SCADA

SCADA supervises geographically spread assets — pipelines, water networks, well pads, power grids. Learn RTUs and PLCs, telemetry communication (radio, cellular, satellite, fibre), protocols (Modbus, DNP3, IEC 60870-5-104, OPC UA), polling and bandwidth, historians and alarms, and cybersecurity.

A gas pipeline hundreds of kilometres long, or a city's water network with hundreds of pump stations, can't be wired to one control room. SCADA uses remote terminal units and communications links to gather data and send commands over long distances.

Where a DCS controls tightly integrated plants, SCADA focuses on monitoring and supervisory control of dispersed sites.

An analogy

SCADA is like air-traffic control for industrial assets: many remote sites report position and status, and the central centre coordinates and issues instructions.

Components

Field instruments → RTUs/PLCs at remote sites (local logic, data logging) → communications network → SCADA master servers → operator HMI, historian, alarm system. Remote sites often run on solar power.

Communications and protocols

Licensed radio, cellular (4G/5G), satellite, microwave and fibre. Protocols: Modbus (simple), DNP3 and IEC 60870-5-104 (utility-grade with timestamps and report-by-exception), OPC UA and MQTT for IIoT. Report-by-exception and timestamped events reduce bandwidth.

Applications

Pipeline leak detection (mass balance, pressure-wave), compressor and pump station control, water-network pressure management, well-pad monitoring, electricity distribution (with EMS/DMS).

Security and modernisation

SCADA networks are prime cyber targets (e.g. Ukraine grid, Colonial Pipeline IT impact). IEC 62443, network segmentation, VPNs, secure protocols (DNP3-SA, OPC UA security), monitoring and incident response are essential. Cloud historians, edge computing and analytics add predictive maintenance.

Formulas

• Polling bandwidth: BW = N_sites × bytes per poll × 8 / t_poll
• Poll cycle time: t_cycle = N_sites × (t_request + t_response + latency)
• Report-by-exception savings: traffic ∝ fraction of values changing beyond deadband

Worked example: Polling bandwidth

Given: 200 RTUs; 500 bytes per poll, every 10 s

Find: Average bandwidth.

• 200 × 500 × 8/10 = 80 000 bit/s = 80 kbit/s

Answer: ≈ 80 kbit/s (plus protocol overhead)

Worked example: Report by exception

Given: Only 10 % of values change beyond deadband per cycle

Find: Traffic reduction.

• Traffic ≈ 10 % of full polling + periodic integrity polls

Answer: ≈ 80–90 % less traffic

Common mistakes

• Wrong: Assuming SCADA networks are isolated and safe. — Right: Apply IEC 62443 defence-in-depth; many are internet-reachable by mistake.
• Wrong: Polling everything fast. — Right: Use report-by-exception and appropriate intervals per data type.

Applications

• Oil & gas pipelines: Compressor stations, leak detection, custody metering.
• Water & wastewater: Pump stations, reservoirs, treatment plant supervision.
• Power grids: Substation automation, feeder control and outage management.

Key takeaways

• RTUs/PLCs + telemetry + master + HMI/historian.
• Protocols: Modbus, DNP3, IEC 104, OPC UA.
• BW = N × bytes × 8/t; use report-by-exception.
• Cybersecurity is critical.

Going further

• Real-time transient models: Hydraulic models detect and locate pipeline leaks.
• MQTT & IIoT: Publish-subscribe telemetry for edge devices.
• Grid SCADA/EMS: State estimation and automatic generation control.