A model of power consumption by autonomous vehicles predicts that once widely adopted, they will consume as much power worldwide as data centers do.
By Bill Schweber 02.27.2023
My summary:
> “1 billion autonomous vehicles, each driving for one hour per day with a computer consuming 840 W, would consume enough energy to generate about the same amount of emissions as data centers currently do.”
> Well… maybe yes, maybe not. Despite the academic veneer and a detailed table explaining their model and its embedded assumptions, the inherent nature of the model has many rough estimates, sophisticated guesses and outright assumptions. I’d compare it more to a fancy “back of the envelope” assessment (Figure 2) rather than a rigorous academic analysis. Step back, and there’s the reality that any assessment of the future for a situation such as this is simply not possible.
> Call me cynical, but it seems that the design reality is that system demands and thus power use often expand to the maximum the system can provide or tolerate. Remember, when transistors and ICs were devised and then went to ever-lower voltage/power, the per-function dissipation dropped by several orders of magnitude—you can’t argue with that fact.
> However, coincidentally with this good news, user requirements increased at a greater rate, so now you have server racks pushing almost 10 kW each—an almost unthinkable figure just a decade ago, up from about 2.4 kW each in 2011 according to one survey (I have seen others, and those numbers are similar) (Figure 3).
>> Call me cynical, but it seems that the design reality is that system demands and thus power use often expand to the maximum the system can provide or tolerate. Remember, when transistors and ICs were devised and then went to ever-lower voltage/power, the per-function dissipation dropped by several orders of magnitude—you can’t argue with that fact.
>> However, coincidentally with this good news, user requirements increased at a greater rate, so now you have server racks pushing almost 10 kW each—an almost unthinkable figure just a decade ago, up from about 2.4 kW each in 2011 according to one survey (I have seen others, and those numbers are similar) (Figure 3).
So the calculation is based on 0.840 kWh per day per car for an hour driving.
When driving 100 km/h for an hour, a reasonably efficient car will use 5 liters of petrol. The energy in 5 liters is about 5 * 34 = 170 MJ. Hence, 47 kWh.
So yes, it’s a lot of energy used by the computations, but still less than the actual force required to move the car forward.
So roughly 5 cents of energy to autonomously pilot a 2 ton vehicle with several human occupants through a relatively hostile and uncontrolled environment…
At scale safe autonomous driving will be saving 100k lives and eliminating over a trillion of dollars in economic losses per year.
This is a minuscule amount of energy expended for the results achieved by any reasonable standard!
I agree with your opinion, but disagree with measuring power use in terms of monetary cost. Doing it that way excludes most of the actual cost of power generation and use.
> A model of power consumption by autonomous vehicles predicts that once widely adopted, they will consume as much power worldwide as data centers do.
It's referring to the energy to power the brains of self-driving, not to move the vehicle.
But that seems... fine? I can't tell if the article wants us to think that's a lot.
Data centers and networks account for something like 1% of global electricity use.
If autonomous driving saves a ton of lives then an extra 1% is well worth it. Just from productivity increases alone it may also be well worth it. But also, the "brains" will presumably become more energy-efficient over time (as electronics have generally), and the share of renewables is increasing.
I think that pretty much sums it up on many levels. I think of not just better route planning, but that one delivery driver can service 100's of customers who would otherwise make a trip themselves.
For delivery you're right but we could have this today already. There's no reason not to put minimum wage drivers in a car and have them do this today. We have computers and mapping already, self-driving isn't needed for this at all and paying someone 100 dollars a day or whatever is guaranteed to be much cheaper than the self delivering truck - at least for now and in the foreseeable future.
So why isn't it a thing? Clearly not for the lack of self-driving tech.
It's more palatable to be moving at only 65mph or even 55mph if you're a passenger and don't have to pay attention, as opposed to being an active driver going 75-80mph to finish the task as soon as possible. And I doubt self driving cars will be programmed to speed and then slow down when they spot a cop, where human drivers are generally overspec'd enough to do that.
If this is a serious question then you can look to eight years of autonomus vehicle records held by Rio Tinto in the mega scale mining sector (800 million tonne / annum moved) running dump trucks, trains, etc.
Continous 24/7/365 operations show operational energy amd maintenance efficiences from supervised autonomous trucks over human operated trucks.
Besides being fine, I also doubt it will stay true.
Datacenters have been consuming more and more energy because the work they are expected to do is always increasing. Self-driving cars have some phenomenally large problem to solve... and that's it. Any efficiency gain on those computers will realize as reduced consumption.
The problem will get bigger and bigger, requiring more and more compute power. You see, whatever system you design to autonomously control a car will never be "enough", or perfectly safe. There are simply too many variables in the environment, too many possibilities. So a certain design, taking a certain amount of compute power, will give you a certain level of safety. But it won't be perfect. People will want more safety and better systems. That'll take more power (probably exponentially more, as we've already seen with other simulations). The cycle will never end.
that cycle will break at the target markets and their price sensitivity.
similar to how current safety mechanisms are implemented. * two airbags, low weight constructions in entry segments targetting higher mileage versus 5+ airbags, advanced manufacturing processes that score high in crash tests and more as the segment price increases.
But sometimes it's saved elsewhere. Home office often increases energy usage in data centers through chat software, video conferencing etc. But it saves much more in energy otherwise wasted for commute.
Likewise, other computations in data centers might also save energy used somewhere else. Given all that servers do, 1% of electricity really doesn't seem much.
> If autonomous driving saves a ton of lives then an extra 1% is well worth it
Your total scope is about 40,000 road deaths. Off the top 1/6 of those are pedestrians most often hit at night. Another 1/6 is motorcycles. A little more than 1/2 are those who are drunk or otherwise impaired. Around a 1/3 involve someone under the age of 24 and male.
People presume that driving is exceptionally dangerous. If you actually examine the data you realize it's one of the safest things we do, given the _trillions_ of miles that are driven every year in the US alone.
From this perspective, it's not at all clear that greater levels of automation are going to improve anything. From the history of introduction into airlines, it will initially make travel _less safe_ until drivers are retrained to use the new systems in a way that doesn't enhance risk. These will often be non-intuitive lessons.
If you put a human in a machine also designed by a human and use it around other humans, you will never eliminate human factors.
I don't know where you're getting 40,000 road deaths from, because:
> According to the World Health Organization (WHO), road traffic injuries caused an estimated 1.35 million deaths worldwide in 2016. That is, one person is killed every 26 seconds on average. [1]
I also don't know what the breakdown you give has to do with. If we can save the lives of pedestrians, drunk people, and males under 24, then amazing.
Also, a quick Google search reveals that in the US:
> Deaths and injuries resulting from motor vehicle crashes are the leading cause of death for persons of every age from 3 through 33 (based on 2003 data).
That sounds pretty dangerous to me if it's the leading cause of death. Maybe not that dangerous per-mile, but clearly those miles add up.
> I don't know where you're getting 40,000 road deaths from, because:
US road fatalities. This is because NHTSA catalogs every one of them and the circumstances leading up to them. This product is called the FARS.
> If we can save the lives of pedestrians, drunk people, and males under 24, then amazing.
You're assuming they'll use the automation. Or you're presuming it will be forced on people. Which ignored motorcycles. Or bicycles. People use those.
> Deaths and injuries resulting from motor vehicle crashes are the leading cause of death for persons of every age from 3 through 33 (based on 2003 data).
I'd like to know where you got that statistic from. In the US, it is not likely to be true. Accidental self inflicted injury, a separate category, is often the most common until Heart Disease and Cancer become factors.
The reason to use the US is because it is a developed country, a wide variety of different states and road conditions, and it has a single reporting agency covering it. Presuming that automated vehicles are going to lead to major improvements outside of developed nations is a bit unfounded.
True. And now we contrast this with the additional pollution deaths the increased energy and resource consumption will lead to. 1.3 million people isn't very many.
Currently we're in a paradoxical situation where there are two very contrasting mantras in society: 1. We need to reduce our carbon emissions, cut down on resource consumption and move to only producing energy from renewable sources. And 2. We should have more automation, more AI, more data, more computation in general.
Both directly contradict each other, anyone who's looked into the state of batteries and renewable energy knows we simply don't have a way for both to be true. In fact, we drastically have to reduce energy consumption if we want to make that green dream work, there's no practical way as of today to store all the solar and wind energy that would be needed for it.
I was ready to go do the math and prove you wrong, but you're right!
Some random infographic I found suggests that the mortality associated with driving of all road vehicles is about 1.14 Fatalities per 100 Million Vehicle Miles Traveled. If we say that the average driver is mostly urban and mostly arterial driving, with a health amount on highway on a per hour basis that puts at us at about 1.14 fatalities per (100,000,000 * 1 hour / 45 miles ) 2 million hours.
I tried to find other routine activities with any element of risk that were safer than that and had trouble doing so.
Airline travel is safer - but the amount of "safety stuff" done there is much more extreme, considering how there's little vehicle training, not much maintenance, and driving continues at all times and in all weather.
This isn't to say we shouldn't try to make it safer, but that we should correctly understand where the risks are.
Fatalities/VMT isn't the right stat to use to evaluate safety because injuries and deaths cluster around conflict points(intersections, crossovers to oncoming lane, offramp exits, etc.). It can be simultaneously true that most of the miles driven are safe, but also that there are danger zones which some road users are repeatedly subjected to. And when we talk about AVs being safe, we aren't saying "they'll be safer at avoiding a crash when racing", a situation where one might credibly consider using VMT. It's relatively mundane situations of making a clean left turn, stopping at stop signs, decelerating when exiting the freeway, and so forth. Those are things you encounter with higher frequency when you are making relatively shorter trips.
> If autonomous driving saves a ton of lives then an extra 1% is well worth it. Just from productivity increases alone it may also be well worth it
I think that's a pretty optimistic assessment of it. So far they've only been capable of driving in conditions where humans don't have accidents anyway, and driving in more difficult conditions is a bit of a "ladder to the Moon" scenario.
I don't buy into the whole "think of all the time you'll save" thing either.
The real problem isn't the autonomous part (although it is an issue). Full electrification of just the US will require a doubling of our power generation and transport capacity. This isn't trivial at all.
The US currently produces 1200 GW (power, not talking energy, that is not the correct conversation).
For context, a typical nuclear power plant produces 1 GW. Which means that this doubling effect would be the equivalent of building 1200 new nuclear power plants. I am not saying that's what we have to do, it's a useful mental tool to be used in understanding the scale of the problem.
Solar isn't going to make this happen. For starters, a solar system has to be overbuilt at a minimum of 10 times the required steady-state output. In other words, if you are after 1 GW steady state you need a minimum of 10 GW of solar. The scale is so massive that you start running into regional, weather and MTBF and other issues.
The right solutions is likely a combination of lots of nuclear along with lots of solar, wind and hydro (where possible). We are going to need a serious shift in culture to be able to accomplish this, otherwise, electric cars are going to end-up requiring burning all kinds of fuels just to generate the power to keep them charged. And then you have all the battery issues. At some point you have to honestly ask for an honest accounting of what it might all lead to. I have these images in my mind of massive land areas devastated by strip-mining and other activities just to be "clean" in lands far and away from the mines.
Electric cars are a good idea. We all want them. We just need to be sure we have honest discussions in order to ensure we do it right and don't create a bigger long-term mess than what we manage today.
EDIT:
Why do I say that energy isn't the correct conversation? Because energy allows one to fall under the fallacy of thinking that things can happen over time. That is not the case here. When you have a thousand cars plug in to charge at the same time, you need enough power to service that load. Yes, of course, the objective is to store a certain amount of energy. However, the critical path, the thing that will break systems and overload power generation and transmission is their simultaneous power requirement. The conversation has to be about power. In a place like Los Angeles you might have a million cars simultaneously connecting to the grid to charge. That imposes an instantaneous power requirement that has to be met or things go wrong very quickly. This is how the electric transition is more complicated than a lot of these hand-wavy energy estimates which conveniently ignore what needs to happen instantaneously.
The charging infrastructure will absolutely be smart enough to moderate instantaneous demand as needed to keep the grid healthy.
Likewise, all these batteries will in aggregate form virtual power plants that also serve the grid and make it more resilient.
It is definitely “complicated” in the sense that it’s technically sophisticated, but there’s no cold fusion level obstacle to making it work. Just another couple decades of relentless innovation and technology evolution.
I absolutely agree that the whole world needs to be hard at work scaling up clean electricity generation and storage as this transition occurs!
I do think that the EVs themselves have to play a role in the solving the storage challenge, which requires battery tech to support “spare” cycles to use for storage and delivery alongside the actual driving. The really nice thing is that the economics can work out where dedicating X% of your battery cycles to power storage and delivery pays for Y% of your battery lifespan, where Y > X
— in other words it’s profitable for consumers to participate.
The St James Bay project is a sequence of hydroelectric dams covering an area the size of Scotland -- and it was done in the sixties. We're going to need the power. It's unit-profitable. So really, the issue is where to stick it.
(It would, of course, be much more reasonable to simply cover the rooves of all existing buildings, which is a larger area, and have some accessory generation as well.)
Under absolutely ideal conditions, the power output of a solar array is an inverted parabola spanning 12 hours. This means you need 1.5x nominal capacity (integral if parabola is 2/3 area of constant power rectangle).
Of course, now you need batteries to “fill-in” around the parabolic output curve.
This takes care of daytime needs. Double it to cover nights. You have to more than double batteries because you need them 100% of the time during this period.
We are up to 3x nominal steady-state power. It’s a bit worse than that given system and other losses. Call it 4x. If you need 1 GW, you have to build 4 GW.
However, it doesn’t end there. A single cloud formation can cut output by 50%. I see it all the time on my system. Rain? Up to 95%. I’ve seen my 13 kW system go down to 600 W for days with moderate rain. Dirt? 5% or thereabouts, depending on how bad it is. And then there’s the negative temperature coefficient and seasonal realities, which can drop you by up to 25%.
If you do the math (as I have) the reality of solar is shocking. A 10x overbuilding estimate might be way low. I’ve seen plausible calculations calling for multiples of that.
But if you add in existing hydro + nuclear, a good proportion of wind (which is more consistent and stronger at night), some good long distance interconnects, and a modicum of storage, the number comes out a lot smaller.
Without going too deep: If you need 100% backup for solar, why use solar?
I built a 13 kW array. During recent rains I saw it go down to 600 W peak output. Without 100% backup it would be pointless. I can’t charge electric cars with this, it isn’t enough. I’d have to install a ridiculous amount of storage to to make it viable for a range if use cases. Four years ago, when I installed it, I was idealistic about solar. Four years later I understand the hard-cold reality vs. the fantasy of solar.
> Without going too deep: If you need 100% backup for solar, why use solar?
In addition to this being a lie, if we pretend it's true then there's a sufficient reason. Solar costs less than the marginal cost of other generation except wind.
What is a lie? That solar is so unreliable that you need 100% backup or accept shutting down entire buildings or towns?
That is far from a lie.
I just got back from a business trip to Singapore. Among other places, I visited a food production building where we have installed some of our technology for more efficient indoor farming. This building houses some 240 companies. They have a massive solar array sourcing a good deal of their power.
It rained almost constantly for two weeks. That massive array went to nearly zero output during that time.
The only way that building could function was with 100% backup for solar, provided by the grid.
Batteries? The batteries were drained within a day. They could only be replenished by the use of grid power. Once again, 100% backup.
Not sure how you can possibly call it a lie.
This backup can take many forms. Where it makes sense, batteries might be able to help (yet batteries are not magic and they need a reliable power source to charge). A reliable power grid that can provide 100% of the load at any given moment is a must. This grid might be fed by solar (at a different geographic location), wind, hydro, nuclear or burning something. I believe in Singapore's case it is mostly burning gas.
It might seem sensible to think in terms of a wide geographic distribution of solar arrays to supply city A when weather events render the local solar resources unusable. OK, well, at a minimum, then, you need a 100% backup array located elsewhere. Let's call it 1000 km away.
Sounds great. You are betting on the statistical improbability of both arrays being unusable. Weather patterns being what they are, this might be a sensible bet.
What nobody brings into these conversations is: What happens with cities and towns B, C, D, E, F, G, etc.
If both A and C rely on a B's array for weather event mitigation, B's array has to be overbuilt by a massive factor to, at any given time, supply all of A, B and C's power requirements (or combinations thereof). The same become the case for A and C, as they might be required to supply their 1000 km neighbors.
At the end of the day, I think everyone is starting to understand the ways in which solar and wind are unreliable. The idea that mitigation requires 100% backup can't really be disputed at scale.
Put a different way, as I have said elsewhere, one would have to overbuild solar by an obscene multiplier (some have estimated as much as 70x) to achieve the equivalent of a reliable power source across a wide geographical area.
The UK have a massive solar and wind project bringing them power from Morocco, some 4000 km away. Sounds great...until there's a problem. If the cities using this power don't have 100% reliable backup there will be trouble.
Frankly, it does not help that articles such as this one provide false information that get repeated and becomes widely distributed:
Here, the energy department claims 3.125 million 320 W panels give you 1 GW.
Not even close. First, 320 W panels NEVER make 320 W.
Why?
That rating is at 25° C and ideal lighting conditions. Panels have a negative temperature coefficient. At real operating temperatures they do not output the rated peak output. At actual operating temperatures you will lose 5% to 7% of that. Add other real-life installation factors and 10% is likely a very safe estimate.
That means a 320 W panel will, at best, under ideal condition, produce somewhere in the range of 290 to 300 W peak for a few seconds at the top of the solar day.
That would bring the 3.125 million panel installation down to 938 MW (assuming 300 W peak per panel) for 15 to 30 seconds per day.
Just to deliver a 1 GW peak for a few seconds we have to multiply the number of panels by a factor of almost 7%. That brings the count u...
An incredibly incoherent and long winded way of saying you don't know the term 'capacity factor' and don't understand what dispatchable loads are isn't evidence you weren't lying.
You excel at insulting people online. We've been here before. This is what you do instead of having conversations. Not taking your bait.
My claim is that solar requires 100% backup.
OK, let's break it down:
Solar Output:
Under ideal conditions, roughly inverted parabolic output
1.00 ┼ ╭────╮
0.90 ┤ ╭─╯ ╰─╮
0.80 ┤ ╭─╯ ╰─╮
0.70 ┤ ╭╯ ╰╮
0.60 ┤ ╭╯ ╰╮
0.50 ┤ │ │
0.40 ┤ ╭╯ ╰╮
0.30 ┤ ╭╯ ╰╮
0.20 ┤ ╭╯ ╰╮
0.10 ┤ │ │
0.00 ┼──────────────╯ ╰─────────
Clouds and weather can cause up to 100% drop in output any time
Rain will drop your peak output practically down to zero all day
Dirt will drop total output by a variable amount until cleaned
1.00 ┼ ╭───╮
0.90 ┤ ╭─╮│ ╰╮
0.80 ┤ ╭─╯ ││ │ ╭╮
0.70 ┤ ╭╯ ││ │ │╰╮
0.60 ┤ ╭╯ ││ │ │ ╰╮
0.50 ┤ │ ││ │ │ │
0.40 ┤ ╭╯ ││ ╰─╯ ╰╮
0.30 ┤ ╭╯ ││ ╰╮
0.20 ┤ ╭╯ ││ ╰╮
0.10 ┤ │ ││ │
0.00 ┼──────────────╯ ╰╯ ╰─────────
Daytime:
You need up to 100% backup to be available at any time.
Night:
To state the obvious, 100% backup is required.
Conclusion:
If you need reliable 100% backup at any point
during 24 hours, this backup capacity has to be
available 24/7/365.
The above is absolutely true.
Regarding the link you provided. I am familiar with this article, which, of course, confirms my point.
Figure 5, which is titled "Maps of electricity system reliabilities under the most reliable solar-wind mix without excess generation or energy storage", puts the US at less than 88%.
This means that, ON AVERAGE, 12% of the time you need 100% backup for the most optimal form of solar and wind combined.
I loudly highlighted ON AVERAGE because it is important to understand this is a problem when dealing with statistics. You cannot apply the average of the average of the average to an entire population. This is known as the tyranny of the averages.
Averages are fantastic to wave hands around and talk about generalities. They do not cover my little town of 300K residents going dark for two hours because the solar system powering it had no backup.
The real numbers, once you start to get closer to local realities, are not represented by these average. The real numbers are much worse.
Solar cannot exist without external backup. Solar and wind cannot exist without external backup. That's the real conversation we have to have.
Somehow people read this to mean that I am saying solar and wind are bad and not usable. That is not AT ALL what I am saying. I am simply highlighting that we cannot "go green" just with these technologies. And it gets much, much worse, when we start to throw in hundreds of millions of electric vehicles. My point, in the aggregate, is that we, in the US, desperately ne...
Read the article I linked. Then stop posting lies and long-winded demonstrations that you don't understand that weather can be anything other than 100% correlated over a region and is forecastable.
Then note w2e, hydro and other average-energy-limited dispatchable sources exist. A hydro system which can provide 100MW on average can provide 1GW peak. There are a few tens of GW of waste stream methane available in the US alone. It can be stored and burnt at any time to meet unshiftable demand.
Then note that dispatchable loads such as EV charging exist and exceed non-dispatchable ones. They make VRE easier to use, not harder as you claim.
Then note when pools are pumped, water towers are fied, thermal storage in buildings is charged, zinc refineries and so on are run. It's not done in the middle of the night in regions with thermal generation because people just love noise and graveyard shifts.
If you have 1 watt of constant load, 2 watts (average over 6 months) of dispatchable load, 2 net watts of solar (ie. 4 nameplate watts), long distance transmission, 2 net watts of wind, and 1 net watt of hydro and green fuel burnable in an OCGT, then you have plenty of power.
You're willfully ignoring systems and demand shifting and trying to claim only always-on thermal generation works. This is a lie.
Even assuming fossil fuel for the week a year the system has low output, a VRE dominated grid still cheaper and greener than alternatives and you still don't need to run your electrolysers or zinc refineries or arc furnaces or thermal storage heaters. And if you have a day warning the Al smelters can clear the lines too. This is done all the time in systems dominated by thermal generation to meet demand spikes. Claiming these loads need backup instead of being backup is a lie.
You're trying to pretend turning a fossil fuel generator for the non-dispatchable loads on ever eliminates any gains from the whole system just because it has the same power as the un-shiftable peak demand. This is also a lie.
You're trying to pretend that fixed loads are constant, so a power-limited thermal generator would need no storage or overprovision. This is also a lie.
You're trying to pretend weather events where there is zero wind or sun across a large region for over a day are common. This has never happened. This is a lie.
You're trying to pretend thermal generators never go offline in a correlated way so that any correlated downtime is exclusively VRE. This is a lie.
You're trying to pretend building a system that can eliminate 85-95% of emissions in a year or two for a tenth of the price and then startig on the boondoggle which eliminates 90% of emissions and exports the rest to Niger isn't the obvious and objectively correct answer even if your above lies were accurate.
Wow, you have a consistent need to diminish and insult others. Remarkable.
You continue to resort to insulting me over many threads across many months. Any time there's a thread on a related subject you seem intent in demonstrating your ability to insult me.
I read the article. Of course I did. I read it a couple of years ago, or whenever it was published.
It actually supports my point!
Solar requires the availability reliable power
for up to 100% backup.
That can be batteries, another solar installation a thousand kilometers away, wind, fossil, nuclear, a bunch of people pedaling power-generating bikes. Anything, really.
I am not saying solar or solar + wind are a bad idea, pointless or that we should not use them. Heck, I have made a non-trivial investment in solar myself. And one of the companies I own actually develops related technologies.
Once again, the very article you posted clearly states that the COMBINATION of solar and power in the US is, at best, ~ 88% reliable. Solar alone? Less so.
I can support my claims with actual data from my own solar array. I just visited a six million dollar installation in Singapore. The results are even worse.
Here in So. Cal, we are about to get hit with yet another large storm. All solar power will be reduced to nothing during this time.
I have no clue what your problem is. You are consistent in the use of insults in place of conversation, math and science. Lots of hand-waving. No substance at all. It's exhausting, really, but I am not going to allow you to bully me into discussing this perspective in the context of facts-based discussions.
If I am wrong, I am more than willing to be shown where, why and how. Always. My life and career has been about learning, constantly. That happens with respect and facts, not insults.
EDIT:
Further to the article you seem to be leaning on. From figure 2, and using Photoshop to identify and match the color scale:
"Shading in each panel represents the 39-year average estimated reliability (% of total annual electricity demand met)"
Solar + 12 hours of storage gets up to only 82% reliability (% of demand met). In other words, cities go dark 18% of the time.
Solar requires 100% reliable backup.
Even with the most optimal configuration (close to 50/50 wind and solar with 12 hours of storage), reliability comes up to about 86%.
Solar requires 100% reliable backup.
If we overbuild both solar and wind, add 12 hours of storage and shift the mix to about 40% solar and 60% wind, we can get to about 95% of demand met. Solar alone, in this overbuilt scenario, would only be about 88%. So, we overbuild by 50% and only get about 6% better performance.
Solar requires 100% reliable backup.
And, once again, these are numbers that are presented in terms of averages, which have almost nothing whatsoever to do with local realities. You cannot apply an average figure calculated from multiple averages to an entire country and conclude that those numbers represent local realities.
The fact that you have to use wind and storage to achieve performance above the 53% of demand achievable by solar alone fully supports the point I have been making:
Solar requires 100% reliable backup.
I am eager to see what new insults you are going to use to divert from the inconvenience of having the articl...
The usual motte and bailey nonsense. Look back at your words. You said it needs 100% backup and repeatedly affirmed you meant something with 100% of the output and 100% of the average output and that this made solar useless.
You're now trying to gaslight by saying you meant the whole system needs redundancy.
You're also somehow claiming that anyone advocating solar is advocating for zero wind. This is another unhinged lie.
How do you power a city at night if all you have is solar?
Solar requires 100% reliable backup.
Batteries?
That's backup.
Oh, wait. The batteries need backup because they cannot be charged by a solar array that is sized to only fit the load. You need to oversize it. By a lot. In fact, you need to at least double your array size to charge batteries to be used at night.
In other words, backup for the backup.
Wind?
That's backup.
Hydro?
That's backup.
More solar from thousands of miles away where there might still be light?
That's backup.
Nuclear?
That's backup.
Gas/Coal/Fossil?
That's backup.
Oh, wait, here's Elon Musk talking about something you claimed to be a lie, when I said you need to overbuild solar arrays by 7x or more for continuous power. Here he is, less than two weeks ago:
It's also great that he mentions solar and wind can coexist. Some of backup can be co-located at some locations. Fantastic.
Please stop resorting to insults. Please? Insulting people does not change the reality around you. It never will.
Do the math and understand what you are talking about. Please.
Nobody is saying solar is bad or impossible to use. This interpretation you seem to prefer is entirely fabricated. My claim that solar requires 100% backup is --quite literally-- as true as day and night and as necessary as what happens with solar between day and night. That you continue to attack and insult me over this is difficult to understand.
BTW, he is not talking about weather effects and only thinking about local semi-ideal conditions. I'm sure that will be covered at some point. The multiple is much higher than 5x.
Note to @dang, if you are reading this, please do not delete this thread.
This person really needs to let the world know, in no uncertain terms, who he is. Deleting his relentless reach for insults and attacks would simply provide him with opportunities to do it again and again. If this is the person he wants to be, so be it. Please know it does not affect me and I will not descend to his chosen level of discourse. I will continue providing facts as I think them to be so for anyone to challenge respectfully. I actually want to know if I am wrong, because that is the only way one learns anything. However, this does not happen through insults. It happens in the context of a conversation based on mutual respect and a foundation had in the scientific process, with math and physics as very good starting points.
Hopefully there's a lesson here for others. Treat each other with respect and consideration. Do not react to someone who does not offer you the same. This is their overwhelming objective. To have to descend to their level. Don't do it. Poise in the face of adversity. That's the real proof of a person's character.
> Without going too deep: If you need 100% backup for solar, why use solar
This is where you started. Now you're doubling down on the gaslighting,
You were very explicit and batteries or a resilient VRE system with other renewables were notnl what you were talking about.
And you are consistently disrespectful, arrogant and condescending. The only person stupid enough to think the cry bully act will work is you. Thinking that your snide condescention is subtle is not the same as it not being blindingly obvious.
It's nighttime, my solar array is making zero power right now. Thankfully I have 100% backup from the grid.
Regarding: "Without going too deep: If you need 100% backup for solar, why use solar"
Science and engineering are about always asking questions and challenging assumptions.
Is solar --which requires 100% backup-- better, the best, ideal?
Is it, really?
How about this. Let's use the article you posted to challenge what I've been saying [0]. You know, the one that backfired and actually supported my all my claims so far. Let's use it one more time. Maybe you are right and I am stupid.
Again, table #2.
My question was: Why use solar?
The table shows the "39-year average estimated reliability (% of total annual electricity demand met)".
In other words, what percentage of the time the lights in your town will stay on given a power generation and storage technology mix.
Here's what it says (comments mine):
Without storage ($):
Solar alone: 54% (half the time you go dark)
Wind alone: 82% (wow)
With 3 hours of storage ($$):
Solar alone: 66% (meh)
Wind alone: 85% (nice)
With 12 hours of storage ($$$$$):
Solar alone: 81% (finally)
Wind alone: 88% (better)
Overbuilt by 1.5x with just 3 hours of storage ($$$$):
Solar alone: 70% (meh)
Wind alone: 95% (double-wow)
Wind alone, without storage, is massively better at delivering reliable power than solar. The city goes dark 18% of the time, rather than 46% for solar.
If we just add 3 hours of storage, wind goes up a few percent, same for solar. Not too exciting.
The real find is what happens if we overbuild generation capacity by 1.5x and add just 3 hours of storage:
Wind goes up to 95% reliability, while solar only 70%.
95% RELIABILITY FOR WIND!
70% RELIABILITY FOR SOLAR!
That is a massive find, one deserving saying it out loud.
The city is dark only 5% of the time with wind, vs. 30% with solar.
Both require backup for the percentage they cannot deliver reliably. Wind requires massively less backup.
Is wind cheaper, simpler, easier to maintain over the long term? Not sure, I would have to do the math on that. The questions are compelling enough for me to explore further.
There's another element that is often not explored: Solar requires almost 100% reliance on China for the entire supply chain. Does wind technology inshore most of the supply chain? The world today is a nasty place. If the US, Europe and other regions can inshore manufacturing and technology, rather than send trillions of dollars to China, this would be a good thing, something worth pursuing.
So, again, I ask:
Given the above. Why use solar?
Your article, once again, supports my perspective.
> you are consistently disrespectful, arrogant and condescending.
You are confusing those things with stating what I know because I am sure of it. There is a difference. A professor isn't disrespectful, arrogant and condescending because she asserts what she knows to be true. Neither is a doctor, lawyer or, for that matter, an engineer.
I urge you to, once again, stop the insults, fire-up Excel and do some math before posting. I just used the numbers you provided to show that questioning solar as the best option likely has merit.
Open for discussion? Of course. Get to the point where you have done enough work to have a reasonable understanding of the subject matter. Then, use numbers. Not insults. Anyone would gladly have a conversation with you under those condition.
None of this means solar is useless and should not continue to scale. It does mean we really need to be objective and check our assumptions at the door. My guess is that, yes, of course solar is worthwhile and valuable. Yet, no, it is unlikely to be the best solution for everything. It needs backup, lots of it. It needs wind. Lots of it. And...
Everybody here understands what a system is and understands how to do the math and what the tradeoffs are except you, chum. Your arrogance is still not a substitute for truth.
I notice you've backed off even further from your claims of needing 1200GW of inflexible thermal generation that isn't even useful for peaking (not that this will stop you from posting them again).
Here's yet another example in the comment chain:
> Sure, you did the math for local solar only.
> But if you add in existing hydro + nuclear, a good proportion of wind (which is more consistent and stronger at night), some good long distance interconnects, and a modicum of storage, the number comes out a lot smaller.
You don't understand that you can have a mix of power generation whilst claiming that everyone else is stupid and ignorant. Or you're intentionally lying.
Neither is worthy of respect or of treating you better than you treat others.
If there's one thing the US has enough of, it's space. For Europe, that's going to be much harder. But the US has ample space to not only satisfy their own demand with renewables, but also export it in one form or another.
>That is not the case here. When you have a thousand cars plug in to charge at the same time, you need enough power to service that load.
It seems to me that an obvious and relatively easy fix is for the charging stations to be more complicated that a simple cable. Since the most common use case is one where a car returns to the same place every day, a station with batteries could sip power throughout the day and then deliver it to the car at any rate necessary while maintaining a constant load on the network.
>That is not the case here. When you have a thousand cars plug in to charge at the same time, you need enough power to service that load.
Batteries at point of consumption are a thing. Said batteries can be charged up when power is available, spreading the load. Not talking about the batteries in the cars, in case that wasn't clear. Think Tesla Powerwall. Things like this will become more ubiquitous and likely much cheaper as well. Not to mention grid-scale batteries, which help spread the load in other scenarios.
Also, nuclear and other things (including fossil fuels) will be part of the mix. So "solar isn't going to make this happen" is kind of missing that.
How do we deal with winter now? We do just fine. Heat pumps will only help since they are so much more efficient. The gas can still be used as long as needed, but in a centralized way. And as to the grid, it’s not impossible to improve it in the places where that is needed. I mean, what would be the alternative?
Yes and no. Whatever you take out of a battery has to be replenished quickly. We quickly get to doubling or quadruple the energy station’s batteries. Yet, one can’t get away from having to quickly top off the batteries as they are consumed. At the end of the day it is hard to escape physics.
All that is fine. Humans are good at solving problems and getting shit done. The path to solving this one has been clearly laid out by people who do in fact deal with physics and get shit done, in the most recent Tesla investor day presentations.
> I have these images in my mind of massive land areas devastated by strip-mining and other activities just to be "clean" in lands far and away from the mines.
Because extracting petrol is so much cleaner...
Our civilization is going to use more and more energy as it grows, that's just a fact of life. Might as well get on with the program and install whatever capacity is needed to sustain them. If the US can't be arsed to upgrade its infrastructure to accomodate that (see Texas), then I'm sorry but they deserve what's happening to them.
Also, main issue in terms of capacity in the near-term is mainly linked to time-of use (i.e instantaneous power at a given time); EVs help a lot with this because they can easily charge at a more convenient time.
> So "solar isn't going to make this happen" is kind of missing that.
That's a common fallacy in thinking about this. When you get to hundreds of millions of vehicles there is no such thing as a more convenient time. You'll have massive power requirements 24/7/365.
People tend to think about this in terms of their own micro view of the universe as opposed to the problem at scale. Once you do that, and apply numbers to it, not just wave a magical wand, the reality of the matter becomes visible.
In my case, I wanted to understand this in some detail. I wrote a simulation many years ago to do just that. It was interesting to see my numbers confirmed years later once the question started to be asked. BTW, my model predicted a range between 900 GW and 1400 GW. The accepted estimate today is approximately 1200 GW.
Another popular simplification is to say something like: You have 300 million vehicles. Divide them such that 1/24th of them start charging every hour and that solves your problem. Only 12.5 million cars are plugged in at any given time.
Well, no, that's not the way reality works. This is the equivalent of over-simplified academic physics problems where everything starts with the equivalent of "assume a cow is a uniform sphere of milk one meter in diameter". Works great for simplifying problems. Sucks at painting reality.
> Another popular simplification is to say something like: You have 300 million vehicles. Divide them such that 1/24th of them start charging every hour and that solves your problem. Only 12.5 million cars are plugged in at any given time.
> Well, no, that's not the way reality works. This is the equivalent of over-
simplified academic physics problems where everything starts with the equivalent of "assume a cow is a uniform sphere of milk one meter in diameter". Works great for simplifying problems. Sucks at painting reality.
Yeah, you're right. Grid aware charging and price signals will mean charging will shift to times when energy is cheapest. So rather than being a constant additional load it will fill in all the times with low load reducing the peak requirement to lower than the average.
> Yeah, you're right. Grid aware charging and price signals will mean charging will shift to times when energy is cheapest.
To some extent, sure. However, once you pass a certain number of cars on the road things quickly become stochastic. Just like going to the gasoline station. People will charge when they need to charge. The aggregate loads will not be trivial at all and they will not happen when we with they did. A simple example of this are long-haul trucks travelling the roads at all hours of the day. They don't have a lot of options in terms of charging schedules.
Yes, some percentage of the population will have the ability to behave as nice little robots and charge when they are told they should. Most everyone else will not behave in that way, either by choice or circumstance.
I love the ridiculous hyperbole. Calling the behavior of anyone who currently uses a block heater a 'nice little robot', as well anyone who reacts to information (like it costing $100 to fast charge on the way home but $5 to get a charge by parking daily in a spot with a L1 charger), but anyone who follows your simulation blindly is a free thinker. Around 90% of cars in the US cover under 60 miles a day. This can easily be provided by an L1 charger during off peak.
The overwhelming majority of cars are parked the overwhelming majority of the time. Adding level 1 chargers (ie. A standard household extension cord with a bit of weatherproofing) to parking spots is about two to three orders of magnitude cheaper than the nonsense of massive overprovision of thermal generators, and massive overbuild of transmisikon to only run them 12% of the time.
Margins on trucking are razor thin. Nobody in their right mind will pay your 8x overprovisioned $200/MWh nuclear power at $1600/MWh to pay for the down time or even $100/MWh overprovisioned fossil fuel power when they are already crossing hundreds of miles and anyone with a few hectares can build a battery buffered charging station at $30/MWh.
They'll just book a slot on a charger as soon as they hit the highway to match the mandatory rest breaks and they'll have 4-6 hours and several hundred miles of flexibility to avoid peak charging in addition to the arbitrage supplied by the battery on the charger which is necessary to provide the 3MW or so to fast charge.
You have been chasing me around HN insulting me at every turn. Please. Stop it.
I am more than willing to have a conversation with you. Use facts. Use math. Stop with the insults. OK?
I'll repeat what I said:
"once you pass a certain number of cars on the road things quickly become stochastic"
If you were to model things and play with the numbers you would learn quite a bit, including gaining an understanding of how and why the things you are saying do not align with reality at scale. Once things go beyond a threshold (which is regionally dependent), there is no way to avoid the need for massive amounts of power that we do not currently produce.
You seem to require an explanation for my sarcastic "behave as nice little robots" comment, next time I'll label it as such.
It refers to the ideas pushed by people who seem to think they will be able to control how everyone behaves and, therefore, through this control, in turn, have some degree of control or mitigation for the power issues we are sure to have to face. Another example of this is people saying things like "just drive less".
<sarcasm> Sure. Of course. Why didn't I think of that? </sarcasm>
If I keep seeing the same idiotic lie, I'll keep calling it an idiotic lie. I don't care if it's coming from the same person or not.
You seem to want to denigrate other people, then put words in their mouth and insult them for what you imagined. Then arrogantly insult other people's ability to perform basic logic, but when your straw man is pointed out you cry victim.
Your numbers are only as good as your assumptions, and when you keep insisting on ridiculous assumptions like people treating EVs exactly like ICEs in spite of there being cheaper, more convenient options that don't lead to the conclusion you decided on before you started you will get ridiculous numbers.
Driving less is similarly an option to contribute. People do it everywhere the infrastructure lets them.
Please chill, stop with the insults and go learn something.
I have ignored your attacks and continued to provide data and enough math to understand my perspective (and refute it with math and physics if desired). I have used articles you provided to show I am likely right. I am more than open to dialogue and a discussion on any of this. I am more than open to being wrong. That cannot happen if you refuse to engage in conversation.
Rather than repeat myself, here's the parallel thread where I use your own article to, multiple times, show you that your perspective might not be aligned with reality:
You've been spoon fed the answer over and over, but you still do things like pretend the only options are 100% wind or 100% solar, and pretend there is zero flexibility in load. Try adding existing hydro, existing flexible loads and w2e to the highlighted parts of the paper you're reading rather than only looking at the corners of the graphs.
Or try doing the actual math with actual reality which isn't the bizarro absolute world of a single generation technology where EV owners prefer to charge at precisely 5pm, pay 10x as much and sit around for 20 minutes rather than plugging in while parked you've invented.
Then try not referring to an offhand comment by a serial liar as some kind of argument from authority.
Or try having a single shred of integrity or honesty or respect for anyone around you.
That would be great, but Americans simply won't stand for it. They all want to drive 3-ton SUVs everywhere. And people in many other countries aren't that much better.
Is wanting to drive 3-ton SUVs everywhere something inherent to those humans or do you think there are other circumstances encouraging that? And do you think those circumstances could be changed?
Sometimes I get the impression that despite all the futurism talk most tech people's solution to a trash fire is to say nothing can be done about it and just go back to the line of people throwing more trash into the fire.
Nothing can be done about the trash fire when everyone else actually likes the trash fire. One person can't change the world (not when everyone else is opposed to him).
Sure, you can blame the 3-ton SUV proliferation on the car-centric society Americans have created for themselves, but that isn't going to change anytime soon, unless Americans are forced to change it somehow. Sure, a lot of Americans complain about it, some of them even try to do something about it, but it isn't enough, because most other Americans actually like it that way and work against their efforts. Even the ones who complain mostly don't understand that it's a product of their own choices and how much they're contributing to the problem personally.
To be fair, the sparseness and geographic distribution of population, business and industrial centers in a great deal of the US is such that driving a car quickly becomes the only sensible option. A large SUV, sure, that's a valid question. Some of it is also driven by geography/topology. These cities did not evolve with mass transport as an important option.
To give you an idea of what this means in practice. I can get to the nearest major airport in about 30 minutes by car (at night, no traffic). If I want to use mass transit, it will take a minimum of 3 hours, only if you time it perfectly. It could take you six hours.
It's worse than that. Here's what I would have to do:
Walk 30 minutes
Get on a bus for 35 minutes
Walk for another 5 minutes
Get on another bus for 54 minutes
Walk another 5 minutes
Get on a bus for 20 minutes
Walk 30 minutes
Arrived
Remember that, these days, you have to be at the airport two to three hours before your flight.
Now add weather, heat, rain, carrying luggage, perhaps travelling with kids, etc. I'm sorry, even the most ardent environmentalist is going to just throw everything into a large SUV or minivan and just get there.
This reality will not change. These cities are simply not designed for what exists in other parts of the world. The only hope we have is a future where shared self-driving robot cars can make a difference. I can't see anything else that might change this reality.
If every passenger vehicle were a ford lightning and every truck or commercial van were a class 8 electric semi the average charging load would be under 200GW.
Fuel demand doesn't vary much by week or by month, so only diurnal storage is needed.
Just plug most of the vehicles in when parked, and then duplicate or separate 1 day worth of battery (5-20% extra) and put it at the charger or generator for the rest.
Suddenly your 1200GW 'requirement' shrinks by 90%.
But computing applications also tend to use more or all of the available power... so by itself, tech improvements do not mean that power used will go down correspondingly. (or else thanks to processor Flops/watt improvements, we'd have most laptops running for days on their current batteries, right?)
It seems we'd only expect power use to go down when autonomous driving applications are "feature complete" and stop having a use for additional compute power.
In the meantime, discussions of limits based on achievable/acceptable power draw (a la phones which I would guess have been in a consistent range for years now) are more likely to be meaningful.
With some exceptions, increased efficiency has historically been used to increase capabilities rather than reduce consumption.
It’s by no means ridiculous to assume that this might happen here too, and in fact publishing analyses that make the assumption might be exactly what helps make conservation a priority and breaks those same assumptions.
> It’s by no means ridiculous to assume that this might happen here too, and in fact publishing analyses that make the assumption might be exactly what helps make conservation a priority and breaks those same assumptions.
Power consumption has become a serious problem with auto "infotainment" systems.
There are cars running Unreal Engine in the dashboard.
There's a noticeable reduction in EV range.
(Also a tendency to have too much electronics running while the vehicle is parked. Teslas are noted for this. There's "sentry mode", and "summon mode", and WiFi, and remote updates, and telemetry, and apps phoning home...[1] This can be a worse problem than power consumption for the electronics while driving, since most vehicles are driven a small number of hours per day.)
A modern driverless vehicle like those being deployed by Waymo and Cruise will have considerably more compute than virtually any home computer. A pretty typical setup would be something like a high spec Xeon (or equivalent), 1-2 massive GPUs (or equivalent), lots of big FPGAs/ASICs, and some beefy networking on top, duplicated for redundancy. You wouldn't have to go very far back before the 'supercomputer' moniker became appropriate.
I don't think this person realizes there are AVs with active cooling for their compute (and not dinky PC gaming style active cooling, automotive cooling but for compute)
That’s another problem with modern cars: moving two tons of metal and plastic to, oftentimes, move a single human around is a tremendous waste of energy.
That an increase in energy usage is relatively small compared to an already huge waste of energy isn’t a strong argument to declare it insignificant.
Note that the electric power consumption corresponds to about 1.1 horsepower.
It's been an observation with self-driving container transporters at ports that the self-driving ones use less energy, and get about twice the tire life, than the human driven ones. The self-driving ones are dull, boring, smooth drivers. Which is just fine for container ports.
If AVs become as good as they proclaim, then overall power consumption and emissions will dramatically decline when people no longer have to buy cars to get around.
The early NVIDIA drive based “fully autonomous concept cars” had like 2-3kw worth of computing, Drive Pegasus which is still probably is not enough for L4/5 is 500W per unit and it seems like the minimum viable configuration is 2 of them in a vehicle.
It still isn’t a lot compared to moving the car but it is quite a bit of power.
For a car with 100kw/h battery pack and a range of 500km a 1kw driving computer would reduce the range by about 5% which is about as much as having the air conditioner on.
Looking at the aircon heat pumps for cars that seems about right, it might seem a lot but you are driving a metal box with huge glass windows without any insulation on a sunny day… cars get hot enough to melt plastic if parked at direct sunlight for hours.
It's all relative. They'll use a lot less energy than cars use today. And they won't blast out most of that energy out as heat either. Non autonomous ice cars just aren't very efficient.
And because they are autonomous cars, they will get utilized a lot more meaning we need less of them. People can only drive cars that they own so much. An autonomous car can drive itself, all the time. When it is not driving, it's losing money. So, the most economical thing to do with autonomous car is to utilize the hell out of them and have them driving around people and stuff 24/7 and earning money. Basically, uber without the drivers, tips, etc.
So, we gain some more energy in terms of manufacturing efficiencies because we need less vehicles to transport people and goods the same distance.
And of course the vast majority of the power used for this is going to come from cheap, clean, renewable sources instead of 100% of it being fossil fuels being transformed into CO2 that is blasted into the atmosphere via the tail pipe.
So yes, there is a bit of power involved. But vastly less than what we are spending on transport today. And it will be generated in a vastly cleaner way.
Autonomous cars don't have to be electric cars and electric cars don't have to be autonomous. These are two different things. An autonomous non-ev car will blast out more heat than a non-autonomous ev.
Strictly speaking not. But ice cars are a lot more complex. And they are more expensive to operate. So, I doubt there will ever be a large market, or even any market, for autonomous ice vehicles.
I could travel coast-to-coast in a weekend in an autonomous ICE vehicle (35-45 hours on the road, depending on traffic), but likely not in an autonomous EV.
114 comments
[ 3.4 ms ] story [ 214 ms ] threadHow Much Power Do Autonomous Vehicles Use? A Lot
A model of power consumption by autonomous vehicles predicts that once widely adopted, they will consume as much power worldwide as data centers do.
By Bill Schweber 02.27.2023
My summary:
> “1 billion autonomous vehicles, each driving for one hour per day with a computer consuming 840 W, would consume enough energy to generate about the same amount of emissions as data centers currently do.”
> Well… maybe yes, maybe not. Despite the academic veneer and a detailed table explaining their model and its embedded assumptions, the inherent nature of the model has many rough estimates, sophisticated guesses and outright assumptions. I’d compare it more to a fancy “back of the envelope” assessment (Figure 2) rather than a rigorous academic analysis. Step back, and there’s the reality that any assessment of the future for a situation such as this is simply not possible.
> Call me cynical, but it seems that the design reality is that system demands and thus power use often expand to the maximum the system can provide or tolerate. Remember, when transistors and ICs were devised and then went to ever-lower voltage/power, the per-function dissipation dropped by several orders of magnitude—you can’t argue with that fact.
> However, coincidentally with this good news, user requirements increased at a greater rate, so now you have server racks pushing almost 10 kW each—an almost unthinkable figure just a decade ago, up from about 2.4 kW each in 2011 according to one survey (I have seen others, and those numbers are similar) (Figure 3).
>> However, coincidentally with this good news, user requirements increased at a greater rate, so now you have server racks pushing almost 10 kW each—an almost unthinkable figure just a decade ago, up from about 2.4 kW each in 2011 according to one survey (I have seen others, and those numbers are similar) (Figure 3).
https://en.wikipedia.org/wiki/Jevons_paradox
When driving 100 km/h for an hour, a reasonably efficient car will use 5 liters of petrol. The energy in 5 liters is about 5 * 34 = 170 MJ. Hence, 47 kWh.
So yes, it’s a lot of energy used by the computations, but still less than the actual force required to move the car forward.
Edit: Fixed number.
At scale safe autonomous driving will be saving 100k lives and eliminating over a trillion of dollars in economic losses per year.
This is a minuscule amount of energy expended for the results achieved by any reasonable standard!
It's referring to the energy to power the brains of self-driving, not to move the vehicle.
But that seems... fine? I can't tell if the article wants us to think that's a lot.
Data centers and networks account for something like 1% of global electricity use.
If autonomous driving saves a ton of lives then an extra 1% is well worth it. Just from productivity increases alone it may also be well worth it. But also, the "brains" will presumably become more energy-efficient over time (as electronics have generally), and the share of renewables is increasing.
I don't really see a problem here.
So why isn't it a thing? Clearly not for the lack of self-driving tech.
Continous 24/7/365 operations show operational energy amd maintenance efficiences from supervised autonomous trucks over human operated trucks.
Datacenters have been consuming more and more energy because the work they are expected to do is always increasing. Self-driving cars have some phenomenally large problem to solve... and that's it. Any efficiency gain on those computers will realize as reduced consumption.
The problem will get bigger and bigger, requiring more and more compute power. You see, whatever system you design to autonomously control a car will never be "enough", or perfectly safe. There are simply too many variables in the environment, too many possibilities. So a certain design, taking a certain amount of compute power, will give you a certain level of safety. But it won't be perfect. People will want more safety and better systems. That'll take more power (probably exponentially more, as we've already seen with other simulations). The cycle will never end.
similar to how current safety mechanisms are implemented. * two airbags, low weight constructions in entry segments targetting higher mileage versus 5+ airbags, advanced manufacturing processes that score high in crash tests and more as the segment price increases.
Likewise, other computations in data centers might also save energy used somewhere else. Given all that servers do, 1% of electricity really doesn't seem much.
Your total scope is about 40,000 road deaths. Off the top 1/6 of those are pedestrians most often hit at night. Another 1/6 is motorcycles. A little more than 1/2 are those who are drunk or otherwise impaired. Around a 1/3 involve someone under the age of 24 and male.
People presume that driving is exceptionally dangerous. If you actually examine the data you realize it's one of the safest things we do, given the _trillions_ of miles that are driven every year in the US alone.
From this perspective, it's not at all clear that greater levels of automation are going to improve anything. From the history of introduction into airlines, it will initially make travel _less safe_ until drivers are retrained to use the new systems in a way that doesn't enhance risk. These will often be non-intuitive lessons.
If you put a human in a machine also designed by a human and use it around other humans, you will never eliminate human factors.
> According to the World Health Organization (WHO), road traffic injuries caused an estimated 1.35 million deaths worldwide in 2016. That is, one person is killed every 26 seconds on average. [1]
I also don't know what the breakdown you give has to do with. If we can save the lives of pedestrians, drunk people, and males under 24, then amazing.
Also, a quick Google search reveals that in the US:
> Deaths and injuries resulting from motor vehicle crashes are the leading cause of death for persons of every age from 3 through 33 (based on 2003 data).
That sounds pretty dangerous to me if it's the leading cause of death. Maybe not that dangerous per-mile, but clearly those miles add up.
[1] https://en.wikipedia.org/wiki/List_of_countries_by_traffic-r...
US road fatalities. This is because NHTSA catalogs every one of them and the circumstances leading up to them. This product is called the FARS.
> If we can save the lives of pedestrians, drunk people, and males under 24, then amazing.
You're assuming they'll use the automation. Or you're presuming it will be forced on people. Which ignored motorcycles. Or bicycles. People use those.
> Deaths and injuries resulting from motor vehicle crashes are the leading cause of death for persons of every age from 3 through 33 (based on 2003 data).
I'd like to know where you got that statistic from. In the US, it is not likely to be true. Accidental self inflicted injury, a separate category, is often the most common until Heart Disease and Cancer become factors.
The reason to use the US is because it is a developed country, a wide variety of different states and road conditions, and it has a single reporting agency covering it. Presuming that automated vehicles are going to lead to major improvements outside of developed nations is a bit unfounded.
Currently we're in a paradoxical situation where there are two very contrasting mantras in society: 1. We need to reduce our carbon emissions, cut down on resource consumption and move to only producing energy from renewable sources. And 2. We should have more automation, more AI, more data, more computation in general.
Both directly contradict each other, anyone who's looked into the state of batteries and renewable energy knows we simply don't have a way for both to be true. In fact, we drastically have to reduce energy consumption if we want to make that green dream work, there's no practical way as of today to store all the solar and wind energy that would be needed for it.
Some random infographic I found suggests that the mortality associated with driving of all road vehicles is about 1.14 Fatalities per 100 Million Vehicle Miles Traveled. If we say that the average driver is mostly urban and mostly arterial driving, with a health amount on highway on a per hour basis that puts at us at about 1.14 fatalities per (100,000,000 * 1 hour / 45 miles ) 2 million hours.
I tried to find other routine activities with any element of risk that were safer than that and had trouble doing so.
This isn't to say we shouldn't try to make it safer, but that we should correctly understand where the risks are.
I think that's a pretty optimistic assessment of it. So far they've only been capable of driving in conditions where humans don't have accidents anyway, and driving in more difficult conditions is a bit of a "ladder to the Moon" scenario.
I don't buy into the whole "think of all the time you'll save" thing either.
The US currently produces 1200 GW (power, not talking energy, that is not the correct conversation).
For context, a typical nuclear power plant produces 1 GW. Which means that this doubling effect would be the equivalent of building 1200 new nuclear power plants. I am not saying that's what we have to do, it's a useful mental tool to be used in understanding the scale of the problem.
Solar isn't going to make this happen. For starters, a solar system has to be overbuilt at a minimum of 10 times the required steady-state output. In other words, if you are after 1 GW steady state you need a minimum of 10 GW of solar. The scale is so massive that you start running into regional, weather and MTBF and other issues.
The right solutions is likely a combination of lots of nuclear along with lots of solar, wind and hydro (where possible). We are going to need a serious shift in culture to be able to accomplish this, otherwise, electric cars are going to end-up requiring burning all kinds of fuels just to generate the power to keep them charged. And then you have all the battery issues. At some point you have to honestly ask for an honest accounting of what it might all lead to. I have these images in my mind of massive land areas devastated by strip-mining and other activities just to be "clean" in lands far and away from the mines.
Electric cars are a good idea. We all want them. We just need to be sure we have honest discussions in order to ensure we do it right and don't create a bigger long-term mess than what we manage today.
EDIT:
Why do I say that energy isn't the correct conversation? Because energy allows one to fall under the fallacy of thinking that things can happen over time. That is not the case here. When you have a thousand cars plug in to charge at the same time, you need enough power to service that load. Yes, of course, the objective is to store a certain amount of energy. However, the critical path, the thing that will break systems and overload power generation and transmission is their simultaneous power requirement. The conversation has to be about power. In a place like Los Angeles you might have a million cars simultaneously connecting to the grid to charge. That imposes an instantaneous power requirement that has to be met or things go wrong very quickly. This is how the electric transition is more complicated than a lot of these hand-wavy energy estimates which conveniently ignore what needs to happen instantaneously.
Likewise, all these batteries will in aggregate form virtual power plants that also serve the grid and make it more resilient.
It is definitely “complicated” in the sense that it’s technically sophisticated, but there’s no cold fusion level obstacle to making it work. Just another couple decades of relentless innovation and technology evolution.
I absolutely agree that the whole world needs to be hard at work scaling up clean electricity generation and storage as this transition occurs!
I do think that the EVs themselves have to play a role in the solving the storage challenge, which requires battery tech to support “spare” cycles to use for storage and delivery alongside the actual driving. The really nice thing is that the economics can work out where dedicating X% of your battery cycles to power storage and delivery pays for Y% of your battery lifespan, where Y > X — in other words it’s profitable for consumers to participate.
Seems like a large number. But if that replaced all gasoline, diesel and natural gas consumption, we'd repay that $6T of solar in under 5 years.
The St James Bay project is a sequence of hydroelectric dams covering an area the size of Scotland -- and it was done in the sixties. We're going to need the power. It's unit-profitable. So really, the issue is where to stick it.
(It would, of course, be much more reasonable to simply cover the rooves of all existing buildings, which is a larger area, and have some accessory generation as well.)
Under absolutely ideal conditions, the power output of a solar array is an inverted parabola spanning 12 hours. This means you need 1.5x nominal capacity (integral if parabola is 2/3 area of constant power rectangle).
Of course, now you need batteries to “fill-in” around the parabolic output curve.
This takes care of daytime needs. Double it to cover nights. You have to more than double batteries because you need them 100% of the time during this period.
We are up to 3x nominal steady-state power. It’s a bit worse than that given system and other losses. Call it 4x. If you need 1 GW, you have to build 4 GW.
However, it doesn’t end there. A single cloud formation can cut output by 50%. I see it all the time on my system. Rain? Up to 95%. I’ve seen my 13 kW system go down to 600 W for days with moderate rain. Dirt? 5% or thereabouts, depending on how bad it is. And then there’s the negative temperature coefficient and seasonal realities, which can drop you by up to 25%.
If you do the math (as I have) the reality of solar is shocking. A 10x overbuilding estimate might be way low. I’ve seen plausible calculations calling for multiples of that.
But if you add in existing hydro + nuclear, a good proportion of wind (which is more consistent and stronger at night), some good long distance interconnects, and a modicum of storage, the number comes out a lot smaller.
Here's the math: https://mitpress.mit.edu/9780262545044/electrify/
It comes up with a 20% overbuild.
Without going too deep: If you need 100% backup for solar, why use solar?
I built a 13 kW array. During recent rains I saw it go down to 600 W peak output. Without 100% backup it would be pointless. I can’t charge electric cars with this, it isn’t enough. I’d have to install a ridiculous amount of storage to to make it viable for a range if use cases. Four years ago, when I installed it, I was idealistic about solar. Four years later I understand the hard-cold reality vs. the fantasy of solar.
In addition to this being a lie, if we pretend it's true then there's a sufficient reason. Solar costs less than the marginal cost of other generation except wind.
That is far from a lie.
I just got back from a business trip to Singapore. Among other places, I visited a food production building where we have installed some of our technology for more efficient indoor farming. This building houses some 240 companies. They have a massive solar array sourcing a good deal of their power.
It rained almost constantly for two weeks. That massive array went to nearly zero output during that time.
The only way that building could function was with 100% backup for solar, provided by the grid.
Batteries? The batteries were drained within a day. They could only be replenished by the use of grid power. Once again, 100% backup.
Not sure how you can possibly call it a lie.
This backup can take many forms. Where it makes sense, batteries might be able to help (yet batteries are not magic and they need a reliable power source to charge). A reliable power grid that can provide 100% of the load at any given moment is a must. This grid might be fed by solar (at a different geographic location), wind, hydro, nuclear or burning something. I believe in Singapore's case it is mostly burning gas.
It might seem sensible to think in terms of a wide geographic distribution of solar arrays to supply city A when weather events render the local solar resources unusable. OK, well, at a minimum, then, you need a 100% backup array located elsewhere. Let's call it 1000 km away.
Sounds great. You are betting on the statistical improbability of both arrays being unusable. Weather patterns being what they are, this might be a sensible bet.
What nobody brings into these conversations is: What happens with cities and towns B, C, D, E, F, G, etc.
If both A and C rely on a B's array for weather event mitigation, B's array has to be overbuilt by a massive factor to, at any given time, supply all of A, B and C's power requirements (or combinations thereof). The same become the case for A and C, as they might be required to supply their 1000 km neighbors.
At the end of the day, I think everyone is starting to understand the ways in which solar and wind are unreliable. The idea that mitigation requires 100% backup can't really be disputed at scale.
Put a different way, as I have said elsewhere, one would have to overbuild solar by an obscene multiplier (some have estimated as much as 70x) to achieve the equivalent of a reliable power source across a wide geographical area.
The UK have a massive solar and wind project bringing them power from Morocco, some 4000 km away. Sounds great...until there's a problem. If the cities using this power don't have 100% reliable backup there will be trouble.
Frankly, it does not help that articles such as this one provide false information that get repeated and becomes widely distributed:
https://www.energy.gov/eere/articles/how-much-power-1-gigawa...
Here, the energy department claims 3.125 million 320 W panels give you 1 GW.
Not even close. First, 320 W panels NEVER make 320 W.
Why?
That rating is at 25° C and ideal lighting conditions. Panels have a negative temperature coefficient. At real operating temperatures they do not output the rated peak output. At actual operating temperatures you will lose 5% to 7% of that. Add other real-life installation factors and 10% is likely a very safe estimate.
That means a 320 W panel will, at best, under ideal condition, produce somewhere in the range of 290 to 300 W peak for a few seconds at the top of the solar day.
That would bring the 3.125 million panel installation down to 938 MW (assuming 300 W peak per panel) for 15 to 30 seconds per day.
Just to deliver a 1 GW peak for a few seconds we have to multiply the number of panels by a factor of almost 7%. That brings the count u...
As to the amount of backup needed: https://www.nature.com/articles/s41467-021-26355-z
Then consider all the dispatchable loads that decrease the gap even further.
My claim is that solar requires 100% backup.
OK, let's break it down:
The above is absolutely true.Regarding the link you provided. I am familiar with this article, which, of course, confirms my point.
Figure 5, which is titled "Maps of electricity system reliabilities under the most reliable solar-wind mix without excess generation or energy storage", puts the US at less than 88%.
This means that, ON AVERAGE, 12% of the time you need 100% backup for the most optimal form of solar and wind combined.
I loudly highlighted ON AVERAGE because it is important to understand this is a problem when dealing with statistics. You cannot apply the average of the average of the average to an entire population. This is known as the tyranny of the averages.
Averages are fantastic to wave hands around and talk about generalities. They do not cover my little town of 300K residents going dark for two hours because the solar system powering it had no backup.
The real numbers, once you start to get closer to local realities, are not represented by these average. The real numbers are much worse.
Solar cannot exist without external backup. Solar and wind cannot exist without external backup. That's the real conversation we have to have.
Somehow people read this to mean that I am saying solar and wind are bad and not usable. That is not AT ALL what I am saying. I am simply highlighting that we cannot "go green" just with these technologies. And it gets much, much worse, when we start to throw in hundreds of millions of electric vehicles. My point, in the aggregate, is that we, in the US, desperately ne...
Then note w2e, hydro and other average-energy-limited dispatchable sources exist. A hydro system which can provide 100MW on average can provide 1GW peak. There are a few tens of GW of waste stream methane available in the US alone. It can be stored and burnt at any time to meet unshiftable demand.
Then note that dispatchable loads such as EV charging exist and exceed non-dispatchable ones. They make VRE easier to use, not harder as you claim.
Then note when pools are pumped, water towers are fied, thermal storage in buildings is charged, zinc refineries and so on are run. It's not done in the middle of the night in regions with thermal generation because people just love noise and graveyard shifts.
If you have 1 watt of constant load, 2 watts (average over 6 months) of dispatchable load, 2 net watts of solar (ie. 4 nameplate watts), long distance transmission, 2 net watts of wind, and 1 net watt of hydro and green fuel burnable in an OCGT, then you have plenty of power.
You're willfully ignoring systems and demand shifting and trying to claim only always-on thermal generation works. This is a lie.
Even assuming fossil fuel for the week a year the system has low output, a VRE dominated grid still cheaper and greener than alternatives and you still don't need to run your electrolysers or zinc refineries or arc furnaces or thermal storage heaters. And if you have a day warning the Al smelters can clear the lines too. This is done all the time in systems dominated by thermal generation to meet demand spikes. Claiming these loads need backup instead of being backup is a lie.
You're trying to pretend turning a fossil fuel generator for the non-dispatchable loads on ever eliminates any gains from the whole system just because it has the same power as the un-shiftable peak demand. This is also a lie.
You're trying to pretend that fixed loads are constant, so a power-limited thermal generator would need no storage or overprovision. This is also a lie.
You're trying to pretend weather events where there is zero wind or sun across a large region for over a day are common. This has never happened. This is a lie.
You're trying to pretend thermal generators never go offline in a correlated way so that any correlated downtime is exclusively VRE. This is a lie.
You're trying to pretend building a system that can eliminate 85-95% of emissions in a year or two for a tenth of the price and then startig on the boondoggle which eliminates 90% of emissions and exports the rest to Niger isn't the obvious and objectively correct answer even if your above lies were accurate.
You continue to resort to insulting me over many threads across many months. Any time there's a thread on a related subject you seem intent in demonstrating your ability to insult me.
I read the article. Of course I did. I read it a couple of years ago, or whenever it was published.
It actually supports my point!
That can be batteries, another solar installation a thousand kilometers away, wind, fossil, nuclear, a bunch of people pedaling power-generating bikes. Anything, really.I am not saying solar or solar + wind are a bad idea, pointless or that we should not use them. Heck, I have made a non-trivial investment in solar myself. And one of the companies I own actually develops related technologies.
Once again, the very article you posted clearly states that the COMBINATION of solar and power in the US is, at best, ~ 88% reliable. Solar alone? Less so.
I can support my claims with actual data from my own solar array. I just visited a six million dollar installation in Singapore. The results are even worse.
Here in So. Cal, we are about to get hit with yet another large storm. All solar power will be reduced to nothing during this time.
I have no clue what your problem is. You are consistent in the use of insults in place of conversation, math and science. Lots of hand-waving. No substance at all. It's exhausting, really, but I am not going to allow you to bully me into discussing this perspective in the context of facts-based discussions.
If I am wrong, I am more than willing to be shown where, why and how. Always. My life and career has been about learning, constantly. That happens with respect and facts, not insults.
Further to the article you seem to be leaning on. From figure 2, and using Photoshop to identify and match the color scale:"Shading in each panel represents the 39-year average estimated reliability (% of total annual electricity demand met)"
https://i.imgur.com/9OMK9eZ.png
This shows that 100% solar (without mixing-in wind) in the US only has a reliability (% of demand met) of about 53%.
If we add 12 hours of storage (also in figure 2 of the article YOU provided):https://i.imgur.com/ut52ZbI.png
Solar + 12 hours of storage gets up to only 82% reliability (% of demand met). In other words, cities go dark 18% of the time.
Even with the most optimal configuration (close to 50/50 wind and solar with 12 hours of storage), reliability comes up to about 86%. If we overbuild both solar and wind, add 12 hours of storage and shift the mix to about 40% solar and 60% wind, we can get to about 95% of demand met. Solar alone, in this overbuilt scenario, would only be about 88%. So, we overbuild by 50% and only get about 6% better performance. And, once again, these are numbers that are presented in terms of averages, which have almost nothing whatsoever to do with local realities. You cannot apply an average figure calculated from multiple averages to an entire country and conclude that those numbers represent local realities.The fact that you have to use wind and storage to achieve performance above the 53% of demand achievable by solar alone fully supports the point I have been making:
I am eager to see what new insults you are going to use to divert from the inconvenience of having the articl...You're now trying to gaslight by saying you meant the whole system needs redundancy.
You're also somehow claiming that anyone advocating solar is advocating for zero wind. This is another unhinged lie.
Solar requires 100% reliable backup.
Batteries?
That's backup.
Oh, wait. The batteries need backup because they cannot be charged by a solar array that is sized to only fit the load. You need to oversize it. By a lot. In fact, you need to at least double your array size to charge batteries to be used at night.
In other words, backup for the backup.
Wind?
That's backup.
Hydro?
That's backup.
More solar from thousands of miles away where there might still be light?
That's backup.
Nuclear?
That's backup.
Gas/Coal/Fossil?
That's backup.
Oh, wait, here's Elon Musk talking about something you claimed to be a lie, when I said you need to overbuild solar arrays by 7x or more for continuous power. Here he is, less than two weeks ago:
https://www.youtube.com/live/Hl1zEzVUV7w?feature=share&t=121...
You need to 5x for continuous power, he says.
Oops.
It's also great that he mentions solar and wind can coexist. Some of backup can be co-located at some locations. Fantastic.
Please stop resorting to insults. Please? Insulting people does not change the reality around you. It never will.
Do the math and understand what you are talking about. Please.
Nobody is saying solar is bad or impossible to use. This interpretation you seem to prefer is entirely fabricated. My claim that solar requires 100% backup is --quite literally-- as true as day and night and as necessary as what happens with solar between day and night. That you continue to attack and insult me over this is difficult to understand.
BTW, he is not talking about weather effects and only thinking about local semi-ideal conditions. I'm sure that will be covered at some point. The multiple is much higher than 5x.
This person really needs to let the world know, in no uncertain terms, who he is. Deleting his relentless reach for insults and attacks would simply provide him with opportunities to do it again and again. If this is the person he wants to be, so be it. Please know it does not affect me and I will not descend to his chosen level of discourse. I will continue providing facts as I think them to be so for anyone to challenge respectfully. I actually want to know if I am wrong, because that is the only way one learns anything. However, this does not happen through insults. It happens in the context of a conversation based on mutual respect and a foundation had in the scientific process, with math and physics as very good starting points.Hopefully there's a lesson here for others. Treat each other with respect and consideration. Do not react to someone who does not offer you the same. This is their overwhelming objective. To have to descend to their level. Don't do it. Poise in the face of adversity. That's the real proof of a person's character.
This is where you started. Now you're doubling down on the gaslighting,
You were very explicit and batteries or a resilient VRE system with other renewables were notnl what you were talking about.
And you are consistently disrespectful, arrogant and condescending. The only person stupid enough to think the cry bully act will work is you. Thinking that your snide condescention is subtle is not the same as it not being blindingly obvious.
Hmmm. OK.
It's nighttime, my solar array is making zero power right now. Thankfully I have 100% backup from the grid.
Regarding: "Without going too deep: If you need 100% backup for solar, why use solar"
Science and engineering are about always asking questions and challenging assumptions.
Is solar --which requires 100% backup-- better, the best, ideal?
Is it, really?
How about this. Let's use the article you posted to challenge what I've been saying [0]. You know, the one that backfired and actually supported my all my claims so far. Let's use it one more time. Maybe you are right and I am stupid.
Again, table #2.
My question was: Why use solar?
The table shows the "39-year average estimated reliability (% of total annual electricity demand met)".
In other words, what percentage of the time the lights in your town will stay on given a power generation and storage technology mix.
Here's what it says (comments mine):
Wind alone, without storage, is massively better at delivering reliable power than solar. The city goes dark 18% of the time, rather than 46% for solar.If we just add 3 hours of storage, wind goes up a few percent, same for solar. Not too exciting.
The real find is what happens if we overbuild generation capacity by 1.5x and add just 3 hours of storage:
Wind goes up to 95% reliability, while solar only 70%.
95% RELIABILITY FOR WIND!
70% RELIABILITY FOR SOLAR!
That is a massive find, one deserving saying it out loud.
The city is dark only 5% of the time with wind, vs. 30% with solar.
Both require backup for the percentage they cannot deliver reliably. Wind requires massively less backup.
Is wind cheaper, simpler, easier to maintain over the long term? Not sure, I would have to do the math on that. The questions are compelling enough for me to explore further.
There's another element that is often not explored: Solar requires almost 100% reliance on China for the entire supply chain. Does wind technology inshore most of the supply chain? The world today is a nasty place. If the US, Europe and other regions can inshore manufacturing and technology, rather than send trillions of dollars to China, this would be a good thing, something worth pursuing.
So, again, I ask:
Given the above. Why use solar?
Your article, once again, supports my perspective.
> you are consistently disrespectful, arrogant and condescending.
You are confusing those things with stating what I know because I am sure of it. There is a difference. A professor isn't disrespectful, arrogant and condescending because she asserts what she knows to be true. Neither is a doctor, lawyer or, for that matter, an engineer.
I urge you to, once again, stop the insults, fire-up Excel and do some math before posting. I just used the numbers you provided to show that questioning solar as the best option likely has merit.
Open for discussion? Of course. Get to the point where you have done enough work to have a reasonable understanding of the subject matter. Then, use numbers. Not insults. Anyone would gladly have a conversation with you under those condition.
None of this means solar is useless and should not continue to scale. It does mean we really need to be objective and check our assumptions at the door. My guess is that, yes, of course solar is worthwhile and valuable. Yet, no, it is unlikely to be the best solution for everything. It needs backup, lots of it. It needs wind. Lots of it. And...
I notice you've backed off even further from your claims of needing 1200GW of inflexible thermal generation that isn't even useful for peaking (not that this will stop you from posting them again).
Here's yet another example in the comment chain:
> Sure, you did the math for local solar only.
> But if you add in existing hydro + nuclear, a good proportion of wind (which is more consistent and stronger at night), some good long distance interconnects, and a modicum of storage, the number comes out a lot smaller.
> Here's the math: https://mitpress.mit.edu/9780262545044/electrify/
> It comes up with a 20% overbuild.
Yet another example of someone pointing out your straw man. Then you arrogantly insulting them.
Everyone you're talking down to has either done the math, or read journals by competent people who have.
Wow.
I can't imagine what aspect of your life drives you to behave in this manner.
Please stop chasing me around HN to insult me. OK? You can disagree with me all you want. Just stop with this troublesome behavior. It isn't healthy.
This conversation is over.
You don't understand that you can have a mix of power generation whilst claiming that everyone else is stupid and ignorant. Or you're intentionally lying.
Neither is worthy of respect or of treating you better than you treat others.
It seems to me that an obvious and relatively easy fix is for the charging stations to be more complicated that a simple cable. Since the most common use case is one where a car returns to the same place every day, a station with batteries could sip power throughout the day and then deliver it to the car at any rate necessary while maintaining a constant load on the network.
Batteries at point of consumption are a thing. Said batteries can be charged up when power is available, spreading the load. Not talking about the batteries in the cars, in case that wasn't clear. Think Tesla Powerwall. Things like this will become more ubiquitous and likely much cheaper as well. Not to mention grid-scale batteries, which help spread the load in other scenarios.
Also, nuclear and other things (including fossil fuels) will be part of the mix. So "solar isn't going to make this happen" is kind of missing that.
I guess you didn't read my post then.
Because extracting petrol is so much cleaner...
Our civilization is going to use more and more energy as it grows, that's just a fact of life. Might as well get on with the program and install whatever capacity is needed to sustain them. If the US can't be arsed to upgrade its infrastructure to accomodate that (see Texas), then I'm sorry but they deserve what's happening to them.
Also, main issue in terms of capacity in the near-term is mainly linked to time-of use (i.e instantaneous power at a given time); EVs help a lot with this because they can easily charge at a more convenient time.
That's a common fallacy in thinking about this. When you get to hundreds of millions of vehicles there is no such thing as a more convenient time. You'll have massive power requirements 24/7/365.
People tend to think about this in terms of their own micro view of the universe as opposed to the problem at scale. Once you do that, and apply numbers to it, not just wave a magical wand, the reality of the matter becomes visible.
In my case, I wanted to understand this in some detail. I wrote a simulation many years ago to do just that. It was interesting to see my numbers confirmed years later once the question started to be asked. BTW, my model predicted a range between 900 GW and 1400 GW. The accepted estimate today is approximately 1200 GW.
Another popular simplification is to say something like: You have 300 million vehicles. Divide them such that 1/24th of them start charging every hour and that solves your problem. Only 12.5 million cars are plugged in at any given time.
Well, no, that's not the way reality works. This is the equivalent of over-simplified academic physics problems where everything starts with the equivalent of "assume a cow is a uniform sphere of milk one meter in diameter". Works great for simplifying problems. Sucks at painting reality.
> Well, no, that's not the way reality works. This is the equivalent of over- simplified academic physics problems where everything starts with the equivalent of "assume a cow is a uniform sphere of milk one meter in diameter". Works great for simplifying problems. Sucks at painting reality.
Yeah, you're right. Grid aware charging and price signals will mean charging will shift to times when energy is cheapest. So rather than being a constant additional load it will fill in all the times with low load reducing the peak requirement to lower than the average.
To some extent, sure. However, once you pass a certain number of cars on the road things quickly become stochastic. Just like going to the gasoline station. People will charge when they need to charge. The aggregate loads will not be trivial at all and they will not happen when we with they did. A simple example of this are long-haul trucks travelling the roads at all hours of the day. They don't have a lot of options in terms of charging schedules.
Yes, some percentage of the population will have the ability to behave as nice little robots and charge when they are told they should. Most everyone else will not behave in that way, either by choice or circumstance.
The overwhelming majority of cars are parked the overwhelming majority of the time. Adding level 1 chargers (ie. A standard household extension cord with a bit of weatherproofing) to parking spots is about two to three orders of magnitude cheaper than the nonsense of massive overprovision of thermal generators, and massive overbuild of transmisikon to only run them 12% of the time.
Margins on trucking are razor thin. Nobody in their right mind will pay your 8x overprovisioned $200/MWh nuclear power at $1600/MWh to pay for the down time or even $100/MWh overprovisioned fossil fuel power when they are already crossing hundreds of miles and anyone with a few hectares can build a battery buffered charging station at $30/MWh.
They'll just book a slot on a charger as soon as they hit the highway to match the mandatory rest breaks and they'll have 4-6 hours and several hundred miles of flexibility to avoid peak charging in addition to the arbitrage supplied by the battery on the charger which is necessary to provide the 3MW or so to fast charge.
I am more than willing to have a conversation with you. Use facts. Use math. Stop with the insults. OK?
I'll repeat what I said:
"once you pass a certain number of cars on the road things quickly become stochastic"
If you were to model things and play with the numbers you would learn quite a bit, including gaining an understanding of how and why the things you are saying do not align with reality at scale. Once things go beyond a threshold (which is regionally dependent), there is no way to avoid the need for massive amounts of power that we do not currently produce.
You seem to require an explanation for my sarcastic "behave as nice little robots" comment, next time I'll label it as such.
It refers to the ideas pushed by people who seem to think they will be able to control how everyone behaves and, therefore, through this control, in turn, have some degree of control or mitigation for the power issues we are sure to have to face. Another example of this is people saying things like "just drive less".
<sarcasm> Sure. Of course. Why didn't I think of that? </sarcasm>
You seem to want to denigrate other people, then put words in their mouth and insult them for what you imagined. Then arrogantly insult other people's ability to perform basic logic, but when your straw man is pointed out you cry victim.
Your numbers are only as good as your assumptions, and when you keep insisting on ridiculous assumptions like people treating EVs exactly like ICEs in spite of there being cheaper, more convenient options that don't lead to the conclusion you decided on before you started you will get ridiculous numbers.
Driving less is similarly an option to contribute. People do it everywhere the infrastructure lets them.
I have ignored your attacks and continued to provide data and enough math to understand my perspective (and refute it with math and physics if desired). I have used articles you provided to show I am likely right. I am more than open to dialogue and a discussion on any of this. I am more than open to being wrong. That cannot happen if you refuse to engage in conversation.
Rather than repeat myself, here's the parallel thread where I use your own article to, multiple times, show you that your perspective might not be aligned with reality:
https://news.ycombinator.com/item?id=35131368
You've been spoon fed the answer over and over, but you still do things like pretend the only options are 100% wind or 100% solar, and pretend there is zero flexibility in load. Try adding existing hydro, existing flexible loads and w2e to the highlighted parts of the paper you're reading rather than only looking at the corners of the graphs.
Or try doing the actual math with actual reality which isn't the bizarro absolute world of a single generation technology where EV owners prefer to charge at precisely 5pm, pay 10x as much and sit around for 20 minutes rather than plugging in while parked you've invented.
Then try not referring to an offhand comment by a serial liar as some kind of argument from authority.
Or try having a single shred of integrity or honesty or respect for anyone around you.
Sometimes I get the impression that despite all the futurism talk most tech people's solution to a trash fire is to say nothing can be done about it and just go back to the line of people throwing more trash into the fire.
Sure, you can blame the 3-ton SUV proliferation on the car-centric society Americans have created for themselves, but that isn't going to change anytime soon, unless Americans are forced to change it somehow. Sure, a lot of Americans complain about it, some of them even try to do something about it, but it isn't enough, because most other Americans actually like it that way and work against their efforts. Even the ones who complain mostly don't understand that it's a product of their own choices and how much they're contributing to the problem personally.
To be fair, the sparseness and geographic distribution of population, business and industrial centers in a great deal of the US is such that driving a car quickly becomes the only sensible option. A large SUV, sure, that's a valid question. Some of it is also driven by geography/topology. These cities did not evolve with mass transport as an important option.
To give you an idea of what this means in practice. I can get to the nearest major airport in about 30 minutes by car (at night, no traffic). If I want to use mass transit, it will take a minimum of 3 hours, only if you time it perfectly. It could take you six hours.
It's worse than that. Here's what I would have to do:
Remember that, these days, you have to be at the airport two to three hours before your flight.Now add weather, heat, rain, carrying luggage, perhaps travelling with kids, etc. I'm sorry, even the most ardent environmentalist is going to just throw everything into a large SUV or minivan and just get there.
This reality will not change. These cities are simply not designed for what exists in other parts of the world. The only hope we have is a future where shared self-driving robot cars can make a difference. I can't see anything else that might change this reality.
Fuel demand doesn't vary much by week or by month, so only diurnal storage is needed.
Just plug most of the vehicles in when parked, and then duplicate or separate 1 day worth of battery (5-20% extra) and put it at the charger or generator for the rest.
Suddenly your 1200GW 'requirement' shrinks by 90%.
It seems we'd only expect power use to go down when autonomous driving applications are "feature complete" and stop having a use for additional compute power.
In the meantime, discussions of limits based on achievable/acceptable power draw (a la phones which I would guess have been in a consistent range for years now) are more likely to be meaningful.
It’s by no means ridiculous to assume that this might happen here too, and in fact publishing analyses that make the assumption might be exactly what helps make conservation a priority and breaks those same assumptions.
Power consumption has become a serious problem with auto "infotainment" systems. There are cars running Unreal Engine in the dashboard. There's a noticeable reduction in EV range.
(Also a tendency to have too much electronics running while the vehicle is parked. Teslas are noted for this. There's "sentry mode", and "summon mode", and WiFi, and remote updates, and telemetry, and apps phoning home...[1] This can be a worse problem than power consumption for the electronics while driving, since most vehicles are driven a small number of hours per day.)
[1] https://teslamotorsclub.com/tmc/threads/battery-drain-while-...
Hot damn, didn't knew modern Tesla have so much computer power it could easily run VR in glorious 4k/120Hz
...or, apparently the article is full of shit and just throwing random numbers at the wall.
In Europe, we have (IMO stupid) rules not to have rear ligts on at daytime , to save power (<10W I'd guess) when daytime running lights are used.
That’s another problem with modern cars: moving two tons of metal and plastic to, oftentimes, move a single human around is a tremendous waste of energy.
That an increase in energy usage is relatively small compared to an already huge waste of energy isn’t a strong argument to declare it insignificant.
It's been an observation with self-driving container transporters at ports that the self-driving ones use less energy, and get about twice the tire life, than the human driven ones. The self-driving ones are dull, boring, smooth drivers. Which is just fine for container ports.
As a point of reference, the input power connector on Tesla's new HW4 computer is rated for 10A @ 16v, or 160 watts.
In fairness to the authors, HW4 was not yet released when this paper was published.
It still isn’t a lot compared to moving the car but it is quite a bit of power.
For a car with 100kw/h battery pack and a range of 500km a 1kw driving computer would reduce the range by about 5% which is about as much as having the air conditioner on.
And because they are autonomous cars, they will get utilized a lot more meaning we need less of them. People can only drive cars that they own so much. An autonomous car can drive itself, all the time. When it is not driving, it's losing money. So, the most economical thing to do with autonomous car is to utilize the hell out of them and have them driving around people and stuff 24/7 and earning money. Basically, uber without the drivers, tips, etc.
So, we gain some more energy in terms of manufacturing efficiencies because we need less vehicles to transport people and goods the same distance.
And of course the vast majority of the power used for this is going to come from cheap, clean, renewable sources instead of 100% of it being fossil fuels being transformed into CO2 that is blasted into the atmosphere via the tail pipe.
So yes, there is a bit of power involved. But vastly less than what we are spending on transport today. And it will be generated in a vastly cleaner way.