PLCs (Programmable Logic Controllers) grew up as replacements for multiple relays and are used primarily for controlling discrete manufacturing processes and standalone equipment. If integration with other equipment is required, the user or his system integrator typically has to do it, connecting human-machine interfaces (HMIs) and other control devices as needed.
PLCs are fast. Response times of one-tenth of a second make the PLC an ideal controller for near real-time actions such as a safety shutdown or firing control. PLCs are best applied to a dedicated process that doesn’t change often. If your process is complex and requires frequent adjustments or must aggregate and analyse a large amount of data, a DCS (Distributed Control System) is typically a better solution.
A DCS, sometimes called Decentralised Control System, is simply put a control method in which we have some independent CPUs. Of course, the very flexibility of a DCS system also makes it much more vulnerable to unwanted access by operators that can cause shutdowns.
DCSs were developed to replace PID controllers and is found most often in batch and continuous production processes, especially those that require advanced control measures. The vendor handles system integration, and HMIs are integral.
DCSs take much longer than PLCs to process data, so it’s not the right solution when response times are critical. In fact, safety systems require a separate controller. But a DCS can handle many thousands of I/O points and more easily accommodate new equipment, process enhancements and data integration. If you require advanced process control and have a large facility or a process that’s spread out over a wide geographic area with thousands of I/O points, a DCS makes more sense than a PLC, which can only handle a few thousand I/O points or less. It’s just not as scalable as a DCS.
Another problem with PLCs is redundancy. If you need power or fault-tolerant I/O, don’t try to force those requirements into a PLC-based control system. You’ll just end up raising the costs to equal or exceed those of a DCS.
The network architecture of the DCS is way easier to implement using its integrated software and hardware package. However, DCS software and hardware packages cost a lot more than PLC equivalents.
As users demanded more production information, PLCs with more processing power and networking became more common. PLC-based control systems began to function like a mini-DCS. At the same time, the DCS hybridised to incorporate PLCs and PCs to control certain functions and to provide reporting services. The DCS supervises the entire process, much like the conductor in an orchestra. Protocols, like OPC, have eased interactions between the two control systems.
The complex nature of many continuous production processes, such as oil and gas, water treatment and chemical processing, continue to require the advanced process control capabilities of the DCS. Others, such as pulp and paper, are trending toward PLC-based control. In Batch or hybrid processes automation like some steel-producing industries in which we have to use DCS and PLC together, there would be some overlaps in their functions.
Process control has become increasingly complex. It’s difficult to know everything about these systems, increasing the need for vendor and service partner support. Manufacturers also continue to reduce factory staff and a generation of experienced process control personnel has begun to retire. As a result, the quality of support has become a critical factor in vendor and service partner selection.