Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

This site may earn commission on affiliate links.
This is very interesting

 
One important point is Elon's window for selling has closed and won't open till after Q2 ER. By then, Shanghai should have ramped back to full capacity. A split is on its way. We can also hope the macro picture will have improved a bit. For the next 2 months, the shorts are down 1 attack vector.
What is this window of which you speak?
Note that Google returns a search result for a specific company's (Global Future City Holding) insider trading which looks like an SEC rule, which it isn't.
SmartSelect_20220525-215401_Firefox.jpg
 
  • Like
Reactions: UCF3

The Renewables Revolution

Why Mainstream Projections for Future Human Energy Consumption are Wrong by at Least an Order of Magnitude


***************************************************************************************************************************

Introduction: The Energy Situation
You and I are engaged in a perpetual war against entropy. The tyrannical Laws of Thermodynamics fundamentally guarantee the usefulness of concentrated controllable energy because it can do work for us by generating:
  • Concentrated heat
  • Mechanical force
  • Photon emission
  • Endothermic chemical reactions
  • Electromagnetic field changes
As it turns out, these impressive and physically irreplaceable capabilities are why we need continual energy supply to have a modern industrial civilization. The blog Wait but Why has a couple of fun and informative summary articles on energy for dummies and Tesla’s role in the future of energy.

Understanding Tesla Energy’s potential first requires understanding the total addressable market, and I’m not talking about the 180 petawatt-hours currently used by humans annually (link).

Around 1800, humanity began to transition from a traditional biomass energy economy (wood mainly) that had been in place for millions of years of hominid evolution. The new industrial sector of the economy used industrialized energy sources, starting originally with coal and water wheels used primarily in Great Britain initially and shortly after in Continental Europe and the United States and then later worldwide. A century later we added more sources to the mix: oil, natural gas, hydroelectric, and finally nuclear. As the chart below clearly shows, the biomass energy market at the dawn of the Industrial Revolution was two orders of magnitude smaller than today's total market for post-industrial energy sources. On the one hand, world population is up 8x since 1800, so per capita energy consumption globally is up "only" about 13x. However, most of us would aspire that everyone on the planet can have a standard of living like the middle class in OECD countries.

Global average per capita power used in 1800 was ~200 watts, whereas in the developed countries today it's more like 4,000-10,000 watts. This has been a roughly 20x-40x increase in consumption per capita and there is plenty of appetite for more, as shown by data indicating that the most affluent consumers within developed nations tend to use far more energy than those with lesser means. This is also significantly understating the true energy cost of a typical OECD lifestyle, because much of the energy usage is occurring in countries with heavy manufacturing bases like China and India that sell primarily in OECD markets.

If we want a super prosperous future for humanity even just with today's energy usage patterns, then 10 billion people in 2050 will consume ~10 kW of power per capita which is 100 TW or about 5x what humanity consumes today.

Many environmentalists, including me, have found this fact alarming, because of the problems with our current energy mix and because of the inextricable links between per capita energy consumption and economic prosperity, human rights fulfillment, health, life expectancy, social harmony, and scientific progress. For decades it has seemed as though the Universe handed us a cruel choice between impending catastrophe from peak oil and ecological collapse or self-inflicted catastrophe from economic and agricultural collapse. Was modern industrial civilization doomed from the start to be a brief party before the inescapable reality of thermodynamics and finite fossil fuel reserves put us into a death spiral of energy shortages, famine, war and extreme weather events?

After years of anxious researching of this topic that led to me investing in TSLA originally, I have changed my mind entirely. That energy consumption curve is about to go vertical like we've never seen before, and surprisingly that’s good news not only for the economy and poverty, but also for the biosphere, world peace, and of course Tesla Energy's cash flow statement.

The Stone Age did not end because we ran out of stones, and the Oil Age will not end because we ran out of oil.

global-energy-substitution.png

Cost Collapsing & Doomed Legacy Energy “Assets”
Nuclear fission energy was famously hailed in 1954 by Lewis Strauss, then serving as the Chairman of The Atomic Energy Commission, for being on course to provide energy “too cheap to meter” in the near future. Infamously, Strauss’ prediction was wrong. The details of why are beyond the scope of this forum, but we've been here before with grandiose claims of energy abundance. Why is it different with the solar, wind and batteries (SWB) electric grid model?

At this point the main thing to recognize is that SWB electric generation, unlike the nascent nuclear industry in 1954, has decades of proven empirical cost decline trends with no slowdown in sight. While nuclear development has long been unfairly limited by social and political opposition by well-meaning but misinformed people, the time has passed for that because all legacy power plant architectures are headed straight for obsolescence in the 2020s. All of them, end of story. The SWB revolution is disrupting the entire category of thermal power stations, irrespective of the heat source being coal, natural gas, oil, atomic fission, atomic fusion, geothermal heat, herds of hamsters running on wheels, or magic Vibranium crystals gifted to us by Wakanda for $0. The old power plant model involves vaporizing water to crank a turbine to spin a dynamo to drive transformers for voltage step-up to transmit power long distances on high-voltage AC lines to power more transformers for voltage step-down before finally something useful happens with the energy. On average two thirds of the electric power generated at the dynamo is lost as waste (mostly resistive heating) en route to the end user.

Legacy power plants need to be big, centralized and far from population centers, necessitating this ridiculous electricity distribution apparatus. Solar PV panels in particular are notoriously scalable and emission free, so they can go on buildings or be built in medium size farms or in big farms that can be located close to the demand.

The old model is steadily losing economic competitiveness via exponential decay (reverse S-curve/negative logistic function), like all technologies being disrupted by a new paradigm. They will predictably flounder, attack SWB with media disinformation and FUD, whine and cry for government bailouts, possibly receive some degree of bailouts, and then fail anyway when they can't even achieve positive gross margins and their competition gets 10% cheaper every year.

Moreover, before long even functional coal/gas/nuclear/oil/geothermal plants with capital costs already paid for will start retiring earlier than TSLA hodlers. Why? Their operating costs will be more expensive than additional replacement SWB expansion, even if we continue the glorious time-honored tradition of refusing to pass laws to charge appropriate fees to operators of such plants for the unpriced negative externalities of their activities. It gets even worse for legacy plant operators when we add in the fact that localized solar and batteries generation will be cheaper in most populated areas than the mere operating costs of the big regional grids needed for legacy centralized power stations to be practical. All those towers, transmission wires, and substations ain’t cheap and the existing infrastructure is aging. RethinkX calls this price threshold “GOD parity” (Generation On Demand). Game over.

This is not theoretical, seeing as wave after wave of early coal plant shutdowns in the United States have already been happening for several years even during the Trump Administration when many EPA regulations on coal were relaxed, and gas plants are now facing an unstoppable battery onslaught. Tesla’s big splash into the South Australian energy market was a warning shot. The Hornsdale Battery Reserve wiped out the vast majority of the local Frequency Control Auxiliary Services market—previously ruled by a gas plant operator cartel—and it happened almost as soon as the switch was flipped on. Even in Texas, that great bastion of low regulation and unbridled free market capitalism, that place with a uniquely Wild West electricity bidding market in ERCOT, that place which happens to have an endowment of some of the world’s cheapest natural gas, is leading the USA in total SWB deployment. Tony Seba and the RethinkX team are correctly predicting the greatest stranded asset write off in the history of capitalism happening in the 2020s.

The writing is on the wall, or at least the writing is on this chart that you might want to print out and affix to your wall to gaze at for peace and serenity whenever you stare at your brokerage account feeling panic and dismay about a TSLA drop:

256CC10A-59A3-4E89-969A-523C70A515E7.jpeg


This empirical trend has been remarkably steady for 7 decades since the invention of the modern silicon photovoltaic cell in 1954 by researchers at Bell Labs, which was basically the Tesla of the 20th century. That's Wright's Law for you.

Similar trends have been seen for wind and batteries.

Wind hit the scene with competitive prices sooner than solar did by about 10 years. However, wind is much more exquisitely sensitive to local microclimate and to daily weather fluctuations, because the extractable power is roughly proportional to the wind speed cubed. 10 knots wind speed can provide about 8x more electricity than 5 knots wind speed. Wind also has more impact on landscape aesthetics, long distance plant reproduction, and bird migration patterns. Worst of all, compared to solar, wind has a much slower rate of cost decline and vastly less total available power globally by about 4 orders of magnitude. Still, it's good enough to help the cause and there's lots of wind out there in places like the North American Great Plains and the European North Sea.

Batteries are the dark horse coming to market the latest of the three. The first prototype lithium-ion cell was made by Stanley Whittingham in 1974, 20 years after the first solar PV cell. Exxon bought the technology IP and gave up on it, so modern Li-ion cells really didn't get going until John Goodenough and Akira Yoshino's work in the late 1980s presented a better design using cobalt oxide and carbon for the cathode and anode. Using these advances, Sony in 1991 began sales of the first practical Li-ion battery cell, kicking off development of an industry that led to Whittingham, Goodenough and Yoshino jointly winning a Nobel Prize in Chemistry. The costs since then have been plummeting according to a Wright's Law relationship between volume and cost:

1653268716018.png

Source

Consider the technology and cost projections Tesla disclosed on Battery Day, and further consider that:

1) They were probably sandbagging​
2) They were probably withholding information and not revealing all their technology​
3) Two years is a LONG time for Tesla’s technology to progress​

I think it's safe to say that the cost decline trend is going to hurtle well past $100/kWh and could conceivably hit less than $50/kWh with known plausible technology development pathways.

Nonsensical Mainstream Forecasts
The UN Sustainable Development Goals for 2050 contain some excellent ideas that we might want to try. However, I disagree with important aspects of the energy goals. They are aiming for moderate increases in total energy consumption as the developing world industrializes and they focus a lot of attention on improving energy efficiency and consuming less energy in the rich world (good luck convincing average people to do that or vote for people who will force them to do it). I believe this is a severe projection error that is happening because the UN analysts are discounting the implications of the solar energy cost trend, either because of some judgment error on their own part or because the implications seem so outrageous to a casual observer that it's not politically viable to publish what they actually think.

If you look into forecasts from other major government and NGO think tanks, you get similar estimates, such as:
This ridiculously inaccurate BS is influencing opinions, influencing the economy and business planning, and influencing public policy. I don't normally like to throw shade at strangers, but this is dangerous incompetence that is increasing existential risk for humanity and it needs to be called out. Improperly influencing public sentiment, lawmaking and capital allocation from these positions of responsibility will indirectly kill millions of people and thrust tens or hundreds of millions more into years of extreme poverty they didn't need to suffer. This is morally unacceptable and I hope it's due to mere incompetence instead of malice. In any case, the outcome is the same and I am heavily voting with my capital that these forecasts are wrong.

Just look at some of these slides from organizations that Wall Street considers credible sources for energy projections. They are forecasting:
  • Linear growth of renewables for a quadrupling by 2050 to 25% of the global energy supply
  • Linear growth of the world economy
  • Stable global demand for coal
  • 40% growth of oil and gas demand by 2050
  • 30% EVs in the global fleet by 2050
  • Increasing global CO2 emissions
  • Declining energy intensity per $ of GDP of 30% by 2050
It needs to be said: These projections are pathetic. These analysts are getting 1st-year undergraduate level statistical analysis wrong. Solar, wind and battery growth trends have been for many decades following almost perfect and instantly recognizable exponential growth trends. If you plot it on a log scale it's almost a perfect line with a linear regression having R^2 of more than 0.9! How can these official reports be published where the long term historical growth trends aren't even shown and where they assume without any justification that the trend will suddenly switch from exponential to linear? If you're going to project a radical change in the growth trajectory then you had better tell the world why you think that instead of just riding off your credibility. That's how science is supposed to work! You have a hypothesis, then provide data and analysis to prove it.

It's no wonder the finance and consulting industries don't understand Tesla. They actually believe this nonsense.

1653505794249.png


1653505914629.png

1653506079688.png


1653504933656.png

Disruptive Technology Majorly Expands Markets
Disruptive technology expands markets, because of higher utility and/or lower costs. It is never a one-to-one substitution, but human psychology tends to make these market expansions and new use cases surprising to most people, including experts. If you haven't seen Tony Seba present on this you need to watch now. We humans often underestimate the collective cleverness of our species to find new ways to exploit resources.


In retrospect, the market for paper information storage and transmission in 1970 was, to say the least, an extremely poor indicator of the size of the future market for digital electronic information storage and transmission. Anyone predicting that computers would merely replace existing paper-based use cases would have been embarrassingly mistaken. And there were many such people, and they were indeed embarrassingly mistaken. Only those with imagination who reasoned from first principles had anything resembling accurate forecasts, and even they tended to majorly underestimate the trajectory of growth in bytes of information generated, stored and transmitted. If I could go back in time and tell Alan Turing that humanity in 2020 would have a worldwide web transferring 3 zettabytes of data annually between over a billion computers made from carefully organized sand, he would’ve probably fainted from shock. But here we are now when I have a handheld pocket computer that can hold 256 GB, which is the equivalent of 54 billion words of written English.

Likewise, the market for horse and buggy transport was an extremely poor indicator of the size of the future market for automobile transport. Same for other technology disruptions, including examples like:
  • Film photography --> Digital
  • Vinyl music storage --> Tape --> CD --> MP3
  • Hunting/gathering food --> Agriculture
  • Human-borne freight with hands and shoulders --> Human-borne freight with baskets & backpacks --> Wheeled freight pulled by humans --> Wheeled freight pulled by domesticated animals --> Wheeled freight hauled by automobiles
Every time something better and more efficient and cheaper comes along, we end up using it far more than the preceding technology generation, unless another subsequent technology disruption ends up eliminating the want/need altogether. The empirical evidence from not only recorded history of the last 8,000 years but indeed from millions of years of hominid evolution dating back to our stone tool technology disruptions indisputably shows that technological disruptions always tend to follow this pattern. This is fundamentally the main reason why GDP per capita improves over time. The fact that regulated market economies tend to promote such technological advancement is one of the primary reasons they produce more overall prosperity than other social systems, despite having some well documented downsides.

When illustrated this way, the point seems like common sense, but human psychology is not built to intuitively understand exponential growth and has a bias for assuming that life in the future will look similar to life today, because the future is intangible and requires imagination and because making bold predictions about the status quo changing comes with risk to social status.

This pattern is also predicted by basic economic theory: when technology improves the combined "total factor productivity" of land, labor and capital such that supply of a certain good or service has increased, the market clearing price decreases and the total quantity sold increases. The “price elasticity of demand” for a good/service measures how much the quantity demanded increases for a given decrease in market price (link). More formally, it’s the slope (1st derivative with respect to price) of the demand curve at any point. We can only speculate about what the true shape of the demand curve for energy looks like at the costs we're talking about, because it's uncharted territory with no readily available econometric data. However, we have some clues.

The inflation-adjusted price of energy has actually been increasing since WW2 with gradual exploitation of the easy fossil hydrocarbon reserves and the easy hydroelectric opportunities and with the rise in population and industrialization, but the quantity sold has still been growing. Another factor in demand has been the advances in other complementary technologies that have increased the economic utility per unit of energy consumed. So, what happens when the population still keeps rising and the rate of industrialization accelerates globally and technology continues to increase the utility per Joule consumed, while at the same time a new type of energy drastically increases supply?

Energy consumption patterns in such cities as Dubai, Riyadh and Abu Dhabi may provide a glimpse of what changes are coming. For instance, Dubai gets 90% of its freshwater supply from desalination of the Arabian Gulf and Dubai also has a 2000 square meter indoor ski resort/snow park at the Mall of the Emirates. Furthermore, perhaps we should keep watching what happens in major Middle Eastern cities they will be on the leading edge of the SWB revolution due to their strong sunshine and cheap land giving them the cheapest solar power of any big cities in the world.

So...why are professional analysts and organizations assuming that the SWB revolution will leave a market comparable to today's energy market? This contradicts all available evidence and theory about the dynamics of technology disruptions and specific potential use cases for cheap SWB energy. This is the same type of flawed thinking that Tony Seba’s team has shown to typically occur for any tech disruption, and it's unsurprisingly happening again.

As far as I can tell:

We are not headed for a net-zero carbon economy with moderate total energy consumption increases by (hopefully) 2050 if and only if governments and citizens cooperate in an unprecedented display of human trust, solidarity, and selfless cooperation.

I think we are headed for a carbon negative economy by 2050 because of SWB, even if we forecast with the most cynical possible assumptions about selfish profit-seeking and personal-quality-of-life-improving motives dominating behavior. I also think the transition to the new economy will increase energy consumption roughly similar to the 100x jump from our ancestors had in transitioning from biofuels to fossil fuels/hydro/nuclear.

The Sun continuously bestows 173 petawatts of power to Planet Earth’s crust and currently we use about 10,000x less power than this (4 orders of magnitude). Additionally, in outer space at Earth’s distance from the Sun, solar energy flux of 1.4 kW/m2 is waiting for us in practically unlimited quantities if we eventually want to reserve Earth’s precious surface area for other purposes, and can be beamed down to Earth’s surface with microwaves and captured with rectifying phased array antennas. Starship and off-Earth manufacturing may make this economical someday. With enough energy at a sufficiently affordable cost, there is practically no limit to the human economy in this solar system or any other solar system. Stars ☀️ are mind bogglingly powerful.

We are also not going to reduce our average energy intensity per unit of GDP in the process.

In developed economies, like the USA for example, the energy intensity of GDP (average joules per $ of GDP produced) has been slowly declining for decades, but I think the SWB revolution will lead to a monumental reversal of this trend. The steepest drop in energy intensity occurred during the 1970s energy shock, and energy prices have on average creeped up in the time since. With SWB power the price of energy falling will mean more energy is used for relatively low value applications compared to now, which implies that value produced per joule will drop. Although some individual energy use cases like computation and transportation will get more efficient per FLOP or per mile, this effect will be heavily outweighed by the growth of the market and by the development of new energy-intensive industrial capacity and infrastructure that exploits the delightfully cheap SWB energy of the future.

Moreover, this staggering increase in energy supply will help us actively fix the biosphere without crashing our economies, while delivering all the other benefits of SWB that the UN goals already include, like achieving a wide variety of social justice objectives, avoiding hazardous pollution, reducing radiation exposure from coal emissions, reducing root causes of warfare, etc. It will also eliminate our terrible dependency on oil and gas for chemical purposes while increasing the supply of said chemicals. The SWB revolution will be as profoundly disruptive as the Scientific Revolution and subsequent First Industrial Revolution.

What new potential is waiting to be unlocked by lower prices? Some have been discussed already by others and me, but the math is telling me that almost EVERYBODY is underestimating the magnitude of what’s coming.

Primary Chemical Production
We need chemicals. All kinds of them. Lots of them.

Too much of our chemical supply comes from oil and gas, but if there's one thing I know in my limited understanding of chemical engineering, it's that almost everything is negotiable if you’re willing to pay the price of fighting entropy by expending enough energy. Many chemical synthesis processes are known and fairly well researched but have been rightfully shelved because of their wasteful energy consumption. When energy suddenly explodes in abundance and affordability, that changes everything. We are about to enter a new era of brute force chemical synthesis driven by cheap SWB power.

Energy-intensive processes exist to synthesize all of the following primary chemicals using solar power and abundant material inputs including primarily air, water, salt (NaCl) and limestone (calcium carbonate CaCO3):
  • Methane
    • CH4
    • Sabatier Process
  • Ammonia
    • NH3
    • Haber-Bosch Process
  • Hydrogen
    • H2
    • Water electrolysis or Chloralkali Process
  • Hydroxide
    • OH-
    • Water electrolysis
  • Carbon monoxide
    • CO
    • Steam reforming of CH4 or High-temperature electrolysis of CO2
  • Lye / Sodium hydroxide
    • NaOH
    • Chloralkali Process
  • Chlorine
    • Cl2
    • Chloralkali Process
  • Ethanol
    • CH3CH2OH
    • Fermentation of sugar with yeast or new electrochemical process published in 2017
  • Sodium carbonate
    • Na2CO3
    • Obtainable by mining, but also can be synthesized with the Solvay Process
  • Quicklime
    • CaO
    • High-temperature calcination of limestone (CaCO3)
  • Ozone
    • O3
    • Coronal Discharge Method or Electrolysis of H2O
Out of all the chemicals on this list, only sodium hydroxide and ozone are currently mass-produced with electric power, with the rest directly relying on coal and natural gas as either feedstocks or heat sources.

With these basic building blocks and ultra-low cost of energy, almost any imaginable chemical can be synthesized. As a matter of fact, today’s chemical industry already makes extensive use of the chemicals in this list to knit together bigger molecules. I would encourage you to go explore Wikipedia for any of these chemicals to see the possibilities. These primary chemicals can also be used to synthesize a wide variety of derivative chemicals currently produced by cracking petroleum, because syngas of CO + H2 + steam can go through the Fischer-Tropsch process to form liquid hydrocarbons we typically get from oil. Other pathways exist too, and I'm not enough of an expert in this area to know which is best. However, I am fairly confident that cheap SWB energy and cheap H2 and CH4 will make it economically competitive to produce key chemicals like olefins, higher alkanes, carbon monoxide, formaldehyde, and aromatics.

I do not have the requisite expertise in chemical engineering to estimate how much of the cost is affected by energy compared to other factors, but the outlook is good especially as the hydrocarbon mining industry dwindles and loses economies of scale. Also, there may be ways to exploit cheap energy with tradeoffs that reduce these other costs.

It could get even crazier. If we get to a point where we can use machine learning models like AlphaFold to reliably design proteins in a reasonable timeframe and then use genetic engineering techniques like CRISPR to insert the corresponding gene sequences into microorganisms for precision fermentation, in principle we could theoretically make a practically unlimited supply of our own customized organic catalysts to unlock even more potential for a sustainable abundant chemicals industry. Among the benefits of customized enzymes would be lower cost, higher yield, improved energy efficiency, reduced undesirable byproducts, and higher throughput. This is a long shot because protein folding and enzyme behavior are extremely complicated to model computationally and even advanced AI may not be able to solve protein design in a practical computational budget. However, the known laws of physics don’t preclude the possibility and if it works it will literally catalyze the growth of the chemical synthesis industry even more than cheap SWB energy alone.

Methane and hydrogen synthesis from CO2 and H20 could by themselves drive a 10x or more bump in energy consumption because it requires splitting H20 and CO2 molecules which has very low thermodynamic efficiency since those molecules are extremely stable. As Terraform Industries founder Casey Handmer wrote in a whitepaper explaining the purpose of the startup and doing the physics calculations:

"Indeed, the quantity of electricity required to synthesize 100% of California’s natural gas demand exceeds grid consumption by more than 10:1. This means that synthetic fuel price parity will drive enormous increases in demand for solar power deployment, providing the demand necessary to keep the learning curve bending downwards."

Plus, this 10x estimate is probably understating the impact of increased consumption of methane and hydrogen driven by lower cost, again in line with the Law of Demand that predicts higher quantity demanded when price is lower.

Methane is just one of the many chemicals we need, and a huge variety of other chemicals can be derived from H2 and CH4. As far as I'm aware, we can in theory replace the main outputs of the entire petrochemical stack with air, sunshine, water, and common minerals at potentially lower cost. This matters.

Terraform Industries is a very early stage startup, and they have a competitor who also is surely aware of where solar prices are headed: SpaceX.


We already knew this though, because this method is the only possible means of in situ methane production for fuel of habitat and rockets on Mars. Elon said on the recent Everyday Astronaut interview with Tim Dodd that when Starship/Super Heavy is fully and rapidly reusable, the propellant expenses become most of the cost structure. Elon also said that the sole figure of merit for the rocket, booster, launch and manufacturing teams is to work together to reduce cost per ton to orbit. Obviously this means they will soon devote HEAVY, RELENTLESS attention and engineering prowess towards developing high throughput, low cost, environmentally friendly CO2 + H20 --> CH4 + O2 factories. What people are missing is that this also implies that SpaceX might expand to adjacent markets, which happens to include...all uses of oil and gas that can't be directly replaced with SWB power. Methane is methane. CH4 is a global commodity. There's nothing special about "rocket fuel" methane except perhaps higher purity. Ironically, the Musk Foundation may technically need to award the $100M X-Prize for carbon capture technology to SpaceX!

No one is likely to outcompete SpaceX in any market segment where they devote the majority of their engineering resources, and once full and rapid reusability is working, methane production will be the next target. I think only Terraform has a shot at beating SpaceX by getting a head start while SpaceX focuses on Starship/Super Heavy, but then again maybe they'll be acquired by SpaceX like Hibar, Grohmann and Maxwell were acquired by Tesla.

Yes, you read that right. SpaceX looks to be in the early stages of becoming the world's largest and most profitable hydrocarbon producer. Potentially in the long run SpaceX Hydrocarbons could make today's Saudi Aramco look tiny, because SpaceX will have even lower production and distribution costs than Saudi oil & gas, and there's vastly more available carbon in the atmosphere than in Saudi crustal oil & gas reserves. SpaceX will likely get even more bonus pricing power vs. mined hydrocarbons to the extent that governments provide carbon sequestration subsidies. This is probably 20 or 30 years away but it seems almost inevitable if you look at the historical track record of SpaceX and other Musk companies and the overarching goal to "maximize the probability that the future is good".

SpaceX will need to make green hydrogen in the process, so they'll also probably be the leader in that area too. Note that Elon has always said that H2 is a poor choice for energy storage, but he has never commented on its many critical applications in chemical production...

Just as Tesla used their experience and expertise with batteries to pivot into development of a whole new business unit (Tesla Energy), SpaceX will probably make a whole new business unit to leverage their hardcore engineering and manufacturing expertise in carbon sequestration / synthetic hydrocarbon production. SpaceX Energy would effectively become a synthetic oil and gas company that mines carbon from the atmosphere and hydrogen from the hydrosphere instead of getting both from the crust.

Note that SpaceX is conveniently headquartered in Texas, the undisputed leader of the hydrocarbon industry outside OPEC nations. Texas has a huge local supply of hydrocarbon processing engineers, organic chemists, technicians, and hydrocarbon finance and consulting specialists who are all going to be looking for new work pretty soon. Texas also has relatively minimal bureaucratic hurdles for permitting for chemical factory construction, many existing organic chemical factories, gas pipelines, gobs of wind and sunshine, and cheap land. Southeast Florida was another decent option for Starbase but they chose Texas and I believe a future plan for hydrocarbon synthesis was probably one of the undisclosed reasons in favor of Texas. Mordor gentrification is proceeding at a speed that will catch almost everyone off guard.

Transportation
This one probably needs little argument on this forum but the point is worth repeating.

Cheap autonomous EVs and cheap tunnels will reduce the cost of transportation by an order of magnitude in the long run while also increasing the utility of the transportation service provided by increasing convenience, reliability, speed, comfort/luxury, and relaxation and by freeing up people's attention from attending to the road.

Although this new system will substantially improve upon the energy consumption per passenger mile (or per shipping container mile) of today's transportation options, the increase in overall miles demanded will likely overwhelm the efficiency gains and increase the energy demands of transportation by an order of magnitude or maybe two. Consumers will commute longer distances, go out on trips more, take more faraway vacations, and order more deliveries. Businesses and other institutions will exploit the new technology in a similar manner.

Transportation currently accounts for about 20% of human energy use, so if transportation demand per capita increases by 10-100x and efficiency per passenger-mile improves by 3x, the net result is a 3-33x increase in human energy consumption from this factor alone. Then multiply this by bringing the other 80% of humanity to developed-world living conditions and the net change is 17-170x increase!

Hyperloop, long distance tunnels, Starship orbital transport, and/or EVTOL aircraft may also come about and provide similar benefits for long-distance transportation. I can promise you that the primary reasons I don't fly to Hawaii or visit my family more often are that I'd have to consume a bunch of oil, spend a bunch of money, and lose most of a day to travel. If those constraints were removed I'd do it probably monthly.

Heating, Ventilation & Air Conditioning
In order to save money or to help the environment (or in today’s world, to avoid funding Putin and Russia’s oligarchy) many homes and other buildings are kept at less than perfectly comfortable temperatures. In fact, I’ve been in several factories and warehouses and most of them didn’t have any climate control at all because of energy cost. If you’ve read this far, by now I’m sure you can see where this is going.

The HVAC market is going to explode, especially as extreme temperatures become more common and in particular as Europeans, Canadians and Russians start to install AC to cope with their increasingly uncomfortable summer heat.

As with cars, there will be some work to improve efficiency from heat pumps and radiant heat systems for example, but honestly I’m expecting the attitude to shift towards not caring because it isn’t worth the time and effort. Consider the average person’s relationship with electricity for low consumption devices. Turning off unused lights, computers and TVs was important 20 years ago, but these devices are an order of magnitude more efficient today so it’s arguably not even worth the time anymore unless you’re going on vacation.

Humans are organisms seeking to maintain homeostasis. Clever retail businesses have figured this out and used it as a trick to attract foot traffic. At Disney World it’s common to leave the front door of a shop or restaurant wide open in the middle of a Floridian summer because the cool air leaking out attracts customers like a porch lamp attracts moths. Energy consumption is the main reason why more places don’t do this already; it makes sense for Disney today because they have so much pedestrian traffic in front of the door and high average gross profit per customer entering the building.

The return on investment for building thermal insulation will be reduced, which will probably reduce the amount of insulation in the average building. A couple caveats include:
If designs using structurally insulated panels, like Boxabl walls with expanded polystyrene, end up becoming common, then the average thermal insulation level will increase.
Thermal insulation usually improves noise insulation so that may become the main driver for continuing the insulate buildings. However, if thermal insulation becomes less common then that will further drive up energy consumption for HVAC.

Even outdoor heating and cooling will increase. For example, outdoor heating lamps for restaurant and bar patios have a lifetime ownership cost dominated by the energy cost of usage. At least in Seattle, it is obvious that COVID majorly accelerated the move for extending restaurants/bars out onto the road where parking spaces used to be. Patrons and staff would prefer these tents be 72F/22C, all else being equal, and pedestrians passing by may appreciate the free relief from the weather. Imagine all the applications where people would want this: waiting lines at amusement parks, open air stadiums, festivals, etc.

This is a sector which will be especially impacted by economic growth in developing nations. India, for instance, is hot, but less than 10% of the 1.3 billion Indians currently have AC. Ouch. Here's some data on AC market penetration by nation. And if you haven't heard, India is getting hotter. They will install AC. As it turns out, developing nations tend to be closer to the Equator and have higher temperatures and are being disproportionately impacted by global warming. There's a similar story for most of South Asia, the Middle East, the entire continent of Africa, and most of the Americas south of the Rio Grande.

There is a bit of a natural limit to how much HVAC demand would likely increase, because we are mainly interested in keeping buildings comfortable, and there's only so many buildings. Heating and cooling currently accounts for a bit over half of global final energy consumption, and my rough guess is that energy used for HVAC will quadruple from today's levels in the mid-to-late 21st century, becoming in total about double today's total human energy consumption.

Information & Communications Technology
For the last century, there's been steady exponential growth of the energy efficiency of computers in terms of floating point operations per second per watt. The improvement rate has been an order of magnitude every 5 years, in accordance with Koomey's Law (link1 & link2).

Likewise, information transmission technology has been getting more efficient in terms of bytes transmitted per second per watt by around an order of magnitude every 5 years (see fig 3 in this study for an example from Finnish cell networks).

Despite this impressive rate of improvement, the total amount of energy consumed by Information and Communications Technology (ICT) has also been growing exponentially, to the point that this sector which barely existed 25 years ago now accounts for more than 10% of overall human electricity use and 2% of overall carbon emissions with the exponential growth likely to accelerate in the coming decades with continued growth in users, 5G, cloud services, HD video streaming, Internet of Things, and AI services.

This is extremely difficult to predict precisely because the trajectory is highly sensitive to assumptions about the trajectory of the overall ICT industry. However, it is very safe to predict than even in the lowest end scenarios the ICT sector will consume vastly more energy than it does today because the ICT industry is still just getting started in the grand scheme of things. Probably between 10x and 1000x by 2050, making it consume 0.2x to 20x more power than current human activities. I would guess 2x as a decent middle estimate.

Manufacturing & Construction
Energy costs constitute a substantial portion of the cost structure for many manufacturing industries, such as metallurgy and ceramics where energy can be 30% of the total cost.

This is really hard to estimate the impact of but it's safe to say that factories will find ways to use more energy. This would probably include selection of more energy-intensive materials, usage of higher performance machines that are less efficient, and using energy to squeeze out more product yield with less scrap material. Some of it will also just come from being able to sell more widgets by passing on the energy savings to the customer.

Construction is essentially outdoor manufacturing so it has similar economics.

Desalination & Pumping Water against Gravity
I posted about this recently.

I’ve been thinking about this a lot lately and I think the only viable plan is desalination on a tremendous scale.

We have more water on this planet than we could possibly use. The problem is that almost all of our H2O molecules are jumbled together with salt ions and other impurities in big pools located downhill from where we live.

The Sun’s energy has always solved this problem for us by generating evaporation, wind, clouds, rain and rivers. This has always been frustratingly inconsistent, unpredictable and unevenly distributed across Earth’s landmass. With climate change, desertification, deforestation, erosion and biodiversity loss these issues are getting significantly worse every year, and our food supply is currently dependent on unsustainable water draw rates. That’s very bad news.

Some very good news is that we can actively use the same solar energy to bypass the natural hydrological cycle to synthesize as much freshwater as we want and pump it to wherever we want and even put it in solid or gaseous form if we want. In principle, it’s physically possible to artificially refill all our lakes, rivers and streams with clean, 100% reliable water supply. With electricity we can remove the salt and other impurities and then pump the water uphill. The process is already pretty well established, with over 21,000 desalination plants in operation worldwide today, mostly concentrated in the Middle East where they have nearby saltwater, cheap oil and gas, and severe freshwater scarcity.

The problem with desalination and water transportation has always been the cost and environmental impact of the energy and pipeline requirements. A breakthrough in cheap sustainable energy and underground pipeline construction costs is needed. How else will we irrigate the Midwest when we’ve finishing sucking the Ogallala Aquifer dry? How else will East Africa survive when Lake Victoria is nothing more than a memory?

Solar energy is soon going to be cheaper per Joule, by an order of magnitude, than hydrocarbon fuels ever have been anywhere in the world. So the energy portion is already likely to be solved in a timely manner.

Underground aqueducts already exist. Mexico City, for instance, relies heavily on their pipeline infrastructure and they recently completed a 39-mile wastewater removal tunnel. Sadly, these subterranean aqueduct projects take years to complete, cost around $1 billion per mile and have insufficient flow rates to really make a dent in the overall global water crisis.

Can we fix that part? Suppose that Boring Company hits their technical goals for Prufrock and then follows that up with a MegaPrufrock variant with a 4-meter radius. It might cost 5x more per mile than regular Prufrock for a cost of about $25M per mile fully outfitted with pumps. With a 50 m^2 cross sectional area and a 20 m/s flow rate (we don’t care much about pump energy consumption in this future) it could transport 1000 m^3 of water per second.

For comparison, the discharge rate of the Amazon River is 200k m^3/s. If we wanted to deliver an equivalent flow of water, we could do it with just 200 of these aqueducts at a cost of just $5B per mile. No crazy technology involved, just big pipes with big internal propellers.

Building one such aqueduct from the Gulf of Mexico to Nebraska 1000 miles away would cost $25B. Nebraska today has about 1M irrigated acres of farmland. Our pipeline can deliver 1 acre-foot of irrigation (1200 m^3) every 1.2 seconds. Corn needs about 2 acre-feet per season.
1M acres * 2 feet irrigation * 1.2 sec/acre-ft / 86400 sec/day = 28 days of aqueduct flow to grow 1M acres of corn.​
Wow, that actually sounds feasible!

Lake Mead has a maximum capacity of 32 billion m^3. At our estimated 1000 m^3/s flow rate, Lake Mead could be completely filled from empty in one year with a single pipe. It’s 270 miles from the ocean. At $25M/mile the pipe would be $7B to construct.

I am fairly convinced by now that if Boring Co can actually hit their preposterous goals and Wright’s Law for solar development holds true for another decade or two, we can provide plenty of water for our own needs, directly fix our wildlife reserves and reverse desertification. Tesla would play a massive role in powering this infrastructure because they will be the leading energy company by the time any of this would happen.

I want to add some more hard numbers.

The theoretical minimum energy consumption of reverse osmosis filtration is about 1 kWh per cubic meter of water. Actual efficient reverse osmosis plants today typically achieve about 3 kWh/m^3 (link).

Generating an Amazon River worth of freshwater via desalination would require (220k m^3/s) * (3 kWh/m^3) = 660,000 kWh/s = 2.4 TW.

In the USA, freshwater consumption is approximately 280 billion gallons per day, which is 390 billion m^3/year or 12k m^3/s (link). In other words, we collectively use about 5% of an Amazon River. Producing all of this via desalination would require on the order of 130 GW or on the order of 1 PWh annually. If SWB power costs get down to $1/MWh, this would cost $1B annually in electricity expenses.

However, since we're imagining a future in which industries are exploiting low energy costs, so reverse osmosis would probably lose to multistage flash distillation, which in a design made to be simple and cheap at the expense of energy efficiency would consume more like 25 kWh/m^3. This would drive up the electricity need to more like 8 PWh per year which would cost more like $8B/year, but that's still a tiny amount of cost compared to the value to the US economy.

The theoretical minimum energy consumption of piping water one kilometer vertically is about 2.7 kWh/m3 for doing the work against gravity. In reality it might require 10x or 100x more because in most cases requiring lifting water 1 km above sea level will involve energy loss from imperfect propellers and fluid friction. I don't know how to estimate that. Roughly I would estimate 10 PWh for pumping required annually or around the same as the distillation energy consumption.

Total would be about 20 PWh.

This could scale easily another order of magnitude if it starts to become necessary to supplement our glaciers, rivers, streams and lakes with additional water to compensate for climate change impacts and actively fight desertification and wildfire risk. The US gets about 8 trillion m^3 of precipitation annually, which is 20x more than our freshwater consumption.

Indoor Farming
Conventional agriculture is in a crisis state on many fronts.
  • Poor soil health and erosion
  • Shifting climate and with it drought, flooding, storms, plant and livestock diseases, and more
  • Water shortages
  • Wildlife habitat encroachment
  • Pesticides, herbicides and fungicides
  • Gradually declining nutritional value and flavor
  • Waste from failed crops, slight product imperfections, transportation damage, spoilage
  • Lack of interest from rural youth to enter the business
  • Natural gas dependence for fertilizer synthesis
  • Fertilizer runoff and downstream problems like algal blooms
  • Oil dependence for machinery like tractors
  • Countryside landscapes rendered boring and ugly
  • Pollinator and insect population collapse
  • Long supply chains, especially during off season locally
Some of these problems can be solved directly by replacing fossil fuels with SWBs, but the overall system is still a slow-moving catastrophe that's getting worse every year.

Indoor farming, or controlled environment agriculture (CEA), has clear benefits for many of these problems.

  • Soil isn't even needed for some types of CEA and when it is required, the soil is easy to maintain in controlled conditions
  • Total weather protection
  • 90-99% water efficiency such that the primary way H20 molecules exit the facility is via the produce itself
  • High density per acre/hectare even when accounting for required solar panel area
  • Pesticides, herbicides, and fungicides generally aren't necessary because the environment is treated as a clean room which prevents intrusion of weeds and pathogens
  • Perfect growing conditions, high freshness at time of consumption, and ability to select different varieties of consumer favorites make nutrition and flavor much higher
  • Crops rarely if ever fail, produce is extremely consistent and marketable, less transportation needed as produce can be grown locally year round, and less time between harvest and consumption drastically cuts spoilage
  • Can be located in urban areas where many young people are interested in farming but can't find affordable land
  • Near zero fertilizer waste and negligible amounts escape facility into the environment, and as mentioned in the chemical synthesis section ammonia can be produced with solar and that can feed the fertilizer supply chain
  • Traditional farm machinery not even needed; all equipment in indoor farm will be electrical
  • Can give countryside back to wildlife or at least make pretty parks and communities for people who prefer rural living
  • Long supply chains generally unnecessary
Here is video interview of someone who farms citrus and pomegranate trees in a cheap greenhouse he made after retiring from a career at the Post Office. In Nebraska. All year, even when it hits -20F temperatures outside. His system is low tech and uses passive geothermal heating with a fan blowing air through a big tube loop buried underground. I would bet someone can improve on something this and scale it up. If nothing else, it's proof of concept that places like Montreal or Helsinki could potentially grow tropical crops locally 365 days a year. The question is cost.


Depending on who you ask, electricity costs are usually around 60% of the total cost structure for controlled environment agriculture. Energy cost is also embedded in the materials and construction for the building itself. Labor is another ~10-30% depending on the crop and the level of automation. Water input is usually 90-99% less than conventional agriculture and this advantage is growing over time as indoor farms improve efficiency while field agriculture is increasingly vulnerable to heat waves, periods of drought, deteriorating soil health, and long-term desertification which mitigates the benefits of innovation in conventional water management for farms. Human labor is not required by the laws of physics for growing and harvesting plants and being in a factory environment makes automation easier than traditional farms. In theory with sufficient automation much of the labor component of cost could be removed. That means that the whole cost structure for indoor farming may be poised to plummet in the next couple decades and it will need a lot of electricity.

Now what if the farms also have electric transport via long distance underground freight tunnels that maybe go 100+ mph and electric trucks for last mile delivery? The most energy-intensive crops could be grown in places like Arizona and North Africa and shipped out to less sunny places regionally for significantly less cost, better freshness, more resilience to weather disruptions to surface transport, and negligible environmental impact compared with refrigerated diesel trucks in use today. Plus, solar farms in hot arid deserts can help restore ecosystem health on land damaged by prior human activities by providing shade and concentrating rainwater. For the crops with the most energy intensity, this could be a winning strategy.

The biggest question for CEA in my opinion is whether it will be able to win in the really big markets like wheat, soy, corn, rice, sorghum, potatoes and yams. Thus far CEA is mainly competitive only for fruits, vegetables and cannabis.

It’s hard to estimate the total impact of CEA on future energy demand but I think it’s probably at least 200W/m^2 of continuous power for lighting and temperature control, so one hectare of total grow area would require at least 1 MW of electricity. For a sense of scale, 5 billion hectares are cultivated globally today.

Rough napkin math:
  • Assume CEA will have 10x more net productivity per hectare of cultivated area because of its conditions perfectly optimized for growth 24/7/365, no loss of crops to diesease/pests/weather, consistently perfect marketable produce, and lower post-harvest food waste
  • Assume human population will grow to 10 billion and malnutrition will be solved, increasing food consumption by 30% from today
CEA to replace all legacy plant growing would require 5B hectares * 1.3 * (1/10) = 650M hectares of grow area (note: not land area because farms will have like 10-1000 levels of plant shelves depending on the plant height and the factory size)

At 1 MW/hectare, this CEA industry would consume 650 TW of power, which is about 30x more power than humanity uses today, yet it would require only about 1.3B hectares of land for solar panels with 20% efficiency compared with 5B hectares currently cultivated. Also, unlike today’s monoculture row crop farms, solar farms can coexist with moderately healthy ecosystems with proper biodiversity. This is all very rough estimation but it suggests than CEA energy consumption could someday be at least an order of magnitude more than current human energy consumption.

Optimus
This is a wildcard but is straightforward to estimate.

What if the Bot actually works really well and there’s eventually an army of ten billion of them (wild guess of one bot per person) and each one consumes 5 kWh per day on average? That’s another 50 TWh of electricity needed daily or about 20 PWh (peta) per year, which is 10% of our current consumption.

If we have 10 bots per person it’d be 200 PWh per year which would by itself double human energy consumption.

100 bots per person --> 2000 PWh/yr --> 20x human energy consumption

Moonshot Possibilities Not Worth Writing About in Detail
  • Atomic recycling, usually via application of extreme heat (e.g. pyrometallurgy)
  • Oceanic trash cleanup
  • Antimatter synthesis
  • Asteroid mining for precious metals (assuming Starship works and hits targets for the other costs not related to fuel and assuming other technical challenges solved)
  • Other ideas yet to be invented or so obscure I don’t know about them
  • Green mining and refining with tunnel boring machines instead of open pit techniques and brute force separation of atoms, Get notes from boring co spreadsheet

Conclusion
There is almost no understanding yet about how many problems the SWB revolution will solve, how big the energy market will get, and the stupendous potential of Tesla Energy to be the biggest player in this new energy future.

Adding up the rough estimates in the preceding sections leads to a very rough prediction of 30x-200x of increase in energy demand from today's levels.

The profitability potential is reduced by the fact that margins on a low cost commodity product can’t be large because the unit revenue is low. Will the explosive volume growth outweigh this factor? It’s hard to guess in advance. Even if energy cost falls 10x, if the overall volume grows 100x, then the revenue will be up 10x. Considering the nature of a SWB grid and the importance of load scheduling, weather simulation and other software services, I could easily see Tesla turning Autobidder and its ancillary battery management software into a dominant global energy management platform like AWS is for cloud computing.

Tesla's skill with integration and smooth consumer product experiences, akin to Apple's consumer electronics ecosystem, may give them a sustainable competitive advantage in making fully integrated home energy systems including Solar Roof, Cars, Powerwall and HVAC. I see no one likely to catch them in the next 15 years minimum, and we know how fast they can scale production.

Do not sleep on Tesla Energy. It will be way bigger than car hardware if you look out to 2040 and beyond. While estimating with precision is difficult, I see potential for Tesla Energy to be a $10+ trillion business by itself.

This is not investment advice. The author is not a certified professional financial advisor. Make your own decisions.
 
Last edited:
dunno, I bought some more at $642 but clearly I could have waited for $625. I had fear of missing out (FOMO) though and just picked a number.
A Douglas Adams inspired pick?😎

I bought another 25 at 625. I didn’t remember putting in that limit order until it executed! Thank you Silver Oak. 🤪

BTW, how many times can one look in the mirror and say “NOW I’ve finally reached the number of shares of TSLA with which I’m satisfied and I can simply sit back and HODL?” Asking for a friend.
 

The Renewables Revolution

Why Mainstream Projections for Future Human Energy Consumption are Wrong by at Least an Order of Magnitude


***************************************************************************************************************************

Introduction: The Energy Situation
You and I are engaged in a perpetual war against entropy. The tyrannical Laws of Thermodynamics fundamentally guarantee the usefulness of concentrated controllable energy because it can do work for us by generating:
  • Concentrated heat
  • Mechanical force
  • Photon emission
  • Endothermic chemical reactions
  • Electromagnetic field changes
As it turns out, these impressive and physically irreplaceable capabilities are why we need continual energy supply to have a modern industrial civilization. The blog Wait but Why has a couple of fun and informative summary articles on energy for dummies and Tesla’s role in the future of energy.

Understanding Tesla Energy’s potential first requires understanding the total addressable market, and I’m not talking about the 180 petawatt-hours currently used by humans annually (link).

Around 1800, humanity began to transition from a traditional biomass energy economy (wood mainly) that had been in place for millions of years of hominid evolution. The new industrial sector of the economy used industrialized energy sources, starting originally with coal and water wheels used primarily in Great Britain initially and shortly after in Continental Europe and the United States and then later worldwide. A century later we added more sources to the mix: oil, natural gas, hydroelectric, and finally nuclear. As the chart below clearly shows, the biomass energy market at the dawn of the Industrial Revolution was two orders of magnitude smaller than today's total market for post-industrial energy sources. On the one hand, world population is up 8x since 1800, so per capita energy consumption globally is up "only" about 13x. However, most of us would aspire that everyone on the planet can have a standard of living like the middle class in OECD countries.

Global average per capita power used in 1800 was ~200 watts, whereas in the developed countries today it's more like 4,000-10,000 watts. This has been a roughly 20x-40x increase in consumption per capita and there is plenty of appetite for more, as shown by data indicating that the most affluent consumers within developed nations tend to use far more energy than those with lesser means. This is also significantly understating the true energy cost of a typical OECD lifestyle, because much of the energy usage is occurring in countries with heavy manufacturing bases like China and India that sell primarily in OECD markets.

If we want a super prosperous future for humanity even just with today's energy usage patterns, then 10 billion people in 2050 will consume ~10 kW of power per capita which is 100 TW or about 5x what humanity consumes today.

Many environmentalists, including me, have found this fact alarming, because of the problems with our current energy mix and because of the inextricable links between per capita energy consumption and economic prosperity, human rights fulfillment, health, life expectancy, social harmony, and scientific progress. For decades it has seemed as though the Universe handed us a cruel choice between impending catastrophe from peak oil and ecological collapse or self-inflicted catastrophe from economic and agricultural collapse. Was modern industrial civilization doomed from the start to be a brief party before the inescapable reality of thermodynamics and finite fossil fuel reserves put us into a death spiral of energy shortages, famine, war and extreme weather events?

After years of anxious researching of this topic that led to me investing in TSLA originally, I have changed my mind entirely. That energy consumption curve is about to go vertical like we've never seen before, and surprisingly that’s good news not only for the economy and poverty, but also for the biosphere, world peace, and of course Tesla Energy's cash flow statement.

The Stone Age did not end because we ran out of stones, and the Oil Age will not end because we ran out of oil.

View attachment 807416

Cost Collapsing & Doomed Legacy Energy “Assets”
Nuclear fission energy was famously hailed in 1954 by Lewis Strauss, then serving as the Chairman of The Atomic Energy Commission, for being on course to provide energy “too cheap to meter” in the near future. Infamously, Strauss’ prediction was wrong. The details of why are beyond the scope of this forum, but we've been here before with grandiose claims of energy abundance. Why is it different with the solar, wind and batteries (SWB) electric grid model?

At this point the main thing to recognize is that SWB electric generation, unlike the nascent nuclear industry in 1954, has decades of proven empirical cost decline trends with no slowdown in sight. While nuclear development has long been unfairly limited by social and political opposition by well-meaning but misinformed people, the time has passed for that because all legacy power plant architectures are headed straight for obsolescence in the 2020s. All of them, end of story. The SWB revolution is disrupting the entire category of thermal power stations, irrespective of the heat source being coal, natural gas, oil, atomic fission, atomic fusion, geothermal heat, herds of hamsters running on wheels, or magic Vibranium crystals gifted to us by Wakanda for $0. The old power plant model involves vaporizing water to crank a turbine to spin a dynamo to drive transformers for voltage step-up to transmit power long distances on high-voltage AC lines to power more transformers for voltage step-down before finally something useful happens with the energy. On average two thirds of the electric power generated at the dynamo is lost as waste (mostly resistive heating) en route to the end user.

Legacy power plants need to be big, centralized and far from population centers, necessitating this ridiculous electricity distribution apparatus. Solar PV panels in particular are notoriously scalable and emission free, so they can go on buildings or be built in medium size farms or in big farms that can be located close to the demand.

The old model is steadily losing economic competitiveness via exponential decay (reverse S-curve/negative logistic function), like all technologies being disrupted by a new paradigm. They will predictably flounder, attack SWB with media disinformation and FUD, whine and cry for government bailouts, possibly receive some degree of bailouts, and then fail anyway when they can't even achieve positive gross margins and their competition gets 10% cheaper every year.

Moreover, before long even functional coal/gas/nuclear/oil/geothermal plants with capital costs already paid for will start retiring earlier than TSLA hodlers. Why? Their operating costs will be more expensive than additional replacement SWB expansion, even if we continue the glorious time-honored tradition of refusing to pass laws to charge appropriate fees to operators of such plants for the unpriced negative externalities of their activities. It gets even worse for legacy plant operators when we add in the fact that localized solar and batteries generation will be cheaper in most populated areas than the mere operating costs of the big regional grids needed for legacy centralized power stations to be practical. All those towers, transmission wires, and substations ain’t cheap and the existing infrastructure is aging. RethinkX calls this price threshold “GOD parity” (Generation On Demand). Game over.

This is not theoretical, seeing as wave after wave of early coal plant shutdowns in the United States have already been happening for several years even during the Trump Administration when many EPA regulations on coal were relaxed, and gas plants are now facing an unstoppable battery onslaught. Tesla’s big splash into the South Australian energy market was a warning shot. The Hornsdale Battery Reserve wiped out the vast majority of the local Frequency Control Auxiliary Services market—previously ruled by a gas plant operator cartel—and it happened almost as soon as the switch was flipped on. Even in Texas, that great bastion of low regulation and unbridled free market capitalism, that place with a uniquely Wild West electricity bidding market in ERCOT, that place which happens to have an endowment of some of the world’s cheapest natural gas, is leading the USA in total SWB deployment. Tony Seba and the RethinkX team are correctly predicting the greatest stranded asset write off in the history of capitalism happening in the 2020s.

The writing is on the wall, or at least the writing is on this chart that you might want to print out and affix to your wall to gaze at for peace and serenity whenever you stare at your brokerage account feeling panic and dismay about a TSLA drop:

View attachment 807406

This empirical trend has been remarkably steady for 7 decades since the invention of the modern silicon photovoltaic cell in 1954 by researchers at Bell Labs, which was basically the Tesla of the 20th century. That's Wright's Law for you.

Similar trends have been seen for wind and batteries.

Wind hit the scene with competitive prices sooner than solar did by about 10 years. However, wind is much more exquisitely sensitive to local microclimate and to daily weather fluctuations, because the extractable power is roughly proportional to the wind speed cubed. 10 knots wind speed can provide about 8x more electricity than 5 knots wind speed. Wind also has more impact on landscape aesthetics, long distance plant reproduction, and bird migration patterns. Worst of all, compared to solar, wind has a much slower rate of cost decline and vastly less total available power globally by about 4 orders of magnitude. Still, it's good enough to help the cause and there's lots of wind out there in places like the North American Great Plains and the European North Sea.

Batteries are the dark horse coming to market the latest of the three. The first prototype lithium-ion cell was made by Stanley Whittingham in 1974, 20 years after the first solar PV cell. Exxon bought the technology IP and gave up on it, so modern Li-ion cells really didn't get going until John Goodenough and Akira Yoshino's work in the late 1980s presented a better design using cobalt oxide and carbon for the cathode and anode. Using these advances, Sony in 1991 began sales of the first practical Li-ion battery cell, kicking off development of an industry that led to Whittingham, Goodenough and Yoshino jointly winning a Nobel Prize in Chemistry. The costs since then have been plummeting according to a Wright's Law relationship between volume and cost:

View attachment 807614
Source

Consider the technology and cost projections Tesla disclosed on Battery Day, and further consider that:

1) They were probably sandbagging​
2) They were probably withholding information and not revealing all their technology​
3) Two years is a LONG time for Tesla’s technology to progress​

I think it's safe to say that the cost decline trend is going to hurtle well past $100/kWh and could conceivably hit less than $50/kWh with known plausible technology development pathways.

Nonsensical Mainstream Forecasts
The UN Sustainable Development Goals for 2050 contain some excellent ideas that we might want to try. However, I disagree with important aspects of the energy goals. They are aiming for moderate increases in total energy consumption as the developing world industrializes and they focus a lot of attention on improving energy efficiency and consuming less energy in the rich world (good luck convincing average people to do that or vote for people who will force them to do it). I believe this is a severe projection error that is happening because the UN analysts are discounting the implications of the solar energy cost trend, either because of some judgment error on their own part or because the implications seem so outrageous to a casual observer that it's not politically viable to publish what they actually think.

If you look into forecasts from other major government and NGO think tanks, you get similar estimates, such as:
This ridiculously inaccurate BS is influencing opinions, influencing the economy and business planning, and influencing public policy. I don't normally like to throw shade at strangers, but this is dangerous incompetence that is increasing existential risk for humanity and it needs to be called out. Improperly influencing public sentiment, lawmaking and capital allocation from these positions of responsibility will indirectly kill millions of people and thrust tens or hundreds of millions more into years of extreme poverty they didn't need to suffer. This is morally unacceptable and I hope it's due to mere incompetence instead of malice. In any case, the outcome is the same and I am heavily voting with my capital that these forecasts are wrong.

Just look at some of these slides from organizations that Wall Street considers credible sources for energy projections. They are forecasting:
  • Linear growth of renewables for a quadrupling by 2050 to 25% of the global energy supply
  • Linear growth of the world economy
  • Stable global demand for coal
  • 40% growth of oil and gas demand by 2050
  • 30% EVs in the global fleet by 2050
  • Increasing global CO2 emissions
  • Declining energy intensity per $ of GDP of 30% by 2050
It needs to be said: These projections are pathetic. These analysts are getting 1st-year undergraduate level statistical analysis wrong. Solar, wind and battery growth trends have been for many decades following almost perfect and instantly recognizable exponential growth trends. If you plot it on a log scale it's almost a perfect line with a linear regression having R^2 of more than 0.9! How can these official reports be published where the long term historical growth trends aren't even shown and where they assume without any justification that the trend will suddenly switch from exponential to linear? If you're going to project a radical change in the growth trajectory then you had better tell the world why you think that instead of just riding off your credibility. That's how science is supposed to work! You have a hypothesis, then provide data and analysis to prove it.

It's no wonder the finance and consulting industries don't understand Tesla. They actually believe this nonsense.

View attachment 808677

View attachment 808678
View attachment 808682

View attachment 808666

Disruptive Technology Majorly Expands Markets
Disruptive technology expands markets, because of higher utility and/or lower costs. It is never a one-to-one substitution, but human psychology tends to make these market expansions and new use cases surprising to most people, including experts. If you haven't seen Tony Seba present on this you need to watch now. We humans often underestimate the collective cleverness of our species to find new ways to exploit resources.


In retrospect, the market for paper information storage and transmission in 1970 was, to say the least, an extremely poor indicator of the size of the future market for digital electronic information storage and transmission. Anyone predicting that computers would merely replace existing paper-based use cases would have been embarrassingly mistaken. And there were many such people, and they were indeed embarrassingly mistaken. Only those with imagination who reasoned from first principles had anything resembling accurate forecasts, and even they tended to majorly underestimate the trajectory of growth in bytes of information generated, stored and transmitted. If I could go back in time and tell Alan Turing that humanity in 2020 would have a worldwide web transferring 3 zettabytes of data annually between over a billion computers made from carefully organized sand, he would’ve probably fainted from shock. But here we are now when I have a handheld pocket computer that can hold 256 GB, which is the equivalent of 54 billion words of written English.

Likewise, the market for horse and buggy transport was an extremely poor indicator of the size of the future market for automobile transport. Same for other technology disruptions, including examples like:
  • Film photography --> Digital
  • Vinyl music storage --> Tape --> CD --> MP3
  • Hunting/gathering food --> Agriculture
  • Human-borne freight with hands and shoulders --> Human-borne freight with baskets & backpacks --> Wheeled freight pulled by humans --> Wheeled freight pulled by domesticated animals --> Wheeled freight hauled by automobiles
Every time something better and more efficient and cheaper comes along, we end up using it far more than the preceding technology generation, unless another subsequent technology disruption ends up eliminating the want/need altogether. The empirical evidence from not only recorded history of the last 8,000 years but indeed from millions of years of hominid evolution dating back to our stone tool technology disruptions indisputably shows that technological disruptions always tend to follow this pattern. This is fundamentally the main reason why GDP per capita improves over time. The fact that regulated market economies tend to promote such technological advancement is one of the primary reasons they produce more overall prosperity than other social systems, despite having some well documented downsides.

When illustrated this way, the point seems like common sense, but human psychology is not built to intuitively understand exponential growth and has a bias for assuming that life in the future will look similar to life today, because the future is intangible and requires imagination and because making bold predictions about the status quo changing comes with risk to social status.

This pattern is also predicted by basic economic theory: when technology improves the combined "total factor productivity" of land, labor and capital such that supply of a certain good or service has increased, the market clearing price decreases and the total quantity sold increases. The “price elasticity of demand” for a good/service measures how much the quantity demanded increases for a given decrease in market price (link). More formally, it’s the slope (1st derivative with respect to price) of the demand curve at any point. We can only speculate about what the true shape of the demand curve for energy looks like at the costs we're talking about, because it's uncharted territory with no readily available econometric data. However, we have some clues.

The inflation-adjusted price of energy has actually been increasing since WW2 with gradual exploitation of the easy fossil hydrocarbon reserves and the easy hydroelectric opportunities and with the rise in population and industrialization, but the quantity sold has still been growing. Another factor in demand has been the advances in other complementary technologies that have increased the economic utility per unit of energy consumed. So, what happens when the population still keeps rising and the rate of industrialization accelerates globally and technology continues to increase the utility per Joule consumed, while at the same time a new type of energy drastically increases supply?

Energy consumption patterns in such cities as Dubai, Riyadh and Abu Dhabi may provide a glimpse of what changes are coming. For instance, Dubai gets 90% of its freshwater supply from desalination of the Arabian Gulf and Dubai also has a 2000 square meter indoor ski resort/snow park at the Mall of the Emirates. Furthermore, perhaps we should keep watching what happens in major Middle Eastern cities they will be on the leading edge of the SWB revolution due to their strong sunshine and cheap land giving them the cheapest solar power of any big cities in the world.

So...why are professional analysts and organizations assuming that the SWB revolution will leave a market comparable to today's energy market? This contradicts all available evidence and theory about the dynamics of technology disruptions and specific potential use cases for cheap SWB energy. This is the same type of flawed thinking that Tony Seba’s team has shown to typically occur for any tech disruption, and it's unsurprisingly happening again.

As far as I can tell:

We are not headed for a net-zero carbon economy with moderate total energy consumption increases by (hopefully) 2050 if and only if governments and citizens cooperate in an unprecedented display of human trust, solidarity, and selfless cooperation.

I think we are headed for a carbon negative economy by 2050 because of SWB, even if we forecast with the most cynical possible assumptions about selfish profit-seeking and personal-quality-of-life-improving motives dominating behavior. I also think the transition to the new economy will increase energy consumption roughly similar to the 100x jump from our ancestors had in transitioning from biofuels to fossil fuels/hydro/nuclear.

The Sun continuously bestows 173 petawatts of power to Planet Earth’s crust and currently we use about 10,000x less power than this (4 orders of magnitude). Additionally, in outer space at Earth’s distance from the Sun, solar energy flux of 1.4 kW/m2 is waiting for us in practically unlimited quantities if we eventually want to reserve Earth’s precious surface area for other purposes, and can be beamed down to Earth’s surface with microwaves and captured with rectifying phased array antennas. Starship and off-Earth manufacturing may make this economical someday. With enough energy at a sufficiently affordable cost, there is practically no limit to the human economy in this solar system or any other solar system. Stars ☀️ are mind bogglingly powerful.

We are also not going to reduce our average energy intensity per unit of GDP in the process.

In developed economies, like the USA for example, the energy intensity of GDP (average joules per $ of GDP produced) has been slowly declining for decades, but I think the SWB revolution will lead to a monumental reversal of this trend. The steepest drop in energy intensity occurred during the 1970s energy shock, and energy prices have on average creeped up in the time since. With SWB power the price of energy falling will mean more energy is used for relatively low value applications compared to now, which implies that value produced per joule will drop. Although some individual energy use cases like computation and transportation will get more efficient per FLOP or per mile, this effect will be heavily outweighed by the growth of the market and by the development of new energy-intensive industrial capacity and infrastructure that exploits the delightfully cheap SWB energy of the future.

Moreover, this staggering increase in energy supply will help us actively fix the biosphere without crashing our economies, while delivering all the other benefits of SWB that the UN goals already include, like achieving a wide variety of social justice objectives, avoiding hazardous pollution, reducing radiation exposure from coal emissions, reducing root causes of warfare, etc. It will also eliminate our terrible dependency on oil and gas for chemical purposes while increasing the supply of said chemicals. The SWB revolution will be as profoundly disruptive as the Scientific Revolution and subsequent First Industrial Revolution.

What new potential is waiting to be unlocked by lower prices? Some have been discussed already by others and me, but the math is telling me that almost EVERYBODY is underestimating the magnitude of what’s coming.

Primary Chemical Production
We need chemicals. All kinds of them. Lots of them.

Too much of our chemical supply comes from oil and gas, but if there's one thing I know in my limited understanding of chemical engineering, it's that almost everything is negotiable if you’re willing to pay the price of fighting entropy by expending enough energy. Many chemical synthesis processes are known and fairly well researched but have been rightfully shelved because of their wasteful energy consumption. When energy suddenly explodes in abundance and affordability, that changes everything. We are about to enter a new era of brute force chemical synthesis driven by cheap SWB power.

Energy-intensive processes exist to synthesize all of the following primary chemicals using solar power and abundant material inputs including primarily air, water, salt (NaCl) and limestone (calcium carbonate CaCO3):
  • Methane
    • CH4
    • Sabatier Process
  • Ammonia
    • NH3
    • Haber-Bosch Process
  • Hydrogen
    • H2
    • Water electrolysis or Chloralkali Process
  • Hydroxide
    • OH-
    • Water electrolysis
  • Carbon monoxide
    • CO
    • Steam reforming of CH4 or High-temperature electrolysis of CO2
  • Lye / Sodium hydroxide
    • NaOH
    • Chloralkali Process
  • Chlorine
    • Cl2
    • Chloralkali Process
  • Ethanol
    • CH3CH2OH
    • Fermentation of sugar with yeast or new electrochemical process published in 2017
  • Sodium carbonate
    • Na2CO3
    • Obtainable by mining, but also can be synthesized with the Solvay Process
  • Quicklime
    • CaO
    • High-temperature calcination of limestone (CaCO3)
  • Ozone
    • O3
    • Coronal Discharge Method or Electrolysis of H2O
Out of all the chemicals on this list, only sodium hydroxide and ozone are currently mass-produced with electric power, with the rest directly relying on coal and natural gas as either feedstocks or heat sources.

With these basic building blocks and ultra-low cost of energy, almost any imaginable chemical can be synthesized. As a matter of fact, today’s chemical industry already makes extensive use of the chemicals in this list to knit together bigger molecules. I would encourage you to go explore Wikipedia for any of these chemicals to see the possibilities. These primary chemicals can also be used to synthesize a wide variety of derivative chemicals currently produced by cracking petroleum, because syngas of CO + H2 + steam can go through the Fischer-Tropsch process to form liquid hydrocarbons we typically get from oil. Other pathways exist too, and I'm not enough of an expert in this area to know which is best. However, I am fairly confident that cheap SWB energy and cheap H2 and CH4 will make it economically competitive to produce key chemicals like olefins, higher alkanes, carbon monoxide, formaldehyde, and aromatics.

I do not have the requisite expertise in chemical engineering to estimate how much of the cost is affected by energy compared to other factors, but the outlook is good especially as the hydrocarbon mining industry dwindles and loses economies of scale. Also, there may be ways to exploit cheap energy with tradeoffs that reduce these other costs.

It could get even crazier. If we get to a point where we can use machine learning models like AlphaFold to reliably design proteins in a reasonable timeframe and then use genetic engineering techniques like CRISPR to insert the corresponding gene sequences into microorganisms for precision fermentation, in principle we could theoretically make a practically unlimited supply of our own customized organic catalysts to unlock even more potential for a sustainable abundant chemicals industry. Among the benefits of customized enzymes would be lower cost, higher yield, improved energy efficiency, reduced undesirable byproducts, and higher throughput. This is a long shot because protein folding and enzyme behavior are extremely complicated to model computationally and even advanced AI may not be able to solve protein design in a practical computational budget. However, the known laws of physics don’t preclude the possibility and if it works it will literally catalyze the growth of the chemical synthesis industry even more than cheap SWB energy alone.

Methane and hydrogen synthesis from CO2 and H20 could by themselves drive a 10x or more bump in energy consumption because it requires splitting H20 and CO2 molecules which has very low thermodynamic efficiency since those molecules are extremely stable. As Terraform Industries founder Casey Handmer wrote in a whitepaper explaining the purpose of the startup and doing the physics calculations:

"Indeed, the quantity of electricity required to synthesize 100% of California’s natural gas demand exceeds grid consumption by more than 10:1. This means that synthetic fuel price parity will drive enormous increases in demand for solar power deployment, providing the demand necessary to keep the learning curve bending downwards."

Plus, this 10x estimate is probably understating the impact of increased consumption of methane and hydrogen driven by lower cost, again in line with the Law of Demand that predicts higher quantity demanded when price is lower.

Methane is just one of the many chemicals we need, and a huge variety of other chemicals can be derived from H2 and CH4. As far as I'm aware, we can in theory replace the main outputs of the entire petrochemical stack with air, sunshine, water, and common minerals at potentially lower cost. This matters.

Terraform Industries is a very early stage startup, and they have a competitor who also is surely aware of where solar prices are headed: SpaceX.


We already knew this though, because this method is the only possible means of in situ methane production for fuel of habitat and rockets on Mars. Elon said on the recent Everyday Astronaut interview with Tim Dodd that when Starship/Super Heavy is fully and rapidly reusable, the propellant expenses become most of the cost structure. Elon also said that the sole figure of merit for the rocket, booster, launch and manufacturing teams is to work together to reduce cost per ton to orbit. Obviously this means they will soon devote HEAVY, RELENTLESS attention and engineering prowess towards developing high throughput, low cost, environmentally friendly CO2 + H20 --> CH4 + O2 factories. What people are missing is that this also implies that SpaceX might expand to adjacent markets, which happens to include...all uses of oil and gas that can't be directly replaced with SWB power. Methane is methane. CH4 is a global commodity. There's nothing special about "rocket fuel" methane except perhaps higher purity. Ironically, the Musk Foundation may technically need to award the $100M X-Prize for carbon capture technology to SpaceX!

No one is likely to outcompete SpaceX in any market segment where they devote the majority of their engineering resources, and once full and rapid reusability is working, methane production will be the next target. I think only Terraform has a shot at beating SpaceX by getting a head start while SpaceX focuses on Starship/Super Heavy, but then again maybe they'll be acquired by SpaceX like Hibar, Grohmann and Maxwell were acquired by Tesla.

Yes, you read that right. SpaceX looks to be in the early stages of becoming the world's largest and most profitable hydrocarbon producer. Potentially in the long run SpaceX Hydrocarbons could make today's Saudi Aramco look tiny, because SpaceX will have even lower production and distribution costs than Saudi oil & gas, and there's vastly more available carbon in the atmosphere than in Saudi crustal oil & gas reserves. SpaceX will likely get even more bonus pricing power vs. mined hydrocarbons to the extent that governments provide carbon sequestration subsidies. This is probably 20 or 30 years away but it seems almost inevitable if you look at the historical track record of SpaceX and other Musk companies and the overarching goal to "maximize the probability that the future is good".

SpaceX will need to make green hydrogen in the process, so they'll also probably be the leader in that area too. Note that Elon has always said that H2 is a poor choice for energy storage, but he has never commented on its many critical applications in chemical production...

Just as Tesla used their experience and expertise with batteries to pivot into development of a whole new business unit (Tesla Energy), SpaceX will probably make a whole new business unit to leverage their hardcore engineering and manufacturing expertise in carbon sequestration / synthetic hydrocarbon production. SpaceX Energy would effectively become a synthetic oil and gas company that mines carbon from the atmosphere and hydrogen from the hydrosphere instead of getting both from the crust.

Note that SpaceX is conveniently headquartered in Texas, the undisputed leader of the hydrocarbon industry outside OPEC nations. Texas has a huge local supply of hydrocarbon processing engineers, organic chemists, technicians, and hydrocarbon finance and consulting specialists who are all going to be looking for new work pretty soon. Texas also has relatively minimal bureaucratic hurdles for permitting for chemical factory construction, many existing organic chemical factories, gas pipelines, gobs of wind and sunshine, and cheap land. Southeast Florida was another decent option for Starbase but they chose Texas and I believe a future plan for hydrocarbon synthesis was probably one of the undisclosed reasons in favor of Texas. Mordor gentrification is proceeding at a speed that will catch almost everyone off guard.

Transportation
This one probably needs little argument on this forum but the point is worth repeating.

Cheap autonomous EVs and cheap tunnels will reduce the cost of transportation by an order of magnitude in the long run while also increasing the utility of the transportation service provided by increasing convenience, reliability, speed, comfort/luxury, and relaxation and by freeing up people's attention from attending to the road.

Although this new system will substantially improve upon the energy consumption per passenger mile (or per shipping container mile) of today's transportation options, the increase in overall miles demanded will likely overwhelm the efficiency gains and increase the energy demands of transportation by an order of magnitude or maybe two. Consumers will commute longer distances, go out on trips more, take more faraway vacations, and order more deliveries. Businesses and other institutions will exploit the new technology in a similar manner.

Transportation currently accounts for about 20% of human energy use, so if transportation demand per capita increases by 10-100x and efficiency per passenger-mile improves by 3x, the net result is a 3-33x increase in human energy consumption from this factor alone. Then multiply this by bringing the other 80% of humanity to developed-world living conditions and the net change is 17-170x increase!

Hyperloop, long distance tunnels, Starship orbital transport, and/or EVTOL aircraft may also come about and provide similar benefits for long-distance transportation. I can promise you that the primary reasons I don't fly to Hawaii or visit my family more often are that I'd have to consume a bunch of oil, spend a bunch of money, and lose most of a day to travel. If those constraints were removed I'd do it probably monthly.

Heating, Ventilation & Air Conditioning
In order to save money or to help the environment (or in today’s world, to avoid funding Putin and Russia’s oligarchy) many homes and other buildings are kept at less than perfectly comfortable temperatures. In fact, I’ve been in several factories and warehouses and most of them didn’t have any climate control at all because of energy cost. If you’ve read this far, by now I’m sure you can see where this is going.

The HVAC market is going to explode, especially as extreme temperatures become more common and in particular as Europeans, Canadians and Russians start to install AC to cope with their increasingly uncomfortable summer heat.

As with cars, there will be some work to improve efficiency from heat pumps and radiant heat systems for example, but honestly I’m expecting the attitude to shift towards not caring because it isn’t worth the time and effort. Consider the average person’s relationship with electricity for low consumption devices. Turning off unused lights, computers and TVs was important 20 years ago, but these devices are an order of magnitude more efficient today so it’s arguably not even worth the time anymore unless you’re going on vacation.

Humans are organisms seeking to maintain homeostasis. Clever retail businesses have figured this out and used it as a trick to attract foot traffic. At Disney World it’s common to leave the front door of a shop or restaurant wide open in the middle of a Floridian summer because the cool air leaking out attracts customers like a porch lamp attracts moths. Energy consumption is the main reason why more places don’t do this already; it makes sense for Disney today because they have so much pedestrian traffic in front of the door and high average gross profit per customer entering the building.

The return on investment for building thermal insulation will be reduced, which will probably reduce the amount of insulation in the average building. A couple caveats include:
If designs using structurally insulated panels, like Boxabl walls with expanded polystyrene, end up becoming common, then the average thermal insulation level will increase.
Thermal insulation usually improves noise insulation so that may become the main driver for continuing the insulate buildings. However, if thermal insulation becomes less common then that will further drive up energy consumption for HVAC.

Even outdoor heating and cooling will increase. For example, outdoor heating lamps for restaurant and bar patios have a lifetime ownership cost dominated by the energy cost of usage. At least in Seattle, it is obvious that COVID majorly accelerated the move for extending restaurants/bars out onto the road where parking spaces used to be. Patrons and staff would prefer these tents be 72F/22C, all else being equal, and pedestrians passing by may appreciate the free relief from the weather. Imagine all the applications where people would want this: waiting lines at amusement parks, open air stadiums, festivals, etc.

This is a sector which will be especially impacted by economic growth in developing nations. India, for instance, is hot, but less than 10% of the 1.3 billion Indians currently have AC. Ouch. Here's some data on AC market penetration by nation. And if you haven't heard, India is getting hotter. They will install AC. As it turns out, developing nations tend to be closer to the Equator and have higher temperatures and are being disproportionately impacted by global warming. There's a similar story for most of South Asia, the Middle East, the entire continent of Africa, and most of the Americas south of the Rio Grande.

There is a bit of a natural limit to how much HVAC demand would likely increase, because we are mainly interested in keeping buildings comfortable, and there's only so many buildings. Heating and cooling currently accounts for a bit over half of global final energy consumption, and my rough guess is that energy used for HVAC will quadruple from today's levels in the mid-to-late 21st century, becoming in total about double today's total human energy consumption.

Information & Communications Technology
For the last century, there's been steady exponential growth of the energy efficiency of computers in terms of floating point operations per second per watt. The improvement rate has been an order of magnitude every 5 years, in accordance with Koomey's Law (link1 & link2).

Likewise, information transmission technology has been getting more efficient in terms of bytes transmitted per second per watt by around an order of magnitude every 5 years (see fig 3 in this study for an example from Finnish cell networks).

Despite this impressive rate of improvement, the total amount of energy consumed by Information and Communications Technology (ICT) has also been growing exponentially, to the point that this sector which barely existed 25 years ago now accounts for more than 10% of overall human electricity use and 2% of overall carbon emissions with the exponential growth likely to accelerate in the coming decades with continued growth in users, 5G, cloud services, HD video streaming, Internet of Things, and AI services.

This is extremely difficult to predict precisely because the trajectory is highly sensitive to assumptions about the trajectory of the overall ICT industry. However, it is very safe to predict than even in the lowest end scenarios the ICT sector will consume vastly more energy than it does today because the ICT industry is still just getting started in the grand scheme of things. Probably between 10x and 1000x by 2050, making it consume 0.2x to 20x more power than current human activities. I would guess 2x as a decent middle estimate.

Manufacturing & Construction
Energy costs constitute a substantial portion of the cost structure for many manufacturing industries, such as metallurgy and ceramics where energy can be 30% of the total cost.

This is really hard to estimate the impact of but it's safe to say that factories will find ways to use more energy. This would probably include selection of more energy-intensive materials, usage of higher performance machines that are less efficient, and using energy to squeeze out more product yield with less scrap material. Some of it will also just come from being able to sell more widgets by passing on the energy savings to the customer.

Construction is essentially outdoor manufacturing so it has similar economics.

Desalination & Pumping Water against Gravity
I posted about this recently.



I want to add some more hard numbers.

The theoretical minimum energy consumption of reverse osmosis filtration is about 1 kWh per cubic meter of water. Actual efficient reverse osmosis plants today typically achieve about 3 kWh/m^3 (link).

Generating an Amazon River worth of freshwater via desalination would require (220k m^3/s) * (3 kWh/m^3) = 660,000 kWh/s = 2.4 TW.

In the USA, freshwater consumption is approximately 280 billion gallons per day, which is 390 billion m^3/year or 12k m^3/s (link). In other words, we collectively use about 5% of an Amazon River. Producing all of this via desalination would require on the order of 130 GW or on the order of 1 PWh annually. If SWB power costs get down to $1/MWh, this would cost $1B annually in electricity expenses.

However, since we're imagining a future in which industries are exploiting low energy costs, so reverse osmosis would probably lose to multistage flash distillation, which in a design made to be simple and cheap at the expense of energy efficiency would consume more like 25 kWh/m^3. This would drive up the electricity need to more like 8 PWh per year which would cost more like $8B/year, but that's still a tiny amount of cost compared to the value to the US economy.

The theoretical minimum energy consumption of piping water one kilometer vertically is about 2.7 kWh/m3 for doing the work against gravity. In reality it might require 10x or 100x more because in most cases requiring lifting water 1 km above sea level will involve energy loss from imperfect propellers and fluid friction. I don't know how to estimate that. Roughly I would estimate 10 PWh for pumping required annually or around the same as the distillation energy consumption.

Total would be about 20 PWh.

This could scale easily another order of magnitude if it starts to become necessary to supplement our glaciers, rivers, streams and lakes with additional water to compensate for climate change impacts and actively fight desertification and wildfire risk. The US gets about 8 trillion m^3 of precipitation annually, which is 20x more than our freshwater consumption.

Indoor Farming
Conventional agriculture is in a crisis state on many fronts.
  • Poor soil health and erosion
  • Shifting climate and with it drought, flooding, storms, plant and livestock diseases, and more
  • Water shortages
  • Wildlife habitat encroachment
  • Pesticides, herbicides and fungicides
  • Gradually declining nutritional value and flavor
  • Waste from failed crops, slight product imperfections, transportation damage, spoilage
  • Lack of interest from rural youth to enter the business
  • Natural gas dependence for fertilizer synthesis
  • Fertilizer runoff and downstream problems like algal blooms
  • Oil dependence for machinery like tractors
  • Countryside landscapes rendered boring and ugly
  • Pollinator and insect population collapse
  • Long supply chains, especially during off season locally
Some of these problems can be solved directly by replacing fossil fuels with SWBs, but the overall system is still a slow-moving catastrophe that's getting worse every year.

Indoor farming, or controlled environment agriculture (CEA), has clear benefits for many of these problems.

  • Soil isn't even needed for some types of CEA and when it is required, the soil is easy to maintain in controlled conditions
  • Total weather protection
  • 90-99% water efficiency such that the primary way H20 molecules exit the facility is via the produce itself
  • High density per acre/hectare even when accounting for required solar panel area
  • Pesticides, herbicides, and fungicides generally aren't necessary because the environment is treated as a clean room which prevents intrusion of weeds and pathogens
  • Perfect growing conditions, high freshness at time of consumption, and ability to select different varieties of consumer favorites make nutrition and flavor much higher
  • Crops rarely if ever fail, produce is extremely consistent and marketable, less transportation needed as produce can be grown locally year round, and less time between harvest and consumption drastically cuts spoilage
  • Can be located in urban areas where many young people are interested in farming but can't find affordable land
  • Near zero fertilizer waste and negligible amounts escape facility into the environment, and as mentioned in the chemical synthesis section ammonia can be produced with solar and that can feed the fertilizer supply chain
  • Traditional farm machinery not even needed; all equipment in indoor farm will be electrical
  • Can give countryside back to wildlife or at least make pretty parks and communities for people who prefer rural living
  • Long supply chains generally unnecessary
Here is video interview of someone who farms citrus and pomegranate trees in a cheap greenhouse he made after retiring from a career at the Post Office. In Nebraska. All year, even when it hits -20F temperatures outside. His system is low tech and uses passive geothermal heating with a fan blowing air through a big tube loop buried underground. I would bet someone can improve on something this and scale it up. If nothing else, it's proof of concept that places like Montreal or Helsinki could potentially grow tropical crops locally 365 days a year. The question is cost.


Depending on who you ask, electricity costs are usually around 60% of the total cost structure for controlled environment agriculture. Energy cost is also embedded in the materials and construction for the building itself. Labor is another ~10-30% depending on the crop and the level of automation. Water input is usually 90-99% less than conventional agriculture and this advantage is growing over time as indoor farms improve efficiency while field agriculture is increasingly vulnerable to heat waves, periods of drought, deteriorating soil health, and long-term desertification which mitigates the benefits of innovation in conventional water management for farms. Human labor is not required by the laws of physics for growing and harvesting plants and being in a factory environment makes automation easier than traditional farms. In theory with sufficient automation much of the labor component of cost could be removed. That means that the whole cost structure for indoor farming may be poised to plummet in the next couple decades and it will need a lot of electricity.

Now what if the farms also have electric transport via long distance underground freight tunnels that maybe go 100+ mph and electric trucks for last mile delivery? The most energy-intensive crops could be grown in places like Arizona and North Africa and shipped out to less sunny places regionally for significantly less cost, better freshness, more resilience to weather disruptions to surface transport, and negligible environmental impact compared with refrigerated diesel trucks in use today. Plus, solar farms in hot arid deserts can help restore ecosystem health on land damaged by prior human activities by providing shade and concentrating rainwater. For the crops with the most energy intensity, this could be a winning strategy.

The biggest question for CEA in my opinion is whether it will be able to win in the really big markets like wheat, soy, corn, rice, sorghum, potatoes and yams. Thus far CEA is mainly competitive only for fruits, vegetables and cannabis.

It’s hard to estimate the total impact of CEA on future energy demand but I think it’s probably at least 200W/m^2 of continuous power for lighting and temperature control, so one hectare of total grow area would require at least 1 MW of electricity. For a sense of scale, 5 billion hectares are cultivated globally today.

Rough napkin math:
  • Assume CEA will have 10x more net productivity per hectare of cultivated area because of its conditions perfectly optimized for growth 24/7/365, no loss of crops to diesease/pests/weather, consistently perfect marketable produce, and lower post-harvest food waste
  • Assume human population will grow to 10 billion and malnutrition will be solved, increasing food consumption by 30% from today
CEA to replace all legacy plant growing would require 5B hectares * 1.3 * (1/10) = 650M hectares of grow area (note: not land area because farms will have like 10-1000 levels of plant shelves depending on the plant height and the factory size)

At 1 MW/hectare, this CEA industry would consume 650 TW of power, which is about 30x more power than humanity uses today, yet it would require only about 1.3B hectares of land for solar panels with 20% efficiency compared with 5B hectares currently cultivated. Also, unlike today’s monoculture row crop farms, solar farms can coexist with moderately healthy ecosystems with proper biodiversity. This is all very rough estimation but it suggests than CEA energy consumption could someday be at least an order of magnitude more than current human energy consumption.

Optimus
This is a wildcard but is straightforward to estimate.

What if the Bot actually works really well and there’s eventually an army of ten billion of them (wild guess of one bot per person) and each one consumes 5 kWh per day on average? That’s another 50 TWh of electricity needed daily or about 20 PWh (peta) per year, which is 10% of our current consumption.

If we have 10 bots per person it’d be 200 PWh per year which would by itself double human energy consumption.

100 bots per person --> 2000 PWh/yr --> 20x human energy consumption

Moonshot Possibilities Not Worth Writing About in Detail
  • Atomic recycling, usually via application of extreme heat (e.g. pyrometallurgy)
  • Oceanic trash cleanup
  • Antimatter synthesis
  • Asteroid mining for precious metals (assuming Starship works and hits targets for the other costs not related to fuel and assuming other technical challenges solved)
  • Other ideas yet to be invented or so obscure I don’t know about them
  • Green mining and refining with tunnel boring machines instead of open pit techniques and brute force separation of atoms, Get notes from boring co spreadsheet

Conclusion
There is almost no understanding yet about how many problems the SWB revolution will solve, how big the energy market will get, and the stupendous potential of Tesla Energy to be the biggest player in this new energy future.

Adding up the rough estimates in the preceding sections leads to a very rough prediction of 30x-200x of increase in energy demand from today's levels.

The profitability potential is reduced by the fact that margins on a low cost commodity product can’t be large because the unit revenue is low. Will the explosive volume growth outweigh this factor? It’s hard to guess in advance. Even if energy cost falls 10x, if the overall volume grows 100x, then the revenue will be up 10x. Considering the nature of a SWB grid and the importance of load scheduling, weather simulation and other software services, I could easily see Tesla turning Autobidder and its ancillary battery management software into a dominant global energy management platform like AWS is for cloud computing.

Tesla's skill with integration and smooth consumer product experiences, akin to Apple's consumer electronics ecosystem, may give them a sustainable competitive advantage in making fully integrated home energy systems including Solar Roof, Cars, Powerwall and HVAC. I see no one likely to catch them in the next 15 years minimum, and we know how fast they can scale production.

Do not sleep on Tesla Energy. It will be way bigger than car hardware if you look out to 2040 and beyond. While estimating with precision is difficult, I see potential for Tesla Energy to be a $10+ trillion business by itself.

This is not investment advice. The author is not a certified professional financial advisor. Make your own decisions.

You have more useful content here than the sum of all my posts.
 
I feel safer investing at $630 today than $54 in 2019 when Tesla still had much to prove.

Let me tell you what happened this evening during my drive to our downtown Rec Centre. I was travelling at 50 kmh in a 50 zone w. Adaptive Cruise Control (ACC) enabled. All of the sudden, the car (seemingly spontaneously) beeped loudly while slamming on the brakes, and came to a complete stop in the middle of my lane.

My 1st reaction was, WTF?! What's going on? Why are we stopping? Only then did I see that a pedestrian had RUN out across the street (jaywalking), and out in front of my car which was in his opposing lane! I did nothing, this save was full credit to Tesla Vision, and Automatic Emergency Braking.

So I glanced out the driver's side window, and I saw 2 people standing at a city bus stop from where this person had bolted out onto the street. The people were shaking their heads, like they'd seen this before. My 'after-the-fact' impression is that this person was trying to get run over to get an insurance settlement (or at least free room'n'board at the local hospital for a spell).

So this was NOT a phantom braking event. This was my Tesla saving him, me, and my car from a lot of damage, pain and paperwork. Much appreciated, Elon and the Tesla Autopilot Team. U Rock!

Tomorrow, I think I'll buy 30 shares of TSLA in exchange for a job well done.

Cheers!
 
was my Tesla saving him, me, and my car from a lot of damage, pain and paperwork. Much appreciated, Elon and the Tesla Autopilot Team.
Hopefully insurance scammers learn not to run in front of Teslas as they find out about autopilot/AEB and that the dashcam feature. Another advantage for the Tesla insurance business.
 

The Renewables Revolution

It has been a far too-long time since a well-presented document on the future of energy has crossed my desk. It is an absolute delight to see this fire hydrant of a post.

I do notice a number of items upon which I care to comment. Few of my remarks will be of the “well done!” variety, so in advance of those, let me say “Well done!”. Rather, most will be ones with which I take issue.

This post looks at your first two chapters, “The Energy Situation“ and “Cost Collapsing”. I’ll try to keep up with others over the next days, as I find time.

By the way, as long-timers are aware but newcomers likely are not, back in the mid-Pliocene I received my PhD in Resource Economics; my earlier degrees were in (Hi-T / Hi-P) geochemistry, in geology and in….East Asian Studies. But I am primarily self-taught.

Energy Situation -

A fine précis of the history of energy consumption. The sole - but very large - statement of yours with which I take issue is this paragraph:

If we want a super prosperous future for humanity even just with today's energy usage patterns, then 10 billion people in 2050 will consume ~10 kW of power per capita which is 100 TW or about 5x what humanity consumes today.

This does not address other consumptives, albeit ones that also use some of the power you discuss. I do not discuss “nature” but rather mention only the Big Three traditionally used to represent humankind’s bare necessities:
*Food - meaning of course agricultural production, with its demand of land and other inputs;
*Water - use by man of which both stresses other parts of the biosphere and otherwise, or in addition, will consume an immense chunk of that ~10kW/pp;
*Shelter - the demand for wood, concrete and other inputs also rise immensely were humanity to enjoy a super-prosperous future. Whence comes the global wood inputs? And concrete entails far more than the petajoules addressable by an energy revolution. For example: WHERE are the requisite sources of sand?


Cost Collapsing & Doomed Legacy Energy “Assets”

The graph of Ramez Naam’s that you provide is a log-log graph. I’ve never met a log-log that I didn’t love look at with as much askance as with a kaleidoscope on a roller coaster. The reason is that just about anything looks like a straight line when plotted log-log. Amusingly, the BNEF data on that chart gives the lie to what I just wrote (though with the correspondingly lowest Rˆ2).

But I also don’t find data corroborating that something that costs less gets more customers as hugely illuminating. I think you provide a similar such chart, I think out of ARK, further on.

What definitely is from ARK is the graph of Li-ion battery costs over time. As shown, it appears that they enjoyed a massive diminution in cost from 1985 to 2005…and then not only stagnated, but are provided only to 2012. Yet you write of the data as demonstrating Wright’s Law in action. I think the subsequent chart was (ARK’s not yours) attempt to Invoke Wright’s Law, but overall it seemed rather….goopy…to me.

To be - possibly - continued.
 
https://aseannow.com/topic/1260627-elon-musk’s-tesla-registers-subsidiary-in-thailand/

Great news for those of us living in Thailand! I have been reluctant to pull the trigger without Tesla establishing an official presence and all that entails here in The Land of Smiles. Perpetually gridlocked Bangkok in particular desperately needs to transition to EVs to reduce its toxic air quality. As for those of us in Phuket, it’s the perfect island vehicle.

Some collateral impacts may be:

1) increased speed of deliveries with Thailand being a member of ASEAN. Shipping from Shanghai will be quite speedy and efficient.

2) Heightened evaluation of Thailand as another manufacturing base for Tesla in Asia. It’s a huge auto market with an extensive vehicle manufacturing base, affordable labor and supply chain expertise. This could help reduce risks associated with the predictable increase in tensions between the PRC and the West in the coming decade.

3) I’ll be really really happy! 😎
 
Follow up:

Page 23 of this slide deck:

Energy Revenue: $616M
Energy Costs: $688M

@The Accountant does great projections and summaries of actual results in this thread:
There is very little worthwhike data collection in near real time regarding either solar or storage sales, whether by Tesla or other players . We struggle even to get an approximate global data point annually and about 3-4m late, let alone any decomposition of global data that can be trusted.

This is an important gap for wider understanding, but not actionable in an investment context by almost anyone, which is one reason for the data gap.

Previously I have assessed (from memory as I am on a phone) that Tesla is getting about a third of global utility storage, but a trivial % of global solar, and maybe a tenth of residential storage.

At current electricity prices (which imho are an abnormalshort term aberattion to the high side that will somewhat unwind) non-tesla residential storage is becoming cost effective as an unsubsidised personal investment. In the UK one can fit 10kwh of non tesla storage for gbp 5k giving a 10-yr payback or better, with combined 4-6ke solar and 10kwh storage systems at gbp 10k with IRR at 7-10%. Not shabby and you can see explosive growth beginning.

We can deduce that Tesla is likely capacity constrained more than its residential storage competitors by the combination of price (Tesla is 2x the competition in price), and the lead time ( Tesla is 10-12m with difficult availability, whereas competition are 2-3m with fair availability).

What we don't know is to what extent Tesla growth rate is lagging or leading competition and whether the residential sector story (with respect to trsla) is replicated in the utility storage sector. I find it very difficult to get reliable data to assess that.

On margins we can see that NM is neutral (0) so if GM us good (?) Then overheads must be a concern. Since the Asian competitors are expanding quickly then someone is doing something right.

Availability and price of the competition vs tesl on the domestic storage sector is so difficult that I am myself seriously considering ordering non-Tesla.

This sectir is Teslas to lose, but I'm afraid they seem to be losing ground. But as I have said data is not available and not very believable. I understand why auto gets priority, but it is a shame to watch this play evolving as it seems to me to be more a matter of poor execution than of lack of capital.
 
This is not investment advice. The author is not a certified professional financial advisor. Make your own decisions.
Amazing write-up! Kudos to your effort. I think very very few utility professionals understand this, and most of those who understand are not in a position of great influence. Utilities are moving at the speed of ICE Auto right now, and will get majorly disrupted!