Your Green Home Scored a HERS 40. Its Embodied Carbon Is Identical to the Code-Minimum House Next Door.

A homeowner in Amherst, Massachusetts, did everything right: R-49 attic insulation, triple-pane windows, a variable-speed air-source heat pump sized to a proper Manual J calculation, and an envelope so tight it needed an ERV to keep the air breathable. Her HERS Index score came back at 42, meaning her house would use 58% less energy than a standard new home, and she framed the certificate and hung it in the mudroom next to a photo of the solar array on her garage roof.

Fifty-four tons of carbon dioxide equivalent had already entered the atmosphere before she plugged in her first lamp.

Her HERS rating measured none of it, because Ekotrope, REM/Rate, EnergyGauge, and every other tool her rater used track operational carbon exclusively: the energy consumed over decades of heating, cooling, and hot showers, converted to emissions via grid factors from NREL's Cambium dataset. What those platforms systematically ignore is embodied carbon, which is the emissions baked into her concrete foundation, the spray foam sealed inside her walls, the copper plumbing branching through her floors, and the heat pump compressor sitting on a concrete pad outside her kitchen window. For her house, that ignored portion represents roughly a third of all the carbon it will produce in its first quarter-century of existence.

A 100-Home Study That Broke an Illusion

In December 2025, a coalition led by the Northeast Home Energy Rating System Alliance, working with Ekotrope, Builders for Climate Action, and NMR Group, published what amounts to an indictment of how the residential building industry measures environmental performance. Funded by MassCEC, National Grid, and Eversource, the Massachusetts 100-Home Embodied Carbon Study is the first large-scale effort to measure both operational and embodied carbon in newly built homes, and its conclusions should unsettle anyone who assumed a high-performance envelope automatically meant a low-carbon building.

One hundred homes. Certified HERS raters collecting real construction data and feeding it through the BEAM (Building Emissions Accounting for Materials) tool, developed by Builders for Climate Action, against environmental product declarations and the draft RESNET/ICC 1550 Standard, which is the first embodied carbon standard ever written for low-rise residential buildings.

Collectively, those 100 homes produced 5,555 net tons of CO2 equivalent in embodied carbon emissions, averaging 55.5 tons per home before a single thermostat was ever set, and over a 25-year horizon, embodied carbon accounted for as much as 32% of a home's total carbon footprint. Extrapolate to Massachusetts' 11,390 new homes built in 2024, and you get 632,145 tons of embodied CO2e from a single year's construction in a single state.

But here is what should keep green builders awake at night: embodied carbon intensity showed no meaningful correlation with HERS Index scores or Carbon Index ratings, which means above-code homes performed no better, on average, at limiting embodied carbon than homes built to minimum code. In some cases, the additional materials required to achieve exceptional operational performance, the extra insulation layers, the more complex mechanical systems, the thicker foundation walls supporting heavier assemblies, actually increased upfront embodied emissions beyond what a simpler, less efficient home would have produced.

Where Carbon Actually Hides in Your Walls

Three material categories account for 68% of all embodied emissions across the study's hundred-home sample, and knowing which three tells you almost everything you need to know about where the residential sector is failing on materials-phase carbon.

Concrete leads at 39%. Foundation walls, footings, slabs, and other structural elements drive the bulk of it, and the national average ready-mix concrete used in most of these homes, a standard 3,001 to 4,000 psi blend specified by structural engineers who never think about global warming potential, carries a carbon intensity 46% higher than commercially available low-GWP alternatives that cost roughly the same per yard and meet identical structural requirements. Almost nobody specifies the cleaner mix, because nobody's energy model flags the dirtier one.

Mechanical, electrical, and plumbing systems account for 18%, with HVAC hardware alone contributing 12% of net embodied emissions. Dual-fuel homes running both a heat pump and a propane furnace averaged 11.3 tons of MEP-related carbon, while all-electric homes with ductless heat pumps averaged 7.6 tons, because fewer distribution components and no gas piping means less metal, less plastic, and less carbon embedded in the mechanical infrastructure of your house.

Insulation is where the numbers go from troubling to absurd. XPS rigid foam board, a product specified routinely in high-performance residential envelopes because of its moisture resistance and compressive strength, carries approximately 9,948 kilograms of CO2e per 1,000 square feet of coverage. Dense-pack cellulose and wood fiber board, which achieve comparable R-values per inch, are net-negative because they store atmospheric carbon in their fiber structure rather than releasing fossil carbon during manufacturing. We are not talking about a 10% or 20% improvement by switching materials. We are talking about the difference between emitting nearly ten tons of carbon and sequestering a fraction of a ton, for the same thermal performance, in the same wall cavity.

At the extremes, an 8,400-square-foot residence with a walkout basement, extensive poured concrete, open-cell spray foam insulation throughout, and a dual-fuel heating system produced 144.3 tons of embodied CO2e, while a compact home under 700 square feet with minimal concrete and a single heat pump registered 18.5 tons. A factor of 7.8 separating worst from best, driven entirely by material and design choices rather than operational efficiency.

Every AI Energy Model Ignores This

Ekotrope, REM/Rate, EnergyGauge, and BEopt all run some version of the same calculation: simulate envelope thermal performance, estimate annual energy consumption, convert to carbon using grid emissions factors, and output a HERS score that becomes the basis for code compliance, utility rebates, green certifications, and the homeowner's sense of environmental virtue. Not one of these platforms calculates embodied carbon, which means the AI-powered and physics-based energy models that have gotten remarkably sophisticated at predicting operational performance, including the Johns Hopkins Decision-Focused Learning model published at E-Energy '25 that outperformed traditional simulations on HVAC cost prediction, are optimizing for roughly 68% of a new home's 25-year carbon footprint while treating the remaining 32% as if it does not exist.

Researchers at the University of Bath recently built what appears to be the first AI tool that predicts embodied carbon from plain-text descriptions of buildings, using machine learning and natural language processing to correctly identify key materials 80% of the time and produce real-time embodied carbon estimates at the earliest design stage, where carbon-saving potential is highest. Forty-three building professionals tested it on real projects with overwhelmingly positive feedback, and the approach represents what the residential energy modeling industry should have integrated a decade ago. It does not yet connect to any US residential energy rating platform.

Fragments of a complete system already exist: the BEAM estimator handles embodied carbon assessment, Ekotrope handles operational modeling, and the draft RESNET 1550 standard would bridge the two through HERS raters who already collect 60% to 70% of the data an embodied assessment requires. Raters in the Massachusetts study reduced their assessment time from 3.5 hours to 2.7 hours after completing five to ten projects, proving that the incremental burden is manageable and shrinks quickly with practice. Infrastructure ready, integration absent.

Your Green Home Starts Life in Carbon Debt

Consider what the numbers mean for a specific homeowner. A HERS 40 all-electric home produces approximately 4 tons of operational CO2e per year, extrapolating from the study's average of 57.7 tons over 25 years for all-electric homes and adjusting for above-average efficiency. That home carries 55.5 tons of embodied carbon on day one, which means nearly 14 years of living in that ultra-efficient house must pass before cumulative operational emissions even equal the embodied "debt" that was already in the atmosphere on move-in day.

Now imagine an alternative: a slightly less operationally efficient home, HERS 60 instead of HERS 40, built with low-GWP concrete, cellulose insulation, and a streamlined all-electric mechanical system. Using the study's data on material substitution potential, embodied carbon could plausibly drop from 55.5 to roughly 30 tons, and that home's total 25-year carbon footprint, combining somewhat higher operational emissions with dramatically lower embodied emissions, would be competitive with or better than the HERS 40 trophy home that cost more to build and used more carbon-intensive materials to achieve its superior rating.

Nobody runs this comparison today because no tool makes it easy, the operational model and the embodied model live in separate software packages maintained by separate organizations using separate data standards, and neither the building code nor the green certification programs require looking at both numbers on the same page.

What You Do About This If You Are Building a Home

Specify low-GWP concrete. Ask your ready-mix supplier for a blend with supplementary cementitious materials, fly ash or slag substitution, or a formulation meeting EPA low-carbon concrete benchmarks. Your supplier likely stocks these mixes already, they cost roughly the same per cubic yard, they meet identical structural requirements, and specifying one instead of the default cuts the largest single source of your home's embodied carbon by up to 46%.

Kill the XPS foam. If your builder is specifying XPS rigid insulation anywhere in your wall assembly, foundation wrap, or sub-slab application, demand a substitute: mineral wool, EPS with roughly one-tenth the embodied carbon, or wood fiber board that is genuinely carbon-negative. Dense-pack cellulose in wall cavities performs thermally and stores carbon rather than emitting it. No energy-performance justification exists for choosing the insulation material with the highest embodied carbon when equally performing alternatives sit in the same distributor's warehouse.

Go all-electric and ductless where possible. All-electric homes in the study produced 60% lower total 25-year carbon emissions than fossil-fuel homes, and ductless heat pump systems carry substantially lower embodied carbon than ducted dual-fuel systems because they require fewer distribution components, no gas piping, and no combustion venting, meaning less material embedded in the structure of your house.

Build smaller. Conditioned floor area is the single strongest predictor of embodied carbon in the dataset, and townhomes in the study produced 38% less operational carbon and 32% less enclosure-related embodied carbon than average single-family detached homes, not because of superior materials but because shared walls mean less exterior surface and smaller footprints mean less foundation concrete.

Ask your HERS rater about RESNET 1550. It is still in draft, the BEAM tool is live, and Massachusetts is running a pilot, but if your rater has never heard of the standard, that tells you exactly how far behind the residential industry remains on the embodied side of the carbon equation.

A Real Counterargument, and Why It Weakens Every Year

Operational emissions compound while embodied carbon is a one-time charge, which means over 50 or 75 years a HERS 40 home eventually wins on total lifecycle carbon regardless of what materials built it, and this is the strongest honest case for continuing to prioritize operational efficiency above all else. Every year that ultra-efficient envelope runs, it emits less operational carbon than a less-efficient alternative, and given enough decades, the cumulative savings dwarf any upfront materials penalty.

That argument was strong a decade ago. It weakens every year the electrical grid decarbonizes, because the ISO New England grid serving Massachusetts is getting measurably cleaner, future operational emissions for all-electric homes will be lower than current models predict, and embodied carbon, by contrast, is already in the atmosphere and does not improve retroactively. As grids get greener, the proportion of lifetime emissions attributable to materials only grows, and the breakeven point where operational savings overcome embodied penalties stretches further and further into the future.

RMI's analysis of 921 model homes across the US, Canada, and Europe found 30% to 50% reductions in embodied carbon achievable with commercially available, code-compliant, cost-competitive materials. American builders construct roughly 167 million square meters of new residential space annually, producing an estimated 50 million tons of embodied carbon, equivalent to the yearly emissions of Norway, Peru, or Sweden. Cutting that by a third requires no new technology, no experimental materials, and no code changes. It requires specifying different products on a purchase order.

What This Analysis Cannot Tell You

The Massachusetts study assessed 100 homes in one state with an unusually high HERS rating adoption rate (88% of new homes, twenty percentage points above any other state), and the sample was not randomly selected. Embodied carbon calculations covered only cradle-to-gate emissions (lifecycle stages A1 through A3), excluding construction-phase transport and waste, decades of maintenance and component replacement, and end-of-life demolition. MEP product categories remain limited because many mechanical components lack detailed environmental product declarations, which means the 18% figure for MEP embodied carbon is likely an undercount. Wood's biogenic carbon storage was excluded entirely due to methodological uncertainty around forestry accounting, which probably understates the benefit of wood-frame construction over concrete-heavy designs. Cost parity claims from RMI are generalized across national markets; your local ready-mix supplier may not stock low-GWP alternatives, and your builder may genuinely never have been asked for them.

Early data. Also the best data the residential sector has ever produced on a question it has spent decades ignoring, and the question will only get louder as states follow Massachusetts toward mandatory embodied carbon disclosure for new homes.