The big top-line goal: retrofit my 150-year-old house to minimize carbon emissions over the next 30 years, including the upfront carbon of the new building materials and the carbon emissions from operating the home.
The solution also needs to be cost-effective. While not part of the top-line goal per se, I do have a budget; and besides, if I want this to be a replicable model, it needs to be affordable. (I must confess, however, that the idea that keeping the planet habitable for human life is “too expensive” strikes me as fundamentally odd. Something is clearly wrong with an economic system that produces such outcomes.)
So. How to go about calculating this?
The basic equation is this:
And here’s the research plan for calculating (and minimizing) each variable:
1. Develop retrofit options
There are a thousand possible ways to retrofit my house — exterior or interior insulation of many potential thicknesses; replacing or keeping the windows; multiple types of HVAC systems to consider; and endless combinations of these. It would be impossible to analyze them all. How do I identify a relevant subset to study, without missing an optimum option?
The short answer is that I’m using a combination of three approaches:
- Create scenarios myself based upon best practices and past experience, ranging from “lighter” to “deeper” retrofits (including, hopefully, a “Passive House” option that meets PHIUS+ standards).
- Run an optimization model using BeOPT (a free NREL tool) to identify any cost-optimal combinations I may have missed.
- For each of the scenarios identified above, run a sensitivity analysis to make sure I’ve dialed in the right insulation levels (comparing 30-year energy savings vs carbon impact of the insulation itself).
The result will be a list of possibilities for which I will calculate both operational and upfront carbon.
2. Vet insulation options
For each retrofit option identified in Step 1 above, I will look for available insulation products that are “Red List free” and avoid blowing agents with a high Global Warming Potential (GWP). If I can’t find such a product, the option will get ruled out, or the insulation type will be switched.
“Red List free” refers to products that do not include chemicals on the Living Futures Institute’s Red List. This criteria is not related to the 30-year carbon reduction goal, but rather to the goal of creating a healthy home (and, more broadly, a healthy world, free of persistent, bioaccumulative toxins that are doing untold damage to our own health and the health of our ecosystems).
Regarding blowing agents: It turns out that some insulations use blowing agents with such high GWP that it effectively cancels out any carbon benefit from improved energy efficiency. Building Green has done good research on this; their results (which I’ll dig into in future posts) show that some blowing agents are unlikely to have a positive carbon payback over their service life — namely closed-cell spray foam and extruded polystyrene (XPS).
One important caveat: Building Green’s research uses the most common blowing agent for each type of insulation. Some manufacturers have switched to better blowing agents, which can change the results significantly. So true vetting of insulation options requires finding actual products.
3. For each option, calculate operational CO2e
I will run an energy model for each option to calculate its annual energy use. To convert that figure to carbon emissions, I need to track the energy used on-site back to its source — in the case of electricity, the power plant; and in the case of natural gas, its initial extraction.
For electricity, obtaining the carbon intensity of the current system is straightforward, and many energy modeling tools already incorporate the carbon intensity of electricity based upon a project’s location.
But the grid is likely to change over the next 30 years — in fact, if we are successful at meeting the challenge of climate change, it will transition from being fossil fuel dominated to 100% carbon-free. And even if we don’t quite meet that goal, market forces and regulation will push our fuel mix in that direction (in fact, they already are).
So, I will be running each option through three scenarios:
- Current Conditions: Assumes today’s energy mix, using readily-available carbon intensity factors. I consider this an unlikely scenario; however, this is the standard practice for energy modeling, and is the approach used by certifications like Passive House.
- Clean Power Plan: Carbon emissions if states comply with the Clean Power Plan, or meets its objectives (which, as of this writing, we’re on track to achieve, despite the plan being scrapped by the Trump administration). The EPA estimated that the CPP would reduce CO2 emissions 32% below 2005 levels by 2030; and the EIA projections show emissions remaining essentially flat from 2030 to 2050.
- Meeting Paris Targets: Carbon emissions assuming a reduction in grid carbon intensity of 50% between 2020 and 2030, and then 100% between 2030 and 2050. (This trajectory keeps us within the 500Gt emissions limit required for a 50% chance of meeting the Paris targets.)
For scenarios that maintain a natural gas furnace, I intend to account for the full impact of natural gas, not just the emissions from on-site combustion. That means accounting for leakage that occurs during production and transportation. Because methane is ~24 times more potent than CO2 as a greenhouse gas, leakage during extraction and transportation can make it as bad as coal from a climate perspective.
To keep the scope perfectly consistent between electricity and on-site natural gas, I would need to take into account the extraction and transportation impacts for electricity fuels as well. IF I can find an existing reputable methodology and data source(s) for including these impacts, I will do it.
4. For each option, calculate upfront CO2e
This will be done using a Life Cycle Assessment (LCA) tool such as Tally or Athena Impact Estimator. These tools calculate the carbon emissions incurred from the extraction of raw material through the manufacturing process — a scope known as “cradle to gate” (as in, the factory gate).
One open question is whether to include carbon emissions from the transportation of the material from the manufacturing facility to the project site (“cradle to site” in LCA parlance). While this would allow me to investigate important trade-offs — for example, whether it’s worth it to import Passive House windows from Germany versus settle for less-efficient local windows — the data is difficult to calculate reliably and consistently. While I have not ruled out this “cradle to site” approach, it will depend on whether I can find good data with reasonable effort.
5. For each option, calculate CO2e of insulation blowing agents
I will calculate the impact of insulation blowing agents based either on the data published by BuildingGreen, or based on manufacturer-provided information if available.
Should I be including “lifetime” impacts that could extend beyond the 30-year scope of my analysis? There could be an argument for approximating how much of the blowing agent escapes from each insulation type over a 30 year span. But this would be a huge effort — one based on a large number of assumptions — that is out of scope for this project. I prefer a more conservative (precautionary) approach. Similar to upfront carbon, the global warming impact of blowing agents is guaranteed as soon as it is produced. We can quibble over how long it will take to reach the atmosphere — but it will get there eventually. And given high uncertainty around the rates at which blowing agents escape and the life span of insulation, the conservative approach is to include the full Lifetime GWP of the insulation.
Well … it appears that I have a lot of work to do. Better get started!