Better to use transistor outputs to drive diode protected external relays. Much easier to plug in a new relay in the middle of the night than try and find a PLC output card, or solder a new relay into the card and burn off a track because you cannot see. The transistor outputs can be driven equally safely into bi-directional diode protected multi chip super bright LEDs for indication.
Also bear in mind these systems are invariably operating on 24VDC. Relays have a tendancy to finish up with badly burned contacts on DC. External relays generally are larger physically and, therefore, have bigger air gaps to quench the DC arc.
I NEVER USE INTERNAL PLC RELAYS TO DRIVE DC!!!!! Rarely even have a relay failure these days using this method. Use Omron G2R type relays. Extremely reliable. They are also available re-badged Allen Bradley if you have trouble getting Omron.
Designed and implemented a base load 11kV generation power station in '96/'97. The system is working constantly, as you can imagine. Sets starting and stopping (load and capacity control, duty order being changed from the SCADA screen, etc etc.
THEY HAVE NOT HAD A RELAY FAILURE SINCE INSTALLATION!!!! THEY HAVE NOT HAD A PLC COMPONENT FAILURE SINCE INSTALLATION EXCEPT 1 ANALOGUE CARD WHERE A TRANSDUCER POWER SUPPLY SHORTED BELTING 240VAC INTO THE CARD!!!!
Show me a timer or electro mechanical device that can boast that sort of reliability. By the way, there are 9 networked PLCs on the site with remote I/O all over the place. In that time there have been battery failures, governor failures, cracked cylinder liners, fuel pump failures, fan motor failures, battery charger failures, mag pick up failures, cables snapped from engine vibration turning the copper conductors work hard, oil leaks, 2 fires from diesel leaks. The PLCs are still going.
If the SCADA running on one of Mr Gates operating systems crashes, the power station continues to operate automatically. If a set fails, or is failing, the PLC system starts another set, closes outgoing high voltage circuit breakers to supply the residents and industry, operates the telephone dialler to summon an operator to a fault in the power station. The only problem is that it is not possible to change engine duty order without the SCADA. However, each set can be turned into manual mode and operated through the PLC. Alarms, load information etc continues to be shown on the touch screen attached to each PLC.
If the PLC system were to collapse, all sets can be turned into "dead in the water" mode and operated from the instruments on the control panels. However, the critical single points of failure remain - governors, relays, timers, mag pick ups etc.
I would suggest that the doom and gloom attitudes towards PLCs expressed here are
TOTALLY UNFOUNDED!!!!
I might also add that I implemented a power station for the Royal Australian Air Force in '98/'99. The system is automatic including automatic dummy load control, automatic reactive load control, 11kV underground cables are highly capacitive and drive the power factor in to lead, small sets do not like leading power factor (AVRs mainly). Therefore a reactive load bank is required to control power factor back towards unity. When the Air Force hit site (it is a standby air base), they turn everything to manual. They adjust load sharing, var sharing etc manually. Sets go out on reverse power, under frequency, under voltage, over voltage, etc etc. They forget to open HV breakers before starting and dead bus closing sets and wonder why the stes "bounce" due to the enormous inrush of trying to supply 4 HV underground cable systems that are highly capacitive.
There are many more blackouts than when the system is allowed to run automatically under PLC control.
Give me a PLC any day!!!