Inferensys

Glossary

Black Start Capability

The ability of a generating unit to start up from a de-energized state without relying on an external power supply, a critical resource for energizing the grid during a system restoration following a total blackout.
Developer building agentic RAG system, retrieval pipeline diagram on laptop, technical workspace with notes.
SYSTEM RESTORATION

What is Black Start Capability?

The ability of a generating unit to transition from a completely de-energized, shutdown state to full operational power without relying on any external electricity supply from the grid.

Black Start Capability is the critical design feature of a generating unit that allows it to start up from a total station blackout, using an on-site auxiliary power source—typically a diesel generator or a small micro-turbine—to energize its own auxiliary systems and prime mover. This self-contained restart sequence is the foundational first step in a top-down system restoration plan following a catastrophic grid collapse.

Units with this capability, often hydroelectric plants or specific gas turbines, are designated as cranking sources by the transmission operator. They provide the initial energization of transmission corridors to supply station service power to larger, non-black-start capable thermal units, allowing them to sequentially restart and re-synchronize, thereby methodically rebuilding the interconnected grid from a de-energized state.

SYSTEM RESTORATION ANATOMY

Core Characteristics of a Black Start Resource

A black start resource must possess specific engineering attributes and auxiliary systems to transition from a de-energized state to full generation without external grid power, forming the foundational seed for power system restoration.

01

On-Site Prime Mover

The resource must possess a dedicated auxiliary power source, typically a diesel generator or battery energy storage system (BESS) , capable of energizing critical station service loads. This prime mover provides the initial power to crank lube oil pumps, fuel systems, and excitation controls before the main unit can synchronize. For hydro units, this may be a smaller house turbine or a diesel-driven governor oil pump.

100-500 kW
Typical Auxiliary Gen Size
02

Island Mode Frequency Control

During the initial energization of a dead bus, the unit's governor and excitation system must operate in isochronous mode to maintain a stable frequency and voltage without an external reference. The control system must suppress oscillations when energizing the first transmission line, managing the Ferranti effect and line charging capacitance without tripping on overvoltage or overexcitation limits.

60.00 Hz
Target Frequency Stability
03

Voltage Control & Reactive Absorption

The unit must be capable of manual voltage control to gradually raise voltage from zero during a soft-start energization sequence. Critically, it must absorb the line charging reactive power of the first transmission path. This requires an automatic voltage regulator (AVR) with a wide under-excitation limiter (UEL) range to prevent self-excitation and terminal overvoltage when connected to a lightly loaded, high-capacitance circuit.

0.0 to 1.05 pu
Voltage Ramp Range
04

Load Rejection & Pickup Capability

The resource must withstand sudden load rejection without tripping if the energized transmission path faults or the downstream breaker opens unexpectedly. Conversely, it must handle block load pickup, where large blocks of cold load are added in discrete steps. This requires robust governor response and sufficient inertia constant (H) to ride through transient frequency dips without triggering under-frequency load shedding (UFLS) relays.

> 3.0 s
Minimum Inertia Constant (H)
05

Cranking Path Coordination

The resource must be strategically located to energize a defined cranking path—a transmission corridor connecting it to a large, non-black-start unit (the target). The black start unit must provide sufficient short-circuit current to operate protective relays along the path and supply the massive inrush current required to start the auxiliary motors of the target thermal unit, such as boiler feed pumps and forced draft fans.

10-20 MVA
Min. Short-Circuit Contribution
06

Communication & SCADA Independence

Standard SCADA and ICCP telemetry may be unavailable during a total blackout. The resource must have a resilient, off-grid communication system, often satellite-based or point-to-point microwave, to coordinate with the reliability coordinator. The plant's distributed control system (DCS) must allow for local/manual synchronization without relying on the central energy management system (EMS) for breaker close commands.

Satellite/VHF
Backup Comms Medium
BLACK START CAPABILITY

Frequently Asked Questions

Essential questions about the specialized procedures and equipment required to energize a de-energized grid from scratch following a total or partial blackout.

Black start capability is the ability of a generating unit to start up from a completely de-energized state without relying on an external power supply from the grid. During a total blackout, the transmission network has zero voltage, meaning standard generators—which require station service power for cooling pumps, fuel systems, and excitation—cannot restart. A black start unit overcomes this by using an on-site auxiliary power source, typically a diesel generator or a battery energy storage system, to energize its own auxiliary loads and begin generating. Once the unit is online and stable, it establishes an energized cranking path to a nearby non-black-start plant, providing the station service power that plant needs to restart. This sequential process, known as system restoration, gradually re-energizes transmission corridors, synchronizes additional generators, and carefully picks up load blocks to rebuild the interconnection from isolated islands into a fully operational grid.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.