Inferensys

Glossary

Spinning Reserve

The portion of contingency reserve provided by generating units synchronized to the grid that can begin delivering their full committed capacity within 10 minutes of a dispatch instruction.
Moody home-office setup in a converted highrise loft, analyst working late with multiple screens showing knowledge graph visualizations, city lights through large windows behind.
CONTINGENCY RESERVE

What is Spinning Reserve?

Spinning reserve is the portion of operating reserve provided by generating units that are synchronized to the grid, unloaded, and capable of delivering their full committed capacity within 10 minutes of a dispatch instruction.

Spinning reserve is a critical ancillary service comprising generation capacity that is already online and synchronized to the power system but not fully loaded. These units operate with a governor droop response and headroom, allowing them to rapidly increase output via their turbine controls within the mandatory 10-minute window following a sudden loss of generation or transmission element.

This synchronized capacity is distinct from non-spinning reserve, which consists of offline resources requiring startup time. Spinning reserve provides immediate primary frequency response support, arresting frequency decay before slower supplemental reserve can be deployed, and is essential for meeting NERC Disturbance Control Standard (DCS) recovery obligations.

CONTINGENCY RESERVE CLASSIFICATION

Core Characteristics of Spinning Reserve

Spinning reserve is the most immediate and reliable form of contingency reserve, provided exclusively by synchronized generation that can ramp to full committed capacity within 10 minutes. Its defining characteristics center on speed, synchronization status, and governor response capability.

01

Synchronized & Online Status

The defining prerequisite of spinning reserve is that the generating unit must be electrically synchronized to the grid and already online. The generator's rotor is locked to the system frequency (60 Hz in North America, 50 Hz in Europe), and the unit is injecting or capable of injecting active power. This contrasts with non-spinning reserve, which can be provided by offline quick-start units like combustion turbines or diesels. Synchronization ensures that power delivery begins the instant the governor responds, eliminating the startup and breaker-closing delays inherent to offline resources.

< 1 sec
Initial Response Time
02

10-Minute Full Deployment Window

NERC reliability standards require spinning reserve resources to deliver their full committed capacity within 10 minutes of a dispatch instruction. This window is not arbitrary—it aligns with the Disturbance Control Standard (DCS) recovery period. The unit's ramp rate must be sufficient to bridge the gap between its current economic dispatch point and its maximum emergency capacity within this timeframe. For example, a 500 MW steam turbine with a 5% droop setting and a 15 MW/min ramp rate can provide approximately 150 MW of spinning reserve if it is currently dispatched at 350 MW.

10 min
Max Deployment Time
03

Governor Droop & Primary Frequency Response

Spinning reserve is fundamentally linked to the governor droop characteristic. When a sudden generation loss causes frequency to decay, the governors on all synchronized units autonomously increase mechanical power input in proportion to the frequency deviation. The droop setting—typically 5% for thermal units—defines the percentage speed change required for a 100% change in valve position. This inherent negative feedback provides the primary frequency response that arrests frequency decline within seconds, buying time for Automatic Generation Control (AGC) to deploy secondary regulation and restore nominal frequency.

5%
Typical Droop Setting
04

Headroom & Economic Dispatch Interaction

To provide spinning reserve, a unit must operate with headroom—the difference between its current output and its maximum emergency rating. This creates a trade-off with economic dispatch efficiency. A unit held back at 80% of capacity to provide 20% spinning reserve is not operating at its most efficient heat rate. System operators must balance the reliability requirement for contingency reserve against the economic penalty of part-load operation. Advanced unit commitment algorithms co-optimize energy and reserve markets to minimize total system cost while ensuring sufficient spinning reserve is geographically distributed across the interconnection.

05

Contingency Event Triggering

Spinning reserve is specifically designated for contingency events, not routine load following. The most severe single contingency—typically the sudden loss of the largest online generator or transmission line—defines the Most Severe Single Contingency (MSSC). NERC standards require each balancing authority to carry enough contingency reserve to cover its share of the interconnection's MSSC. When a 1,200 MW nuclear unit trips offline, the immediate frequency drop is arrested by the combined governor response of all synchronized spinning reserve across the entire interconnection within 2-3 seconds.

2-3 sec
Frequency Nadir Arrest
06

Qualifying Resource Types

Not all synchronized generation qualifies equally. Steam turbines (coal, nuclear, combined-cycle) provide inherent spinning reserve through their large rotating inertia and governor response. Hydroelectric units are exceptional spinning reserve providers due to their rapid ramp rates—often exceeding 50 MW/min. Combustion turbines synchronized at minimum load can also qualify. However, variable renewable resources like wind and solar typically do not provide spinning reserve unless they are deliberately curtailed and equipped with fast-acting power electronics to emulate governor response, a capability known as synthetic inertia.

CONTINGENCY RESERVE CLASSIFICATION

Spinning Reserve vs. Non-Spinning Reserve

A technical comparison of the two primary classifications of contingency reserve based on synchronization status, response time, and operational characteristics.

FeatureSpinning ReserveNon-Spinning Reserve

Synchronization Status

Synchronized to grid

Offline or load-interruptible

Response Time Requirement

Full output within 10 minutes

Full output within 10 minutes

Typical Resource Types

Partially loaded thermal units, hydro units, pumped storage

Fast-start combustion turbines, diesel generators, interruptible load

Immediate Governor Response

Provides Inertial Support

Start-Up Time Overhead

0 seconds (already online)

5-10 minutes

Minimum Deployment Duration

30-60 minutes typical

1-2 hours typical

NERC Classification

Contingency Reserve - Spinning

Contingency Reserve - Non-Spinning

SPINNING RESERVE ESSENTIALS

Frequently Asked Questions

Clear, technical answers to the most common operational and market questions regarding synchronized contingency reserves in power systems.

Spinning reserve is the portion of contingency reserve provided by generating units that are synchronized to the grid and can begin delivering their full committed capacity within 10 minutes of a dispatch instruction. It works by maintaining a margin of unused, ready capacity on generators that are already online and connected. These units operate at a level below their maximum output, creating headroom. When a sudden loss of generation or transmission occurs, the Automatic Generation Control (AGC) system or a manual dispatch signal instructs these units to rapidly increase their power output by opening their valves or gates further. Because the turbine-generator mass is already rotating in synchronism with the grid frequency, the response is immediate and physically coupled, providing critical inertial support during the initial moments of a disturbance before the full 10-minute capacity ramp is completed.

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.