Redundant PLC vs Non-redundant PLC

Dear,

usually on what basis a redundant PLC is selected vs a non-redundant PLC, I am working in designing a automation system for water pumping station (WPS) and sewage pumping stations (SPS).

the water pumping station is huge around 250KW each pump, consisting of 3 pumps,

and sewage pumping stations are scattered all around the site for taking sewage and send it to a sewage treatment plant

on what basis, shall I select redundant PLC vs non-redundant PLC,

the selected PLC will be fault-tolerant.

both of the system are considered to be critical, I mean if WPS PLC fails, no water will reach the city and if SPS PLC fails, we will get sewage overflow.

I appreciate any advice on this issue.

Redundant at first glance, but is the rest of the system redundant? The PLC is one of the most reliable parts of any system but often seems to be the only bit where someone has considered redundancy. If the rest of the system has no redundancy, and isn't likely to ever have it, then it seems a bit pointless forking out for a nice redundant PLC system. If it does then you'll have to follow suit with your control system.

I agree completely with what amberman said. A Redundant PLC has an important tool in creating a redundant system, but it does not turn a standard system into a redundant one. The IO and networks are much more likely to fail, as well as the actual devices they control.

I was once asked to put together a system where the customer wanted a triple redundant PLC to control just one pump, even though the pump is far more likely to fail than the PLC. It just didn't make sense.

To decide if you need a redundant system or not, you need to look at your process and decide what faults could happen, how likely they are, and how hard they would be to recover from. Pumps are probably more likely to fail than any of the solid state components, so my guess is that you already have teams of pumps to share the load, but with spare capacity for redundancy.

In my experience redundant PLCs are a waste of money. The CPU is the least likely device to fail. If you have full redundant PLCs with duplicate I/O you add tremendously to complication, and add very little to reliability. The most common point of failure in the control system is in the field devices.

You should have redundant pumps - US regulations, for example, require the system to be able to deliver design flow with the largest unit out of service. To accommodate controller failure you need to have manual overrides. For your application that probably means high and low level float switches in the reservoir to turn the pumps on and off - at constant speed if necessary.

If you are looking at redundant PLC systems I would suggest dual processor and dual power supply type with a changeover unit in the middle and single I/O racks. Omron have the CS1 if that helps. The usual thing to go in a PLC rack is a power supply. These have 2 CPUs and power supplies and fault signals built in for a failure of either. That is as far as I would go - dual I/O means dual everything or relays to feed both racks. And then there is the issues of, for example, Modbus where there is only one master on the network. You finish up adding a whole more single points of failure - hard to get that through to people sometimes.

Tom Jenkins said:

In my experience redundant PLCs are a waste of money. The CPU is the least likely device to fail. If you have full redundant PLCs with duplicate I/O you add tremendously to complication, and add very little to reliability. The most common point of failure in the control system is in the field devices.

You should have redundant pumps - US regulations, for example, require the system to be able to deliver design flow with the largest unit out of service. To accommodate controller failure you need to have manual overrides. For your application that probably means high and low level float switches in the reservoir to turn the pumps on and off - at constant speed if necessary.

Click to expand...

+1.
EVERY time I have had an inquiry which included a phrase like "redundant PLC processors shall be quoted as an optional item", the resultant optional price has been rejected.

BobB said:

If you are looking at redundant PLC systems I would suggest dual processor and dual power supply type with a changeover unit in the middle and single I/O racks. Omron have the CS1 if that helps. The usual thing to go in a PLC rack is a power supply. These have 2 CPUs and power supplies and fault signals built in for a failure of either. That is as far as I would go - dual I/O means dual everything or relays to feed both racks. And then there is the issues of, for example, Modbus where there is only one master on the network. You finish up adding a whole more single points of failure - hard to get that through to people sometimes.

Click to expand...

dear,

thank you for you reply,
I understand from the above that we have PS1 CPU 1 - changeover - PS2 CPU2
the changeover would be automatic and will have one source of power supply for example from UPS. Is this correct?

Generally CPU1, changeover, CPU2, PS1, PS2. You can use power to the 2 power supplies from 2 different locations if you wish. I have used the Omron CS1 dual system on job before with common I/O racks.