The "waiting 2-3 years until the incremental power has 0 marginal cost" makes sense in theory, but in practice, things get built when they are net positive or better. Waiting 2-3 years for lower costs only makes sense if the value generated during those 2-3 years would be net negative vs waiting.
These projects are planned years out. That is, the 2-3 years I"m talking about will automatically occur. That is, the project costs will always lag the current market prices by 2-3 years at least.
@@thelimitingfactor I believe you could get price reduction in batteries in those 2-3 years because they're also being increasingly implemented in cars, but you wouldn't get that in solar PV. Wright's Law applies for an increasing production (hence, increased market) but if you don't have an increasing demand there is no noticeable price reduction. Only triggering the transition in a short time frame (2030) would accelerate production & price reductions.
Awesome video! - thanks for the reality check. I love what RethinkX are doing and your perspective is a step in making their high-concept view into a practical roadmap.
The grid interconnector 'gridlock' issue is an interesting problem. In the UK, the backlog for an interconnector can be as much as a decade. This did not matter much when new FF thermal power plants took ~10 years to build, but is much more problematical when renewable generation can be built out much faster. My take is that shuttered (say) coal fired power stations (the last coal generator in the UK closes this year) represents significant value due to their existing redundant grid interconnector. Every retired power station had a grid interconnector. Now, wind or solar farms might not be co-located beside a decommissioned thermal power plant, but building transmission lines to that site might be quicker than waiting for a new interconnector. Separately, building out a grid scale battery farm on the site may be a profitable solution for buffering and grid stabilisation.
Also, the inverter based renewables (wind and solar) require stabilisation. Massive flywheels in the old turbine halls make a cheap and efficient way to stabilise the grid frequency.
If you spread storage out around the grid, there's less peak demand on grid transmission. During lower demand periods, charge storage. Then use stored power locally. Abandoned fossil fuel and nuclear sites are likely great places for large scale storage installations. Not only are large transmission systems in place, the ground is often contaminated. Land used for storage, solar, or wind does not need to be cleaned up to residential standards.
@@dominicgoodwin1147 Although this would be a low cost solution, I am not sure it will be needed. Batteries are capable of providing "virtual inertia" to replace generators and flywheels.
@@dominicgoodwin1147 - The UK is using battery storage for renewable stabilisation (as well as curtailment mitigation) - with nearly 5GW installed already, a further 4GW due to come one stream shortly and 95GW in the pipeline for installation by 2030. There is one 400MW flywheel in the UK, but it serves a different role - as a massive peak power source for the JET fusion research facility near Oxford. Batteries are also being used to address the problem of grid congestion that in the past required gas peaker plants to switch on on the under supply of the congestion point and the power generators to be switched off on the over supply side. With both generating parties being paid higher spot rates to both generate and not generate power. Batteries provide an elegant and more cost effective solution than current apporach and mitigate the expensive cost of upgrading grid networks at congestion points that only periodically need to carry the extra power.
@@Travlinmo You ought to as he hasn't got a clue. What he thinks and what is possible are as far removed from each other as is North and South. Renewables have none of the criteria essential; for a stable, economic and reliable grid. Just how can weather dependant and uncontrollable sources of generation feed a grid that has to be minutely controlled on an instantaneous basis, its just nonsense.
You probably know this more than me, but from what I understand, cost of installation is higher than the cost of solar panels themselves, which means that timeline for replacing fossil fuels highly depends on cost of labor. If in next two years we will have automated robots for ~20k like most companies predict, it will likely decrease cost of solar by 2-3 times, which likely will very quickly outpace costs of burning fuel, as installation costs is big part of the capital expense needed for solar. This will also likely decrease cost of solar + batteries, which will increase demand for batteries and hopefully will allow for much faster ramp up of battery storage systems, especially if battery production will benefit from automated robots as well.
@@thelimitingfactorand at the same time there’s 7-8 TWh/a overcapcity of LFP cell production in China. It might be all that capacity can be diverted to BESS deployments. Big news coming on that front quite soon 😁
Very much true. For example rooftop solar in Switzerland didn't become cheaper dunrig the past 4 years in spite of the panel prices dropping was visible even though not as dramatic as elsewhere. Rooftop solar booming anyway due to increased grid energy prices motivating potential owners to start a project.
My brother has been designing and installing solar for off-grid users for 30 years. About 15 years ago he was talking about how panels had dropped in price so much that the (lead acid car) batteries were more expensive than the panels. The result is they were designing larger solar output and smaller battery system. The result was that the batteries might be recharged by noon and they were just dumping excess power for the rest of the day. (Not actually dumping power, but pulling out less power than the panels were capable of producing.)
I'm off the grid, have been for about 30 years. Many days my batteries are full by lunchtime. But I need that extra panel input during cloudy periods. It's cheaper, at this point in time, to overbuild generation than to add more storage. BTW, solar panels have fallen in price from ~$100/watt to ~$0.20/watt over the last 40 years. And they should continue to get cheaper.
I've seen a couple comments miss the point of Superpower. It isn't about being able to do things with "extra" power*, it's about the marginal cost of the power being used. By pairing installed generation with the worst case generation scenarios, you will have insanely cheap energy the rest of the year. How much more affordable can consumer products, agricultural products, etc. be made with energy that is basically free for 9 months (depending on latitude) out of the year? *Practically speaking, there is no such thing as excess power. That is because excess power being used means it is no longer excess and would counteract the reduction in marginal energy cost. The moment you try to do something with the excess all the benefits of superpower disappear.
Maybe it sould have been called excess capacity. Becuase that is what it will be most of the year. Since not all the power it could generate will be needed, the plants will be partially curtailed most of the year.
Solar (PV) modules are 6.1cent/W (€) right now, so a new 410W module is 25€, here at a local dealer, no taxes, only local shipping, amazing, 10 years ago I paid 36cents/W and that was considered cheap. Now, solar is nearly zero cost. Solar works for me at 53°N, >=75% of all electricity comes from my roof, all electric here of course. With a south facing solar array you can go 100% 24/7/365. Batteries are included :-) And that up here in the north
my country had tariffs on chinese modules 10 years ago, today no more, plus there is zero Vat this year (instead of 19%). That pushes solar and shows everybody here the true price. Solar rocks
Tesla should produce magnesium in the Gulf of Mexico (sea water) using super power. Gigacastings out of cheap magnesium. See Jordans video about tixo moulding.
Capital intense equipment dosn't work very well if it isn't used most of the day if not 24/7. This said in many instances it will be most cost effective to curtail the superpower.
Really enjoyed seeing Tony's recent work🥰and your perspective on it. Lots of new AI driven data centers are going to require lots more electricity....a great use for new renewable energy generation coming online before the US has time to mostly trade-up to EV's and home heat pumps.
Another flaw in the analysis is that it assumes the old model of a few large generators feeding many (smaller) loads. The old utility model. But the system is now more like the internet: Millions of nodes generating and consuming power at random times.. So the economic choices are made not just by utility companies, but by millions of individual node owners. The grid is like a network not a distributorship.
And we now have the ability for moment by moment pricing which can be used to move some demand off peak demand times, cutting the overall cost of electricity.
This is a great point and the power of rooftop solar and home storage is really underplayed. They don't need to be the cheapest form of energy for mass adoption, they just have to be cheaper than the market rate, to make sense for the consumer. This will then in turn completely change the form of the demand curve that the utilities have to cover.
I think that there are multiple standards for vehicle-to-grid (V2G) communication, including ISO 15118 and SAE J3068. Plus there are two UL standards that apply to V2G technologies: UL 1741 and UL 9741. The former applies to all types of generation equipment, while the latter applies only to vehicles and associated supply equipment. Someone needs to set out a universal V2G system standard, just as NACS is becoming a universal plug standard, accepted across America's charging network (including home and work charging and discharging). Battery chemistries continue to improve, and future EV owners won't be so concerned about battery degradation. An affordable, universal V2G standard needs to be developed and accepted, then the grid will be able to easily use BEVs themselves to become the buffer for renewable resources.
NACS itself isn't a universal standard, it's a design that is tied to the small number of countries that operate on split phase 110/120v domestic or combined single phase 240v electricity supplies. Whereas most of the world operates on 220/240V single and three phase supplies. This is why CCS2 (AC or AC/DC combo) is the standard in many parts of the world, plus China with its GB/T standard. The same applies to V2G as the electrical standards around the world differentiate the standards seeks to address these differences. Similarly, vehicles have a specific supply characterisitc of high power / short duration supply, whereas other techologies are addressing medium power / long duration needs that differing standards address. It is these differences in systems and uses that are driving supporting standards, as one size doesn't fit all in this case. What is positive is the small number of standards that apply meaning things are being tightly controlled to bring cost and implementation advantages that the alternative of a plethora of competing, often company/industry specific, standards would previal holding back adoption.
Another important RethinkX omission is lack of analysis on what industries can profitably use the excess power ("superpower") produced by overbuilding renewables. Industries that use lots of power like crypto mining, datacenters, and aluminum refining all use expensive capital assets that no company is going to be willing to have sitting idle half the time. I bet the capital equipment to produce hydrogen via electrolysis isn't cheap. And it's expensive to pay your workforce while your factory is shut down without power. With anything made by superpower, companies couldn't plan their production and promise delivery dates to customers. If there were good superpower industries waiting to go, why wouldn't we already be seeing them since solar in California and Australia already overproduce in the spring?
Some possibilities (not necessarily good possibilities, but possibilities nonetheless) that come to mind for intermittent excess power: (1) desalination of water; (2) produce green hydrogen, and then combine that green hydrogen with CO2 to produce methane to be used as fuel in SpaceX flights; (3) pump water uphill and back into a reservoir to be reused in hydroelectricity generation; (4) generate heat that is pumped back into the ground for later extraction via a ground-source heat pump; (5) extract CO2 from the atmosphere. The possibilities I suggest are wild guesses, but they have have something in common. They suggest that superpower may not be soaked up by already-existing-and-widespread uses, but rather that intermittent superpower may be soaked up by not-yet-widespread uses, simply because the benefits of superpower's low cost will outweigh the drawbacks of superpower's intermittent nature (in contrast, already large consumers of electricity typically would see intermittent power as being a deal breaker).
@@thelimitingfactor "There's only so much of a market for superpower"... today... When things get cheap, industries get founded to take advantage of low costs.
@@CiaranMcHale The last one is the best one. When power is free, convert CO2 to elemental carbon or some way to sequester it. The reality is when we get to zero emissions, we'll need to clean up some of the excess that we added.
@@thelimitingfactor Other possibilities, most metal plants could produce a good deal more and could go on overdrive when electricity pricing is lower. AI training centers and digital coin centers could boost computation during times when energy production is highest. They probably could do this now with a well coordinated grid. By the sounds of what OpenAI would like to do, they want 5-7 5 GW plants. they could soak up a lot of power during the day
Great video, There are two concepts that Rethink X talked about that you did not touch on: 1) Capacity Factor: Nuclear capacity factor will start dropping when solar saturates the market in autumn and spring. This will push up the cost of running existing nuclear and gas on a kWh basis. 2) Distribution Parity: It will be cheaper to go Solar, Wind and Battery for individual properties/communities than what it will cost to transmit power from central generation facilities. There are cases where the business case for grid connection failed vs on site solar and storage.
The fuel cost for nuclear is pretty low. Nuclear is not very dispatchable, so unless there's going to be periods of weeks and months where the generation is needed it's likely they will stay in operation. Plus if reactors are curtailed for a significant period of time it spreads their fixed costs over far fewer MWh of electricity sales. It's likely the government will find a way to make sure reactors get to sell their goods. Coal and gas will get turned off.
We need to account for the EROEI for the overall systems, That is, solar, wind and battery, compared to mass produced advanced little nuclear reactors. Which requires more materials to generate a constant reliable watt? Which requires more land? And, on the plus or minus side (I'm not sure) which will _cost more energy to also_ recycle?
@@fireofenergy Another consideration is that nuclear tech is also advancing, and molten salt reactors can use existing spent fuel AS fuel. Considering how much is currently being paid to store all that waste, it could conceivable cost negative money to operate these new reactors.
@@fireofenergy First, we can't make a comparison with "advanced little nuclear reactors". They have not yet been built and evaluated. What we have is the nuclear industry telling us again that 'We have a new idea and this time it will make nuclear affordable". They've told us that over and over and over for more than 50 years and have yet to make it come true. Low credibility. Maybe this new design will be cheaper. The Union of Concerned Scientists did an analysis and found that smaller may be more expensive. Time will tell. But, remember, new nuclear is around $0.15/kWh and needs to come under $0.04/kWh to be acceptable to the grid. That's about a 3/4 cost drop. Pretty hard to imagine. We've been closing paid off reactors that need more than $0.04/kWh to stay in business. The market won't buy their power. Now, you talk about the overall system. Fine. Let's talk about the overall system for nuclear. First, like wind and solar, nuclear needs storage. Storage is used to move supply from low demand periods to high demand periods. Back when we were building nuclear we built a lot (over 100 just in the US) pump-up hydro energy storage plants. Add that cost in. Then, there's the need for backup generation for when reactors "break" and go offline. Happens and happens more than most think. When a reactor goes offline it can be off for days, weeks, and even more than a year. The grid has to have excess generation standing by to replace one or more reactors. When both SONGs reactors in SoCal went down the grid wasn't prepared for that large loss and had severe supply problems. That's another cost to be added into the overall nuclear system. Hint: A 100% 24/365 reliable wind, solar, and storage system is cheaper than nuclear, storage, and backup. Considerably cheaper.
Good job. I do this type of analysis myself and came to the same conclusion. In my models the cheapest energy today includes solar/wind/batteries/pumped hydro/hydro with gas peakers. As costs come down for batteries/solar, you can add more and more “super-power” to offset peaker use .. so these peakers get used less and less and less. In effect superpower is a type of “peaker” .. so IMHO that is how we can easily assess its cost effectiveness.
See, now THAT makes sense. You could actually extend [I assume] the lives of peaker plants in this way, making the grid more robust as renewables come online.
@@jamesengland7461 Is kind of the definition of super power. And we already have some small measure of it. Renewable heavy regions sometimes go into negative power prices.
Peakers need to be dispachable. I don't think superpower from excess wind and solar is suitable. Batteries can replace speakers at lower cost, so I expect that to happen instead.
Tony Seba is selling hype via theoretical extrapolations of cost curves and so on, as if they don't have any unforeseen bottlenecks or don't need a practical buildout of supporting industry and supply chain. Great work adding some of such practical considerations and showing more realistic version of the future. Much needed dose of skepticism.
Thank you for the thoughtful reflection on the RethinkX presentation. Two questions: 1) Do the energy demand assumptions used account for new high demand categories (eg. AI compute) that would scale demand above the current trend line? We’ve seen with computers that as processing speed, memory capacity, storage capacity, and connection bandwidth became cheaper and more available that demand for each was induced at a rate that many people found surprising. 2) In terms of rate of transition and buildout, is there any accounting of the coming reality check on deficits and debt loads? You can run cost benefit analyses till you’re blue in the face, but if you can’t finance the debt (eg. corporate or gov’t bonds) you don’t get to buy the equip or do the work.
Large AI model training is an interesting case for super power. Unlike many other use cases, interruptions are not a big deal, you just continue from the last checkpoint. The training runs are also reaching the point where they're talking about building nuclear reactors just to power them.
by the way solar cell prices (not panels) fallen down to single digit cents... makes 10 usd for a 2 m2 panel which gives out 400 to 420 w/hrs... for meeting a household's electricity need, let's say 10 kw/hrs capacity, core tech just costs less then 300 usd of investment... rest of the cost are cables, glass, films, frame, and workmanship... there are design options to squeeze those down too and it will eventually get down...
"The average useful life of a combined cycle natural gas plant is typically considered to be 25 to 30 years. However, with proper maintenance and potential component upgrades, some plants can operate for up to 40 years or even longer." The average useful life of US coal plants is 40 years. Almost all our coal plants are older than 30 years. Some of our CCNG plants are likely over 30 years old, the average age is about 20 years. All these plants will have to be replaced with something. Wind and solar have vastly lower installed cost per MWh generated than do coal, nuclear, and CCNG. Additionally, since wind and solar are cheaper than buying fuel for a paid off CCNG plant as more wind and solar are added they will be given priority over natural gas in order to enjoy fuel savings. That means that when CCNG plants have a serious repair issue a utility might decide it's time to retire that plant early and spend the money on more wind/solar or storage.
You made a good point about electric power needing less input because about 70% of the fossil fuel input is consumed by the process before it ever go to work. Two other factors will have a big impact. The second and third cars that we typically have sitting in the driveway at most homes not only might utilize the super power or even be charged with the super power we usually have with our solar panels at home in the day time. They might even feed the grid,in effect replacing some of the battery expansion for grid level backup. The other factor is that many of us are already weaning ourselves off the grid with home solar panels and batteries. When the sodium batteries get into mass production, price and availability of LFP should improve also.
Wow! Lots of factors affecting existing energy grids. I see it as a dynamic energy ecosystem. 1. Some assumptions in Wrights rule regarding a fixed decrease in cost over time deploying sustainble energy. Perhaps a greater cost decline would occur beyond the typical 'cost of material and production volume/scale' due to improved battery design (increase in storage capacity) and (ongoing) decrease in mining costs (supported by recycling battery materials, etc) 2. California building codes for new homes require solar panels. Say this transition takes place in other states over a few years. While adopting this new building code (along with battery cost improvements), it would be cost affective to have the federal government to provide financing to home owners to install battery storage devices. Battery storage in homes would provide a primary layer of energy source to buffer any major power outage that would be eventually covered by the local energy utility's storage system. 3. Meanwhile, continued conversions of existing homes to (newer) heat pump technology reducing energy consumption and costs.
Excellent video! If Wright's law is any indication (20% cost reduction for each cumulative doubling of total battery production), cost will likely sink slower and slower in time.
Critically the Lazard report also shows the very high cost of power firming. Casio has much higher electrical costs despite higher renewable penetration and lower generation cost. Transmission, capacity firming, and distribution are all large costs than generation. Overhauling the grid to hand intermittent load and moving that power is costly. The power queue as you mentioned is already extreme. More than 8 fold increase in the last 10 years to more than double current generation is in the power queue and time delay has almost tripled. Also the cost for energy storage assumes 4hr of capacity. With grid connection and lots of base load supply this works, but multi day let alone multi week storage is highly cost prohibitive to say nothing of seasonal variation. My current project is on evaluation of space solar power which can provide highly competitive generation, transformative logistics, and 24/7 renewable base load. It serves as a perfect complement to variable renewable sources with the capture to beam power to multiple different ground locations as needed.
I think you are missing the Micro Grids and Virtual Grids options in your analysis. I feel by 2030 there WILL be many homes producing surplus energies from newer/ higher energy dense batteries and more efficient tandem perovskite solar panels. Also, most of the EVs on the roads will have bidirectional capabilities. This would drive Micro Grids and Virtual Grids. At least that is what I feel like. The challenge is going to be building smart grids that can take care of these growths.
One thing you touched on but didn't expand is the energy companies wanting to maintain their market shares. This pushes into the political arena such as the Ohio nuclear plant bribery of government officials. Depending on where you read, oil companies make up to 6 million dollars per hour. That's 24 hours/day year round. You can bet they will not die quietly. As to your point on waiting a couple years for prices to decrease even more, I believe that the sooner we can eliminate the need for fossil fuels the better even if it costs a few dollars more. However I do agree that it is unlikely that we will reach full decarbonization by 2030. There are too many obstacles and too many people who either are afraid of change or just flat don't want change in their lives even if it makes their lives better. Good job dissecting Mr. Seba's forecast.
Limiting factor - no you want to charge grid battery & commercial & power walls @ peak solar, homes will need to charge overnight @ super off peak. they can still export to local grid , after battery is charged (4H), and discharge is up to (6H).
Here in Sweden I get more or less 0 kWh solar for 4 months when it also is super cold. I would need a lot of super power batteries to cover 4 super cold months. We need to be nuanced, this is not a solution for all.
US Dollar value is no longer pegged to Gold, it’s currently pegged to the price of Oil. What happens to the Dollar when Oil is of little value? Does that make Lithium or Solar Panels the new Oil? There is no shortage of Lithium or the materials necessary to produce Photovoltaic Cells…
Has regulations changed from 2010 when solar and wind power were subsidized by nuclear, gas and coal power by requiring utilities to assume the green energy was being provided 24/7 everyday of the year in the balance books. This requires that the extra cost of going green would be shouldered by nuclear, gas and coal. The income from nuclear, gas and coal would have to have their cost increased to offset the difference in the real cost of green energy and the subsidized cost on the utilities books. This started in Europe and came to the US decades ago. Very few really understood this requirement for the utilities to do this off-setting of the true cost of green energy. I worked in the utility industry for the last 45 years. The more coal, nuclear and gas power plants are shut down due to the increasing off-set to dirty energy into a smaller and smaller amount will keep showing dirty energy is getting more and more expensive. This will essentially stop this off-setting cost burden since the less dirty energy is provided there will be less ability to subsidize green energy; the real cost of going green will appear.
Provided the predictions regarding cost erosion are going to become a reality, then the true cost of renewables including storage will be even lower than today. Subsidies will not be required anymore, maybe with the exception of the last few plants to be built for the super power, when it needs to be financed in spite of the uncertainty how much in 1 year it will be used.
Thanks, does your analysis consider domestic energy generation from residential rooftop solar? There is normally not much regulation stopping households installation a system which mostly covers there needs, which overtime could reduce the need for utility and grid build out?
This confirms my confusion… I don’t understand why Tesla isn’t doubling down on Megapack. Seems like they’re overdue to break ground on a few more Megapack factories. Per an older video of yours, there’s easily going to be demand for 10-20 more Megapack factories this decade yet I do not see this happening. At this rate, we’ll probably go into 2030 with Lathrop and Shanghai only. There has to be a limited factor that I’m missing. Tesla has the cash. Demand is/will be there.
Excellent analysis, thanks. I figure superpower can happen much sooner than you think in super renewable energy resource States like Texas and California and many in the midwest. Then their grid can export the excess to neighboring states. Wait till geothermal Drilling gets to become mainstream then your base load power problem goes away and fossil fuel plants will disappear quickly.
@@bobwallace9753 that is rapidly changing. Search RUclips for the new plasma laser drilling technology that is almost ready for prime time that can vaporize holes deep quickly and cheaply into the Earth's crust leaving a glass tube for the geothermal energy flow. It is being pilot tested and developed as we speak...
Intelligent control systems allow demand management, so meeting the energy they need should be an interaction between producers, grids and consumers. Octopus Energy in the uk is already doing this with intelligent tariffs, as is Tesla of course.
Advantages both ways. Scale economics of utility production & storage is obvious, more so as panels drop to almost free while re-wiring the house costs. Present use case is immense GWh scale battery farms in outer suburbs.
Good analysis. I have a couple of thoughts. 1) I agree that the multitude of decision makers and entities in a market and regulated monopolistic market make super rapid change more difficult thus arguing for 2040 versus 2030 - at least from a directional perspective. 2) I think you second point regarding waiting for costs to drop further to lessen overall cost is weaker. It depends on how you factor the externality cost of the exponential impact of further GHG emissions and the potential for catastrophic non-reversible exponential climate change. There are solid scientific reasons to view a delay of even a decade as a real and potentially incalculably expensive risk. In that case, governments might take a more 'central planning' and less cost efficient approach just to avoid the existential risks of a delay in achieving net zero (and net negative because we need power to power the carbon sucking machines)....
Thanks man! As for #2, that's not a weakness in my analysis, that's an unpriced negative externality that I can't control for in my analysis I'm not here to tell society what it should do, only what the likely path is based on what we know right now
V2G will be a great part of balancing the grid. Over production will be stored in hydrogen and other similar storage options. A stable grid with great backup is the way to get low price. Small nuclear plants that easily can be turned off and on is another base production that needs to come in place.
@@markplott4820 My BEV has LFP batteries and rated for 6000 cycles compared to NCM that is rated 3000 cycles. But i do agree a small LFP home battery as a complete is also good when the BEV is not home. That is the setup I use here.
@@beatreuteler - V2H is a GREAT way to burn out your very EXPENSIVE vehicle battery. (+$30,000).....lol. you are BETTER of with a HOME battery like Powerwalls , even w/o Solar.
@@markplott4820 Home batteries are too expensive for the time being and it is quite well known that there is no detectable damage to a vehicles battery life when using VTH or even VTG. The only reason not to do it is if the benefit isn't there or if the manufacturer makes limitations through warranty clauses. Where I live, it doesn't make economic sense due to regulations causing extra cost. That's eating up all the possible benefits which makes it a nogo for a very long time. But in other countries that can do a good contribution to grid stability.
One obvious solution to excess renewable energy is desalination and pumping water uphill. In the western USA we are horribly short on water. Why waste all that extra energy when it could so easily be put to good use?
@@thelimitingfactor While this may be correct, in some cases, if such a project is of mutual use far beyond the energy supply, it may still be justified. However if that is of mutual benefit, it's just a matter of convincing everyone so you can have it supported financially by the local authorities. What the idea doesn't cover, however, is the fact such convincing and planning may take well more than 6 years in most communities.
Don't forget you will have to install a lot of solar or wind to recharge the batteries to supply the electric loads when the wind or solar is not available. That is not free energy!
Overbuilding solar and wind is the right idea and Toni's projection is likely to be close. China is doing this now and they are actively managing excess solar PV curtailment with grid interactive customers (demand response) i.e. the smart grid
Overbuilding wind and solar saves on storage cost. And, to some extent, that extra generation can find a use. Charging EVs is an obvious use since EV charging is highly dispatchable. Another is desal. Lots of salt water could be turned into potable water during sunny mornings and windy nights with the fresh water stored in reservoirs for later use.
RethinkX was wrong but in a good way - just come to Australia where Superpower is already here! Between 10am and 3pm on most days including winter, the National Electricity Market spot price for generators is NEGATIVE, in other words, baseload power costs the generator up to A$50 per MwH to put it into the national electricity grid. Renewables (mainly solar and wind farms) are being curtailed. In Victoria, retailers are offering FREE electricity between 12pm and 2pm and you can get 50% off standard full-day rates between 10am and 3pm. In South Australia those free windows are wider. The challenge has been a) getting consumers to soak up the excess in the short term and b) building storage and interconnect capacity in the medium to long term. Believe!
The notion of storing 90 hours of energy use in batteries at an affordable cost is ludicrous. It depends on a) Only converting the existing grid which is only about 20% of total energy use i.e. solving only 1/5 of the problem b) No economic or populations growth c) Fanciful assumptions about a huge acceleration in the rate of improvement of battery production. Remember that batteries have only gotten about 7%/pa better for 130 years. The easy wins are past us.
Jordan, thanks for this thought provoking piece. I agree with your conclusion that Seba is overly optimistic in terms of getting to 100%. That said, Seba has continued to surprise and I think we will do better than your at projecting. Here are some things that push the timeline in his favor. 1) Grid upgrades and interconnecting the three US grids would lower the curve by aggregating the variable renewables making their aggregate output smoother. The SEIA timeline plot you showed was from beginning to end. But there are a number of transmission line projects that are far along in their timelines so could be built much more quickly. 2) The old model is to provide all power requested WHEN it is requested. Many customers would happily alter their demand curves if they could recoup the economic benefits of that. This requires time of use metering. Tesla is addressing this on the supply side with power walls being aggregated into virtual utilities to arbitrage power. But that is the Seba model. It can also be done on the demand side and BEVs provide an excellent way for the grid to gain storage essentially for free. I, and many others already do this by charging at night when the grid in my area has more power available. 3) You use an example of solar only being available in the daytime. Aggregating the grid would increase solar availability by about 3 hours (3 time zones) and wind peaks at night making the window of low production much smaller and requiring less battery capacity. Further, wind turbines have continued to grow in height, this increases capacity factor (on time). Massive wind farms being built of the eastern seaboard are looking like 50% - 60% CF which is far in excess of what most aggregate models have utilized. I agree with your overall premise that there are constraints that make 2030 impossible, I think the middle of next decade will see us at least 90% of the way there for transportation and electric with the exception being HVAC which will take considerably longer. I did a deep dive into this stuff about 10 years ago. My conclusions were pretty radical with prognostication that we could get to zero carbon by 2050. I couple of years later I stumbled onto Seba's book, and agreed with his methodology but thought he was over optimistic by decades. NOBODY thought he was even close on BEVs. Here we are 10 years later and his estimates were closest to correct. I still don't think 5 years, but 10, I won't bet against that. PS - Most of my modeling is conceptually similar for Europe as they have access to massive solar in Spain and across the Mediteranian, lots of hydro storage in Norway and massive wind power off shore - Transmission lines the access to wind power handlemanpost.wordpress.com/2014/01/15/compare-maps-of-the-grid-and-renewables/ - Wind Power CF for onshore and off shore is even better handlemanpost.wordpress.com/2015/07/26/new-nrel-numbers-game-changer-for-wind-power/
Limiting factor - keep in mind, we dont need any Nickel cells for home battery, powepack or Megapack storage battery. All can be built w/ LFP instead of expensive Ni cells. and will last 50+ years, and can be charged to 100% , discharge to 1%. unlike Ni cells. we will need more LFP battery factory 🏭.
Limiting factor - super power is defined 2 ways #1 - enough renewable energy is produced to meet 100% of industry & consumer. #2 - super low cost renewable energy in mass quantity, net ZERO cost.
There's a feed back loop. Bumping the time will reduce the rate of increase in economics of scale, likely making it necessary to bump it further... and so on.
7:30 This mismash of operators and regulators only exists in the US. For many places in the world the electrical grid is government owned and operated. Like in Canada. Tony is not looking only at US power grids. Australia,for example,is already going through this transition from coal to solar. Skipping gas and nuclear completely.
Have to realize battery tech. has been suppressed for 140 years, so you have to give them a break. Soon, there will be a battery, that takes all night to charge, but will run all day, just not there yet, things must evolve. For now, ANY kind of power plant will do, once greener tech. catches up, then the older means can be retired, right now, it is about enough power, no matter where it comes from.
Limiting factor - home solar & battery are under used, same for commercial & industry. also wind. most farmer in usa under use SWB, too. Disruption potential is everywhere.
You made a good point about electric power needing less input because about 70% of the fossil fuel input is consumed by the process before it ever go to work. Two other factors will have a big impact. The second and third cars that we typically have sitting in the driveway at most homes not only might utilize the super power or even be charged with the super power we usually have with our solar panels at home in the day time. They might even feed the grid,in effect replacing some of the battery expansion for grid level backup. The other factor is that many of us are already weaning ourselves off the grid with home solar panels and batteries. When the sodium batteries get into mass production, price and availability of LFP should improve also. Even renewable energy has a big percentage of loss in transmission. Distributed generation and storage mitigates that.
Could you comment on the requirement to replace capacity at the end of its useful life please? I do not know if Rethink X has included these costs including cost of disposal / re-use
Nuclear power is not economically "dispatchable". Capital cost are so high that the plants need to stay at full power for as long as possible. After decades in that business, I am not optimistic for new nuclear generation.
I believe that Rethink went with the current electrical power because it is not dependent on any hardware changes, meaning no one has to change their gas stoves, hot water heaters, furnaces, cars, semi trucks, industrial furnaces, and any other hardware which uses direct fossil fuels and has a long lifespan. Changing from fossil fuel hardware to electric hardware will take place gradually overtime as people and businesses need to replace the appliance or machine. Even if Texas did wait for 2040 to replace the entire grid because it would be so cheap, many people could not use this power because there has not been the gradual change I spoke of above. In addition, the projected increase in energy demand needs to be met in the moment, not years later. So its either build more fossil fueled energy plants which are expensive to build and operate and would be a poor long term investment over its multi-decade life or go forward building the solar and battery capacity as electrical energy demand incrementally increases.
With the coming wave of AI data centers and their huge requirements for power, could it be AI buildouts dwarfs the infrastructure's ability to produce and deliver power regardless of the generation source? Perhaps the crux of the issue in 2030 may be how much of a nation's electrical capacity is directed to AI and how much is allocated to the rest of industrial, commercial and residential power needs.
RethinkX is right that low-cost economics will eventually win out. Frustratingly, they don't attempt analysis of what the rate-limited step, or dare I say what "The Limiting Factor," will be. This is critical to projecting the timeframe, as you point out. Just look at the grid interconnect queue as an example, with many new solar projects in OH now not even attempted following legislative changes that make new renewable generation projects easy to stop while new fossil projects continue to be hard to stop.
As a general question. Are Tony Seba and Adam .. talking about Texas (or maybe the US) or about the whole world? On whats defined as intermittency challenge, for South Germany you have around 2-3 months where there is not enough solar power produced. Thats for 48th parallel north. I am sure its similar in the US and other places. Well there is wind energy. But they have higher running expenses (higher marginal costs) and because you can't plaster the whole country but also environmental issues (there is more and more microplastic in the air, birds, microclimate changes) it won't help that much. Australia could make it mainly through solar. Looking to places like Canada and Scandiavia its more realistic that there are 3-5 months where there is a problem. It would be interesting to see whats going on on Denmark. So the intermittency problem can only go away when having fossil plants (nuclear is not enough) running for 1-4 months. It must not be bad. Having a battery storage for 30-100 hours makes sense. But as you demonstrated, its not even possible for Texas. Also i expect worldwide in 2030 maybe in average 4-8 hours can be stored. We will be surprised how quick the capitalistic world can increase the capacity but thats an extremely high demand. Whats missed (but right now i don't see much activities) is high-voltage direct current (HVDC) between continents. From north to south (solves the summer winter issue) and from west to east (solves the daytime issue). Through the Atlantic yes. But the Pacific seems to large. It case it is economically feasable it will take min 10-20 years. Also several nations will keep their (nuclear) power plants running for the coming 20 years. You can't easily regulate or switch them off for a few days. I agree on your points that a lot will be possible until 2035 or 2040. But it will mean at this time nearly all the current fossil power plants will be worthless. Maybe i have some errors. Later i will look again on the video. And the Tesla presentation you mentioned. Its good to think about these questions.
They use several case studies from different markets 🤠 The result is slightly different for different markets, but the overall gist is the same. But, there are wildcards here such as the massive increase in demand from AI, etc.
Another person who has modeled 100% renewables is Mark Jacobson of Stanford University. Here is a pdf for Germany Completion for 100% is 2050, but 80% by 2030 web.stanford.edu/group/efmh/jacobson/Articles/I/143Country/20-WWS-Germany.pdf
As I understand Tony Seba's vision, installing 2x-3x the solar PV required in summer will provide the needed power output in winter. In summer the extra capacity can be curtailed. This is really no different than how fossil power plants are curtailed when demand is not there to use their full capacity. Batteries can cover nighttime demand, and cloudy days are covered by a combination of excess PV capacity and batteries.
Hey Jordan, remember when Elon said that the entire US could be powered by a postage stamp of solar panels that are approximately 100 miles X 100 miles? That's about 10,000 miles, now imagine Solar Fencing along every Interstate Highay in the US. The US has about 70 Interstate highways which total approximately just under 50,000 miles. Solar Panels can be installed as a fence with Solar Panels on both sides with both sides absorbing solar power & turning that into electricity. And obviously Solar Fencing would be installed on both sides of the highway. 50,000 miles of Interstate Highways with a double sided Solar Fencing is approximately 100,000 miles & Solar Fencing on both sides of Interstate Highways is approximately 200,000 miles of Solar Fencing which is 20 times more electrical power generation than Elon said was necessary to power the entire US. What is great about Solar Fencing is that it's easy to install, it doesn't take up any new land & is along Interstate Highways which is already owned by governments so they don't need agreements with anyone other then maybe State Governments. Because Interstate Highways run through every State it means every state is producing Solar energy & along with Mega-Packs can store as much energy as they want. Solar Fencing & Battery Storage can be scaled as they can afford it. And because Interstate Highways run through every state electricity can be transmitted everywhere by building transmission lines within the Solar Fence or by placing transformers, Battery Storage & transmission lines in Boring Tunnels under the Solar Fencing & away from environmental disasters. Most Interstate Highways have mediums separating the two directions & if they wanted to create even more energy they could install Wind Turbine Fencing down the medium of these highways doubling the power generation again & producing energy 24/7 from wind & solar. They could also place small wind turbines on every fence post again producing more power. Solar Fencing IMO is the answer to electrical generation to power the world & we don't have to wait they could start doing this now. Here is a crazy idea, what if all 50 states offered one billion dollars per yr for 5 yrs to install Solar Fencing in their state & the Federal Government matched it dollar for dollar. That would be 100 billion dollars per yr for five yrs & would cover the entire country. Harris the Broadband Czar just blew 50 billion dollars with zero results so finding 50 billion is not that big a deal for the Feds. What do you think of that?
Thank you, was thinking the same thing. Energy is not only electricity. Decoupling environmental damage from human activity is likely to be found in space applications. The only way to power heavy industry in space is nuclear(fission or fusion), there just isn't anything that comes close.
It knocks me out that they are refurbing nuclear reactors to power AI data centers. If plants that are near EOL are taken off the grid by private investment, instead of being extended like Diablo Canyon, could that cause more SWB build out? (I have no idea, just thinking about it all...)
We already have curtailment in 2024, just because we don't know how to deal with excess. It will get worse. Until someone would figure out how to profit with fully automated plants. Might be quemical energy storage (sort of peaker), H2 refinery, decarbonization, or desalinization. Will get there regardless of timelines.
Even if batteries drop to 1/10 of their current price per kWh, a solar/wind and battery infrastructure would cost 5X (AT LEAST) what the current gas, coal, nuclear base load system costs.
@@thelimitingfactor I think tycurtin is referring to the fact in this concept you need a minimum of 4x the nominal capacity which certainly by now is still almost 4x more costly than grid parity. This said, this would be my hint to tycurtin: Please go back to minute 7:40 to look at the graph once more. The cost graphs also for PV and Wind (not only for batteries) are logarithmic. The projected cost reductions are foreseen to keep it at parity throughout the timeline, e.g. in 2030 it is foreseen to be 4x to 5x more cost effective as the competing sources, making up for your sorrow. @thelimitingfactor, please correct me if I'm wrong.
@@waltermcphee3787 The only high cost for nuclear is the red tape of US rules and regulations designed to make nuclear practically impossible. These were the result of 3 Mile Island and psyops like "The China Syndrome." The footprint and maintenance of solar and wind is immense. As energy sources, solar and wind are so diffuse they take up vast regions compared to traditional power plants. Not only is it unsightly, but it takes huge resources to maintain. They just don't have very long lifespans either. This is all of course excluding the batteries required to make them equivalent to gas/nuclear. That's a whole other story that makes solar./wind economically unviable. Take for example, if there were no solar/wind subsidies from the government, how many solar and wind farms would be built??? NONE
Meh, it's going to continue to accelerate faster than any predictions. Tony Seba was spot on with EVs a decade before it happened. *Another* 10 years is a super long time. Maybe it's a 2-3 year delay, but I don't see 10.
I'm just repeating what their analysis says, but highlighted the caveats that they didn't or people missed. That is, where have I misrepresented their research?
Limiting factor - in CA as case study, CA generally overproduce solar & wind , but are lacking enough battery storage in State, so pge & others waste it by exporting to nearby states, for a profit. they can fix this in short time, by adding more battery storage. and in CA hydro, geo are maxed out. Solar thermal is not working. CA has not explored off shore wind or tidal. IRELAND is 100% tidal turbine power.
I have been a big fan of this concept for many years. It describes what *could* be done in 2030 by the engineers, if they were 'king of the world'. They focus on 2030 because we need decarbonization ASAP. Your take on the concept is correct, focusing on the practical issues in the making. And there is another issue: We don't have too many use cases for 'superpower' (yet). Our industrial processes are usually designed to run 24/7 and most of them do not like to be interrupted at all ;) So we will have to invent new processes that can be easily turned on/off.
Australia solar already being curtailed. Utilities are turning off. Up to 50% already. Else feed in tariff negative. Not enough batteries but currently build of batteries. Australia already there. But need another 200GW. Cost of operations of coal and gas. Lastly you forget behind the meter. Economics not as determinants
If you were to do the same kind of analysis on Tony Seba’s predictions from 2013 -2024, would the same pragmatic approach have led us to where we are today? Would we have looked at 2013 and said it was even possible we could do it in 10 years? It would be great to see your breakdown, as you do an absolutely transcendent job at explaining this stuff.
The biggest difference and more important difference, and frankly, the metric that would prove you wrong or Rethink X wrong is the growth rate. They have 100%, and you have 50%. That is worlds apart. I would like to see a deeper dive into that, projections of how things would change as you model based on 50% towards 100%. Maybe chat it as a graph so we can try to predict how things will go based on the actual cost curves and utility transition rates
I think in many countries, licensing for new mines is very time consuming. In some places it takes 20 years to open a new mine. It is +/- transparent how many and how big of the new mines are in the pipeline and where they stand. I believe the materials to be mined do not support the 100% growth path rethinkX sketched and it might be even critical if it is capable to support the 50% rate. While I wouldn't like to make the judgment on either of the 2 who is right, to me at least it looks of the 50% rate is closer to reality.
An astute analysis. However I would like to point out that this doesn't take into account ANY other developments like Ai and Robots, which will be in full swing by the early 2030s. And then there is Nuclear Fusion etc. As even more Wildcards.
I'm somewhat skeptical of the "Super Power" RethinkX is predicting. Today, in 2024, here in Australia as we switch from Winter to Spring, the energy prices have already switched to negative on the wholesale market. I have solar and a home battery and participate in the energy market. I'm already struggling to know what to do, during the middle of the day when all home power requirements are done, the car is fully charged, the home battery is fully charged, and it actually costs me money to export energy to the grid!!!! I just don't know there will be a demand at certain times of the day to make use of this Super Power. I think what is more likely is to curtail product, which is how I am dealing with it today, by either disabling my solar system or "going off grid" via my home battery system. We would need some industries or other uses to step in to make use of this free Super Power for it to make sense. I do agree that overbuilding wind and solar will have positive impacts and will mean that less fossil fuel backup is required in the event of dunkelflaute events.
The "waiting 2-3 years until the incremental power has 0 marginal cost" makes sense in theory, but in practice, things get built when they are net positive or better. Waiting 2-3 years for lower costs only makes sense if the value generated during those 2-3 years would be net negative vs waiting.
And if a lot of people put off buying until the price drops that would mean that the price would drop more slowly anyway.
These projects are planned years out. That is, the 2-3 years I"m talking about will automatically occur. That is, the project costs will always lag the current market prices by 2-3 years at least.
@@thelimitingfactor I must have misunderstood your point. In any event, very interesting video.
@@thelimitingfactor Exactly. If it automatically accurs then your point isn t invalid. If people wait, costs don t come down. I m confused.
@@thelimitingfactor I believe you could get price reduction in batteries in those 2-3 years because they're also being increasingly implemented in cars, but you wouldn't get that in solar PV. Wright's Law applies for an increasing production (hence, increased market) but if you don't have an increasing demand there is no noticeable price reduction. Only triggering the transition in a short time frame (2030) would accelerate production & price reductions.
Awesome video! - thanks for the reality check. I love what RethinkX are doing and your perspective is a step in making their high-concept view into a practical roadmap.
Good summary!
I'm always amazed by your deep analysis! Your breakdown was incredibly informative and gave me a new perspective.
Already just 4 hours from posting the sheer quality of the comments is a tribute to the respect in which your work is held. Rockstar Gordon!
Thanks man! I agree, great viewers 😁
Looking forward to this. Thanks Jordan
Sure thing man!
This channel is just getting better and better. ..❤👌👍
Thanks man 😊
You nailed it, Jordan! Thanks for all you do.
The grid interconnector 'gridlock' issue is an interesting problem. In the UK, the backlog for an interconnector can be as much as a decade. This did not matter much when new FF thermal power plants took ~10 years to build, but is much more problematical when renewable generation can be built out much faster.
My take is that shuttered (say) coal fired power stations (the last coal generator in the UK closes this year) represents significant value due to their existing redundant grid interconnector.
Every retired power station had a grid interconnector.
Now, wind or solar farms might not be co-located beside a decommissioned thermal power plant, but building transmission lines to that site might be quicker than waiting for a new interconnector.
Separately, building out a grid scale battery farm on the site may be a profitable solution for buffering and grid stabilisation.
Also, the inverter based renewables (wind and solar) require stabilisation. Massive flywheels in the old turbine halls make a cheap and efficient way to stabilise the grid frequency.
If you spread storage out around the grid, there's less peak demand on grid transmission. During lower demand periods, charge storage. Then use stored power locally.
Abandoned fossil fuel and nuclear sites are likely great places for large scale storage installations. Not only are large transmission systems in place, the ground is often contaminated. Land used for storage, solar, or wind does not need to be cleaned up to residential standards.
@@dominicgoodwin1147 Although this would be a low cost solution, I am not sure it will be needed. Batteries are capable of providing "virtual inertia" to replace generators and flywheels.
@@dominicgoodwin1147 a flywheel is happening at an old coal plant in County Clare, Ireland and I think another in East Anglia in Britain.
@@dominicgoodwin1147 - The UK is using battery storage for renewable stabilisation (as well as curtailment mitigation) - with nearly 5GW installed already, a further 4GW due to come one stream shortly and 95GW in the pipeline for installation by 2030. There is one 400MW flywheel in the UK, but it serves a different role - as a massive peak power source for the JET fusion research facility near Oxford.
Batteries are also being used to address the problem of grid congestion that in the past required gas peaker plants to switch on on the under supply of the congestion point and the power generators to be switched off on the over supply side. With both generating parties being paid higher spot rates to both generate and not generate power. Batteries provide an elegant and more cost effective solution than current apporach and mitigate the expensive cost of upgrading grid networks at congestion points that only periodically need to carry the extra power.
10 years since I first heard the name Tony Seba, bearing in mind the 2 "Global Issues" we've seen, he is still pretty spot on.
I am not going to bet against Seba.
@@Travlinmo
You ought to as he hasn't got a clue.
What he thinks and what is possible are as far removed from each other as is North and South.
Renewables have none of the criteria essential; for a stable, economic and reliable grid.
Just how can weather dependant and uncontrollable sources of generation feed a grid that has to be minutely controlled on an instantaneous basis, its just nonsense.
@@Travlinmo I am - not enough copper.
You probably know this more than me, but from what I understand, cost of installation is higher than the cost of solar panels themselves, which means that timeline for replacing fossil fuels highly depends on cost of labor. If in next two years we will have automated robots for ~20k like most companies predict, it will likely decrease cost of solar by 2-3 times, which likely will very quickly outpace costs of burning fuel, as installation costs is big part of the capital expense needed for solar. This will also likely decrease cost of solar + batteries, which will increase demand for batteries and hopefully will allow for much faster ramp up of battery storage systems, especially if battery production will benefit from automated robots as well.
While true, not enough impact quickly enough to hit a 2030 target
@@thelimitingfactor but by 2040.. ;)
@@thelimitingfactorand at the same time there’s 7-8 TWh/a overcapcity of LFP cell production in China. It might be all that capacity can be diverted to BESS deployments. Big news coming on that front quite soon 😁
Very much true. For example rooftop solar in Switzerland didn't become cheaper dunrig the past 4 years in spite of the panel prices dropping was visible even though not as dramatic as elsewhere. Rooftop solar booming anyway due to increased grid energy prices motivating potential owners to start a project.
My brother has been designing and installing solar for off-grid users for 30 years. About 15 years ago he was talking about how panels had dropped in price so much that the (lead acid car) batteries were more expensive than the panels. The result is they were designing larger solar output and smaller battery system. The result was that the batteries might be recharged by noon and they were just dumping excess power for the rest of the day. (Not actually dumping power, but pulling out less power than the panels were capable of producing.)
💯
I'm off the grid, have been for about 30 years. Many days my batteries are full by lunchtime. But I need that extra panel input during cloudy periods. It's cheaper, at this point in time, to overbuild generation than to add more storage.
BTW, solar panels have fallen in price from ~$100/watt to ~$0.20/watt over the last 40 years. And they should continue to get cheaper.
@@bobwallace9753 Same here in Australia. Better to over supply with panels.
I've seen a couple comments miss the point of Superpower. It isn't about being able to do things with "extra" power*, it's about the marginal cost of the power being used. By pairing installed generation with the worst case generation scenarios, you will have insanely cheap energy the rest of the year.
How much more affordable can consumer products, agricultural products, etc. be made with energy that is basically free for 9 months (depending on latitude) out of the year?
*Practically speaking, there is no such thing as excess power. That is because excess power being used means it is no longer excess and would counteract the reduction in marginal energy cost. The moment you try to do something with the excess all the benefits of superpower disappear.
Maybe it sould have been called excess capacity. Becuase that is what it will be most of the year. Since not all the power it could generate will be needed, the plants will be partially curtailed most of the year.
Solar (PV) modules are 6.1cent/W (€) right now, so a new 410W module is 25€, here at a local dealer, no taxes, only local shipping, amazing, 10 years ago I paid 36cents/W and that was considered cheap. Now, solar is nearly zero cost. Solar works for me at 53°N, >=75% of all electricity comes from my roof, all electric here of course. With a south facing solar array you can go 100% 24/7/365. Batteries are included :-) And that up here in the north
HOW? I did an admittedly lazy and cursory search days ago and $1/W seems to be the price here. I need to shop further 😂
my country had tariffs on chinese modules 10 years ago, today no more, plus there is zero Vat this year (instead of 19%). That pushes solar and shows everybody here the true price. Solar rocks
Good for you.
I think where I live we still pay some equivalent to €180.-- or so for a 410W module.
Tesla should produce magnesium in the Gulf of Mexico (sea water) using super power. Gigacastings out of cheap magnesium. See Jordans video about tixo moulding.
Capital intense equipment dosn't work very well if it isn't used most of the day if not 24/7.
This said in many instances it will be most cost effective to curtail the superpower.
Really enjoyed seeing Tony's recent work🥰and your perspective on it. Lots of new AI driven data centers are going to require lots more electricity....a great use for new renewable energy generation coming online before the US has time to mostly trade-up to EV's and home heat pumps.
Amen!
Excellent review ! Thanks for the deep insight !
Another flaw in the analysis is that it assumes the old model of a few large generators feeding many (smaller) loads. The old utility model. But the system is now more like the internet: Millions of nodes generating and consuming power at random times.. So the economic choices are made not just by utility companies, but by millions of individual node owners. The grid is like a network not a distributorship.
And we now have the ability for moment by moment pricing which can be used to move some demand off peak demand times, cutting the overall cost of electricity.
Great point.
This is a great point and the power of rooftop solar and home storage is really underplayed.
They don't need to be the cheapest form of energy for mass adoption, they just have to be cheaper than the market rate, to make sense for the consumer. This will then in turn completely change the form of the demand curve that the utilities have to cover.
I think that there are multiple standards for vehicle-to-grid (V2G) communication, including ISO 15118 and SAE J3068. Plus there are two UL standards that apply to V2G technologies: UL 1741 and UL 9741. The former applies to all types of generation equipment, while the latter applies only to vehicles and associated supply equipment. Someone needs to set out a universal V2G system standard, just as NACS is becoming a universal plug standard, accepted across America's charging network (including home and work charging and discharging). Battery chemistries continue to improve, and future EV owners won't be so concerned about battery degradation. An affordable, universal V2G standard needs to be developed and accepted, then the grid will be able to easily use BEVs themselves to become the buffer for renewable resources.
Amen!
NACS itself isn't a universal standard, it's a design that is tied to the small number of countries that operate on split phase 110/120v domestic or combined single phase 240v electricity supplies. Whereas most of the world operates on 220/240V single and three phase supplies. This is why CCS2 (AC or AC/DC combo) is the standard in many parts of the world, plus China with its GB/T standard.
The same applies to V2G as the electrical standards around the world differentiate the standards seeks to address these differences. Similarly, vehicles have a specific supply characterisitc of high power / short duration supply, whereas other techologies are addressing medium power / long duration needs that differing standards address.
It is these differences in systems and uses that are driving supporting standards, as one size doesn't fit all in this case.
What is positive is the small number of standards that apply meaning things are being tightly controlled to bring cost and implementation advantages that the alternative of a plethora of competing, often company/industry specific, standards would previal holding back adoption.
And on top of the generated power it avoids the probelm of 60% energy wasted as heat when using other methods.
It also neglects the emerging exponential curve in demand for power.
Correct!
Yes. But partially this wil be compensated by savings thanks to efficiency gains in transition from fossil to electric.
Another important RethinkX omission is lack of analysis on what industries can profitably use the excess power ("superpower") produced by overbuilding renewables. Industries that use lots of power like crypto mining, datacenters, and aluminum refining all use expensive capital assets that no company is going to be willing to have sitting idle half the time. I bet the capital equipment to produce hydrogen via electrolysis isn't cheap. And it's expensive to pay your workforce while your factory is shut down without power. With anything made by superpower, companies couldn't plan their production and promise delivery dates to customers. If there were good superpower industries waiting to go, why wouldn't we already be seeing them since solar in California and Australia already overproduce in the spring?
My thinking as well!
There's only so much of a market for superpower
Some possibilities (not necessarily good possibilities, but possibilities nonetheless) that come to mind for intermittent excess power: (1) desalination of water; (2) produce green hydrogen, and then combine that green hydrogen with CO2 to produce methane to be used as fuel in SpaceX flights; (3) pump water uphill and back into a reservoir to be reused in hydroelectricity generation; (4) generate heat that is pumped back into the ground for later extraction via a ground-source heat pump; (5) extract CO2 from the atmosphere.
The possibilities I suggest are wild guesses, but they have have something in common. They suggest that superpower may not be soaked up by already-existing-and-widespread uses, but rather that intermittent superpower may be soaked up by not-yet-widespread uses, simply because the benefits of superpower's low cost will outweigh the drawbacks of superpower's intermittent nature (in contrast, already large consumers of electricity typically would see intermittent power as being a deal breaker).
@@thelimitingfactor "There's only so much of a market for superpower"... today...
When things get cheap, industries get founded to take advantage of low costs.
@@CiaranMcHale The last one is the best one. When power is free, convert CO2 to elemental carbon or some way to sequester it. The reality is when we get to zero emissions, we'll need to clean up some of the excess that we added.
@@thelimitingfactor Other possibilities, most metal plants could produce a good deal more and could go on overdrive when electricity pricing is lower. AI training centers and digital coin centers could boost computation during times when energy production is highest. They probably could do this now with a well coordinated grid. By the sounds of what OpenAI would like to do, they want 5-7 5 GW plants. they could soak up a lot of power during the day
Great video,
There are two concepts that Rethink X talked about that you did not touch on:
1) Capacity Factor: Nuclear capacity factor will start dropping when solar saturates the market in autumn and spring. This will push up the cost of running existing nuclear and gas on a kWh basis.
2) Distribution Parity: It will be cheaper to go Solar, Wind and Battery for individual properties/communities than what it will cost to transmit power from central generation facilities. There are cases where the business case for grid connection failed vs on site solar and storage.
Both good points! Interestingly, more nuclear powerplants require more storage, lol
The fuel cost for nuclear is pretty low. Nuclear is not very dispatchable, so unless there's going to be periods of weeks and months where the generation is needed it's likely they will stay in operation. Plus if reactors are curtailed for a significant period of time it spreads their fixed costs over far fewer MWh of electricity sales. It's likely the government will find a way to make sure reactors get to sell their goods. Coal and gas will get turned off.
We need to account for the EROEI for the overall systems, That is, solar, wind and battery, compared to mass produced advanced little nuclear reactors. Which requires more materials to generate a constant reliable watt? Which requires more land? And, on the plus or minus side (I'm not sure) which will _cost more energy to also_ recycle?
@@fireofenergy Another consideration is that nuclear tech is also advancing, and molten salt reactors can use existing spent fuel AS fuel. Considering how much is currently being paid to store all that waste, it could conceivable cost negative money to operate these new reactors.
@@fireofenergy
First, we can't make a comparison with "advanced little nuclear reactors". They have not yet been built and evaluated. What we have is the nuclear industry telling us again that 'We have a new idea and this time it will make nuclear affordable". They've told us that over and over and over for more than 50 years and have yet to make it come true. Low credibility.
Maybe this new design will be cheaper. The Union of Concerned Scientists did an analysis and found that smaller may be more expensive. Time will tell. But, remember, new nuclear is around $0.15/kWh and needs to come under $0.04/kWh to be acceptable to the grid. That's about a 3/4 cost drop. Pretty hard to imagine. We've been closing paid off reactors that need more than $0.04/kWh to stay in business. The market won't buy their power.
Now, you talk about the overall system. Fine. Let's talk about the overall system for nuclear. First, like wind and solar, nuclear needs storage. Storage is used to move supply from low demand periods to high demand periods. Back when we were building nuclear we built a lot (over 100 just in the US) pump-up hydro energy storage plants. Add that cost in.
Then, there's the need for backup generation for when reactors "break" and go offline. Happens and happens more than most think. When a reactor goes offline it can be off for days, weeks, and even more than a year. The grid has to have excess generation standing by to replace one or more reactors. When both SONGs reactors in SoCal went down the grid wasn't prepared for that large loss and had severe supply problems. That's another cost to be added into the overall nuclear system.
Hint: A 100% 24/365 reliable wind, solar, and storage system is cheaper than nuclear, storage, and backup. Considerably cheaper.
6 years? He's silly. The decisions to pursue this won't even be made in 6 years.
Cool topic
Following Tony since 2017
Good job. I do this type of analysis myself and came to the same conclusion. In my models the cheapest energy today includes solar/wind/batteries/pumped hydro/hydro with gas peakers. As costs come down for batteries/solar, you can add more and more “super-power” to offset peaker use .. so these peakers get used less and less and less. In effect superpower is a type of “peaker” .. so IMHO that is how we can easily assess its cost effectiveness.
See, now THAT makes sense. You could actually extend [I assume] the lives of peaker plants in this way, making the grid more robust as renewables come online.
Kind of, but the big difference is that peaker power is very expensive, while super power is practically free.
@@mrleenudler only once that amount of capacity has actually been built, many years away.
@@jamesengland7461 Is kind of the definition of super power. And we already have some small measure of it. Renewable heavy regions sometimes go into negative power prices.
Peakers need to be dispachable. I don't think superpower from excess wind and solar is suitable. Batteries can replace speakers at lower cost, so I expect that to happen instead.
Tony Seba is selling hype via theoretical extrapolations of cost curves and so on, as if they don't have any unforeseen bottlenecks or don't need a practical buildout of supporting industry and supply chain. Great work adding some of such practical considerations and showing more realistic version of the future. Much needed dose of skepticism.
but me MUST try, & then can START with just 25% SWB and add more later, as cost comes down.
Thank you for the thoughtful reflection on the RethinkX presentation. Two questions:
1) Do the energy demand assumptions used account for new high demand categories (eg. AI compute) that would scale demand above the current trend line? We’ve seen with computers that as processing speed, memory capacity, storage capacity, and connection bandwidth became cheaper and more available that demand for each was induced at a rate that many people found surprising.
2) In terms of rate of transition and buildout, is there any accounting of the coming reality check on deficits and debt loads? You can run cost benefit analyses till you’re blue in the face, but if you can’t finance the debt (eg. corporate or gov’t bonds) you don’t get to buy the equip or do the work.
Nicely composed Jordan
thanks man!
Large AI model training is an interesting case for super power. Unlike many other use cases, interruptions are not a big deal, you just continue from the last checkpoint. The training runs are also reaching the point where they're talking about building nuclear reactors just to power them.
Super power doesn't really work for data centres because they need baseload, not intermittent to max out capital efficiency
by the way solar cell prices (not panels) fallen down to single digit cents... makes 10 usd for a 2 m2 panel which gives out 400 to 420 w/hrs... for meeting a household's electricity need, let's say 10 kw/hrs capacity, core tech just costs less then 300 usd of investment... rest of the cost are cables, glass, films, frame, and workmanship... there are design options to squeeze those down too and it will eventually get down...
"The average useful life of a combined cycle natural gas plant is typically considered to be 25 to 30 years. However, with proper maintenance and potential component upgrades, some plants can operate for up to 40 years or even longer."
The average useful life of US coal plants is 40 years. Almost all our coal plants are older than 30 years. Some of our CCNG plants are likely over 30 years old, the average age is about 20 years. All these plants will have to be replaced with something. Wind and solar have vastly lower installed cost per MWh generated than do coal, nuclear, and CCNG.
Additionally, since wind and solar are cheaper than buying fuel for a paid off CCNG plant as more wind and solar are added they will be given priority over natural gas in order to enjoy fuel savings. That means that when CCNG plants have a serious repair issue a utility might decide it's time to retire that plant early and spend the money on more wind/solar or storage.
Yeah, it's very much situation and geographically determined
You made a good point about electric power needing less input because about 70% of the fossil fuel input is consumed by the process before it ever go to work.
Two other factors will have a big impact.
The second and third cars that we typically have sitting in the driveway at most homes not only might utilize the super power or even be charged with the super power we usually have with our solar panels at home in the day time. They might even feed the grid,in effect replacing some of the battery expansion for grid level backup.
The other factor is that many of us are already weaning ourselves off the grid with home solar panels and batteries. When the sodium batteries get into mass production, price and availability of LFP should improve also.
Wow! Lots of factors affecting existing energy grids. I see it as a dynamic energy ecosystem. 1. Some assumptions in Wrights rule regarding a fixed decrease in cost over time deploying sustainble energy. Perhaps a greater cost decline would occur beyond the typical 'cost of material and production volume/scale' due to improved battery design (increase in storage capacity) and (ongoing) decrease in mining costs (supported by recycling battery materials, etc)
2. California building codes for new homes require solar panels. Say this transition takes place in other states over a few years. While adopting this new building code (along with battery cost improvements), it would be cost affective to have the federal government to provide financing to home owners to install battery storage devices. Battery storage in homes would provide a primary layer of energy source to buffer any major power outage that would be eventually covered by the local energy utility's storage system.
3. Meanwhile, continued conversions of existing homes to (newer) heat pump technology reducing energy consumption and costs.
Excellent video! If Wright's law is any indication (20% cost reduction for each cumulative doubling of total battery production), cost will likely sink slower and slower in time.
Critically the Lazard report also shows the very high cost of power firming. Casio has much higher electrical costs despite higher renewable penetration and lower generation cost. Transmission, capacity firming, and distribution are all large costs than generation. Overhauling the grid to hand intermittent load and moving that power is costly.
The power queue as you mentioned is already extreme. More than 8 fold increase in the last 10 years to more than double current generation is in the power queue and time delay has almost tripled.
Also the cost for energy storage assumes 4hr of capacity. With grid connection and lots of base load supply this works, but multi day let alone multi week storage is highly cost prohibitive to say nothing of seasonal variation.
My current project is on evaluation of space solar power which can provide highly competitive generation, transformative logistics, and 24/7 renewable base load. It serves as a perfect complement to variable renewable sources with the capture to beam power to multiple different ground locations as needed.
absolutely excellent
I think you are missing the Micro Grids and Virtual Grids options in your analysis.
I feel by 2030 there WILL be many homes producing surplus energies from newer/ higher energy dense batteries and more efficient tandem perovskite solar panels. Also, most of the EVs on the roads will have bidirectional capabilities. This would drive Micro Grids and Virtual Grids. At least that is what I feel like.
The challenge is going to be building smart grids that can take care of these growths.
Yeah, many will be producing, but that doesn't solve the thrust of the issues I pointed out
This is the future but it's probably off by a few decades at least. Too much infrastructure to rebuild.
First mover advantage, competitive markets will dominate decisions
One thing you touched on but didn't expand is the energy companies wanting to maintain their market shares. This pushes into the political arena such as the Ohio nuclear plant bribery of government officials. Depending on where you read, oil companies make up to 6 million dollars per hour. That's 24 hours/day year round. You can bet they will not die quietly.
As to your point on waiting a couple years for prices to decrease even more, I believe that the sooner we can eliminate the need for fossil fuels the better even if it costs a few dollars more. However I do agree that it is unlikely that we will reach full decarbonization by 2030. There are too many obstacles and too many people who either are afraid of change or just flat don't want change in their lives even if it makes their lives better.
Good job dissecting Mr. Seba's forecast.
Thanks man!
I agree. I avoided getting into the weeds on what we should do because that's political 😁 I have thoughts
Opportunity cost is why it accelerates
Limiting factor - no you want to charge grid battery & commercial & power walls @ peak solar, homes will need to charge overnight @ super off peak.
they can still export to local grid , after battery is charged (4H), and discharge is up to (6H).
Does this estimate of available power take into account the power draw of AI models which are scheduled to require 100s of GW in the near future?
Nope. Another reason why it'll take longer than 2030
What's possible by 2030 isn't what's probable by 2030.
LOL! Great summary for the video 💯
Thanks for this. Good analysis..
You're most welcome!
I think a lot of additional power generation will be private, without a grid connection, since the economics still apply. Date centres,etc.
Here in Sweden I get more or less 0 kWh solar for 4 months when it also is super cold.
I would need a lot of super power batteries to cover 4 super cold months.
We need to be nuanced, this is not a solution for all.
Bingo!
Of course. The closer you get to polar circle...
Is the wind not stronger during those 4 months.
US Dollar value is no longer pegged to Gold, it’s currently pegged to the price of Oil. What happens to the Dollar when Oil is of little value? Does that make Lithium or Solar Panels the new Oil? There is no shortage of Lithium or the materials necessary to produce Photovoltaic Cells…
Purchase of oil must be in US dollars, according to Nixon.
But it’s not pegged. Oil price fluctuates in US dollar terms.
I think you're right there. Due to global megalomaniacs all transitions get delayed and doubted.
Has regulations changed from 2010 when solar and wind power were subsidized by nuclear, gas and coal power by requiring utilities to assume the green energy was being provided 24/7 everyday of the year in the balance books. This requires that the extra cost of going green would be shouldered by nuclear, gas and coal. The income from nuclear, gas and coal would have to have their cost increased to offset the difference in the real cost of green energy and the subsidized cost on the utilities books. This started in Europe and came to the US decades ago. Very few really understood this requirement for the utilities to do this off-setting of the true cost of green energy. I worked in the utility industry for the last 45 years. The more coal, nuclear and gas power plants are shut down due to the increasing off-set to dirty energy into a smaller and smaller amount will keep showing dirty energy is getting more and more expensive. This will essentially stop this off-setting cost burden since the less dirty energy is provided there will be less ability to subsidize green energy; the real cost of going green will appear.
Interesting!
Much of the real cost of going green will be realized when there is sufficient battery purchase to go along with the solar/wind generation
What about the external cost of co2 emissions and the disposal of radio active waste? Right. That’s the tax payers money
Provided the predictions regarding cost erosion are going to become a reality, then the true cost of renewables including storage will be even lower than today. Subsidies will not be required anymore, maybe with the exception of the last few plants to be built for the super power, when it needs to be financed in spite of the uncertainty how much in 1 year it will be used.
Thanks, does your analysis consider domestic energy generation from residential rooftop solar? There is normally not much regulation stopping households installation a system which mostly covers there needs, which overtime could reduce the need for utility and grid build out?
Yes, because it's primarily about scaling
the wind always blows off shore and pumped hydro are excellent batteries
This confirms my confusion… I don’t understand why Tesla isn’t doubling down on Megapack. Seems like they’re overdue to break ground on a few more Megapack factories. Per an older video of yours, there’s easily going to be demand for 10-20 more Megapack factories this decade yet I do not see this happening. At this rate, we’ll probably go into 2030 with Lathrop and Shanghai only. There has to be a limited factor that I’m missing. Tesla has the cash. Demand is/will be there.
Even a new factory announcement for a Megapack factory this year with construction starting next year won’t reach volume production until 2027/2028.
Interconnect queues, etc is my guess. I think this is why they're working on the newer megapack that can be direct connected to High Voltage
JG, does this take personal Solar installation that reduces the grid need?
Installation of any kind. It's simply a matter of manufacturing scale, regardless of implementation
Excellent analysis, thanks. I figure superpower can happen much sooner than you think in super renewable energy resource States like Texas and California and many in the midwest. Then their grid can export the excess to neighboring states. Wait till geothermal Drilling gets to become mainstream then your base load power problem goes away and fossil fuel plants will disappear quickly.
Geothermal is not cost competitive.
@@bobwallace9753 that is rapidly changing. Search RUclips for the new plasma laser drilling technology that is almost ready for prime time that can vaporize holes deep quickly and cheaply into the Earth's crust leaving a glass tube for the geothermal energy flow. It is being pilot tested and developed as we speak...
Mass energy storage should be distributed. Individual homes and businesses need the power/energy they need no matter the state of generation.
Intelligent control systems allow demand management, so meeting the energy they need should be an interaction between producers, grids and consumers. Octopus Energy in the uk is already doing this with intelligent tariffs, as is Tesla of course.
Advantages both ways. Scale economics of utility production & storage is obvious, more so as panels drop to almost free while re-wiring the house costs. Present use case is immense GWh scale battery farms in outer suburbs.
Good analysis. I have a couple of thoughts. 1) I agree that the multitude of decision makers and entities in a market and regulated monopolistic market make super rapid change more difficult thus arguing for 2040 versus 2030 - at least from a directional perspective. 2) I think you second point regarding waiting for costs to drop further to lessen overall cost is weaker. It depends on how you factor the externality cost of the exponential impact of further GHG emissions and the potential for catastrophic non-reversible exponential climate change. There are solid scientific reasons to view a delay of even a decade as a real and potentially incalculably expensive risk. In that case, governments might take a more 'central planning' and less cost efficient approach just to avoid the existential risks of a delay in achieving net zero (and net negative because we need power to power the carbon sucking machines)....
Thanks man!
As for #2, that's not a weakness in my analysis, that's an unpriced negative externality that I can't control for in my analysis
I'm not here to tell society what it should do, only what the likely path is based on what we know right now
Listening to Jordan talk about RethinkX sounds like an adult explaining the faults in a child's fantasy.
V2G will be a great part of balancing the grid. Over production will be stored in hydrogen and other similar storage options. A stable grid with great backup is the way to get low price. Small nuclear plants that easily can be turned off and on is another base production that needs to come in place.
but NOT with vehicles.
NCM too expensive for V2G
you are better off w/ LFP home battery anyway.
@@markplott4820 My BEV has LFP batteries and rated for 6000 cycles compared to NCM that is rated 3000 cycles. But i do agree a small LFP home battery as a complete is also good when the BEV is not home. That is the setup I use here.
@@markplott4820 If it isnt w. vehicles, it's not VTG. And also vehicles are coming w. LFP these days.
@@beatreuteler - V2H is a GREAT way to burn out your very EXPENSIVE vehicle battery. (+$30,000).....lol.
you are BETTER of with a HOME battery like Powerwalls , even w/o Solar.
@@markplott4820 Home batteries are too expensive for the time being and it is quite well known that there is no detectable damage to a vehicles battery life when using VTH or even VTG. The only reason not to do it is if the benefit isn't there or if the manufacturer makes limitations through warranty clauses. Where I live, it doesn't make economic sense due to regulations causing extra cost. That's eating up all the possible benefits which makes it a nogo for a very long time. But in other countries that can do a good contribution to grid stability.
One obvious solution to excess renewable energy is desalination and pumping water uphill. In the western USA we are horribly short on water. Why waste all that extra energy when it could so easily be put to good use?
Because capital efficiency. In some cases, even if you get the power for free, you'll still lose money.
@@thelimitingfactor While this may be correct, in some cases, if such a project is of mutual use far beyond the energy supply, it may still be justified. However if that is of mutual benefit, it's just a matter of convincing everyone so you can have it supported financially by the local authorities.
What the idea doesn't cover, however, is the fact such convincing and planning may take well more than 6 years in most communities.
Don't forget you will have to install a lot of solar or wind to recharge the batteries to supply the electric loads when the wind or solar is not available.
That is not free energy!
think on this , entre USA can be SWB powered , with just 1% total land needed, size or Rhode island. but needs even distribution.
Overbuilding solar and wind is the right idea and Toni's projection is likely to be close. China is doing this now and they are actively managing excess solar PV curtailment with grid interactive customers (demand response) i.e. the smart grid
Overbuilding wind and solar saves on storage cost. And, to some extent, that extra generation can find a use. Charging EVs is an obvious use since EV charging is highly dispatchable. Another is desal. Lots of salt water could be turned into potable water during sunny mornings and windy nights with the fresh water stored in reservoirs for later use.
RethinkX was wrong but in a good way - just come to Australia where Superpower is already here! Between 10am and 3pm on most days including winter, the National Electricity Market spot price for generators is NEGATIVE, in other words, baseload power costs the generator up to A$50 per MwH to put it into the national electricity grid. Renewables (mainly solar and wind farms) are being curtailed. In Victoria, retailers are offering FREE electricity between 12pm and 2pm and you can get 50% off standard full-day rates between 10am and 3pm. In South Australia those free windows are wider. The challenge has been a) getting consumers to soak up the excess in the short term and b) building storage and interconnect capacity in the medium to long term. Believe!
The notion of storing 90 hours of energy use in batteries at an affordable cost is ludicrous. It depends on
a) Only converting the existing grid which is only about 20% of total energy use i.e. solving only 1/5 of the problem
b) No economic or populations growth
c) Fanciful assumptions about a huge acceleration in the rate of improvement of battery production. Remember that batteries have only gotten about 7%/pa better for 130 years. The easy wins are past us.
Jordan, thanks for this thought provoking piece. I agree with your conclusion that Seba is overly optimistic in terms of getting to 100%. That said, Seba has continued to surprise and I think we will do better than your at projecting. Here are some things that push the timeline in his favor. 1) Grid upgrades and interconnecting the three US grids would lower the curve by aggregating the variable renewables making their aggregate output smoother. The SEIA timeline plot you showed was from beginning to end. But there are a number of transmission line projects that are far along in their timelines so could be built much more quickly. 2) The old model is to provide all power requested WHEN it is requested. Many customers would happily alter their demand curves if they could recoup the economic benefits of that. This requires time of use metering. Tesla is addressing this on the supply side with power walls being aggregated into virtual utilities to arbitrage power. But that is the Seba model. It can also be done on the demand side and BEVs provide an excellent way for the grid to gain storage essentially for free. I, and many others already do this by charging at night when the grid in my area has more power available. 3) You use an example of solar only being available in the daytime. Aggregating the grid would increase solar availability by about 3 hours (3 time zones) and wind peaks at night making the window of low production much smaller and requiring less battery capacity. Further, wind turbines have continued to grow in height, this increases capacity factor (on time). Massive wind farms being built of the eastern seaboard are looking like 50% - 60% CF which is far in excess of what most aggregate models have utilized. I agree with your overall premise that there are constraints that make 2030 impossible, I think the middle of next decade will see us at least 90% of the way there for transportation and electric with the exception being HVAC which will take considerably longer. I did a deep dive into this stuff about 10 years ago. My conclusions were pretty radical with prognostication that we could get to zero carbon by 2050. I couple of years later I stumbled onto Seba's book, and agreed with his methodology but thought he was over optimistic by decades. NOBODY thought he was even close on BEVs. Here we are 10 years later and his estimates were closest to correct. I still don't think 5 years, but 10, I won't bet against that. PS - Most of my modeling is conceptually similar for Europe as they have access to massive solar in Spain and across the Mediteranian, lots of hydro storage in Norway and massive wind power off shore - Transmission lines the access to wind power handlemanpost.wordpress.com/2014/01/15/compare-maps-of-the-grid-and-renewables/ - Wind Power CF for onshore and off shore is even better handlemanpost.wordpress.com/2015/07/26/new-nrel-numbers-game-changer-for-wind-power/
Limiting factor - keep in mind, we dont need any Nickel cells for home battery, powepack or Megapack storage battery. All can be built w/ LFP instead of expensive Ni cells.
and will last 50+ years, and can be charged to 100% , discharge to 1%.
unlike Ni cells.
we will need more LFP battery factory 🏭.
Limiting factor - super power is defined 2 ways
#1 - enough renewable energy is produced to meet 100% of industry & consumer.
#2 - super low cost renewable energy in mass quantity, net ZERO cost.
I just bought solar panels through ebay for less than half what i paid almost 2 years ago. How cheap can they get?
Bumping the time seems to make sense. All the more with the expected heavy ramp in power needs for AI in the coming years.
There's a feed back loop. Bumping the time will reduce the rate of increase in economics of scale, likely making it necessary to bump it further... and so on.
7:30
This mismash of operators and regulators only exists in the US. For many places in the world the electrical grid is government owned and operated. Like in Canada. Tony is not looking only at US power grids. Australia,for example,is already going through this transition from coal to solar. Skipping gas and nuclear completely.
Have to realize battery tech. has been suppressed for 140 years, so you have to give them a break. Soon, there will be a battery, that takes all night to charge, but will run all day, just not there yet, things must evolve. For now, ANY kind of power plant will do, once greener tech. catches up, then the older means can be retired, right now, it is about enough power, no matter where it comes from.
Limiting factor - home solar & battery are under used, same for commercial & industry.
also wind.
most farmer in usa under use SWB, too.
Disruption potential is everywhere.
You made a good point about electric power needing less input because about 70% of the fossil fuel input is consumed by the process before it ever go to work.
Two other factors will have a big impact.
The second and third cars that we typically have sitting in the driveway at most homes not only might utilize the super power or even be charged with the super power we usually have with our solar panels at home in the day time. They might even feed the grid,in effect replacing some of the battery expansion for grid level backup.
The other factor is that many of us are already weaning ourselves off the grid with home solar panels and batteries. When the sodium batteries get into mass production, price and availability of LFP should improve also.
Even renewable energy has a big percentage of loss in transmission. Distributed generation and storage mitigates that.
Could you comment on the requirement to replace capacity at the end of its useful life please? I do not know if Rethink X has included these costs including cost of disposal / re-use
Nuclear power is not economically "dispatchable". Capital cost are so high that the plants need to stay at full power for as long as possible. After decades in that business, I am not optimistic for new nuclear generation.
Same!
You couldn't be any more spot on, I think.
I believe that Rethink went with the current electrical power because it is not dependent on any hardware changes, meaning no one has to change their gas stoves, hot water heaters, furnaces, cars, semi trucks, industrial furnaces, and any other hardware which uses direct fossil fuels and has a long lifespan. Changing from fossil fuel hardware to electric hardware will take place gradually overtime as people and businesses need to replace the appliance or machine.
Even if Texas did wait for 2040 to replace the entire grid because it would be so cheap, many people could not use this power because there has not been the gradual change I spoke of above.
In addition, the projected increase in energy demand needs to be met in the moment, not years later. So its either build more fossil fueled energy plants which are expensive to build and operate and would be a poor long term investment over its multi-decade life or go forward building the solar and battery capacity as electrical energy demand incrementally increases.
For "Super" read "Unprofitable".
Who is going to invest in new solar and wind, when there's no profit to be had?
For me the main limiting factor is that solar cell and component production / manufacturing is dominated by China
With the coming wave of AI data centers and their huge requirements for power, could it be AI buildouts dwarfs the infrastructure's ability to produce and deliver power regardless of the generation source? Perhaps the crux of the issue in 2030 may be how much of a nation's electrical capacity is directed to AI and how much is allocated to the rest of industrial, commercial and residential power needs.
Agreed.,Use the "I" of AI to help with this!!
RethinkX is right that low-cost economics will eventually win out. Frustratingly, they don't attempt analysis of what the rate-limited step, or dare I say what "The Limiting Factor," will be. This is critical to projecting the timeframe, as you point out. Just look at the grid interconnect queue as an example, with many new solar projects in OH now not even attempted following legislative changes that make new renewable generation projects easy to stop while new fossil projects continue to be hard to stop.
🎯
Tony is aware of Agenda30 a milestone of Agenda21.
I’ve been waiting for a video like this for a decade critiquing RethinkX et al. Thank you!
✊🏼🤠
As a general question. Are Tony Seba and Adam .. talking about Texas (or maybe the US) or about the whole world?
On whats defined as intermittency challenge, for South Germany you have around 2-3 months where there is not enough solar power produced. Thats for 48th parallel north. I am sure its similar in the US and other places. Well there is wind energy. But they have higher running expenses (higher marginal costs) and because you can't plaster the whole country but also environmental issues (there is more and more microplastic in the air, birds, microclimate changes) it won't help that much.
Australia could make it mainly through solar.
Looking to places like Canada and Scandiavia its more realistic that there are 3-5 months where there is a problem.
It would be interesting to see whats going on on Denmark.
So the intermittency problem can only go away when having fossil plants (nuclear is not enough) running for 1-4 months. It must not be bad. Having a battery storage for 30-100 hours makes sense. But as you demonstrated, its not even possible for Texas. Also i expect worldwide in 2030 maybe in average 4-8 hours can be stored. We will be surprised how quick the capitalistic world can increase the capacity but thats an extremely high demand.
Whats missed (but right now i don't see much activities) is high-voltage direct current (HVDC) between continents. From north to south (solves the summer winter issue) and from west to east (solves the daytime issue). Through the Atlantic yes. But the Pacific seems to large. It case it is economically feasable it will take min 10-20 years. Also several nations will keep their (nuclear) power plants running for the coming 20 years. You can't easily regulate or switch them off for a few days.
I agree on your points that a lot will be possible until 2035 or 2040. But it will mean at this time nearly all the current fossil power plants will be worthless.
Maybe i have some errors. Later i will look again on the video. And the Tesla presentation you mentioned.
Its good to think about these questions.
They use several case studies from different markets 🤠
The result is slightly different for different markets, but the overall gist is the same.
But, there are wildcards here such as the massive increase in demand from AI, etc.
Master Plan part 3 has underground cavern hydrogen storage from summer excess to winter shortage.
Another person who has modeled 100% renewables is Mark Jacobson of Stanford University.
Here is a pdf for Germany
Completion for 100% is 2050, but 80% by 2030
web.stanford.edu/group/efmh/jacobson/Articles/I/143Country/20-WWS-Germany.pdf
As I understand Tony Seba's vision, installing 2x-3x the solar PV required in summer will provide the needed power output in winter. In summer the extra capacity can be curtailed. This is really no different than how fossil power plants are curtailed when demand is not there to use their full capacity. Batteries can cover nighttime demand, and cloudy days are covered by a combination of excess PV capacity and batteries.
@@georgepelton5645 It's really more like 4-7X capacity
Hey Jordan, remember when Elon said that the entire US could be powered by a postage stamp of solar panels that are approximately 100 miles X 100 miles? That's about 10,000 miles, now imagine Solar Fencing along every Interstate Highay in the US. The US has about 70 Interstate highways which total approximately just under 50,000 miles. Solar Panels can be installed as a fence with Solar Panels on both sides with both sides absorbing solar power & turning that into electricity. And obviously Solar Fencing would be installed on both sides of the highway. 50,000 miles of Interstate Highways with a double sided Solar Fencing is approximately 100,000 miles & Solar Fencing on both sides of Interstate Highways is approximately 200,000 miles of Solar Fencing which is 20 times more electrical power generation than Elon said was necessary to power the entire US.
What is great about Solar Fencing is that it's easy to install, it doesn't take up any new land & is along Interstate Highways which is already owned by governments so they don't need agreements with anyone other then maybe State Governments. Because Interstate Highways run through every State it means every state is producing Solar energy & along with Mega-Packs can store as much energy as they want. Solar Fencing & Battery Storage can be scaled as they can afford it. And because Interstate Highways run through every state electricity can be transmitted everywhere by building transmission lines within the Solar Fence or by placing transformers, Battery Storage & transmission lines in Boring Tunnels under the Solar Fencing & away from environmental disasters.
Most Interstate Highways have mediums separating the two directions & if they wanted to create even more energy they could install Wind Turbine Fencing down the medium of these highways doubling the power generation again & producing energy 24/7 from wind & solar. They could also place small wind turbines on every fence post again producing more power. Solar Fencing IMO is the answer to electrical generation to power the world & we don't have to wait they could start doing this now. Here is a crazy idea, what if all 50 states offered one billion dollars per yr for 5 yrs to install Solar Fencing in their state & the Federal Government matched it dollar for dollar. That would be 100 billion dollars per yr for five yrs & would cover the entire country. Harris the Broadband Czar just blew 50 billion dollars with zero results so finding 50 billion is not that big a deal for the Feds. What do you think of that?
yes! That 50 billion could have worked magic *sigh*
98% should be goal, as that last 1% would be long term storage or inter connect.
15:40 the reason to work on and produce fusion power would be for interstellar travel, not for electricity on Earth's sake alone.
Correct
Thank you, was thinking the same thing. Energy is not only electricity. Decoupling environmental damage from human activity is likely to be found in space applications. The only way to power heavy industry in space is nuclear(fission or fusion), there just isn't anything that comes close.
@@stijn2644 Industry in space is a giant leap.
Thank you!
Sure thing man!
Thanks for yet another excellent video
It knocks me out that they are refurbing nuclear reactors to power AI data centers. If plants that are near EOL are taken off the grid by private investment, instead of being extended like Diablo Canyon, could that cause more SWB build out? (I have no idea, just thinking about it all...)
We already have curtailment in 2024, just because we don't know how to deal with excess. It will get worse.
Until someone would figure out how to profit with fully automated plants. Might be quemical energy storage (sort of peaker), H2 refinery, decarbonization, or desalinization.
Will get there regardless of timelines.
I think we know how to deal with a certain amount of excess, but I think what you're saying is that it'll take time to adapt, which I agree with
Even if batteries drop to 1/10 of their current price per kWh, a solar/wind and battery infrastructure would cost 5X (AT LEAST) what the current gas, coal, nuclear base load system costs.
Nah, they're already near parity (depending on the lattitude)
@@thelimitingfactor Then why is China building so many coal power plants?
@@thelimitingfactor I think tycurtin is referring to the fact in this concept you need a minimum of 4x the nominal capacity which certainly by now is still almost 4x more costly than grid parity. This said, this would be my hint to tycurtin: Please go back to minute 7:40 to look at the graph once more. The cost graphs also for PV and Wind (not only for batteries) are logarithmic. The projected cost reductions are foreseen to keep it at parity throughout the timeline, e.g. in 2030 it is foreseen to be 4x to 5x more cost effective as the competing sources, making up for your sorrow.
@thelimitingfactor, please correct me if I'm wrong.
Is the cost of continued gas and nuclear cheaper than renewables, I doubt that maintenance and burning gas is in fact cheaper.
@@waltermcphee3787 The only high cost for nuclear is the red tape of US rules and regulations designed to make nuclear practically impossible. These were the result of 3 Mile Island and psyops like "The China Syndrome." The footprint and maintenance of solar and wind is immense. As energy sources, solar and wind are so diffuse they take up vast regions compared to traditional power plants. Not only is it unsightly, but it takes huge resources to maintain. They just don't have very long lifespans either. This is all of course excluding the batteries required to make them equivalent to gas/nuclear. That's a whole other story that makes solar./wind economically unviable. Take for example, if there were no solar/wind subsidies from the government, how many solar and wind farms would be built??? NONE
Meh, it's going to continue to accelerate faster than any predictions. Tony Seba was spot on with EVs a decade before it happened. *Another* 10 years is a super long time. Maybe it's a 2-3 year delay, but I don't see 10.
I'm just repeating what their analysis says, but highlighted the caveats that they didn't or people missed.
That is, where have I misrepresented their research?
Limiting factor - in CA as case study, CA generally overproduce solar & wind , but are lacking enough battery storage in State, so pge & others waste it by exporting to nearby states, for a profit.
they can fix this in short time, by adding more battery storage.
and in CA hydro, geo are maxed out.
Solar thermal is not working.
CA has not explored off shore wind or tidal.
IRELAND is 100% tidal turbine power.
Thanks for the input!
I have been a big fan of this concept for many years. It describes what *could* be done in 2030 by the engineers, if they were 'king of the world'. They focus on 2030 because we need decarbonization ASAP. Your take on the concept is correct, focusing on the practical issues in the making.
And there is another issue: We don't have too many use cases for 'superpower' (yet). Our industrial processes are usually designed to run 24/7 and most of them do not like to be interrupted at all ;) So we will have to invent new processes that can be easily turned on/off.
The "superpower" does not need to be used all up. Wind power stations as well as PV arrays can temporary be disabled.
Australia solar already being curtailed. Utilities are turning off. Up to 50% already. Else feed in tariff negative.
Not enough batteries but currently build of batteries.
Australia already there.
But need another 200GW.
Cost of operations of coal and gas.
Lastly you forget behind the meter. Economics not as determinants
I already said that at the end of the video
Depends on the lattitude
If you were to do the same kind of analysis on Tony Seba’s predictions from 2013 -2024, would the same pragmatic approach have led us to where we are today? Would we have looked at 2013 and said it was even possible we could do it in 10 years? It would be great to see your breakdown, as you do an absolutely transcendent job at explaining this stuff.
I already answered this below.
I'm just repeating what their analysis says, but highlighting the caveats that they didn't or people missed.
The biggest difference and more important difference, and frankly, the metric that would prove you wrong or Rethink X wrong is the growth rate. They have 100%, and you have 50%. That is worlds apart. I would like to see a deeper dive into that, projections of how things would change as you model based on 50% towards 100%. Maybe chat it as a graph so we can try to predict how things will go based on the actual cost curves and utility transition rates
I think in many countries, licensing for new mines is very time consuming. In some places it takes 20 years to open a new mine.
It is +/- transparent how many and how big of the new mines are in the pipeline and where they stand.
I believe the materials to be mined do not support the 100% growth path rethinkX sketched and it might be even critical if it is capable to support the 50% rate. While I wouldn't like to make the judgment on either of the 2 who is right, to me at least it looks of the 50% rate is closer to reality.
An astute analysis. However I would like to point out that this doesn't take into account ANY other developments like Ai and Robots, which will be in full swing by the early 2030s. And then there is Nuclear Fusion etc. As even more Wildcards.
Correct! Event horizon
I'm somewhat skeptical of the "Super Power" RethinkX is predicting. Today, in 2024, here in Australia as we switch from Winter to Spring, the energy prices have already switched to negative on the wholesale market. I have solar and a home battery and participate in the energy market. I'm already struggling to know what to do, during the middle of the day when all home power requirements are done, the car is fully charged, the home battery is fully charged, and it actually costs me money to export energy to the grid!!!! I just don't know there will be a demand at certain times of the day to make use of this Super Power. I think what is more likely is to curtail product, which is how I am dealing with it today, by either disabling my solar system or "going off grid" via my home battery system. We would need some industries or other uses to step in to make use of this free Super Power for it to make sense. I do agree that overbuilding wind and solar will have positive impacts and will mean that less fossil fuel backup is required in the event of dunkelflaute events.
Temporary Disabling is the way to go.
So, you're saying, that instead of building fusion plants, we can simply borrow fusion power from the sky. Nice.
💯😁