On a napkin
Building the grid back up from zero 1. Self-start unit Hydro / battery+GFM on-site fuel Starts itself. Energizes line. 2. Cranking path Transmission line energized to neighbor Carries power to wake up next generator 3. Wake larger gen Gas / steam plant station service Now self-sustaining, adds capacity 4. Pick up load Load areas restored in chunks, balanced against available gen Tie to neighboring restored islands Typical timeline 0h blackout ~1h first units ~4–12h major restoration ~24–72h most load served days+ full normalization NERC EOP-005 mandates documented restoration plans · annual drill exercisesBlack start is the process of restoring electric service from a completely de-energized state without relying on external power. After a system collapse — a complete blackout of an interconnection or large region — the grid cannot simply be reconnected. It must be rebuilt piece by piece, starting with specially designated generators that can begin operation without grid power. These black-start units energize transmission paths to neighboring generators, gradually bringing the broader system back online. The process can take hours to days depending on the scale and the damage that caused the initial collapse. NERC reliability standards mandate that every balancing authority maintain sufficient black-start capability and documented restoration plans, with annual training exercises and equipment testing.
Resilience is more than reliability. Reliability metrics — SAIDI, SAIFI, CAIDI — measure the frequency and duration of typical outages. Resilience refers to the system's ability to recover from extreme events: hurricanes, ice storms, cyberattacks, equipment failures cascading across the network. A grid can be reliable in normal conditions and still fragile when stressed. The growing emphasis on resilience reflects awareness that extreme events are becoming more frequent and that traditional reliability investment doesn't necessarily produce robust recovery.
The defining characteristic of a black-start resource is the ability to begin operation from a completely de-energized state — without electricity from the surrounding grid. This is harder than it sounds. Most large thermal generators (coal, gas combined-cycle, nuclear) require significant electrical service for boiler feed pumps, coolant systems, fuel handling, and control electronics. They cannot start themselves; they need grid power to wake up. In normal operation this is fine, but in a black-start scenario it makes them dependent on other resources energizing them first.
Traditional black-start resources include hydroelectric plants (which can start using a small backup generator and the kinetic energy of flowing water), some gas turbines specifically configured with on-site fuel storage and self-start systems (often diesel-started), and dedicated diesel generators. Pumped storage hydro can black-start in generating mode. More recently, battery storage paired with grid-forming inverters is increasingly recognized as a black-start resource — and several utilities are now deploying batteries specifically for black-start service, including Southern California Edison, Dominion Energy, and various ISO-mandated reliability deployments. Black-start contracts pay generators for being available to perform this service.
A cranking path is the sequence of transmission lines, substations, and intermediate generators used to bring power from black-start resources to other generators that need station service to start. The path must be carefully planned — energizing too much transmission at once can cause voltage and frequency instability before generators can respond, while energizing too little leaves the next-stage generators unreachable. Each step requires manual switching, voltage control, and stability verification before proceeding.
The classic restoration sequence works outward in stages. A black-start unit (typically hydro or gas turbine) starts and produces a small amount of power. That power energizes a transmission line to a larger nearby generator, providing the station service it needs to start. Once that generator is running, it provides additional capacity and helps stabilize frequency. The combined units can then energize more transmission, picking up additional generators and starting to serve small islanded load. Multiple islands grow and eventually merge into a coherent restored system. Each merger requires synchronizing the phase angles, voltages, and frequencies of the joining systems — a delicate operation that can fail if not done carefully.
A cascading blackout occurs when an initial grid disturbance causes power flows to redistribute in ways that overload other equipment, which then trip offline, triggering further redistribution and additional trips in a chain reaction. The 2003 Northeast Blackout is the classic example: a relatively minor initial event in Ohio (a transmission line contacting a tree) cascaded across the Eastern Interconnection over the course of about an hour, ultimately blacking out 55 million customers across eight states and parts of Canada. Underlying issues included inadequate tree trimming, malfunctioning state estimator software, and failure to detect deteriorating conditions in time.
Modern cascading risks include similar transmission-related events, but also new categories: cyberattacks targeting industrial control systems, GPS spoofing disrupting time synchronization across the grid, inverter-based resource trips during disturbances, and unprecedented weather events causing simultaneous equipment failures across wide areas. NERC reliability standards (particularly the TPL and IRO series) impose strict requirements for maintaining N-1 contingency capability — meaning the system can absorb any single equipment failure without cascading — but N-2 and beyond contingencies remain partial coverage with various exceptions.
A microgrid is a localized energy system that can operate either connected to or isolated from the main grid. When the main grid fails, a properly designed microgrid continues powering its served loads in "island mode" using on-site generation, storage, and load management. Microgrids are increasingly deployed at hospitals, military bases, data centers, university campuses, and critical commercial facilities. The pairing of solar, battery storage, and grid-forming inverters has reduced microgrid costs and complexity over the past decade, making them practical for a broader range of facilities than the historical diesel-only model.
For commercial buyers concerned about resilience, microgrids represent an alternative to depending entirely on grid reliability. A typical commercial microgrid might include rooftop or carport solar, behind-the-meter battery storage, backup natural gas or diesel generation, an EMS with islanding capability, and automated switching at the utility service connection. Total project costs vary widely ($1.5M-$15M+ for typical commercial deployments) depending on scale, redundancy requirements, and existing infrastructure. The economic justification has historically been challenging for facilities outside critical-mission categories, but as utility rates rise and grid resilience concerns intensify, microgrid economics are improving for a broader set of facilities.
Three implications for commercial procurement and facility planning. First, for facilities where extended outages would be operationally catastrophic (data centers, healthcare, manufacturing with continuous processes), independent resilience capability is increasingly worth modeling — not just as backup generation, but as integrated microgrid systems with extended duration capability. Second, geographic location matters: ERCOT and weather-exposed regions face higher cascading risk than well-connected regions in the Eastern Interconnection. Site selection for new facilities increasingly factors grid resilience alongside cost. Third, the resilience investment landscape is evolving: federal programs including DOE Grid Resilience and Innovation Partnerships (GRIP) and various state programs provide funding for resilience deployments. Tracking these programs has become a meaningful procurement function for critical infrastructure operators.
Common questions
Related reading on The Outlet