What Size Air Conditioner Do I Need? (Manual J vs. Square-Foot Rules)

A homeowner in Coral Canyon called us last August. Her contractor had walked the property for eleven minutes, eyeballed the floor plan, and quoted a 5-ton replacement for her 2,300 square foot two-story. No measurement of windows. No inspection of the attic insulation. No duct leakage test. Just a confident “you need a 5-ton” and a $14,200 number scrawled on a clipboard.

We ran a Manual J the next morning. Her actual cooling load was 30,800 BTU/hr: a 2.5-ton system, maybe 3 tons if we built in margin for the western exposure on her great room. The 5-ton would have short-cycled itself into an early grave, left her humid during monsoon storms, and cost her an extra $3,500 upfront for the privilege. This is not a rare story. It is the default outcome when a contractor sells equipment instead of engineering a system.

“Bigger is better” is the most expensive myth in residential HVAC. Oversized air conditioners run in short violent bursts that cool the air without running long enough to dehumidify it, leaving a clammy 76 degrees that the thermostat reads as “satisfied.” They wear out compressors years early. They draw enormous inrush current every time they start. And in Southern Utah, where dust loading on coils and 108-degree design days already punish equipment, oversizing accelerates every failure mode we see in the shop.

The basics: tonnage and BTUs

Air conditioning capacity is measured in BTU per hour, or British Thermal Units, the energy needed to raise one pound of water one degree Fahrenheit. One ton of cooling equals 12,000 BTU/hr. The name comes from the era when a “ton” of ice meant exactly that much daily cooling capacity.

Residential systems in Washington County run from 1.5 tons (18,000 BTU/hr) up to about 5 tons (60,000 BTU/hr). Anything bigger gets split into multiple systems or zoned VRF. The capacity stamped on the condenser nameplate is the AHRI rated capacity at standard conditions: 95 degrees outdoor, 80 degrees indoor return, 50 percent indoor humidity. At St. George’s actual 108-degree design day, that same condenser delivers about 8 to 12 percent less. Real-world capacity is always lower than nameplate. Sizing has to account for it.

The load on your home is also measured in BTU/hr, broken into sensible load (heat that changes air temperature) and latent load (heat tied up in moisture). The job of a load calculation is to add up every BTU of heat entering your home on the worst day of the year, then pick equipment that matches it. Not exceeds it. Matches it.

Why the square-foot rule fails

You have probably heard the rule of thumb: “600 square feet per ton” or “20 BTU per square foot.” It is wrong. Not “a rough approximation” wrong. It is fundamentally wrong, because it treats every home as identical when no two homes are.

A 2,000 square foot home in 2026 Desert Color with R-49 attic insulation, 2x6 walls, Low-E argon-filled windows, and a HERS-rated tight envelope might have a 24,000 BTU/hr load, about 2 tons. That same 2,000 square feet built in 1985 Bloomington, with R-19 attic insulation, single-pane aluminum-frame sliders facing west, and 12 ACH50 air leakage, can hit 42,000 BTU/hr. Same square footage. Nearly double the load.

The rule of thumb ignores everything that actually matters:

• Insulation R-values in the attic, walls, and floors
• Window area and orientation: west-facing glass gains four times the heat of north-facing
• Window quality: single-pane, double-pane, Low-E coating, SHGC rating
• Infiltration rate: how much outside air leaks through gaps
• Ceiling height: a 12-foot vaulted ceiling has 33 percent more volume than 9-foot
• Occupants and internal gains: people, lights, ovens, computers all produce heat
• Duct location and leakage: attic ducts in 140-degree summer attics are a load source
• Shading: neighboring homes, trees, patio covers all change solar gain

A square-foot rule cannot see any of that. So it defaults to the safe-for-the-contractor side, which is always too big.

Manual J: how real load calcs work

Manual J is the ACCA-published residential load calculation protocol, the actual engineering standard, referenced in the International Residential Code. It is what every legitimate HVAC contractor uses and what every junk contractor skips.

Here is what we measure on a real Manual J:

1. Exterior wall area, by orientation. North walls. South walls. East walls. West walls. Each gets its own heat gain calculation based on construction type and solar exposure.
2. Window schedule. Every window, measured. Width times height. Frame type. Glass type. SHGC and U-factor pulled from the NFRC label or estimated from age and construction.
3. Ceiling and roof. R-value, color (dark roofs gain more), whether the attic is vented, ducts present.
4. Floor. Slab-on-grade, crawlspace, or above unconditioned space.
5. Infiltration. Either measured by blower door (ideal) or estimated from construction era and tightness category.
6. Internal gains. Number of occupants, kitchen appliances, lighting load.
7. Duct system losses. Where the ducts run (attic vs. conditioned space), insulation level, estimated leakage.
8. Design conditions. Outdoor 108 dry-bulb / 67 wet-bulb for St. George; indoor 75 dry-bulb / 50 percent RH.

That data goes into Wrightsoft Right-J, CoolCalc, Elite RHVAC, or a similar ACCA-approved program. The software outputs sensible load, latent load, total cooling load, and a recommended equipment range, typically a window of about 95 to 115 percent of total load. A real Manual J takes a competent technician 60 to 120 minutes per home. It is not a five-minute napkin sketch. Anyone telling you otherwise is selling something.

Southern Utah specifics

Our climate has its own personality, and equipment sizing has to respect it.

Design temperature for cooling in Washington County runs 105 to 108 degrees outdoor, depending on which weather station and which percentile you use (ASHRAE 1 percent vs. 0.4 percent). Indoor design is 75 degrees, sometimes 78 for extreme efficiency designs. The temperature difference at design, your delta-T, is around 30 degrees, which is large but predictable.

What gets overlooked: humidity is not zero here. From mid-July through early September the monsoon brings dewpoints into the 50s and occasionally 60s. An oversized AC that runs in 7-minute bursts cannot pull that moisture out. Homeowners report “the AC is set to 74 but it feels muggy.” That is a sizing problem, not a thermostat problem. Right-sized equipment runs longer cycles, often 25 to 40 minutes, and dehumidifies as it cools.

The other Southern Utah factor: dust loading. Our outdoor coils accumulate dust and pollen at 3 to 5 times the rate of a coastal climate. An oversized system that short-cycles also cycles its outdoor fan more, pulling more dust onto the coil per hour of cooling delivered. We see oversized 5-ton condensers in St. George with coils so plugged at year four that capacity has dropped 20 percent. A right-sized 3-ton in the same neighborhood runs cleaner because it runs steadier.

Common sizes by Washington County home archetype

Real numbers from real Manual Js we have run in the last twelve months:

1,200 sq ft St. George ranch (1990s, single-story, R-30 attic, double-pane vinyl windows, attic ducts): Total load 21,400 BTU/hr. Equipment selection: 2-ton 14.3 SEER2 condenser. Many of these homes have an existing 3-ton unit that short-cycles all summer. Downsizing on replacement is common.

1,650 sq ft Washington City rambler (2008, R-38 attic, vinyl Low-E windows, conditioned-space ducts): Total load 24,800 BTU/hr. Equipment: 2-ton or 2.5-ton depending on how much margin we want for west-facing master bedroom. Two-stage 16 SEER2 is the sweet-spot here.

2,400 sq ft Coral Canyon two-story (2015, R-49 attic, 2x6 walls, Low-E argon windows, two zones): Total load 36,200 BTU/hr split across two zones. Typical solution: 3-ton single system with proper zoning, or 1.5-ton plus 2-ton dual-zone setup. Almost never the 4-ton or 5-ton these get spec’d at by builders.

3,100 sq ft Sienna Hills custom (2020, R-49 attic, foam-insulated walls, 12-foot great room ceiling): Total load 41,800 BTU/hr. Equipment: 3.5-ton variable-speed inverter system. The vaulted ceiling and west-facing glass push it just past 3 tons.

4,000 sq ft Ivins custom (2018, ICF walls, R-60 attic, deeply shaded courtyard): Total load 44,500 BTU/hr. Equipment: 4-ton variable-speed, often with a small zone for the master suite. Build quality matters more than square footage. A 4,000 sq ft tight house can have less load than a 2,500 sq ft leaky one.

The pattern: most Washington County homes need 0.5 to 1 ton less than the equipment that gets installed by default. Builders oversize. Replacement contractors match the old (oversized) unit. The cycle continues until someone runs an actual load calc.

What happens when you oversize

Oversized equipment is expensive in five different ways at once:

1. Short-cycling. The compressor runs for 5 to 10 minutes, satisfies the thermostat, shuts off, restarts 12 minutes later. Compressors are designed for 15 to 30 minute run cycles. Short-cycling kills them. We replace short-cycled compressors at year 6 to 8 routinely; right-sized systems run 12 to 15 years.
2. Comfort. Air feels cold but clammy. Temperature stratifies, cold near the floor and warm near the ceiling. Rooms farthest from the air handler never catch up.
3. Humidity. Latent removal requires runtime. Oversized = no runtime = monsoon-season stickiness even at 73 degrees.
4. Electric bills. Inrush current at startup is 5 to 7 times running current. More starts per hour means more inrush. Combined with poor part-load efficiency, bills run 15 to 25 percent higher than right-sized.
5. Equipment cost. A 4-ton condenser, coil, and air handler costs $1,500 to $2,500 more than a 3-ton, plus larger linesets, larger breakers, sometimes a service panel upgrade.

What happens when you undersize

Undersizing is rarer than oversizing, since most contractors err the other way, but it has its own failure mode. An undersized system runs continuously on the hottest days and never reaches setpoint. You set the thermostat to 75 and the house holds at 79. The compressor is fine; it is just outmatched.

The fix: target equipment selection at 100 to 115 percent of design load, never below 95 percent. A small margin lets you handle the rare 110-degree day without leaving capacity on the table the rest of the year.

Other factors that change sizing

Manual J is the start, not the finish. Three things commonly change the equipment selection after the basic load calc:

Duct leakage and Manual D. ACCA Manual D sizes the actual duct system to deliver the load. A 3-ton AC connected to leaky undersized ducts delivers maybe 2.4 tons to the rooms. We test static pressure with a manometer and run a duct leakage test before quoting any system replacement. If the ducts are 25 percent leaky in the attic, the conversation changes. Sometimes we seal the ducts and downsize the equipment, sometimes we replace both.

Window upgrades and attic insulation. A homeowner who is replacing single-pane sliders with Low-E argon at the same time as the AC has dropped their cooling load 10 to 15 percent. We model the new envelope, not the old one. Same for blowing R-19 attic up to R-49, which drops load measurably.

Additions and remodels. A 400 sq ft sunroom addition on a Bloomington home is often glass-heavy and west-facing, and it can add a full ton of load on its own. Treat additions as their own load calc, not a percentage of the existing house.

How to evaluate a quote that does not include Manual J

Red flags that should kill a quote before you sign:

• Contractor never measured windows, never went in the attic, never asked about insulation
• Quote arrives the same day from a brief phone call
• Sizing matches your existing equipment exactly with no questions asked
• “We always install 4-ton on this size house,” which is sizing by neighborhood, not by home
• No mention of duct condition, static pressure, or duct leakage test
• Refusal to share the Manual J printout when asked
• Recommended size jumps a half-ton or full ton between two contractors with no explanation

A good contractor will hand you a Manual J report, walk you through the line items, and explain why they chose 3 tons instead of 3.5. If they cannot explain the load number, they did not calculate it.

Get a real load calc before you buy

M&M Mechanical Inc. has been sizing systems in Washington County since 1992. Every replacement quote we write includes a Manual J load calculation, a static-pressure and duct-leakage check, and a real explanation of why the equipment we recommend matches your home. We do not sell oversized equipment. We do not match the old unit by default. We measure, we calculate, and we install systems that run long, quiet, and dry.

If you are sizing a replacement or a new build in St. George, Washington, or Ivins, call us at (435) 674-1275 for a free Manual J assessment. We will walk your house, run the numbers, and show you the report. Whether you ultimately go with us or someone else, you will at least know the right answer.

Learn more about our air conditioning services, new construction work, how SEER2 ratings affect your decision, and how to prevent frozen coils once your system is installed.