What Is the Basis for the BCP “72-Hour Rule”? | Self-Reinforcing Failure Cycles and the 24-Hour Evacuation Decision in the 2026 Kumamoto Earthquake

First 72 Hours

From the initial shock through 72 hours, failures persisted and reinforced one another

As time passed, the most visible problem shifted from the initial shock and emergency response to electricity and communications, medical continuity, damage to public facilities, and finally water, fuel and the ability to sustain daily life. This does not mean that one problem was resolved before the next appeared. Power and communications failures remained while medical deterioration, water outages, fuel shortages, road congestion and prolonged displacement accumulated. Each failure delayed the recovery of the others.

Shared Failure Domain

The basic structure revealed during the first 72 hours

The following functions were affected across the same geographic region.

The critical point is not simply that information was released after damage occurred. Electricity and communications were themselves required to collect, transmit and integrate damage reports. When those systems failed, the actual extent of the disaster remained invisible for an extended period.

“Damage status unknown” does not necessarily mean that an inspection is merely taking time. It may mean that communications are unavailable, power has been lost, roads are blocked, staff cannot reach the site, or the receiving organization is unable to process reports. The route used to obtain the information may itself have disappeared.

“No response” is not an empty field. It is a hazardous condition that requires the highest confirmation priority.

Dual Upper-Layer SPOF

Electricity and communications created a self-reinforcing failure cycle

The first losses were not limited to individual buildings or pieces of equipment. Electricity, which enabled the region to operate, and communications, which enabled the region to be understood, were lost at the same time.

A mutually reinforcing cycle

Power outage
  ↓
Base stations, LANs and terminals stop
  ↓
Damage information cannot be transmitted
  ↓
Deployment of personnel, fuel and power vehicles is delayed
  ↓
Power and communications outages continue longer

Communications outage
  ↓
Outage locations, equipment status and required resources are unknown
  ↓
Restoration priorities cannot be determined
  ↓
Deployment of personnel, fuel and materials is delayed
  ↓
Restoration of electricity and communications is delayed further

Electricity failure stopped communications, and communications failure delayed electricity restoration. This was not merely a simultaneous outage. It was a self-reinforcing failure cycle.

Broadcast and Upstream

Broadcast can survive while upstream reporting fails

During a disaster, one-to-many communication from governments and broadcasters to residents is comparatively easier to maintain.

Government, broadcasters and communications operators
  ↓
Emergency alerts, public-address systems, television, radio and web
  ↓
Large numbers of residents

Returning information from the affected area requires a functioning communication route for every shelter, hospital, resident and field officer.

Residents, shelters, medical institutions and field officers
  ↓
Safety, damage, rescue needs, remaining power and road conditions
  ↓
Base stations, access lines and relay networks
  ↓
Municipalities, fire services, police and infrastructure operators

The ability to distribute an alert does not mean that conditions inside the affected area are understood.

Disaster communications should prioritize a few hundred bytes of standardized data transmitted redundantly over multiple routes before images and long-form reports.

Common Operating Picture

Disaster response requires a common operating picture

Immediately after a disaster, government agencies, municipalities, electricity utilities, communications operators, road authorities and medical institutions release fragmented information. However, there is no sufficiently integrated public operational view that presents those conditions by region and supports restoration decisions.

Publishing information on websites or social media does not mean that the generation, collection, integration and mutual verification of disaster information have been digitized.

The first task in disaster response is not to explain the damage. It is to establish two-way information routes and identify the areas that remain invisible.

Facility Continuity

A running emergency generator does not guarantee facility continuity

Whether an emergency generator started is not enough to assess the continuity of a hospital or government facility.

Keeping life-support equipment running is not the same as keeping a hospital operational.

The same applies to government buildings. Even if a building does not collapse, its administrative function has stopped when safety cannot be confirmed, terminals and cooling cannot operate, communications are unavailable, staff cannot reach the building, or surrounding roads are blocked.

The performance of a government building should be evaluated not only by whether it avoided collapse, but by whether public disaster services continued after the event.

Regional Survivability

The 72-hour period is not simply the time that stored power must last

The 72-hour period used in disaster planning is not a guaranteed electricity restoration period. It is the initial period before rescue operations and external support can become fully established.

Owning 72 hours of batteries or fuel is not, by itself, a BCP. The following functions must be maintained together throughout the initial 72 hours.

Begin preparing for wide-area evacuation within five hours

The complete extent of damage cannot be understood within the first several hours. However, evacuation preparation should not be postponed until all damage has been confirmed.

The question to be assessed during the first five hours is whether the affected region can sustain life for the next 24 to 72 hours.

1. People who have a receiving location and can travel independently
2. People who have a receiving location but need transport assistance
3. People without a receiving location who can travel independently
4. People without a receiving location who need medical, welfare or transport assistance

Priority should be given to people who depend on electricity, medical services or cooling for survival, including users of home oxygen, ventilators and dialysis, people requiring nursing care, pregnant women, infants and young children.

The 24-Hour Decision Boundary

The first 24 hours define whether supplies should move in or people should move out

Water and fuel repeatedly remain among the final unresolved problems after major disasters. This is not merely a shortage of relief supplies.

Can water and fuel continue to be delivered while the affected population remains inside the disaster area?

Water and fuel must be assessed as continuing flows, not stored quantities

• Water is required every day for drinking, food preparation, sanitation, medical care, toilets and cleaning.
• Fuel is continuously consumed by vehicles, generators, water and wastewater systems, refrigeration, communications equipment and restoration machinery.
• As long as residents remain in the region, demand returns the next day and the day after.
• Transport requires roads, drivers, vehicles, unloading, storage, distribution, security and information management.
• Private vehicles seeking water and fuel compete with ambulance, fire, restoration and logistics traffic.

Twenty-four hours is not the evacuation completion time; it is the deadline for choosing the primary strategy

By the end of the first 24 hours, at least the following outlook should be established.

• Expected water restoration or alternative water sources inside the region
• Ability to maintain wastewater, drainage and toilet functions
• Fuel delivery volume, delivery routes and controlled distribution capacity
• Expected duration of utility and emergency power
• Cooling and heat-protection measures under extreme temperatures
• Continuity of hospitals and welfare facilities
• Availability of roads, bridges, railways, ports and airports
• Receiving facilities and transportation outside the affected area

The 72-hour concept is not a reason to keep residents inside the affected area for 72 hours. Viability should be assessed by 24 hours. When it cannot be established, the remaining 48 hours should be used to execute wide-area evacuation.

The realistic choice is neither “everyone stays” nor “everyone leaves”

1. Give priority water and fuel to rescue, medical and restoration personnel.
2. Transfer medically dependent people, older people, infants, young children and pregnant women first.
3. Allow people with receiving locations who can travel independently to leave through designated routes and time periods.
4. Reduce the remaining population and therefore the total water, fuel, medical and sanitation demand inside the affected area.
5. Establish transfer hubs at the edge of the affected area and manage incoming supplies and outgoing evacuees under the same transportation plan.

Distributed Shelter and Fuel Logistics

A shelter is not merely a building; it is a base for sustaining life and supporting wide-area evacuation

When indoor shelters cannot be confirmed as safe and parking areas or school grounds are opened for people sleeping in vehicles, the vehicle functions as a private room, bed, cooling space, means of transportation and information terminal.

What is required is a distributed life-support and wide-area evacuation base that can keep vehicle-based evacuees safe and move them outside the affected area when necessary.

Fuel panic can obstruct restoration routes

Fear of fuel shortages
  ↓
Private vehicles converge on fuel stations
  ↓
Main roads, intersections and entrances become congested
  ↓
Ambulances, fire services, restoration vehicles,
water trucks and fuel deliveries are delayed
  ↓
Restoration slows and fuel anxiety increases further

The solution is not uncontrolled storage of portable fuel cans at shelters. It requires managed fuel storage, supply agreements with mobile refueling vehicles and tankers, priority distribution to emergency vehicles, medically dependent residents and evacuation vehicles, controlled quantities and operating hours, and waiting areas that keep queues off public roads.

Continuing Hazard

The period after an earthquake is not merely restoration; it remains an active disaster phase

When explosions, fires or collapses occur some time after the earthquake, the causes of individual incidents must await investigation. Specific causal claims involving gas systems, shutoff devices, ventilation or electrical equipment should not be made before the evidence is confirmed.

However, the possibility of serious incidents after a delay is itself an important design issue.

Initial earthquake
  ↓
Latent damage to pipes, joints, electrical equipment and structures
  ↓
Gas or combustible-material leakage, short circuits or pressure abnormalities
  ↓
Passage of time, stopped ventilation and residual energy
  ↓
Ignition, explosion, fire or collapse

Stopping the input does not, by itself, create a safe shutdown. A safe condition exists only when residual energy can be detected, isolated and removed.

Equipment that survives the initial shock must still be shut down, isolated and reinspected on the assumption that aftershocks and latent failures remain possible. Re-entry, re-energization and restart require controlled decisions.

Design Transition

Past lessons improved individual systems, but disasters spread horizontally across them

Power outage
  ↓
Communications, water, hospitals and shelters stop
  ↓
Blocked roads prevent fuel, water and medicine from arriving
  ↓
Communications failure hides damage and resource requirements
  ↓
Private vehicles seeking evacuation and fuel obstruct restoration routes
  ↓
Government, medical and logistics capacity declines

Emergency generators, mobile phones, alternative government offices, other shelters and nearby hospitals are not true redundancy when they depend on the same electricity grid, communications network, water system, roads and fuel logistics.

Four required design transitions

Verification after 72 Hours

Questions that must be examined after the initial 72 hours

Updated casualty figures and final incident causes will not produce structural lessons unless the following questions are examined.

Conclusion

Separate required functions from the shared failure domain

The first losses in the 2026 Kumamoto earthquake were not individual buildings. They were the electricity that moved the region and the communications that made the region visible.

The problem was not the absence of modern equipment. The problem was that the entire region, including its modern equipment, depended on the same electricity, communications, water, roads and fuel.

FAQ

Frequently asked questions

Related Articles

Related articles and terminology

Sources and Editorial Note

Principal sources and editorial note

This article is a contemporaneous record reconstructed from information available during the first 72 hours after the earthquake, including outage notices from electricity, communications and gas operators, publications by national and local governments and medical institutions, and field reporting by news organizations. Its purpose is to preserve the structural issues that could be recognized at that time.

• Earthquake information issued by the Japan Meteorological Agency
• Power outage information issued by Kyushu Electric Power Transmission and Distribution
• Outage information issued by NTT DOCOMO, KDDI, SoftBank and Rakuten Mobile
• Gas supply suspension and safety information issued by Kyushu Gas and Saibu Gas
• Publications by the Cabinet Office, Kumamoto Prefecture and affected municipalities
• Medical service, closure and patient-transfer information issued by medical institutions
• Reporting by NHK, TBS NEWS DIG, The Yomiuri Shimbun and Kumamoto Nichinichi Shimbun