A study of 100 new homes in Massachusetts found that each one released an average of 55.5 tons of CO₂e before anyone turned on a light switch, ran the dishwasher, or adjusted the thermostat. That carbon was baked into the concrete foundations, the lumber framing, the insulation between the studs, the glass in the windows. It entered the atmosphere at the cement kiln, the steel mill, the foam-blowing facility. By the time the buyers walked through the front door, more than half of their home's lifetime environmental damage was already done.
Now California wants to count it.
What "Embodied Carbon" Actually Means for Your Foundation
Operational carbon is what your house emits while you live in it. Furnace running in January, air conditioner grinding through August, water heater firing every morning. You can measure it on your utility bill, and building codes have been squeezing it downward for decades.
Embodied carbon is different. It is the sum of every emission generated to extract raw materials, manufacture building products, transport them to the job site, and assemble them into a structure. For commercial buildings, operational carbon dominates, but for homes that relationship inverts: according to the California Air Resources Board, residential embodied carbon accounts for 59% of total lifecycle emissions. Your foundation alone, per an Atlanta-based lifecycle assessment study, represents roughly 40% of all the embodied CO₂ in your home.
That Massachusetts study measured extreme variation across home types. A compact residence under 700 square feet with minimal concrete and a single heat pump registered 18.5 tons of CO₂e. An 8,400-square-foot residence with a walkout basement, extensive poured concrete, and open-cell spray foam insulation hit 144.3 tons. Townhomes averaged 39 tons and showed 32% lower building-enclosure emissions than the sample mean. Single-family detached homes with two or more stories averaged 59.2 tons.
Extrapolate those numbers to Massachusetts's 2024 building pace of 11,390 new homes and the state's residential construction added an estimated 632,145 tons of CO₂e in embodied carbon in a single year. California builds roughly three times as many homes annually.
The December Deadline Nobody in Residential Knows About
Assembly Bill 1704, building on AB 2446 (Holden, 2022), sets three legislative deadlines. By December 31, 2026, CARB must adopt a framework for measuring the carbon intensity of building materials used in new construction, including residential. By December 31, 2028, CARB must develop a comprehensive strategy for 40% reduction. By 2035, that reduction must be achieved, using 2026 EPD industry averages as the baseline.
That framework applies to any project involving five or more new residential units or 10,000 square feet of non-residential space, and it requires submission of a lifecycle assessment aligned with ISO 14040, focused on Product Stage phases A1 through A3: raw material extraction, transport to manufacturer, and manufacturing. Not a voluntary reporting program. A measurement mandate with legislative teeth.
Six months remain.
CalGreen already imposed mandatory embodied carbon requirements on non-residential buildings over 100,000 square feet in July 2024, expanded to 50,000 square feet in January 2026. Those rules build on the Buy Clean California Act of 2017, which set GWP limits on structural steel, reinforcing steel, flat glass, and mineral wool for state projects, and added concrete to the list. Drew Shula of Verdical Group told Marketplace the current requirements "aren't especially difficult to meet, builders just haven't had to think about them before." Commercial builders adapted. Residential builders, for the most part, do not yet know any of this exists.
EPDs: The Document Your Builder Has Never Seen
An Environmental Product Declaration is a standardized report disclosing the environmental impact of a specific building material, verified by a third party and published according to ISO 14025. Think of it as a nutrition label for concrete.
Most residential builders have never read one. Most have never heard the term. Commercial construction has been collecting and comparing EPDs for years, driven by LEED credits and corporate sustainability commitments. Residential has not, because until now nobody required it.
Existing tools for working with EPDs are not new either. Building Transparency's EC3 calculator, launched in 2019 and now used by over 19,000 registered users in 70 countries, provides a free, searchable database of more than 17,000 digitized EPDs covering concrete, steel, wood, glass, aluminum, insulation, gypsum, carpet, and ceiling tile. Microsoft used EC3 during its campus redevelopment and reported a 30% reduction in embodied carbon. It integrates with Autodesk Revit through the tallyCAT plugin and connects to Autodesk Construction Cloud for takeoff-to-carbon workflows.
EC3 is free. It works. It was built for firms that model buildings in Revit and run takeoffs in Autodesk Construction Cloud.
A residential framing crew does not do that.
The AI That Reads a Napkin Sketch
Researchers at the University of Bath published a tool in 2026 that predicts embodied carbon from plain-text building descriptions. No BIM model required. No material takeoff spreadsheet. Describe a building in conversational English, specifying materials, dimensions, and intended use, and the machine-learning model returns a carbon estimate that refines itself as the description gets more specific. Professor David Coley's team trained it on 150,000 synthetic buildings and tested it with 43 building professionals. One team used it to evaluate a glass-and-masonry building near Exeter across the full design cycle, adjusting insulation and glazing decisions based on the carbon feedback without ever knowing precise material quantities.
Published in Architectural Engineering and Design Management (DOI: 10.1080/17452007.2026.2613773), the paper drew industry feedback that was, in the authors' words, "overwhelmingly positive, especially around how easy the tool is to use and how well it fits into existing workflows."
Two caveats matter. Because publicly available real-world embodied carbon data for buildings barely exists, the model was trained entirely on synthetic data; it can be retrained on better datasets as they become available, but right now the predictions are approximate, and the gap between synthetic training data and messy real-world residential construction in the U.S. is not trivial. And 43 testers is a pilot, not a validation study. Promising for early-stage design screening, but it is not an LCA.
What This Costs
Run the compliance math for a California production builder putting up a 50-unit subdivision. Under AB 1704's framework, that project trips the five-unit threshold and requires an ISO 14040-aligned LCA focused on A1 through A3 phases.
| Compliance Path | Estimated Cost | Who Does It |
|---|---|---|
| EC3 (free tool + manual EPD matching) | $0 software, 40-80 hours labor | In-house sustainability staff or consultant |
| One Click LCA (commercial platform) | $3,000-$12,000/year depending on tier | Design team or LCA consultant |
| Third-party LCA consultant | $15,000-$50,000 per project | Specialized environmental firm |
| Bath NLP tool (research prototype) | Free (not yet commercially available) | Designer at concept stage only |
At $600 per home for a 50-unit project, that regulatory overhead is a rounding error; at $7,500 for a custom builder doing four houses a year, it is a line item that bites, landing on exactly the segment of the industry least equipped to absorb new compliance costs.
Meanwhile, the Department of Energy released an analysis in June 2026 finding that updated green building codes would add approximately $14,000 to the cost of a new home. Energy Secretary Chris Wright called the mandates "nonsensical." In response, the International Code Council said DOE was "completely contradicting its own findings" from three decades of code analysis, noting that a now-removed DOE webpage had shown $182 billion in savings between 2010 and 2040. Political headwinds against green building regulation at the federal level blow strong, even as California accelerates.
What Argues Against Panic
AB 1704's December 2026 deadline is for CARB's framework, not for builder compliance. Mandatory reductions do not arrive until the 2028 strategy and the 2035 target. Production builders doing five or more units trip the threshold. Custom builders mostly do not. But the five-unit line means that subdivisions, townhome developments, and apartment-over-retail projects are squarely in scope, and small-volume custom work stays out.
Moreover, the framework focuses on Product Stage emissions (A1-A3), which are the easiest to measure because manufacturers produce the EPDs. Builders do not need to calculate transportation emissions, construction-process energy, or end-of-life disposal carbon. They need to look up the EPDs for their concrete, steel, lumber, insulation, and glass, and submit them in a standardized format. If CARB designs the submission system well, this could be a spreadsheet exercise, not an engineering study.
That is the optimistic scenario. It depends on CARB delivering a clear, simple framework by December, and the agency is still conducting public engagement with no public draft yet circulated. It depends on EPD coverage being complete enough that builders can find declarations for the specific products they actually buy from their actual suppliers, not just generic industry averages that represent nobody's real material choices. And it depends on a residential supply chain cooperating with a transparency mandate that some of its members view as regulatory overreach during a housing affordability crisis where median home prices in California already exceed $800,000.
If You Are Building in California
For production builders doing five or more units: register for EC3 now, because it is free and because the builders who know their numbers before the framework drops will be positioned to influence the baseline rather than merely be measured against it. Start pulling EPDs for your standard material specs and see where your carbon-intensity baseline sits against the industry average that CARB will use to set the 2026 benchmark.
For custom builders doing four or fewer units: you are likely exempt from the initial framework, but the trajectory is unmistakable, with Colorado, Oregon, and Washington developing similar programs, and LEED v5 tightening embodied carbon credits, and major material suppliers publishing EPDs whether you ask for them or not. Learning to read an EPD takes an afternoon. Finding the lower-carbon concrete mix from your local ready-mix supplier might save you money if the carbon-intensive option also happens to be the over-engineered one.
For homebuyers: ask your builder what the embodied carbon of your foundation is. They will almost certainly not know, and that is precisely the point, because the question itself begins to create the demand signal that makes measurement worth doing. A 2,000-square-foot slab foundation is roughly where the biggest single carbon decision in your entire home gets made.
What This Analysis Does Not Cover
The 55.5 tons-per-home figure comes from Massachusetts, a cold climate with a different grid carbon intensity than California. California homes use less heating energy, which means embodied carbon's share of total lifecycle emissions may be even higher than the 32% measured in that study. We do not have a comparable dataset for California new construction. The 59% residential embodied carbon figure from CARB is a sector-wide estimate, not a per-home measurement.
EC3's 30% reduction claim comes from Microsoft's commercial campus redevelopment, not from anything residential, and whether those savings translate to a market where the product mix is narrower and the procurement process is a phone call to a local supplier rather than a formal bid process remains unproven. Bath's NLP tool was trained entirely on synthetic building data. Its accuracy on real residential projects in the U.S. has not been independently tested. CARB's final framework requirements remain unknown, and everything in the compliance cost table above is an estimate based on current commercial LCA pricing, not on a residential-specific compliance pathway that does not yet exist.
What we do know: the concrete in your foundation generated more carbon than any other single material decision in your home, California is building the system to measure it, and the tools to measure it already exist. Most of the people who will need to use them have, for the most part, never heard of them.
That gap has six months to close.