That sequence prevents a common mistake: increasing tonnage because a distant room is warm while leaving roof gain, solar glass, duct leakage, insulation, and static pressure unmeasured. The correct framework is Manual J for the load, Manual S for the equipment combination, and Manual D plus field verification for delivery. Square footage can frame an early comparison, but it cannot close any of those steps.

Trace the load before choosing a desert tonnage

The NOAA/NCEI 1991–2020 monthly normals for E Mesa station USC00022782 report a July normal maximum of 105.9°F and a normal minimum of 78.9°F. January normals are much milder at 67.5°F and 41.0°F. Those values establish an intensely cooling-dominant climate context; they are monthly averages, not Manual J design temperatures, and they cannot be converted into a required capacity.

The load calculation still belongs to the property. It must identify the roof assembly, ceiling insulation, exterior-wall construction, window area and orientation, shade, infiltration, internal gains, thermostat assumptions, and ducts outside the conditioned enclosure. A broad Arizona AC sizing guide can explain statewide principles, but the Mesa house needs its own inputs. Two nearby homes of equal area can have different afternoon room loads because their west glass, roof absorptance, and attic duct routes differ.

Sensible gain is likely to dominate the peak discussion, yet latent load does not disappear in a dry climate. Occupants, ventilation, infiltration during seasonal moisture events, and indoor moisture sources remain part of the calculation. The designer should retain the sensible and latent components rather than replace them with a desert label.

Separate equipment output from attic delivery

A warm west room can result from a high room load, inadequate equipment output at the selected condition, or a delivery shortfall. Those mechanisms require different corrections. More nominal capacity may help only the first two, and it can make the third harder by demanding additional airflow from an already restricted path.

Condition-specific equipment capacity is measured at the coil; delivered room capacity is affected by the blower operating point, duct leakage, conductive gain through duct insulation, fittings, branch resistance, and register performance. A long attic run may supply less useful cooling than the design expects even while the system satisfies a thermostat near the central return. That is why equipment performance and distribution losses must appear as separate lines of evidence.

A useful diagnosis compares room temperatures and load targets with total external static pressure, duct leakage information, and branch airflow. It does not infer “undersized” from runtime alone. Long runtime can be normal and efficient during a demanding hour; persistent room imbalance points toward the load map and air path.

A hypothetical west-facing tract home with long attic runs

Consider a hypothetical Mesa case, not evidence about citywide housing stock: an approximately 1,850-square-foot, 1990s single-story stucco tract home. It has a vented attic with long flex-duct runs, west-facing great-room glass, one central return, and mixed exterior shading. An aging single-stage split system is being compared with a variable-capacity replacement and a larger single-stage proposal.

The question is not whether one proposal is a nominal size larger. First, the west great room needs a room load that reflects its glass, shade schedule, roof exposure, and infiltration. Next, the proposed outdoor unit, indoor coil, and blower must be evaluated as a matched combination. Finally, the attic branch must be checked for length, compression, fittings, leakage, insulation, and available pressure. The owner should not pay for additional compressor capacity until those three records agree about the source of the discomfort.

The published 2,000-square-foot sizing comparison is a useful warning against treating floor area as a final answer, even though this illustration is somewhat smaller. It supplies comparison context; it does not assign a load to this hypothetical house.

Put roof, glass, and duct location into Manual J

ACCA Manual J organizes the building inputs into room-by-room sensible and latent loads. In the great room, glass orientation and external shading should be entered as observed, not averaged across the house. The ceiling and roof need their actual assembly assumptions. If supply ducts cross a vented attic, their location, insulation, leakage treatment, and surface area belong in the model instead of being hidden in a generous safety factor.

Manual J is also where the team resolves scope. Roof or window work that is complete or contractually included can be modeled; a possible future upgrade should not be credited without documentation. Indoor set points, occupants, appliances, ventilation, and infiltration require defined assumptions. The report should show them so a reviewer can distinguish a conservative input from an accidental exaggeration.

The output is not merely a whole-house total. Room loads provide delivery targets for the great room and the distant bedrooms. If their calculated shares differ from the existing airflow pattern, the mismatch becomes a design problem for the ducts rather than a reason to round up the outdoor unit.

Read the high-ambient rows in Manual S

Matched output at the selected outdoor condition

ACCA Manual S begins after the load is stable. The comparison must identify the exact outdoor unit, indoor coil, blower, and approved operating setup. Manufacturer expanded performance data—not the model-name tonnage—show total and sensible capacity at the selected outdoor and indoor conditions. In Mesa, the high-ambient row can matter materially, but the condition itself must come from an approved project weather source rather than the monthly normal cited above.

Airflow changes the sensible/latent split and coil behavior, so a proposal should not quote a capacity row that assumes an airflow the existing ducts cannot support. A variable-capacity system may offer a wider operating range and better part-load matching, while a single-stage unit may have a narrower choice. Neither product type is automatically correct. The evidence is the calculated load beside the matched combination’s performance table and operating range.

Reject undocumented substitutions. Changing the coil or air handler can change certified performance, pressure drop, and airflow requirements even when the outdoor-unit label looks similar. The final submittal should keep every component and the selected data row traceable.

Use Manual D to test static pressure and duct leakage

Critical attic path: leakage, insulation, and static pressure

ACCA Manual D converts room loads and blower capability into a distribution design. The filter, coil, grilles, trunk, fittings, flex branches, and registers all consume available pressure. The longest run to the west room is a critical path; nominal duct diameter alone does not describe its resistance.

The U.S. Department of Energy’s guidance on sealed and insulated ducts explains why leakage and conductive transfer outside conditioned space deserve explicit attention. Field inspection should look for compressed or sagging flex, sharp bends, disconnected sections, damaged insulation, and accessible sealing defects. Leakage testing and measured static pressure can then show whether the installed path resembles the design assumptions.

Balancing is not a license to choke easy branches until noise and pressure rise. The design must allocate air to the room-load targets while staying within the blower’s verified operating range. Where the path cannot deliver, a branch, return, register, or duct-location change may be more valuable than larger equipment.

Document the permit and commissioning evidence

The City of Mesa residential construction guidance identifies the local permit, mechanical-plan, code, and inspection pathway. That administrative record does not calculate the load. It should sit beside—not replace—the Manual J report, Manual S selection, and Manual D design.

Startup should document the exact installed components, refrigerant-charge procedure, blower setting, total external static pressure, relevant pressure drops, and room delivery checks. ENERGY STAR HVAC Quality Installation guidance reinforces that sizing, installation, airflow, and commissioning work together. A high-efficiency label cannot compensate for an undocumented air path.

The Mesa decision closes only when three proofs line up: a calculated building and room load, matched-system output at the approved condition, and measured delivery at the distant register. If the west room still lacks capacity, the records should reveal whether the next action belongs to the envelope, equipment selection, or attic distribution. That is a defensible answer; an extra half-ton chosen from a warm-room complaint is not.