President Joe Biden wants to electrify the federal vehicle fleet. Tesla Inc. just sold half a million plug-in cars and expects to increase that by 50% a year. General Motors Co. and Ford Motor Co. are even working on electric Hummers and F-150s.
So where’s the power going to come from? EVs are already greener than cars running on gasoline even with our current electricity grid. But mass adoption only makes sense as part of a do-over of the entire energy system, so the question of what will power all these plug-ins is valid.
And here’s a surprising answer: Electrifying U.S. vehicles wipes out the equivalent of our entire current power demand.
The U.S. consumes a lot of energy; last year, about 100 quadrillion BTUs (equivalent to 17 billion barrels of oil; which, we’ll admit, is only marginally less abstract). But only about a third of that is ultimately used in terms of actually lighting lights, turning wheels and so forth. The second law of thermodynamics means, for every unit of thermal energy we actually put to useful work, roughly another two end up wasted as heat.
How we don’t use energy is just as important to understand as how we use it. Here’s a simplified version of a Sankey diagram from the Lawrence Livermore National Laboratory showing the various inputs to the U.S. energy system and where they end up.
U.S. Energy Inputs: 2019
15
Hydro,
geothermal,
biomass,
nuclear
2
37 quads
32
11
Solar,
wind
Petroleum
Natural gas
Coal
Petroleum inputs
Electricity
Oil accounts for 38% of U.S. energy inputs
96 quadrillion BTUs
U.S. energy inputs, 2019
Virtually all oil goes directly to consumer segments (i.e. hardly any used to generate electricity)
Electricity
generation
Two-thirds of energy inputs for generating electricity is lost as heat
Consumers
Transport accounts for 70% of oil inputs (just 0.1% of transport’s energy comes from the grid)
ReSidential,
commercial,
industrial
Transport
Most oil energy for transport is lost as heat, with only about a fifth actually turning the wheels
32
Usable
64 quad BTUs
Lost as heat
Ultimately, two-thirds of “primary energy consumption” is actually wasted, and about a third of that relates to transport Source: Calculations based on data from Lawrence Livermore National Laboratory, Otherlab and U.S. Department of Energy
Large-scale waste is unavoidable with a thermal energy system, or one where we mostly burn stuff or split atoms (97% of the inputs in 2019). Burning fossil fuels also generates the carbon emissions causing climate change; so wasted energy is a proxy for the damage being done (apart from nuclear power). In contrast, renewables such as wind, solar and hydropower capture energy directly from infinite sources. While a small amount is lost in transmission, the vast majority is used.
So here’s a thought experiment: What if the entire U.S. light-duty vehicle fleet (currently about 270 million cars and trucks) were electrified by 2030 and we expanded wind and solar generation at a rapid pace, while eliminating coal power, at the same time?
The result is that we not only end up with a drop in U.S. carbon emissions of almost 30%, but also a far more efficient system overall.
Let’s go back to that Sankey diagram[1] and show what happens to the inputs of 2019 by 2030 under a few simple assumptions:
- • Solar generation rises by 20% a year; wind by 10%; coal and oil generation drop to zero.
- • Energy consumption rises as per the Energy Information Administration’s long-term forecasts[2]…
- • …But with the added adjustment that light-duty vehicle energy demand converts entirely to electricity.[3]
- • The additional electricity required that isn’t met by solar or wind is generated with natural gas.[4]
U.S. Energy Inputs: 2030
Imagine we change the energy mix, reducing oil by 13 quads, cutting coal almost entirely and raising gas by 8 and solar and wind by 4
–2
Hydro,
geothermal,
biomass,
nuclear
+4
–13
+8
–10
Solar,
wind
2019 levels
Petroleum
Natural gas
Coal
Petroleum
Other fuels
Not all 2019
flows are shown
throughout
83 quadrillion BTUs
U.S. energy inputs, 2030
83 quadrillion BTUs
U.S. energy inputs, 2030
Roughly the same amount of energy goes into generating electricity
Electricity
generation
Electricity
generation
But now less than half is wasted (goodbye coal) and more goes to power transport
But now less than half is wasted (goodbye coal) and more goes to power transport
Consumers
Electrifying all passenger vehicles cuts oil inputs to transport by more than half
Effectively, about 14 quads of oil energy is replaced by 4 quads of electricity
Residential,
commercial,
industrial
Transport
Transport’s wasted energy is reduced by more than half
35
Usable
48 quad BTUs
Lost as heat
35
Usable
48 quad BTUs
Lost as heat
The result is a bigger economy running on 13% less energy inputs. In a word: efficiency Source: Calculations based on data from Lawrence Livermore National Laboratory, Otherlab and U.S. Department of Energy
The power-generation system transforms from one dominated by fossil-fuels both in terms of inputs and useful energy to one that is essentially half natural gas and half non-fossil, with the majority of that being wind and solar. Despite the electrification of light-duty vehicles, inputs to the grid actually fall slightly. The replacement of coal-fired power by more efficient gas turbines and the rapid expansion of non-thermal renewable power means useful electrical energy rises by more than a third anyway.
That efficiency gain feeds into an even bigger one: the replacement of inefficient internal combustion engines.
Despite the assumed retention of these by heavier vehicles — an unsafe assumption, but just keeping it simple — and increased use of petroleum in industrial processes, the amount of oil funneled into the top of the energy system drops by more than one-third. Assuming that’s all gasoline, it equates to more than six million barrels a day of demand dropping away. That’s peak oil demand right there. Along with that, the wasted energy from transportation, which accounts for more than a third of the total today, drops by more than half.
Throw in the efficiencies on the grid itself, and the amount of wasted energy saved is equal to one-sixth of current U.S. energy consumption. Overall, U.S. primary energy consumption drops by 13%.
That saving is bigger than the entire amount of electricity we draw from the grid today — despite a bigger population, a bigger economy and an utter transformation of the American vehicle fleet.
Usable electricity generated in 2019
13 quadrillion BTUs
Efficiency gain (net of inputs lost as heat, 2019 vs. 2030)
+16 quadrillion BTUs Source: Calculations based on data from Lawrence Livermore National Laboratory, Otherlab and U.S. Department of Energy
Even under this scenario, more than half the energy inputs of 2030 would still be wasted as heat. But with the grid and the vehicle fleet now much more efficient, the industrial sector becomes the single biggest user of fossil fuels and source of wasted energy.
Such an enormous project requires enormous investment; in EVs, of course, but also in everything from new wind turbines to electric-vehicle chargers. At current capital costs,[5] the build-out of solar, wind and gas-fired capacity required under our simple projection adds up to about $80 billion a year. But “current” does a lot of work there; renewable technology costs have dropped precipitously over the past decade and BloombergNEF projects a further drop of 40% and 20% for solar and wind-power, respectively, by 2030.[6] Moreover, focusing only on costs ignores the benefits of investment: At a notional $50 a tonne, the value of negated carbon emissions adds up to $83 billion in 2030. At $50 a barrel, $115 billion worth of annual oil demand disappears.
Gas producers, with all that extra demand for power, would no doubt be happy. But since gas is just the plug in this simplistic model, don’t go buying that plot in Appalachia just yet. Resulting higher gas prices would have their own impact. Also, wind and solar-power might grow faster than our assumptions as prices keep falling. Meanwhile, lithium-ion battery pack costs, having dropped by almost 90% since 2010, are projected to drop another 60% by 2030.[7]
The point here is that alternatives to the thermal energy system that has powered us simultaneously to modernity but also a gathering climate crisis are available. And their less-is-more efficiency gains offer a compelling reason to embrace them.
[1] You will notice our version adds up to 96.3 quadrillion BTUs of primary energy consumption rather than Lawrence Livermore National Laboratory’s 100.2 quadrillion. This is because the latter treats renewable energy sources as if they were thermal in order to make them comparable to the dominant sources such as oil, natural gas and coal. This grosses-up the renewable energy numbers; we’ve adjusted them down to remove this.
[2] Forecasts are taken from the Reference case in the EIA’s Annual Energy Outlook 2020, published in January 2020.
[3] This forecast is derived as follows. Light-duty vehicles account for 54% of U.S. petroleum-derived energy demand, according to the “Supersankey” chart created by Otherlab for the Advanced Research Project Agency of the Department of Energy. This implies 13.9 quadrillion BTU of primary petroleum demand, for useful energy consumption of 2.9 quadrillion at assumed 21% thermal efficiency. That 2.9 quadrillion is equivalent to 3.8 quadrillion of electricity from the socket, assuming all vehicles are electrified at 77% efficiency (see this). Vehicle miles traveled are forecast by the EIA to grow by 9% through 2030, implying the electricity required from the socket for the fleet would be just over 4.1 quadrillion BTUs.
[4] Apart from coal, oil, natural gas and renewable power, all other electricity sources are kept constant to keep things simple.
[5] As per “Capital Cost and Performance Characteristic Estimates for Utility Scale Electric Power Generating Technologies” (Energy Information Administration, February 2020).
[6] Forecast change in levelized cost of electricity for U.S. utility-scale solar and onshore wind power from 2020 to 2030, as per BloombergNEF’s “New Energy Outlook 2020” (November 2020).
[7] Source: BloombergNEF’s “New Energy Outlook 2020” (November 2020). The projected decline in battery costs assumes continuation of an 18% learning rate.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
By Liam Denning and Elaine He
