When the utility drops, a hospital does not get to wait. Operating rooms, intensive care, life support, fire systems, and egress lighting all depend on emergency power coming up and carrying the load. That is the whole reason healthcare standby power exists, and it is why servicing it is held to a tighter bar than almost any other building you work on.
If you service generators for hospitals or other healthcare facilities, the commercial baseline does not simply carry over. Several authorities apply at once, the system is engineered to prioritize load in ways an ordinary building is not, and the record you leave behind has to survive a surveyor. This guide lays out how the pieces fit — and is careful about what it does not pin down, because the exact numbers live with the standards and the authorities, not in a figure you can memorize.
Why healthcare is held to a tighter bar
In a typical commercial building, standby power is about keeping the lights on and the business running. In a hospital, it is about keeping people alive. The loads tie directly to human safety, so the requirements are more demanding and there are more of them.
Healthcare emergency power sits under additional standards and additional oversight layers that a normal occupancy never sees. That is the practical difference: more authorities, a tighter standard, and a much lower tolerance for a record that cannot be defended.
The overlapping authorities
The thing that trips up operators coming from commercial work is that healthcare emergency power is not governed by one rulebook. Several apply at the same time, and they overlap.
- NFPA 110 — the standard for the emergency and standby power system itself: the engine-generator set, the transfer equipment, and the controls and accessories, and how that system is installed, maintained, operated, and tested.
- NFPA 99 — the health-care facilities standard, which addresses how the essential electrical system and other critical systems serve a healthcare facility, including how power is distributed and prioritized.
- An accrediting organization — such as The Joint Commission, a widely recognized U.S. healthcare accrediting organization that evaluates facilities against its own published requirements and looks for evidence that emergency power is tested and documented. Confirm the specifics with the facility's accrediting body.
- The authority having jurisdiction (AHJ) — enforces the adopted codes locally and determines which editions apply and how they are enforced.
- CMS — for hospitals participating in Medicare and Medicaid, the conditions of participation sit on top of the fire and life-safety framework as a condition of being paid.
The honest framing
NFPA 110 and NFPA 99: two different jobs
These two get conflated constantly, and keeping them straight makes you sound like you know the building.
NFPA 110 — the system
NFPA 110 is about the emergency and standby power system itself. The engine-generator set, the transfer equipment, the controls and accessories — and how that system is installed, maintained, operated, and tested. If it is part of producing and switching standby power, it lives here.
NFPA 99 — the facility
NFPA 99 is the health-care facilities standard. It addresses how the essential electrical system and other critical systems serve the facility — how power is distributed and prioritized across the building so the most critical loads are handled first. Where NFPA 110 is about the source, NFPA 99 is about how that power is organized to serve patient care.
How the load is split: the branches
Broadly, you will hear about a life-safety branch — systems tied directly to occupant safety, such as egress lighting, exit signs, alarm and fire systems — a critical branch serving patient-care equipment and areas, and an equipment branch serving other building systems that can tolerate a short delay before they come back. The point for a service company is that not all load is equal. The system is engineered to prioritize, and the testing has to confirm that the prioritization actually works when power is lost.
The precise definitions of each branch, what connects to it, and how each is tested are defined in the governing standards and the adopted editions. The takeaway to carry into the building is the concept: the essential electrical system is not one undifferentiated load, and a test has to prove the right things come back in the right order.
The shape of healthcare emergency power testing
Emergency power testing generally follows a cascade — routine inspection, operational testing, and periodic load testing — from frequent and light to less frequent and more demanding.
Routine inspection
Routine inspection is the regular hands-on checking of fluids, batteries, connections, and components. It is the walk-the-machine work that catches small problems before they become a no-start at the worst possible moment.
Operational testing
Operational testing means running the system on a regular cadence to confirm it starts, runs, and transfers — exercising the transfer sequence so you know the switch actually moves the load when the utility drops.
Periodic load testing
Periodic load testing is the less frequent, more demanding test under load. Where adequate building load is not available, supplemental load testing — such as a load bank — commonly applies, on terms set by the governing standards and the AHJ.
Confirm the specifics
The permanent record is what a surveyor asks for
The governing authorities expect the testing to happen — and they expect a permanent record that it happened. Surveyors ask to see documentation of inspection, testing, and maintenance. A test you ran but cannot prove you ran does very little when someone comes asking.
A defensible record generally identifies:
- The equipment by identity — make, model, serial, and location.
- When the work was performed.
- Who performed it, as a named person.
- The readings and results.
- Any deficiencies found and corrected.
The exact contents and retention of records are governed by the editions and the authorities, but those elements are the backbone of a record that holds up. The through-line is the same one the whole occupancy turns on: a reading with no named person, no timestamp, and no equipment identity is hard to defend in front of a surveyor.
Setting a default cadence on your healthcare agreements
Most service shops put a default test cadence in their service agreements so the schedule is clear before the first visit — then adjust per site to match the governing standards, the adopted editions, the accrediting organization, and the AHJ. Stating your default plainly keeps the facility's expectations and your dispatch calendar aligned.
Whatever the default is, write it down, and write down that it yields to the site: the agreement states the cadence, and the cadence defers to the governing standards, the adopted editions, the accrediting organization, and the AHJ wherever they say otherwise. An agreement that is explicit about both halves protects the shop and the facility at survey time.
Where PowerOps fits
PowerOps does not set test intervals, system classification, or which editions apply — the governing standards and the authorities do. What PowerOps does is make the record defensible. Completed PMs, load bank results, and transfer tests become records written by a named, authenticated technician at sign-off via an explicit click — not a backfillable checkbox.
Field readings, photos, and serial numbers are captured in the field rather than retyped at the office. And the canonical data — the facility, the site, the equipment list, access notes, prior findings — auto-flows from one visit to the next, so technicians are not re-entering it every time. The result is an attributable, timestamped, immutable record: a named person, a timestamp, the equipment identity, the readings, and the deficiencies — the kind of record a healthcare surveyor actually asks to see.