Show Notes
Episode Overview
In this episode of Built, Wired, and Secured, Alex Morgan speaks with Michael Harrington about a problem many building owners, facilities teams, and IT leaders underestimate: the gap between what a UPS system is expected to do and what it actually delivers during an outage. The discussion opens with a realistic three-minute neighborhood power event where critical systems were supposed to stay online, but tenant leasing card readers stopped working, lobby doors failed to respond, and point-of-sale terminals rebooted unpredictably. No one was injured, but operations slowed immediately and tenants wanted answers.
That scenario frames the central point of the conversation: a UPS is not the same thing as guaranteed runtime. Assumptions around battery health, testing quality, ownership, and system architecture can create hidden risk inside commercial buildings. This episode focuses on the practical questions decision-makers should ask before the next outage exposes weak planning.
Why UPS Assumptions Fail in Real Buildings
Michael explains that UPS failures in the field usually fall into three broad categories.
- Battery aging: Batteries lose capacity over time, and real-world performance can degrade much faster than paperwork suggests. A battery rated for ten years may behave like a much shorter-duration asset after four to six years if conditions are poor or maintenance is weak.
- Testing that does not reflect reality: A quick staged test at a light load can make a UPS appear healthy even when it would fail under actual operating conditions.
- Ownership blind spots: If no one clearly owns load profiles, runtime targets, and maintenance signoff, the UPS becomes a silent single point of failure.
That ownership issue matters because it connects technical performance to business consequences. When the lights come back on, teams still have to explain why doors, access control, payment systems, or tenant-facing technology failed. The problem is rarely just hardware. It is usually a mix of aging equipment, weak validation, and unclear accountability.
Testing for Real Runtime, Not False Confidence
One of the clearest takeaways from the episode is that not all UPS tests are meaningful. Michael argues that quick vendor run tests often create more confidence than they should. If tenants expect ten minutes of runtime, the building team should be testing for ten minutes at a representative load, not for thirty seconds at twenty percent load.
He recommends:
- Testing with realistic loads and realistic durations
- Including battery discharge tests under actual environmental conditions
- Logging voltage, current, and temperature during testing
- Reviewing trend data over time instead of relying on one-time pass/fail checklists
The episode also draws an important distinction between commissioning and long-term assurance. Acceptance tests are useful when a system is first installed, but they do not replace scheduled, instrumented verification over the life of the asset. A building can pass its original test and still drift toward failure if no one is tracking actual performance.
Centralized vs. Distributed UPS Architecture
The conversation then shifts into design trade-offs. Centralized UPS systems can make sense when a building wants one battery bank, one maintenance contract, and a single long-run objective. That model can be efficient, but it also concentrates risk. If the central battery bank fails, multiple systems can go dark at the same time.
Distributed UPS deployments offer a different resilience model. By placing UPS units closer to door controllers, communication closets, or critical racks, teams can reduce exposure to a single point of failure and align runtime more closely to the systems that matter most.
Michael highlights the trade-offs directly:
- Centralized UPS: simpler to manage in some environments, but higher concentration of risk
- Distributed UPS: better alignment to critical loads and reduced failure exposure, but more units to monitor and maintain
The right choice depends on criticality and tenancy. In other words, leaders need to ask what they are protecting and how much downtime they can tolerate. Protecting a few emergency-service racks is not the same as protecting an entire floor of shared meeting rooms or tenant-facing building systems.
The episode also touches on redundancy. Parallel UPS strings can provide immediate failover, but they require matched equipment and coordinated maintenance to avoid common-mode failures. A staged redundancy approach may be more cost-effective, but only if load prioritization and automated transfer logic are clearly documented and tested.
Operational Discipline Matters as Much as Equipment
For facilities and IT leaders, one of the most valuable sections of the episode is the operational checklist Michael outlines. He recommends an approach that combines maintenance cadence, documentation discipline, and stronger vendor oversight.
- Run an annual full-load or representative-load discharge test
- Perform quarterly health checks that record capacity and internal temperatures
- Set battery replacement triggers based on measured capacity, not just time in service
- Require vendors to provide trend logs and a failure-mode report after each service visit
- Formalize handoffs between construction, commissioning, and operations
- Transfer load studies, single-line diagrams, and runtime target documents with the asset
- Assign clear accountability for maintenance, testing, and uptime commitments
That last point is especially important in multi-stakeholder environments. If no one knows who signs the uptime expectations for tenant-critical systems, the argument happens during the outage instead of before it.
Examples from the Field
Michael shares two practical examples that illustrate how different levels of investment can improve resilience.
The first is a low-cost mitigation: adding distributed UPS modules to communication closets that supported tenant access control and payment terminals. Those units were inexpensive, remotely monitored, and reduced outage complaints by 80 percent for a single-digit percentage of the original capital estimate.
The second is a capital-level correction at a large multi-tenant campus. There, an aging centralized battery bank was replaced with a modular, serviceable system, and critical loads were separated onto a dedicated UPS. The upfront spend was meaningful, but it removed a recurring failure mode and extended usable runtime during sustained outages. Over a three-year horizon, the reduction in tenant disruption justified the investment.
In both cases, the solution started with the same foundational question: what are we trying to protect, and how much downtime are we willing to tolerate?
Five Takeaways for the Next Site Walk
- Identify critical loads and document runtime targets
- Verify that tests are performed at representative loads and durations
- Require trend data, not just a completed checklist
- Set battery replacement criteria based on measured capacity
- Consider distributed UPS for tenant-facing, high-impact systems and assign clear ownership
Michael closes with one more operational recommendation: run tabletop exercises around realistic failure scenarios. Walk through a three-minute outage, a sustained fifteen-minute outage, and a battery failure during warm weather. Those scenario discussions reveal weak assumptions faster than lengthy manuals and help facilities and IT teams align on what breaks first and what should be protected.
For anyone responsible for building operations, tenant experience, or critical business systems, this episode is a reminder that resilience is not created by buying a UPS and assuming the problem is solved. It comes from validating runtime expectations, documenting priorities, and making sure accountability is clear before the next outage tests the plan.
UPS Expectations Fail Quietly Until the Outage Happens
Commercial buildings depend on more technology than many owners and operators realize. Access control, lobby entry, communications closets, payment systems, and tenant-facing infrastructure all rely on power continuity at moments when interruptions are most disruptive. That is why uninterruptible power supply systems are often treated as a safety net. The problem, as discussed in this episode of Built, Wired, and Secured, is that many teams mistake the presence of a UPS for proof of resilience.
Alex Morgan and Michael Harrington use a short outage scenario to show how quickly that assumption breaks down. In the example, a neighborhood-wide outage lasts only three minutes. The building’s UPS is expected to carry critical systems through the event. Instead, leasing card readers go offline, lobby doors stop responding, and point-of-sale terminals reboot unpredictably. No catastrophic damage occurs, but operations stall and confidence in the building’s readiness disappears almost immediately.
That scenario captures the real issue. UPS performance is not just about whether a unit exists. It is about whether the system has the battery health, runtime capacity, maintenance discipline, testing rigor, and ownership structure to perform under real conditions.
Why a UPS Can Still Be a Hidden Risk
One of the strongest points from the conversation is that decision-makers often assume a UPS equals guaranteed runtime. In practice, that assumption is dangerous. Michael breaks the most common failure modes into three categories: aging batteries, unrealistic testing, and ownership blind spots.
Battery aging is the most obvious technical issue, but it is also one of the most misunderstood. A battery may be rated for a long service life on paper, yet actual capacity can deteriorate far sooner. Michael notes that a unit described as a ten-year battery may act like a much shorter-duration asset after four to six years when maintenance is inconsistent or environmental conditions are poor. Heat, deferred service, and weak monitoring all accelerate that decline.
The second failure mode is testing that does not reflect actual building conditions. A short staged test on a light load may show that the UPS turns on and appears functional, but that does not mean it will carry critical systems for the duration tenants expect. It is easy to pass a checklist and still fail during a real event.
The third issue is operational, not electrical. Someone must own runtime targets, load documentation, maintenance signoff, and escalation when performance trends look weak. If those responsibilities are scattered or undefined, the UPS becomes a silent single point of failure. Everyone assumes someone else has it covered until the outage proves otherwise.
Testing Must Reflect the Real Load
One of the most practical lessons in the episode is the need to test UPS systems the way they are expected to perform in reality. If a building expects ten minutes of coverage for critical systems, testing at twenty percent load for thirty seconds is not meaningful. It may confirm basic functionality, but it does not validate runtime expectations.
Michael recommends testing with representative loads and for representative durations. He also emphasizes instrumented testing that records voltage, current, and temperature, because trend data tells a much more useful story than a one-time pass/fail result. This matters because batteries often degrade gradually. If no one is logging performance over time, early warning signs may be missed.
Another important distinction is the difference between acceptance testing and operational assurance. A system may perform acceptably when first commissioned, but that does not guarantee similar results years later. Building teams need scheduled verification that reflects current conditions, current loads, and current battery health. Commissioning proves the system started in a known state. Ongoing verification proves it stayed there.
Architecture Choices Shape Risk
The episode also explores how UPS design decisions affect resilience. Centralized UPS architecture can be attractive because it allows one battery bank, one maintenance contract, and a single long-run objective. In some environments, that creates efficiency and simplifies administration.
But centralized design also concentrates risk. If the central battery bank fails, multiple critical systems can drop at once. That may include tenant-facing technology, communications infrastructure, or building systems that affect operations far beyond the data room.
Distributed UPS design takes a different approach. Instead of protecting everything from one point, it places UPS capacity closer to specific systems such as communication closets, door controllers, and critical racks. That can reduce exposure to a single point of failure and align runtime more precisely with the needs of each load.
The trade-off is straightforward. Distributed systems can improve resilience, but they also increase the number of devices that must be monitored, maintained, and replaced over time. There is no universal right answer. The right architecture depends on what the building is trying to protect, how much downtime is acceptable, and how much operational complexity the team is prepared to manage.
The discussion on redundancy follows the same logic. Parallel UPS strings can provide immediate failover, but they require coordinated maintenance and matching equipment to avoid common-mode failure. Staged redundancy may be more cost-effective, but only if load priorities are clearly mapped and transfer logic is tested. If teams do not know which systems stay live and which are shed during a transfer event, they cannot predict tenant impact with confidence.
Operations, Documentation, and Vendor Oversight
Technology resilience in a commercial building is rarely just a hardware problem. It is usually an operations problem first. That is why Michael’s owner checklist is one of the most valuable parts of the episode.
He recommends an annual full-load or representative-load discharge test, plus quarterly health checks that record internal temperatures and capacity. Battery replacement should be tied to measured thresholds, not simply to the calendar. A battery that has lost too much capacity is a risk whether it is technically within its age range or not.
Vendor oversight is another recurring theme. Service visits should produce more than a generic maintenance summary. Teams should require trend logs and a failure-mode report after each visit so that degradation patterns, weak components, and unresolved issues are documented clearly.
Documentation handoffs matter just as much. Load studies, single-line diagrams, and runtime target documents need to move from construction and commissioning teams into operations. If that transfer never happens, the people responsible for ongoing performance inherit a critical asset without the information needed to manage it well.
Finally, accountability must be explicit. Someone needs to own the runtime targets and uptime expectations for tenant-critical systems. If ownership is vague, conversations about service levels turn into disputes during the outage instead of planning decisions beforehand.
Small Changes Can Deliver Real Resilience
One of the most useful aspects of the conversation is that it does not frame resilience as only a large capital project. Michael shares a low-cost mitigation that had measurable results: adding distributed UPS modules to communication closets supporting tenant access control and payment terminals. The modules were inexpensive, remotely monitored, and reduced outage complaints by 80 percent, all for a single-digit percentage of the original capital estimate.
That example matters because it shows how targeted investment can solve a high-impact problem without a full infrastructure overhaul. It also reflects a broader planning lesson: resilience improves fastest when teams identify the specific systems that create the most tenant pain when they fail.
The capital-level example is equally instructive. At a large multi-tenant campus, an aging centralized battery bank was replaced with a modular serviceable system, and critical loads were split onto a separate dedicated UPS. The upfront spend was significant, but it removed a recurring failure mode and extended usable runtime during longer outages. Over a three-year horizon, the reduction in tenant disruption justified the cost.
In both examples, the process started with the same question: what are we trying to protect, and how much downtime are we willing to tolerate?
A Practical Checklist for Building Leaders
This episode leaves listeners with a simple but effective framework for the next site walk or infrastructure review.
- Identify critical loads and document runtime targets
- Verify that testing is performed at representative loads and durations
- Require trend data, not just test completion
- Set battery replacement triggers based on measured capacity
- Consider distributed UPS for tenant-facing, high-impact systems
- Assign clear ownership for maintenance, testing, and uptime accountability
Michael adds one more recommendation that building and IT teams should not ignore: run realistic tabletop exercises. Walk through a three-minute outage, a sustained fifteen-minute outage, and a battery failure during warm weather. Those scenarios reveal operational assumptions quickly and help teams understand where the real points of failure are.
Resilience Is a Business Decision, Not Just an Electrical One
The larger message behind this episode is that UPS planning is not a narrow facilities topic. It directly affects tenant experience, operational continuity, vendor accountability, and the credibility of the teams responsible for the environment. When building leaders treat the UPS as a primary asset instead of an afterthought, they make better decisions about maintenance, testing, architecture, and capital planning.
For commercial properties, that shift has real business value. Fewer outages mean fewer tenant complaints, less disruption to leasing and access systems, more predictable operations, and better alignment between facilities and IT. In other words, resilience is not just about backup power. It is about protecting trust, continuity, and the day-to-day functionality tenants depend on.
If you want to hear the full conversation and the practical examples behind these recommendations, listen to this episode of Built, Wired, and Secured. It is a useful reminder that backup power only protects the business when the assumptions behind it are tested, documented, and owned.