This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


An Electric Car Revolution Will Require Perpetual Subsidies

Schalk Cloete's picture
Research Scientist Independent

My work on the Energy Collective is focused on the great 21st century sustainability challenge: quadrupling the size of the global economy, while reducing CO2 emissions to zero. I seek to...

  • Member since 2018
  • 999 items added with 341,231 views
  • Apr 9, 2018 2:00 pm GMT

Your access to Member Features is limited.


  • Battery electric vehicles (BEVs) can be economically attractive as commuter cars in city traffic.
  • Attractiveness fades for longer-distance travelling due to larger battery packs, a smaller efficiency advantage, and more fast charging at on-peak rates.
  • As a result, displacing just a low single-digit percentage of global oil consumption through BEVs will require perpetual subsidies. 
  • A strong argument can be made that future overall cost reductions could be greater for hybrids than BEVs.
  • Autonomous driving technology does not change this outlook. 


This article follows up from a widely read earlier article on why a fully electric transportation future is unlikely to arrive any time soon. Battery electric vehicles (BEVs) certainly represent an important wedge in any transportation emissions reduction plan, but they are no holy grail.

BEVs have the same ideological appeal as wind and solar power. They are very easily marketed as perfectly clean and sustainable alternatives with rapid cost reductions that will soon relegate dirty fossil fuels to the dustbin of history. As in the case of wind and solar, this ideological appeal has attracted a broad fan-base and an equally broad range of technology-forcing policies.

Similar to wind and solar, however, the truth is that market penetration beyond a certain, relatively modest, level will require perpetual subsidies. A previous article quantified this point for wind and solar. This article will do the same for BEVs.

BEV saturation point

All energy technologies have strengths and weaknesses. That is why a healthy mix of technologies is generally the optimal solution, allowing each technology to serve the market segment it suits best.

BEVs have totally different strengths and weaknesses to ICE vehicles, implying that this technology class will diffuse easily into certain market segments, but have a much tougher time in others. Indeed, BEVs will penetrate the mobility market according to the well-known S-curve, struggling not to saturate well below 10% of current oil consumption.

Greater clean mobility contributions will likely come from much less hyped solutions such as car-free lifestyle options, efficiency improvements (including hybrids), sustainable fuels, conservation (downsizing), and fuel cell technology. It is vital that we leave ideology aside and replace current BEV technology-forcing policies with technology-neutral policies creating a level playing field for all clean mobility options.

Quantifying BEV competitiveness

As an illustrative example, we will calculate the combined fuel and drivetrain costs of a BEV and a hybrid as a function of driving patterns. The following graph shows the assumptions employed.

Here we assume that BEVs will be available with ranges between 200 miles (sales below this range have been poor even with large incentives) and 500 miles (comparable to a conventional car). At the low end, BEVs will be used as commuter cars, accumulating 10000 miles per year. At the high end, BEVs are often used for longer trips, racking up 15000 miles per year. Based on fuel economy of the Hyundai Ioniq hybrid and electric versions, BEVs are assumed to be 3x more efficient than hybrids at the low end (primarily city driving) and 2x more efficient at the high end (primarily highway driving).

Finally, electricity is assumed to cost $100/MWh at the low end due to off-peak charging at home (cost of home charger included), but $180/MWh at the high end due to more peak-time charging at more expensive fast charging stations (half way between home charging and $260/MWh at Tesla’s “non-profit” California superchargers).

Other important assumptions are as follows: BEV and hybrid drivetrain costs of $2500 and $7000 respectively (1, 2, 3), BEV efficiency of 270 Wh/mile (including charging losses),  gasoline price of $2.1/gal before taxes ($60/bbl oil), 20 year lifetime, 6% discount rate, and 30% gross margin on drivetrain and battery packs.

Three technology improvement levels are assessed for BEVs and hybrids. For BEVs, it is all about battery pack costs over a range of $50/kWh to $150/kWh. For hybrids, a scenario of 50% higher efficiency as well as a scenario of 50% increase in efficiency and power/cost ratio is assessed. The results look as follows:

BEV vs hybrid drivetrain and fuel costs

As shown, BEVs just start to become competitive at the low end for the three different levels of technology advancement. Thus, BEV economics are attractive when the required range is small, driving occurs in stop-start city traffic and charging happens at home during off-peak hours. On the other end of the spectrum, BEV costs are generally more than double hybrid costs.

It should also be mentioned that future car-free lifestyle options may well remove a large chunk of demand for shorter city trips where BEVs are at their most attractive.

Technology uncertainties

BEV technology development is heavily focussed on batteries. Simple extrapolation of learning curves (e.g. the method used in BNEF‘s projections) yields impressively low numbers:

However, lithium ion batteries face harsher physical limits than microchips or PV cells. Even though cumulative battery production will need to increase 1000-fold (10 doublings) from today’s level to displace 10% of oil consumption, lithium ion batteries are already approaching their commercially achievable energy density limits. Recent impressive cost declines were primarily due to the establishment of global value chains and economies of scale, and are not repeatable. Future cost reduction efforts are therefore likely to be hampered by diminishing returns as technology development encounters physical limits.

More importantly, lithium ion batteries rely on several relatively rare technology metals. Given the absolutely enormous scale-up that will be required to have a substantial impact, this potential limitation is getting increased attention. The recent spike in the price of cobalt, a critical technology metal produced primarily in the unstable Democratic Republic of Congo, is shown below as an example.

Interestingly, oil offers a very good analogy in this case. At the start of the ICE revolution, oil was very cheap. One simply needed to drill a hole in the right place and oil would just come gushing up. At that point, oil was a minor component of gasoline prices next to refining and distribution costs. This is the point where we are now with BEVs: technology metals are a relatively minor cost in battery packs.

Of course, we all know what happened next. Despite continued technological advancements, oil prices eventually increased 5-fold as demand boomed and the easy resources dried up, making oil the dominant component in gasoline prices. The same is likely to happen with technology metals. At current spot prices, only the raw materials required for Tesla’s battery cathodes already cost $50/kWh. If the oil price history is anything to go by, material costs could really spoil the battery cost reduction story going forward.

As a result of physical limits on technological advancement and technology metal availability, sustainable production of $100/kWh lithium ion battery packs may well be impossible, with even the $150/kWh level proving difficult to maintain at the massive production volumes required to make a significant dent in oil consumption. Of course, oil also offers a good example of the potential geopolitical implications of technology metals.

I am therefore less confident about the optimistic technology assumptions for the BEV than the hybrid in the graph repeated below. Hybrids still have significant headroom for improvements in efficiency and cost, and are much less exposed to battery material limitations due to much smaller battery packs.

BEV vs hybrid drivetrain and fuel costs

As an example, Mazda is targeting 56% thermal efficiency in the longer term for its new SPCCI engine technology illustrated in the video below (current hybrid engines are about 40% efficient). The first generation SPCCI engine is due to enter commercial production next year and beta version test drives have been encouraging (1, 2, 3).

As outlined in an earlier article, my view is that such advanced combustion engines will eventually end up in a hybrid configuration where the electric motor is much more powerful than the engine. This will allow the engine and transmission to be strongly downsized, saving costs and allowing for operation within the optimal operating range almost all the time. As a result, efficiency, reliability and longevity will go up, while cost and emissions will go down. A drivetrain cost breakdown from the previous article is repeated below as an example.

Such a hybrid configuration can conceivably increase overall efficiency by 50% over the current state of the art with no cost penalty relative to conventional ICE cars, thus achieving the most optimistic hybrid technology scenario considered in this study. The most optimistic BEV scenario, on the other hand, will most likely require a new battery technology. If such a technology emerges, it will take 2-3 decades and hundreds of billions of dollars to establish a new fully cost-optimized global value chain and scrap outdated lithium ion battery infrastructure.

Autonomous vehicles

The previous article outlined why autonomous driving technology may well favor ICEs more than BEVs. Besides, the competitiveness picture does not change much if we increase the distance traveled per year by a factor of 5, reduce the vehicle lifetime to 7 years and add $2000 for autonomous hardware costs:

The hybrid will probably perform even better than suggested by the above graph. Autonomous traffic flow will be much smoother, allowing hybrids to also achieve highway efficiency at the low end. Also, such high-utilization applications will encourage developers to further enhance hybrid efficiency through waste-heat recovery systems, thus further reducing fuel costs. The ability of hybrids to refuel very rapidly at any time (without having to worry about electricity price fluctuations) will also be a significant advantage in an autonomous fleet.

That being said though, I maintain that we are unlikely to achieve broad deployment of full autonomy within the timeframes demanded by climate science.


Adding all of these observations together, it is difficult to see BEVs displacing more than 10% of LDV fuel consumption (2% of oil consumption) without perpetual subsidies. Displacing 10% of oil consumption may be possible at a substantial subsidy cost, but pushing beyond that point really sounds increasingly wasteful.

As mentioned earlier, substantially larger cuts in oil consumption and greenhouse gas emissions can be achieved through a range of other channels such as car-free lifestyle options, efficiency improvements (including hybrids), sustainable fuels, conservation (downsizing), and fuel cell technology. It will be a real shame if we get so distracted by BEVs that we fail to harness the great potential of these clean mobility options.

Schalk Cloete's picture
Thank Schalk for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Schalk Cloete's picture
Schalk Cloete on Apr 14, 2018

Yes, it is clear that fossil fuels have substantial externalized costs. Relative to efficient hybrids, however, BEVs do not offer significant benefits in this regard (see my earlier article:

The issue of oil wars is very tricky and subjective. As mentioned in the article though, the battery technologies currently being scaled up will also impose large energy security concerns based on the materials they require if they are to make a significant dent in oil consumption.

I should also point out the most important benefit of oil: its great economic surplus. Average global oil production, refinement and distribution costs are about $30/bbl. The economic value of of oil is probably around $200/bbl (rough estimate of the oil price where economic growth stalls). This results in an economic surplus of about $170/bbl, which sums to about $300 trillion to date or $350/ton of CO2 emitted.

I agree that most car trips are short. However, people buy cars so that they have total freedom to go wherever they want whenever they want. A short-range BEV simply does not give this freedom, putting it in the same league as public transport (at a much higher cost). Obviously, sales of low-range BEVs have been poor despite subsidies that are much larger than the cost of a small battery pack:

Schalk Cloete's picture
Schalk Cloete on Apr 14, 2018

My previous article showed that BEVs offer no significant reduction in externalities relative to hybrids:

Wireless charging is certainly interesting from a practicality point of view, but its deployment will probably be prohibitively expensive. Its lower charging efficiency will also erode much of the efficiency benefit of BEVs.

Engineer- Poet's picture
Engineer- Poet on Apr 14, 2018

The biggest downside to hydrogen is that the round-trip efficiency from electricity, “renewable” or otherwise, is about 40%.  This strongly favors steam-methane reforming over any non-fossil source.  SMR also deals with the storage issue, mostly.  Time and tech have not changed Dr. Ulf Bossel’s conclusion.

More transportation fuel is used for these long/heavy applications than for short-distance light duty vehicles.

Perhaps in Europe, but in the USA it’s the exact opposite:  9.3 million bbl/d of motor gasoline vs. 2.99 million for distillate (out of 3.94 million bbl total distillate consumption).

storing excess wind/solar energy as hydrogen is more practical than using it to power BEVs.

It really depends on the duration of storage.  Batteries are far more practical than hydrogen for buffering on the scale of hours.

In a solar dominated system, electricity will be cheap in the middle of the day, which is a much less practical time to charge a car than during the night.

Perhaps not AS practical, but most commuter vehicles are driven in the morning and evening and are parked around noon.  Parked vehicles make very good sinks for excess generation.

The problem is the size of the sink required.  I recall calculating that levelling Germany’s PV peaks would require on the order of half a Chevy Volt per capita (considerably less in Tesla-equivalents).  Current efforts to build BEV fleets fall woefully short of anything sufficient to solve this problem.

This doesn’t mean hydrogen is a solution.  Having to build 2.5x as much generation to get the same net energy out is a major hit to EROI as well as LCOE.  Last, H2 is a very slippery molecule with low volumetric energy density and is a favored food of certain bacteria (whereas methane is a waste product of anaerobes).  For gaseous fuels, geological storage is the only possibility for levelling over seasonal variations and it’s not going to be possible to hold H2 in many formations currently used for natural gas.  Making this work is going to be far more difficult than the status quo if it’s possible at all.

None of these things are secrets.  What I draw from this is that the current dogma is a failure because it is intended to fail.  It is a creation of fossil-fuel interests, who finance the “environmental” lobby and calls its tune.  Any real solution is going to (a) require a different technical direction (almost certainly aimed at making room-temperature liquids, not gases) and (b) a very different set of interests to back it.

faster charging of BEVs can easily become very expensive.

This really doesn’t matter if most parked vehicles are on the grid.  The occasional fast charge when going far from home gets swallowed in the savings from not paying the 3x premium for the losses of power-to-gas-to-power.

Engineer- Poet's picture
Engineer- Poet on Apr 14, 2018

(1/2) It pains me when someone who has obviously never studied the subject tries to lecture on the physics of AC power transmission.  Nothing about is intuitive except to someone like Tesla.  Case in point:

Local grids are “constant voltage like a copper sheet” (ignoring step ups & downs); power flow is varied by varying the impedance of the load (same as for HVDC long distance), which assumes the distance is short enough that the phase shift in transit can be ignored or corrected

First thing:  AC power does not travel from the point of greater to lesser voltage.  It travels from the point of leading phase to the point of lagging phase.  AC lines (of all sizes) have complex impedance but are most easily modelled as a pure inductor.  If you have the same phase at both ends, no real power will flow along the line.  You will have reactive power (VARs, volt-amperes reactive) flow from higher to lower voltage, but no actual energy.

What happens when you turn on an AC load is that the local phase sags enough to transfer the energy to feed it.  When you turn it off or add local generation, the phase advances.

You can generate VARs without a watt of power.  By convention, capacitor banks generate VARs (inductive loads consume them).  VARs provide voltage support; you can actually have a higher voltage at the load end of a line than the generating end by adding a bunch of VARs at the load.  The flow of real power and VARs can be in opposite directions (NOTHING is intuitive about AC transmission!).  Synchronous condensers generate VARs.  They provide power-factor† correction and voltage support, not phase shifting.

if the line gets too long, you add phase shifters of some sort as Bas suggests (e.g. synchronous condensers).

Wrong.  Many transformers shift phase as a normal part of their operation.  Three-phase transformers are connected either as delta (3 windings connected like a triangle) or wye (3 windings connected at a common point on one end and to the line at the other).  A transformer can have one side delta-connected and the other side (usually the higher-voltage side) wye-connected.  A delta-wye transformer has a 30° phase shift between the two sides.  (Even a wye-wye transformer will have a delta-connected “idler” winding to dissipate the 3rd harmonic currents generated by non-linear magnetic properties.)

Phase-shifting transformers are something completely different.  They allow the phase relationship between input and output to be varied on demand.  They are used to control the flow of real power across lines; by making one end appear to have a different phase than actually exists at that point in the grid, the flow of power can be changed over what would otherwise be the case.  Germany’s neighbors have used such transformers to prevent Germany from dumping its own surges of intermittent power across borders and thus externalizing the problem of managing them.

Fixed capacitor banks in series with a long line can offset the line’s inherent inductance (preventing large phase shifts at small currents which limit the power capacity), but using such tricks risks problems like sub-harmonic instabilities in the system.

you’re correct that HVDC terminals costs more than HVAC terminals, and making bidirectional terminals costs more than single directionals. That will likely always remain true.

That’s only part of the issue.  Point-to-point links are not a GRID and can’t substitute for one.  In theory, you could have a number of interlinked HVDC lines with circuit breakers and everything else running along the eastern seaboard and west along the St. Lawrence to Chicago and points west.  There’s no intrinsic limit to how much power such a network could carry; it could provide firm transmission capacity from wind farms west of Des Moines straight to NYC and Boston… but no such thing is even in planning.  From all appearances, we lack the technical ability to manage what appears to be a rather simple DC system you could model at scale on a tabletop.

† Power factor is the ratio of real power to volt-amperes; total VA is the root of the sum of the squares of watts and VARs.  Capacity of transformers and losses in resistance scale as VA, not watts; things are most efficient at a power factor of 1, with no VARs moving at all.  The more reactive power travelling through a line or transformer, the lower the power factor.  It makes sense to generate VARs as close as possible to where they are consumed, to avoid losses in points between.

(continued for length)

Engineer- Poet's picture
Engineer- Poet on Apr 14, 2018

(2/2) Utilities bill for VARs; VARs don’t consume watts directly, but they require more transformer capacity and lose energy in transit.  Large industrial users will often use capacitors or synchronous condensers to offset the VARs consumed by their induction motors and minimize their cost for VARs.

Don’t get me started on induction generators, especially not capacitor-excited induction generators.  That’s a whole ‘nother can of worms.

Confused yet?  The beauty of our system is that despite being so non-intuitive, it’s wonderfully easy to use.

Sean OM's picture
Sean OM on Apr 16, 2018

Hydrogen has several other advantages over BEVs aside from faster refill. The first is that an H2 storage tank is much cheaper (and lighter) than a battery. Thus, for transport applications over long distances and/or with heavy loads (both of which require large quantities of energy stored on-board), FCVs will be cheaper than BEVs. More transportation fuel is used for these long/heavy applications than for short-distance light duty vehicles.

The definition for long haul in europe is 200-300km (125-185 miles), and in the US it is generally over 250 miles. The US trucking industry uses about 400B gallons of distillates a year. A lot of the trucking especially corporate is short haul.

The Tesla semi will supposedly have a range of 300 (480km) or 500 miles (800km), and can be refilled without additional permitting at all destinations, and very little special equipment.

EngineerPoet really covered it well. HFC really is greenwashing by the NG industry. You basically have to rebuild the NG infrastructure which I am sure they are waiting for lots of government handouts. Then you start to figure out it is 1/3 cheaper to use NG.

Engineer- Poet's picture
Engineer- Poet on Apr 16, 2018

The US trucking industry uses about 400B gallons of distillates a year.

Try 2.990 million bbl/d * 365 days * 42 gal/bbl = 45.8 billion gallons/yr.

Sean OM's picture
Sean OM on Apr 16, 2018

I know what you are saying. Your math is correct. The only reason why I could think it is off is political as I lifted that number from the American Trucker Association website. “It also takes over almost 398 billion gallons of diesel fuel to move all of that freight. ”

Geoff Thomas's picture
Geoff Thomas on Apr 18, 2018

An extremely small, (ie chain saw size motor) dc generator, can generate power, it may not be able to drive the car, it only has to charge the car batteries, when the batteries are low, the owner starts it, – may be connected to the drive chain, may be manual, when he is low, it then increases the range, and when the batteries are lower the owner has to pause a while whilst the small (probably two stroke) (could be bolted to the rear bumper bar etc) petrol device brings up the batteries to proceed.
Ideologically – so what, gets you there.
Most motoring is very small range, a cheap extender is the best solution.
Schalke is a bit too black and white, imho.


Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »