Embodied carbon and reuse

Mei Takahashi, the life cycle assessment consultant who taught the students to count the carbon in building materials, arrives in the tenth week with a spreadsheet and a provocation. 'Your client's board has asked a fair question,' she says. 'Greenway Homes could build forty-six new homes on an empty lot for about the same money as converting this warehouse. New homes would be airtight, well insulated and all-electric. Old homes in a brick warehouse will leak more heat for as long as they stand. So which is better for the climate? You have heard the slogan that the greenest building is the one already built. This week you find out whether it is true for this building.'

Two kinds of carbon

A building's contribution to climate change comes from two sources. Operational carbon is the greenhouse gas emitted by the energy used to heat, cool, light and run the building over its life. Embodied carbon is the greenhouse gas emitted in making, transporting, installing, maintaining, replacing and eventually disposing of the building's materials. For most of the past half century, energy codes and green building programs focused on operational carbon, because it was the larger share for most buildings. As buildings have become more efficient and electricity grids cleaner, embodied carbon has become a growing share of the total, and for a new, efficient building on a clean grid it can be the majority of the building's emissions over decades.

Life cycle assessment divides a building's life into stages, using a scheme standardized in Europe and now used worldwide. The product stage, called A1 to A3, covers extracting raw materials, transporting them and manufacturing building products. The construction stage, A4 and A5, covers transport to site and construction itself. The use stage, B1 to B7, covers maintenance, repair, replacement and refurbishment of materials, and the operational energy and water used by the building. The end-of-life stage, C1 to C4, covers demolition, transport, waste processing and disposal. A further module, D, reports potential benefits beyond the building's life, such as materials reused or recycled. The emissions released before a building is occupied, stages A1 to A5, are often called upfront carbon, and they matter especially because they are released now, when reducing emissions quickly matters most.

What reuse saves

When an existing building is reused, its structure, foundations and most of its envelope are kept, and the carbon spent making them, long ago, is not spent again. The upfront carbon of the project is only the carbon of the new materials: the new partitions, finishes, services, insulation, windows, roofing and, in the Stanfield's case, the steel for the addition and the topping on the floors. Structure and envelope typically account for the largest share of a new building's upfront carbon, because they contain the most concrete, steel, masonry and glass. So a renovation that keeps them usually has much lower upfront carbon than a new building of the same size, often less than half, depending on how much new material it adds.

Demolition adds its own emissions: the energy of demolition itself, the transport of the debris and the disposal of materials that are not reused or recycled. Construction and demolition debris is one of the largest waste streams in North America, and a brick warehouse like the Stanfield, once demolished, would mostly go to landfill or be crushed for fill, with only a small share of its timber and brick salvaged. Its materials would also be lost as a resource. The century-old Douglas fir in its floors, for example, could not be replaced at any price today in the sizes and quality the building contains.

The evidence on reuse

The most cited study of the carbon value of reuse is The Greenest Building, published in 2011 by the National Trust for Historic Preservation's Preservation Green Lab in the United States. It compared the life cycle impacts of reusing and retrofitting existing buildings with demolishing them and building new ones, for several building types in several American cities. Its central finding was that, for buildings of the same size and function, reuse almost always produced lower environmental impacts than new construction, and that a new building designed to be more efficient than the existing building it replaced would take from ten to eighty years to overcome, through its lower operating emissions, the climate impact of its construction. For most building types and climates, the period was twenty to thirty years. The study also found that renovations requiring large quantities of new materials could reduce or even cancel the benefits of reuse.

The finding is now repeated widely, and many architects quote the line coined by the preservation architect Carl Elefante, that the greenest building is the one that is already built. Takahashi warns the students that the line, like any slogan, can be misused. Its truth depends on how much new material a renovation adds, how much better the new building would perform, how clean the energy grid is, how long the building will last and how long the comparison runs. A reuse project that strips a building to its frame and rebuilds everything else may save little; a new building on a clean grid with low-carbon materials may pay back its upfront carbon faster than the old studies assumed. The way to know is to count.

Embodied carbon policy

Policy on embodied carbon in the United States has moved in different directions at different levels of government. A federal Buy Clean initiative that favored low-carbon materials in federal construction lost its policy basis when the executive order behind it was rescinded in January 2025. States have continued their own programs. Minnesota's Buy Clean Buy Fair law requires environmental product declarations and limits on the global warming potential of certain materials in large state-funded buildings, beginning with concrete in 2026. Several cities have adopted embodied carbon requirements or incentives for new construction, and some, recognizing the value of reuse, have studied ways to discourage demolition of sound buildings. Green building rating systems award credits for reuse and for whole-building life cycle assessment that shows reductions against a baseline.

Counting the Stanfield

Takahashi helps the students build a simple model of the two options for Greenway's board. Option 1 is the conversion of the Stanfield, with the addition, as designed. Option 2 is demolition of the Stanfield and construction of a new mid-rise building of the same floor area on the same lot, built with a concrete podium and mass timber above, all-electric, to a high efficiency standard. For each, the model estimates the upfront carbon and the annual operational carbon, and then finds the year in which the new building's lower operating emissions would make up for its higher upfront emissions, if ever.

# Reuse vs demolish-and-replace for the Stanfield: carbon payback (feet, kg CO2e)
area = 47360 + 5600                   # ft2: existing floors plus the addition
# upfront carbon A1-A5, kg CO2e per ft2 (typical values, to be replaced by the project LCA)
reuse_upfront = 13.0                  # new services, finishes, toppings, windows, roof, addition steel
demolition = 2.3                      # demolition, transport and disposal of the old building
new_upfront = 39.0                    # concrete podium + mass timber, envelope, services, finishes
# operational energy, kWh per ft2 per year (all-electric in both options)
reuse_eui, new_eui = 8.8, 6.5
for grid_name, grid in {"current Minnesota grid (assumed)": 0.35, "a nearly clean grid": 0.02}.items():
    up_reuse = reuse_upfront * area
    up_new = (new_upfront + demolition) * area
    op_reuse = reuse_eui * area * grid
    op_new = new_eui * area * grid
    saved_per_year = op_reuse - op_new
    print(grid_name)
    print(f"  upfront: reuse {up_reuse / 1000:,.0f} t, new {up_new / 1000:,.0f} t")
    print(f"  operating: reuse {op_reuse / 1000:,.1f} t/yr, new {op_new / 1000:,.1f} t/yr")
    print(f"  years for new to catch up: {(up_new - up_reuse) / saved_per_year:,.0f}")

current Minnesota grid (assumed)
  upfront: reuse 688 t, new 2,187 t
  operating: reuse 163.1 t/yr, new 120.5 t/yr
  years for new to catch up: 35
a nearly clean grid
  upfront: reuse 688 t, new 2,187 t
  operating: reuse 9.3 t/yr, new 6.9 t/yr
  years for new to catch up: 615

The result is striking. Reuse saves well over a thousand tons of upfront carbon, roughly two thirds of what the new building would emit before anyone moved in. On the current grid, with the assumed emission factor, the new building would need several decades to catch up through its better efficiency, close to the twenty to thirty years The Greenest Building found typical. As the grid becomes cleaner, which Minnesota's law requiring carbon-free electricity by 2040 is meant to achieve, the operational savings of the new building shrink and the payback stretches toward centuries. For the Stanfield, reuse is clearly the lower-carbon choice.

Takahashi points out the assumptions that could change the answer and asks the students to test them. If the renovation required much more new material, for example replacing all the floors with concrete, the reuse upfront figure would rise and the gap would narrow. If the reused building were much less efficient, the operating gap would widen. If the new building were built with very low-carbon materials, its upfront carbon would fall. But for a building with a sound structure and envelope, on a clean or cleaning grid, the advantage of reuse is large and robust, and it comes when it matters most, now rather than decades from now.

Making the reuse lower carbon still

Counting carbon does more than settle the argument with the board. It shows the students where the carbon of the conversion lies and how to reduce it. Takahashi's breakdown of the reuse option shows that the largest items are the steel for the addition, the floor topping, the new windows and storefront, the services and the finishes. The students respond with a list of changes. The addition's structure is redesigned in mass timber on steel columns, reducing its steel. The floor topping is specified in a gypsum-based or low-cement mix, and its thickness reduced where acoustic needs allow. Original windows are repaired rather than replaced, as Lesson 6 recommended, saving the carbon of new units. Salvaged timber from the light court is reused in the building. Finishes are kept minimal, with exposed brick and timber in place of new linings. Heat pumps are chosen with low-impact refrigerants and their refrigerant charge counted, because leaking refrigerants can add significantly to a building's emissions.

These choices also serve the conservation goals of the project. Repairing windows keeps original fabric, exposing brick and timber keeps the building's character, and reusing salvaged material keeps the building's own material in the building. Bérubé notes that conservation and carbon reduction are not always aligned, the risk of interior insulation to the masonry being one example, but on most questions they point the same way: keep what exists, repair it, add as little as necessary.

Exercise 10: counting carbon

• Estimate the upfront carbon of a renovation of a building near the school, listing the new materials it would need.
• Run the payback model with your own values for upfront carbon, energy use and grid intensity, and find the crossover year.
• Identify the three largest sources of upfront carbon in the Stanfield conversion and propose a reduction for each.
• Explain to a lay audience, in one paragraph, why reuse can be better for the climate than a new efficient building.

Self-check

• Do I distinguish between operational, embodied and upfront carbon?
• Have I counted the carbon of demolition as well as new construction?
• Have I tested how the answer changes with the grid, the amount of new material and the building's efficiency?
• Have I used the carbon breakdown to improve the design?