The Energy Challenge Is Huge–But Only Half as Huge as Pielke Estimates
I’m reviving my dormant blog out of fear that I’m missing something, perhaps something obvious. So, take your mind off of the coronavirus pandemic and relax by worrying about the climate instead. (Who’d have thought?) Then let me know if I’m off base. Here’s the situation:
Pure Cap-and-Dividend would require net zero greenhouse gas (GHG) emissions by 2042. While reassessing the feasibility of that timeline, I’ve lately been knee-deep in data, mostly from the U.S. Energy Information Administration (EIA), and slogging through related academic debates.
So it happens I was especially interested when I came across a column on Forbes.com by Roger Pielke, a professor at the University of Colorado-Boulder, in which Pielke calculates that achieving net-zero emissions for the globe by 2050 will require building the equivalent of three large-scale, traditional nuclear power plants—each approximately the size of the Turkey Point nuclear plant in Homestead, Florida—every two days. Since nuclear power is out of favor with many, Pielke provides a renewable energy equivalent: installing a wind farm with 1,500 turbines (2.5 MW each) over about 300 square miles every single day until 2050.
Narrowing the scope to the United States (my focus for the remainder of this post), Pielke calculates that we need approximately one new large-scale nuclear plant every six days. Or, extrapolating from his global figures, a new 1,500 turbine wind farm about every nine days. Wind farms would eat up a whopping 10% of the nation’s land area.
Pielke isn’t arguing against action on climate. Nor is he advocating antiquated nuclear technology. (The nuclear energy industry is moving toward smaller, more uniform power generation facilities.) Rather, Pielke is seeking perspective. “The scale of the challenge is huge,” he states, “but that does not make achieving the goal impossible. What makes achieving the goal impossible is a failure to accurately understand the scale of the challenge and the absence of policy proposals that match that scale.” Agreed.
That said, the very idea of adding so much new electricity generation is indeed, as Pielke puts it, “absolutely, mind-bogglingly huge.” Further, Pielke is talking about achieving net zero in 2050. That’s eight years later than Pure Cap-and-Dividend requires. Gulp. Am I advocating the impossible?
I haven’t seen anyone question Pielke’s assessment. Andrew Nikiforuk generally seconded it in the Canadian online magazine The Tyee, emphasizing the need to reduce our energy demand, not simply rely on renewables to save us. So, already working through EIA data, I took up Pielke’s invitation to do the math myself and started in on things from his angle.
Unfortunately, Pielke’s approach seemed to check out. His numbers, from BP’s Statistical Review of World Energy 2019, appear generally consistent with the EIA’s data on United States energy use (at least after backing out non-combustion use of fossil fuels from the EIA numbers). And Pielke’s calculations are fairly straightforward. In 2018, according to BP, the United States consumed roughly 2,300 million tons of oil equivalent (mtoe) of primary energy (EIA says 2,400 mtoe), about 1,900 mtoe of which was in the form of fossil fuels. Thus, Pielke reasons, we need to replace that 1,900 mtoe of fossil fuel derived energy with energy from GHG free sources.
Pielke uses Turkey Point as a benchmark because he says it produces roughly 1 mtoe of energy per year. That looks right. Turkey Point has two nuclear generating units with a combined capacity of about 1,685 MW. Assuming a 92% capacity factor (the recent average for nuclear plants), Turkey Point’s nuclear units crank out around 13,580,000 megawatt-hours (Mwh) of electricity per year. One million tons of oil equivalent converts to about 11,630,000 Mwh. Let’s call that close enough.
By my calculations, Pielke’s wind farm equivalency is a bit further off. He implies that two of his hypothetical wind farms (again, comprised of 1,500 turbines rated at 2.5 MW each) are equivalent to three Turkey Points, meaning each wind farm would generate about1.5 mtoe of energy annually. As described, though, Pielke’s wind farms should each crank out just shy of one mtoe of energy annually, assuming a 34.5% capacity factor (wind farms generate electricity a much smaller portion of the time than do nuclear plants). That said, upsizing the turbines to 4 MW each (a simple task with imaginary wind farms) brings us in right around 1.5 mtoe of energy annually. We’d need about 1,270 such wind farms.
As for the rate at which we need to build this new generating capacity, that obviously depends upon one’s deadline for replacing all fossil fuel use. Twenty years is roughly 7,305 days (depending on where your leap years fall). 30 years is approximately 10,957 days. From there you can calculate days per Turkey Point (or wind farm). Whether you envision building 1,900 Turkey Point-sized nuclear plants or 1,270 wind farms over such a period, the task sounds daunting because it is daunting. My own separate analysis of EIA data suggests adding 1,900 mtoe of GHG-free electricity even over a 30-year time frame would require an unprecedented build out of new generating capacity. (More about that another time.) I felt deflated.
Aha! That’s Way Too Much Electricity
I kept staring at my spreadsheets. Double checking math. And generally trying to wrap my head around the numbers. I looked back at Pielke’s article. And Nikiforuk’s.
Then, with three screens of spreadsheets before me, including data on energy measured in British thermal units, million tons of oil equivalent, and megawatt-hours, I noticed something odd. By chance, on one screen, I was looking at 1,900 mtoe as converted to megawatt-hours: 22,097,000,000. On another screen, total electricity generated in 2018 by the U.S. electric power sector: about 4,018,253,000 megawatt-hours. Wait, 22 billion to 4 billion?
That would mean we need to build enough new GHG free capacity to generate about 5.5 times the total amount of electricity generated by the electric power sector in 2018. That can’t be right, I thought to myself, I must have messed up somewhere. After all, the electric power sector today uses about 40% of the primary energy (fossil fuels and otherwise) consumed in the US. But where did I go wrong?
Finally, I recognized the flaw in my thinking and, I believe, correspondingly in Pielke’s analysis. A look at the electric power sector best illustrates the error, so let’s start there. (EIA breaks energy data out into five sectors: electric power, transportation, industrial, residential, and commercial.) The electric power sector consumes approximately 585 mtoe of the 1,900 mtoe of primary fossil fuel energy consumed by the United States. (Electric power generation is relatively clean in terms of GHG emissions; roughly 20% of our electricity comes from nuclear energy and about 17% from renewables, including hydro.)
Following Pielke’s logic, we’d need 585 new Turkey Points to replace the 585 mtoe of primary fossil fuel energy used by the electric power industry. But hold on! From that 585 mtoe of primary fossil fuel energy, the electric power industry generates roughly 2,350,000,000 megawatt hours of electricity—the equivalent of about 202 mtoe. The other 383 mtoe of primary fossil fuel energy is lost. (See “Electrical System Energy Losses” in tables 2.2 through 2.6 of the EIA’s Monthly Energy Review.)
Pielke’s calculation conflates primary energy with what I’ll call “useful energy.” (Electricity is sometimes categorized as “secondary energy,” as are gasoline and other refined fuels, but that term is less useful here.) As far as the electric power sector is concerned, we don’t need to replace all of the energy in the fossil fuels that we burn—we need to replace the electricity derived from burning them. Thus, we do not need 585 Turkey Points nuclear plants, we need 202 of them. That means 383 fewer Turkey Points needed.
The Transportation, Industrial, Residential, and Commercial Sectors
Next we need to consider the 1,315 mtoe of fossil fuel-based primary energy consumed by other sectors, keeping in mind that—as with the electric power sector—we don’t need to replace all of the primary fossil fuel energy with electricity, we need to replace the actual useful energy with enough electricity to accomplish the same tasks. Unfortunately, this will require some guesswork. For the electric power sector, we can simply compare the use of fossil fuels with the amount of electricity generated from it. I’m not aware of comparable data for the other sectors. Still, we should keep two important factors in mind: First, we lose a great deal of fossil fuel energy when converting the heat from fossil fuel combustion into other types of energy e.g., electricity or mechanical energy. (When fossil fuels are used for heating or, better yet, combined heat and power (CHP), far less energy escapes unused. Generally speaking, as we move off of fossil fuels, we should shut down the CHP systems last.) Second, in contrast, we can efficiently convert electricity into both heat and mechanical energy.
The transportation sector consumes about 676 mtoe of primary fossil fuel energy. (Again, based on EIA data after backing out non-combustion fossil fuel use.) Most of that fossil fuel energy is lost as heat from internal combustion engines. According to the US Department of Energy, conventional gasoline vehicles convert, at best, about 30% of the energy in gasoline to power the vehicle; electric vehicles, by comparison, put over 77% of the electricity they take from the grid to work. That means that 100 units of gasoline energy can be replaced by 39 units of electric energy (probably less).
Is that wishful thinking? Consider a real-world comparison: a BMW M340i xDrive and a Tesla Model 3 with the Performance configuration. These two all wheel drive cars are roughly the same size with similar performance metrics and prices. BMW claims the M340i xDrive will hit 60 mph in 4.1 seconds. It’s probably faster; Car and Driver clocked the rear wheel drive version at 3.8 seconds. Car and Driver reported a 3.5 second time for the Model 3—but that was in 2018 and the car has since undergone continual improvements. Tesla’s claimed 3.2 second time is realistic. Both will set you back around $60,000.
As for energy usage, EPA rates the BMW’s fuel economy at 25 MPG combined. The Tesla? 113 MPGe (miles per gallon equivalent) combined. (Of course, neither car will come close to its mileage estimate if you take advantage of the acceleration figures.) Put another way, the Tesla will take you a mile on about 22% of the energy the BMW requires.
All that considered, how much electricity do we think we will need for the transportation sector? First, let’s set aside approximately 13% of the sector’s primary fossil fuel energy that is used for jet fuel and aviation gasoline (about 88 mtoe) and assume we need to replace that with an equivalent amount of energy in the form of electricity. There is plenty of reason to believe the number will be lower, but that is a topic for another time. Now, for the remaining 588 mtoe of fossil fuel primary energy consumed for surface transportation, let’s conservatively assume we need electric energy in an amount equal to 40% of the fossil fuel primary energy used today. That’s about 353 fewer Turkey Points required.
The industrial sector accounts for about 372 mtoe of primary fossil fuel energy consumption. Primary fossil fuel energy use varies significantly within the sector and is beyond the scope of this analysis. Here, I will simply assume a modest 25% reduction in the energy needed across the sector if that energy is provided in the form of electricity instead of in the form of fossil fuels. That’s about 93 fewer Turkey Points.
Finally, let’s turn to the primary fossil fuel consumption in the residential and the commercial sectors, approximately 154 mtoe and 114 mtoe respectively. In both cases, the bulk of the fossil fuel use is for heating, e.g., homes and businesses heated by oil boilers and natural gas furnaces.
Though higher efficiencies are possible, an average boiler or furnace is running well if it is 85% efficient –meaning only 15% of the energy from the fuel burned is escaping up the chimney. (Remember, we are estimating actual efficiencies of existing systems, not possible efficiencies.) Still, by the time that heat is delivered through heating ducts or via steam or water radiators, another 35% of the heat can be lost. Electric heating, by contrast, is 100% efficient (at least if indoors). That might sound surprising because electric heating is traditionally viewed as expensive, but remember, under our existing fossil fuel-heavy energy system, much fossil fuel energy escapes (as heat) when generating the electricity in the first place. In any case, when we replace the combined 268 mtoe of primary fossil fuel consumption by the residential and commercial sectors with electricity, we don’t need to replace the heat energy lost up the chimney or through the duct work. Given the above, it strikes me as reasonable to assume that 30% of the primary energy delivered to the residential and commercial sectors is lost—and would not be if electricity were delivered. That knocks out about 80 more Turkey Points.
A Thousand Turkey Points Avoided?
In sum, once we recognize that we do not need electricity in an amount equivalent to the primary energy expended to produce useful energy—but rather we simply need enough electricity to replace the useful energy currently eked out of the fossil fuels, we eliminate the need for about 909 mtoe of electricity (and 909 Turkey Points).
Might my assumptions about how much electricity we need to replace the useful energy each sector derives from fossil fuels be wrong? Absolutely. Other than the electric power sector figures, they are educated guesses at best. But they are estimates we must make if we want perspective on the challenge. Pielke’s estimate—that we need electricity in a quantity equal to our primary fossil fuel energy consumption, notwithstanding all of the energy lost when fossil fuels are put to use—was simply hidden.
Moreover, I have not even addressed conservation and efficiency. I will argue elsewhere that we can easily slash our energy use (and overall natural resource gluttony) by 25% to 50%. For context, according to BP’s data, on a per capita basis, the Germans, French, and Japanese consume just 56%, 53%, and 51% of the energy we consume. Here, though, let’s assume that Americans can, over 20 years, reduce our end-point energy use by a meager 10%. Such a modest reduction will eliminate the need for another 99 mtoe of energy. That would mean we need fewer than 900 Turkey Points.
So now, instead of fearing Pielke is right, I’m worried I’m wrong. After all, underestimating the amount of new clean energy we will need helps no one. Are my assumptions off base? Is electricity so much less useful that, if we want to leave fossil fuels behind, we actually need over 5.5 times the electricity we generate today?
Adding the equivalent of 900 Turkey Points is still no small task. Creating that much new energy generation in the United States over 30 years would require building the equivalent of a new Turkey Point every 12 days. To do so within the 20-year term of Pure Cap-and-Dividend would mean adding a new Turkey Point about every 8 days. And I’m sure you’ve noticed that this discussion does not even touch on myriad other challenges involved in transforming our energy system, from energy storage to intermittency to land use to new infrastructure to…well, you get the idea.
Importantly, though, adding 900 mtoe of electricity generation by 2042 is within reach, a topic I’ll explore further another day. For now, suffice to say that the challenge remains huge–but less than half as huge as Pielke estimates