Great to see someone actually doing some calculations instead of just presenting a cosy vision. For example, they argue that - contrary to what Elon Musk implies with his illustrations - the solar cells on the roof of the gigafactory will only be able to generate a small fraction of the energy required to run it.
> The average solar insolation in Arizona is 1,964 kWh/m2/yr (in Phoenix). If we assume a solar PV efficiency of 15%, one m2 of solar panels would generate at most 295 kWh per year. Consequently, almost 68 square kilometers of solar panels (6,800 ha) would be required to power the factory -- a calculation that also relies on the assumption that solar energy is equally distributed throughout the days and the seasons (or stored in batteries), and that there's no open space between the panels. Remarkably, Tesla shows an illustration of the factory with solar panels on the roof. Knowing that the factory will occupy a surface of 1 ha, while 6,800 ha of solar panels is required to run it on renewable energy, Tesla's claim is an obvious example of greenwashing -- and everyone seems to buy it.
These are the kind of calculations that make me weep - somebody getting happy with numbers without engaging their brain as to the answer. Obviously the Tesla factory is not going to require 20,000,000 Megawatthours to operate - that is as much as the whole electricity production of the State of New Hampshire.
Assuming their input data is actually correct, I will speculate they are calculating the energy cost of the entire lithium battery production from raw materials, such as extracting lithium from mineral deposits, the majority of which would happen before the materials arrived at the Tesla factory.
I also expect Elon Musk knows what his electricity bill is. Given his investment in green technologies and his general success in delivering projects, I would be rather surprised if he advertised a solar factory if there was a chance he would be out by 4 orders of magnitude on how much power he would need.
>Assuming their input data is actually correct, I will speculate they are calculating the energy cost of the entire lithium battery production from raw materials, such as extracting lithium from mineral deposits, the majority of which would happen before the materials arrived at the Tesla factory.
If you look at one of their sources[1], it is clear that this is the case.
According to this study[0], it takes about 116 kWh of input energy to create a 100 kWh battery, given the materials. The majority of the total energy usage is in material production, not battery production. Assuming Tesla is not making their own materials, this deflates the required input energy substantially.
> Assuming Tesla is not making their own materials, this deflates the required input energy substantially.
And also inflates that amount of greenwashing involved...... Why crow over renewable energy when your use of energy is irrelevant relative to your suppliers?
But their statement seems accurate and clear: they will operate the GigaFactory using renewable energy. They don't claim or imply that they will extract the materials using renewable's or that all their suppliers will use renewable's.
They should source their lithium from solar-powered contractors. Maybe even sell them some panels for a good price.
(I realize the brine is already evaporated in the sun, but there's electrical separation after that.)
Reportedly, the gigafactory will import ore from the mines directly, and process it. So, optimization might be possible, especially given that they can choose more expensive extraction processes because they don't have to compete on the cent.
They don't have any magical technology that allows them to dramatically reduce energy needed. The obvious conclusion is that they won't make anywhere near as many batteries as assumed.
Transportation uses very very little energy compared to everything else that needs to be done.
You will not find your magical technology in shipping.
Additionally ocean shipping is cheap enough that it can take less energy to ship something to the other side of the world than to the other side of Nevada.
I had a dig around, you are right that transport of lithium is relatively negligible in the energy budget (back of an envelope calculations suggest around 0.5 kwh for a 1 kwh capacity battery), transport of other material such as nickel is actually more important, but even when you add them all together it is only maybe a couple of percent or so of the energy involved.
It seems that the drying process is the one to target as it accounts for around 50% of the energy budget and uses big electric ovens. Electric ovens are something that respond very well to economies of scale though, given the losses are a function of the volume versus the surface area, so you could see significant reductions there with a factory this size.
I think this is a very uncharitable interpretation of that 'illustration'. The illustration makes no explicit claim that the energy is produced by the solar panels on the roof, let alone solar power. And I think it goes without saying that heavy industry cannot be supported by a small cluster of panels. There are even wind turbines in the picture, which further discredits the idea that Musk is trying to imply the solar panels can do it alone.
The claim of "1 ha" stumped me, and I looked it up - the numbers you can find on the interwebs are conflicting, but it should take up about 200-500 ha. The assumption of 15% efficiency also seems quite conservative to me - I can imagine that they will use new cells fresh from the lab that are more in the 18-20% range. Another thing that Elon mentioned is that they might be able to drive energy consumption down because of synergy effects.
But yeah, they will probably not produce more than 20% of their needs with these solar panels.
Phoenix, Arizona is actually not the most sun-soaked part of the country. The Mojave desert gets a much greater amount of solar insolation[0]. Using those numbers, we can calculate ~511 kWh/m2/year, assuming 15% efficiency as well. That lessens the gap somewhat, but there's still quite a distance - perhaps wind and hydroelectric would account for the remainder?
It may just be flat-out wrong, which is unfortunate.
Except his calculations seem all wrong. From the article picture above there's no way that factory only has ~100x100 sqm (1ha - a bit bigger than a soccer field). Maybe 100ha. And that seems to be the case:
Money quote: Using the worst assumption in each case, solar cells with storage emit 631 gCO2e/kWh vs 337 gCO2/kWh for normal hydrocarbon fuel in spain.
Important takeaway: A solar cell embodies a huge amount of energy, yet can be shipped. Make solar cells in places with lots of clean energy and ship them elsewhere.
True, but that's only half the story. The other half is that high-load and low solar-insolation means batteries degrade faster / require larger batteries and requires more panels, respectively. Even batteries built with clean energy isn't necessarily always the best application versus taking energy from the grid (especially a nuclear powered grid).
Shipping is negligible and when compared by energy per tonne per km, can be up to almost 100x better than road, so going 200km by road can compare to going nearly 20000km by sea in some cases.
Putting everything in one place could get savings from many factors, but transport doesn't appear to be that major compared to the others.
The conclusion of this article is that a life cycle analysis of solar with storage is better than the marginal contribution of greenhouse gases from operating a traditional plant but not by a lot. Has anyone seen a life cycle analysis of decommissioning a coal or nuclear plant or managing tailing ponds? I commend the effort to make apples to apples comparisons but that would need to be added in to make this genuinely apples to apples.
"the manufacture of 1 kWh of lithium-ion battery storage requires 400 kWh of energy, the factory would require 20,000 GWh of electricity per year to manufacture all these batteries."
The 20,000 GWh figure is clearly bollocks, one factory a hectare in size is not going to be using the same amount of electrical power as the whole of Nigeria.
edit - also, one thing I do not understand with their argument, is if Tesla's illustration is really supposed to show that the factory can run entirely on rooftop solar, why would it feature around 100 large wind turbines dotted all across the hillside?
Here's an offtopic thing which annoys me and many people do wrong. It is important to do significant figures right.
2.28 GW is the most you should ever write with that calculation. What you wrote went down to hundredths of a watt (if I can count).
Nobody should care about the third significant figure in such a comparison, much less the twelfth. The second digit is even of questionable accuracy. Strict significant figure rule following isn't important in casual contexts, but deciding which digits are actually relevant is imporant. Is accounting for leap days really important here?
Also, the output is roughly the installed maximum capacity of the Hoover dam, but about 5 times larger than it's actual yearly production.
So currently for batteries used by Tesla, that is getting the lithium from Chile and Argentina, then transporting it to battery factories in China, then shipping them to their car factory in California, as opposed to getting lithium from Nevada, to a factory in Nevada, to the other side of the same Nevada factory, as you also make the cars there.
The total embodied energy is going to get absolutely slashed.
edit - btw, I accidentally fat fingered the down button on your comment that had the link to - http://www.withouthotair.com/ - sorry. I hope that isn't what prompted you to remove the comment.
No that wasn't the reason, though its nice of you to mention it.
I'd re-read the withouthotair site the other evening after seeing Elon's presentation, as he made a similar case regarding the area of the USA needed to be powered by solar. I was a bit bummed out by the conclusion for the UK and wondered what the global equivalent was, as the UK is low on solar. I then noticed I'd skipped a chapter that does the rough calculations for other regions (http://www.withouthotair.com/c30/page_231.shtml) the basic answer being that Solar may just save us, but it'll be an incredible amount of work, both political and engineering.
(Which is basically the same answer as for the UK, though in the UK solar is mostly replaced with wind/wave/hydro as we're relatively blessed with those).
I think he may be underestimating the future efficiency gains.
There is already 30% efficient directly illuminated multi-junction PV on the market ( http://www.emcore.com/wp-content/uploads/ZTJ-Cell.pdf ), they are just so damn expensive at the moment that only the space industry are using them.
Also, while it is a huge amount of work, it is no worse than many other industries, the economics are starting to make it a good investment for big finance, and we are not likely to run short of either silicon or lithium along the way.
The Without Hot Air site seems to have got its sums wrong. It continually makes the point that we'd need to cover the whole UK in renewable energy generation, and that still wouldn't be enough.
However, we already generate 7.5% of our energy from renewables and clearly 7.5% of the country isn't covered in windfarms/solar/biofuels. Therefore he must have made a mistake somewhere: if reality disagrees with theory, the theory must be wrong.
7.5% of our energy or 7.5% of our electricity? It's the latter. We're still consuming the output of the North Sea plus a little bit more in petrol and diesel.
> In order to fill these gaps [ night, cloud ] , a storage solution
> or a backup infrastructure of fossil fuel power plants is required
> -- a factor that is often ignored when scientists investigate the
> sustainability of PV systems.
If you're approaching the problem with solar power with the assumption you need storage or fossil fuel, you're basically arguing with yourself.
Note they did say 'solar power', not 'solar cells / PVC / etc'. Solar thermal plants ride through periods of cloud, and well into the night, reasonably well. But the big problem is that assumption that you must either store solar, or you burn coal / oil ... followed by a complaint that scientists ignore important factors.
I'm not sure I follow your point? Even without solar power, storage is an important part of many grids e.g. meeting peaks with pumped storage hydro.
Are you saying that solar thermal is enough storage for a predominantly solar grid? I'd suggest it isn't. I'd also suggest that solar thermal is a type of storage solution and it's time-shifting properties are one of the things that make it useful.
You're right that solar thermal is technically a type of storage system for solar. But that wasn't countenanced in the article -- wherein we're told the (availability) problem of solar can only be solved by either fossil fuel power plants or storage systems.
My point is that this is either a flawed or a disingenuous premise.
There's myriad ways we could deal with gaps in sunlight availability: nuclear fission (not a fan, but undeniably one way), biofuels (also not a fan, but ditto), reduced demand / opportunistic usage patterns (to ameliorate the effect - not easy, but disturbs me that questions about energy usage never include 'reduce' or 'adapt' in the list of considered answers), and, most importantly, a stack of renewable options that are not constrained by the vagaries of sunlight -- geothermal, wind (terrestrial and tethered), wave, tidal, and hydroelectric spring to mind.
You don't need pumped hydro. The wonderful thing about hydro is it throttles easily, and leaving water in the lake is equivalent to putting more water into the lake.
When the reservoir is full, Cruachan Power Station in Scotland can operate for 22 hours before it is emptied. Also, it is required to leave 12 hours water in the lake at all times to provide emergency cold start backup on the grid. If it wasn't pumped hydro and was only working from runoff, then it wouldn't work.
I don't agree with this argument. The sun doesn't shine at night, period. Any solar system which supplies energy after the sun is down relies on storing residual energy collected during the daytime. Solar thermal is storing energy in heat just as a lithium battery would using chemical reaction.
Are you going to fully discharge that 7560kWh lead-acid system every day?
No, you are not. You are actually going to buy five times as many batteries and work in the 20% of the storage range which is suitable for continual cycling. So that's 90 batteries off the bat, or about $6750, almost double the cost of the Powerwall, not including the ancillary equipment you haven't budgeted for to bring your batteries into the same league as the Powerwall.
You still need the battery room to install them in, and the fire protection system to put out electrical or chemical fires when things go wrong.
The Powerwall includes thermal and fire containment (the batteries are swimming in gel), the Powerwall also takes care of the "battery room" issue by storing the battery in a container on the wall.
For those with greater power requirements, there are rack mounted power cells, again with thermal and combustion mitigation already in place.
The integration and packaging is why everyone is excited about these batteries at this price.
Lead acid batteries don't last as long, and suffer much more from large depth of charges, meaning you need to replace them 2-3x more often. This doesn't just have a cost to it financially, but also environmentally.
It also means that if you want your batteries to last, you can reduce your depth of charge by purchasing excessive capacity. This way instead of discharging 100% of X capacity, you can discharge 50% of 2X capacity, and this lower discharge depth makes the battery last longer. Which means you tend to have to buy a lot more capacity than li-ion.
Beyond that, look at it like any system. e.g. do you want 4 small fridges that give you x liters of volume, or one large fridge that gives you 4x liters of volume for a slightly higher price? Do you want 10 harddrives that give you 100gb each for $250, or would you buy a $300 1 TB drive?
There are non-financial reasons to go for simplicity, something that works out the box, plug and play. Especially for consumer grade equipment. Beyond that, Tesla's solutions offers some integrated solutions e.g. concerning fire hazards that you'd need to do yourself otherwise. And lastly, it appears batteries benefit from economies of scale. Their battery is already for sale today, and considering the factory they're building, it's quite exciting to see how cheap batteries will get by 2025, but that's a different story.
Your AGM batteries are Lead Acid. Tesla Powerwall is a LiOn Battery System, designed for 10 years. Implies somewhat more than 10 kWh "actual" in order to deliver 10 kWh, day-in, day-out for 10 years, and avoid deep discharge. The value is greater efficiency, great life span.
The price for the 10 kWh model is basically 50% of what everyone was expecting, which is what has everyone excited. You don't often see a 50% drop in price in systems that are this closely followed.
I keep imagining a world where Solar is our primary source of electricity and I keep seeing that world lose its mind when a volcanic ejection leaves us powerless.
I get that it should be part of a much larger plan, but Solar in its current iteration doesn't seem like that safe of a bet.
Nuclear is the safe bet. However, in your block out the sun scenario, we will have lots and lots of problems beside just power generation. Such as crops.
I keep imagining a world where people eat food from agriculture, and I keep seeing that the world lose its mind when a volcanic ejection / tsunami / fire / mudslide / drought leaves us foodless.
Come on man. There's more to it. There are real challenges and a volcanic ejection certainly is one of them, it's just not anywhere up the list.
This new push towards off-grid, battery-powered solar power just doesn't sound right. Batteries might have improved, but they are high maintenance and polluting. Moreover this requires the addition of a lot of new electrical circuitry at the point of installation - a waste of space, time and money.
It's the responsibility of the world's governments to work out ways to keep the grid-connected solar solutions working. Maybe change the 'Net metering' rules to buy the solar power from customers at a lower price, and let the supply-demand work itself out.
It'll be very stupid if we let go a well-functioning grid based solar power solution just because some power distribution companies don't find it profitable.
Maybe change the 'Net metering' rules to buy the solar power from customers at a lower price, and let the supply-demand work itself out.
It won't be that simple, since even this (allowing customers to contribute power back to the grid at scale) requires expensive infrastructure upgrades.
I'd argue that one implication of a smarter grid would be that there'd have to batteries all over the place, exactly because batteries can adapt quickly and 'smartly' (with a .05$ microchip) to energy supply and demand.
Whereas there is a limit to the maximum intelligence of refrigerators, airconditioning devices, heat pumps, and washing machines.
> Batteries might have improved, but they are high maintenance and polluting.
My brother came home from elementary school one day telling us how modern electricity production causes pollution (sure, that's true) and if we'd go back to powering our society by burning wood, the way people did it in ancient times, we wouldn't have the pollution.
Burning wood, of course, causes plenty of pollution; a major way to see the decline in civilization after the fall of Rome is to notice how the level of particulate pollution from fires cratered. If we were to produce the same amount of power we produce now, but from wood instead of coal and oil, our pollution situation would be much worse. Also, we'd quickly run out of wood.
I think people today have imprinted on the message "coal and oil production of electricity is bad". Making your own batteries from a non-fossil-fuel source isn't a way of powering your home more efficiently, or of reducing pollution, it's a way of refusing to partake in coal consumption -- the hijab of the environmentalist movement.
Net positive below a certain scale (especially for the whales), negative above it. The contentious bit is where the tipping point is. For coal, the first time it was restricted in London due to pollution killing the populace, was 1306.
It's a bit silly to argue over where materials are made when considering sustainability, as that requires an assumption that the "dirty" productive capacity would otherwise not be used for anything else.
I may be being overly cynical, but given the blog is dedicated to showing how high technology solutions cannot work and has articles about PV generation leading with the claim that solar can be often worse than coal, before you drill down and realise that they are not talking about generation, but have segued into running the figures on what happens when they are glued to the casings of laptops and mobile phones ( http://www.lowtechmagazine.com/2008/03/the-ugly-side-o.html ), I think they are trying to find numbers to back an established view, rather than adjusting the viewpoint to the numbers.
---
edit - Actually, to be honest, I do not think in this case I am being overly cynical. The article I linked to, titled "The ugly side of solar panels" starts with:
"New research shows, albeit unintentional, that generating electricity with solar panels can also be a very bad idea. In some cases, producing electricity by solar panels releases more greenhouse gases than producing electricity by gas or even coal."
This claim is then not really addressed till right at the end of the article, where we find the following:
"For rooftop and ground-base installations, the eco-friendliness can be good or doubtful, depending on the solar insolation and the life expectancy. But if we consider solar panels mounted on gadgets like laptops or mobile phones, solar energy becomes a plainly bad idea.
If we take a life expectancy of 3 years (already quite optimistic for most gadgets) and a solar insolation of 900 kWh/m² (quite optimistic too, since these things are not lying on a roof), the result is 1,038 gram CO2 per kWh in the worst case scenario (high-efficient mono-crystalline cells produced in the US). That means that it is better for the environment to power a gadget with electricity generated by coal, rather than by a solar panel."
So, if you buy a laptop or mobile with a solar panel on it and then throw it away after three years, then it is worse than coal.
Which is obviously a completely rational point to make on the subject, given all those masses of mobile phones and laptops sold everywhere with solar panels on them.
Then it gets even better:
"All this does not mean that PV solar energy should not be promoted. For one thing, it’s much better using silicon wafers to make energy generating equipment instead of energy guzzling equipment (like computers, mobile phones and car electronics)."
And by this point I am not sure if this is satire, or if the author thinks their website is hosted on a steam powered abacus.
Yes, we are clearly better off using the traditional coal-fired laptops, even though the extra bulk of the thermal shielding and fire extinguisher can sometimes be an inconvenience.
Actually, now that I think of it, I can't remember ever seeing a laptop with its own coal fired power plant. I wonder why this isn't done?
He's completely ignoring the effects on the grid, which will be huge. Batteries are also a complement to wind, for example. Here's one piece discussing that dynamic; there are many others at e.g. Greentechmedia.com.
The article mentions getting 7000 charge/discharge cycles from lithium ion batteries. That doesn't sound right to me. Most consumer electronics are lucky to get 300 or so at 100% discharge.
n.b. most laptop / phone / tablet batteries have fairly sophisticated controllers built into them and the % charge you see on the device may not be the exact % charge in the cells. So running your laptop to 0% may still not be the full discharge (presumably to prevent this exact problem).
That said, I did own a Sony laptop that had a battery saver option where it would limit the charge to 80% in order to prolong battery life. I can't remember if this actually meant "displayed charge = real charge - 20" or was capping the charge to 80%
I had to abandon my reading before I reached the end, I was too angry at the disgustingly sloppy referencing. In particular his second paragraph put me in a bad mood straight away:
> In order to fill these gaps, a storage solution or a backup infrastructure of fossil fuel power plants is required -- a factor that is often ignored when scientists investigate the sustainability of PV systems.
Um. Pardon? Either he's having an argument with himself, or he's reading some pretty lame journals. Or is he just confused? Let's see:
> Obviously, this strategy requires a backup of fossil fuel or nuclear power plants that step in when the supply of solar energy is low or nonexistent. To make a fair comparison with conventional grid electricity, including electricity generated by biomass, this "hidden" part of the solar PV system should also be taken into account. However, every single life cycle analyse of a solar PV ignores it. [3, 2].
Yeah, okay. That's a perfectly debatable paragraph. So let's see the science behind the claim: [2] has nothing to do with anything! It addresses energy storage, but is not about energy storage; and it does NOT claim to be a literature review, nor does it claim to reference "every single life cycle analysis". Am I missing something? Is this citation of "[3, 2]" supposed to represent the entire sum of human scientific knowledge on this matter? Am I an idiot? Surely I'm the idiot, I didn't waste my time reading this article - that point must be shooting straight over my head!
Okay, let's read the paper. It's on energy payback/cannibalism - a logical presentation from what I skimmed - but is absolutely focused on the energy life cycle analysis of a whole technology experiencing rapid growth - Eg. nuclear back in the day - whereby the energy invested to rapidly establish new technologies may be greater than or at least massively offset any efficiencies (or indeed, "zero emissions") they may have over existing (think "sunk cost") incumbent energy production.
Which has nothing to do with central thrust of this article which seems to be that we should point and laugh at all those idiot scientists who forgot that the sun disappears each night!
But this writing forgets one thing - who is saying that PV (or wind for that matter) can replace established baseload power generators in a 1:1 swap?
Nobody sane, that's for sure. So I'll give you a hint: it's electricity buyers. They don't give a damn that they're destroying the planet by using PV solar, the fundamental fact is that in countries like Australia, even though you seemingly can't swing a cat without tripping over high-grade thermal coal ideal for cheap power, we pay among the highest electricity rates in the world. It's that kind of corrupted and/or government-regulated inefficiency that is the only thing to blame for the fact that PV solar can compete at all, even when subsidies are withdrawn.
> The average solar insolation in Arizona is 1,964 kWh/m2/yr (in Phoenix). If we assume a solar PV efficiency of 15%, one m2 of solar panels would generate at most 295 kWh per year. Consequently, almost 68 square kilometers of solar panels (6,800 ha) would be required to power the factory -- a calculation that also relies on the assumption that solar energy is equally distributed throughout the days and the seasons (or stored in batteries), and that there's no open space between the panels. Remarkably, Tesla shows an illustration of the factory with solar panels on the roof. Knowing that the factory will occupy a surface of 1 ha, while 6,800 ha of solar panels is required to run it on renewable energy, Tesla's claim is an obvious example of greenwashing -- and everyone seems to buy it.