How to Automate an Electrical Substation That Cannot Be Switched Off

A substation does not become automated when a few devices start talking to a SCADA screen. It becomes automated when protection, control, monitoring, communication and operator response are engineered as one system.

That distinction matters because almost none of this work happens on a dead station. The plant is in service, it has a duty it must keep meeting, and the automation has to be built around that rather than the other way round.

What the system is made of

Three layers, and they fail differently, which is why they are kept separate.

The bay. Protection relays, meters, instrument transformers, and a bay control unit per bay that reads every switch position, measures the electrical quantities, enforces the interlocking rules and timestamps what it sees. One controller per bay is not redundancy, it is containment: a fault in one bay's controller must not take the visibility of the whole station with it. On the Musaffah 220, 33 and 11 kV station that came to ninety Schneider C264 units.

The station. Servers, an operator workstation, a separate engineering workstation, a station clock and the UPS that keeps all of it running when the auxiliary supply does not. This is the layer that turns dozens of bays into one picture.

The boundary. Gateways that publish the station outward. On the Musaffah job that meant redundant pairs, one publishing the 220 kV side to the load dispatch centre and another publishing the 33 and 11 kV side to the distribution management system. Two different audiences, two different data sets, and a good reason not to serve both from one device.

The order that works

Projects that go wrong usually went wrong in the sequence, not the engineering.

1. Decide what the station has to do differently. Faster fault response. Remote switching. Visibility from a central operations room. Event records good enough to settle a disturbance investigation. This decides the architecture, the alarm philosophy and the protocol, and it has to be settled before anybody lists hardware. A system specified from a hardware list is usually overbuilt in one place and thin in another.

2. Survey what is actually installed. Not what the drawings say. Legacy relays, hardwired control, panel logic, auxiliary supplies, cable routes, panel space and earthing all constrain the design, and on a station of any age the record and the copper have drifted apart. We do this at grid scale too: the ADWEA enterprise GIS upgrade was a communications as built survey against site condition across 45 power stations and 31 water stations, and its whole value was the differences it found.

3. Plan the outage backwards from the duty. The station has to keep meeting its obligation. Every task is arranged around that constraint, bay by bay, not around the convenience of the works programme.

4. Keep the old safety alive while the new system is proven. At Al Khazna the upper and lower ring system was upgraded at 33 and 11 kV with the hardwired interlocking deliberately retained through the change. The station was never dependent on the new system being healthy while it was being installed. That is the difference between a staged cutover and a leap.

5. Land the control centre integration in the same contract. A station that is automated locally but invisible to the dispatch centre has moved the problem rather than solved it.

6. Train and document at handover. Operators and engineers separately, because they need different things, and as built drawings plus configuration backups so the next engineer inherits a system rather than a mystery.

What it looks like across six stations

On TRANSCO project N-15223 we replaced the substation control and monitoring system across six grid stations as main contractor, running from 400 kV down to 11 kV: 400/220 kV, 400 kV, 220/33 kV and three 132/11 kV stations.

Each took servers, human machine interface and engineering workstations, IEC 101 gateways, bay control units, station UPS and fibre optic cabling, plus integration back to the load dispatch centre, the emergency control centre and the distribution management system. Station by station, over several years, without interrupting transmission service.

The same architecture went into three primary substations at once at Khalifa Port, for a client whose cranes do not stop, with the site survey, detailed design, shop drawings, civil works and the IT and fibre network all inside one contract.

Retrofit or replace

Not every station needs a full replacement. Where the primary plant is sound and the weakness is visibility, a retrofit that adds bay level intelligence, a station layer and a gateway gives most of the benefit for a fraction of the outage.

The trigger for a full replacement is rarely failure. It is unsupportability: the platform is obsolete, the spares are gone, nobody holds the configuration, or the station cannot speak to the systems the utility now runs. Those are the same four tests that apply to any legacy controller or drive.

Working with us

Almafor delivers substation control and monitoring systems as main contractor on live transmission and distribution stations from 400 kV down to 11 kV, for TRANSCO, TAQA Distribution, Al Ain Distribution Company and AD Ports Group, including the panels, the fibre and the civil works around them.

If you have a station on an unsupported platform, the useful first conversation is about what is installed, what the interlocking currently depends on, and what outages you can realistically get. Call +971 55 954 5801 or email info@almafor.com.