It may be presumptuous of me to try and write a course Distributed Control Systems (DCS), and I feel my short experience (+ - 3 years) using these systems do not give me that right. But I think if I have the right to show what they've learned, share my experiences, my mistakes and keep learning all great books like "Lessons In Industrial Instrumentation" by Tony R. Kuphaldt and "Process Control and Optimization" of Bela Liptak with you, so with my short experience with the help of these two magnificent books and my DARE we will draw out this project "to write a course of DCS Distributed Control Systems." So not tired but with chatter, let's begin.
Although we have not written one way on a little old concept called DDC (Direct Digital Control), the DCS were imposed on this and revolutionized the concept of control. The direct digital control (DDC) during that time suffered from a substantial problem: the potential danger that there is a flaw in a single digital computer that controlled or run multiple PID control loops, functions that should never stop. The digital control brought many benefits, but not worth doing if there was a risk that the operation completely stopped (or fail catastrophically) followed by a hardware failure or software on one computer.
Distributed controls are intended to address this concern by having multiple computers, each responsible for a group of PID loops, distributed and linked facilities to share information among themselves and with the operating consoles. Now there was the worry of having all links on a single computer. The distribution of computers or controllers also ordered the signal wiring, because now hundreds or thousands of instrument cables only have to get to the distributed nodes, and not all the way down to the centralized control room. Only had network cables is bound to the controls, representing a drastic reduction of cable needed. In addition, the distributed control introduced the concept of redundancy in industrial control systems, where the digital signal acquisition and processing units were equipped with a "spare" or "spare" to automatically take control of all critical functions In the event of a primary failure.
The following figure shows a typical architecture of a Distributed Control System (DCS):
Each rack contains a processor to implement all necessary control functions with individual card input and output (I / O) to convert analog signals to digital or vice-versa. The redundancy of processors, redundant network cables, and even redundant I / O cards is implemented to prevent a component failure. DCS processors are usually programmed to perform a routine self-revision system redundant components to ensure availability of spare equipment in case of failure.
If even there were a total length of the racks in a control, PID loops only of this single rack will be affected, no other system loop. On the other hand, if the network cables fail, only the flow of information between these two points would be damaged, the rest of communicating the information system continues normally. Therefore, one of the "laws" or key features of a DCS is its tolerance to serious flaws: no matter what hardware or software failure impact on the control of the process is minimized by design.
Some modern distributed control systems to this date (2011) are:
- ABB 800xA
- Emerson DeltaV and Ovation
- Invensys Foxboro, I / A Series and InFusion
- Honeywell Experion PKS
- Yokogawa: CENTUM VP and CENTUM CS
The following figure shows a rack or cabinet DCS I / A Series, Invensys Foxboro:
Here's another picture of Emerson DeltaV DCS with a processor and multiple I / Os:
A photograph of an Emerson Ovation DCS below embedded in a vertical enclosure:
Several modern DCS as the I / A Series, Invensys Foxboro third party use computers instead of their own brands as operator station. This leverages the existing technologies in computers and screens work without sacrificing the reliability of control (as the control hardware and software continue to be industrial type).
The PLC (Programmable Logic Controllers) are given more prominence in the PID control due to its high speed, functionality and relatively low cost. It is now possible with modern PLC hardware and network to build a "copy" of a distributed control system as individual PLCs as nodes, and build redundancy with these nodes and not affect the operation of critical controls. Furthermore, these systems can be purchased at a very low considering the initial cost of a DCS.
However, it currently lacks the PLC is the same level of integration of hardware and software needed to build distributed control systems functional, ie it really comes standard DCS today: ready to use and pre systems built. In other words, if a company chooses to build their own DCS using programmable logic controllers, they must be prepared to make and spend many hours of programming work to try to emulate the same level of functionality and power of a pre-configured and pre- DCS developed.
Any engineer or technician who has experienced the power of modern DCS (with self-diagnosis, management of intelligent, event auditing, advanced control, redundancy, data collection and analysis, alarm management, etc..) Will realize that these characteristics are not at all easy to implement for any engineer. Woe to him who believes that these features can be implemented or created by a staff of engineers at a lower cost and less time! (Guerra advised not kill people: D)
