Most multistory buildings have enclosed exit stairways somewhere within them to provide safe egress from their non-discharge floors. In a building that relies on hydronic heating/cooling, steam heating, or direct (resistance) electrical heating, the US model building Codes are rather clear that the heating/cooling equipment for the stair can live within the stair enclosure, in addition to having the ventilation equipment for that stair enclosure placed within the stair enclosure, connected directly to outside air.
This is as per the penetration rules in NFPA 101 220.127.116.11.1(10), which explicitly permit steam and water heating and cooling lines to serve the stairshaft in addition to electrical conduits, as well as IBC 1023.6 point 2, which permits exit stair ventilation equipment (in my case, a "single room" or "decentralized" type HRV or ERV in addition to an air handler/indoor unit) to be located within the stair enclosure. Furthermore, the IBC's definition of ventilation includes means for supplying conditioned air to the space being ventilated, which supports interpreting 1023.6 point 2 to permit the placement of an air handler dedicated to an exit enclosure within that exit enclosure.
However, it seems to me that this argument falls down as soon as you provision direct-expansion air conditioning to the exit enclosure using a traditional split-system architecture. This is due to NFPA 101 18.104.22.168.1(10) not permitting refrigerant lines to enter stairshfts, the Handbook commentary stating that 22.214.171.124.1(10) is intended to be a comprehensive list of permitted exit enclosure penetrations, and IMC 1109.2.3 prohibiting refrigerant lines from entering exit enclosures (exit passageways and exit stairs).
This, however, has the effect of forcing the air handler for the exit enclosure to be moved outside the exit enclosure, or to be migrated into a unitary system such as a through-the-wall or rooftop terminal type unit. However, through-the-wall and packaged terminal units struggle with unwanted air leaks between the unit chassis and the wall sleeve, and RTUs have efficiency issues as well due to the entire unit living outside, forcing ducts to penetrate walls into exterior space. Furthermore, there may not be adjacent indoor spaces available to house said air handler if it is desired to place it indoors. Even if such spaces are available, that requires fire dampers on the duct penetrations thru the stairshaft as well as shaft-rated construction around the ducts and the air handler itself, which makes maintenance access to said air handler much more difficult.
This leaves me wondering: is it the intent of the model Codes to prohibit exit enclosures from being served by direct-expansion air handlers (split system indoor units) located within them? It seems to me that the combination of a small ductless mini-split heat pump and a distributed-type HRV or ERV would provide an inexpensive way to fully service an exit enclosure in a fashion that's compatible with a low-load, high-efficiency building design, yet on their face, the IMC's refrigeration rules seem to apply to all split refrigeration systems, from a one-ton minisplit heat pump to a massive supermarket's remote chiller system, despite the vast difference in risk between the two. Furthermore, hydronic and steam lines can leak too; worse yet, some heating systems that are allowed, such as one-pipe steam radiators, are more prone to fluids leakage during common malfunction conditions (any sort of stuck radiator vent on a one-pipe system will cause the radiator in question to spit cruddy water on a regular basis) than modern direct-expansion split systems ever are.