Demystifying Energy Storage System Costs

 |  Includes: AONEQ, AXPW, TSLA
by: John Petersen

Energy storage technology developers frequently veil their costs in mystery by talking about the direct cost of cells, modules or packs, when end users just want to understand fully installed system costs. Lux Research recently went a long way toward taking the mystery out of the energy storage sector in a report titled "Grid Storage Battery Cost Breakdown: Exploring Paths to Accelerate Adoption." The report is an extraordinary piece of work and I'm grateful that Lux gave me a copy so that I can share their methodology with readers who are confused by public statements that focus on trees and even branches while ignoring the forest.

The basic premise of the Lux report is that an analysis of energy storage device costs is meaningless at anything less than a systems level that includes the all-in costs of cells, packs, thermal management systems [TMS], battery management systems [BMS], power conditioning systems [PCS], and, in the case of stationary products for the electric grid, land, construction and interconnection costs.

In four years of blogging on the energy storage sector, this is the first time I've seen anyone identify and quantify most of the costs associated with turning raw materials and components into a functional energy storage system that's ready to perform useful work. While the Lux model does not include transportation costs, facility depreciation, profit margins, and research and development costs, that issue is easy to resolve with a couple of simple gross margin assumptions.

The following table begins with the essential cost classes for a 1.5 MW, 4 MWh stationary energy storage system using lithium-ion energy cells, an air-cooled TMS and shipping container construction. Instead of trying to separately identify the costs that were excluded from the Lux report, my revised table adds a 25% margin to direct cell and pack costs and a 15% margin to TMS, BMS and PCS costs to cover transportation, depreciation, overhead and profit for the companies that do the work.

7.8.12 LiFePO4 Energy.pngClick to enlarge

While the Lux Report went through the same exercise for stationary systems using Lithium-Ion Power Cells, Zebra Battery Cells and Vanadium Redox Flow Batteries, and I went through the same exercise of adding appropriate gross margins for various system elements, printing all the tables in the body of this article would add too much bulk and complexity. So I've decided to simply provide a link to my Excel spreadsheet for those who want to drill down deeper into the differences in system costs.

Returning to the table it's easy to see how a cell manufacturer like A123 Systems (AONE) can honestly talk about cutting cell costs to $500 per kWh while reporting average cost of goods sold of over $1,000 per kWh. It's also easy to see how EV manufacturers like Tesla Motors (NASDAQ:TSLA) can talk about paying $250 per kWh for commodity grade cells, charge $500 per kWh for battery pack upgrades on the Model S and bury the costs of their TMS, BMS and PCS in the price of their products without fully or fairly disclosing the true economic cost of electric drive and the path those costs are likely to follow for the foreseeable future.

In January of this year I published an article titled "Why The Electric Vehicle House of Cards Must Fall," which included a bottom-up cost walk analysis from Bernstein Research and Ricardo PLC that started with a $19,000 gasoline powered vehicle, deducted the costs of internal combustion drivetrain components and then added the incremental costs of electric drivetrain components. The end result of their bottom up cost walk analysis was a $38,800 electric vehicle.

1.8.11 Cost Walk.pngClick to enlarge

To date, that article has drawn 1,067 reader comments; a personal record for me that may be a website record for Seeking Alpha. A huge percentage of the comments criticized the Bernstein-Ricardo analysis based on claims of cell level or pack costs that are floating around in the ether but only tell part of the story because they omit petty details like profit margins for component manufacturers and balance of system costs for companies that want their batteries to do useful work. When you put it all together and recognize that everybody in the value chain has to either earn a profit or fail in business, the truth is damned ugly. That doesn't make it any less true.

The bottom line for investors is that energy storage product developers and electric vehicle manufacturers who disclose part of their costs without discussing issues like markups and balance of system costs are telling half-truths that become whole-lies when a reader doesn't know enough to understand he's only getting part of the story.

I'm sick to death of happy talk news reports about how automotive and utility grade lithium-ion battery systems will cost $250 per kWh a decade from now. Those claims are true if the only costs you consider are the materials and labor that will go into the cells of tomorrow. When you include pack costs and the necessary TMS, BMS and PCS, the stories are low-balling the true cost by sixty to seventy percent and making the absurd seem plausible. Investors who don't understand that there are three discrete elements in every functional battery system are essentially playing a game of Three-Card Monte where they never see more than part of the truth.

The Lux Report was brutally frank in noting, "Lead acid batteries currently have the lowest capital cost ($/kWh) and the widest market penetration of all batteries, despite weaker performance characteristics, especially in terms of cycle life." It was even more direct in its assessments that "Despite its maturity, Li-ion batteries remain prohibitively expensive for all but the most lucrative grid storage applications" and "Even with mass production, large systems, and decreased material costs, Li-ion and molten-salt batteries and VRFBs will not hit the $450/kWh mark by 2022 unless substantial improvements are made up and down the value chain. Next-generation technologies that can achieve similar levels of performance for ultra-low costs will have an opportunity to snatch up market share, even a decade from now."

My old team at Axion Power (NASDAQ:AXPW) recently sold a half-million dollars of PbC batteries to Norfolk Southern (NYSE:NSC), which plans to use them in a battery-powered switching locomotive. A million dollar follow on order for a battery-powered long-haul locomotive is expected later this year. While future expansion of the battery-powered locomotive program will depend on system performance on the rails, the implementation decision follows two and a half years of double redundant testing by NS, Axion and Penn State that was part of a broader technology evaluation program where NS tested Axion's PbC battery and East Penn's Ultra Battery, along with NiMH batteries, LiFePO4 batteries, Zebra batteries, Fuel cells and Lead acid traction batteries. The performance metrics that ultimately made the difference were high charge acceptance rates, greater depth of discharge range and longer cycle life in one of the most demanding applications imaginable. The $800 per kWh NS paid for the PbC batteries and BMS is currently comparable to the cost of competitive lithium-ion and Zebra systems, but costs are expected to fall as manufacturing technology matures and supply chains get more robust.

Over the last several years, the media's focus on of the wonders of lithium-ion batteries and electric cars has gravely distorted public perceptions about both the likelihood of commercial success and the potential societal value of battery-powered cars. The most telling graphs I've seen in a long time come from a recent McKinsey report titled "Resource Revolution: Meeting the world's energy, materials, food, and water needs."

The first graph compares the potential resource savings from several key technologies with the average societal cost of implementing those technologies. Of the eight identified energy technologies, electric and hybrid vehicles are the least efficient by a wide margin.

7.8.12 McKinsey RS.pngClick to enlarge

The second compares the relative resource benefit of productivity technologies with the number of publications that discussed those technologies. Once again, less than 9% of the potential energy benefit snagged 33% of the energy press.

7.8.12 McKinsey Hype.pngClick to enlarge

Together, these two graphs are a perfect example of the Hype Cycle in action. Electric cars are without question the most expensive conservation opportunity identified in the first graph, but they get the lion's share of the press coverage because the EV illusion has tremendous emotional appeal despite the fact that the implementation is devoid of economic merit because it costs more than it saves.

I regularly remind readers that there is no such thing as a silver bullet energy storage solution. I'm also the first to point out that Axion's PbC will never be used in electric and heavy hybrid vehicles. With its extraordinary charge acceptance rates, wide depth of discharge range and long cycle life, however, the PbC will be a formidable competitor in building energy efficiency, transport efficiency, road freight shift and power plant efficiency, markets that represent an order of magnitude more economic potential and an order of magnitude less hype. That strikes me as a fair trade and brings me back full circle to a thought I first expressed in November 2008, that investors who want market beating performance from their energy storage investments should focus on companies that manufacture cheap products for the masses instead of cool and sexy products for the 1%.

Disclosure: Author is a former director of Axion Power International and holds a substantial long position in its common stock.