#1 cleantech news, reviews, & analysis site in the world. Subscribe today. The future is now.


Published on May 9th, 2015 | by Zachary Shahan


Tesla Powerwall & Powerpacks Per-kWh Lifetime Prices vs Aquion Energy, Eos Energy, & Imergy

May 9th, 2015 by  

Update (May 10): Note that none of the prices below include any subsidies or incentives.

[Full Disclosure: I’m long TSLA & SCTY.]

Since the Tesla Powerwall news came out, it seems the world has been racing to understand energy storage. It’s easy to understand the concept of storing electricity and using it at a later point in time. We’ve all been living with batteries of various sorts essentially our whole lives. However, understanding what the purchase price means in relation to how much electricity the Tesla Powerwall will store and how much that will affect our electricity bills is a new endeavor for many. Additionally, understanding how Tesla’s utility-scale battery offering competes with top players on the market is a mystery to many.

The other day, some other curious people and I ran numbers comparing the per-kWh price of the Tesla Powerwall & Powerpacks (the utility-scale battery options described on the bottom of this page) with top competitors on the market. Admittedly, that was too simplistic a comparison. The kWh rating provided for all of these products is simply the maximum amount of electricity they can store at one point in time. So, in the case of the Powerwall, 7 kWh means that the battery can hold up to 7 kWh of electricity at one time, similar to how a 5-gallon jug of water can hold up to 5 gallons of water.

You have to multiply that capacity rating by # of cycles (# of times the battery will be filled up and then emptied), depth of discharge (whether the battery can be fully emptied during each cycle or needs to be only 80% emptied, 70% emptied, etc), and efficiency (how much electricity is actually transmitted, not lost, in each cycle), and then divide by price to determine a per-kWh price for all of the kilowatt-hours your system is expected to produce… before degrading to 80% of its rated capacity, that is (at which point it’s actually still useful, but that’s apparently the global standard for “end of product life”).

As you can see, there are a number of assumptions you have to make to perform these calculations, and even if all of your assumptions are correct, it’s not like the products are completely dead at the end of the studied time period. This also leaves out operational costs (which we’ll assume to be $0 in the calculations below).

Anyhow, this is the best method I’ve found for comparing Tesla’s Powerwall and Powerpacks to top products on the market. More importantly, on the residential side, the numbers should help a consumer to evaluate the cost-effectiveness of getting a Powerwall (should you get commercial access to one) — that’s the main aim in the next section of this article. Note that I’ve actually left out “competing” lithium-ion and lead-acid batteries in the residential section. Basically, even at a glance, it’s clear that they don’t compete with the Powerwall, so I didn’t bother finding all of the specs and doing the calculations. If you want to do so for any particular battery, I’m happy to add the info in, but I’ll need links or company spec sheets indicating cycle life, expected DoD, efficiency, and price in order to do so.

With a ridiculous amount of help from three wonderful CleanTechnica readers, below are the assumptions and results, split into a “residential” section and a “utility-scale” section.

Residential Battery Storage — Tesla Powerwall x 4 vs Aquion Energy x 2 vs Iron Edison x 1

Subheading have you confused? I ran the numbers for 4 Powerwall purchase scenarios, 2 Aquion Energy products, and 1 Iron Edison product. Since the intro above was too long already, I’ll jump into the table first and list some of the takeaways and the assumptions underneath it:

Tesla Powerwall vs Competitors

What do the “$/kWh used” figures actually mean on a practical level? Here are some examples, just using the final figure from the first Powerwall column:

  • If you buy electricity from the grid for $0.35/kWh (with no unavoidable fixed costs) and generate excess electricity from your solar panels that you can’t sell back to the grid, you can save $0.10/kWh by purchasing a Powerwall from SolarCity for $5,000.
  • If you buy electricity from the grid for $0.35/kWh (with $0.10/kWh unavoidable fixed costs) and generate excess electricity from your solar panels that you can’t sell back to the grid, you break even by purchasing a Powerwall from SolarCity for $5,000… but are using more of your own solar electricity and relying less on electricity from the grid.
  • If you buy electricity from the grid for $0.35/kWh and are not allowed to send any electricity from rooftop solar panels into the grid, and you would generate that electricity at a cost of $0.15/kWh, you would essentially pay $0.05/kWh more by purchasing a Powerwall from SolarCity for $5,000 (but you’d perhaps have to have another source of backup in winter if you don’t have much sunshine then, but you’d be protected against rising grid electricity rates).
  • If you buy electricity from the grid for $0.15/kWh (with $0.05/kWh unavoidable fixed costs) and generate excess electricity from your solar panels that you can sell back to the grid for $0.15/kWh, you pay $0.25/kWh more by purchasing a Powerwall from SolarCity for $5,000… but are using more of your own solar electricity and relying less on electricity from the grid.
  • If you buy electricity from the grid for $0.15/kWh in the middle of the day (with no unavoidable fixed costs), buy 5.8 kWh of electricity from the grid for $0.40/kWh in the evening each day, and generate excess electricity from your solar panels that you can sell back to the grid for $0.15/kWh or store for later use, you break even by purchasing a Powerwall from SolarCity for $5,000… but are using more of your own solar electricity and relying less on electricity from the grid.

As you can see, the financial benefit depends a great deal on individual circumstances related to your local utility and solar power options. If you are considering a Powerwall (or a product from Aquion Energy or Iron Edison), hopefully this helps you to make a decision.

Tesla Powerwall price

Aquion big battery

Iron Edison LiFePO4

Assumptions used above:

  • Powerwall: 92% efficiency, capacity = average of 90% rated 7 kWh over product life (due to assumed degradation over time), 5,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”). Also, SolarCity prices come from statements from SolarCity, the wholesale price comes from Tesla Energy, and the retail price from a distributor includes an assumed 20% markup.
  • Aquion Energy S20P: 85% efficiency, capacity = average of 90% rated 2.366 kWh over product life (due to assumed degradation over time), 3,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”). Price from retailer linked above (and seems to be on sale).
  • Aquion Energy M100-L082P: 85% efficiency, capacity = average of 90% rated 28.4 kWh over product life (due to assumed degradation over time), 3,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”). Price from retailer linked above (and seems to be on sale). Admittedly, far larger than most homes would need — more appropriate for some businesses.
  • Iron Edison 24V Lithium Battery: 96% efficiency, capacity = average of 90% rated 4 kWh over product life (due to assumed degradation over time), 2,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”), 80% depth of discharge. Price from retailer linked above.


Utility-Scale Battery Storage — Tesla Powerpack vs Eos Aurora vs Imergy Flow Battery

Before jumping into this table, let me note that the Eos Aurora battery isn’t supposed to be available until 2016. Also, when it comes to Imergy, it has states a current price of $500/kWh (capacity) and a projected future price of $300/kWh (capacity), so I’ve run numbers for both. It’s not clear when Imergy expects to reach $300/kWh from what I’ve read.

Lastly, Imergy’s vanadium flow batteries reportedly have “unlimited” ability to cycle (no degradation like with the other batteries here), but the warranty period is just 5-10 years. So, in two examples, I give Imergy a product lifetime of 15 years (same as with Tesla, which offers a 10-year warranty but projects a 15-year lifespan) and 30 years. Whether a utility or other business customer is going to consider the cost over 30 years or bet on such a lifespan is a question to consider.

Tesla Powerpack vs Imergy

As you can see, the assumptions here are quite critical to determining the lowest-cost product. Will startup Eos Energy deliver as promised? Will Imergy’s battery work and be useful to you for 30 years? Will Imergy’s cost come down to $300/kWh as projected? Will Tesla’s Powerpack last 5,000 cycles and cost $250/kWh?

I imagine some utilities are going to hedge their bets a bit and buy various storage products. And I imagine other utilities will just go for the products that they most trust or prefer for their specific needs.

Of course, there are many other battery-storage companies on the market and winning contracts, as well as some reportedly coming to market in the coming year, but I couldn’t find prices or important specs for them so they were not included above. Such companies include Alevo, AES, Bosch, CODA Energy, EnerVault, and many more.


Image Credit: Tesla Energy

Tesla Powerblocks

Image Credit: Tesla Energy

EOS Aurora battery

Image Credit: Eos Energy

Imergy Power Systems

Image Credit: Imergy Power Systems


Assumptions used in the table above:

  • Tesla Powerpack: 92% efficiency, capacity = average of 90% rated 10 MWh (example) over product life (due to assumed degradation over time), 5,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”), $250/kWh.
  • Eos Aurora 1000 | 6000: 75% efficiency, capacity = average of 90% rated 6 MWh (example) over product life (due to assumed degradation over time), 10,000 cycles before degrading to 80% of rated capacity (theoretical “end of product life”), $160/kWh.
  • Imergy Vanadium Flow Battery: 75% efficiency, capacity = 100% of rated 7.5 MWh, 5,475 cycles over 15-year lifespan or 10,950 cycles over 30-year lifespan (it could also cycle twice a day, in which case the 30-year examples could apply to 15 years), $500/kWh current pricing and $300/kWh projected future pricing.

Have more to add? Drop me a note.

If you want to read more along these lines, this is a good piece: Tesla Battery Economics: On the Path to Disruption.

Tags: , , , , , , , , , , , , , , , , , ,

About the Author

Zach is tryin’ to help society help itself (and other species) with the power of the word. He spends most of his time here on CleanTechnica as its director and chief editor, but he’s also the president of Important Media and the director/founder of EV Obsession and Solar Love. Zach is recognized globally as a solar energy, electric car, and energy storage expert. He has presented about cleantech at conferences in India, the UAE, Ukraine, Poland, Germany, the Netherlands, the USA, and Canada.

Zach has long-term investments in TSLA, FSLR, SPWR, SEDG, & ABB — after years of covering solar and EVs, he simply has a lot of faith in these particular companies and feels like they are good cleantech companies to invest in. But he offers no professional investment advice and would rather not be responsible for you losing money, so don’t jump to conclusions.

  • John Buck

    They don’t have the gigafactory though, and the tesla model 3 will provide most of the revenue to research electric batteries.

  • Denet Lewis

    Can anyone help me with these questions regarding a residential battery system?

    1.What are the requirements for selling home batteries or systems
    with inverters e.g. national / international norms, association rules and
    quasi-binding recommendations (e.g. UL1943, IEEE, etc.)?

    2.What are the requirements for redemption / disposal?

    3.What are the requirements for transport (does ADR apply)?

    4.What are the special safety requirements, must haves and
    recommended (UL, CE, etc.…)?

    5.Are there any technical design requirements for stationary storage
    (3-phased, USV, remote control, etc.)?

    6.What are the registration / approval processes with the network

    7.What are the requirements to qualify for subsidy programs
    (technical, warranty)?

    8.Are there any regulatory requirements with regard to minimum
    warranty times by law or incentive programs?

    9.Are there any regulatory requirements for the manufacturer with
    regard to ongoing monitoring obligations?

    10.Are there any fire regulation requirements? (indoor vs. outdoor)

    11.Are there any insurance issues with battery systems? (house

    2.Installation Requirements

    1.Are there any special requirements / conditions for installers
    (OSHA Requirements, certifications)?

    2.Do installers need to demonstrate any minimum special training /
    certifications / safety training?

    3.Is there any final approval by Utility or City authorities

    3.Documentation Requirements

    1.What types of documents have to be provided to installers and/or

    4.Technical Recommendations/Preferences

    1.What is the desired voltage for battery pack(from Customer/Market)

    2.What is the desired voltage of the inverter, one or more phases,
    AC or DC?

    3.Are there any min. requirements for the inverter regarding angular
    phase shift (differences of phases (cos(φ))

    4.What is the required min. and max. performance (Amps) of the
    system (AC and DC)

    5.What are the mechanical/environmental (temperatures, humidity,
    noise, FCC regulation) Requirements for the System specified in indoor vs.
    outdoor (Housing outdoor rating, IP Codes, UL codes,?, etc..)

    6.What are the interconnections standards and dominant designs (AC
    vs. DC)

    7.What are the desired System topologies (AC-coupled vs. DC-coupled)

    8.What are required installation provisions for indoor vs.
    outdoor (e.g. wall hanging or standing, weight restrictions, etc.)

    5.Customs Requirements

    1.What are the local customs regulations?

    2.What are the import taxes?

    3.What are the local content requirements?

    6.Servicing requirements

    1.Are there any special requirements on servicing (Service Hotline,
    Service personnel, etc..)

    Is there a service “over
    the air” required incl. system performance info to customer

  • Phillip Hannam

    viction energy already has a system for sale the eco

  • Michael Overturf

    Nice job

  • eveee

    Wait. Are you saying Tesla claims battery expertise? Please find where they said that. They don’t claim anywhere that they know how to make cells.
    You keep comparing apples and oranges and complaining that oranges aren’t red.
    I mean really, Tesla says, here it is, this is how we do it.
    Like I said, the real issue is experts like you with a grudge that keep resisting a success for some emotional reason.
    Who cares and why does it matter whether Tesla is smarter or not?
    As I said before, commitment is what it takes.
    You tout Nissan, but they did not build a GigaFactory or a SuperCharger network.
    But EVs are not going to happen without those.
    You denied that volume will lower cost. Please. Nobody real claims that.
    You are back to your emotional diatribes with claims about cheating.
    Forget it.

    • Robert Haylar

      Resisting a success? A battery that carries 3 times the cell weight in supporting material is a ‘success’? I will not mention that car again after this, but you don’t seem to understand what I am saying. If the battery is vanilla, that does undermine quite a lot, doesn’t it?

      Tesla claim to have the ‘most advanced BMS’ etc. It is common to read of journalists praising Tesla’s high-tech battery. But do they know that to be so, or are they just parroting what every other journalist says?

      Tesla don’t say ‘how they do it’, but spin stories around what are simply consequences of a particular design.

      Take the patented ‘fuse’.

      1) The wire (fuse) is there because it is otherwise impossible to tab weld the heavy gauge bus-bar to the battery cap.
      2) The bus-bar is heavy gauge in the hope of reducing battery resistance, as a consequence of the long path through the battery that spreads over the entire floor pan.

      3) The’ flat battery cap’ is there, because a PTC would limit the current that needs to be forced from the cell, not intended for that purpose.

      4) Without a PTC, there must be a ‘fuse’

      As a fuse, the wire is useless. Under acceleration, cell current will be at least 10A. The fuse will be around 20A when allowances are made for temperature etc. The cell can go into thermal runaway well before 10A is reached.

      It is no surprise that mechanical disruption of the packs results in fire. Fires that were avoidable, not “bad luck” or

      “no worse that a gasoline car”, but avoidable,

      The Gigafactory…that will be an ecological disaster if it succeeds. A mountain of cells put to inappropriate use.
      But, you will see my opinions on that, the Leaf, and Tesla-only charging stations, as ’emotional’

      I have been looking into the cost calculations from other papers. I will provide details, but for now, it is clear that currently, no battery is cheap enough to be of benefit to a 4kW solar installation, net-metering or not. For larger systems, pay-off only occurs when capacity is around 48kWhr, and where power must be around 20kW.
      I would agree, that if the Powerwall could be bought for $3k with no additional costs, that may at least contain the necessary expenditure to a minimum.

      • eveee

        You have no idea what you are talking about. You just stepped into the zone of battery applications where you have no experience. Get out while the going is good.

        • Robert Haylar

          You have no idea what areas of battery and power electronics in which I have worked. But, there is the matter of your experience and expertise to consider.

          In the Australian example you provided as evidence for the Powerwall, adding a Powerwall to a small solar installation brings the leveled cost from $0.14/kWhr for solar to $0.30kWhr – which is the same as grid-only. But it’s worse than that.

          The author of the reference has used the battery to shift stored solar energy, but the daily tariff is flat, That means it does not matter when the battery is used, because the grid cost/Kwhr remains the same throughout the day. Using the stored energy at night, is an attempt to justify the battery.

          The 4kW panels produce 16kWhr/day, to supply a 12kWhr/day load. Some of the excess solar energy is sold to the grid, and some stored in the battery,

          Selling returns $0.08/kWhr, while grid energy costs $0.30kWhr. It would be better to use the stored energy for use in the household, so offsetting the $0.30/kWhr grid cost, than to sell it for $.08/kWhr. That makes sense, doesn’t it?

          Unfortunately, the Powerwall is too expensive.

          By selling the excess 4kWhr to the grid, the owner can earn 4*$0.08 = $0.32/day, or by offsetting grid energy 4*$0.30 = $1.20/day. The latter is only effective if the battery cost is less than $1.20 – $0.32 = $0.88/day. The Powerwall will cost at least AUD$4000. Over 10 years, that becomes $1.09/day – which is more than the $0.88/day that could be earned by adding the battery to the solar installation.

          When efficiency is considered, the battery cost will increase to $1.30/day. It is more economical not to have the battery, and sell directly to the grid, or not sell the excess at all.

          Matters get worse for the inclusion of a battery, if there is also a low tariff period. Adding more storage in an attempt to remedy the problem will not work, because adding more Powerwalls adds the same daily cost per battery.

  • Bob Fearn

    Thanks for this Zachary. Very useful info but it does point out, again, how complex batteries as well as inverters and solar modules are for the average consumer.
    I think it in the best interest of solar related manufacturers to cooperate and adopt similar standards and terminology whenever possible.
    This probably won’t happen without some government assistance (interference) however this complexity is not helping solar.

  • Jacob

    I honestly do not believe their 100% DoD claims.

  • Carl Parrish

    I read this article hoping to see the comparison to ambri. Before I heard about the powerwall I was expecting to get an ambri for my house but now I’m not sure if I should just go and get a powerwall.

    • JonathanMaddox

      Without providing any source or study to back me up, my gut feeling is that the lifetime cycle calculations for molten metal Ambri batteries are going to be a lot simpler than those for “interesting” lithium chemistries and that the batteries will last an awful lot longer in heavy-duty applications. But they’re not available at retail at all yet, as far as I know, and simply due to the issue of manufacturing scale they may not be competitive on capital cost, at least in the early days.

  • eveee

    Not offering the 7kwhr version? Is that due to availability? I mean they advertised it. GIven the tremendous response, one figures they are sold out.

  • doughouseman

    A technical question – is your “kWh/cycle” assuming round trip efficiency (e.g. charging and discharging) or is it assuming that you have 7KW(Tesla) in the battery and of that you get 5.8 back out of the battery?
    If it is the latter – what are the losses putting the energy in the battery (e.g. how many KWH do I need to have to put 7 KWH in the battery)?

    • eveee

      Efficiencies are considered in the tables. 92% battery efficiency. Inverter efficiency is 97%. See comment below on costs for references.

  • peter

    As an average consumer i just want to know based on average consumption what is the payback or roi

  • Environment: Zero

    Thanks for the article!

    Could you update this based on the fact that $3000 is the retail price for clarity/consistency? As per your subsequent article…
    And if for us non-Americans if you could put what prices do and don’t include tax? Presumably the purchase prices don’t, I’m not sure if your kWh prices for electricity do though for the use case examples?

  • heinbloed

    A fresh price comparison from Germany, published today:



    Tesla is half price (costs per kWh) of the next competitor Varta, a traditional German battery manufacturing company(+100 years).


  • Adam Devereaux

    It’s obvious that Tesla has brought a LOT of publicity to the storage market. News article from sources I never would have expected are talking about cost per kWh of storage, cost per kWh stored and are looking at storage as something that is at the cusp of affordability versus some pie in the sky pipe dream. I’m sure Sonnenbatterie is thrilled with the publicity as is every other (competitive) storage provider.

  • Sascha

    Wowzer – quite a heated topic! To the best of my knowledge (based on our conversation with them) IronEdison offers Nickel-Iron batteries that discharge 100% for 11,000 cycles. I don’t believe they have a Lithium option but I could be wrong.

    Aquion, as someone correctly stated in the comments cycles 3000 times at 100% depth of discharge where capacity gets reduced to 80% of original rating. If you discharge 50% then cycle life extends to 6000 and you still have 80% of the original capacity rating.

    In our calculations here at altE, the cost per watt-hour/cycle, was the least expensive with the IronEdison because it can cycle so many times. Aquion was a little more than Tesla, but as we found out the Tesla batteries that will be available in the short term aren’t made for daily cycling after all – that’s a big “catch” in our book. It’s no longer an apples-apples comparison when the batteries can’t be used in a daily manner.

    Additionally the Aquion batteries have the advantage of being completely non-toxic.

    • eveee

      Go over to the Iron Edison web page. They do have Li. They also have Iron. The model used in the article is on the web page.
      Are you able to discuss the Tesla batteries? It sounds like you are saying you don’t have access to the daily cycle batteries. Sounds like they are in short supply.
      Whats your take on the Iron batteries?

  • eveee

    Whoa pardner.

    “no details are known on functionalities, inverter, and installed cost”

    Thats because you did not look at the SolarCity page.

    Thats where you find out that SolarCity installs complete systems in the US for $7140 purchased, or $5000 for a 9 year lease including DC-AC inverter.


    The first four columns of the first table could have had a better description. The first two are for compete systems installed with DC-AC inverter. The next two are battery only, retail and wholesale.

    There is a bunch of manufacturers besides SonnenBatteries that were not included like Balqon for instance. Its meant to be representative, not a comprehensive list of every storage worldwide. Somebody was bound to get left out. But I can understand why there needs to be reporting about a PowerWall system that got 800 million dollars worth of order in a week.


    It would be helpful if you include some well researched links to Sonnebatterie and some math. That offering sounds interesting and we would all like to hear more about it.

    • Jacob

      But now it is official that $3000 is indeed the final retail price. Not wholesale.

  • heinbloed

    An old 6/2013 price list from Germany comparing various manufacturers and technologies:


    ( no one beats Tesla’s price)

  • Marion Meads

    Well, if you have a link to pricing and specs from the manufacturers, it would be included.

    • bink

      he not the one that should be worried about, integrity, and intellect as a journalist here. This is obviously biased and not well researched because if ti had been well researched, it would have never fit into the narrative we see before us, the oversimplification is mind blowing

      • eveee

        Can you create a table with the real numbers as you see it and show us?

  • Paul in Austin

    I give you a “B” for effort. However, unless I missed something, you left out important information.

    What is the interest cost? You don’t buy a power plant in cash, so you don’t buy a storage battery with cash either, at least in an apples-to-apples comparison.

    Are there wiring costs? Will the battery need temperature conditioning? What are the shipping and sales tax costs?

    Will the batteries need housing? I personally only have one car in the garage, so that leaves a lot of room. Other people do not have this extra space.

    Will there need to be an inverter separate from the PV inverter? (I asked this question of the Tesla sales desk and they could not answer.)

    The 92% efficiency you gave for Powerwall was for DC-to-DC. Are there extra losses for DC-to-AC?

    • bink

      Paul, you are whistling Dixie in the wind with this guy, he is putting out so much misinformation it is ridiculous.

      His over simplification of the commercial units is malpractice. The Tesla unit is DC versus the others AC systems for price comparison.

      He also makes widely disputable claims about a 5,000 cycle life for the Tesla battery at 100% discharge with no facts to back it up.

      The government research will tell you different based on actual field operating units and not manufacturer claims

      Also what he is leaving out in his oversimplification is, application use. The lithium technology has to be customized. It cant do power and energy at the same time, it is one or the other.

      A commercial or industrial application could have many duties to perform; demand response, peak shaving, power quality, load factoring, volt support, providing reactive power, emergency back up

      Some are power applications and some are energy. In order for the lithium to provide these services it would have to be paired in a hybrid configuration with some other technology or over sized in a limiting discharge design with two different battery banks with different performance profiles .

      • Robert Haylar

        Quite right. There are all sorts of cells, because no one cell can be used for all applications. But, Tesla are like the proverbial man armed only with a (noisy) hammer.
        They are the biggest consumer of a cell that originated in 1992.
        I am sure that Panasonic are delighted to make such a profit from a product that is near its end of life. Of course they are cheap!
        Funny, huh. Who’s the Daddy?

        Anyway, the article’s calculations are inane. Circular, and without any meaning.
        The formula is $/kWhr = Total Cost/(cycles*capacity)
        where capacity is de-rated for ‘efficiency’

        For example, the Powerwall:

        Cost = $7140
        Total energy = Cycles*Capacity = 28,980kWHr
        Cycles = 5000
        Capacity = 7kWhr ( de-rated to 5.8KWhr to account for “efficiency”)

        The formula is;
        $/kWhr = Total Cost/(cycles*de-rated capacity)

        $/Kwhr = (7140/5000*5.8) = $0.246/Kwhr

        1) The number of cycles must occur over some period of time, but that time has not been quantified at all.

        2) Power is the rate at which energy is expended, but power has not been considered.

        Without any quantified cycle time or intervals between, we get:

        Power = Energy/Time = Total kWhr/Cycles = 28980/5000 = 5.796 = De-rated efficiency.

        Forget the commercial units, that will take care of itself.
        Insurance companies don’t like flammable batteries. Nor will Tesla’s vast experience in building 45,000 batteries over 3 years, from over-sized AA cells, carry much weight. Except where the brand may be of conspicuous use. Some offices, that sort of thing.

        • bink

          finally someone who gets it. I am just afraid how this will affect the Industry, when the rose colored glasses fall off. I always believed Elon was good for the world, but now he has been drinking his own kool aide for profit.

          He has damaged the rest of us with his minions and their claims, its blasphemy

          • Robert Haylar

            There are some here, and elsewhere, who won’t hear anything that disturbs their delusion that a box containing some cells wired together, is really high tech.
            Bogus claims last as long as the time they can evade examination. Telsa appeals to a rather limited set. 38,000 orders is not a lot considering the population and publicity.

            I was initially bothered by the disgraceful and self-serving cynicism of that battery, but I doubt the battery will go very far.

            I know what the consequences of that cell are, and how Musk has used a deficiency to his advantage. Only works on the ignorant. There are engineers who evaluate storage upon its merits. I doubt that Johnson Controls are losing any sleep.

          • bink

            very true, I guess I am tired of hearing about it

        • eveee

          If your comment had any semblance of admission that Tesla just got massive orders or already has battery packs in the field with utility partners,.. it might be believable.

          Keep in mind that the vast majority of orders are for the large power packs.

          “Musk said the company has received 2,500 reservations for the commercial-scale batteries and that the typical installation comes with “at least 10 Powerpacks.” So that’s 25,000 units totaling 2.5 million kilowatt hours.

          Musk used Twitter last week to disclose pricing for the Powerpack at $250 per kilowatt hour.

          Total Powerpack Orders So Far: $625 million.


          You claim Tesla is obsessed with small cell size. Looks more like you are.

          Straubel couldn’t be clearer. Cost and performance. Hard nosed proof. Thats all he cares about. So I ask the same thing. Show me.

          “For the immediate future we see 18650 as the most compelling. Believe me, we challenge it constantly. It always ends up being very controversial, for reasons I don’t totally understand. Nobody gives a damn about the shape and size of your fuel tank! But for some reason the shape and size of what you put your chemicals in to carry your energy in an EV is super controversial. What people should really debate is what are the nature of the chemicals inside; they’re what determines the cost and performance.”

          ” Nobody wants to talk about cost—they always leave that to the end of the discussion. That’s silly. For EVs, there are some key safety and performance metrics that are foundational. They have to be there. Beyond that the most important thing is cost efficiency of energy storage. So if anyone has a more cost-efficient cell architecture, we’d be all ears. Right now nobody has proven they have a more cost-effective cell architecture than ours.”


      • eveee

        Imergy’s products contain DC-AC inverters. That is public information. But Imergy’s $500/kwhr or %250/kwhr cost claims do not include any information about whether that includes DC-AC inverter or not. Its just a raw statement. If you have a reference to a quote, that would be helpful.
        Perhaps you could include facts regarding Imergy’s 25 year lifetime? See, no one can. Those are projected for all these systems. And IMO, they are reasonable. But you are talking like its malpractice to quote 5000 cycles from Tesla, which came from Musk investor call statement of 15 year life, 10 year warranty, versus Imergy claim of 25 year life. I think if the relative CEOs make those statements where all can view, and investors and technical people can view, thats fairly equal.
        A calendar life of 15 years and 5000 cycle life are reasonable for a lithium battery. There is no claim of 100% discharge, and we don’t know if it is or not. The charge controller can limit it so that externally we only see the cycles. Details like that wait for more specifics. Until then, 100% discharge is only speculation. But if you want to add that, a 80% discharge will bring cycle life up to thousands of cycles.

        • bink

          I know for a fact Imergy is selling a DC-AC complete system for installation, and I know Tesla is for a fact not.

          I do not believe the Imergy claim in regards to their short entry price point but the redox platform makes me believe it can be achieved, mid term just not today, they have to execute first.

          VRB’s have a long history with just a few failures of smaller systems but not the larger ones. VRB’s since that time has advanced beyond any major concerns due to unreliability with the development of a new electrolyte from DOE’s PNNL which is now the industry gold standard.

          Imergey and any VRB vendor has an advantage in that they are able to sell the system hardware and lease the energy (electrolyte) component. reducing customer purchase price.

          A rVRB platform also allows you to perform energy and power services without customization. This opens up savings opportunities or value add services to the customer .lithium cant compete on that basis

          The failures and life cycles for lithium utility and commercial batteries are well known. why do you think they are mostly used only for ancillary services. they are too expensive to perform energy services and life is reduced when trying to.

          • eveee

            You don’t know if the $500/kwhr includes inverter or not.

            You have provided no evidence that utilities are shunning lithium batteries for the reasons stated.

            On the other hand, utilities are using Tesla batteries and have just responded massively. It looks a lot more like the cost has dropped substantially and utilities, and others are responding to $250/kwhr.

            VRBs are great. Like all technologies they have their best fit applications. Even BIll Watson, Imergy CEO admits that.

            “Lithium has won a lot of projects for short, quick duration and it may be fine for that.”

            Read more: http://www.pv-magazine.com/news/details/beitrag/flow-batteries–more-cost-effective-for-residential-and-utility-scale-installations_100016565/#ixzz3ZsI9mHZ1

            The market is large. IMO, there is room for them all.

          • eveee

            Can you show us a link to the fact that Tesla does not offer an inverter with their Power unit?

    • eveee

      All the comparisons are without interest, so they are compared on the same basis, so its still good for a simple relative comparison. It will be off where lifetimes differ, say where one entry has a longer calendar life than another. Interest reduces the value of future payments the more one goes into the future.

      Interest cost depends on the application. Cycle life has less economic effect the further into the future you go.
      There is a rough comparison without out it, but going into those economics would get very complicated in a hurry.
      To simplify, here is a generic situation. Here we try to figure out what a series of kwhr outputs in the future is worth on a present value basis, so we can compute cost/kwhr.
      The concept is that a future kwhr is like a future payback or payment, so its value is less than a kwhr today based on interest.
      A series of 10 equal payments (or paybacks) over ten years at 5% interest is 7.722 times one payment, not 10 as calculated in the tables.
      If interest is less, say 3%, its 8.35. You get the idea.
      For our case, simplifying to end of year kwhrs rather than daily makes a small difference. Its the difference between continuous interest and annual compound interest.

      You see how complicated it gets? You have to figure each and every case based on not only the cycle life, but the intended application, years of operation, assumed interest, and so on. And then you would have to see if they really could be compared directly or if their applications would be different so they could not be compared directly.

      But that would bore most people to tears. Are you still reading, or did I cure your insomnia? 🙂

      • bink

        cylindrical cell is a limiting platform. flow battery is not. You are guys have no clue about real world application in the utility and commercial sector and that is what you should have been working off of and you left out the revenue opportunities to give a real sense of the best ROI

        • eveee

          Can you fill in the details on those? Perhaps you can provide some numbers for us.

    • eveee

      About the tables. The first table’s first two entrants left to right are the complete PowerWall systems with DC-AC inverters. The others are battery systems, no inverter.

      In the second table, all the figures are for battery only. In the case of Imergy, we don’t know what their costs include, we only have limited public information of $500/kwhr and projected $250/kwhr. We don’t know if that includes inverter or not, but their products do. Take that any way you like. Its uncertainty. To make the comparison close to equal, we need to look at public information. Otherwise, we are looking at rumor, and how do we compare a rumored cost vs a public one?

      That means the numbers are certainly subject to some judgement, EOS and Imergy make claims, but there are no concrete cost figures on their websites. I don’t doubt them, but the devil is in the details.

      Heres Imergy’s Watson’s take:


      He is talking mostly from the perspective of utility storage. When he says long duration, he means hours. Sounds strange, eh?
      Thats because lithium is better for short duration, even seconds.
      That has value in the frequency and voltage regulation or reactive power services markets.
      Thats a pretty honest admission from him.
      I like Imergy and think VRB has a bright future. Frankly, IMO, it looks like all three have a spot in an energy future with different niches.
      VRBs biggest problem is the availability of vanadium. If you scratch deeper, you find that it does have a life limitation, 25 years, and some O and M for pumps and PEM, but really pretty good. After the lifetime Imergy says you get %40 salvage value of the vanadium fluid. The specifics are more complicated than the summary stating cost/kwhr and we don’t have access. Like the cost of working fluid, for example.

      • Robert Haylar

        41,400,000Kwhr? From 18650’s? I have a really good idea.

        Button cells. Think of how long they last in watch.

        Do you really think that a utility is going to build a storage unit by summing the 2K dribbles from each Powerwall, even if the Powerwall met its implausible claims? Totally unrealistic, evee.

        Telsa make one battery, and only one battery, which, because of its large volume of cells, allows talk of Mega this and that. They make 15000 batteries a year, A total of about 45000.


        The 60kW car battery is just a simple battery, where one part of it is repeated 1500 times. Clusters of 4 cells, having a hug.

        Put your time and effort into supporting innovative companies, where the measure is not how pleased some are by the really fast car.

        • eveee

          Do I think a utility will use them?

          They are using them.

          You can google or go on Teslas website to find out the utilities Tesla is partners with.

          • Robert Haylar

            There isn’t a battery company that does not have some partnership of one sort or another. All of them, past and present. A123, Valence, all promising lithium ion storage.

            The Powerwall is wasteful. Underpowered, and with a lot of over-capacity. Under-utilization is known to extend cell life.
            Not very clever at all. What, other than low price, does it offer- and isn’t even that due to Panasonic?

            Find one patent, or anything else, that makes this battery anything but a pimped box of cells.

            Are you up for the challenge, evee?

          • eveee

            Mr. Haylar – I have been watching your language get worse and worse. I don’t mind that you have your interests and passions, after all, we all do. Thats OK. But there are limits.
            Its not my challenge, and I am not affiliated with any of these companies. There are quite a few that I like now and have liked in the past. That is irrelevant.
            What is relevant is that I like it when different folks chime in with different ideas and contribute to the discussion and maybe help solve some problems. Please stick to that.
            I can see things in each product that could be improved.
            You seem to hold a grudge against Tesla. Really, I don’t care if you do. They sell cars. They sell boxes with batteries in them. The fact that 800 million dollars worth of orders were made for their storage product is newsworthy and notable regardless. It has an effect on the whole market including all the other storage players.
            Many of those storage companies have interesting and noteworthy products like EOS, Imergy, and Aquion.
            I am interested in all of them.
            I really don’t care about their profits or patents or any of that. All of that competitive stuff is not any more interesting than listening to yet another iPhone Android debate. Who cares really?
            I care about what that will do for the future.
            I also have this annoying habit of being picky and researching the heck of out of minutiae with references to make sure something is right and I am not always right, but its ok to be human.
            So when I point out that you overlooked that a statement you made is factually incorrect, don’t feel bad. I do that with people I agree with. And they tolerate it sometimes.

          • Robert Haylar

            Sorry if I have offended, eveee, I don’t hold a grudge against Musk, nor you. Believe that or not.

            There are always limits in manufacturing, but the Powerwall is not what it claims to be. That bothers me, not Musk.

            When did I overlook a factual statement? Perhaps I missed the posts. On your part, I showed that the $/Kwr price obtained by the methods employed here, result in a nonsense quantity of Capacity*efficiency. That means the calculation does not model use. That was ignored.

            Try it. Compare the energy that may be processed into usable power, and then work out how much it will cost – or even if it can work at all. There are a litany of problems that go beyond simple $/kwhr.

            Tesla get an inappropriate cell cheaply, but even so, power output could have been doubled, but then, their inappropriate cell is not cheap enough. That’s right. Not cheap enough for quantity to overcome the cell’s disadvantages, so those disadvantages remain in place.
            kWhr cost, does not adequately define a cell, nor does “chemistry”.

            I need not speculate why Telsa have chosen that route. Facts are facts. The Powerwalls are technically inadequate for their intended purpose.

            I know that, because I could engineer that same battery with ease. You would not want to live in an area where those units would provide emergency power, even if the utility agreed, let alone pay for what has been left on the cutting room floor.

            Recall what William Ruskin said of paying too little.

          • eveee

            Stick to the facts, then. Your language and manner speak otherwise.

            Its specious to say the calculation does not model use. The calculation is not for use, its for relative comparison. If you want the calculation for use, you need to do a lengthy analysis of the exact situation or even a dumbed down one, but there has to be a lot more detail. The purpose of the article wasn’t to do every detail that bored the entire audience and only appealed to a select set of nerds and accountants.

            And please, don’t make unsupported or unsupportable statements. The internet is too full of opinions already.

            Take this statement and find references to back it up. You might have to reform the statement into one that has more precise definition to make it meaningful. Do it like an engineer designing a system. Then it can be analyzed critically.

            “The Powerwalls are technically inadequate for their intended purpose.”

            That statement is too loose. You have to assume or find out what Tesla intends. Thats mind reading unless you can find statements. Worse, it involves just people factors, not physical design.

            Make it into an examination of a market or application, assume nothing about intentions or mind reading, and go from there. Show where it could be used best and least effectively, strengths and weaknesses, and then it will be a balanced assessment.

            Assessments with an axe to grind or grudges have not place in rational thought. Assume your reader is intelligent and can make up their own mind.

          • bink

            who are you? battery installations are complex and Zach should have just reported the story but he went too far because he did not understand the complexity of the situation stay away from oversimplifications when something is dependent upon use unless you model every use for comparative analysis

          • eveee

            And what have you added to the discussion? Any of what you propose? Any specifics? Any analysis? Show us numbers and analysis. This crowd can handle it.
            Go ahead. Complex me. But you better make it good. and don’t waste my time. I am not interested in slogans. Keep it engineering.

          • Robert Haylar

            The method of calculation employed does not take into account the effect of power output capacity of any of the batteries.

            It is easy to understand that the rate of return on investment is related to the rate at which energy may be stored and returned for use.

            Since power is defined as the rate of energy consumption, it is also a direct indicator of the possible rate of return of investment,
            For those who find lengthy analysis to be tedious, power output capacity offers a quick means of comparison.

          • eveee

            You are not doing very well with that philosophical reply. Its not wrong. Its just that it isn’t anything like a real engineering discussion with numbers and facts, and a discussion of a real application. Thats what you seem to be requesting in your complaint. I asked you to address it by providing your own analysis with numbers. If you did that, I would take it seriously. If you just respond with words, it sound like you want to debate or merely complain, which I am not at all interested in. I am asking for your lengthy, tedious analysis. Give me it. In numbers. Real engineering. Give me an example application. Take grid tied solar.

          • Robert Haylar

            You are right – my reply is not wrong, but lacking the straw man that you seek.

            Your method does not yield an accurate value of cost/Kwhr, because there is no means by which it could.
            Now that I have alerted you to the error, perhaps you will desist from promoting it. or do you need me to further explain why you should?


          • eveee

            Why oh why, after multiple requests, can’t you respond with the very thing you are complaining about and provide some detailed numbers and calculations?
            So provide your calculations. Quit complaining and do something useful.
            Nobody is promoting anything. Nobody is preventing you from providing those numbers.
            I can’t take you seriously anymore because you have shown no interest in advancing the discussion in a real way.

          • Robert Haylar

            I did not offer to provide any detailed calculations, You did that for me. If your method works, then one should be able to derive a formula or result that relates to power,

            The desired result, Cost/Energy*Time, can be expressed in terms of power, because Power = Energy/Time. That result may not be obtained from the data, or by algebraic manipulation of the formula, because it is a cartoon,
            My objection is to that method.

            Do you think that I can post a complete procedure for a complex problem in this box? How helpful would it be, when there are countless combinations of equipment and circumstance? But, generally.

            You do not measure the usefulness of your car, by how many times you fill the gas tank. The real goal is how much useful work you get for the money you spend.

            Similarly, the number of battery cycles is largely irrelevant. It does not matter how many times charge passes through the cell’s separators. That’s just a secondary effect of one particular technology.

            Quantify and consider only the useful work delivered by the battery – the power delivered. Replace cycles with the
            total power delivered by the battery, alone. Otherwise, periods when the battery is not (potentially) earning you money, will artificially lower the cost.

            The problem is data. Estimation will lead to large errors and inappropriate purchases.

            I suggest that any prospective buyer obtain some cheap power meters. They can be fitted between loads and wall socket. Many have wireless links that upload to a PC.
            Now you have real data. You can find maximum loads, peak load periods, averages etc. There will be software, that can do a lot for you, and graphically display data,

            Selection of the battery becomes more rational. Capacity and power can be selected from what is available.
            It’s then easier to determine battery life.

            I think, though, that the data will show that reducing waste energy will be the better deal. It is not all that rational to spend more money, so that you can continue to waste more money.

          • eveee

            There is no power limit for grid tied applications. For peak shaving and TOU, a 10 year use (ignoring interest), and sizing the storage so that its less than the loads, the storage is about 12c/kwhr, but the TOU differential is over 19c/kwhr in places like California. I count that as economic storage in the continental US. Its not even with solar. Just TOU peak shifting.

          • Robert Haylar

            Sure…you would need 5 Powerwalls to get 10kW. + inverter
            How much would that save? Say your household is average- around 30kWhr/day. One Powerwall stores 7kWhr, that’s only 7/30 = 23%.
            It’s posible to make it work, but the margin is slim.

            That’s the problem. To minimize capital expenditure, you need to know how much storage and how much power output you need. Averages will not tell you that.
            Google have a metering project. There are meters that you can install by yourself, and monitor the house. Like TED Home Pro etc. It’s a small investment that may pay back a lot.
            You know, when “you” buy energy at a low tariff, store it then “sell it back to yourself” to make a “profit”, there is a risk that enthusiasm for that idea, may result in you shooting yourself in the foot. Best to get the data.

          • eveee

            Why would I need 5 PowerWalls for 10kw when I am already grid tied and get 10kw from the grid? I wouldn’t.
            As long as your storage is less than the load, all the storage can get used. You want to undersize it in grid tied application for best economy.
            One could have solar charging the storage or you could have just storage and charge it from the grid, converting AC to DC with a simple unregulated full wave bridge supply.
            A SolarEdge inverter is pretty cheap. A 5k one is about $1330. One of those inverters, a cheap unregulated supply, and a PowerWall is what you need to peak shift. There is a $0.19/kwhr TOU price difference in some areas of the country. Although the PowerWall pencils out to $0.12.kwhr without inverter, adding those other items is only going to increase this about one third. Thats still comfortably low enough to make money against the difference in power prices between peak rates and lower rate times.
            You don’t need 5 PowerWalls. The only sizing necessary is to make sure you don’t get more PowerWalls than your load.
            For Grid tied Solar plus Storage, you want to use up the storage in the evening higher cost TOU peak loads.
            7 kwhr is a pretty good size for that.
            For peak shifting, you can charge at night for at least 6 hours and discharge anytime during daytime TOU when there are loads available to absorb the stored energy. Thats most of the day and evening.

          • Robert Haylar

            Then, I am not sure I understand. If you already have a grid-tied system, then what is it? Solar is again different –

            performance depends on so many factors other than the battery.

            I was addressing the means of determining battery size, where the battery’s load is the household.

            Grid-tie does allow the battery to discharge at a more-or- less constant rate – so your method of cost calculation could be applied – but that is not really the point, because then you are engaged in a commercial practice of selling energy back to those who supplied it, and relying on very high utilization.

            Using batteries in grid-tied TOU is possible, but how much do you hope to sell back? Say, the low tariff rate applies for 8 hours, and you may output back to the grid for the remainder of the day at the batteries maximum C rate. If that were done, the charging rate must be a minimum of 2C, More capacity could bring that down to an acceptable level, but then cost rises.

          • eveee

            What grid tied system? One that uses batteries and inverter to avoid peak power prices. At a price differential of $0.19/kwhr, there is plenty of room for arbitrage to pay back the system.

            If the system is charged overnite for 7 hours, the charge rate can be 1C, not 2C. Same for discharge. The 2kw limit is a 3.5 hour full charge, but you don’t need to charge that fast. You have plenty of time to charge and discharge slowly. In fact, thats the whole point. If you do that, its easier to get the most self consumption. Its more critical the the energy storage incremental amounts are small enough than the power output as far as economics goes.

            The matter of whether the energy gets sold back to the grid is a matter of whether arbitrage is needed or self consumption. Arbitrage is easiest. The inverter can easily be programmed to be off during charging and on during peak use. You save by avoiding peak rates and by selling into the peak. If self consumption is desired, that takes load management.

            Take the simplest case. A battery and inverter. The PowerWall needs no charge controller or charger. Thats built in. You only need a simple full bridge rectifier and capacitor to convert AC to DC to power the input of the PowerWall. Basically, a cheap unregulated supply. It can have transformer isolation or not, because the input of the PowerWall can take the voltage. And you need a DC to AC inverter. A 5kw inverter from SolarEdge is surprisingly cheap, at about $1330.
            That does not even take much hook up.

            Now if you want solar, its a different deal. Maybe you want to avoid tier pricing. If you are on this pricing plan, the math is different. Its like the old cell phone limits. You go over and pay double. Solar makes sense to lower your tier rates. You can look at solar plus inverter and see if it makes economic sense in your area for that.
            Solar City guarantees a lower price than the utility in some places.

            A lot of people are asking if their existing PV solar can add a PowerWall. Thats easy if the existing solar is string tied, not micro inverter.

            The most expensive is emergency backup. A new breaker box and wiring critical loads is some work and cost. Thats not free.

          • Robert Haylar

            How did you calculate the $0.19W/hr, in detail? If a large differential is available, then based upon that differential, it’s not difficult to calculate a profit if the investment in the storage system is low enough. I don’t dispute that.
            But, there are other considerations. Like power output, battery C rate will limit the return, because it will limit the amount of energy that may be stored and retrieved.

            I will refer to the 10kWhr unit, because the cell type is known in that case. There are also differences in use of terminology between the cell and battery markets.

            A battery’s C rate is derived from the A/hr rate (capacity) of the cells that comprise it. In the case of the 10kW/hr Powerwall, the cell’s capacity is 3A/hr.

            C represents the unit of charge, the Coulomb. One amp is 1Coulumb/sec, so a 3A/hr cell will deliver 3A for 1 hour.

            That is the cell’s maximum rating, but also from which the other ‘capacity’ kWhr, is derived. Fully charged, an NCA cell’s voltage is 4.2V. The Watt is the unit of power (volt*amps), energy is (power*time), leading to 12Whr for the cell. The battery’s ‘capacity’ is the sum of all of the cells. A lot of cells are needed to obtain 10kWhr. That’s why the car has 6,600 cells.

            Like many, many, cells, Panasonic’s cell can’t be used at its A/h rating. – it would expire within a few hundred cycles, so must be de-rated according to conditions and life expectancy.

            The Powerwall’s specifications are vague, but assuming a nominal output voltage of 350V, and maximum 2kW power output, battery output current s 2kW/350V = 5.7A. The current at the cell level depends upon the way the cells are arranged, but from the number of cells necessary for 10KWhr, that would be around 0.7A or 0.7/3 = 0.23C.
            The maximum discharge rate would be 5.7A. Charge rate has no been specified.

            The battery management system (BMS) will disconnect the cell stack, if either the charge or discharge C rate are exceeded. Charge and discharge rates are set by the manufacturer, and can’t be arbitrarily determined. Charge rate may be lower or higher than the discharge rate,

            That’s another design level decision.

            The Powerwall’s certainly have a BMS, which is the ‘charge controller’ – but limited to on and off control. All lithium-ion batteries, except for incendiary bombs, have a BMS,

            There are ‘dumb’ and ‘smart’ chargers. The first type have an internal current limit that does not exceed the BMS charge limit. They are specific to the battery.
            Smart chargers use digital communication ( SMBus), to read the settings of the batteries BMS, and adjust the charge rate accordingly. That system was developed by Intel.
            What is the charge rate of the Powerwall? Does it have SMBus? The charge rate will limit the amount of energy that may be stored over some given time period, and indeed the number of full cycles that may be achieved.
            Be wary of “x cycles OR x years” warranties, for all batteries.

            The other end, the discharge rate, will also influence effective energy throughput, especially where the load varies as it does in a household. It’s not just Telsa’s battery that may not fulfill requirements, but others too.

            There is a difference, though. Bosch and Varta’s systems, for example, are sophisticated, and can control charge rates and manage loads, because they are designed from ground up, and where an appropriate cell was selected

            Telsa began with the cell, and are selling the outcome.
            It’s up to the purchaser to determine if that outcome is fit for purpose.

          • eveee

            The $0.19/kwhr in detail is simple. Its the difference between TOU and off rate. TOU rates in some states are well over $0.30/kwhr vs base rates in the low teens.

            lest you think that is too high, take a look at this SCE rate plan where the TOU differential is a whopping $0.37/kwhr !

            San Diego is EV country and those TOU rates will bring all kinds of solar, storage, and other methods in a hurry.

            The link is in



            Forget the 10wkr unit. It can’t cycle so its no good for economics.
            The 7kwhr unit is the one that an economic case can be made.

            You know, I already know about all that battery stuff you wrote. Instead of calculating it, look it up. Its in the public domain on Tesla discussion sites. Bottom line is, its in the box. What do you think Tesla is already doing with cars? You need to do some research on what the are already doing to catch up. Yes, they use BMS. And they use CANBUS to communicate. And there is a DC-DC in the box. To operate a BMS there has to be a controller, and a BMS unit on every cell. Thats a sophisticated system.

            That Tesla can price that at $3000 for 7kwhr, or at $250/kwhr for large systems selling to utilities and produce packs for thousands of cars on the road means its more than likely its no crude system.

            As I said before, you can’t predict every situation and maybe not all of the stored charge can be used for self consumption as opposed to driving the grid, but the charging and discharging can easily be done at levels lower than the 2kw rate. And doing the discharge at a low rate makes it easier and more likely to be self consumed. Thats a demand management issue. It can be handled crudely and still have a high percentage self consumption, but the best way is to upgrade to a Nest type system. Tesla is working with them.
            For residential grid tied its easier to lose economy by having too much energy storage capacity than too little power capacity.

            At $0.37/kwhr rate differential, you can do all kinds of thing to make money. If the calculations show its cheaper than grid over 10 years, it means the break-even is less than 10 years.

            Under those circumstance, residential market switch over is at a tipping point. We are there in some areas of the continental US. Count San Diego in that. And most of California, too.

          • Robert Haylar

            I took a look at one of those references. A great deal of discussion concerns changing rates and plans. Well, yes,
            That’s the risk you take when engaging in the commercial practice of selling energy, especially when the market is in a constant state of flux.

            The matter concerns selection of a battery, fit for purpose.

            If you already knew what I last wrote concerning the cell, then you are acting as if you didn’t, by suggesting that the battery could be charged by connecting a DC supply, and that the charge rate could simply be pulled out of the air. There is no evidence for the Powerwall suggesting you may do either. I am fully conversant with all of the means of capacity and state of health monitoring, and BMS construction. The car is a car, and not a storage battery.

            I wrote only of the general, in the hope of explaining why the cell itself governs the entire battery, A cell designed for portable use, will not make a good storage cell, and it shows in the poor A/h capacity.

            The ratio of capacity to power output must be determined
            when the system is first designed – and that includes the cell as much as it does batteries. Battery behavior is a reflection of cell behavior, Later changing either, would be difficult.

            If you can earn with one 7kWhr Powerwall, then adding more units will not reduce the kWhr unit cost, because for each incremental increase in power or capacity, there is the same incremental purchase cost.

            Double the power or capacity, and the purchase cost doubles. Adding more Powerwalls will not reduce Kwhr unit battery cost. ,

          • eveee

            No. The cell is designed for stationary use and is not the same chemistry as the car.
            You could be forgiven for not knowing that if you were not such a battery expert, but as an admitted expert, you should know better. You need to read more about Tesla batteries to understand why they are superior. You seem to be unaware of the different chemistry used in the 7kwhr PowerWall.
            The 7kwhr PowerWall is NMC.
            The 10kwhr PowerWall is NCA, like the car.


            Now that its been shown that power is not a fundamental limitation to grid tied applications, its pointless to continue claiming that it is without any explanation.

            Further, the system costs includes the inverter. Increasing the battery packs by 2 does not increase the inverter cost by 2x.
            The assertion that costs scale linearly is not borne out. Therefore, system cost per kwhr is lowered when increasing storage.

            The claim that charge rate could not be controlled is speculative. On who claims knowledge of BMS, and controller technology should know that the DC-DC converter could easily be programmed to control charge rates.

            Whether or not that is so, is left to be determined as Tesla reveals further details. Logically, it would be pointless to add the DC-DC for turnkey operation and make such a sophisticated system and yet have no provision for charge control.

            While we have not been given full details yet, in the investor conference call, Musk and Straubel gave notice that the PowerWall was designed to be added to existing solar systems seamlessly.
            We shall see.

            I need to spend time on other matters.


          • Robert Haylar

            The cell in the 7kW unit is NMC, but capacity will be less than NCA, Typically 2.2Ah to 2.4Ah, compared to 3Ah for NCA. (2.2/3 = 0.73 so 70% of 10kWhr fits within the limits)

            NMC cells are bit cheaper because they use less cobalt
            Tesla are using the cells at much the same C-rate, which is necessary to cell life, so the output is still 2kW.
            We use similar cells from E-One Moli. There is nothing Telsa can obtain, that is not available to everyone else.

            Indeed, purchasing 2 Powerwalls does not increase inverter cost, but doubles the battery cost. Some do amortize their cost over the entire system, which may lower the apparent cost of the inverter. Battery cost doubles.

            The DC/DC converter will not have infinite capacity. Do you know what that limit is?

            Power output is not a fundamental limit?
            I can calculate the return based upon income per hour, because: Differential*kWhr = Differential*hours*Kw

            There are only so many hours in a day or tariff period, and that (fundamentally) limits both the total and rate of income. Increasing power output will increase rate of return.

            Compare the Powerwall and Edison’s Lithium batteries, for a large system, that is designed from one appropriate battery, and not an assembly of inappropriate batteries.

            Powerwall 7kWhr, 2kW, $3000

            Edison 240V Lithium-ion Storage 37.4kWhr 80kW, $28,806

            40 Powerwalls to equate power output. $120,000
            5 Powerwalls to equate capacity = $15,000,

            The 5 Powerwalls will take 3.5hrs to discharge.
            The Edison will take 0.46hrs to discharge, so will offer higher return per hour of use.

            The Edison unit does all that with 16 cells. LiFePo4,
            Longer cycle life than NMC.

            Grid Storage with 18650?
            10MW/hr, 2MW output. Hey, 5 hours of storage.
            But what if the customer wants only 2.5 hours?
            What is the customer wants 4MW for x hours

            Tesla are stuck with the unnecessary cell cost, brought about by the clockwork relationship of capacity, C-rate and time, of their one and only cell.
            I don’t expect Seagate to announce that their new hard drive will be made from thousands of floppy disks, because they are oh so clever.

          • eveee

            Read how Tesla is doing it. You say Tesla cells are the same as off the shelf. Thats not quite true. They don’t actually use an off the shelf version of the cells that you can buy. For example, Tesla cells do not contain overpressure mechanisms at the cell end cap that individual use batteries do. We don’t really have full information about what they are doing.
            Do I know what the DC/DC capacity limit is? The battery is advertised as 2kw. Now thats not clear, because it doesn’t say which way we are talking, charging or discharging. So we don’t know for sure.

            What matters into end is whether it is economic or not.

            I showed that the 2kw limit does not limit charging over an 8 hour period of either night arbitrage, daytime discharge, or daytime solar charge.

            It all depends on the application what you need. I proved beyond a shadow of doubt, that grid tie does not need high power to be useful and economic. Now some other applications, that could be a different story.

            If anything, it shows how a high power rating can be of limited value for grid tied application. That is not always what you need to optimize for that application. It can go the other extreme where power is high, but energy is too low, too.

            Hey, here is a fundamental question. Here is Tesla selling all these systems to all those utilities. Did they all just miss this? Did the utilities and businesses that are already testing PowerWalls and PowerBlocks in the field completely miss this? Is the mass stampede purchase of so many PowerBlocks by power professionals a huge mistake? Aren’t they using them as utility grid storage? Isn’t that grid tied after all? Have all those professionals been duped, except you?

            Lets be real here. They didn’t make a mistake. It beggars belief that all those companies missed this flaw that only you could discern. Thats a bit hard to swallow.

          • Robart Haylar

            Good grief, eveee. Yes, the car uses a variant of the cell, where the PTC ( current protection), CID ( pressure protection) are removed to cut costs. The other item in the cap, the pressure relief valve, is replaced by burst points scored into the can. Will probably operate, but only at high pressure. Tesla attempt to replace the PTC with a wire that acts as a fuse, but the fuse limit must be high to prevent it blowing under normal conditions. Not a good idea.

            If a short occurs against the cell package ( the un-fused side), as may happen in an accident, the cell is unprotected. That is why protection is usually inside the cell. Also, If a cell or group of cells becomes weak, there may not be enough current to blow the fuse, yet enough to heat the cell.

            The internal PTC limits current, because over-current heats it. so there is also a degree of thermal protection there.
            Cell manufacturers included protection to cover themselves against risk, and have covered all that is practical at the cell level.

            Tesla’s car cell is an 18650 NCA. Packaging and cell chemistry define it. The cell has higher energy than some cells, because Panasonic managed to thin the electrodes and separator, so more could fit in – entailing a precision production process that adds to cost.

            NMC. Lower costs means NMC cells are marketed to the battery hand-tool sector as “high power cells”, but with lower capacity than NCA, Not so important when C-rate will limit use, anyway. NMC cells have a somewhat higher cycle life than NCA, but only at the expense of low utilization.

            Tesla supporters can do more harm than good, by raising expectations that Tesla can walk on water.

            No, I don’t know what the DC/DC converter is, but that says nothing about the 18650. I would agree that the Powerwall is likely to charge overnight, but that time can’t be changed either. I do not agree that you have demonstrated that power is not an issue, For one thing, I need not consider my ROI to be based upon the battery’s expected life, but shorter.

            Grid tie or not, low power output places a limit on earnings, because it limits the possible rate of earning. Of course, if a battery’s power output is too high, you may pay more for the battery than necessary. The same applies to capacity.

            For any system there will be an optimum capacity/power ratio.

            Tesla’s is fixed, while many manufacturers change the cell type when making small and large capacity batteries.

            You can see that 40 Powerwalls equal the power capacity of the Edison. I did choose the 80kW battery as an extreme example, but actually, it’s not the most extreme,

            The 80kW battery may well be too much for many, but there are many other batteries where the Powerwall fails when power is considered. Also, some of Edison’s ‘Solar’ batteries offer $/Kwhr costs close to the Powerwall’s; $475/kWhr v $428/kWhr, There will be many others.

            There is no one ‘best’ battery, and Tesla can only honestly share the marker sector that suits their one cell – yet they are claiming the battery is widely applicable.

            Did the utilities miss what I wrote? No, but there will be clauses in that contract concerning proof of performance.
            It will not matter much to the utility once all of the inverters and other equipment are in place, should the remainder fail. There will be many willing competitors who can replace the cell stacks. The utilities are besieged with proposals and bids, so know the market value.
            If Tesla’s system works, the utilities do not bear the costs incurred by a company over-eager to dispose of their mountain of cells. I don’t know of your views, but I don’t think engineering success comes at the expense of the profligate waste of cells, or materials in general.
            It’s not the first time that success has been touted according to apparent interest, agreements or cooperation, only to find that to be short lived.

            Tesla may just be able to compete when the ratio of storage and power are just so. Cell stacks using 18650’s were built 10 years before Tesla were even an idea, Nobody missed the opportunity to extend that idea, but let it be. Tesla do not know more about batteries, nor the markets, than their more experienced competitors. I would say less.
            I could ask you – the 18650 was long ago available to well established large battery manufacturers – why did they not take the opportunity? Some manufactures consider profit to be a good thing. Tesla are losing money.

            Ah…but Telsa get their cells cheaply!

            Think about Tesla’s position. At $200/kWhr, the dead cell cost of the 10k unit is 57% of the sticker price – it’s probably worse for the 7k unit – with no room to move
            because Tesla have bottomed out their cell cost. There is little opportunity for added value upon what is basically a box of cells. All battery manufacturers face that problem.

            Currently, for Tesla, buying more cells does not result in lower incremental cost, and there won’t be much of a gain, even when the gigafactory comes on line. Some analysts say cost will increase, It’s easy to see why Tesla can’t offer reduced incremental cost according to battery size.
            Telsa’s battery is costly to assemble, because of the large number of cells, the fire risk of NCA, and the need for cooling to maintain cell life, BMS is relatively cheap. The market is saturated with devices produced by the semiconductor manufacturers, and available to all.

            On the other hand, LiFePo4 is safer, and runs cooler, so the cells may be stacked closely together.

            (Yes, they do burn at a higher temperature, but the total combustive energy is much less than NCA.)
            It’s harder to start a fire with a LiFePo4, and they don’t liberate oxygen until temperature is high. Fire is still a possibility, but tends not to propagate. With NCA propagation is all but certain. I have seen many battery fires, because we are often required to test packs to destruction. LiFePo4, will also pass crush and penetration tests. I do not want to exaggerate the fire risk at Tesla’s expense, but the measures taken against propagation used in each individual pack, failed. It’s better not to pull the tiger’s tail.

            It is more cost-effective to pay more for LiFePo4, than to overcome the problems of NCA. The difference in $/Kwhr is not so great, China makes a lot of those cells, and they are not going to go away. Tesla get a low cost on a cell that interests few. We use them in small portable packs, but nowhere else.

            It’s a technical failure to proceed with that cell, but Tesla have no choice, I am often amused, and dismayed, by Straubel’s pseudo-technical chat.

            But, it’s hard to discount the effect of marketing and branding. If Tesla succeed, that will be the reason, not technical excellence – but at the expense of gouging resources to make that mountain of cells.

          • eveee

            I didn’t say power is not an issue. I made a case that your blanket characterizations did not apply in all cases. Thats all. And there are plenty of questions about those cases.
            And I also noticed that Tesla has done quite a bit of engineering, some of which you were unaware of.
            Teslas battery design places a limit on earnings?
            Every design places a limit on earnings.
            The thing is to do the math.
            But there you go again saying that Tesla has only one cell after I already showed you there are two chemistries.
            And you have not shown how or why an 18650 cell and all the rest of what they are doing is inferior.
            In fact, numbers just don’t bear out your complaint.
            The energy density is better than any other car manufacturer. So is the cost per kwhr. Those are essential figures of merit.
            They are essential, particularly cost per khwr, for both stationary and vehicular applications.
            I don’t care how. I don’t care why. And i don’t care who does it.
            Then you tell me that the GigaFactory won’t lower costs much.
            Are you really trying to be one in a long line of experts that have underestimated Tesla?
            I mean if I asked you to write on a slip of paper what your estimate for the raw cost/kwhr of the cells in the PowerWall, do you mean to tell me you would have written down a figure under $200/kwhr for a complete battery pack with BMS and controller sans DC-DC?
            You would be the only one besides Tesla.
            You say fires are certain with NCA, and ignore what Tesla did in their pack as if its not there.
            And you say LiFePo is cheaper. I think you are in denial. I don’t see that anywhere, from any data, any source.
            I totally get that you think LiFePo is safer. And I believed and thought everything you said long ago. And then I found out Tesla’s success and said, I have to understand why they are succeeding. You can’t just live in denial forever and pretend your competition has not left you in the dust and tell yourself its all a fraud and marketing when your vehicle just doesn’t compete with Tesla.
            You say LiFePo is more cost effective. Somebody should have already made such a vehicle already and beaten Tesla. They haven’t. You know why. They can’t. LiFePo just won’t give you the energy density or cost/kwhr. Thats why A123 folded. Great cell. Wonderful power. Great safety. Just too expensive for a long range EV. And too heavy to make into an 85kwhr battery back.
            A 30kwhr LiFePo pack will weigh close to 1000 pounds using standard Chinese LiFePo. You can go with pouches, but mechanical packaging and cooling are a problem. So is expansion. By the time you are done putting mechanical features around the pouches to keep them in place, they are not much less heavy than prismatic.
            Here. I will tell you right out. The reason Tesla is succeeding, even despite throwing all their patents open to the public, and still no one can follow their act is precisely because of all the experts that just can’t get over their notions of why they think Tesla can’t succeed. Really. Seriously.
            I was in your camp. But I can’t deny reality forever.
            Sure the Model S is expensive. And its popular and selling. The others could have done that. Still could. Instead all you see is stuff like the Cadillac ELR thats frankly a disappointment next to the Model S.
            So after all is said and done, the questing isn’t are you right and Tesla is just a fraud, but how did Tesla achieve success?
            Its a little late to be claiming they don’t really know what they are doing and have fooled not just a few rube residential owners, but how many power professionals?
            Sorry, not buying it.

          • Robert Haylar

            I did not make a blanket statement regarding power, but that there will be an optimum value for any system.
            There are means of calculating that optimum, but not by the means expressed in this article or generally found.
            There are papers from various academic bodies tackling that issue, Look them up.

            I acknowledged the NMC variant. How about LiFePo4?
            18650 capacity is around 1.2A/h, so 2-3 times as many cells, yet again. Tesla offer only 2kW output power, and there is no cost benefit from scaling, whereas other batteries offer a number of power options, and reduced incremental cost with scale.

            For all but the most limited of systems, there are better options than Tesla’s,

            Compare 2 similar batteries
            Tesla 7Kwhr, 2kW, $3000
            $/Kwhr $428
            $/Kw $1,500

            Edison 160Ahr
            8.32kWhr, 4kW, $5011
            $/kWhr $602
            $/kW $1,252

            But, the ratio of capacity to power cost is

            Tesla 428/1500 = 0.28
            Edison 602/1252 = 0.48
            That ratio is kWhr/kW = hr.

            Expressed in earnings $/hr, the Edison offers x1.7

            That can be further improved;
            36.4kW, 18kW, $17,167
            $/Whr $471
            $/Kw $953
            471/953 = 0.79 x2.82

            Benefits include those of LiFePo4. Low voltage means self-installation is possible, and there are a larger range of inverters and chargers. The battery may be purchased as is. Greater throughput can earn more. For TOU households there is no use in buying a battery that can’t drive high loads, otherwise, one must pay the grid price for the remainder. Select the battery accordingly.

            Notice that in all the examples, the $/kW > $kWhr.
            That’s because high discharge rate cells are more expensive than high capacity cells.

            The Powerwall is only a bargain if you follow the misdirection towards capacity. Many say Musk is a brilliant salesman. QED.
            The Tesla units are expensive in terms of useful output – and that is power, Adding more units does not improve the situation, because that’s just like adding more cells to the same unit.

            Nobody is mystified by what is simply cells in a box.
            You said it yourself. The costs of adopting one type of cell can incur costs that exceed the benefit.
            I checked with a reliable supplier. I can buy the Panasonic cell for $6.34. Single cell price. At $200/kWhr, Tesla pay $2.40. Do you not see the profit margin that has disappeared? There has been no improvement in the technology at all.

            You or anyone else can construct a 10kWhr battery by simply wiring those cells together. Test the performance at various C-rates. That is what you will get, and no more.
            The remainder of the design concerns packaging, BMS and fire protection. They add nothing to the cell’s behavior, but are means of mitigating side effects.
            What ‘high tech’ do you think Tesla have incorporated, that mystifies engineers?
            What advantage is there in taking a cell with a volumetric density of 676W/litre, and then, because of the necessary mitigations, reduce that to 50W/litre? (The 10kWhr unit is around 200 Litre, taken from the published dimensions.)

            In each module or pack, the cells are grouped in clusters of 4. Take two AA cells and place then together. The contact will be a narrow strip. Turn one cell through 90 degrees and the contact is just a patch. Take the one just turned, to be a cooling pipe.
            What to do? Make that pipe rectangular section, in an attempt to restore the lost contact area. Patent that.
            The pipe must be insulated to avoid electrical contact with the cell. Wrap Kapton tape around it. But, the thermal contact will not be good if a gap develops between the cell and the cooling pipe. Fit a piece of thermally conductive rubber, known at “gap pad” in the electronics world, to fix that. The remainder is a repetition.
            Free patents? You couldn’t sell them, that’s for sure.

            Even when other matters are ignored, Tesla are not doing so well.

            The car is a car. The battery is but one component.
            BMW 20kWhr/80mile
            Tesla 80kWhr/270mile
            Simple as that.

            Fires. The first was difficult to assess, because of the general flame, but the other in Mexico showed the flame front that is a hallmark of a propagated cell fire.
            Comparing that failure to the shortcomings of the 100-year old gasoline engine is weak, at best. A better comparison would be between batteries in a test chamber, subjected to identical conditions – which is difficult to ensure when in a car. Tesla have not published any data demonstrating they have improved upon what was already available. Is there any data to support their battery’s claims other than whatever they say?

            We make industrial machines of one type or another.
            For fun, we made a go-kart from those components and a large battery. First time drivers are initially shocked by the silent but enormous acceleration. Then, when accustomed to it, grin. Of course high performance electric vehicles will attract buyers.

            The apparent deals with the utilities may not be what they seem. Tesla are paying the bill to try and get a foot in the door,

          • eveee

            I just thought of some interesting observations. VRB flow battery maker Imergy CEO Watkins observes correctly that batteries are better at short term storage. His perspective applies to utilities, where long storage is 4 hours. Short term storage is under an hour, perhaps even minutes.

            But that kind of storage is high C rate.

            C is given in A-hrs. If its a 40A-hr battery that can discharge at 1C, it can discharge the whole amount in 1hour. Now that is actually short term storage.

            But the real issue here is that batteries can be designed for 25C, like the A123 types, but the trade off decreases energy storage.

            Based on that C, Lithium battery power is relatively easy to come by in capital cost.

            Not so VRB, which has an expensive PEM fuel cell. The truth is, VRB manufacturers want more energy than power because it suits VRB economics much better.

            So really, the PowerWall has 3.5 hours of storage at maximum power output. Just enough to compete in the 4 hour utility storage market.
            Their other offering has enough for 5 hours at max power.
            More C than that, and it starts to become a short term peak power source, rather than energy storage. Thinking of that, in off grid application, high power peaks would be shunned because they would entail unnecessary spending for peak power, increasing cost for an un needed feature. In fact, only 3.5hours storage at peak power output seems a rather short period of off grid. Perhaps it would be necessary to add more capacity to get energy storage even before one considers maximum power, particularly if batteries are used for a day or more. In that case, the PowerWall might have to be expanded by more units to increase energy storage, not power.

            But the short term storage market is really not the best market for VRB. VRB wants to be a very low C battery and give more long term storage and less power than a classic battery.

            Thats what Imergy CEO is saying when he says Lithium Ion is better for short term storage.

            Read his statement as, I can’t compete on a cost/kw basis. I have to compete on a cost/kwhr basis by lowering the cost of vanadium, getting it from the lowest grade, cheap sources, and operating for a very long time in order to pay back the high capital cost of the PEM with lots of energy storage.

            This is just the reverse of your argument that Tesla’s battery does not have enough power.

            Looks like lithium batteries win with many combinations of C and storage times up to 4 hours. At 1 hour and above VRB gets better, but its best applications are longer term.

            A VRB manufacturer would shun short term peak power applications.

            VRB has a number of shortcomings. It relies on abundant supplies of cheap vanadium and is vulnerable to PEM cost. Its best used where it can avoid those penalties.


            Looking at Watkins pronouncements that Imergy VRB is clearly better than lithium and his estimates of over $1,000/kwhr for same. one gets the notion that the PowerWall announcements $300/kwhr could come as quite a shock and result in some panicked board room meetings.

            At least its a contest now.


          • Robert Haylar

            I don’t take from the references you supplied that lithium ion is superior for short term storage, but that large scale makes VRB more economical?

            The 2MW output of Tesla’s proposed system is a limit in all cases. If the maximum grid demand were 2MW, and the utility’s generators capable supply that demand, there would be need of grid storage. Therefore, Tesla’s battery is of no use for systems where the maximum demand is 2MW or less, so can only be applied to larger systems. What about 4MW?

            In that case, the 2MW battery can only supply half the total demand, so increasing the cost per kWhr in comparison to a battery that can supply 4MW. Doubled, I would say. Better not to bother at all. The same applies as demand is increased, and the battery can supply an even smaller percentage.
            Grid demands are not conveniently ‘averaged’, but contain peaks several times that average. A useful storage system must be capable of meeting those demands.

            But, if what you suggest were true, there are already cheaper alternatives than Tesla’s. Earlier, I provided an example from Edison, where the storage cost is $474/kWhr for a built battery, compared to Tesla’s $428/Kwhr for the built battery,

            Tesla’s cell cost is $200/kWhr. If Telsa can be granted their cell cost to battery cost ratio, then so can the competition, arriving at a cell cost of $221/Kwhr for the Edison. Volume purchases should surely see that fall.
            The claimed general cell cost of $350/Kwhr for ‘lithium ion’ must be wrong.

            Why then are these LFP batteries not in use? Power output is much higher, cycle life is better than NCA, fire risk is lower. Compared to Tesla’s battery, specif energy is higher. Few cells, means fewer enclosure smaller BMS etc..
            There is no proof of performance from Tesla to support claims that their battery is otherwise superior.

            How well does Tesla’s use of 18650 stack up in the car?
            Not well. According to the SAE and other sources, Nissan’s Leaf has an energy density of 140Whr/kg. The precise weight of Tesla’s 85kWhr battery is hard to find, but appears to be around 600kg, resulting in 147Kwhr/kg.

            Compare the specifications of the cells used by each

            ( Panasonic) and AESC.
            Energy densities are (630) 309Whr/kg and

            (233) 155Whr/kg in favour of the Panasonic.

            In the case of the Leaf, the cell’s 155Whr/kg density is reduced by about 10% to 140Whr/kg when Nissan produce a battery from those cells, but Tesla manage a reduction of 75% Which is the more advanced battery?

          • eveee

            “But, if what you suggest were true, there are already cheaper alternatives than Tesla’s”
            Hardly. Didn’t bother to read the Imergy article?
            Says they expect $300/kwhr some time in the future but do $500 now.

          • eveee

            What on earth are you talking about built batteries? There are no circumstances upon which Edison has a lower cost/ kwhr than Tesla and no one else does either.

          • eveee

            Here are two answers.

            1. A study of Sydney Aus, household use, 92% battery use factor.
            2. Estimates of Solar plus storage costs competing with grid from

          • Robert Haylar

            None of that exclusively supports Tesla’s units, nor adresses any of my other points concerning the more able competing batteries.

            Your first reference shows all household loads to be conveniently less than 2kW. Total consumption is 12kWhr, or around a third of the average US consumption, corresponding to an average monthly bill of $43. Saving 100% of that amount would return the cost of the system in
            13 to 14 years.

            Sydney University concluded that going completely off-grid was not economically viable. There are similar studies from Spain ( where there is a high daily tariff differential), concerning TOU storage. The conclusion is that such systems will not be viable until system costs fall below 5c/kWhr. They also conclude that battery capacity should be around twice the average daily household consumption, and the discharge rate 5C where C represents the average daily power consumption. I did seek out papers myself, rather than to just suggest that you do.

            The dilution of Panasonic’s cell capacity at the hands of Tesla is not very flattering. The Powerwall similarly dilutes
            the cell’s volumetric efficiency to 50Wh/l. Battery experts? High tech? Hardly,
            There are some alarming errors to be found in the battery packs. For example, the BMS and the control gear are in the same enclosure as the cells, so exposing the first line of defense to heat or fire from the cells. Rather like putting the fire escape in the fire.
            There are more errors that are not so easily described, or so obviously clear.

          • eveee

            Facts bothering you?

          • Robert Haylar

            Facts such as the less than 2KW household power demand? Or that the calculation assumes that the battery can be 100% discharged to give 7kWhr? How about that 12kWhr daily consumption? The spreadsheet calculates the electricity bill for that 12KWhr, as $3.60 day – which just happens to result in the average US bill of $109/month, But for 30kWhr/day, of course. Since you support that calculation as fact, you must now agree that the US user will need 3 Powerwalls, rather than the one given in the calculation.

            The other facts, such as the low cost of the Edison cells, do tend to contradict Telsa’s claims that everyone else pays $350/Kwhr. Oddly, a 24kWhr replacement battery for the Leaf costs $5500, or $229/Kwhr. How do they do that? Perhaps it’s the headlights blinding them?

            Telsa can continue to bang the rocks together, while other companies quietly forge ahead with this Century’s batteries. I do wish Tesla success with that rock factory being built in Nevada. There is nothing wrong with Tesla’s engineers, I am sure, but with the one technical manager, who’s name appears again and again on a string of trivial patents.

          • eveee

            Facts such as proof that a Sydney household benefits from a PowerWall with 2kw max power despite your pleading to the contrary and that over 90% of the energy storage capacity is used daily.

            This is impossible according to your theory.

            Get adjusted to the cost of electricity in Sydney. Its higher.

            And no, the US user doesn’t have to use 3 PowerWalls. What the study shows is that savings can be had with a system sized less than the load.

            The size of the savings is proportional to the size of the load.

            Naturally, a larger load can be countered by more storage and solar.

            But it doesn’t have to. Thats up to the user.

            What is the capacity cost/kwhr of the selected Liithium Ion Edison battery you find so agreeable because of power output?


            160Ahr x 24V = 3.84kWhr, cost? $2,761.

            energy capacity cost/kwhr = 2,761/3.84 = $716/kwhr.

            But thats not their best energy capacity battery.

            Their eponymous battery is .. well. Iron after all.

            a 24V, 100Ahr= 2.4kwhr NiFe battery is $2,540.
            So thats 2,540/(2.4)= $158/kwhr.

            Considering that none of those comes with a DC-DC converter, there is not much comparison to the PowerWall at
            $3000/7kwhr = $428/kwhr
            or PowerBlock
            at $250/kwhr.

            Here is what I suggest. Consider the notion that the Edison battery is less suitable because its cost/kwhr is not optimized. It has too much power compared to energy. You have it wrong. Households don’t need C of 4 for optimum storage efficiency. The Syndne study proves it.

            Finally, a replacement battery for a Leaf does not constitute a battery for sale to anyone. You have to purchase a car to get that price and surrender the existing battery which still has value at 80% remaining capacity, not to mention salvage value of materials. It indicates that a Leaf battery is indeed becoming lower in cost, but lets not jump the gun here and compare unlike circumstances.

          • Robert Haylar

            OK. I have access to Sydney University’s paper concerning the viability of going off grid. The paper is copyright.

            I will make a summary of their findings.

          • eveee

            Robert. Your problem is that you lack an understanding of applications and are biased. Based on your replies, I don’t trust your judgement.
            The matter of C rates shows that in real applications, real lithium batteries, a C rate of less than 1 yields lowest cost/ energy or cost/ kwhr and meets power needs for both on and off grid storage. Apps can vary. Maybe an off grid app is operating a sawmill. Sure an LiFeP would do well there. Generally, tho, LiFeP loses on cost/ kwhr to the PowerWall. I am a fan of LiFeP, but not for every application, just as the PowerWall might not be the best chemistry for power tools.
            Your comments say it all. You bear a grudge. Why, I really don’t know. You think Tesla has done nothing. You ignore that they have brought low cost storage to many. And you ignore that thousands of power professionals have endorsed them and ordered 100s of millions of dollars of them, in a record one week.
            To ignore all that is just too much. Cheers.

          • Robert Haylar

            Well, eveee, the article you used as a reference has some errors. There is one that you can verify yourself, if you don’t trust me, The battery is discharged overnight, at an average power consumption of 0.12kW, There is some variation, but the total is exactly 7kWhr.
            Firstly, depth of discharge will be a more realistic 80%

            When a 4kW inverter is lightly loaded (0.12kW) efficiency falls, because there are losses that do not scale with output power. The load is about 3% of the inverters capacity. Efficiency will be around 75%.
            Add those together, and the effective delivery is about 4kWhr. The author states that solar+battery is neck and neck with grid-only. Now, the lead goes to grid only.
            There are also 25 tariff plans to consider. It’s not easy.
            I was not planning to do that, but take Sydney University’s calculated battery sizes, and compare.

            Tesla have not developed or improved batteries, not even on price. Enerdel, a US manufacturer, has prismatic cells. of 16A/h(2C). From an online retailer, they are available for $15 in 500+ quantities. That results in $260/kWhr. From that, any practical power level can be produced, using 58 cells. What would the price be from the factory if purchased in MWhr? $200/Kwhr would be easily met.

            Tesla’s forum puts the Leaf’s modules at $660/0.5KW. The battery would then cost $31,000. More than the cost of the car. If Edison can retail a battery at $471/kWhr against Telsa’s, $428/kWhr, what must cell’s cost? The higher C rate LFP would outgun the Powerwall in any application.

            So, where is the bias? Disinformation concerning prices is the only thing that leads anyone to think that Tesla have an exclusively low cell price. Anyway, there is competition from Australian energy (grid) producer AGL. They will be introducing their battery in June. It’s not a Tesla. We have at least one group of professionals who are not accepting the offer. Nor those choosing AES for grid storage.

          • eveee

            still banging that drum. enerdel raw volume battery costs vs a complete cabinet , bms, dc-dc system. cmon, you can do better than that. thats a quote from a company for what they could do if they had volume for raw cell costs. who generated that kind of volume? and you know what? enerdel almost went out of business waiting for an EV to soak up their planned volume that never materialized. and same with Fisker and A123.
            you are tripping all over yourself trying not to notice that Tesla succeeded and business acumen matters. Its also incorrect to say that Tesla did not design anything. Sure, not the cells themselves.
            You can say the same thing about any car company. Most don’t design many of the major subsystems, seats, interiors, even transmissions.
            Car companies are systems businesses. They figure out how to put it all together. They don’t make tires. GM doesn’t make cells for the Volt, either. No other car company even has a battery factory.
            GMs efforts on the first Volt were lackluster, with a Cd of 0.30. The Spark was a stopgap. The Bolt is a shared platform with an ICE vehicle. Commitment and dedicated EV platforms is what separates the men from the boys in EVs.
            Standalone battery companies like Panasonic or Enerdel and all the mainstream car manufacturers (except maybe Nissan) failed to notice that there must be a cell manufacturer, a pack manufacturer, and an end product user.
            Sony tried to make batteries for utility storage. At least its a system company, but they don’t have the volume to get the business humming without an EV to boost numbers.
            None of the other car manufacturers were bright enough or gutsy enough to take their EV packs and use them for utility storage. Some talked. But none acted. You snooze, you lose.
            You showed your bias with your emotional language long ago. Thats what bias is. Emotion before reason. Pre – judge.

  • Sohail Hasnie

    Great article. Quick question about theoretical end of product life assumption / definition.

    My understanding is that’s a concept for the car industry, the so called first life use for Lithium Batteries. At home, some us may use these batteries up to the point when the DOD drops to 40% or below (some people do that with mobile phones). Two Powerwalls after its 10,000 cycle use (capacity dropped to 40% DOD) will be close to a new one somehow, right? Lead acid batteries are used for a much lower depth of discharge

    Even after 40% DOD is reached presumably after 40,000 cycles (if I remember some of the charts I have seen from top Japanese manufacturers), there may still be a salvage value of the battery for UPS use or use in developing countries (I know some people are looking for used Nissan Leaf batteries for second life use).

    That may improve the numbers. Really appreciate a revised table for people who will like to run these batteries to real end of life.

    • PW

      I agree with you Sohail.
      I’m pretty sure the 80% mark is established by the range impact of the shifting power-to-weight ratio. There has been a bit of coverage recently on a study looking at the use patterns of drivers. IIRC the study concluded that even with declining range the autos will still be useful over many more years for a substantial percentage of owners without replacing the batteries.

      I’d also like to point out one extremely important aspect of a grid connected battery, whether it’s in a vehicle or home, is grid support in the regulation power market. This is a high value power market that, by some estimates could defray the costs of batteries by as much as $3000 per year. More information can be found by investigating vehicle to grid (V2G) or building to grid (B2G) power. Willet Kempton at the University of Delaware is a leading researcher in the area and maintains an informative website on the topic. While the technology for this service and the associated income stream is still developing, former FERC Chair Wellinghoff has mentioned it often as part of the future we can expect to see.

      • bink

        You people are not dealing with a full deck. You will not be providing grid services from you vehicle. Utilities need to rely on the services, energy and capacity being there. Someone needing to unplug their vehicle doesn’t quite fit into that equation nor should it.

        Why would you intentionally shorten the life of your EV battery for that. Grid services would destroy your battery in an instant but it would never get that far because you need complex battery control software and hardware components to make that work. You novice are killing me

    • vensonata

      Sohail. Absolutely. The specs on batteries are usually “cycle life to 80% remaining capacity.” I have cycled my own lead acid banks to 40% capacity before disposing of them. The 80% is just a way of evenly comparing the batteries not ultimately the bottom line for actual price per kwh. It is like a car in that way. Warranty to a certain number of miles, but some drive them into the ground, some go for a fresh new model after 50,000 miles.
      At some point though cycle life starts to become irrelevant. Ultracapacitors may have a million cycles. If you cycle them 3 times a day, they will last a thousand years. Do you want to pay for a thousand years of energy storage up front? So I suppose even 10,000 cycles like vanadium, is already getting out there. A good 10 year battery or even 5 is fine for most people since we live in an age of rapidly changing technology and sometimes drastic improvements and falling prices. So liquid long lifespan batteries might face unexpected psychological resistance.

      • eveee

        Hey vensonata. Can you give some detail on the lead acid experience. Some guys are keeping them over a decade. You seem to be saying you can go to 40% capacity. The simple analysis to 80% isn’t enough for that. Specifically, what calendar life and amount of daily cycling. That might show some concrete numbers on likely costs.

    • eveee

      Thats right. Cycle life degradation is arbitrarily assume to be to 80% as an informal standard. There is calendar life, which inevitably leads to failure, and probably increasing series resistance.
      Nissan made statements about repurposing used EV packs for utility storage, for example.
      The tables are a relatively simple comparison without interest. Take the second table, first entry column, the Powerblock.
      The second entry from the top is 10kwhr unit x 0.92 round trip efficiency x 0.9 average capacity degradation. The average capacity degradation takes into account a slide from 100% to 80% as the battery cycles.
      So that comes to 8280 khwr.
      The next entry is the total # kwhr by multiplying by a daily cycle for 5000 cycles. That comes to 4140000kwhr.
      The next entry is the value of the kwhr stored, based on the initial cost/kwhr stated by the sources and the kwhr capacity used.
      So $250/kwhr x 8250kwhr = 2070000.

      From that, you can see how changing the number of cycles can change the calculations. You can create your own spreadsheet. To change the calculation for degradation to 40% on a straight line, you need to put the average capacity in the first calculation as 0.7 or [(1.0 + 0.4)/2] instead of 0.9 used.

      What left out is that the series of kwhr paybacks is not equal to a lump sum. For real world numbers, you can approximate this as the sums at the end of each year.

      For 5% interest over 10 years, the sum is 7.722 instead of 10 times each annual sum. From there it gets more complicated, because you have to apply different math to each situation, interest rate, and years of life.
      Since fuel costs and electricity costs rise with inflation, the actual difference is less than the interest rate by the difference between inflation and interest. Thats usually much less than interest. Then the simple sum becomes a better approximation.

      In general, the more the cycles, the lower the cost. One caveat. When the calendar life gets really long, like 25 years, interest makes future payments worth much less, the higher the interest rate.

  • Billi

    —–Note that I’ve actually left out “competing” lithium-ion and lead-acid
    batteries in the residential section. Basically, even at a glance, it’s
    clear that they don’t compete with the Powerwall, so I didn’t bother
    finding all of the specs and doing the calculations. If you want to do
    so for any particular battery, I’m happy to add the info in, but I’ll
    need links or company spec sheets indicating cycle life, expected DoD,
    efficiency, and price in order to do so.—-

    hi zachary , i just got a quotation for a 630 ah at 48 volt industrial traction lead acid battery that would cost about 2650 dollar or 2400 euro from a company called bater in poland , traction /forklift batteries are rated capacity wise with c5 , so those 630 ah could be drained in 5 hours and that would stand then for a 30 kwh total capacity battery so a 5 kw draw constantly over 5 hours -to reach 80% dod , if one discharge this battery over a period of 24 hours , then the total capacity is about 850 ah or 40 kwh one can take then 1500 times 32 kwh out -80% dod-

    that makes 48000 kwh that costs then 0,055 dollar per kwh—without efficiency details taken into account — until the warranty runs out , that means that 80% of the total capacity is still available in this battery , so i have still about 32 kwh total capacity left after warranty time run out and a lead value of about 600 dollar ……

    two questions i have

    1, why should lead acid batteries not compete with the tesla battery

    2. where did you find the details , that a tesla battery can be cycled 5000 times

    tesla s 60 kwh battery in the sedan is only covered by warranty for up to 125,000 miles , this is about 500 cycles not 5000

    sure the 85 kwh model has unlimited milage ….. still where did you get your figures from

    thanks for the good work

    • eveee

      1.You want the short answer why lead acid batteries don’t compete with lithium?
      The lead acids won’t last as long. Even tho they can be 100% discharged, if they are used aggressively, their life suffers.
      By the time one lead acid set is done, and you replace it with another, the lithium would still be going.
      Home made EV builders have known this for some time now and have switched to lithium.
      Lead acid batteries are cheap, but they don’t deliver the calendar life or cycle life of lithium batteries. Theres lots more, like the Pukert effect, which reduces the available power output, but then the conversation can be really long.
      2. The statement of 5000 cycles is from 15 year life, in the investor call.
      You can look up many sources with cycle life like that. Its not uncommon at all. Depends on DoD and chemistry. Sony has one with 10,000 cycles. Tesla warrants the PowerWall for 10 years. Thats about 3000 cycles. A lithium battery is normally rated for a cycle life to 80% capacity after it degrades. Its still usable. So that rating is not a real limitation. Calendar life is more important. That can be extended by not leaving the battery fully charged for long periods. The warranty period includes both.
      This kind of performance is not unique to Tesla. What is, is the cost.
      This is not just a battery. Its a charge controller, BMS, and DC-DC converter. So its worth a lot more than just a battery.

      • Billi

        no short answer needed eveee 🙂 cause you expect only that lixxx battery’s will perform like you expect , that is a difference to my 12 year old lead acid one that powers our home , so there is expectation and real live figures , and i would wish that those 2 different shoes would fit !
        If you want to provide me a short answer , then provide a link , that documents that the Tesla battery is good for 5000 cycles …..

        I am just saying that the kWh taken out costs can be about 5-6 cents from a lead acid battery , a proven technology that is since decades available …. possibly a cent more pricy than stored hydro power , but far cheaper than that Tesla BMS banana

        • vensonata

          Lead acid, both flooded and sealed AGM have been analyzed to death by people who live with them (like me). Endless forum discussion are on Boating and RV sites. Engineers and mathmaticians, and starry eyed inventors share back and forth, sometimes decades long variations of experience with different types and brands under different conditions. Basically the topic has been exhausted. And I would say that your estimate of 5-6 cents is completely unrealistic. The very best possible price is on industrial quality lead acid “forklift” flooded lead acid batteries (see Hup solar one) any battery retailer will give a list of price per kwh hour. The most optimistic I have seen is 14 cents Kwh. That is the raw battery. No inverters etc. And with flooded lead acid the time required for maintenance is conveniently left out. Here is a new T-shirt slogan “lead is dead”.

          • eveee

            Oops. I just asked for the answer you just gave. Whats the deal with the 10 year life? Degradation to 40% or something?
            All the EV guys had to replace theirs in 3 years. Some managed five, but off grid can nurse them with light use apparently especially if you stay away from heat and excess charge/discharge. if its flooded, you gotta stick to maintenance. Screw up and its dead.

          • Billi

            …… you like to get me wrong , i did attach a quotation for a battery of 40 kwh capacity at 1500 cycles for 2650 dollars combined witha pip inverter charger from maximum_solar , would make a system that could work … in the us , with that high electricity consumption , a 7 kwh tesla battery is unrealistic and way behind the times to be grid interactive to become a buffer for pv peaks

            so please install sun power first , and then lets talk about storage and do not get fooled about a tesla battery , that costs only and produce nothing

          • eveee

            Heres a reference. What Vensonata is referring to is what he knows and I forgot. Lead Acid does not deliver what you pencilled in. Your assumptions are wrong. You think the rated numbers are what you get. They are not. For example, here is a nice reference you can read. I mentioned Pukert effect for a reason. You can milk lead acid for many years with very low DoD and low discharge current. But that is very limiting and you are not getting many kwhr out of it. But a Li battery can get much more DoD per cycle and high current and still have a long life.

            ” It can be seen that the AGM pack must be limited to a 30% depth of discharge to get comparable life to a lithium-ion that is at 75% depth of discharge. This means that the AGM battery must be 2.5 times larger in capacity than the lithium-ion to get comparable life”


            So when you count kwhr and cycle life, you get the wrong number unless you take into account that the AGM battery must be 2.5 times larger to get comparable life.

            That in no way decreases the other notable limitations of lead acid, like low efficiency during higher rate charge/discharge and many others.

            So when Vensonata says the debate is over, he’s speaking of all those articles and papers, but more importantly, he uses lead acid as does other people on this blog. If he could get more out of them, he would. And he is going with lithium. Personally, if I was going off grid, he would be one of the first people I would want to hear from.

          • Billi

            nothing unrealistic about my 5-6 cents per kwh , i have the facts , of prices , life , cycles …. you have nothing in your hands to show me that i can compare that tesla battery….

  • Anon

    This article tries with a lot of words to summarize something that cant be summarized. Batteries (no matter the chemistry) are so incredibly unique and tailored to a given application as to render any exercise like this mundane. I applaud the author for the effort, but there’s so much missed I caution the reader into drawing any meaningful conclusion.

    Big problems with the analysis, to name a few:
    1) no mention of power/energy ratios and how this will affect installed cost (for a 6 hr system, Imergy sells bigger tanks but Tesla sells 6X the batteries)
    2) no O&M costs included, but a conclusion posited on some mishmashed version of a lifetime cost.
    3) No distinction between DC system cost and installed cost inclusive of power electronics, shipping, construction, permitting, etc. (Eos quotes DC module only at $160/kWh)
    4) not even a hail mary toss or a wink at what any given installed function might due to first or operational costs……..freq regulation will price very differently vs. solar ramping vs. peak load shifting vs. etc.
    Utilities have 18 month procurement cycles for a reason. Even if all of the author’s many assumptions were 100% true, there are applications right now where each of the 3-4 options above would win in a competitive bid. Trying to compare these technologies without first laying out a specific application and understanding the requirements for that application is pretty worthless.
    This article is basically saying “whats better, a Camry, a Model S, or a BMW 5 series?” It depends on what the customer wants.

    • vensonata

      What you are suggesting might require a book, not an article. The intent, I believe, of the article is not to be exhaustive, but to introduce people, not “experts”, but ordinary people, to the idea of home energy storage and commercial energy storage, since it seems that the intermittent energy sources such as solar and wind will soon play a large part in our life. It is only sensible to start easy, a few prices, how do batteries work, what is their basic durability, their size etc. Nothing fancy. Many of you are perhaps “over qualified” for this article, your more refined and thoughtful ideas will be quite useful in later discussions.

      • geoff

        yeah good introductory article, adding some balance of system costs might be good though, there’s an article by tom Lombardo adding $1500 for inverter, one ‘rule of thumb’ is that balance of system costs are around the same cost as the cells, although the tesla powerwall has some system attributes. Maybe $1500 for inverter plus another $1000 for installation etc. 5000 cycles is definitely do-able using lithium titanate anodes, energy density drops due to lower voltage but not critical in home applications. I think tesla is giving 10 year warranty on cells. important to note its not just tesla with this stuff.

        • eveee

          SolarCity offers a system with the PowerWall and DC-AC inverter installed, for $5000 for a 9 year lease or $7140 purchased outright.

      • Robert Haylar

        No matter how well intended, misinformation is worse than none at all.

        For those not ‘over qualified’, perhaps another rule of thumb should be considered. “If something sounds too good to be true, it probably is”

        There is now sufficient information from this article and the comments, to say that none of the proposed systems would pay,

        • eveee

          Do you have any specific useful information to provide?

      • bink

        then you need to educate and not advocate. This article served no purpose in the education department. A lithium battery will not 100% discharge and last 5000 cycles.

        And before you say I am lying look up the EPRI Smart Grid Update report from 2014.

        The Southern Company (uitlity) tested a lithium 50kW battery for ancillary and peak shaving.

        Test Period: 1 year Test result: successful performance Battery life : severely degraded, end of life. Reason: temperature and application use

        In other words what Anon said

        • vensonata

          Bink, I would never suggest you were lying. The article gives numbers from the manufacturers, thats all. I am in no position to second guess them, however it is clear that several reputable companies such as Sony, Bosch, Saft, and Fronius clearly state that their complete lithium battery packs do give 5000 and more cycles at 100% depth of discharge. The “depth of discharge” though is a little tricky. From the customers perspective if they buy a 7kwh battery they can cycle it to a full 7kwh 5000 times. However the battery is actually a 9 or 10 kwh battery, therefore it is being cycled to 70-80%. This is to simplify the math for the customer, that is all. Fronius gives all this info clearly in their battery and inverter pdf. They state basically ” 9 kwh battery gives 6000 cycles at 7.2 kwh per cycle” with 80% remaining capacity. That 80% remaining can still be used and that quantity may in fact cancel the 20% loss of capacity having to be figured in to the final real cost per kwh. Bosch/Saft specifies 7000, Sony/Sonnenbatterie 8000-10,000!
          Many have suggested that the Tesla 7kwh battery pack is “really” a 10 or even 12 kwh battery cycling to 70 or 80%. The reason given is the weight of the pack. It corresponds to the Tesla car batteries weight as approximately 12 kwh. Beats me, but ingenious. Eventually somebody will take one apart and give the details. Or it would be nice if Tesla just cleared it all up.

          • bink

            let me explain something to you when it comes to utility grid applications.
            A peak shaving application requires a 100% dod at 100% SOC, there cant be any drop off.

            So you are right in the sense that they oversize the battery and limit discharge of each battery bank to 50% to accomplish peak shaving, but that was not my statement. 100% DoD was my statement and if you look at those brochures what do they all say max 80% DoD or no warranty coverage.

            Once the battery degrades in a utility peak shaving application it is toast. A contract may call for you to move 4 hours of energy capacity not 3.2 hours. You have just incurred a huge financial penalty from which there is no recovery.

            Now you send out the replacement team to swap out the battery or batteries along the way. Of course this is all covered under a very expensive warranty contract

            lets get real poeple

          • vensonata

            Except none of that applies to residential use. And commercial use such as a Walmart also uses milder demand applications for these batteries. You are referring to these gigantic grid peak shaving batteries which call for other characteristics which indeed sound inappropriate for lithium format.
            This article seems directed at the guy who wants a battery other than lead acid for his RV or yacht, or perhaps the off grid people. Then there are the German and Australians who already have solar and inverters and can actually save money by putting in a lithium unit.
            The utility battery will not be voted on by the general public. Electrical engineers and MBA’s will thankfully making those decisons.

          • bink

            You trying to pick and chose now, did Zach not include the PowerBox which Tesla is trying to sell to utilities ? and by the way the pricing for the Eos system and Imergy is AC systems not DC (Tesla) so at best Zach is disingenuous cause he knows this and did not make the distinction.

            This was a comparison price article do not try and change the narrative to suit your argument. The fact remains he did not do a apples to apples.

            glad you brought up commercial application and a Walmart. Walmart will need a battery that can provide power for uninterrupted power supply and power quality (motors for refrigeration) in addition to peak shaving (demand response) and emergency back up (energy applications)

            As I have already stated lithium is limited in providing all of these services. With a Eos or Imergy battery you can chuck the diesel generator and fuel. Lithium no

          • eveee

            The PowerWall does say 10 year warranty and 7kwhr.

            No warranty cost stated.


          • bink

            Lets talk about the commercial 100kWh battery so it is apples to apples comparison

          • eveee

            Do you have that information?

          • bink

            I was referring to Zach’s chart. I know that Imergy is selling a complete system, but none of these is installed prices and that can make the difference as well. Have you seen a lithium battery installation HVAC required

          • eveee

            Lets. What do you know about it?

          • Adam Devereaux

            let me explain something to you when it comes to utility grid applications.
            A peak shaving application requires a 100% dod at 100% SOC, there cant be any drop off.

            No it doesn’t. If I have a 100kWh battery and I have a peak shaving application where I only need 20kWh at each demand peak why did I need to discharge completely? There are many peak cap capacity requirements where you may partially charge and discharge several times in a day. Which is something some lithium battery chemistry EXCEL at.

            You are creating distortions of requirements. The reality is these batteries compete and have similar application profiles. Who cares if in reality your battery stack theoretically holds 120kWh but you only ever utilize 100kWh of it. As far as you know it is a 100kWh battery because that is what your peak-shave PPA is based on.

            This is similar to a 120GB SSD actually having 128GB of raw NAND capacity that it uses in the background.

            There is no problem having a preference. But you don’t have to spread FUD and vehemently defend your poorly explained reasoning. The reality is largely lifetime performance and LCOE is the king here.

          • bink

            This is why I hate dealing with you guys no clue!!!!!!!!!

            Did I mention kWh’s? or was I strictly taking capacity ?

            I don’t care if you have 5 peaks you cant over size the system, that is a costly equation and you are actually making my point because that is exactly how a standard lithium battery is installed for peak shaving application.

            All that extra battery that you are not utilizing is costly, vanadium 1:1 ratio to get 1kWh of stored energy, lithium 1:2 ratio to get 1kWh stored energy

            Utility puts out RFP for 4MWH’s of energy for peak shaving

            industry standard installation for redox battery 1:1 installed rating 4MWH’s + 4MWH’s 100% discharge of energy rating

            industry standard installation for lithium; 1:2 installed rating of 4MWH’s = 8MWH’s 50% discharge of energy rating to get to 4MWH’s

            This is not me, so you have to ask yourself why are they doing that? yo uand I both know why but the bottom line is you just paid for an oversized battery so how is that cheaper ? Its not. the ratios It even gets worse as you scale the capacity because the power is coupled with the energy and the balance of system start to scale in cost as well

          • Adam Devereaux

            “I don’t care if you have 5 peaks you cant over size the system, that is a costly equation”

            Doesn’t matter, IF the lithium batteries are cheaper for the useable capacity. You are arguing your preference. As an end user I don’t care what the “raw” capacity of the cells are, I care what is the specified and warrantied capacity during my PPA, contract or warranty period so I can calculate my ROI.

            “This is not me, so you have to ask yourself why are they doing that? yo
            uand I both know why but the bottom line is you just paid for an
            oversized battery so how is that cheaper ?”

            Doesn’t matter. Period. I get that you prefer the flow battery because you seem to be bothered by the concept of there being technically more capacity then you are using. But that’s all that is, your preference. If its cheaper its cheaper. If the flow battery manufacturers are able to make a cheaper battery then absolutely fantastic. Let the market prove it.

            I take it you are not a Volt owner. After all it’s such a waste that they only use a portion of the battery right? All Volt owners must be fools because of that? Give it up, stop making nebulous argument with reasoning that is simply based on your aesthetic preferences and stick to cold hard facts.

          • Adam Devereaux


            “the ratios It even gets worse as you scale the capacity because the
            power is coupled with the energy and the balance of system start to
            scale in cost as well”

            This is a function of the particular Lithium cells chosen, there are products that are capable of 1C output or much higher with no issues. You know that means a 1:1 output to storage ratio or higher right? I.E. you can buy a lithium battery system that could hold 100kWh but output 200kWh for 30 minutes (2C). Or 400 kWh for 15 minutes (4C).

          • eveee

            Tesla and other manufacturers probably won’t reveal it. I can tell you this. The charge controller handles it all. Just as you say, if the box says 8kwhr on it, and the charge controller stops charging at 80% of a 10kwhr battery pack, from the outside you can’t tell it from a 100% 8kwhr pack. In reality, the 80% operated pack gives way more cycle life. A 1000 cycle 100% discharge lithium can give 2000 cycles or more when discharged only 80%. Thats a big difference.

          • bink

            wow!!!, what a leap

          • eveee

            Everyone else has been curteous and respectful. You can discuss the matter and disagree respectfully and civilly.
            You have yet to explain how you know the content of a sealed container with a charge controller is discharging batteries inside to 100% DoD or is limited to 80% DoD.
            Everyone gets that full discharge is no good for lithium life.
            Even if you don’t see that, explain how a 20% increase in capacity to allow 80% DoD and double or more cycle life makes lithium costs unmanageable.
            Its quite instructive that VRB company, Imergy’s CEO has kinder words for lithium.

          • bink

            actually he has been saying what i have all along it is limiting in certain applications.For you to ask that question about the state of the battery is why I come at you guys, you speak as though you have some in depth knowledge because you pulled an article when there is more to it and you would only gain that knowledge through experience or knowing someone with experience.

            Everything in a utility and commercial battery installation is declared and measured via sensors controllers and remotely monitored from state of charge (SOC) to discharge capacity

            That wasn’t a good look for you to ask that question

            and the CEO from Imergy did not confirm anything Zach put forth

          • eveee

            Read it. And say it out loud.

            “Lithium has won a lot of projects for short, quick duration and it may be fine for that.”

            You seem to have trouble with words the head of Imergy has no problem with.

  • Carl Borrowman

    Great comparative article. My only hesitation on Powerwall is the relative output: I don’t think one unit would last that long at all for most households in the event of a cloudy week…

    “your $3,500 investment (plus inverter, plus electrician, plus solar panels) gives you one 15-amp circuit running draw for a quarter of a day. Time to think about chaining together two or three Powerwalls, and doing high-drain applications (laundry, dishwasher) early in the day on sunny days.

    The average American house uses 30 kilowatt hours a day, according to the US Department of Energy. A Japanese home, smaller and more efficient, uses about a third as much power. This suggests a typical US home might need 2-3 Powerwall units” (~$10.5K before the cost of inverter and installation is factored in, total ~$15K before the cost of solar panel installation, probably $20K-$25K total system cost… one would have to have a very hefty monthly electric bill for this to make sense, as it could be nearly 21 years before the system paid for itself with $100/mo average electric bill, so one would also need to figure in the cost of an extended warranty for 3 powerwall units, then perhaps replacement costs by the time the system finally begins to pay dividends.)


    Still, this is only first gen, and I have no doubt Tesla will greatly improve on this over time.

    • vensonata

      Average yearly household electricity use per household (Kwh)
      U.S. 11,698
      Germany 3,512
      UK 4648
      Italy 2,777
      Australia 7,227
      So before anyone worries about how they are going to use their new battery maybe they should wonder about their total electrical use.
      Secondly, they may need to understand ‘when’ they use electricity. For instance you find out that a fuse blows if you overload it.
      The age of thoughtless resource use is over. There will be plenty of electricity, but a little thought is required.

      • Robart Haylar

        It’s a complex matter isn’t it? One wonders what would possess 38,000 to immediately place an order on battery that does not exist,

        • vensonata

          Perhaps, just because a reservation is free.

        • Brand. Excitement.

        • eveee

          Do you have proof that the battery does not exist?

          • bink

            have they all been manufactured? kinda like that model X

          • eveee

            So you have no proof, but you made the statement anyway?

          • bink

            yep, where are they?

          • eveee

            I didn’t make the statement. Stop placing burden of proof on statements I didn’t make and back up your own statements with real proof.

          • eveee

            When you have proof they have not been manufactured, we will all listen. Until then, you have only added conjecture. However, we do know that there are systems in the field. So at least those systems have been manufactured. So this is not analogous to the Model X exactly.

          • bink

            wow, demo’s

          • bink

            where are they? reservations?

  • The links/sources are in the article. But here’s the quote for Aquion from the source: “Extremely high cycle life at 100% depth of discharge – 3000 cycles at 100% DoD; 4000 cycles at 80% DoD; 6000 cycles at 50% DoD”

  • Ivor O’Connor

    What is needed is an online calculator built with AngularJS. Then people can plug in the variables and see how it works for them. Perhaps this should a feature, one of many, that are available here?

    Many of the variables could also be chosen with a picklist making it even easier. And the results should probably be displayed in a table. So comparisons can be seen side-by-side.

  • neroden

    So it sounds like the Aquion is price-competitive, but Tesla’s got the largest residential option available.

    And it looks like the utility-scale market is not transparent about their pricing.

    On the whole it seems like the commercial / industrial market will jump for the Tesla option, because they probably get gouged by the “utility scale” guys (who don’t list prices upfront) and are too big for the residential-scale guys.

    • I agree with all of those takeaways.

      Would also add that the utility-scale market is getting quite competitive. It would be useful for me to interview some top people in the utility business and get their perspectives on the different options.

    • shunyata

      The problem with the Aquion, which these analyses miss, is the continuous power output.

      A single Aquion S20 may be rated at 2.4 kWh over a 20 hour discharge, but has a maximum power output of 450W. In other words, to cover a 1 kW load you are going to need more than two of these batteries. Ok if you want a lot of energy backup but far from ideal if you’re trying to load shift.

      You simply can’t compare batteries for residential applications without taking power characteristics into account.

      • eveee

        Thats a power/energy ratio of 0.45kw/2.4kwhr which comes to 0.187C and indeed that is lower than PowerWall, at 0.35C and Balqon at 0.5C. Lets take some residential applications. I used the C because that is a typical battery specification.
        Without going into great detail, I suggest 0.35C is OK for residential grid tied applications like grid solar, or arbitrage because peak power can be supplied by the grid. 8 hours of solar or night time charge is plenty of time to spread out charging. Discharge during the day or evening to avoid peak charges same. But demand management and efficiency are more important to avoid TOU.
        Its only off grid or utility/commercial/industrial where you really get into C.
        C depends on application a lot. A comparison would have to go over domestic energy and power use for each application. Thats not some simple comparison.

    • Offgridman

      On Aquion, the options shown for residential storage can be linked up to ten or twelve units, just like Tesla’s, to equal or be larger than the Tesla storage options shown.
      Also the DOE used them in a grid tied test a couple of years ago that proved out their cycle life claims.
      They are a new company so getting their storage might involve some delays or vary some in price according to distributor, but the same is true of Tesla’s right now.
      But their overall lifetime, and recently received cradle to grave clean environmental rating make them the better choice overall.

  • vensonata

    Aquion has 3000 cycles at 100% depth of discharge. Then they have 80% capacity remaining. And like Lithium batteries, you can continue to cycle them for several thousand more times but at accelerating diminishing storage capacity. The comparison used the same parameter for the different batteries, that is ” how many cycles to 80% capacity” That is the only way to get consistent values with different chemistries. The “price per kwh” might be much lower if every last cycle is used, which I think people with batteries might be inclined to do. I certainly do so, having lived on battery energy for 15 years.

  • Cheerful Clips

    These other Battery Corporations look wonderful but they all lack Elon Musk.
    Stay tuned.

  • lkruijsw

    Hi Zachary,

    I have also struggled in demystifying the costs of batteries. And at certain moment I realized is best to start directly with the cost of cycling one kWh (so divinding the $/kWh by the number of cycli).

    Then you can still add the additional costs, but by taking the price of one cycle you directly have some number that says something.

    Furthermore, there is caveat in battery economics. If you take a larger battery then the cycle load per cell reduces while you can benifit from the additional power and storage.

    Suppose, you make a car with just the batteries for 20 miles, then you have to recharge everyday and you cycle it to death. At the same time you don’t have much power and little range.

    So, Elon Musk realized that a car needs a large battery pack, so, the battery is cycled less, while you gain range and power. With this he started on the high end instead of the low end.

    With the Power Wall he giving the same message. The innovative part is that he sets the standard for 7kWh and 10kWh. Enphase is offering batteries of 1.2kWh. But then you have no power, very limited storage and you cycle it to death.

  • Egnaczak

    Why would we use lightweight lithium batteries for stationary storage ? Shouldn’t they be reserved for vehicles ? How much available lithium is there in the world ? Hopefully silicon PV cells and lithium batteries don’t create a destructive mining boom.

    • Neptune

      I wonder that myself. However, I’m not that worried about silicon PV, only about lithium batteries.

    • Frank

      I used to worry about that too, but it seems there is a fair amount of it, and they found a large deposit in Wyoming. http://www.treehugger.com/cars/vast-reserves-lithium-found-wyoming-could-meet-all-us-demand.html

    • Lithium is actually hugely abundant and not hard to source. Mark Z Jacobson of Stanford has done some good research and presentations on this.

      • eveee

        Also, recyclable.

    • neroden

      There’s plenty of lithium. The lithium refining can be power-intensive. Worry, rather, about the *other* metals in the lithium-ion battery, such as cobalt.

      • Egnaczak

        Are there competing battery systems out there that don”t have “other” metals or other toxic substances ? Seems like we have jumped on the lithium type but are there better batteries in terms of material extraction, processing, and recycling / reuse……..can there ever be a “green” battery ?….and how “clean” are mining operations to produce this stuff.

        How about more hydro that uses multi-use watersheds as its “battery” ? If you don’t have water in your backyard then community hydro……..like they did years ago when there was no grid

        • bink

          there are others, yes, vanadium used in VRB’s is sourced from recycling operations from the steel making industry and oil and gas, but it is not mined for that specific use.

          American Vanadium a VRB manufacturer will began a mining process in Nevada

          As to the toxicity of the electrolyte it is 3 times less acidic than your car battery electrolyte. the vanadium itself is not considered toxic in unconcentrated amounts

        • eveee

          Yes. Aquion, Eos, and LiFePo are a range pretty benign ones.
          Its not really a question of lithium. These all require mining, lithium or not, because they require materials. Steel and aluminum, even plastic have impacts.

      • eveee

        The statement

        ” “lithium” batteries typically use 10 times as much cobalt as lithium;

        is only true of laptop batteries of the LiCo Chemistry.

        Its not true of NCA, for example, where they are about equal, and Nickel is the dominant composition.

        The 7kwhr PowerWall pack is NMC, whose material concentrations are not listed in these references, and the exact mix is proprietary, but IMO, NCA is a reasonable guide for approximation.
        For reference, only a very small amount of a lithium battery is lithium by elemental weight, because lithium is a very light element. That does not mean the active metal is not heavy, however, as you will see next.

        Here they give composition by percentage. This is for LiMn2O2, which is about 33%. By weight, Lithium is much lower. The raw stock before processing to LiMn2O2 is Li2Co3.


        You have to download the GREET paper referenced in the comments, Table 2, Page 10.

        By weight, there is a table in this reference.


        Another for good measure.


        To complete the discussion, its necessary to take those numbers and analyze further.

  • dfgsdgdgsdf

    You’ll never get 5000 cycles out of a Powerwall

    • vensonata

      You may be a little out of date. Batteries are improving by leaps and bounds. 10,000 cycles are claimed by reputable lithium battery makers such as Sony. It depends on the chemistry. The powerwall has is specifically declared by Tesla to have 5000 cycles, it is not a wild guess. Normally that means 5000 cycles and then you still have 80% of the capacity of the battery remaining. There may well be another 2-3000 cycles left at diminished cycle volume but perfectly adequate for household stationary use. For EV it is different, it affects your range and possibly acceleration. But again, the Tesla car battery warranty expects 70% remaining capacity after 8 years, and perhaps even more. At that point you can still drive but have to charge more often. In my own lead acid battery banks I use them down to less than 40% remaining.

      • Efried

        depends on the power drawn – the figures might be right for 20 hrs discharge or 0.1 C

    • Ivor O’Connor

      I used to think this too. Apparently though Li batteries cycles can be tailor made. After learning this I thought why don’t they just make them to last 20 or 25 years like the solar panels and inverters. Then I learned Tesla expects battery technology to drastically change in ten years. In a good way so that you can buy much much more for much less. Hence they are saving you money by not building something meant to last 20 to 25 years.

      • bink

        “Then I learned Tesla expects battery technology to drastically change in ten years.That ‘s got nothing to do with it, the tailor made part is because of the coupling of energy and power in the cylindrical battery platform.

        They cant be separated from each other, so each battery system, bank or module is actually customized.

        This is more expensive and limiting in a grid, commercial or industrial application. In a utility grid application the platform that can stack or perform multiple applications will win. Lithium is limiting in that you can only perform a power or energy application but not both.

        For instance, I am a utility seeking energy storage for duty at a substation, I need to defer my capital transmission capital, investment, shave peak loads, perform area frequency and balancing, and provide reactive or leading power to minimize line losses.

        That is a combination of power and energy capability that is required. Lithium would not be the choice here.

        The battery would never survive because of the heavy duty cycle, it would be very expensive because you have to design two different types of battery banks or design a hybrid configuration with some other technology like lead acid and you would never fully utilize the system because most of these function could not be performed at the same time

        In other words back to the drawing board in regards to utility

        • Ivor O’Connor

          I simply paraphrased what Tesla has openly said. They are focusing on batteries that last 10 years. They expect they will be using better battery technology that will cost less and do more by then. Much like you don’t design computers to last forever because you expect much better equipment will replace it. (I remember paying $1000 for a 10MB SCSI hard drive. Now I pay $160 for a 480,000MB mSATA drive that performs 200x faster. (But despite this everything seemed to perform faster with the 20-30 year old hardware.))

          As for your concerns about shaving peak loads, performing area frequency balancing, reactive and leading power, etc.. You don’t go into what type of heavy duty cycle time is needed. On the surface it seems no different than the Tesla’s used day in, day out, as Taxis. If you have real expectations that need to be met state them clearly and then state why you think the Tesla utility power packs will not perform as designed.

          • bink

            obviously you missed the part about how it is technically impossible for the same lithium battery designed for frequency and balancing to turn around and perform peak shaving which requires 100% discharge of available energy with no loss of SOC (state of charge), cycling several times a day.

            Therefore the lithium battery could only exploit the ancillary services opportunity and not the deferred transmission capital investment, capacity, and transmission congestion revenue opportunities.

            In a utility regulatory cost benefit process the alternative investment has to have a life equal to or exceeding the equipment it is deferring to get a credit

            In this case the lithium battery is is not a good investment it is leaving revenue on the table.
            Ivor you are obviously uninformed on this subject matter to even equate EV driving with what a utility battery has to do and the environment it has to operate in.

          • Ivor O’Connor

            I call bs.

            Begone troll.

          • bink

            how about educating yourself before commenting

          • nakedChimp

            bink is involved with RedoxFlow and got his nose bloody a couple of times here already.. read him with a couple grains of salt.

          • Ivor O’Connor

            I doubt he has any experience with the industry or else he would not have insisted batteries must be designed to discharge 100% several times a day. It seems much more likely he’s been given some impressive sounding jargon barely strung together in the hopes of confusing people. Maybe he’s paid by RedoxFlow but I can’t imagine they’d want their name tarnished by people like him.

          • Damien

            Are you always this rude? Stick to discussing issues and forget about the defensiveness and insults.

        • eveee

          ” the tailor made part is because of the coupling of energy and power in the cylindrical battery platform.

          Cylindridal has nothing to do with it. The relationship between power and energy is expressed in units of C.

          “They cant be separated from each other, so each battery system, bank or module is actually customized.”

          Yes, chemistry can be altered, and some other things to improve C rate on a cell basis.

          “This is more expensive and limiting in a grid, commercial or industrial application. In a utility grid application the platform that can stack or perform multiple applications will win. Lithium is limiting in that you can only perform a power or energy application but not both.”

          This is where you lose me. Sure there are trade offs in battery design. As batteries improve, they increase both power and energy density, series resistance, etc., lowering those limitations.

          While a VRB battery can increase energy by adding a bigger tank, it cannot increase its PEM size without adding more PEM. A lower C can be achieved than the original, but not higher. That can only be achieved by adding more PEM.

          In the end, Power capacity must be met, no matter what. There are some battery types with radically high C, not merely because of chemistry, but materials. A123 produces cells with C rates as high as 25. But that is only one figure of merit.

          A utility can select the C according to application if necessary.

          VRB has the advantage that it can achieve a high C at low cost. Lithium has lowest cost for high cycle life when C is closer to 1.

          • bink

            you have no clue what I am talking about so I know you will not listen. Cylindrical has everything to do with it.

            Ask yourself, what battery architectural platform’s power and energy components are physically decoupled and can be scaled independent of each other ?

            Answer: Redox

            In a peak shaving application 100% energy discharge is required. Obviously discharging a lithium ion battery to 100% DoD will greatly reduce battery life, greatly impacting capacity through degradation.

            It is standard practice in the energy storage industry to limit DoD to 50% for utility peak shaving applications.

            to store 1kWh of energy using lithium you have to install 2kWh (1:2), A redox battery is a 1:1 ratio. per stored kWh of energy. As you scale the system it gets worse.

            Energy and power are coupled in the cylindrical battery cell. That is why the lithium battery is designed for either power or energy applications but not both.

            You are now seeing hybrid battery configurations (lithium + advanced lead acid) trying to widen the revenue opportunity for lithium

          • eveee

            If you would stop being insulting and notice what other people are saying you might get somewhere. Do you know what C is for batteries? It has nothing to do with packaging. If you don’t get that, you have some catching up to do.
            You keep repeating what a flow battery is and does as if everyone is deaf and has no intelligence. We get it.
            Stop hammering the same thing. You have already made your point and we do get it. You don’t get that most people do get it.
            We are not all against VRB or flow in some vast conspiracy.
            I tried to get you to read Imergys CEO statements about flow vs batteries. Apparently, you don’t get what he already knows, that flow and batteries both have a place in the market.
            Trashing batteries is not going to make flow any better or make it work in applications that batteries do better.
            If you are more of a proponent for VRB than him, something is very wrong.
            And you are wrong about some of what you say about flow. You have no choice but to increase PEM to increase power. There is a limit to flow, too. Energy can expand with tank size, but not power.

          • bink

            wow , clueless, it has everything to do with the packaging. Flow is decoupled “No Waste” “Optimal Sizing”

            You can scale power independent of energy in flow platform or vice versa or scale both.

            If I want more power I add more power cells nothing else is required to be done (plug and play),

            If i want more energy you add volumes of electrolyte (that’s it) the tank is oversized already to allow increased capacity

            That is the advantage. You obviously are the one who does not understand battery platforms

            What the guy from Imergy is saying I have been saying for 3 years and he just came on board over there.

            I started in this industry in 2010. I know redox that was my first technology, so stop quoting someone who just got to the party even if he is right

            I agree with him there is a place for lithium in power applications only because to use a redox battery is to underutilize it in that type of application

          • bink

            No limit to flow unless there is no power source, continuous flow, can discharge over 15+ hours

        • olsen

          There are many applications as it relates to the utilities. Currently, there is more than 200 installations totalling more than 275MWh of lithium ion battery based energy storage in operation according to the US DOE Energy Storage database. Of those 36% are related to traditional utility operation and 25% is related to integration of renewable energy. So saying that lithium ion does not fit the bill is not necessarily true. But just like with everything else you buy, the technology must make sense for the application, and lithium ion is not the only technology out there. For T/D deferral you may want to consider the sodium based batteries, and for load shifting you may want to consider flow batteries.

    • Jacob

      Look up the Wikipedia page for Powerwall.

    • eveee

      So how does Sony get 10,000 cycles out of its lithium battery?


      • bink

        really? the problem with your question is lithium’s well documented history of failures in commercial and utility applications due to the type of operations it is asked to perform

        • eveee

          Then provide them with citations.

        • Offgridman

          Bink, it is necessary to agree with Eveee here. You come on saying what an expert you are, and at times get quite crude with the language in expressing your opinions.
          I am not even going to attempt to disagree with you that storage for utilities could be less expensive using VRB, because I think you might be right.
          But half of this piece talks about using storage at the residential level, and due to the monopoly choke hold that they have in a some places the Tesla option with its warranty could be economical.
          But for myself having said goodbye to the grid almost ten years ago the storage option with a life that might extend into two or three or maybe four decades looks quite promising.
          So here for the third time this month, please tell me where I might find that VRB option. It doesn’t even have to be a direct link. Working from my phone I also have a problem getting links to copy and paste correctly. But could you just provide a company or product name that I can look up for myself that is currently selling on the residential or offgrid basis? I have tried all of the commonly known ones and only get a response that they will sell a product that is better than avoided peak costs, or cheaper than using a diesel generator. Neither of which applies to me, because my wind and solar are fairly well balanced and a few Kwh of storage are needed for over night use, with a max of 20-30 so I don’t have to climb on the roof to get the snow off the panels from the rare storms.
          Can you try to help with this? Otherwise it is looking like the Aquion option is going to be the best bet for storage that can last for the time periods I want to allow for (20-30 years).

          • bink

            i actually think it is your best bet to go with a Aquion battery. It is simple and does the job no bells and whistles required.

            A VRB is expensive when “UNDERUTILIZED” not because of pricing, but because you would not be using all its attributes in a residential application.

            currently there is the issue of footprint and weight in a typical size garage. Its not that huge but you wont be hanging on a wall, probably need a couple square feet

          • Offgridman

            Well for myself the size and weight are a minor issue because I am already using over 3,000 lbs of AGM so used to dealing with that.
            From my perspective it isn’t just the Kwh’s that will be available, but for how many years it will be possible to access them to come up with a per day cost long term. So due to light cycling most of the time on the current pack it should be possible to get at least ten years.
            I still have a few years before it will be necessary to get something new but the possibility of over sizing somewhat to get 20-40 years seems very appealing.
            Now Aquion is saying that not going under 50% DOD will mean 80% left after 15+ years, so enough to get me another ten years or so.
            The VRB from Imergy is supposed to be available in a variable 5-30 Kwh size that would suit my needs according to a PDF from them, but no matter the talk of long life they only offer a ten year warranty. Maybe because of the pumps wearing out? And so far no specific pricing.
            Fortunately there is no rush on figuring this out, so I can wait and see what happens with these various markets for lithium, vanadium, or aqueous. But the ones that can get certified for the very long lifetimes would seem to be preferable for the offgrid or micro grid, even if a little more expensive to start with. Unlike a utility thinking of a specific pay off in ten years, we get thinking about systems that can last as long as our solar panels and still be usable for our children.

          • bink

            as far as longevity a VRB is your best bet. the energy component will never degrade and the balance of system is PV pipes and pumps a 10 year industrial pump with propeller replacement every five years will serve you well.

            The cell stack is at least a 10 year life, basically you will be replacing certain components that are least costly

            Imergy is likely to offer a electrolyte leasing program, probably 30% of total system costs. so you can see how they can match or beat lithium capex pricing easily (remember they are selling you a complete system

          • eveee

            I just did a calculation and I can’t remember who it was for, but the topic was shipping costs for Aquion because its heavy. I concluded it did increase the cost by some low double digit percents, so significant, but OK if you need that.
            Then we discussed how they just had a layoff. Darn it. Well I hope they succeed. We need that kind of long life non toxic battery. Looking at their stuff, it seems pretty robust, only thing I didn’t like was the efficiency and the pulse tests plus efficiency make me worry about series resistance some. All that could mean exactly zip to you if you are off grid and already nice to your batteries which is almost certain.
            If you got lead acid to your off grid location, you can get probably get Aquion there.

          • Offgridman

            I have to admit to a preferential attitude towards Aquion, first off as recently proven their very low environmental impact by the cradle to grave certification.. Secondly the model they have developed of being able to make them more locally sourced materials that when expanded will mean smaller factories able to turn out similar capacities of storage without the industrial scale processing needed for lithium or vanadium.
            It is also nice that their cycle life times short and long term were proven out by the DOE a couple of years back in a grid tied application.
            As to the layoffs, I didn’t get a chance to get back to you last time, but searching back through the references it really looks like hype from a local financial reporting source. A layoff of five redundant positions out of a staff of 125 at one facility doesn’t indicate to me a company in dire financial straights.
            As for the shipping, yes right now that is expensive, just this year has production gotten high enough that they can offer sales for smaller applications beyond their grid scale orders. These are all still coming from the one plant in Pennsylvania,and being for immediate delivery of course the highest shipping rates will apply. But from their website it is obvious that they are trying to build out a network of more local distributor/installers. Once these are certified they will be able to send them the packs dry or empty, with the electrolyte (mostly water) being added by them and the packs sealed and tested, which all has to happen at the factory right now.
            The price that Zach found for residential use right now makes them look reasonable, but if you figure that there’s no need to do a 100% depth of discharge on a properly sized pack every day. (something people using lead acid are used to). Then you have to start doing your numbers where these packs will still have over 80% capacity after 6,000 daily cycles, and like with lead acid still be usable for a long time after that at the lower capacity.
            I like what Zach did here for the residential grid tied applications, but it doesn’t account for how like in my offgrid situation and I would suspect for isolated micro grids you not only account for the number of Kwh’s you will be getting, but for how many years out you can plan on accessing them, to end up with a per day cost long term of having storage.
            As an explanation of what I mean, while I have switched to a AGM for my main storage, there are still some of my original cheap deep cycle pack from ten years ago that while diminished are still being used on a regular basis, even though four years ago they were down around their 80% point.
            One last thing on the price point, with Aquion saying that grid scale storage is going to be in the 250$/Kwh range, and just in this first year of residential sales the prices being in the 500$+ area. In a few years with a proper distribution network and more competition among the various storage companies, it would seem reasonable to expect that residential price to come down around the 400$ area. So with longer cycling capabilities and a better environmental impact it will still stay competitive with lithium for people where the size isn’t as much of an issue.
            We will have to wait and see how this all plays out though, I am very happy to see all of these different competitors coming out against the old model of lead acid being the most practical for off grid storage. And equally as happy that there is at least 3-5 years before it is necessary to make a choice about what will be done for myself for storage, so these market forces can play out, before deciding what will be the best option for keeping ourselves offgrid and using renewable power long term.

          • eveee

            Yes to all of the above. A lot of people are noticing that for both EV and off grid purposes, oversizing energy capacity is the way to go for long cell life. For off grid, it sounds like folks get that there is no need to stop at 80% of initial capacity. You can go way past that if you have oversized your capacity and that way it works and lasts much longer.

            Its understandable why some one off grid would not want to replace their system every 5 years.

            Only a while back projections were only showing $250/kwhr by 2020.


          • vensonata

            There is a youtube video by a fellow who spent 2014 searching for a Vanadium battery he could buy for his off grid lifestyle. He tracked down every lead. The smallest he could find was 40 kwh. That would give a full 40 kwh for at least 10,000 cycles but the companies wouldn’t sell it to him. The next size up was 100 kwh. That he could buy. The price was $100,000. So $1000 kwh. But, but, but…. you had to be in their sales service area, because, you see it requires and annual visit by a technician to service pumps etc etc. How much was that service visit? They wouldn’t say. There you have the only known residential price disclosure I have ever seen from behind the Vanadium curtain. I myself did extensive inquiries into vanadium flow batteries 7 years ago. The original company was located in Vancouver, where Zach is visiting at this moment. All fog, fog, fog. Just not ready for prime time.
            As Bink, in his inimitable way says, there are a few working prototypes out there…very few. We can only wait and see but I suspect it will never be for the likes of you and me.

          • bink

            First get your facts right and stop making up stories.

            There are three major VRB manufacturers based in the US, at the beginning of 2014, only one of them had an commercially available product, so I know you are making stuff up, the other is just now demonstrating with project partners, and the other is under the radar.

            No they are not going to sell you an individual system their target market is commercial, industrial, utility.

            Service technician is part of warranty service for commercial, industrial, utility clients which your supposed friend was not

            the original inventor of the VRB was a professor at The University of New South Wales.

            Second, VBB was the second company to acquire the VRB IP.
            They then became Prudent Energy purchased by a Chinese company.

            VBB, did nothing but hold on to the patents and did nothing to advance the technology. Patent protections expired in 2009 about the time VBB became Prudent Energy.

            The Department of Energy at the Pacific Northwest National Laboratory (PNNL) started research into advancing the chemistry and cell stack that year.

            In early 2012, scientists at PNNL announced a breakthrough in the development of a vanadium electrolyte chemistry which addressed three major issues with VRB as identified in prior generations
            1) low energy density (almost doubling of density, now containerized)
            2) limited ambient temperature operating range ((now operates in -40C to 50C+ environmental conditions, (no HVAC required)
            3) electrolyte fouling (no gelling or fouling)

            The inventors of this new electrolyte are the founders of UniEnergy Technologies (UET). the system has been tested at Sandia National Laboratories and two other independent private sector organizations.

            one installed battery in California,and one being developed two in Washington State, and several more being developed elsewhere.

            Get your facts straight

          • vensonata

            I am merely citing a youtube video by an earnest fan of Vanadium. Those were his experiences, not mine. Watch it yourself. I have never met him and he is not a “friend”. My experience personally was also great interest in Vanadium, but despite phone conversations and emails it was not possible to actually purchase one. End of story.
            Now, I realize you have a compulsive habit of flinging accusations around. It is a bad habit indeed. However you are “useful” on this site as a pretext for answering questions about batteries in general. In other words, we need a “loose cannon” on deck to bring information to light. Generally your mannerisms are unsuitable to the atmosphere of a green website, but a joker has a role.

          • vensonata

            Oh yes, did I mention that “American Vanadium” is the Canadian company which operates out of Vancouver. Interesting name. Kind of like “Canadian Solar” which is Chinese. Ah, but what is in a name? A rose smells as sweet by any other name.

          • bink

            I am familiar with and have had business dealings with American Vanadium and no you would not have gotten a battery from them either.

            I wont say anything about them. I will only say that they do not have the benefit of American invention.

            Several IP licenses for the electrolyte were handed out by PNNL for the VRB industry, that is how much the government beleives it is the battery for utility and industrial use.

            In several months, the energy storage world will be seeing something disruptional coming from VRB technology.

            i am not going to say much about my involvement in the vanadium battery world least I give myself away to some of my colleagues

          • eveee

            LOL. I can’t stop laughing.

          • bink

            me too

          • nakedChimp

            hehe, thanks for the chuckle 🙂

  • GCO

    Thanks @ZShahan, good read, although at least the first table compares not-yet-available vs established products, and probably 2016 vs 2014 pricing (or before), so we’ll have to wait another year+ to make a fair comparison.

    Next, estimating whether daily cycling could make storage pay for itself sure is interesting, but this is Cleantechnica, not Financialtechnica, so I can’t help but wonder: what about the environmental footprint?

    I’ve read (some of) the Renault Fluence vs Fluence EV very detailed LCA, and it seems like producing batteries is so resource- and energy-intensive, it takes years for an EV to repay this debt compared to a gasoline vehicle.
    Now, merely time-shifting electricity seems a lot more benign than avoiding burning some oil derivative in a very inefficient engine. Are there situations in which residential batteries can bring enough positive environmental impact to compensate for their own?

    • eveee

      On the it takes years for an EV to repay this debt idea, not so much.

      Try this:

      “Using the Dunn et al. estimate, the Leaf battery’s footprint is suddenly only about 1.2 tons CO2. An ICE (the engine) takes more energy to build than the analogous EV powertrain (sans the battery), by about half a ton (these are Climate Central’s numbers). So using the Dunn et al. numbers for the EV battery instead of the inflated ones, the overall footprint of producing a Leaf-like EV become 8.2 tons CO2, vs. about 7 tons for a Prius-like hybrid and 7.5 tons for a plain-vanilla ICE compact. Not negligible, but not dramatic either. In fact, the difference is arguably smaller than the level of uncertainty about these estimates! Now, the Prius can barely drive 5,000 miles on its starting advantage, and ICE compacts can barely leave the parking lot on theirs.”


      The situation is undoubtedly better today, with a cleaner, less coal grid, and higher energy density batteries.

      • stanson

        For 100% electric vehicles, remember many non-engine components will not be needed at all such as the exhaust system, cooling system, etc. I assume there is a substantial amount of CO2 generated by manufacturing these.

        • stanson

          Meant to say CO2 NOT generated…

        • neroden

          Absolutely. ICE vehicles have a horrible CO2 profile.

          Unfortunately this sort of analysis doesn’t apply to home batteries vs. diesel / natgas generators, which are relatively simple compared to an ICE automobile.

        • eveee

          Yup. An ICE engine creates more carbon than an electric motor, too.
          There is less CO2 in EV repairs, too.

      • GCO

        What about using analysis from the automakers themselves instead, who might know a thing or two about what actually goes in their products?

        Renault: http://group.renault.com/en/commitments/environment/environmental-policy/
        Nissan: http://www.nissan-global.com/EN/ENVIRONMENT/CAR/LCA/

        Note that in both cases, manufacturing an EV has a significantly higher footprint than its ICE equivalent, but electricity is much cleaner and more efficiently used than gasoline so the EV catches up quickly enough — again, see Renault’s and Nissan’s numbers, a couple years, strongly depending on assumptions; faster if driving more, if the grid is cleaner (this is indeed improving), etc etc.

        Anyway, it’s clear that manufacturing batteries to replace refining then burning oil is overall benefiting the environment.

        Now, are batteries worth their environmental toll when their only use will be time-shifting grid power usage instead, ie using less grid electricity at a certain time but more some time little earlier?

        It seems unlikely, but I’d love to see better numbers.

        • eveee

          Yes. Read what it says. It says the Leaf has 40% lower emissions by 100km. Thats about 60k miles. That means it breaks even at less miles than that.
          Nissan does their figures for what they call an equivalent vehicle and comes up with a larger starting delta than the ref I gave. Still, even at that, it looks like it catches up at about half way, or 33k miles.
          So take your pick. It doesn’t take that much for an EV to outpace an ICE. The differential will only grow as the grid gets cleaner and batteries get better, which they seem to have recently.

    • 1) Indeed.

      2) Good point. Something to research.

  • Otis11

    Did you remember to account for the 30% Federal Tax credit? That should lower all of your numbers by 1/3 for home users…

    Also, California has a 60% tax credit – which combine with federal would make all of your numbers 1/10 of what they currently are… I’m not aware of any other states that have similar incentives, but could be uninformed.

    • Nope. Thanks. Will add a note at the top.

    • Jacob

      Most people do not live in USA.

      • Otis11

        Understandable – while these will arguably have a much larger impact in Australia where they are already economical, (I’d assume they are also economical in Europe, but have little hard data on hand to support this assumption) I think the US is a very large potential market for this product (and it is currently the target market – though they have immediate plans for expansion). Given the numbers provided, it seems like this is fairly uneconomical in most of the US, but after the Federal Tax Credit it starts to make sense in a few markets (Mainly HI and CA) and with the California state tax credit it’s highly economic in California…

  • Jacob

    What about Sonnenbatterie. They use Sony LiFePo4 cells and say they can be cycled 10,000 times. With a DoD of 80%.

    Sungevity said they will sell these batteries for $10k in USA. With a usable capacity of 3.5kwh I presume.

    • The price is still too speculative. But would like to include them once they provide a number. Also need efficiency, but that may be online.

      • Robert Haylar

        Sorry, but all of the calculations are incorrect.
        Capacity does not degrade linearly over time, and is temperature dependent. Initial capacity itself, is also temperature dependent.
        Electrochemical activity goes on 24/7, so there is also a battery calendar life. Cell manufacturers can be “economical with the truth”

        Reduced capacity is accompanied by reduced Coulombic efficiency. More heat is generated within the battery. Charging time increases, so more waste heat is also produced by the charger.
        Power output will also reduce as a result of increased cell internal resistance. Overall efficiency will degrade much faster than indicated.

        The value of any of the mentioned batteries cannot be calculated from cycle life. The calculation assumes a Panglosian world, where
        100% utilization is possible, and does not at consider that power demand will fluctuate, but always be a convenient match for the battery.

        To make a more accurate calculation, hours of productive use must be considered. Solar and tariff cycles are periodic. To fully-cycle a 7kWhr battery every 24 hours, the battery must output 7kWhr.
        If the battery is not fully discharged and recharged every 24hrs then your calculations are invalid.
        The homeowner would become a slave of the battery. Always ensuring that just the right load is connected ( 2kW max for the PW), while consuming energy, needed or not, for a period of 13.6 years ( 5000 cycles)
        Failure to comply will result in delayed buyer remorse. Take the short cut and accept it now.

        • heinbloed

          Read the article again.

          • Robert Haylar

            Please expand.

        • vensonata

          This article is a beginning, a good beginning. Your interest in the chaos theory of battery life calculations is more for a specialized forum of battery geeks. Calendar life is mystical even for the leading speicalists… see a full hour lecture on youtube by the leading academic guru of lithium batteries, Jeff Dahn. He is the prof who trained the phd engineers who work for Tesla. You will plainly see the problem is there is at present, no technology on the planet able to provide accurate and complete info about lithium batteries in all their glorious formulations over their complete lifespan. Dalhousie University is investing in partnership with 3m in the expensive and complex equipment needed for just such reliable data. Here is a quote from the esteemed professor Jeff Dahn: ” Analysis of lithium battery life performance is the hardest problem I have ever worked on, I expect it will take the rest of my working life to solve it”. Just saying… its complex.

          • Robert Haylar

            Yes. Professor Dahn’s micro-coulumb counting. It is not relevant to the discussion of existing batteries, because whatever may be claimed from it, already occurs. It’s a tool for accelerating cell design.
            I can only see that you are shooting yourself in the foot, by even suggesting that a back-of-the envelope calculation can be compared to “the hardest problem…”

          • Robert Haylar

            Sorry if my reply was a bit sharp, but do you know, Vensonata, you mention “battery geeks” both in support and denial?
            But, there is no need to be so defensive, Yes, cell life is a difficult problem. Don’t you think a large margin of error should be included to account for the unknown?

            But it’s also the overall calculation that is a problem.
            Take a simple example. I charge the battery at night, then during the day, expend it. If I really need that expenditure of energy, I may save some money, That is the basis of the calculation, but extrapolated to meet the battery’s end of life.
            If I were to buy a system, use if for one day, wait a year, then based upon the time that has passed, calculate my return, I am sure that you would object that I had exaggerated the cost.

            Similarly the given calculations include all possible downtime, to arrive at the final number, so reducing the apparent cost/Kwhr.

            If I spend any time away, or not use the system at all, then I can’t use 100% discharge numbers. Not using the battery at all, is the extreme case, just as 100% is. In practice, the average will be somewhere between 0% and 100% utilization, which is another hard to know quantity..

          • vensonata

            I mention battery geeks not as a dismissal, we need battery geeks. Just that, as I said, this article is a beginning for most people. They don’t deal in this area often. Take laptop batteries, or power tool batteries, the average person who has used these things, possibly for ten years, would have difficulty giving even approximate guesses as to the actual cost be unit of energy and even favorable or unfavorable conditions for their use. I suggest that over the next year or two more ” back of envelope” analysis will emerge along with “rules of thumb” and “complete idiots guide to home battery storage”. Patience and simple approximations are most appropriate at this stage of the game.

          • Exactly. My biggest concern with this piece is that it is already too dense to serve its purpose — getting more people to learn about the issues at hand and what factors need to be considered. Hopefully, someone actually moving forward toward a purchase and expecting it to “pay off” financially will consider how often they may not be at home and use the battery, how much electricity they actually use on a daily basis etc… as well as look at cycle life graphs and learn more about the specifics of the batteries they’re considering.

            I hope this was a super useful first step helping people to get to that stage.

          • Robert Haylar

            Analogies are often useful when tackling a topic not familiar to others. My internet bill is totaled each month. I could divide that by the number of hours in a month, to arrive at
            a $/hr figure. On the other hand, if internet usage were a matter of profit, I should count only the hours that benefit that aim. Similarly, with battery usage.

            Usage is difficult to quantify because demand varies so much, and often in a way that is not convenient to turning a profit. Power output is important in that case, and for another reason.

            A battery offering 4kW output, can be used 50% of the time, yet meet the utilization of a 2kW unit, run at 100%. If battery life were dominated by cycle life, then the 4kW unit would last twice as long as the 2kW under the same conditions.

            A 4kW unit can do whatever a 2kW unit can, but not the converse. One should stay clear of batteries offering low power output.

            Claims for battery life should always be taken with a grain or ton of salt. If you, or anyone else, can do that back-of-an-envelope calculation, then some unscrupulous
            manufacturer may reverse the process to arrive at a plausibly marketable cycle life. They may not be around to be accountable. Perhaps a manufacturer with a proven pedigree may be the better choice.

            One means of helping potential purchasers, would be to

            seek as much proof of performance as can be provided. When a new arrival claims a cycle life substantially in excess of similar batteries, one should take note.

          • I love that video. Watched it a few times. 😀

        • All of these points were considered. I could write an article in response here explaining the rationale behind the assumptions and calcs used, but I have to run. If you want to put together some tables using different assumptions, I’d probably be happy to include them as updates in the article with explanations like I provided above. (Assuming there are no glaring mistakes in content or presentation.)

          • Robert Haylar

            All of my points were considered? Your assumptions are clearly based upon full discharge cycles, and there is no sign of the effect power output may have.

            The “total kWh produced…” is simply;
            capacity*claimed cycle life*efficiency.

            Perhaps you can finish what you started rather the vague offer that you may do so, if I do it for you.,

          • eveee

            They are not all based on full discharge cycles. Its based on data that you can access online. Tesla doesn’t say how they do it, but they provide a 10 year warranty and give you 7kwhr for $3k. All the other figures are available publicly although some of the others are sketchier, because they are claims from manufacturers, like $160/kwhr or $500/kwhr not necessarily today, and not advertised prices. Where there are other factors, its mentioned in the article.
            For example, you can look at Iron Edison in the table. In the information below it says based on 80% DoD. The number of kwhr is based on the cycle life degrading to 80%capacity with an average of 90%, efficiency, and 80% DoD.
            Tesla just states 7kwhr, and the calculation shows that because Tesla states it with no other information.
            I look forward to seeing more detailed information on all of them.

        • bink

          Sorry Zach your Tesla Power Box does not include an AC system

          • eveee

            Can you point to a place in the article where the claim is made?

        • Billi

          500 cycles not 5000 cycles , get the math right

        • eveee

          Read the text. It specifically states that the Iron Edison battery performance relates to 80% DoD. That calculation is not applied to the PowerWall. We have no information to the effect that PowerWall numbers are for %80 DoD.

          We have even sketchier information about Imergy, and EOS, just capital cost of capacity estimates projected by them.

          The calculations don’t assume a homeowner is slave to a system. Its just a table for comparison. All systems would have their cost metric change if they were not full utilized.

          The issue of how well and if such systems can be utilized is better left for another article.

          Battery slaves seems like a catchy title.

          “Failure to comply will result in delayed buyer remorse. Take the short cut and accept it now.”

          sounds like those deceptive letters that scare old people into buying unnecessary items that I throw in the garbage.

    • sjc_1

      Yabo Power Technology
      lithium titanate
      Capacity: 15/18Ah
      Rated voltage: 2.4V
      Cycle lifespan: 12000 to 20000 times

Back to Top ↑