In a fire or smoke event, hospital elevators must do two seemingly opposite things: stop carrying ordinary passengers and start serving the people who manage the emergency. This Article covers the two code-mandated elevator life-safety modes — Firefighters' Emergency Operation (FEO), which recalls cars and hands manual control to responders, and Occupant-Evacuation Operation (OEO), the newer mode that lets designated elevators move ambulatory occupants out of a building during a fire. It also addresses why these requirements are unusually consequential in hospitals, where a meaningful share of "occupants" cannot use stairs at all.

The companion Article Elevator Emergency Power, Recall & Resilience covers the power source that keeps these modes alive (generator transfer, selective car operation under emergency power, shunt-trip interlocks). This Article stays on the control logic, sequences, signaling, and clinical/operational consequences of FEO and OEO themselves.

Why elevators matter more in a hospital fire

Standard building life-safety doctrine is "in a fire, use the stairs, not the elevator." That doctrine assumes a population that can self-evacuate down a stairwell. A hospital violates that assumption structurally:

This is why elevator life-safety in healthcare is not a checkbox. It governs how the fire department fights a fire above grade, how staff defend patients in place, and — in the rare full-evacuation scenario — whether vertical movement is even feasible. The governing framework comes from the International Building Code (IBC) (occupancy, accessible-means-of-egress, and the OEO/FSAE provisions), ASME A17.1/CSA B44 (Safety Code for Elevators and Escalators) (the actual FEO and OEO control requirements in its dedicated emergency-operations section), NFPA 101 (Life Safety Code) and NFPA 99 (Health Care Facilities Code) (the health-care-occupancy egress and defend-in-place model that CMS enforces), NFPA 72 (National Fire Alarm and Signaling Code) (the initiating devices and interface), and NFPA 13 / NFPA 70 (NEC, including Article 517 for health care) for the suppression and electrical interfaces. The local Authority Having Jurisdiction (AHJ) — the fire marshal and elevator inspector — interprets and enforces all of it, and accreditation bodies (The Joint Commission, DNV, CMS surveyors) test it in the field.

Phase I: Emergency Recall Operation

Phase I is the automatic mode. When a fire-detection device triggers, the elevator system removes cars from public service and brings them to a designated level under its own logic, then parks them with doors open and waits for a firefighter.

What triggers Phase I. Phase I is initiated by smoke detectors located in the elevator lobbies, the machine room, the machine space / control space, and the hoistway (where sprinklers are present at the top of the hoistway). Critically, the smoke detector in the lobby at the designated recall level recalls the cars to the alternate level, on the logic that the primary discharge level may itself be the fire floor. Activation of these detectors, or in many systems a building-fire-alarm signal mapped to the elevator interface, drives Phase I. (Sprinkler waterflow and heat detection in the machine room or hoistway also tie into the related shunt-trip / power-shutdown interlock — that interaction is detailed in the resilience Article; the key principle is that detection that recalls a car must precede any detection that removes power from it, so a car never loses power between floors with occupants aboard.)

The recall sequence.

  1. On a Phase I signal, cars cancel all registered car and hall calls and do not answer new ones.
  2. Each car travels nonstop to the designated recall level (typically the main building-egress / fire-command level) — or to the alternate recall level if the triggering detector is at the designated level.
  3. The car arrives, opens its doors, and parks with doors open, out of service. A visual signal (the illuminated firefighters' hat symbol) and audible signal indicate the car is in Phase I.
  4. A car already moving away from the recall level reverses at the next available opportunity; a car moving toward it continues.

The Phase I key switch. At the designated level, a three-position key switch (OFF / ON / BYPASS) lets responders manually recall ("ON"), restore normal-pending ("OFF"), or — with "BYPASS" — return cars to normal service when the initiating device has not automatically recalled them. Modern code keys these to the standardized FEO-K1 key so that any responding department's master key operates any compliant elevator.

In a hospital, Phase I has an immediate clinical consequence: the moment a detector trips, every car in that bank stops serving the floors. Patient transport, code-team movement, OR turnover, supply runs, and family circulation in that bank all halt. This is why clinical-flow separation and redundant bank design (covered in the design-requirements Chapter) matter for resilience — a fire alarm in one elevator lobby should not strand an entire service if banks and detection zones are planned so that not all vertical capacity recalls simultaneously.

Phase II: In-Car Firefighters' Operation

Phase II begins only after Phase I has parked a car. It puts a single car under the exclusive manual control of a firefighter standing inside it, defeating all the automatic protections that normally keep passengers safe — because a firefighter needs to do things (move to the fire floor, hold doors, override door reopening through smoke) that those protections would otherwise prevent.