Published on December 25th, 2017 | by Dr. Maximilian Holland0
Timeline For Electric Vehicle Revolution (via Lower Battery Prices, Supercharging, Lower Battery Prices)
December 25th, 2017 by Dr. Maximilian Holland
Originally published on EV Obsession.
Everyone knows that electric vehicles (EVs) are going to replace internal combustion engine vehicles (ICEVs) in the long run. Many of us are excited about this key transition away from fossil fuels and hope that it comes sooner rather than later, yet are not sure exactly when the big breakthroughs in market share are going to happen. It is often stated that at battery prices of $100/kWh, EVs will successfully compete with ICEVs, but mainstream predictions about how and when this happens vary widely. In this in-depth article, we are going to look in more detail at the figures relevant to different vehicle segments, estimate the most probable timeline for feature and price parity in these segments, and offer a counterpoint to the more conservative timelines that we see from both incumbents (OPEC) and progressives (BNEF) alike. We know the EV disruption is real, since it is already well underway in the premium sedan segment.
The bottom line is this: EVs will be better and cheaper than ICEVs at the value end of the biggest volume segments (small SUV and compact car) by 2022. At the higher-priced end of these segments, EVs will already be at parity by 2019. By 2024–2025, EVs will outcompete on both features and price in pretty much every vehicle segment (except rocket ships).
Feature Parity — What’s Left?
Let’s unpick feature parity. EVs are already superior in many respects to their ICEV counterparts. However, as of 2017, the sticker price for EVs competing in volume segments is typically higher (for most buyers sticker is more influential than total cost of ownership), and there are two remaining features that have room for improvement relative to ICEVs. These are road-trip-ability and fast recharging. Specifically, potential buyers who have gotten used to ICEVs like to own a vehicle that can make road trips — i.e., be driven on the highway for a good 2 or 2.5 hours (a safe duration), take a short rest break (15 minutes), and then get back on the road for another couple of hours (and repeat as necessary).
Even though road trips are in reality a tiny fraction of journeys, the automobile has come to represent personal freedom and the ability to go anywhere at anytime. A vehicle that can’t do this (as an ICEV does it) just won’t cut it for many folks. CleanTechnica’s survey research underlines buyers’ preference for these two key features, even for EV enthusiasts. For true mass-market sales, these features are essential. The good news is that the right mix of price, long range, and fast charging will be here soon. Much sooner than mainstream analysts predict.
Fast charging is something of an independent feature (whereas price and range are closely bound together, as I cover below). In practical terms, to approximate ICEV refueling speeds, fast charging during road trips means 350+ kW chargers, and vehicles able to accept such charge. These charge rates allow a vehicle to add 175+ miles (282+ kilometers) of range in 15 minutes. 175 miles is equivalent to 2.5 hours at 70mph (113kph). Bear in mind that this 175+ miles should equate to 80% of max battery capacity, since the flow of electrons during charging slows down as you approach 100%, and time increases disproportionately. Charging to 80% is the sweet spot for rest breaks on road trips. 15 minutes is a comfortable (and widely recommended) minimum break duration if you’re in the middle of a long journey. Enough time to visit the bathroom and grab a coffee and a bite to eat.
Some may argue that we sometimes average more than 70 mph on a highway, and sometimes can drive for more than 2.5 hours between breaks, which are sometimes as short as 10 minutes. Fair enough, but most sensible folks, especially those with families on board, don’t significantly deviate from these speeds and durations most of the time. Given that a road trip is a very occasional use case anyway, these criteria will be comfortable for the majority of buyers. The small percentage of hard-core road trippers can wait an extra one or two years (literally 1 or 2 years) for the range/price that works for them. For the majority, then, 175 miles range at 80% charge means 220 miles range at 100%. What battery size and cost enables this range depends on the type and size of vehicle, and evolving battery pack prices per kWh, which we will now dive into.
Now let’s look into price parity. I have previously written that to compete on upfront sticker price with an equivalent ICEV, EVs have a margin of about 15% of the total vehicle cost to give over to the battery pack cost. This is the main basis of the frequently cited estimation that batteries must reach $100/kWh for EVs to compete with ICEVs. Basically, while powertrains on ICEVs and EVs are different, much of the rest of the vehicle (chassis, interior, paneling, paint etc) are equivalent. ICEV powertrains are inherently more complex and expensive than EV powertrains. The ICE, transmission, exhausts, and all the peripherals cost around 22–24% of the total vehicle cost (see: McKinsey & Co. 2012, Kochhan et al. 2014).
Ever tightening emissions regulations are making ICEVs ever more expensive, whilst volume and learning are making EV powertrains ever less expensive. Vibration tolerances, engine heat and noise insulation, simplifying auxiliary systems (alternators, etc.), and other savings can also be made by EVs. However, energy storage on an ICEV is basically a cheap gas tank with a pipe attached. An EV battery is much more expensive up front. In the medium term, as EV powertrain production increases and efficiencies are found, when the dust settles, expect this figure of 15% to hold. As a note, this is a key area where most forecasters get the figures wrong, some continuing to calculate that as of 2017, batteries inevitably account for up to 50% of vehicle cost. On the contrary, doing the math, at the end of 2017, Tesla’s Model S 100D needs at most 16% of the vehicle price ($94,000) to cover the cost of the battery pack ($15,000 or under).
Given all of the above, battery pack prices and pricing trends are obviously the crucial element in overall EV price parity with ICEVs. What are battery prices currently and what is their price reduction trend? Most keen EV watchers are aware that, in October 2015, GM stated it was paying LG Chem $145/kWh at the cell level. Factoring an additional 31% to reach battery pack prices (this multiplier will reduce with time), that’s $190/kWh pack price. June 2017, 20 months later, Audi revealed it is paying $114/kWh cell price (likely sourcing from Samsung SDI, LG’s main competitor). Converting to a $150/kWh pack price, that’s a 21% price reduction from the big Korean suppliers over those 20 months, so around 14% per year. Tesla revealed in April 2016 it was below $190/kWh at pack level.
Tesla plays its battery cost cards very close to its chest, but its November 2017 guidance on semi truck pricing, for sale in late 2019, implies a projected battery pack price of $75 to $90/kWh by that date. That’s a 50% to 60% price reduction over 3.5 years, around an 18% cost reduction per year. Given this, calculating that these largest EV battery players at the end of 2017 are below $150/kWh and achieving 15% cost reduction per year over the next several years seems a safe bet. Doing the math, that means breaking the $100/kWh barrier between 2019–2020, $70/kWh between 2021–2022, $50/kWh between 2023–2024, and so on. Battery production volume is quickly increasing, R&D investment is increasing, and we are still pretty early on the technology maturation curve. Recall the history of solar PV pricing? Batteries will likely have a similar journey.
We know that in 2017 a premium sedan like the Model S outsells its ICEV counterparts at a price which already gives enough margin to cover battery costs (of even a 330 mile range battery). But premium sedans are at most 2–3% of the global auto market. The real disruption will occur when EVs out-compete in the much higher volume “compact SUV” and “compact car” segments that make up at least 35% of the market. What size of battery do they need to achieve 220+ mile (road-trip-ready) range? And when will the cost of such a battery enable them to reach sticker price parity with equivalent ICEVs?
So let’s do the math. Premium-end small/compact SUVs (Audi Q5, Jaguar F-pace, BMW X4, and the like) currently sell for an average price of around $51,500, and the larger volume value-end examples (Honda CRV, Ford Escape, Jeep Cherokee, VW Tiguan, and similar) sell for a midpoint of around $29,000. Allowing for drivetrain and aerodynamics efficiency learning, a 65 kWh battery pack on an efficient platform will give these sized vehicles a range of 220–240 miles, thus allowing 175+ miles between 80% fast charges. At the higher end, allocating 15% of the price to pay for the battery means that by 2018–2019, an EV version costing $51,500 with a 65 kWh battery (at $119/kWh pack price) will reach sticker price parity with its ICEV competitors. At the value end ($29,000), this parity with ICEV versions will be achieved by 2022 (65 kWh pack at $67/kWh). Road trip junkies who insist on a 75 kWh battery good for 260–270 miles range will find one by 2023 at $29,000.
In the compact car segment (C-segment), higher-end examples (Audi A3, BMW 2 series, Alfa Giulia, etc.) sell for an average sticker of $40,000, and value-end models (VW Golf, Ford Focus, Mazda 3, Honda Civic, and others) sell for an average of $23,550. At this vehicle size, the Hyundai Ioniq EV released in 2016 already achieves an EPA range of 124 miles from 28 kWh, suggesting that 45 kWh would achieve a 220 mile range. Likely efficiencies can gradually improve further in the coming years. To be conservative, however, I assume 50–55 kWh would give a 220–240 mile (354–386 km) highway range. At the higher end, allowing 15% to pay for the battery means that by 2019, an EV C-segment vehicle costing $40,000 with a 55 kWh battery pack (at $109/kWh) will reach sticker price parity with its ICEV counterparts. At the value end, a $23,550 vehicle with a 55 kWh battery pack (at $65/kWh) will be achieved by 2022.
So, in summary, assuming battery pack prices continue a 15% annual cost reduction trend (safe bet), and most consumers are happy with the road-trip-ready features we have specified (safe bet), value-end small SUVs and compact cars will reach the same price as ICEV equivalents by 2022. At the higher end of both of these segments, the parity will come in 2019. Across these segments, the EV version will feel more responsive, smoother, quieter, and more refined than its ICEV counterpart, will have much lower running costs, will offer much higher reliability, and will be the more desirable vehicle. The only remaining uncertainty is around when (not if) 15 minute charging becomes available. The infrastructure is already being built out in Europe by the IONITY consortium, which includes Audi, BMW, Ford, Mercedes, and Porsche. E.ON has announced plans for a large European charging network, with 350 kW options, to be in place by 2020. Similar initiatives are underway in the US. Obviously, automakers also plan to release vehicles that can charge at these rates, but the release timing of such vehicles is currently an unknown, except for Porsche, which will release such a vehicle in 2019. Tesla is also working towards 350+ kW chargers and vehicles.
The small SUV and compact car segments analysed above are the mass-market segments in global auto sales. Any segments more expensive than these will be disrupted at least a year or so sooner (e.g., the midsize/D-segment that the Model 3 already has some half-million buyers queued up for). The remaining segments such as the smaller B-segment and less expensive categories (with correspondingly less budget to allocate to batteries) will reach parity a year or two later. By 2025, there will be no auto segment where an EV with all the convenience features of an ICEV equivalent, plus all the advantages inherent to EVs, cannot be offered at sticker price parity.
After 2025, EVs in every category will only get cheaper whilst ICEV versions will get more expensive both to purchase and to run and maintain. Cities and states will increasingly feel obliged to outlaw ICEVs for their local and global pollution effects, since no reasonable argument will remain for their continued existence. Will folks still buy ICEVs? Some might do so for sentimental reasons. Will the oil industry enter its final death spiral? Surely. Will analysts still claim that none of this is happening whilst the tidal wave swamps over them? Very likely.
Mainstream analysts have for years been well behind the game on accurately describing the developments in battery prices, and underestimating the learning/cost reduction curve. The kindest excuse is that this is due to a focus on “average” cost rather than those from leaders in the space. In December 2017, Bloomberg New Energy Finance (the most bullish on cleantech among mainstream analysts) still touts a figure of $209/kWh pack price, and $100/kWh by 2025. This is more than 2 years behind the above pricing reveals by Tesla, GM/LG, and Audi/Samsung, and 5 years behind reality for the $100/kWh figure. Given that market pricing has to compete with disruption from the volume leaders, not the uncompetitive laggards, its not clear why “average” figures that include the latter have any relevance. These kinds of dullard pricing analyses then lead to EV market penetration timelines of, for example, 14% of new sales by 2030 (BCG) or 50% in 2040 (BNEF) even at the most radical end. OPEC, Exxon, and others obviously have no choice but to keep repeating that EV sales won’t be significant anytime soon. If they don’t make such claims, investors will bail. They should be ignored.
An analyst source more in tune with real-world battery cost trends seems to be Wards Auto, which figures on $100/kWh by 2020 and the learning curve continuing on the same cost reduction rate thereafter. That’s in line with the reality of statements from major players we noted above. I consider the estimates I have outlined and the underlying math behind them to be fairly robust, but others may feel they are either too conservative or too radical. And yes, of course self-driving vehicles will be an even more significant disruption. Please share your considered thoughts in the comments below.