When Should a PLC or Drive Be Replaced? A 6.6 kV Pump Station Made the Case

Obsolete equipment rarely announces itself. It keeps running. The station meets its duty, the readings look normal, and the case for spending money on a replacement sounds like an argument against evidence.

The evidence arrives later, usually as a question nobody can answer quickly: where do we get the part.

The project that framed the question

Between 2015 and 2016 Almafor worked as main contractor on TRANSCO project N-12611 at the Unit IV pump station. The scope was the replacement of an obsolete cascade drive and its motor with a new 6.6 kV variable speed drive and a squirrel cage induction motor, together with modification of the 6.6 kV switchgear, a 6.6 to 0.4 kV step down transformer for the auxiliary supply, and a SCADA interface so that pump 4210 remained visible from the control system after the change.

Read that scope again and notice what it is not. It is not a control upgrade. The obsolete element was the drive, and once the drive changed the motor had to change, and once the motor changed the switchgear feeding it had to be modified, and the auxiliary supply had to be rebuilt to suit, and the SCADA had to be told about all of it.

This is the pattern that catches asset owners out. Obsolescence is rarely contained to the component that is obsolete.

Four honest tests for whether it is time

Almost every operator we meet has a legacy PLC or drive somewhere and a reasonable suspicion that it is living on borrowed time. These are the questions that settle it.

Can you buy the spare, today, from someone you would trust? Not whether it exists on a broker site. A supported supply chain, with a warranty and a lead time you could survive. When the honest answer is a specialist rebuilder in another country, the asset is already unsupported, whatever its condition.

Does anyone still hold the logic? Source code, the software version that opens it, and a person who understands the sequence. Plenty of stations run perfectly on logic nobody alive has read. That is a functioning system with no ability to change, which is a different asset from the one on the drawings.

What does an unplanned failure actually cost? For a pumping station serving a network the number is rarely the repair. It is the service interruption and the emergency mobilisation. Compare that against a planned replacement executed inside a scheduled outage, which is the comparison the budget conversation usually skips.

Is the equipment blocking something you already want? Remote visibility, energy reporting, condition based maintenance, integration with a wider SCADA. Legacy hardware often has no path to any of it. The cost of the upgrade is then partly the cost of a capability, not only a like for like replacement.

If three of those four answers point the same way, the decision has already been made and only the timing is open.

What a live replacement actually demands

The engineering on N-12611 was not the difficult part. Sequencing was.

Survey what is really installed. Drawings on infrastructure of that age describe intent from years ago. The pre modification survey exists to find the differences between the drawing and the copper, and it always finds some. We do the same thing at grid scale: the enterprise GIS upgrade for ADWEA was a communications as built survey against site condition across 45 power stations and 31 water stations, and its entire value was in the differences it found.

Plan the outage backwards from the constraint. The station has a duty it must continue to meet. Every task is arranged around that, not around the convenience of the works programme.

Prove the new equipment before the old one leaves. Factory testing, pre commissioning checks, and a defined fallback for the point where you commit. Reversibility is worth real money right up until the moment it becomes impossible. Our panels are built and factory tested at our workshop in Mussafah, with the client's engineers present, precisely so that the surprises happen there rather than on site.

Land the SCADA interface as part of the work, not after it. A replaced drive that operators cannot see from the control room has moved the problem rather than solved it. On this project the interface for pump 4210 was in scope from the start.

Leave the site documented. As built drawings, the configuration, the settings, and the operating philosophy in a form the next engineer can use. The obsolescence problem regenerates whenever this step is skipped.

Retrofit or replace

Not every legacy system deserves demolition. We regularly recommend the narrower path when the mechanical plant is sound and the weakness is confined to control.

Where the pumps and motors are healthy and the gap is visibility or sequencing, a retrofit that adds a modern PLC, rebuilds the control panel, upgrades instrumentation and connects the station to SCADA gives most of the benefit at a fraction of the disruption. Our irrigation and pumping work is largely this pattern.

Where the primary equipment is the obsolete element, as at Unit IV, the retrofit path simply relocates the risk. A modern controller wired to a drive that cannot be sourced is a well instrumented liability.

The distinction is worth making early, because the two projects have different budgets, different outage requirements, and different answers to the question of what happens in year ten.

The same judgement at grid scale

Drives are the small version of a problem we spend most of our time on. Substation control and monitoring systems age the same way and fail the same tests.

• On TRANSCO project N-15223 we replaced the SCMS across six grid stations running from 400 kV down to 11 kV, including servers, workstations, gateways, bay control units and the integration back to the load dispatch centre, station by station, without interrupting transmission service.
• On N-15228 at Musaffah the replacement carried ninety Schneider C264 bay control units and redundant gateways to both the load dispatch centre and the distribution management system.
• At Al Khazna the upper and lower ring SCMS was upgraded at 33 and 11 kV with hardwired interlocking retained through the change.

In every one of those the trigger was the same as at Unit IV. Not failure. Unsupportability.

Working with us

Almafor has delivered drive replacement, control system upgrades and SCADA integration on live infrastructure across Abu Dhabi as both main contractor and specialist subcontractor, for clients including TRANSCO, TAQA Distribution, Al Ain Distribution Company, AD Ports Group and the Department of Municipalities and Transport.

If you have equipment you suspect is unsupported, the useful first step is an assessment: what is installed, what is still procurable, what the logic does, and what a planned replacement would actually require. Call +971 55 954 5801 or email info@almafor.com.