Is Working From Home Green?
A back of the napkin Lifecycle Analysis
One phenomenon that’s been discussed at excruciating length following the onset of the Covid-19 pandemic in the spring of 2020 was the rapid shift of many white collar employees to a primarily work from home (WFH) environment. I’m not going to discuss all the socioeconomic and philosophical questions and conundrums brought about as a result of this sudden, dramatic shift in the fundamental way us office workers’ lives (and the associated social structure) changed. Millions of words have been spilled on the topic but I think one particular aspect that is extremely interesting is the sustainability question.
The prima facie assumption by many, if not most, people was that this new world of working from home would be a win for the environment. After all, most office workers in the country drive individual gas powered vehicles every single day, back and forth, to work, where they sit at an office and do computer work that can just as easily be done 50 feet from where they sleep.
As data began to pour in, however, it was clear that consumption habits changed dramatically. This has been most apparent with water consumption. Perhaps the novelty of taking a shower at lunch (I’m guilty as charged) is too good to pass up.
So the emphasis has been: “Carbon, Carbon, Carbon.” But really it’s just “Carbon associated with Transport.” Here’s the thing: Office Buildings are really efficient! We can heat and cool people much more effectively when they are packed in one place. This isn’t just fancy new LEED certified buildings, it’s really all of them. Density, in humans per cubic foot is often dozens of times larger in offices. Commercially sized HVAC systems are also much more efficient as a function of SEER rating. Energy and carbon emissions go hand in hand.
Toilets and sinks installed at office parks are usually much more efficient as well. This alters gallons of water per flush. But perhaps the biggest change (that’s been shown empirically) is that total water consumption per worker goes up much more than just would be expected by the “flush” efficiency. The WFH phenomenon means that people seem to use the bathroom more often, are more likely to take multiple showers and use more water when cooking for themselves.
I’ve had multiple people ask me how I think about problems. I’m not pretending to be an expert in all things but figured the question: “How green is Working from Home?” was a fun opportunity to make a rough, first pass model.
So, what I did is take the best available data (from utilities, EPA, standard emission factors, a few studies) and made some basic assumptions. I wanted to look at a rough framework for carbon impact and water usage for home office workers. Every WFH situation is different so I came up with 3 scenarios and plotted them out.
HEY YOU! SUBSCRIBE! IT’S EASY AND IT MAKES ME FEEL GOOD TO SEE NUMBER GO UP. THANKS!
The Scenarios
We’ve got three people:
An accountant at a large oil company who lives in a 3000 square foot house in a Houston suburb, drives 25 miles each way in a 2019 Ford F-150 to her office downtown. Her husband also works downtown and they sometimes commute together but can’t always due to school schedules, so on these days he stays home. This accountant is considering buying an EV to lessen her impact on climate change.
A Junior analyst who lives in a 700 square foot Washington Heights studio and takes the Subway to his job at Wall Street.
A Google Engineer who lives in a 4 person, 1700 square foot house in Oakland with 3 co-workers. They all take the Google Shuttle to work.
Now, I’m happy to provide the calculations for my model if you’re so inclined, but I don’t want this post to be bogged down with that, when there’s really nothing serious on the line. Assumptions are listed at the end of this post.
So without further delay here are the results:
Climate Impact from Transport
I first charted the transportation Carbon Impact out and then netted out the energy (eg carbon) savings from efficient utilities and scale in a modern office.
The first model shows how much carbon can be attributed to theoretical workers 1 in a variety of scenarios.
There are a few things to note:
Due to power consumption needs in suburban Houston and the higher carbon intensity of Texas’ grid (roughly 2x of California’s per kwh), setting the thermostat at 80 during the day while you go into your office in a modern high rise saves a ton of energy. I estimated a roughly 200 kg/month carbon impact difference for a single house, even factoring in the office energy needs!
Buying a small EV and carpooling, per the model, results in a slightly negative net carbon. But that benefit goes away entirely on days her husband stays at home.
The subway rider’s carbon impact from commuting into work is obviously miniscule given the mode of transport as well as the small size of his apartment. The A/C usage almost makes up for this, but it’s close.
The bay area’s mild climate and close living quarters due to rent as well as its lower carbon grid means that “savings” from working at a brand new Platinum Certified LEED building don’t really matter.
With that said, the dirty diesel bus that takes the Googlers 70 miles, round trip, into the office is tiny compared to even the EV user in Houston driving only 50 miles per day. Public transport FTW.
Same model but in winter:
The Bay area is entirely unchanged. Mild weather all year long is pretty nice.
The NYC apartment consumes significantly less energy in the winter and as such, any marginal benefit from going into the office during summer is wiped out. Although, again, the subway is clearly an energy efficient way to move lots of people.
As for our Texas couple, well, mild weather in most of the winter means that A/C usage plummets and energy bills are generally 50-75% lower in these months. Driving into work, EV or not, will not be less carbon intensive.
Again, the assumption people made that the Covid-19 “Work From Home Revolution” would decrease the carbon footprints of workers holds up using my cobbled together mass balance exercise. Especially when looking at transportation impacts.
It’s when we start talking about water that things get interesting.
Water Usage
Water usage skyrocketed nearly immediately once lockdowns went into place in Spring 2020. It was a global phenomenon. And this wasn’t just demand shift from offices and schools; per capita usage spiked 20-40% in the developed world. People’s homes don’t have as efficient toilets and sinks. It’s easier to pour an extra glass of water, make more iced tea than you need and cook things. There are more dishes to wash. And people took showers a lot more. Lockdown boredom will do that.
There are lots of ways to slice the data I have. I’ve taken a hybrid approach using my own water bills as well as data from municipalities and a few academic studies.
Basically, I assume that the average American WFHer will consume an additional 20 gallons per work day, as a result of normal activities (hand washing, toilets, regular food and drink). 20 days a month at work mean 400 gallons per worker per month. Now, this is conservative. Actual water use for each WFH likely was as high as 40 gallons per day. This was largely because people took more showers, made more homecooked dishes and did things like garden and wash their cars at home. They were bored and lonely. So it’s unclear how much of that full 40 gallons is temporary and how much is durable. But it does raise a question of how our daily routines affect consumption in aggregate.
Climate is a global problem, but water is a local issue (as I’ve covered once or twice). Texas has more than enough surface and clean groundwater at this moment for it to not be a concern. California, however, doesn’t have such luxury. Let’s say 3 million Californians start working from home and as a result consume an extra 800 gallons per month. That’s 2.4 billion gallons of water per month that wasn’t consumed before. Water shortages are funny in that pricing doesn’t reflect true scarcity until you cross the threshold of actual scarcity.
So what’s the economic and environmental impact of the marginal work from home employee’s additional water usage? That’s a complicated problem. And it’s mostly a local one.
For simplicity, I’m going to convert water usage into carbon intensity just to close the loop on the model. Carbon intensity is based on grid composition, water treatment energy costs and extraction and transport costs. I’ve *very roughly* assumed a Ground to Tap residential energy usage of 70,141 kWh/MG in Texas, 79,139 in NYC and 106,852 in California.
Again, we can’t really compare Water usage with CO2 emissions as if there is a fungible “going green currency.” I absolutely want to emphasize this. Being green isn’t just an accounting exercise. That being said, this is a fun hobby for me so, let’s do some unit conversions and tack the water component at the end of our previous chart.
I honestly think the above chart passes the smell test. Someone who drives a single automobile into work probably isn’t being sustainable, no matter how you cut it. Using public transport makes the possibility of leveraging common space into saving resources an interesting point of discussion. But I think the NYC commuter sort of speaks to the root of the issue. The guy living in an apartment in a densely crowded city is going to use less resources regardless, so trying to get him to come into the office to fit a sustainability metric seems like a poor use of time. Especially when so many corporate workers hop into a two thousand pound car and waste time and energy sitting in traffic to just go sit at a computer on the other side of town.
Sorry for the rough format, hope y’all enjoyed.
Until next time, ESG Hound
References:
Griffiths-Sattenspiel, Bevan and Wendy Wilson “The Carbon Footprint of Water.” The River Network (2009)
Deike U. Lüdtke et al. “Increase in Daily Household Water Demand during the First Wave of the Covid-19 Pandemic in Germany” Water (2021)
U.S. Department of Energy Report: Environment Baseline Vol. 4: Energy-Water Nexus (2017)
Carbon Intensities and Energy usage estimates from Dept of Energy datasets
Emission factors and fuel/drivetrain data from the GREET v3.0 Model and EPA AP-42 handbook
Contact me if you want to know more about my silly, not at all serious or robust model
1. Houston to Downtown, Pickup Truck - 2. Same scenario but both house members carpool together. 3 & 4 are same as 1 & 2 but with a generic small EV. 5. is a worker who lives in an apartment and takes subway to work. 6 and 7 are four and one housemates (respectively) riding a bus to work. All values are per person per month