This is an EnergyPlus warning (Line 110). It's pretty clear: although both (paired) interzone surfaces have equal areas (~11.8 m2), the multiplier (20) has only been applied to "Surface=SURFACE 287 Zone=2F-BED-1" (20x ~11.8 m2 = ~236 m2) ... ~11.8 vs 236 m2 - not good. Either avoid multipliers when dealing with interzone surfaces, or ensure that the multiplier is equally applied to both (paired) interzone surfaces.

I echo @Aaron Boranian 's question: does this stem from a zone or floor multiplier? e.g.:

• a 22-storey building with 20x the same similar middle storey?
• 20 apartment units side-by-side, yet only 1 is modelled?

If so, I strongly suggest avoiding interzone surfaces. With multipliers, you're better off with adiabatic boundary conditions, or self-referencing conditions, e.g.:

• middle storey case: have a space's floor as boundary condition for its own ceiling (and vice-versa)
• side-by-side units (e.g. east-to-west layout): have a space's east wall as boundary condition for its own west wall (and vice-versa)

EDIT: The floor and ceiling surfaces illustrated here (in pale green) would normally be adjacent to ceiling or floor surfaces in other zones (e.g. ground floor, top floor). Yet when dealing with zone multipliers (as a proxy for a whole building story), it is safer to avoid interzone surfaces (linking other surfaces in other zones), as suggested in EnergyPlus' IO Reference:

Surfaces in Middle Zones – Middle zones in a building can be simulated using a judicious use of surfaces and zone multipliers to effect the correct “loads” for the building. Thus, middle zone behavior can be simulated without modeling the adjacent zones. This is done by specifying a surface within the zone. For example, a middle floor zone can be modeled by making the floor the Outside Boundary Condition Object for the ceiling, and the ceiling the Outside Boundary Condition Object for the floor.

2nd EDIT: One thing to look out for when setting floors/ceilings to adiabatic (vs interzone) is SolarDistribution. Unless this parameter is set to one of the 2 "WithReflections" options, all "beam solar radiation entering the zone is assumed to fall on the floor, where it is absorbed according to the floor’s solar absorptance". Depending on fenestration layout, internal loads, cooling/ventilation strategies, internal mass settings, etc., trapping beam solar radiation in an adiabatic floor construction (as this energy can't be transferred to the zone below) may indeed tilt cooling vs heating profiles. For the middle storeys, you have the option of setting aside adiabatic boundary conditions, and instead setting the middle storey floor as facing its own ceiling, and vice versa (as described above). This may alleviate the discrepancy. Otherwise, you have variables you can tweak, such as adjusting the inside floor material solar absorptance (a lower value redistributes in a zone a greater portion of beam radiation as diffuse). And in all cases, I would suggest paying attention to InternalMass definitions in each zone, which ideally should include all mass not explicitly-modelled as surfaces ... (more)