When Disaster Hits the 911 Center: Continuity of Operations in the NG911 Era

Every 911 center is one bad day away from finding out how good its continuity plan really is.

A wildfire jumps a ridge line and the evacuation zone now includes your dispatch floor. A hurricane takes down the fiber path feeding your building. A water main breaks two floors

above the comm center. A ransomware incident locks up the county network on a Friday

afternoon. None of these scenarios are hypothetical, and none of them care whether your continuity of operations plan is current.

What has changed is what a good answer looks like. In the legacy 911 world, continuity planning

was largely a physical problem solved with physical answers: a backup building, a diesel

generator, a set of alternate trunk lines, and a binder of phone numbers. Next Generation 911

changes the math. When call delivery rides on an IP-based Emergency Services IP Network

rather than dedicated copper into one specific building, the location of your call takers becomes

far less important than the design of your network and the quality of your routing policies.

That is a genuine operational upgrade, but it is not automatic. Resilience in an NG911 environment is something you architect, document, and test. This guide from NGA walks through what actually fails during a disaster, what modern continuity looks like when it works, and how 911 authorities and PSAP directors can build a continuity plan that holds up on the worst day of the year.

Why Continuity Planning Looks Different Under NG911

Legacy 911 architecture created a hard dependency between a call and a building. Selective

routers delivered calls over dedicated trunks to a specific PSAP at a specific address. If that

address became unusable, the options were limited and slow: activate a backup center, ask the

carrier to reroute trunks, or fall back to a ten-digit administrative line and hope the public found it.

NG911 decouples the call from the building. Calls are routed by policy rather than by hard-wired

circuit. An i3-based system uses functional elements such as the Emergency Call Routing

Function and the Policy Routing Function to decide, in real time, where a call should land. That

decision can account for time of day, current call volume, PSAP status, or an explicit continuity

condition that a supervisor has activated.

In practice, this means a PSAP can be evacuated without the public losing access to 911. The

calls keep arriving. They simply arrive somewhere else, or at someone else's workstation.

NENA's guidance on virtual PSAP management makes this point directly: where an agency has

virtualized its call handling, continuity of operations can mean continuing to function off site

rather than relocating an entire workforce to a second physical facility.

The catch is that policy-based rerouting only works if the policies exist before the emergency,

the receiving agencies have agreed to accept the traffic, and someone has tested the whole

chain end to end. Capability on paper is not the same as capability under load.

The Four Failure Modes Every COOP Plan Should Address

State COOP templates, including the planning guides published by state 911 offices in Wisconsin and Nebraska, tend to converge on the same core scenarios. They are worth thinking about separately because each one breaks something different.

• Loss of facility. Fire, flood, structural damage, hazardous materials, smoke infiltration, or an evacuation order makes the building unusable. Staff are fine, equipment may be fine, but nobody can occupy the room. This is the scenario backup centers were built for, and the scenario remote call handling addresses best.
• Loss of connectivity. The building is fine and staffed, but the network path carrying 911

traffic is severed. A single backhoe, a washed-out bridge carrying conduit, or a failed

aggregation point can do it. Diversity of physical path matters more here than diversity of

building.

• Loss of power. Generators and uninterruptible power supplies cover the comm center, but the supporting telecommunications infrastructure between the caller and the PSAP also needs power. Extended commercial outages have historically been a leading contributor to 911 service disruptions, which is why central office backup power keeps appearing in federal reliability discussions.
• Loss of the system or the workforce. A cyber incident, a failed upgrade, or a public health event that removes a large share of staff at once. This category is often the least rehearsed and increasingly the most likely.

A useful test of any plan: pick one of the four, then trace exactly who does what in the first

fifteen minutes. If the answer depends on one person's memory or one vendor's phone number, the plan has a gap. 

What Resilience Looks Like When It Works

The clearest recent illustration comes out of western North Carolina. When Hurricane Helene

devastated the region in 2024, a 911 center in Madison County lost connectivity entirely.

Because the state had already moved to a shared ESInet, the state 911 Board was able to

reroute that county's 911 calls to another call center roughly 220 miles away, and keep them

there for about a month while local infrastructure was rebuilt. Connectivity back into the ESInet

was later restored using a broadband public safety network connection.

Two details in that story matter more than the headline. First, the rerouting was possible

because the architecture already supported it. Nobody built that capability during the storm.

Second, the receiving PSAP was far enough away to be outside the disaster footprint. North

Carolina's 911 Board has since formalized this thinking in its resiliency guidance, pairing PSAPs

with partner agencies in different regions on the logic that a nearby backup is worth very little

when the event is regional.

The state has also layered in lower-tech redundancy that tends to get overlooked in

modernization conversations, including satellite phones and rugged push-to-talk radios

distributed to PSAPs ahead of hurricane season. Resilience is not only an architecture question.

Sometimes it is a question of whether the supervisor on duty can reach the neighboring county

when everything else is dark.

Continuity Across State Lines

The most interesting development in this space pushes the concept further. In February 2026,

the North Carolina 911 Board, the Washington D.C. Office of Unified Communications, and

Johnston County 911 completed a first-of-its-kind proof of concept that delivered live 911 calls

across state lines.

The exercise validated three things simultaneously: policy-based call routing, interoperability

between separate ESInets, and continuity of operations spanning two jurisdictions hundreds of

miles apart. In other words, a catastrophic regional event no longer has to be answered from

inside the same region, or even the same state.

That is a meaningful shift in how 911 authorities can think about mutual aid. The traditional model assumed a neighboring county would absorb overflow. An interoperable model allows a

state to build relationships with distant partners chosen specifically because they share no

common risk profile, no common weather system, and no common power grid.

The Regulatory Picture Is Moving Toward Resilience

Federal attention has followed the same trend line. Major 911 outages affecting multiple states

have increased as agencies have begun transitioning to NG911 platforms, and in July 2026 the

FCC adopted updated reliability rules aimed squarely at that problem.

For continuity planners, the substance matters more than the rulemaking history. The direction

of travel includes expectations around automatic switchover to geographically diverse facilities,

interoperability between ESInets, clearer definitions of which providers are accountable for

critical NG911 pathways, and more visibility for state and local 911 authorities into the reliability

of the systems they depend on. There is also a process by which authorities can raise identified

deficiencies and receive a response within a defined window.

The practical takeaway is that geographic diversity and documented failover are no longer

differentiators that sophisticated agencies pursue voluntarily. They are becoming the baseline

expectation, and agencies that can evidence them will be in a considerably better position

during their next contract cycle or after-action review.

Building a Continuity Plan That Will Actually Hold

Most PSAPs have a COOP document. Fewer have one that reflects current architecture, current

staffing, and current partner agreements. The following areas are where plans most often drift

out of date.

• Confirm real geographic diversity. Ask your network provider to show the physical path of every connection into your building, not the logical diagram. Two circuits that share one conduit, one riser, or one aggregation site are one circuit with extra paperwork.
• Write the routing policies now, and name who can activate them. Continuity rerouting should be a documented, pre-approved policy with a defined trigger and a short list of people authorized to invoke it. Ten minutes of hesitation over authority is ten minutes of unanswered calls.
• Choose partner PSAPs outside your risk footprint. Map your agreements against the

hazards in your region. If a single hurricane, wildfire complex, earthquake fault, or grid

operator can affect you and your backup at the same time, you need a more distant partner.

• Make remote call handling a tested capability, not a theory. If your plan assumes

telecommunicators can take calls from an alternate site or from home, verify it on a normal Tuesday. Test credentials, headsets, CAD access, mapping, recording, supervisor visibility, and bandwidth. Recording and quality assurance obligations do not pause during a disaster.

• Do not skip the analog layer. Satellite phones, interoperable radio channels, and printed contact rosters remain the last line of defense. Check that they are charged, programmed, and that someone on every shift knows where they are.
• Plan for the people, not just the calls. Telecommunicators working a regional disaster are often personally affected by it. Relief scheduling, food and rest arrangements, family

notification, and post-incident support belong in the plan alongside the network diagrams.

• Exercise annually, and write down what broke. A tabletop exercise once a year with an honest after-action report will surface more real gaps than any amount of document

revision. NENA's operational guidance emphasizes exactly this loop of status reporting,

after-action review, and mitigation of future disruptions.

Conclusion

The agencies that came through recent disasters with 911 service intact were not lucky. They

had made specific decisions in advance: shared or regional ESInet architecture, geographically

separated core components, pre-negotiated partner agreements, and staff who had practiced

the failover before they needed it.

For 911 authorities evaluating NG911 systems, that suggests a different set of procurement

questions. Not just whether a platform is i3 conformant, but where its core elements physically

live, how failover is triggered and how quickly, whether it can hand traffic to a neighboring

ESInet, how remote call taking is secured and licensed, and what the provider is contractually obligated to do at three in the morning during a regional emergency.

Continuity of operations has quietly become one of the strongest arguments for NG911

modernization. Better location data and multimedia capability get most of the attention, and

they deserve it. But the ability to lose a building and not lose 911 is the kind of capability that

only proves its value once, and by then you cannot go back and buy it.

If your continuity plan has not been opened since your last accreditation cycle, that is the place

to start. Pull it out, read it against the system you actually operate today, and find the gap

before the weather does.

Contact NGA to learn more about their NG911 platform and how it provides the foundation for the next generation of emergency communications technology.