Source: https://rmi.org/insight/breakthrough-batteries/ According to evidence detailed in RMI’s Breakthrough Batteries Report, cost and performance improvements are quickly outpacing forecasts, as increased demand for electric vehicles (EVs), grid-tied storage, and other emerging applications further fuels the cycle of investment and cost declines and sets the stage for mass adoption. The report examines how an increasingly electrified, Li-ion battery-dominated world in the near term will open, in the longer term, significant new market opportunities for other emerging battery technologies that are nearing commercial readiness. RMI’s analysis of emerging battery technologies identified six categories (in addition to advanced Li-ion) with significant potential for achieving commercial production by 2025. The below figure shows companies’ anticipated technology commercialization timelines. This is a positive report but I've had credibility problems with the Rocky Mountain Institute in the past. IMHO, they are a little too surface skimmers rather than thinking the problem through. Still, a gift horse. Bob Wilson
This article appears to have some factual basis, based on manufacturing economies of scale. However it also anticipates an 8 X increase in Energy densities or capacity (performance of battery) in a period of 10 years, which is very ambitious and may not be realistic. However, while I cannot dispute the basic premise of the article, that there will a significant decrease in the price of batteries, the amount of decrease they claim will happen, is startling. Many decades ago, Gordon Moore (co-founder of Intel and Fairchild Semi-Conductors) came up with what is know as More's law. Moore's law is the observation that the number of transistors in a dense integrated circuit doubles about every two years (https://en.wikipedia.org/wiki/Moore's_law). What it says is that performance more or less doubles every two years and reality tracked this for many years, though the pace has slowed down in recent years. Now battery chemistry does not follow Moore's law exactly, that performance does not double every 2 years. In an older report (2007) MIT report expects that battery performance will double over a decade rather than 2 years (https://www.technologyreview.com/s/407345/nanobatteries-and-moores-law/) Meanwhile, advances in energy capacity and calendar life are coming from improvements in electrode materials, sometimes using nanoscale particles. (See “3M’s Higher-Capacity Lithium-Ion Batteries,” “Powering GM’s Electric Vehicles,” and “Battery Breakthrough?”) These might lead to a doubling of energy capacity within a decade, which could go far toward improving electronic devices and cars. Battery performance could double in the next 10 years, according to one MIT scientist. (See “How Future Batteries Will Be Longer-Lasting and Safer.”) That’s no Moore’s Law, but, combined with more-efficient devices, it could make a big difference. One part of Rocky Mountain study makes sense. If there is more demand, larger plants are set up, there is more investment in manufacturing technology, there are economies of scale and this drives down cost. There is no expectation here that the base technology has changed, only that it has become cheaper to make. However as the article points out, more interest in the product, more interest in improving the base technology. In the electronics world, by increasing circuit density, manufacturers were able to substantially improve performance and miniaturize devices without a proportionate increase in costs. So cost of ICs came down both due to manufacturing efficiencies (large volumes as the ICs were used more) as well as to the effect of higher packing densities. This is shown in the decline in the costs of the Lithium-Ion batteries over time as shown in the graph below. Most of it is due to Wrights law on volume driven efficiencies, a small portion could be due to increases in battery density, estimated at about 3% a year (source “The equivalent of Moore’s law for batteries is that they improve about 3% every year,” says Mike Toney, a researcher at Stanford Synchrotron Radiation Lightsource who works on battery technology. https://www.forbes.com/sites/mikemontgomery/2018/01/11/get-ready-for-the-battery-revolution/#4370c6db19ca) -2 ) (https://ark-invest.com/research/wrights-law ) While a Moore’s Law style forecast deemed lithium-ion battery technology mature more than ten years ago, Wright’s Law correctly anticipated a reacceleration in cost declines and a resurgence in demand roughly five years ago. The decline in prices has opened up new segments of the auto market to lithium-ion batteries which, in turn, is pushing them toward an even larger market, utility-scale energy storage. While the cost decline in batteries after the launch of Tesla’s Model S appears discontinuous when presented as a function of time, when recast using Wright’s Law – costs presented as a function of unit production – the cost drop appears neither discontinuous nor particularly surprising. So I can see that part of Rocky Mountains Institute hypothesis has factual basis. The second part of the Rocky Mountain article is where Moore's Law analogy comes in is the battery density i.e. improving performance by having a battery be able to hold more charge in the same volume. Experts now believe that energy density or capacity of batteries will improve at greater than the 3% a year where it has historically been (https://www.forbes.com/sites/mikemontgomery/2018/01/11/get-ready-for-the-battery-revolution/#4370c6db19ca), due to the tremendous interest in improving battery performance through higher charge densities. However the Rocky Mountain study is looking for an 800% improvement in 10 years, which means annual average improvement rate of 23% a year for the next 10 years. Now it is not going to be linear, sometimes the growth rate in energy density will be faster, sometimes slower. However, what is assumed is significantly higher than the past. While I will be glad if an 8 fold improvement is achieved in 10 years, but it seems a rather daunting task about battery performance. Yes it happened in IC design, but will it happen in batteries? That is the 64 million dollar question.
My understanding of the 8x improvement from the RMI infographics is not that density will increase by 800%. Their prediction is density will increase by 100% (2x, which is the equilivent to approx 6-7% improvement per year over a ten year period), and price per cell will decrease by 75% (4x), for a cumulintive impact of the price per kWh being 8x cheaper. At some point this level of gains will not be repeatable and will start to trail off as we reach a natural limit based of the cost of materials. I wonder what this limit is?
OK, I see what you are saying and may be I was not looking at it right. A 6-7% increase in battery capacity year over year compared to a current 3% is not unreasonable. Where I have a concern is that both these factors are in a way correlated. The improvements in battery density and decrease in cost are not two mutually exclusive parameters. In electronics industry, a fair amount of cost decreases, especially when manufacturing improvements stalled came up from the density increases. So if you can put more charge into the same size battery, you have decreased cost per Kwh. It appears that Rocky Mountain institute claims that a manufacturer will be able to put twice as much charge in the same battery as today and in addition reduce cost by 75% from where it is today. So if the cost per KWh is about $110 in 2019, it will become about $14 by 2029, which seems very aggressive if you look at the curve. As you rightly point, there is a natural limit. The curve below shows the cost starts to increase after a tipping point, as other industries have shown there is usually a tipping point and hence the Moore's law is starting to fray while Wrights law appears to be more realistic. So the right question may be is "Will the tipping point occur before there is an 8X improvement?" In other words, there has been a dramatic decrease from $270 in 2012 to $110 in 2019, a 60% decrease in about 7 years. Will this continue that in 10 years we get a further decrease to $14 in 2029? This is about 95% decrease from 2012 to 2029. I hope so but am not confident that will happen, that somewhere along the line costs may start to go up. But it will continue to decrease for a few more years, one wishes that it will continue till 2029 and beyond. (https://ark-invest.com/research/wrights-law
I haven't read enough of RMI study, but I'm not going to assume their 8x cost reduction is from 2019. Because, as you say $14 per kWh is optimistic. I think they're applying this cost reduction from 2013 to 2023. In regards to the natural limit, my napkin math goes like this. Nickel $14 per kg Manganese $11 per kg Aluminimum $2 per kg Colbolt $80 per kg Lithium $8 per kg Colbolt is a small percentage so I'm going to round up the raw material costs to that of Nickel $14 per kg. Today we have an energy density of 250 w per kg, with a raw material cost per kWh of $56. We can estimate the high volume cell cost today is $100 per kWh, so approx 50% is raw materials and 50% is manufacturing and packaging. Future we target 500 w per kg, with a raw material cost per kWh of $28, volumes increase significantly so manufacturing and packaging equal $14 per kWh, making a price floor of $32 per kWh. So anyone projecting a price per kWh of around $50 over the next decade I would put in my believable basket. At $14 per kWh I'd struggle to believe it, but would welcome being proven wrong.
When we had Prius, I've had problems with Rocky Mountain Institute and was hesitant to post their reports in the past. I always felt they had some accurate data points that were too often diluted by simplistic analysis. Yet they were the only conservative group that even acknowledged the Prius and what it was doing. Although there are higher energy chemistries possible, they often turn out to be 'primary' batteries (i.e., not rechargeable.) I am retired but from what I'm seeing, the better approaches: high temperature, dry or glass electrolyte - the ions pass through a combination separator and electrolyte. So the lithium (or sodium) ions enter one side and pop out the other. A common cell failure mechanism is a porous, plastic separator overheating and shorting between the anode and cathode. anode - carbon is less dense but low ion storage; silicon higher density but specific volume problem between charged/discharged but; I've not read much about Germanium and Tin. If either avoids the density problem of Silicon or can dope the Silicon to moderate the expansion problem, yea! cathode - lithium is cheap but sodium and potassium are nearly free. Just they become a little heavy. Now if we could just fund a multivalence ion that doesn't become fixed between normal and ion state or precipitate as inert. Well that is how I understand the problem and I'm sticking with it. <GRINS> Bob Wilson
From what I see in the article, it is from a base of 2018/2019 and I do not find reference to 2013. The 2013 number came from another article I found and have quoted above. I may be wrong, but it seems to me that Rocky Mountain Institute is expecting the cost per KWh to come down around $14, which I think is unrealistic. $50 is very much possible. Even coming down to $50 means 8% reduction YOY, which is plausible but higher than the historical average today. And like you, I would welcome $14, not sure it will happen. From what I read and understand, the improvements in energy chemistry have been around 3% a year for the last decade. Now there may be a breakthrough technology that can totally change battery technology, but we do not know about that yet. So I agree with you that I find this report simplistic and too optimistic.
getting into solid states, i mean you can probably get away from cobalt and nickel and get the solid lithium or sodium type, increasing density and decreasing weight in one go. and some solid states are here, just no mass manufacture yet. within a decade it will be figured out. isn't it mercedes who have one in some of their trucks being tested?
Yes, we have improvements to current technology and introduction of new disruptive technology. Solid State is the promising new frontier. However, this technology is not expected to go into production until about 2026. It is possible that Rocky Mountain is expecting this technology to come into play, but if so, they have mentioned it anywhere. What was surprising to me (or may be not) is that Tesla is still hooked (pun intended) on Li-ion and is trying to improve energy density. Toyota wants to showcase a prototype by the 2020 Summer Olympics. https://www.trucks.com/2019/08/13/solid-state-batteries-power-electric-vehicle-breakthrough/ Solid-state technology could grow the global number of electric vehicles from 8 million last year to 100 million by 2028, the report said. The batteries provide many benefits, the researchers said. They include greater longevity, reliability and energy density. The batteries are lighter and have less bulk than lithium-ion batteries. And they eliminate the fire risk posed by lithium-ion batteries’ flammable liquid electrolyte by replacing it with a solid material. TESLA NOT GOING SOLID STATE One outlier is EV pioneer Tesla. The Fremont, Calif., company has bet heavily on lithium-ion technology. It is working on new lithium-ion chemistry with Dalhousie University in Canada. Tesla said the new technology offers the energy density of solid-state batteries while maintaining the same format as today’s lithium-ion cells.................................. Ford, Toyota, Daimler, Volkswagen, Renault, Nissan, Mitsubishi and Hyundai are investing hundreds of millions of dollars in solid-state battery research and development. All build commercial trucks as well as passenger cars and light trucks. BMW and startup luxury EV maker Fisker Inc. also are working on solid-state, as are several Chinese EV companies. Automakers are investing in several U.S.-based solid-state battery-development companies, including Kentucky-based Solid Power and Ionic Materials of Massachusetts. Earlier this year, Volkswagen invested $100 million in QuantumScape, a Massachusetts-based solid-state battery developer. Toyota has pursued solid-state battery development for years. It has a technology-sharing agreement with Suzuki, Subaru and Mazda. The automaker also has a joint-development pact with Nissan, Honda, battery maker Panasonic and the Japanese government.
New Battery Breakthrough Could Replace Lithium-Ion January 23, 2019 TORRANCE, CA—Engineers at the Honda Research Institute here have developed a new type of battery that could replace traditional lithium-ion devices. Fluoride-ion chemistry, developed in collaboration with scientists at the California Institute of Technology and NASA’s Jet Propulsion Laboratory, enables the use of materials with higher energy density and a more favorable environmental footprint than currently available technology. “Fluoride-ion batteries (FIBs) offer a promising new battery chemistry with up to 10 times more energy density than currently available lithium batteries,” claims Christopher Brooks, Ph.D., chief scientist at the Honda Research Institute. “Unlike lithium-ion batteries, FIBs do not pose a safety risk due to overheating. And, obtaining the source materials for FIBs creates considerably less environmental impact than the extraction process for lithium and cobalt. “FIBs provide an attractive alternative to other types of potential high-energy battery electrochemistries, such as those based on lithium or metal hydride chemistries, which are generally limited by the inherent properties of their electrodes,” explains Brooks. “Due to the low atomic weight of fluorine, rechargeable batteries based on the element could offer very high energy densities.” .......................................... When paired with a composite cathode featuring a core-shell nanostructure of copper, lanthanum and fluorine, Brooks and his colleagues achieved reversible electrochemical cycling at room temperature. In the future, fluoride-ion batteries could power electric cars, airplanes and other vehicles. The higher-capacity nature of the battery also makes it a good candidate for cordless power tools and other products.
