Integrating SCADA With Controllers Somebody Else Installed

A pump station does not care that its controller was installed twenty years ago. It still has to start on command, protect the plant and keep the network supplied.

The problem is never that the old controller stopped working. It is that somebody now needs to see the station from an operations room, or report its energy, or get an alarm at three in the morning, and the controller was specified before anyone asked for that.

What integration actually means

Clients asking about SCADA and PLC integration are usually asking four questions at once. Can the existing controller stay? Can it talk to a modern supervisory layer? Can alarms, trends and remote commands be added without disturbing the process? Can this be phased instead of done as one replacement?

The answer to all four is usually yes. What changes between sites is the cost of getting there, and that is decided almost entirely by what is already installed.

Audit before you plan anything

The common mistake is treating an old controller as a problem before establishing what it still does well.

A useful audit records the make, model and firmware, the I/O count, the available ports and protocols, the condition of the panel and its power supply, and the network topology. Then two questions that decide the whole project: does the source code still exist, and does any software that can open it still run on a machine somebody has?

This stage usually finds something unexpected. Plenty of stations run perfectly on logic nobody alive has read. Others have adequate control capacity but no realistic path to Ethernet. And on a surprising number of sites the field wiring and instruments are in worse condition than the controller, which changes the scope from integration to refurbishment.

The three paths, and how to choose

Leave the controller, add a reporting path. A gateway or protocol converter lifts the existing data into the supervisory layer. Cheapest, least disruptive, and correct when the controller is supported and the logic is understood. The limit is that you can only report what the controller already knows.

Keep the controller, rebuild everything around it. New panel, new instruments, new communications, new HMI. This is what most of our irrigation and pumping work looks like, because the pumps and motors are usually sound and the weakness sits in the control and the visibility.

Replace the controller. Right when the hardware is unprocurable or the logic is unrecoverable. The test is not age. It is whether you could buy the spare today from somebody you would trust, and whether anybody can still change the sequence.

The judgement between the second and third is the same one that applies to any obsolete drive or controller, and getting it wrong in either direction is expensive.

Integration into a system that already exists

The most common version of this work is not building a SCADA. It is joining one.

At Shakhbout City the scope was a VFD and instrumentation panel integrated into the client's existing irrigation SCADA. Nothing was replaced at the supervisory layer. The new station simply had to appear in it, behave like the others, and raise alarms the operators already knew how to read.

For Abu Dhabi Investment Council the work went the other way: the existing PcVue platform was upgraded from version 9 to version 16, with field devices replaced, reprogrammed and recommissioned around it. A platform migration rather than a platform replacement.

And on TRANSCO project N-12611 the entire SCADA scope was a single interface, so that pump 4210 remained visible from the control system after its drive and motor were replaced. One signal path, in scope from the start, because a replaced drive operators cannot see has moved the problem rather than solved it.

Where new build differs

When the supervisory layer is new, the constraint moves from protocols to distance. On the Jabal Mibrih potable water pipeline the SCADA design, configuration, testing and commissioning sat in one contract with 17 kilometres of fibre optic cable, because the hard part of that system was never the software. It was getting the signal home. On the Khalifa City "A" irrigation mains, five stations each received a PLC and an RTU alongside the motor control centre, so the PLC runs the pumps and the RTU owns the conversation with the outside world.

Four things that decide whether it works

A signal list built from decisions, not from the I/O schedule. The question is which decisions are currently slow or blind. Every point beyond that becomes an alarm somebody learns to ignore.

Timestamps taken at the source. If the server stamps on arrival, the event record only tells you the order the packets landed in.

Monitoring proven before control is trusted. Run the new visibility in parallel with the existing controls until it agrees with them. Only then move the commands.

A documented handover. Configuration, addressing, alarm philosophy and as built drawings, in a form the next engineer can use. Skip it and the obsolescence problem regenerates.

Working with us

Almafor designs, configures, tests and commissions SCADA for grid stations, pumping networks, pipelines and buildings across Abu Dhabi and the wider UAE, including integration with controllers and platforms somebody else installed. Panels are built and factory tested at our workshop in Mussafah.

If you have plant that runs well and reports badly, the useful first step is an audit of what is installed and what it can already tell you. Call +971 55 954 5801 or email info@almafor.com.