Air Conditioners: A Reference Built on Primary Sources

What an Air Conditioner Actually Does

An air conditioner is a vapor-compression heat pump operating in cooling mode. A refrigerant loop carries heat from indoors (where the evaporator coil sits) to outdoors (where the condenser coil sits), moved by a compressor that elevates the refrigerant's pressure and temperature between the two coils.^[9] The same hardware that defines a heat pump in cooling mode is, with the four-way reversing valve removed, what a dedicated air conditioner is.

The cooling load any AC must handle is the rate at which heat enters the conditioned space. Heat enters through three pathways: conduction through walls, ceilings, floors, and glass; air infiltration through cracks and openings; and internal generation from people, lights, appliances, and solar gain.^[5] A correct AC sizing answers one question — what is the peak rate of heat gain at the design outdoor condition — and provides equipment capacity that matches it.

How AC Capacity and Efficiency Are Rated

AHRI Standard 210/240-2023 defines the test conditions for residential AC capacity and efficiency.^[1] Cooling capacity is measured at 95°F outdoor dry bulb and 80°F indoor dry bulb (67°F indoor wet bulb to fix latent conditions). That single point produces the nameplate "tons" or BTU/hr figure shown on the equipment label.

Two efficiency metrics matter, and they measure different things. EER2 is the steady-state cooling-per-watt at the 95°F design point — a number that lets you compare equipment at peak load. SEER2 is the seasonal weighted average across many outdoor temperatures, capturing how the unit cycles and modulates across a typical cooling season.^[1]

A high SEER2 with low EER2 indicates an inverter unit that performs well at part load but is less efficient at design; a high EER2 with lower SEER2 (rare in current production) indicates the opposite.

The SEER → SEER2 number conversion matters for anyone shopping based on older labels. The new test procedure raised assumed external static pressure from 0.10 in. wc to 0.50 in. wc, which is a more realistic value for installed ducted systems and which lowers measured efficiency by roughly 4-5%.^[1]

A SEER 16 unit under the old rules is roughly SEER2 15.2 under the new rules. Manufacturers republish all current production under the new rating, and the AHRI Certified Reference Number (ARN) on the equipment label maps to a public verified row in the AHRI directory.

Sensible Versus Latent Cooling (and Why It Matters in Humid Climates)

Cooling work splits into two categories. Sensible cooling lowers air temperature — what a thermometer reads. Latent cooling removes water vapor from the air by condensing it on the cold evaporator coil. Both consume cooling capacity, and the split between them depends on the indoor humidity ratio at the entering air.^[7]

The ratio is called sensible heat ratio (SHR): SHR = sensible cooling / total cooling. In a dry climate at design conditions, SHR can be 0.85-0.95 (most of the work is dropping temperature). In a humid climate, SHR drops to 0.65-0.75 (one-quarter to one-third of the cooling work is removing water vapor).^[6]

The capacity-rating implication is concrete. Manufacturer spec sheets publish "total" and "sensible" capacity separately at AHRI conditions; latent capacity is the difference.

A 3-ton (36,000 BTU/hr) unit might publish 27,000 BTU/hr sensible and 9,000 BTU/hr latent at the 80°F / 67°F WB rating point, meaning SHR = 0.75. At a hotter, drier indoor condition (say 78°F / 60°F WB), the same unit might shift to SHR 0.85 with almost no latent capacity.

This is why Manual S equipment selection compares both sensible AND latent capacities against the Manual J loads, not just total tonnage.^[6]

BTU per Square Foot by Climate (and Why Rule of Thumb Misleads)

The most common AC sizing question is "how many BTU per square foot do I need?" The answer is not a single number — it is a range that depends on climate, envelope, ceiling height, sun exposure, and how the space is used.

These are planning numbers, not Manual J replacements. The same 2,000 sq ft home in zone 4 can vary from about 24,000 BTU/hr (a tight 2018 build with R-49 attic insulation and triple-pane windows) to about 48,000 BTU/hr (a 1965 ranch with R-13 attic, single-pane windows, and 10 ACH50 air leakage) — the same square footage, the same climate, double the cooling load.^[5]

The variables that move the load most strongly:

• Window area and U-factor — a south-facing wall of single-pane glass can add 15-25% to total load.
• Attic insulation R-value — R-49 versus R-13 attic shifts total cooling load by 8-15%.
• Infiltration measured in ACH50 — 3 ACH50 versus 10 ACH50 shifts load by 10-20%.

The AC BTU chart article walks through these adjustment factors with worked examples; the Manual J load calculator takes the inputs and produces a planning estimate.

The Four Residential AC Types and How to Pick

Central split-systems dominate US residential cooling because they share ductwork with the heating system. The outdoor condenser unit sits on a concrete pad or roof; the indoor evaporator coil sits above the furnace or air handler; refrigerant lines connect them through a copper line-set. Efficiency varies widely across the SEER2 13.4-22.0 range, and ductwork condition matters: a leaky duct system loses 20-30% of any AC's output to unconditioned space regardless of nameplate efficiency.^[9]

Packaged units combine the condenser, compressor, and evaporator into one outdoor cabinet, with conditioned air ducted into the house through a single supply and return penetration. They are common in manufactured housing, in low-ceiling spaces where there is no room for an indoor air handler, and in rooftop commercial applications.^[9] Efficiency is typically lower than split-systems because the design must fit one cabinet.

Ductless mini-splits put a single indoor head on a wall (or ceiling) connected to an outdoor compressor by refrigerant lines. They are the dominant solution in houses without ductwork (most pre-1960 Northeast housing stock) and in additions where extending the duct system is impractical. ENERGY STAR mini-splits typically achieve SEER2 18-22, well above central split-system equivalents at the same capacity.

Window and portable units are room solutions, not whole-house solutions. A window unit fits a standard double-hung opening and serves a single room of up to about 600 sq ft. Portable units sit on the floor and exhaust hot air through a flexible duct to a window vent; their efficiency is roughly 20-30% lower than window units of equivalent capacity because some of the exhaust air pulls already-cooled room air out of the house.^[4]

Why Sizing Errors Cost Real Money

AC sizing errors cost in three ways: higher equipment cost (oversized), higher operating cost (undersized), and worse comfort (both). The cost magnitudes are not equal — oversizing penalties are typically larger than undersizing penalties in cooling-dominated climates, while undersizing penalties dominate in extreme heat.

Oversizing penalty. A 30%-oversized AC reaches the thermostat setpoint quickly and shuts off, then restarts a few minutes later when temperature drifts up by a degree or two. That short-cycle pattern produces three failures simultaneously: humidity is not removed (because runtime is too short), the compressor wears out from frequent starts (rated for a finite cycle count), and indoor temperature swings 2-3°F between cycles instead of holding steady.^[6] A correctly sized AC runs longer continuous cycles, removes moisture properly, and holds temperature within 1°F.

Undersizing penalty. An undersized AC runs continuously on hot days but cannot pull indoor temperature down to setpoint. The compressor runs at 100% for 8-12 hours straight, then the indoor temperature drifts up another degree by mid-afternoon and stays there until evening cooling. Humidity is typically fine (extended runtime favors latent removal), but sensible cooling fails on the hottest 1-3% of cooling hours. The fix is bigger equipment, which is rarely a small expense.

Costing it out at the US average. A 4-ton AC running 1,500 cooling-season hours at SEER2 14 consumes about 4,286 kWh per year, costing about $699 at the 2024 US average residential rate of $0.163/kWh.^[10] The same load handled by a SEER2 18 unit consumes about 3,333 kWh per year, costing about $543. The $156 annual saving over 15 years totals $2,340 — usually more than the SEER2 18 price premium.

The Operating Cost of Running the AC

Operating cost depends on cooling load, equipment efficiency, hours of operation, and local electricity price. Equipment efficiency and load are fixed once installed; runtime and price drive year-to-year variation.

The same equipment in the same operating condition can run from $65 to $713 per year depending entirely on climate and local electricity price. Phoenix runs the AC year-round and pays moderate rates; Seattle barely uses cooling but its few hours run on cheap hydropower-sourced electricity; Boston has a short cooling season but pays Massachusetts retail electricity prices that are roughly double the national average.^[10]

The variable-speed equipment efficiency premium pays back fastest in long-runtime climates. A SEER2 14.3 to SEER2 20 upgrade in Phoenix saves roughly 1,500 kWh per year (~$200), paying back a $1,500 efficiency premium in 7-8 years.

The same upgrade in Seattle saves about 170 kWh per year (~$20), which never pays back over the equipment's useful life. Match efficiency tier to runtime, not to the salesperson's profit margin on the high-tier unit.

Common Problems and What They Actually Mean

Three patterns account for the majority of AC service calls: short cycling, weak airflow, and high humidity at setpoint. Each has a small set of likely root causes.

Short cycling (runs under 10 minutes, off 5-15 minutes, repeat). Most common cause is oversizing combined with low cooling load conditions — the unit hits setpoint fast and shuts off. Second most common is low refrigerant charge causing the low-pressure switch to trip the compressor. Third is dirty evaporator coil restricting airflow and triggering high-pressure trip. The AC short cycling article walks through the full diagnostic sequence.

Weak airflow at registers. Usually a filter or coil problem, not a compressor problem. Replace filter first (clogged filters can reduce airflow by 30-50%). Check evaporator coil for biofilm or dust accumulation. Check for closed dampers, kinked flex duct in the attic, or collapsed return-side ducts. If airflow is uniform across all registers but low overall, the blower may be wrong-sized for the duct system's static pressure.

High humidity at setpoint (cool but sticky). Almost always oversizing in a humid climate. The AC removes sensible heat faster than latent heat at design conditions; an oversized unit hits the thermostat setpoint before it has been running long enough to condense significant moisture out of the air. Solutions in order of cost: raise the thermostat 2°F (longer cycles, more latent removal), add a whole-house dehumidifier ($1,500-$3,000 install), or replace with right-sized variable-speed equipment that can run continuously at low capacity for dehumidification.

The 2025 R-410A Phaseout

The EPA AIM Act final rule banned the manufacture of new residential AC equipment using R-410A refrigerant after January 1, 2025.^[12] R-410A had been the residential standard since the early 2000s phase-out of R-22 ("Freon"). The reason is global warming potential: R-410A has GWP 2,088, meaning each pound of leaked refrigerant warms the atmosphere as much as 2,088 lb of CO2.

Two replacement refrigerants dominate. R-454B (commercial name Opteon XL41) has GWP 466 — about 78% lower than R-410A. R-32 has GWP 675 — about 68% lower than R-410A.^[12] Both are A2L-classified (mildly flammable, lower-toxicity) under ASHRAE Standard 34, which is a step up from R-410A's A1 (non-flammable) classification but a step down from older R-22's higher-toxicity rating.

Technician implications. A2L refrigerants require nitrogen purge during brazing to prevent ignition of any residual refrigerant, larger outdoor clearance distances to windows and openings, and slightly different line-set sizing rules. None of this is hard or dangerous — the work is well understood from a decade of R-32 use in Asia and Europe — but it does add a few materials cost and a small labor premium that shows up in 2026 install quotes.

What This Cluster Covers

The cluster organizes AC content into three functional areas.

Sizing references

• AC BTU chart by square footage — reference table with climate, ceiling height, insulation, and space-type adjustments
• Mini-split for garage sizing — garage-specific load math (typically 2-3× the BTU per sq ft of interior space)

Troubleshooting

• AC short cycling — diagnostic sequence for the most common service call pattern

Related load and equipment topics

• Heat pump reference — combined heating and cooling from one piece of equipment
• Manual J load calculation — the underlying load methodology
• Manual S equipment selection — capacity matching rules and Manual S tolerances
• Building science fundamentals — envelope drivers (R-value, U-factor, infiltration, climate zones) that move cooling load

Calculators

• AC size calculator — AC-specific sizing with window, portable, mini-split, and central recommendations
• BTU calculator — general BTU sizing for a room or whole house
• Manual J load calculator — full envelope load math approximating Manual J for planning purposes