HVAC Efficiency and Refrigerant Standards Since 2023

You collect three quotes for the same piece of capital equipment. The spread between the highest and the lowest is wide enough to be uncomfortable, and nothing on the front page of any of them explains why.

The instinct is to assume somebody is overcharging or somebody is cutting corners. Occasionally that is true. Far more often all three documents are honest and none of them is quoting the same job, because the parts that differ are the parts nobody writes on the cover page.

This matters more than it used to, because two regulatory changes in the last three years have quietly added decisions to a replacement that did not previously exist.

A Quote Is a Stack, Not a Price

The useful mental shift is to stop reading a quote as a number and start reading it as a sequence of decisions, only the first of which is about the equipment itself.

• Capacity. What the building actually requires, which comes out of a calculation rather than out of the size of the thing being removed.
• Efficiency tier. A genuine choice made on top of capacity, with real running-cost consequences.
• The condition of everything it connects to. Whether existing connections can be reused, whether the electrical supply suits the new equipment, whether the distribution system can carry what the equipment produces, whether the installation meets current code.
• Commissioning. What gets measured and set on site after everything is physically in place.
• Compliance. Permit and inspection where applicable, and disposal of what was removed.

Two bidders can agree exactly on the first two and differ enormously on the third. One included the remedial work the building needs; the other quoted the box. Both numbers are real. Only one of them is a finished job.

Capacity Does Not Come From the Old Nameplate

The single most common shortcut is sizing the replacement by reading the data plate on the equipment being removed.

That assumes two things that are usually wrong. It assumes whoever specified the original equipment got it right, and it assumes nothing about the building has changed since.

In practice buildings change constantly and almost always in the same direction. Insulation gets added. Glazing gets replaced. Air leakage gets sealed. Each of those reduces the load, which means matching the old capacity installs an oversized system deliberately.

The residential standard for working out the real requirement is ACCA Manual J, which calculates room by room rather than treating a building as one box, with Manual S then matching available equipment to that result. One ton of cooling equals 12,000 BTU per hour, and whether a given building needs three tons or two and a half depends on glazing area and orientation, insulation as it stands today, ceiling height, infiltration and internal heat gains.

Oversizing is not a harmless overspend. An oversized system satisfies its control quickly and shuts off, which stops it running long enough to remove moisture from the air. The result is a space at the correct temperature that still feels wrong, and nobody connects that complaint back to the sizing decision made months earlier.

The 2023 Change: Efficiency Numbers Are Not Comparable Across Years

On 1 January 2023, SEER2 and HSPF2 replaced SEER and HSPF as the US efficiency test standard under a Department of Energy test procedure change.

The important detail is not that the numbers moved. It is that the new procedure tests equipment against a higher external static pressure than the old one, because the previous laboratory assumption about the resistance real installed distribution systems impose was too generous.

Two practical consequences follow.

First, comparing the rating on equipment installed before 2023 with the rating on equipment quoted today is not a like-for-like comparison. The figures were produced under different conditions.

Second, and more usefully, the change is a formal admission that the distribution system routinely matters more than ratings implied. If a bid includes remedial work on ductwork or distribution and a cheaper bid does not, the more expensive bid is addressing the thing the standards body revised its own test over.

The 2025 Change: Like-For-Like Often No Longer Exists

The second change is larger and newer.

New residential systems manufactured from 1 January 2025 use lower global warming potential refrigerants such as R-454B and R-32, under the EPA’s Technology Transitions Rule. Those refrigerants are classified A2L, mildly flammable under ASHRAE Standard 34, which changes service tooling, leak detection requirements and technician certification.

Separately, R-22 production and import in the United States ended on 1 January 2020 under the Clean Air Act phaseout, so anything still running R-22 is being serviced from reclaimed or stockpiled supply.

For a replacement bid this produces two concrete decisions that belong on the page.

The existing refrigerant line set is now a real decision rather than a detail. Existing pipework can sometimes be reused and sometimes cannot, depending on condition, sizing and what was previously in it. A bid that reuses it and a bid that replaces it are not quoting the same job, and that single line can account for a large share of the spread.

A like-for-like replacement frequently no longer exists as a model number. Equipment has to be selected against a calculation rather than matched to what was there, which is another reason the load calculation stops being optional.

The Cheaper Bid That Covers Less

There is one specific pattern worth knowing about because it produces the largest gaps and the most disappointed buyers.

In a split system, the outdoor unit and the indoor coil are rated together as a matched pair. Replacing only the outdoor unit and retaining the existing indoor coil is faster and materially cheaper. It is legitimate when the retained component is genuinely compatible and in good condition. It is not legitimate as a default.

Mismatch the pair and the efficiency rating on the box no longer describes what has been installed, delivered capacity can land below requirement, and the metering device may be wrong for the combination. Add the refrigerant transition above and the compatibility question gets sharper still.

If one bid is notably lower than the others, check whether the indoor equipment is in it. That one line explains most large gaps.

Commissioning Is the Part You Cannot See

Two installations of identical equipment can perform differently, and the difference is set on site in steps that leave no visible trace.

Before refrigerant goes in, air and moisture have to come out, and that is verified with a gauge rather than judged by how long a pump ran. Moisture left in a system combines with refrigerant and oil to form acid, which attacks the compressor from the inside. The damage takes years to appear and then looks like bad luck.

The charge is weighed in and verified against conditions on the day rather than estimated from pressures that “looked right”. The blower is set to deliver the airflow the equipment requires, which depends on the measured resistance of the distribution system rather than on a factory default.

The deliverable is a startup record showing charge, airflow and temperature split at handover. A contractor who commissions properly already produces one, so asking costs nothing. A contractor who tells you it is unnecessary has answered the question you were really asking.

Six Questions That Expose the Scope

Ask all six of every bidder. The answers sort the bids faster than the totals do.

• Did you perform a load calculation, and may I see it? A real one lists rooms, design conditions and assumptions.
• Is the matched indoor equipment included?
• Is the existing line set being reused or replaced, and on what basis?
• What static pressure did you measure, and does the bid include any remedial distribution work?
• Will I receive a startup record with charge, airflow and temperature split?
• What are the warranty terms in writing, and does registration affect them?

Two of those six, the load calculation and the static pressure reading, are measurements. A bidder who has neither has not assessed the building; they have priced a product.

Why the Lowest Number Is Not Automatically the Saving

A lower total covering less scope is not a discount. It is a deferral, and the deferred items tend to resurface as either a second invoice or a permanent performance shortfall that nobody traces back to the original decision.

That is not an argument for always choosing the highest bid. It is an argument for making the bids comparable before comparing them, which means getting the scope written down line by line, parts separate from labor, with the assumptions stated.

For an example of that stack set out in full, including the drivers ranked by how much each moves the total, the breakdown behind AC replacement cost in Springfield, MO works through the same structure applied to one category of equipment.

The Short Version

A capital equipment bid is a stack of decisions wearing a single number. Most of the spread between bids lives in the lines that are not on the cover page: the matched components, the existing connections, the remedial work and the commissioning.

Two standard changes since 2023 have added decisions rather than removed them, which means the old habit of matching what was there no longer reliably produces a correct answer. Ask for the calculation, ask for the measurement, and ask for the scope in writing. Then compare.