Does Elon Musk Embody First Principles Reasoning?
I’ve been invited to participate in a short fiction contest for Earth Day 2019. Prior to writing, the organizers suggested participants read an article by James Clear titled, "First Principles: Elon Musk on the power of thinking for yourself".
The article discusses the power of problem solving by reasoning from first principles. According to Wikipedia, first princples are basic, foundational, self-evident propositions or assumptions that cannot be deduced from any other proposition or assumption. The article says of first principles, “[f]irst principles thinking is the act of boiling a process down to the fundamental parts that you know are true and building up from there”. There were two claims in the article I disagree with.
First, the it makes the claim that reasoning from first principles is how scientists think, which is only partially true. Scientists employ several other powerful forms of reasoning to generate new concepts, modify old ones, and solve difficult problems. Dynamic congnitive processes underlying a myriad of reasoning techniques is the hallmark of the tentative nature of science.
Second, the article claims of Elon Musk, “…[that] no one embodies the philosophy of first principles thinking more effectively…”. The article backs this claim up by providing an example involving Elon’s development of SpaceX which has allegedly led to his ability to develop and use rockets cheaper than anyone before him. Despite what we’ve read or heard, the fact of the matter is that the jury is still out on just how much cheaper if at all his rockets actually are. And as charismatic as Elon is, it’d be prudent to anything he claims with a grain of salt.
Vertical takeoff, vertical landing (VTVL) technology has been around for decades, first tested in the early 1990s. One of the biggest hurdles that has faced engineers with regard to this technology is that retaining enough fuel to land, halves a rocket’s efficiency both in energy return on energy invested (EROEI), and in payload capacity. This latter issue has a proportionally negative impact on financial return on investment (ROI). Reusable technology has also been around for decades; e.g. the Space Shuttle.
But these points aren’t where I take contention. Instead, its Elon’s blatant ignorance of what his rockets could do to accelerate an already rapid global warming trend. Elon is well aware of the scientific consensus that the current accelerated global warming trend causing climate change stems from our use of fossil fuels. Fossil fuels are organic, and all perfect organic combustion reactions create carbon dioxide gas; a greenhouse gas. I explain how this gas acts to warm the planet in my blog post, “The ‘Greenhouse’ Gas, Carbon Dioxide”.
One form of rational thought invoked by scientists is critical reasoning, and I’d surmise such thinking would lead one to the conclusion that in order to mitigate global warming would be to substantially reduce our use of fossil fuels. Yet, Elon is promoting unprecedented use of his fossil-fuel-burning rockets worldwide for everything from establishing and supplying a colony on Mars, to expansive commercial international rocket travel.
We can use this so-called first principles thinking to breakdown about how much carbon dioxide gas would be produced from a single Big Falcon Rocket (BFR) launch, then consider that number against the backdrop of the tens of thousands of annual rocket launches Elon invisions in the not-too-distant future.
The balanced combustion equation for methane and oxygen is:
There are 2 booster stages on the BFR; a Super Heavy first stage, and a Starship second stage. The propellant capacity of the second stage is listed online as being 1.1 million kilograms (kg). The propellant capacity of the first stage doesn’t appear to be listed, but I’ve deduced its capacity to be around 2.85 million kg based on other mass figures provided online. Therefore, we can reasonably presume the total propellant mass (methane and liquid oxygen) of a fully-fueled BFR is approximately 3.95 million kg.
We can see from the balanced combustion equation above that for every 1 molecule of methane used, 1 molecule of carbon dioxide is produced for perfect combustion. We can therefore determine how much carbon dioxide is produced from a launch by looking at the molar mass of methane. We do this by adding up the atomic weights of each atom in a methane molecule.
There is 1 carbon atom (12 atomic mass units), and 4 hydrogen atoms (1 atomic mass unit each). Combined, this gives methane a molar mass of 16 grams per mole (g/mol). SpaceX lists the mass of methane in a fully-fueled second stage booster as 240,000 kg. They don’t provide this information for the first stage, but since they list the second stage propellant mass as being 21.8% methane, then we can use that percentage to calculate the first stage methane mass as being about 621.300 kg.
The total mass of methane for a fully-fueled BFR is therefore ~861,300 kg. Since 1 mole of methane is 16 grams, we know a fully-fueled BFR contains 53,831,250 moles of methane. As shown in the balanced combustion equation above, the methane-to-carbon dioxide ratio is 1:1, so we can determine the total mass of carbon dioxided produced by multiplying 53,831,250 by the molar mass of carbon dioxide, and this gives us 2,368,575,000 grams, or 2,368,575 kg of carbon dioxide.
That’s about 2,611 tons.
Despite what some say with regard to how efficient SpaceX’s Raptor engines powering the BFR will be, the 2nd law of thermodynamics tells us there is no such thing as a perfect heat engine. So I’ll cut the total tonnage of carbon dioxide produced by 60%, giving a new total of 1,044 tons per launch.
At first glance, this reduction seems like the lesser of two evils, however an inefficient fossil-fuel-burning heat engine produces a host of toxic products, including carbon monoxide, soot, and oxides that combine with water vapor to form sulfuric and nitric acid. The acids tend to fall back to Earth as acid rain putting carbon sinks like vegetation at risk. Carbon monoxide reacts with hydroxyl radicals in the atomosphere which negatively affects these radicals ability to reduce the lifespan of atmospheric methane gas.
The longer atmospheric methane can survive, the more potent it becomes as a greenhouse gas since it has more time to function as a greenhouse gas. Hydroxyl radicals also react with many pollutants, effectively ‘removing’ them from the air. Because of this, there is an inverse relationship between the amount of hydroxyl radicals in the atmosphere and the amount of pollution.
Also, carbon monoxide can react with hydroxyl radicals to produce carbon dioxide which brings us full circle. The combustion of fossil fuels, regardless of efficiency, invariably results in the production of the greenhouse gas carbon dioxide, which has always been one of the primary drivers of global climate. I explain this latter point in my blog post, “Climate Change - Carbon Dioxide has always been a primary driver”.
The tonnage of carbon dioxide produced per launch, both directly and indirectly is substantial. Worse still, a large portion of it will be directly injected into the sensitive upper layers of the atmosphere. And SpaceX is on track to launch more rockets than any other country in the world.
Furthermore, Elon intends to implement a new space-based Internet communication system comprised of nearly 12,000 satellites. He has dubbed it Starlink, and hopes to have the system fully operational in the next several years. SpaceX expects to transport 20 to 40 satellites per launch, which means it will require 300 to 600 launches. This doesn’t include the number of launches needed for system maintenance requirements.
Elon also wants to implement what he calls Earth-to-Earth BFR, where he intends to use the BFR to transport people around the globe. He has said he foresees a future where you look at your watch, rather than a calendar for launch schedules, which implies his intention to see multiple launches per day.
Then there is his plan to colonize Mars, which would require multiple launches to establish, and sustain. He wants to see a colony of 1 million people living on Mars in the not-too-distant future, and said it will require tens of thousands of launches to support until the colony could become self-sufficient.
Should these plans become reality, the amount of additional carbon dioxide spread throughout Earth’s atmosphere would send global climate careening in a direction the planet hasn’t experienced since the Paleocene-Eocene Thermal Maximum about 56 million years ago.
Elon knows there is scientific consensus that the current accelerated global warming trend is fueled by anthropogenic carbon dioxide emissions, yet he is actively pursuing ventures that will exponentially increase these emissions. In my humble opinion, this sort of rationality is ipso facto not the embodiment of any kind of logical reasoning.
We need to decarbonize industry, not actively promote a new industry based entirely on fossil fuel.
I will contend that Elon’s track record of scientific honesty is likewise subpar. True scientists will present their data without bias for peer review. However, I’ve seen Elon withhold pertinent data, or misrepresent it without being questioned.
An example of withholding pertinent data would be the fact Tesla has not released the vehicle dry (unloaded) weight of their electric semi truck. The United States Department of Transportation (USDOT) enforces a highway weight limit of 36,287 kg for semi trucks. The heavier a semi truck’s dry weight, the less available cargo weight it will be allowed to haul. This has direct implications for this vehicles ROI, and I suspect these semis weigh several tons more than the average semi. This coupled with the fact Tesla’s electric semis have a driving range as little as 1/5 that of other semis, chances are both the ROI and EROEI on these vehicles is well below industry standards.
Solar, wind, geothermal, and hydropower are not viable options for converting electricity in many regions of the United States. As such, many charging stations will likely be powered by coal-burning power plants. The electricity required to keep an electric semi truck with truncated cargo capacity, and limited range on the road could equate to higher carbon dioxide emissions should their electricity come from non-”green” sources. And on that point, there is no such thing as “green energy”. It all comes at an environmental cost, which is why I’ve been writing for 20 years that a powerful first step towards reducing carbon dioxide emissions is to reduce per capita energy consumption. But what do I know?
As for the Tesla semi truck, I’m unable to make any solid claims as to its potential benefits or drawbacks precisely because Elon has not (or will not) share the vehicle’s dry weight. It’s a statistic that is far more critical to know than how fast these trucks can accelerate. It may be there are some cases where going full electric isn’t as environmentally friendly as it seems.
An example of Elon misrepresenting data is when he claimed his Boring Company completed a tunnel under Hawthorne, California for a fraction of the industry standard cost. In fact, he suggested he did it for 1/100 the cost by telling a crowd people it costs the industry upwards of $1 billion per mile, but he did his 1.14-mile-long tunnel for just $10 million.
This was met with applause but no questions. I see a very obvious trend wherein both the media and laity accept what Elon tells them as axioms. We can deduce the truth of Elon’s assertion by simply looking at the diameter of the Hawthorne tunnel. It isn’t more than 4 meters across, therefore it’s diameter is 2 meters which gives it a cross section of about 12 square meters.
The following is a graph published in the Journal of Geotechnical and Geoenvironmental Engineering (2001), showing tunnel excavation costs (Fig. 1). The Y-axis is cost in U.S. dollars per meter, and the X-axis represents tunnel cross section. Costs include excavation and support costs:
Even 200 square meter tunnels in complicated faulted ground requiring extensive structural support doesn’t cost $1 billion per mile. What Elon didn’t share with the crowd, is that the $1 billion/mile figure is an extreme upper-end cost that includes everything: the excavation and support costs, utilitites, conveniences (e.g. restrooms, restaurants, escalators, stations, finish work, etc.).
An example of a $1 billion/mile tunnel might be something like this:
Elon’s tunnel looks like this:
To be fair, the Mayakovskaya subway station, and the Moscow subway system in its entirety cost nowhere near $1 billion/mile. In fact, if we look at some of the most impressive tunnels around the world, we find that the $1 billion-per-mile cost is almost unheard of.
Even the Gotthard Base Tunnel through the Alps, the Channel Tunnel (“Chunnel”) that goes as deep as 75 meters below the sea bed between France and the UK, or the Seikan Tunnel in Japan that goes 100 meters below the sea bed cost anywhere near $1 billion/mile. So Elon’s comparison is patently misleading.
In fact, if we plot Elon’s tunnel on Hoek’s graph as I’ve done below (Fig. 2), we find that his tunnel cost more than even the average tunnel excavation and support costs worldwide:
We see that the estimated minimum cost worldwide for a tunnel with the same cross section as Elon’s Hawthorne tunnel is about $500/meter. That equates to $804,670 per mile. Elon’s tunnel cost $10 million for 1.14 miles, which is about $9 million per mile. This means Elon’s tunnel cost 11 times more than the estimated minimum cost worldwide.
The average tunneling cost worldwide for a tunnel with the same cross section as Elon’s is about $4,500 per meter, which equates to $7,242,030 per mile. That’s over $1.7 million less than Elon’s tunnel. In other words, Elon’s tunnel doesn’t even compete with the industry average. His misrepresentation of data is not a hallmark of what it is to think or act like a scientist.
There are also technologies that Elon has proposed that do not appear to consider some fundamental thermal physics. In his Hyperloop Alpha white paper, he proposes constructing an above-ground tube linking Los Angeles to San Francisco. This tube would contain a near-vacuum environment through which passenger pods would travel. His way around the Kantrowitz limit is to equip the pods with nose-end electric compressor fans capable of transferring high-pressure air build up from in front of the pods to their rears. This would theoretically allow pods to travel faster and more efficiently due to near negligible air resistance.
The tube itself would be welded together from prefabricated sections and set on top of pylons above ground. The tube would invariably experience thermal expansion at different rates throughout its length. For instance, the top of sections of tube exposed to sunshine will experience greater expansion than their shaded bottom halves. This non-linear expansion would result in sections of tube convexing toward the sky.
Elon’s proposed solution is to install “…telescoping tube[s], similar to the boxy ones used to access airplanes at airports … at the end stations to address the cumulative length change of the tube”. But as briefly described above, length change would not be linear, and therefore cannot be compensated for merely at end stations with telescoping sections.
That sort of solution not only assumes expansion would be linear, but that the tube system would be perfectly straight from Los Angeles to San Francsico. We’ve already briefly discussed the non-linear expansion the tube will experience. As for a straight shot from Los Angeles to San Francisco; topography and land use will ensure the tube realizes turns, ascents, and descents along its length. Both these facts cannot possibly be mitigated by end station telescoping sections.
The oil industry deal with dynamic thermal expansion along above-ground pipelines by employing expansion loops. Figure 3 shows 3 examples of thermal expansion loops.
It’s obvious this solution wouldn’t work for a high-speed commerical passenger transportion system, but I’ve yet to hear of any viable solutions at all. Instead, research and development appears to continue to be focused on pod designs, electric sleds, and presentability while what I see as real-world engineering issues are ignored.
Reducing tunnel pressure to 100 pascals throughout a several-hundred kilometer long tube that will certainly experience non-linear thermal expansion, while simultaneously supporting the tremendous forces generated from supersonic multi-ton passenger pods, while simultaneously compensating for differential thermal expansion, and maintaining a near vacuum (100 pascals) will very likely have one of the worse ROIs in the history of public transportation. I wouldn’t be surprised if the high-speed rail Elon mocks in the beginning of his Hyperloop Alpha white paper has a superior financial ROI and EROEI.
Elon has also ignored basic thermodynamics when he promoted Tesla’s solar roof tiles. He has said he wants to see every home in the country have them installed. Solar isn’t a viable alternative energy source in many parts of the country simply because some regions don’t experience the kind of sunshine the desert southwest does. As such, EROEI on solar roof tiles would be atrocious.
But more obvious should be the fact his solar roof tiles do not have any way of mitigating heat because the tiles sit flush against the roof as seen in the following photo:
Heat can reduce a solar panel’s output efficiency by 25%, and converting photons (sunlight) to electrons (electricity) with solar panels already has a low efficiency as is (about 20% or less on average). 25% of 20% efficiency would not bode well for EROEI.
As solar panel temperatures increase, their electric output current increases exponentially, while their voltage output reduces linearly. This causes the solar panel production of power to decrease overall. This is why the most efficienct solar panels have space beneath them for air to flow. Something like the following photo:
The negative effect heat has on solar panels is so well known, that one could theoretically figure temperature by simply looking at a panel’s voltage reduction. Yet Elon’s solar roof tiles ignore this fact completely for reasons unknown to me. Also, not all roof’s, even in sunnier regions, face the Sun. In the United States, north-facing rooftops would make terrible surfaces on which to install solar panels, much less solar roof tiles.
These sort of things should be known by someone who embodies first principle reasoning.