Your Ducts Are Bleeding $320 a Year. Your Contractor Never Did the Math.

Pull a random new-construction home's mechanical plans. Flip to the HVAC sheet. Odds are strong you'll find a rectangle labeled "furnace," some lines scratched toward rectangles labeled "supply," and not a single duct size, friction rate, or CFM value on the page.

That's a duct system designed by vibes.

ACCA publishes Manual D, the engineering standard for sizing residential ductwork. It requires calculating available static pressure, total effective length, friction rate per run, and CFM delivery to each room. Acceptable friction rates fall between 0.06 and 0.18. Every run gets sized. Every fitting's equivalent length gets counted. It's not complicated math, but it is math, and most residential HVAC contractors don't do it.

Instead, they use rules of thumb. Six-inch round for bedrooms. Eight-inch for the living room. Flex duct because it's cheap and bends around anything. Trunk lines become "whatever fits in the chase." Returns get one, maybe two grilles for the whole house, undersized because nobody measured the system's total CFM requirement against the return path.

Predictable result. ENERGY STAR reports that 20 to 30 percent of conditioned air in a typical home never reaches the rooms it's supposed to heat or cool. It leaks through unsealed joints, escapes into unconditioned attics, or gets strangled by undersized runs and excessive fittings that create pressure drops nobody calculated because nobody ran Manual D.

Leaks vs. Bad Design: Two Different Problems

Modern energy codes catch one of these. California's Title 24, the IRC 2021, and ENERGY STAR V3 all require duct blaster testing in new homes. Under ENERGY STAR V3: total leakage no more than 8 CFM25 per 100 square feet of conditioned floor area, with leakage to outdoors capped at 4 CFM25 per 100 square feet.

A duct blaster finds holes. It pressurizes the system to 25 pascals and measures how much air escapes. If you sealed every joint with mastic, your system passes.

But passing a leakage test doesn't mean the ductwork is designed correctly. A duct system can be airtight and still deliver 40% less airflow to the master bedroom because the contractor ran 35 feet of flex duct with four 90-degree bends, creating an effective length that chokes the supply down to half its rated CFM. That bedroom is now chronically warm in summer and cold in winter. The thermostat, located in the hallway that gets plenty of air, reads 72°F and tells the system to shut off.

Testing solves the sealing problem. Only Manual D solves the distribution problem.

The BTU Math Nobody Shows You

The Pacific Northwest National Laboratory's Building America program publishes the loss equation: every cubic foot per minute of leaked air carries 30 BTU/h with it.

Take a 3-ton system for a 2,500-square-foot California home. It moves 1,200 CFM and delivers 36,000 BTU/h of cooling. At the ENERGY STAR "typical" 25% leakage rate, you're losing 300 CFM. That's 9,000 BTU/h gone into your attic. Three-quarters of a ton of cooling, vanishing before it touches a room.

At 25% leakage, the compressor runs roughly 33% longer to compensate for the lost capacity. Your 3-ton system behaves like a 2.25-ton system trying to cool a house that needs 3 tons. It gets there eventually, cycling harder and longer, burning electricity the whole time.

The Dollar Damage Over Ten Years

Average American households spend $1,020 on heating alone, per the EIA. Add cooling, and combined heating plus air conditioning averages about $1,346 per year nationally. California skews higher due to cooling loads and electricity rates; figure $1,600 for a 2,500-square-foot home in the Central Valley or Bay Area.

Run the numbers:

The delta: $320 per year. Over 10 years, $3,200. Over the life of the system, $6,400. A Manual D duct design from a third-party HVAC engineer runs $300 to $800, depending on home size and complexity. That's a payback period under three years, and the savings compound every year after.

Where AI Enters the Picture

Manual D has been around since the 1990s. Contractors know it exists. They skip it because it takes time. Measuring every duct run, counting fittings, looking up equivalent lengths in tables, and iterating until friction rates balance across all branches can eat an afternoon. On a $15,000 residential HVAC install, that afternoon isn't in the bid.

Conduit Tech, acquired by ServiceTitan in 2025, uses a phone's LiDAR sensor to scan a home and auto-generate HVAC load calculations. Wrightsoft's RightSuite has offered integrated Manual J/D/S software for years, producing duct layouts from floor plan inputs. Carrier's HAP includes duct design modules. Plandroid uses intelligent duct placement algorithms.

Newer tools pull from phone scans and floor plan imports instead of manual measurements. What took four hours in 2015 takes 45 minutes now. Conduit's pitch is that a technician with an iPhone can generate a full load calculation during a single site visit. Extend that to duct design and you've removed the time objection that kept contractors from doing Manual D in the first place.

Computational fluid dynamics (CFD) is the next frontier. Autodesk's CFD module ($7,900/year) simulates airflow through duct systems, identifying pressure imbalances and turbulence at fittings that simplified equivalent-length tables miss. That's overkill for most residential work today. But as building models get richer and software gets cheaper, someone will package CFD-lite into a residential HVAC tool that optimizes duct routing the way structural software optimizes beam sizes. No contractor will run a simulation. The software will run it in the background and spit out a layout where every room gets its rated CFM with the shortest total duct run and fewest fittings.

Why This Matters More Than Equipment Upgrades

Homeowners spend $5,000 to $12,000 replacing an HVAC system. They shop SEER ratings obsessively. A SEER 20 unit costs $3,000 more than a SEER 15. At the condenser, efficiency gains run about 25%.

But if 25% of your conditioned air leaks into the attic before it reaches a room, you just erased that upgrade. You paid $3,000 extra for a high-efficiency condenser to push air through ductwork that wastes exactly as much as the old unit did. Your bottleneck was never the equipment. It was the pipes.

A January 2026 press release from MaxSeal, a duct sealing company, claimed that 95% of San Jose homes they inspected had duct leakage issues, with many losing 25 to 40% of conditioned air. Take the specific number with a grain of salt given the source, but the directional finding aligns with ENERGY STAR's national data.

The Best Argument Against This Whole Piece

If you're building new and your jurisdiction enforces duct testing (many do), the worst-case leakage scenarios described above shouldn't apply to your home. A competent installer who knows the duct blaster is coming will seal joints properly, and the 20-30% leakage figures from ENERGY STAR reflect the existing housing stock, not new construction compliant with 2021 IRC or Title 24.

Fair. But the duct blaster tests sealing, not design. A code-compliant new home can still have undersized returns, supply runs that are too long, imbalanced branches delivering wildly different CFM to different rooms, and static pressure high enough to shorten equipment life. Those are design flaws. They don't show up on a leakage test. They show up as the bedroom that's always 4 degrees warmer than the thermostat setting, the bathroom exhaust fan that barely pulls, and the equipment replacement you need at year 12 instead of year 18 because high static pressure burned out the blower motor early.

Duct testing is necessary. It's not sufficient.

What I Didn't Prove

My $320/year savings figure uses a simplified linear model: 25% leakage costs 25% more energy. Real HVAC systems don't scale linearly because compressors, fan motors, and heat exchangers have efficiency curves that vary with load and runtime. The actual savings depend on climate zone, insulation levels, equipment efficiency, and how many hours the system runs. In mild climates the savings will be smaller. In Phoenix, larger.

No national survey measures what percentage of residential HVAC contractors actually perform Manual D calculations. The claim that "most skip it" is well-supported by industry commentary and the sheer volume of badly designed residential duct systems that HVAC forums, energy auditors, and home inspectors document. But I can't cite a study that says "X% skip Manual D" because nobody has published one.

Conduit Tech's AI capabilities and time savings are self-reported. No independent benchmark has compared AI-assisted duct designs against manually engineered Manual D designs for accuracy, code compliance, or real-world comfort outcomes. The tools exist. The evidence that they produce better duct systems than a good technician with a pencil and the ACCA tables is not yet available.

Finally, my energy cost data uses national and California averages. Your home's actual HVAC costs depend on dozens of variables I didn't model, from window orientation to thermostat setpoints. The directional math is sound. The specific dollar figures are estimates, not predictions.