It’s been a while since you’ve heard anything from us at Energy Literacy, but rest assured, our minds are still on energy issues. Lately, we’ve been spending a lot of time thinking about methodologies for doing reasonably accurate whole-picture energy audits using limited data (Think Saul’s energy project, but for any scale). Working with something as complex as the total set of energy flows through the city of San Francisco, for example, we’re bound to rack up all kinds of uncertainty in our estimates. We’re convinced, however, that these calculations are still helpful, even if only to determine the relative orders of magnitude of sources of energy consumption. Just these ballpark estimates can have a remarkable effect on policy conversations, directing focus towards the lowest-hanging fruit and dispelling arguments that have little long-term relevancy.
In that vein, today we talk about the embodied energy of buildings, a topic which is understood hazily at best in most green circles. Whenever there is discussion about enforcing high-efficiency standards for commercial or residential buildings, we have to keep in mind the energy we’ve invested in the infrastructure that already exists. The materials that make up a building as well as the act of constructing it both required energy. Retrofitting or reconstructing a building replaces some or all of the existing structure, and so it’s like we’re paying the energy bill twice for the same building. This additional cost could potentially offset the savings in operating energy coming from the higher efficiency standards. Preservationists rally around this argument, saying that “the greenest building is the one already built.” In reality, calculating the energy embodied by buildings and infrastructure is difficult, and the data to do so accurately doesn’t exist. Because of this, most people punt on the issue, taking a blanket stance in favor of some mixture of preservation and renovation. To determine the most effective policy choices, though, we can’t rely on canned responses. Instead, we need to work towards a system for balancing the energy ledger when it comes to buildings.
Due to the complexity and variability of building construction, accurate accounting is extremely difficult. For a great example of an energy audit of single building, see Catherine Mohr’s TED talk about the construction of her house and her analysis of the trade-offs between operating and embodied energy. Her calculations show that a conventional house takes almost 400 MWh in materials and construction energy costs, but by being mindful of embodied energy, she and her husband cut that figure to 150 MWh. Her accounting uses figures from the house’s architectural plans to determine the material inputs for the construction process. Using published figures (such as those available through the DOE) for the embodied energy of these materials, Mohr totals the material energy in her new house and adds it to her estimates for the energy involved the demolition and construction processes.
This is exactly the kind of analysis we would like to do for larger political units, but there is no feasible way to assess the architectural plans for each building in our area of consideration. To eliminate this need, we would like to know the embodied energy per square foot of many different building types. There have been attempts at this, with somewhat inconsistent results (Dixit, et. al. surveys the literature). The problem is finding a categorization scheme which accurately addresses embodied energy characteristics (i.e., buildings in the same category actually do have similar embodied energy per square foot), but which also is compatible with the data collected by government agencies.
The Advisory Council on Historic Preservation, for example, uses a scheme based on building use (residences, warehouses, hospitals, etc.). Mohr found her micro-scale estimate to be only about two-thirds of that predicted by this study. This might be explained by the fact that this study is based on the manufacturing processes from the time of its publication in the 1970s. Gumaste puts forth another scheme based on the number of stories in a building, arguing that due to structural considerations, this makes a significant difference in the energy per square foot. Ideally, we would like to see benchmarking services like Lawrence Berkeley Lab’s EnergyIQ tool which tracks operational energy costs. Users could do analysis like Mohr’s on the scale of single building and together form a larger picture of the embodied energy of buildings.
Most of these details actually aren’t too important for our purposes — We’d just like to make a rough estimate to guide policy decisions based on government data. To this end, we can try to use the ACHP data referenced above to estimate the embodied energy of buildings in San Francisco. From the city government, we know the total square square footage of buildings broken into categories of use. These don’t correspond exactly to those used by ACHP, but we can make reasonable guesses to form the following table. These estimates account for the energy used in construction, demolition, and the materials themselves.
This isn’t very exact, but it’s a starting point. All the details are contained in the spreadsheet. From this we can see that residential buildings account for roughly half of the embodied energy in buildings. Extrapolating from Mohr’s experience, we might guess that these numbers are overestimates, but at least we have an upper bound.
Ideally, we’d like to derive estimates for the embodied energy per square foot of the construction types specified by the International Building Code. These types are used to determine fire codes by rating the combustibility of buildings. Type I is a concrete and steel structure, while Type V has a wood frame. For each, there are subtypes specifying interior characteristics. In this way, we can make an approximation to the material make-up, construction/demolition energy, and average lifetime of a building from this data. These types are tracked by government agencies and are available for almost every existing structure. Bringing the level of analysis to the individual building in this way, we can track additional variables (such as building age, number of stories, number of bedrooms, and number of bathrooms, all of which are recorded by city government). By referencing this data, we can start to talk quantitatively at the policy-making level about which structures should be retrofitted, renovated, or rebuilt.
Hopefully we’ll have an update for you soon towards this end, but in the meantime feel free to offer comments or links to further studies.