts over. (Some new PLCs have improved this, but there are still limitations on the number of task cycles.) In contrast, DCS can set task cycles. For example, for fast tasks, the sampling time for pressure sensors can be set to 200ms, while for temperature sensors with significant lag time, a sampling cycle of 2s can be used. This allows DCS to reasonably schedule the resources of the controller.
5● From the perspective of network structure:
Generally speaking, DCS typically uses a two-layer network structure, with one layer being the process-level network. Most DCS systems use their own bus protocols, such as YOKOGAWA’s Modbus, Siemens and ABB’s Profibus, and ABB’s CANbus. These protocols are all based on standard serial transmission protocols RS232 or RS485. Field IO modules, especially for analog sampling data, are substantial, and there are many field interference factors, so a network standard with high data throughput and strong anti-interference capability should be adopted. The bus structure based on RS485 serial asynchronous communication meets the requirements for field communication. The IO sampling data is converted by the CPU into integer or real data and transmitted over the operational network (the second layer network). Therefore, the operational network can adopt a network standard with moderate data throughput, fast transmission speed, and convenient connections, and since the operational network is generally located in the control room, the anti-interference requirements are relatively low. Thus, using standard Ethernet is the best choice. The TCP/IP protocol is a standard Ethernet protocol, and we generally use a communication rate of 100Mbit/s.
PLC systems have relatively simple working tasks, so the amount of data that needs to be transmitted is generally not large. Therefore, common PLC systems have a one-layer network structure. The process-level network and operational network are either merged or the process-level network is simplified into internal links between modules. PLCs rarely or never use Ethernet.
6● From the perspective of the scale of application objects:
PLC is generally used in small automatic control locations, such as equipment control or a small amount of analog control and interlocking, while large applications are generally DCS. Of course, this concept is not accurate, but it is intuitive; conventionally, we refer to systems with more than 600 points as DCS and those below this scale as PLC.
Now let us discuss the similarities between them.
1From a functional perspective: PLC already possesses analog control capabilities, and some PLC systems have even quite powerful analog processing capabilities, such as Siemens S7-400, ABB’s ControlLogix, and Schneider’s Quantum systems. DCS also has considerable logical processing capabilities.
2From a system structure perspective:PLC and DCS have the same basic structure. PLC has now fully migrated to computer system control, and traditional programmers have long been eliminated. Small applications of PLC generally use touch screens, while large-scale applications of PLC fully utilize computer systems. Like DCS, controllers and IO stations use field buses (generally based on RS485 or RS232 asynchronous serial communication protocols), and if there are no expansion requirements, meaning only one computer is used, this bus communication will still be beneficial. However, if there is more than one computer in use, the system structure will resemble that of DCS, with the upper computer platform using Ethernet structure. This is one of the reasons why the concepts of large PLCs and DCS have become blurred.3Development directions of PLC and DCS:
Small PLCs will develop towards more specialized usage, such as more targeted functions and greater adaptability to application environments. The boundaries between large PLCs and DCS will gradually blur until they completely merge.
DCS will continue to develop towards FCS. The core of FCS, in addition to a more decentralized control system, is particularly important for instrumentation. The application of FCS abroad has developed to the level of instrumentation. The control system only needs to handle signal acquisition and provide human-machine interfaces and logical control, while the entire analog control is decentralized to field instruments, and there is no need for traditional cable connections between instruments and control systems, using field buses to connect the entire instrumentation system. (Currently, YOKOGAWA has used FCS in the CNOOC Shell Petrochemical project, with intelligent instruments such as EJX, achieving the world’s most advanced control level.)
Source: Jicheng Training