Elon Musk Made the Kessel Run in Less Than Twelve Parsecs

I had to laugh at the stories the other day on the battery backup system Elon Musk and Tesla made for the Australian Power grid:

Tesla has completed its 100 megawatt Powerpack battery backup system in South Australia within 100 days (easily), as Elon Musk had promised. That means the company essentially won the "bet," and won't be on the hook for the entire cost of the project, estimated at $50 million. More importantly, it means that some 30,000 homes in South Australia will have a power backup in case there's no breeze at the Hornsdale Wind Farm located about two hours from Adelaide.

A megawatt is a measure of energy production or transmission rate.  As such, it is a perfectly appropriate way to size the capacity of a power plant that is assumed to have a continuous supply of fuel.  However, it is an extremely odd way to size a battery.  A battery has a fixed energy storage capacity, which is generally measured in watt-hours (or some conversion thereof). For example a 10 Wh battery would provide 10 watts for an hour before running out, or 5 watts for 2 hours, etc.  It is not clear if this is just a typo, that they really mean 100MWh, or if 100 megawatts is the peak discharge rate and they are being silent on exactly how long this lasts (ie how long can those 30,000 homes be powered?)  I checked the first 10 sources in a Google search and not a single media outlet that routinely chastises climate skeptics for being anti-science seems to have questioned the oddball and nearly meaningless 100MW figure.

I was going to compare the number on energy storage here and show that you could actually generate electricity from gas, not just store it, for well less than this.  But it is sort of hard to make the calculation when they don't get the units right.

By the way, if this is required to make wind power work, will we start seeing wind advocates building in $50 million batteries when they present their economics?  Any bets?


  1. NeoWayland:

    Okay, I had forgotten about that bit from the original Star Wars. Good point and good joke.

  2. Zachriel:

    Coyote: how long can those 30,000 homes be powered

    An average home uses about 1 kw, so 30 thousand homes use about 30 mw, but you have to account for peak usage. Figure about an hour or so of backup.

  3. texasjimbo:

    Does that account for transmission loss? Also, if the point is to use "renewable" power, doesn't the battery need its own fleet of windmills/solar panels to keep it toped off at all times? What % of the storage capacity will be lost during cold nights?

  4. johnmoore:

    I see this sort of thing so often that I just assume that scientifically illiterate reporters dropped "-hours" from the units.

  5. Unknown Commenter:

    I see this sort of thing so often illiterate reporters.


  6. SamWah:

    That makes no sense. Musk said it, so it's he who's not making sense. Unless news.com.au said it wrong.

  7. Matthew Slyfield:

    Not sure where the 100 megawatt error comes from, I thought it might be Engaget's error, but the sources they link to (including the Australian Broadcasting Company) have the same error.

    By the way, I found this from The Guardian that has the right info https://www.theguardian.com/australia-news/2017/jul/07/tesla-to-build-worlds-biggest-lithium-ion-battery-in-south-australia

    The battery storage system is 129 megawatt hours.

  8. Bob Long:

    I got attacked for raising the same question (30,000 homes but no mention of hours supported) on another site (https://www.computerworld.com.au/article/630402/tesla-very-big-battery-south-australia-ready-testing/). To be fair, I'd missed the mention that capacity is 129 MWh. But I too could not find any mention of the supported duration in many of the articles I looked at.

    I've been able to work it out (about 27,000 homes for 8 hours at, say, 15 kWh/day each). But that assumes no industry is using power. And to put 30,000 homes in perspective, there were 626,5000 households in South Australia in 2006 (http://www.abs.gov.au/ausstats/abs@.nsf/0/20774DAFD086A4D5CA25773B0017CE22?opendocument).

    The calculation also assumes the entire capacity of the battery can be used. I also suspect there will be distribution areas which would not be reached by that battery.

    I reckon press releases saying simply that the battery will "provide power for more than 30,000 homes" are meant to sound impressive until one does the numbers.

    Of course, such batteries would not be necessary at all if we had proper baseload power stations.

  9. Matthew Slyfield:

    I don't think the battery system at issue is meant as backup against prolonged outages. It's located at a wind farm and is likely intended to smooth out output variances by the wind farm.

  10. Jim Swartwood:

    The Hornsdale website stated that the Tesla battery backup capacity was 129 MWh. The average global household daily electrical consumption is about 10 KWh. Since Hornsdale serves 30,000 homes requiring a total of 300 MWh per day, the battery backup can power them for about 5 hours, assuming no transmission losses. How often will the wind be calm for that long? Often I think.

  11. Q46:

    ‘It is not clear if this is just a typo...’

    No it is deliberate with intent to mislead the public, press and stupid politicians (tautology, I know) who do not realise the difference between energy, kW, and consumption, kWh.

    It is a common ‘typo’ used in relation to windmills in Europe, when politicians and eco-psychos boast about a wind farm having capacity of X GW, ‘enough to power a town/city the size of......... (insert name and number of homes)’.

    The fail to mention that the GWh output in aggregate will be at best 25% of X and intermittent, so the people in that town/city will spend most of the time sitting in the dark burning furniture to keep warm.

    It is why the fiction persists that wind farms can replace fossil fuel generation because they tot up the total capacity of each but ignore the fact that wind only gives at most 25% of rated capacity whereas fossil fuel gives at least 90% or more.

  12. Zachriel:

    The 100 MW would be the maximum capacity. The 129 MWh would be the total energy available.

  13. Zachriel:

    texasjimbo: Also, if the point is to use "renewable" power, doesn't the battery
    need its own fleet of windmills/solar panels to keep it toped off at all

    There are times when the wind blows that electricity needs are low, such as at night. This allows for storing energy in the batteries.

    "The battery storage units were designed to reduce power shortages, minimize intermittencies, and manage summertime peak load to improve the reliability of South Australia’s electrical infrastructure."

  14. Zachriel:

    Q46: No it is deliberate with intent to mislead the public, press and stupid politicians (tautology, I know) who do not realise the difference between energy, kW, and consumption, kWh.

    It's a 100 MW/129 MWh Powerpack system.

  15. Fred_Z:

    So, an hour and a quarter at maximum draw?

  16. auralay:

    Uk blogger Paul Homewood is the go-to site for fisking this sort of claim. He has several posts on this battery, the latest here. https://notalotofpeopleknowthat.wordpress.com/2017/11/25/elon-musk-cons-50m-out-of-south-australians/
    It is 129MWh; would run the state of South Australia for a full 4 minutes!

  17. DaveK:

    Oh, goody! A battery that can store over 100MWH! And it can discharge at the rate of 100MW! What could possibly go wrong?

  18. Matthew Slyfield:

    Exploding, burning, death, dismemberment, weeping, not necessarily in that order.

  19. John Smith:

    kW is power.
    kWh is energy.

  20. John Smith:

    A battery has a energy storage capacity AND a maximum power output. The capacity is measured in kWh and the max power is measured in kW. It is best if both specifications, kW and KwH, for the storage device are given, but it is not "incorrect" to only give one.

    The engadget article links to Reuters which says "129 megawatt hour".

  21. Bill Drissel:

    In all the talk about batteries, I never hear the recovery fraction. If a battery draws 1 KWh from a generator, how much of that can it deliver to a load. For lead-acid batteries, that fraction is about 0.5. Are modern batteries any better. Does the recovery fraction change with age? How much? How fast?
    Bill Drissel
    Frisco, TX

  22. Olivier:

    Han Solo was actually boasting of this shorter route (in distance), meaning that The Falcon Millenium with its superior engines was able to pass closer to the black hole without danger.

  23. Bruce Anderson:

    Might be better to impress the masses by using joules. One joule is a watt-second. So one kWh is 3,600,000 joules, 3.6 MJ. Megajoules, I tell you, megajoules. Plus, joules sound more mysterious and intriguing than watt-hours.

  24. J Crain:

    This seems appropriate to the discussion: http://www.smbc-comics.com/comic/capacity

  25. Granja:

    Excellent graphic for energy consumption and generation:

  26. marque2:

    That bugs me with my kids Chemistry text book. We used to use units that were maybe semi metric, but made sense in real world. 1 calorie is what it takes to heat 1 gram of water 1 degree, and we need about 2 million of these to survive in a day - and it is a common measurement you all use every day. We measured air pressure in atmospheres - about one earth atmosphere at sea level. Now they use jewels for energy, and have no concept, may as well be magic numbers. Even worse for ideal gas problems they use 101.3 kilo-pascals. Wow that has some real relevance, and worse, the US standards bureau didn't like that non metricy looking pascal, so for STP they redefined the atmosphere to be smaller by 1% so you could use an even 100 kPa. And they wonder when they make these measurements more esoteric and more removed from real life, why kids have problems with math and chemistry even with the new common core to save them.

  27. Andrew_M_Garland:

    If 129mWH is correct, then 30,000 homes using say 1kW would have a backup for 4.3 hours, or 8.6 hours at a subsistance 500W each. Awfully thin.

  28. Monsyne Dragon:

    That's the "charge efficiency", and yes, for lithium batteries it's around 80-90%, as opposed to 50-60% for lead acid. The efficiency doesn't change much as the battery ages, but the capacity of the battery slowly decreases over time. (By how much depends on a number of factors, including temperature (batteries degrade faster when hot), and the exact chemistry. (there are really four different types of batteries called "lithium" batteries, some last much longer than others.)

  29. Don:

    "for lithium batteries it's around 80-90%"

    On a good day, with the wind to its back. Even with balanced charging and controlled discharge, lithium batteries develop internal resistance pretty rapidly, dropping that efficiency. I use small ones in my business (2s lipos). A new 1000 mWh lipo battery will last 12 hours in my equipment without any problems when new, but 8-12 months later (~100 charge cycles), it struggles to last 8 hours.

  30. Dan Wendlick:

    Actually, no. A parsec is a parallax second, which is the angular displacement of the position of a body at two measurements six months apart. As such, it is utterly meaningless as either a unit of time or distance if divorced from the size of the Earth's orbit. The distance represented by a parsec would vary on every planet from which the measurement was being made, so it would be a terrible standard of measure for a galactic civilization (as would the light-year, for similar reasons).

    It is more likely that he was referring to a unit of time that just happened to be a homophone or even homograph of the Earth astronomer's term "parsec".

  31. SamWah:

    Chinese Electric Ship--To be used fro shipping coal on a river:


  32. Monsyne Dragon:

    Yah, that's typical for those cells in a heavy usage application. Those lipos are almost certainly Lithium Cobalt chemistry. Typical lifespan for those is 500 cycles, under *ideal* conditions. Add in high current draws, warm environment, and/or high depth-of-discharge, and that will drop fast. (the polymer casings on the lipo cells don't help either. Poor heat dissipation, and they don't keep pressure on the cell like a metal case does when the cell components expand/contract during a charge cycle. )

    It's not the efficiency that's dropping, it's the capacity. After 100 charge cycles, your 1000mWh battery is now a 600-700mWh battery. You get less out because the charger can only put less in. I see this with the Lithium Manganese cells I use in my e-cig, when the cells are worn (after ~200-300 charge cycles) they charge quickly, because there's less energy you can put into them, but discharge quickly for the same reason.

    (IIRC, Tesla uses an variant of the Lithium Manganese chemistry, with active cooling, and a lot of fancy active charge control to keep the batteries in the middle of their charge capacity, this gives them ~1500-2000 charge cycles, but they still expect their EV packs to loose ~30% of their capacity by the end of their service life (~8 years) )

    If you really want lifespan out of a lithium battery, you use Lithium Iron Phosphate cells. They have 10 times the lifespan of Lithium Cobalt, but you take a 15-20% hit on energy density (power storage per unit weight of battery). I know folks with thousands of cycles on Lithium Iron Phosphate RV/Marine house battery packs. Even then they are temperature sensitive, and you can loose half your batteries operating lifespan having them in an non-climate controlled exterior storage bay of an RV that gets to 110F, as opposed to a climate controlled interior storage compartment.

    Ultimately, I think the lesson here is "batteries suck". Lithium chemistries are best-of-breed here, but that just means they suck slightly less than other batteries.

  33. The_Big_W:

    Heh, I'd heard the previous retcon before, but I kinda like yours better. :-)

  34. Don:

    Yeah, been considering the move to LiFe, but the cost of buying all the chargers (I run 11 at once) has made me hesitate for obvious reasons. My last box-full of batteries cost me less than $5 each ($105 for 22 batteries, delivered), so at that price, even though it's a pain, I will probably keep sucking it up for a while.

  35. DaveK:

    Ummm, no.... It's defined in terms of the "Astronomical Unit", and so is tied to the size of the Earth's orbit.

    A parsec (symbol: pc) is a unit of length used to measure large
    distances to objects outside the Solar System. One parsec is the
    distance at which one astronomical unit subtends an angle of one
    arcsecond. A parsec is equal to about 3.26 light-years (31 trillion
    kilometres or 19 trillion miles) in length. The nearest star, Proxima
    Centauri, is about 1.3 parsecs from the Sun.
    [from Wikkipedia, but you can find a slightly more detailed definition at Britannica]

  36. Leonard Jones:

    Elon Musk is famous for stretching the truth to the breaking point. Even with the massive federal and state
    subsidies, Tesla has not shown a profit since it was founded. He is a later day Elmer Gantry who made
    outrageous claims concerning the range and recharge times of his vehicles. He once staged a fake
    video in which a "Modular" battery pack was changed out in just a few minutes. The reality is that
    his cars donot have modular battery packs and it takes two mechanics 12 hours to chage a Tesla's battery.
    The reason for this sham was that by proving quick swap capabilities, he could claim the highest
    level of government subsidies.

    Some electric car guru just called BS on the claims he made regarding range and charge time on his
    recently announced electric semi. I recently read a story about Tesla employees paychecks bouncing,
    pending layoffs and utter chaos on the assembly floor. He is also having to return tens of thousands
    of $1,000 downpayments made by people who complained about the production delays.

    No matter what the stated capacity was, battery power density is at the ragged edge, and all of the
    electromotive elements have been known for 100 years. That means there will never be more
    than marginal increases in power density, so that makes any quantum leap near impossible. In
    America, the average monthly residential electrical load is 897KWh. That divided by 30 days
    equates to about 30KWh per day. X 30,000 and it is 870 thousand KWh per day. My guess is that
    whatever the stated capacity, it will likely be good only for short term outages lasting for a few
    minutes to less than one hour. And even then the size of the battery bank would be so GINORMOUS
    that dozens of Millennium Falcons could not get off the ground with even half the batteries

  37. Leonard Jones:

    At 14, I was an astronomy geek. Since calculators were not yet cheap or readily available,
    I took a pencil and a piece of posterboard and converted the distance in light years to
    Alpha Centuri in miles. Nothing gives someone a sense of distance or scale like this
    little exercise. There is a reason they call big numbers astronomical.

  38. Dan Wendlick:

    Actually, it's exactly what I said. Taking the two readings six months apart gives a baseline of two AU for the parallax measurement. The fact that it is considered to be a constant unit of distance is that for the extremely small angles involved, the sine function can be considered to be linear and equal to the angle of displacement expressed in radians.

  39. Leonard Jones:

    The problem with this thread is that very few people understand the difference between a Watt, a
    KW, a MW and a GW vs a Watt hour, KWh, MWh and a GWh. A Watt is a measure of instantaneous
    output or consumption. A Watt hour is a measurement of energy most often reserved to describe
    energy consumption over a period of time (1 hour.) This is the unit of electrical energy used to
    calculate your monthly electic bill and really the only reliable method of mesuring battery storage

    Before I retired, I worked at a large paper mill in Los Angeles. Paper mills need a crap-load of
    steam to run lineshaft turbines and dry the paper at the end of the process. Along came the
    concept of CoGeneration. We had a GE 747 jet engine spinning a 32MW 12.5KW three phase
    generator. We were making a profit by selling energy to So Cal Edison and the exhaust gases
    were producing the steam was icing on the cake!

    That 32MW was an instantaneous measurement. Now consider that there are 60 seconds in a
    minute and 60 seconds in an hour. That equates to 32,000,000 Watts x60 x60. That is the
    difference between a Watt and a Watt hour!

  40. Physics Bill:

    The demand in South Australia averages 1500 MW, They have bought 4-5 minutes of backup (less at peak load),

    South Australia had a one week power outage last year. They retired their base load power plants (fossil fueled) and invested heavily in wind and solar. The intention was to buy base load power from Victoria via two transmission lines. One line was down for maintenance and the other blew down.The politicians blamed the one week outage on the storm ignoring the fact they put their grid in a single point of failure mode. South Australia has the most expensive power in Australia and close to third world reliability!

    The greens where I live in the Northern Territory are urging more investment in wind and solar, Here in Alice Springs, we have very few cloudy days and are roughly at the Tropic of Capricorn. Even so a solar installation averages about one fifth of its base plate capacity.

    Disclaimer - I am involved with power generation at an off grid site in the outback, We do not use solar or wind.

  41. Wholy One (Seli):

    Elon Musk is a sleaze-bag FRAUD.

  42. John Smith:

    Yes, it is awfully thin. Batteries have never been a serious power source for a good reason.