Issue 028 - Energy transition - Fleet stock and flow
Can Australia's EV imports explain falling gasoline imports?
Reuters reported that Australia's gasoline imports fell about 15%, while imports of Chinese electric vehicles nearly tripled. Estimate whether the additional EVs are large enough to explain the fuel-import drop.
The problem
Estimate how much gasoline the additional Chinese EVs entering Australia could displace over one year.
Convert your result into gasoline-tanker shipments avoided and the percentage of Australia's annual gasoline use or imports.
Could the new EVs plausibly account for a substantial portion of the reported 15% decline, or must other factors explain most of it?
Because Fermi problems target an order of magnitude, I normally use no more than two significant digits and write most calculations in scientific notation; the Fermi reference explains both conventions.
Before checking sources
Matt's first pass
I was not sure where to start with this one. I do not have a good set of memory pegs for Australia. I began by assuming the population of Australia was less than the U.S. but greater than Canada, and landed on about 20% of the U.S. population: 66 million.
Then I assumed a similar household breakdown to the U.S., one household for every 2.5 people:
households ~= 6.6 x 10^7 people / 2.5
~= 2.6 x 10^7 households
vehicles ~= 2.6 x 10^7 households x 1.5 vehicles/household
~= 3.9 x 10^7 vehiclesI assumed the average vehicle uses about 80 L/week, which works out to about 4 x 10^3 L/year. Then I assumed total oil imports probably cover about double total vehicular fuel consumption.
vehicle gasoline ~= 3.9 x 10^7 vehicles x 4 x 10^3 L/year
~= 1.6 x 10^11 L/year
all imported oil uses ~= 2 x vehicle fuel
~= 3.2 x 10^11 L/year
15% decline ~= 4.5 x 10^10 LThen I assumed Australians buy new vehicles about once every 6 years:
new cars/year ~= 3.9 x 10^7 vehicles / 6
~= 6.5 x 10^6 cars/yearI assumed EVs used to be about 1% of cars purchased, and because the article said Chinese EV imports increased by 3x, maybe they are now about 3% of new cars purchased:
EVs in 2025-2026 ~= 6.5 x 10^6 x 0.03
~= 2 x 10^5 EVsI then assumed about 2.5 times that many EVs are now in regular use, about 5 x 10^5 EVs. If each requires no imported oil-based fuel to operate:
fuel displaced ~= 5 x 10^5 EVs x 4 x 10^3 L/EV/year
~= 2 x 10^9 L/yearThat is a large drop, about 2 billion liters, but it only accounts for about 1/22 of the 15% drop in fuel imports. It is not likely to be the chief explanation.
Calibration Score
Matt's Calibration Score: 35 / 100
Higher is better: earn points for accurate pegs, sound models, correct math, and a result close to the sourced answer. The image shows percent full of it: 100 minus the Calibration Score.
Pegs: 0/30. Population, fuel consumption, and import-flow assumptions moved the answer by large factors.
Model: 15/30. The model mixed total vehicle stock with new EV imports and annual gasoline flow.
Math: 0/10. The category mismatch created a substantial numeric miss even though the arithmetic itself was mostly straightforward.
Result: 20/30. The final qualitative result held: extra Chinese EV imports were unlikely to explain the full gasoline-import drop.
Grounding facts
One new EV matters a lot for one household, but the national fuel market is a flow measured in billions of liters. Tens of thousands of extra EVs can displace tanker-scale gasoline volumes and still explain only a small share of a national import swing.
The most important distinction is new EV imports versus total EV stock versus annual gasoline flow. Mixing those categories can create a persuasive but wrong story.
After checking sources
Check and recalibrate
The key correction is stock versus flow. The article is about additional Chinese EV imports, not the entire EV fleet. A rough path from the reported $2.5 billion import value is:
Chinese EV import value ~= $2.5 x 10^9
average import value per EV ~= $2.5 x 10^4 to $3.5 x 10^4
current Chinese EV imports ~= 7 x 10^4 to 1 x 10^5 EVsIf imports nearly tripled, the additional EVs relative to the earlier level are about two-thirds of the current level:
additional Chinese EVs ~= 5 x 10^4 to 7 x 10^4 vehiclesNow estimate gasoline displaced per EV. Use 12,000 km/year and 8 L/100 km:
fuel per gasoline car ~= 1.2 x 10^4 km/year x 8 L/100 km
~= 1.0 x 10^3 L/yearSo the additional Chinese EVs displace roughly:
fuel displaced ~= (5 x 10^4 to 7 x 10^4 EVs) x 1 x 10^3 L/year
~= 5 x 10^7 to 7 x 10^7 L/yearA good central answer is about 70 million liters per year. If you use a broader EV import estimate, the range might be 50 million to 100 million liters per year.
Convert the reported gasoline-import drop:
gasoline import drop ~= 0.9 million tonnes
gasoline density ~= 0.74 kg/L
liters ~= 0.9 x 10^9 kg / 0.74 kg/L
~= 1.2 x 10^9 LCompare the two:
EV share of import drop ~= 70 million L / 1.2 billion L
~= 6%So additional Chinese EVs might explain a small, visible slice of the reported decline, but not most of it. Other factors like refinery operations, inventories, economic activity, fuel prices, driving patterns, hybrids, non-Chinese EVs, and import timing likely explain the majority.
For tanker scale, use about 50 million liters for one medium product-tanker shipment:
tanker shipments avoided ~= 70 million L / 50 million L/tanker
~= 1 to 2 tanker shipments/yearPost-check reflection
Matt's reflection
My biggest errors were a doubled estimate of Australia's population, a 4x overestimate on annual fuel consumption, and a 40x overestimate on the decline in fuel imports. That was mainly because I was considering all potential oil imports on an over-inflated car-count estimate, not just fuel imports.
Looks like I got pretty close on the size of the EV fleet, but only because I underestimated EV share of the total vehicle fleet and lucked into the right number. If my total car count had been correct, my EV fleet estimate would have been much smaller.
Ultimately my final answer held: EVs, and Chinese EVs in particular, are not likely to explain the drop in fuel consumption or imports, and we should look to other explanations. This is a great example of a plausible-sounding explanation that does not hold up to a bit of additional scrutiny, and it should give us pause when considering other plausible but untested explanations that support our darling ideas about how the world works.
Recommended memory peg
Remember Australia has about 28 million people, an average gasoline car might use about 1,000 L/year, and 0.9 million tonnes of gasoline is about 1.2 billion liters. For EV displacement: fuel displaced = EVs x km/year x L/100 km / 100.
Reader results
Bars show how submitted estimates sort into the answer choices from the gut-check prompt.