Who Gets the Power? Prioritizing Critical Loads in Building Outages

Why Power Prioritization Fails in Real Buildings

This episode opens with a familiar but costly scenario: the generator starts, emergency lighting returns, and the building appears stable for a moment. Then the real problems begin. Tenant servers stay dark, elevators remain on recall, HVAC behaves unpredictably, and no one can say with confidence which panel feeds which critical system. The core message is simple: when power is limited, a building cannot rely on assumptions, tribal knowledge, or a policy that only makes sense in a conference room. If the prioritization plan is not short, signed, and testable, it will likely fail when an outage forces fast decisions.

Alex Morgan frames the conversation around one question that should guide every outage decision: what breaks if this goes down? That question pushes teams to move beyond vague categories and define actual consequences. The discussion makes clear that outages become chaotic when facilities, IT, property management, and tenants all operate from different assumptions about who owns what, what is protected, and what can safely be shed.

Where Outage Plans Break Down

• Ownership boundaries are often unclear between property teams, facilities staff, IT, and tenants.
• Facilities may assume tenants protected their own critical equipment with UPS systems.
• Tenants may assume building emergency power covers communications or server loads.
• Undocumented feeders and patched cross-connections create hidden single points of failure.
• Drawings, labels, and maintenance records are often outdated or incomplete.

One of the most important examples in the episode involves undocumented feeder paths. A system may appear to be protected when in reality it is tied to non-emergency power, or worse, connected through an improvised patch no one has recorded. That kind of hidden dependency does not usually show up until a real outage exposes it. The guests describe these cross-connections as silent killers because they undermine both operations and confidence.

A Defensible Prioritization Framework

The conversation lays out a practical sequence for prioritizing loads during an outage. First come life-safety systems: fire suppression, egress lighting, and emergency communications. Next are the communications and monitoring systems that support incident response, such as building alarms and core network functions. After that come business-critical tenant loads, but only when those loads are documented and tied to defined service expectations. Comfort loads, including full HVAC, generally come later unless a process would be damaged without environmental support.

The guests stress that this is not just a philosophy exercise. Priorities must be grounded in actual emergency capacity. If a generator can only support a certain number of kilowatts per feeder, the building cannot promise more than that. In the example shared, a 500 kW generator might reserve 350 kW for life safety and communications, leaving just 150 kW for other essential needs. That kind of limit needs to be documented clearly so operators are not forced to guess under pressure.

Why Sequencing Matters as Much as Priority

A major lesson from the episode is that deciding what should come on first is only half the job. Teams also need to define how power returns in sequence. If too many loads come online at once, inrush current can overload the generator and trip the system. That means an otherwise sensible prioritization can still fail if startup timing is not controlled.

• Bring life-safety systems online first.
• Bring communications and monitoring online next.
• Restore tenant circuits in defined stages rather than all at once.
• Use delays between major feeder startups.
• For large inrush loads, use a confirmation hold before energizing the next stage.

The practical timing guidance is especially useful. The guests suggest 30 to 90 seconds between major feeder transitions, or a 2-minute confirmation step for large inrush loads. These are the kinds of operational details that turn a concept into a plan someone can actually execute at 2 a.m.

What an Executable Policy Looks Like

Michael makes a strong case for simplicity. A workable outage plan should fit on one page. It should tell the operator exactly what to keep, what to shed, which manual bypass steps may be needed, and who to call. The episode emphasizes that if a night technician has to interpret a long policy or escalate basic decisions during a power event, the document has already failed its purpose.

• A one-page prioritization matrix signed by facilities, IT, and property management.
• A switch list that identifies feeders to keep and feeders to shed.
• Manual bypass and rollback steps.
• Emergency contact names and phone numbers in both a binder and a tablet.
• Published emergency capacity per feeder in kilowatts.

Testing and Maintenance Keep Priorities Real

The second half of the episode reinforces that documentation alone is not enough. Reliability comes from maintenance discipline. The team recommends quarterly short transfer tests of about 10 minutes to validate generator start and stage-one load behavior. They also recommend annual full-load tests lasting two to four hours during a low-occupancy window. Every test should use a checklist that covers switching actions, confirmation points, rollback steps, and verification ownership.

Maintenance should also include thermographic checks and labeling audits. These steps help teams find overloaded or mislabeled feeders before an outage turns those weaknesses into business disruption. The message is consistent throughout the episode: policies are only credible when supported by verification.

Two Practical Examples

In a small downtown office retrofit, the owner wanted comfort-grade HVAC on generator power, but capacity was limited. After mapping circuits, the team found critical communications on non-emergency power and reallocated a dedicated tenant breaker to a UPS circuit instead of the generator. The decision worked because the trade-offs were documented and signed, allowing servers to remain online for several hours without upsizing the generator.

At campus scale, sequencing became the key issue. Labs needed HVAC restored within 30 minutes to avoid damaging samples. The team segmented the campus, used staged transfer switches tied to a central controller, and retained manual override steps. An early test revealed that the wrong startup order overloaded the generator, so they added a time delay and an operator confirmation step. That 2-minute hold solved the repeat overload problem.

What To Do This Week

• Map and sign a one-page prioritization matrix across facilities, IT, and property management.
• Publish emergency capacity per feeder in kilowatts so limits are clear.
• Create a one-page switch list for the emergency binder and tablet.
• Schedule quarterly 10-minute transfer tests and an annual 2- to 4-hour full-load test.
• Update drawings, labels, and maintenance records so the plan reflects reality.

The episode closes with a practical offer: GDS Technology has a one-page prioritization checklist covering mapping, sign-off, capacity limits, staged startup timings, and test cadence. For building operators and IT leaders, the takeaway is clear. Better outage outcomes do not come from longer policies. They come from simple rules, verified dependencies, disciplined maintenance, and a plan that works in the dark.

Who Gets the Power First When the Building Goes Dark?

When a building loses power, the first few minutes can create a false sense of relief. The generator starts. Corridor lights return. Stairwells glow again. The property team feels like the emergency system did its job.

Then the phone starts ringing.

A tenant reports that critical servers never came back online. Elevators are stuck on recall. HVAC is short cycling. No one can say for certain which panel feeds the communications closet. What looked like a successful transfer quickly turns into an operational mess, a tenant relations problem, and a business continuity failure.

That is the scenario explored in this episode of Built, Wired, and Secured. The central argument is as practical as it is urgent: if your power prioritization policy is not short, signed, and testable, it will fail when power is scarce.

For commercial buildings, mixed-use properties, and any environment supporting tenant-critical operations, that lesson matters. Outage decisions are rarely made in ideal conditions. They happen under pressure, often at odd hours, with limited staffing and incomplete information. That is exactly why the policy cannot be theoretical. It has to be operational.

Start With One Question: What Breaks If This Goes Down?

One of the strongest moments in the episode is the opening framing question: what breaks if this goes down?

That question sounds simple, but it forces clarity. If a team cannot answer it in one sentence for a given load, system, or feeder, then the building does not really have a prioritization plan. It has assumptions. And assumptions are dangerous in outages.

The value of that question is that it moves the conversation away from labels like important or critical and toward actual consequences. Does shutting down this load create a life-safety issue? Does it block emergency response? Does it interrupt tenant operations tied to revenue, compliance, or perishable processes? Does it only reduce comfort for a limited period?

That consequence-based thinking makes prioritization defensible. It also makes cross-functional alignment much easier because facilities, IT, and property leadership can debate outcomes instead of vague categories.

Why Outage Priorities Break in the Real World

The episode makes it clear that most failures are not caused by one bad device or one bad person. They happen because people, documents, and systems all drift apart over time.

Ownership boundaries get blurry. Property management assumes tenants are protecting their own critical loads. Facilities may believe their responsibility ends with life safety. IT may assume a communications closet is covered by emergency power when it is actually fed from a standard panel. A tenant may have installed a UPS years ago, then repurposed it through an undocumented patch to support something else entirely.

Those undocumented cross-connections are especially dangerous. A hidden dependency can sit quietly for years and only reveal itself during a real outage. By then, the building is already in crisis mode.

This is why mapping dependencies matters so much. A workable outage strategy depends on knowing not just what equipment exists, but what feeds it, what it supports downstream, and which systems have quietly become linked in ways no one intended. If those answers live only in someone’s memory, the plan is fragile by definition.

The Right Priority Order Is Usually Not the Hard Part

At a high level, the priority framework in the episode is straightforward. Life safety comes first: fire suppression, egress lighting, and emergency communications. Then come the communications and monitoring systems needed for response and coordination, including building alarms and core network services. Next are business-critical tenant loads with documented service expectations. Comfort systems such as full HVAC typically follow after that, unless a temperature-sensitive process would be damaged without environmental support.

Most experienced operators will not disagree with that order. The harder part is making the trade-offs explicit and binding before an emergency forces the issue.

That means documenting what the generator can actually carry, feeder by feeder, in kilowatts. A building cannot promise resilience it does not have. If the generator capacity supports life safety and communications but not tenant comfort loads, that limitation needs to be documented in advance. If some tenant-critical systems will be supported while others will not, those decisions need to be visible, signed off, and understood by the right stakeholders.

That is what makes the plan defensible later. When a difficult call has already been mapped, approved, and documented, teams are no longer improvising in the middle of an event.

Sequencing Is Where Good Plans Often Fail

One of the most useful takeaways in the episode is that prioritization and sequencing are not the same thing. It is not enough to know which loads matter most. Teams also need a staged startup plan that controls how those loads return.

If too much equipment starts at once, inrush current can overload the generator and trip the system. That means even a correct priority list can fail if the startup sequence is poorly designed.

The guests recommend staged restoration: bring up life safety first, then communications, then tenant circuits in measured phases. They also give practical timing rules, such as 30 to 90 seconds between major feeder transitions and a 2-minute confirmation hold for large inrush loads.

This is the difference between a policy that sounds intelligent and one that someone can execute reliably at night, under stress, without guessing. Timing matters because real-world equipment does not behave like a spreadsheet. Motors, HVAC systems, and larger electrical loads place demands on backup power systems that have to be managed deliberately.

The Best Policy Fits on One Page

Another strong point from the discussion is that a functional outage policy should be short. In fact, one guest argues that it should be one page maximum.

That page should identify what stays on, what gets shed, what manual bypass steps exist, and who needs to be called. Contact names and phone numbers should be available in both the emergency binder and a tablet or digital copy. The operator should not need to decode a complex policy or escalate obvious actions during a live outage.

This matters because complexity fails under pressure. The more a plan depends on interpretation, the more likely it is to produce hesitation, inconsistency, or outright mistakes. A simple, signed, field-ready switch list is more valuable than a long policy document that only makes sense in daylight.

Maintenance Is What Turns Policy Into Performance

Throughout the episode, maintenance is treated as more than a supporting activity. It is the foundation of reliability.

The recommendations are practical: run quarterly short transfer tests of roughly 10 minutes to confirm generator start and stage-one load behavior. Run an annual full-load test for two to four hours during a low-occupancy period. Use checklists during each transfer that define switching steps, confirmation points, rollback actions, and who verifies each part of the process.

Just as important, keep labels and drawings current. Add thermographic checks and labeling audits to the maintenance plan so overloaded or mislabeled feeders are found before an outage exposes them.

This is where many organizations fall short. They create a policy once, store it, and assume they are prepared. But preparedness is not a document. It is the combination of documented intent, current system knowledge, and repeated verification.

Real Examples Make the Case

The episode includes two examples that show how these principles play out in practice.

In a small downtown office retrofit, the owner wanted comfort-grade HVAC on the generator, but capacity was limited. Circuit mapping revealed that critical communications were tied to non-emergency power. Rather than upsize the generator, the team reallocated a dedicated tenant breaker to a UPS-backed circuit. The result was that tenant servers stayed online for a few hours, and the trade-off was defensible because it had been documented and signed.

At a campus scale, the challenge was sequencing. Labs needed HVAC restored within 30 minutes to prevent sample loss. The team segmented the site, used staged transfer switches tied to a central controller, and retained manual override steps. An early test exposed a bad startup order that overloaded the generator. They corrected it with a time delay and an operator confirmation hold, preventing repeat failures.

Both examples reinforce the same point: resilience is not always about buying more capacity. Often it is about understanding dependencies, making explicit trade-offs, and building a sequence people can trust.

What Building and IT Leaders Should Do Next

If this episode has a practical action list, it is this: map and sign a one-page prioritization matrix, publish emergency capacity per feeder, create a one-page switch list, test quarterly, test annually, and keep drawings and labels current.

Those steps sound basic because they are. That is part of the point. Better outage performance usually does not come from more complexity. It comes from disciplined basics executed consistently across facilities, IT, and property operations.

For organizations that want a faster starting point, the episode mentions a GDS Technology one-page prioritization checklist covering mapping, sign-off, capacity limits, staged startup timing, and test cadence. Tools like that are valuable because they help teams move from concept to execution without overcomplicating the process.

If your building depends on backup power to support tenant trust, operational continuity, and defensible decision-making, this episode is worth a listen. It is a grounded reminder that in an outage, the question is never just whether the generator starts. The real question is whether the right things stay on, in the right order, for the right reasons.

Listen to the full episode to hear the full discussion and use it to pressure-test your own outage prioritization process before the next power event does it for you.