How to use this page
As in the How to know lessons, keep a notebook beside you. Whenever you reach a Pause and answer box, stop and write your answer — a sentence, a rough sum, even a guess — before you read on.
Each pause is followed by a folded panel. Open it only after you have written something. The panel is one line of reasoning to compare with yours, not the official answer. Where yours differs, ask which one fits the facts better.
This lesson has arithmetic in it, but nothing harder than multiplying and dividing. Round freely: we want the right size of answer, not the fifth decimal. Every number we use is listed with its source at the end, so you can check it rather than trust it.
What we assume you already know
We build on the How to know lessons and on two facts that have survived every test anyone has managed to set them. We take these as given:
- An observation is not the same as a belief, and a useful claim names an observation that could show it wrong.
- Mass is conserved. Cooking, burning, smelting, and recycling rearrange atoms; they do not create or destroy them.
- Energy is conserved. It changes form — electrical, chemical, heat, motion — but you never get more out than went in.
- Energy is measured in joules (J). Power is energy per second, measured in watts (1 W = 1 J each second). The “unit” on an Indian electricity bill is one kilowatt-hour: 1,000 W for an hour, which is 3.6 million joules.
- Warming water takes a measured amount of energy: about 4.18 kJ per kilogram per °C. One litre of water has a mass of about one kilogram.
That is all. If you can multiply and you have ever waited for a kettle, you have the foundation.
When we say something is impossible, slow, or expensive — how do we know?
Most of our judgements about what is possible are borrowed: from how things have always been done, or from something else that seems similar. Questions follow:
- When I judge whether something can be done, what am I actually leaning on — habit, a comparison, or the thing itself?
- Is there a lowest possible cost, in energy or in material, for a given job? How would I find it?
- If reality sits far above that lowest cost, what does the gap mean? And if it sits close?
- Which limits are set by nature, and which by process, price, or custom?
- How far can I trust an estimate built from just a few physical facts?
- Write down one thing you believe is “just expensive” or “just slow” — a flight, a house, a phone battery, a hot bath, a bag of cement. Then write why you believe it. Is your reason about what the thing is made of, or about how it has always been?
Keep this example. You will strip it to what is true before the end.
The 30-second kettle
An advert appears on your phone: “Our new electric kettle boils a full litre in 30 seconds. Plugs into any home socket.” Three friends react.
Asha: “Kettles have always taken a few minutes. It’s a gimmick.”
Bilal: “Phones and computers got enormously faster in our lifetime. Why not kettles? I believe it.”
Chitra: “Let’s work out how much energy a litre of water needs, and how much a socket can deliver.”
Facts Chitra can use: the water starts at about 25 °C and must reach 100 °C. A kettle’s heating element turns electrical energy into heat — at best, joule for joule. A heavy-duty household power socket in India — the kind used for a geyser or an air-conditioner — is rated 15 or 16 amperes at 230 volts, which is at most about 3.5–3.7 kW.
- Asha and Bilal reach opposite verdicts. What is each one leaning on? Does either of them say anything about water?
- Do Chitra’s first sum: how many kilojoules does it take to warm 1 kg of water from 25 °C to 100 °C?
- To deliver that much energy in 30 seconds, how many watts are needed?
- Compare with what the socket can give. What is the verdict — and what would have to change for the advert to be true?
After you have written your answers — compare your reasoning
Asha leans on convention: “it has always been so.” Bilal leans on analogy: “something else improved, so this will too.” Their verdicts are opposite, but their method is the same — both talk about something other than the water. Neither can tell you why they are right, or what would change their mind.
Chitra’s sum: 1 kg × 4.18 kJ per kg per °C × 75 °C ≈ 314 kJ. Spread over 30 seconds, that is 314,000 J ÷ 30 s ≈ 10,500 W, or about 10.5 kW.
A heavy-duty socket gives at most about 3.7 kW. So even a perfect element kettle, with not one joule wasted, falls short by a factor of about three. The claim fails on arithmetic, not on anyone’s taste. To make it true you would need roughly three times more power than the socket can supply — or some other source of energy.
And Bilal’s analogy? Computers got faster largely by making each tiny operation use less energy and less material. There is no smaller way to warm a kilogram of water by 75 degrees: the water needs its 314 kJ however clever the kettle. A common 2 kW kettle therefore needs at least 314,000 ÷ 2,000 ≈ 157 seconds — about two and a half minutes. That is roughly what kettles already take. They are near their floor.
Reasoning by convention (“it’s always been this way”) and by analogy (“that other thing improved”) both reason about something other than the problem. First-principles reasoning asks three questions instead: What is this made of? What must physically happen? What does physics charge for it?
The answer is a floor: a minimum, set by conservation of energy, that no design can go below. And it tells you exactly what would have to change — here, more power — instead of leaving you with a feeling.
Where a can’s energy hides
Now apply the same questions to a material: the aluminium in a drink can.
Aluminium is never found as metal in nature. In its ore (bauxite) it is locked to oxygen as alumina, Al2O3. To make a can, a smelter must pull the oxygen off.
Aluminium badly “wants” oxygen. Fine aluminium powder burns with a fierce white light; it is one of the metal powders used in sparklers and fireworks. Chemists have measured the heat released when aluminium and oxygen join to form alumina: about 31 MJ for every kilogram of aluminium (that is 31 million joules).
An old can, on the other hand, is already metal. To recycle it, you melt it and recast it. Aluminium takes about 0.9 kJ per kg per °C to warm up, melts at 660 °C, and then needs about 397 kJ per kg more to turn from solid to liquid.
- Mass is conserved: to get metal from ore, what has to be undone? Energy is conserved: what is the least energy that undoing can take?
- Estimate the energy to melt 1 kg of scrap aluminium, starting from 25 °C.
- Divide your melting estimate by 31 MJ. Write down a prediction: roughly what fraction of the energy of making new aluminium should recycling need? Commit to it before you open the panel.
- What does your prediction say about why people collect scrap cans?
After you have written your answers — compare your reasoning
The aluminium atoms in the ore are the same atoms that end up in the can; nothing is created. What must be undone is the bond between aluminium and oxygen. Energy is conserved, so pulling them apart must cost at least the energy they released when they joined: about 31 MJ per kg, or about 8.6 kWh — nearly nine units of electricity for every kilogram of fresh metal. That floor is set by nature.
Melting: 0.9 × (660 − 25) ≈ 570 kJ to heat it, plus 397 kJ to melt it, so roughly 1 MJ per kg. (Careful tables give about 1.06 MJ, because aluminium needs a little more heat per degree as it gets hotter. Our rough figure is close enough.)
1 MJ ÷ 31 MJ ≈ 1/30, about 3%. Prediction: recycling should need only a few per cent of the energy of making new aluminium.
Now the check against the world. The International Aluminium Institute reports that recycled aluminium needs about 5% of the energy of primary aluminium (8.3 versus 186 GJ per tonne, counted from mine to cast house, in 2019). And real smelters use roughly 14 kWh of electricity per kilogram — less than twice our 8.6 kWh floor. The industry figures count extra things we ignored (mining, refining, power-station losses, sorting scrap), so compare the size, not the decimals. A two-line model built from what a can is made of predicted the right size of answer before we looked.
The energy cost of a material depends on what state its atoms are in, and what state you need them in. Breaking strong bonds is expensive; reshaping metal that is already metal is cheap. That is why scrap cans are worth collecting: each kilogram carries the bond energy that recycling never pays again. A first-principles estimate is a falsifiable prediction: we wrote “a few per cent” down first, and it could have failed.
The emperor’s cutlery
In the 1850s and 1860s, aluminium was a luxury. The French emperor Napoleon III is reported to have served his most honoured guests with aluminium cutlery, while lesser guests dined with mere silver.
In 1884 the Washington Monument was topped with a small pyramid of cast aluminium weighing 100 ounces (about 2.8 kg) — the largest aluminium casting of its time. Before it was installed, it was shown in Tiffany’s jewellery store in New York. That year the whole world produced only about 3.6 tonnes of aluminium, against roughly 2,800 tonnes of silver.
Yet aluminium makes up about 8% of the Earth’s crust by weight. It is the most abundant metal in the crust. Silver is roughly a million times scarcer.
In 1886, Charles Martin Hall in the USA and Paul Héroult in France independently found a way to split alumina dissolved in molten salt using an electric current. Within a few decades aluminium was in pots, foil, wires, and aeroplanes.
- Judged by convention — price, prestige, jewellery shops — what kind of metal was aluminium in 1884? Judged by what the crust is made of, what kind?
- In 1886, did aluminium atoms change? Did the energy of the aluminium–oxygen bond change? What did change?
- Using Observation 2: in the 1880s, was the cost of aluminium close to the physics floor, or far above it? What does a large gap tell an inventor?
- Could any future invention make fresh aluminium from alumina using less than the bond energy? Why or why not?
After you have written your answers — compare your reasoning
Convention said “precious metal”: it was displayed with jewels and served to emperors. The crust said “one of the commonest substances on Earth”. Both were telling the truth about different things. Price described the process of the day; abundance described the material.
The atoms and the bond were exactly the same after 1886. What changed was the method of paying the bond’s energy bill. Earlier processes used metallic sodium to strip aluminium out of its compounds — and sodium itself was costly to make. The Hall–Héroult process pays the bill directly with electricity, and as electricity grew cheap, so did aluminium.
So in the 1880s the real cost sat far above the floor. A large gap is an invitation: it means the limit is in our methods, not in nature, and a better method can close it.
No invention can make fresh aluminium from alumina below the bond energy; energy conservation forbids it. (Step 7 refines exactly what “energy” must be counted.) The only way around the floor is not to break the bond at all — which is what recycling does.
Separate the floor (set by nature: conservation of mass and energy, the strength of bonds) from the gap above it (set by process, scale, habit, price, and law). Gaps can close within a generation. Floors do not move. The question to carry everywhere is: “Is this limit nature’s, or ours?”
A day of muscle
Now measure ourselves in the same units as kettles and smelters.
People have been measured on pedal machines for decades. A healthy adult who is not an athlete can keep up about 75 watts of mechanical work for a full working day of about eight hours. A smaller or less well-fed person manages nearer 50 W. Trained athletes manage considerably more.
A kitchen tool, a factory, or a farm pump measures its work in the same joules.
- How much energy is 75 W kept up for 8 hours? Express it in kilowatt-hours, and compare it with one unit on your electricity bill.
- How long would you have to pedal a perfect generator to supply the 314 kJ for a litre of tea water?
- How many such working days would it take to supply the roughly 14 kWh a smelter uses to make 1 kg of aluminium?
- For most of history, a great deal of paid work was mainly muscle. Once engines and motors exist, what does this arithmetic say about where the value of human work lies?
After you have written your answers — compare your reasoning
75 W × 8 h = 600 Wh = 0.6 kWh. A full, exhausting day of human muscle delivers less energy than one unit on your electricity bill.
314,000 J ÷ 75 W ≈ 4,200 seconds — about 70 minutes of steady, hard pedalling for one litre of tea water. A kettle on the wall does it in under three minutes.
14 kWh ÷ 0.6 kWh ≈ 23 working days of pedalling for one kilogram of aluminium.
So, measured as raw energy, muscle is a very small source. Even when work was mostly muscle, much of its worth lay in what the worker knew — where to dig, how to set the plough — and engines have since taken over nearly all the joules. What remains is skill, judgement, care, trust, and knowing what to do with energy. That is a fact about physics. The arrangement “work in exchange for wages” is something societies built on top of it — and it can be rebuilt.
Measure human effort in the same units as everything else. In joules, muscle is small. Real power over mass — to lift, heat, smelt, pump, and build — comes from directing large flows of energy, and from understanding them well enough to direct them safely. First principles show you where that power actually sits.
Occam: rare metal, or tight bond?
Return to the emperor’s cutlery. Two people try to explain why aluminium was precious in 1884 and ordinary a few decades later.
Model 1. Aluminium was rare. Then huge new deposits were discovered, and at about the same time fashion moved on, and rich buyers simply lost interest.
Model 2. Aluminium was always abundant but tightly bound to oxygen. The old way of breaking that bond was costly. Electrolysis, paid for with electricity that kept getting cheaper, brought the cost down toward the energy floor.
- Check each model against everything we observed: the cutlery, the monument cap, 3.6 tonnes a year, 8% of the crust, 1886.
- Count the separate, unrelated assumptions each model needs.
- Model 2 makes predictions you could check: smelters should be built where electricity is cheap; electricity should be one of the largest costs of making fresh aluminium; recycled aluminium should be far cheaper in energy. Does Model 1 predict anything checkable?
Model 1 needs three unrelated assumptions, and its first one clashes with an observation: aluminium was never rare in the crust. Model 2 uses one mechanism — bond energy, plus the process used to pay it — which we already built in Observation 2. It explains the price fall, explains why recycling is cheap, and sticks its neck out with predictions that could fail.
When two models fit, prefer the one that needs fewer assumptions. That is a rule for choosing, not a proof. Real history is messier — demand, wars, and the spread of cheap electric generators all played parts — and a fuller model would add them. But we add pieces only when an observation demands them.
A working model of first-principles thinking
Gather what the four observations taught. Write your own version in four or five lines before you open ours.
- In your own words: how is reasoning from first principles different from reasoning by convention or by analogy?
- What is a “floor”, and where does it come from?
- What does the gap between the floor and reality tell you?
- How do you check that your estimate is not just a confident story?
After you have written your summary — compare it with ours
| Step | What it means | The question to ask |
|---|---|---|
| Strip | Set aside how it is usually done and what it resembles. | “What is this actually made of, and what must physically happen?” |
| Floor | Use conservation of mass and energy, plus measured properties (heat per degree, bond energy), to find the minimum cost. | “What does physics charge, at the very least?” |
| Compare | Put the floor next to what is observed: time, energy, price. | “How far above the floor is reality?” |
| Read the gap | Close to the floor: limited by physics; expect no miracles. Far above: limited by process, scale, or habit; room for invention. | “Is this limit nature’s, or ours?” |
| Lever | Name what would have to change. | “More power? Stored energy? A new process? Not breaking the bond at all?” |
| Check | Treat the estimate as a falsifiable prediction, written down before you look. | “What measurement would show my estimate wrong?” |
Convention and analogy are not useless. They are good at suggesting ideas: “phones improved; is there a similar trick for this?” But they cannot decide. The floor decides what is possible; observation decides what is true. And a floor is not a price: prices also carry process, scale, transport, taxes, profit, and custom. First principles tell you what is possible and where the room lies, not what the market will charge tomorrow.
And this, too, is a model, not a proven theorem. Every floor rests on a list of principles we chose to use. If we forgot one, the floor is wrong. We will test that in a moment — after you have tried the model on your own kitchen.
Exercise: audit your kettle
Turn Observation 1 into a measurement. You need an electric kettle, a measuring jug, and a clock. A kitchen thermometer helps but is optional.
- Read the power rating P on the kettle’s label or base, in watts.
- Pour in exactly 1 litre (1 kg) of water.
- Measure the starting temperature T. If you have no thermometer, assume 25 °C — and write down that you assumed it.
- Switch on and time, in seconds, until the kettle switches itself off: t.
- Energy into the water: Ewater = 4.18 × 1 × (100 − T) kJ.
- Energy taken from the socket: Ein = P × t ÷ 1000 kJ.
- Efficiency = Ewater ÷ Ein. Floor time = Ewater × 1000 ÷ P seconds.
Worked example with made-up numbers, for illustration only — not a measurement: a 1,500 W kettle, 1 litre at 25 °C, switches off after 240 s. Ewater ≈ 314 kJ; Ein = 1,500 × 240 ÷ 1,000 = 360 kJ; efficiency ≈ 87%; floor time ≈ 209 s.
- Do your own sums. How close to its floor is your kettle?
- Where did the missing energy go? Name two places, and say how you could test each guess (touch, a lid on or off, a second run with a warm kettle).
- If you boil only 250 ml for one cup, how does the floor change?
- Suppose your sums say your kettle is more than 100% efficient. What does that tell you?
After you have done the measurement — compare your reasoning
Many element kettles come out fairly close to their floor; check whether yours does. The missing energy warms the kettle’s own body and element, leaks out as warm air, and leaves as steam in the last seconds before the switch clicks off. A second run in an already-hot kettle should come out a little more efficient — a test of the first guess.
The floor is proportional to the mass of water. A quarter of the water needs a quarter of the energy. Better insulation can trim the small gap; nothing can trim the 314 kJ per litre. So the biggest saving available is a habit, not a technology: boil only what you need.
An efficiency above 100% would break conservation of energy. Far more likely, a measurement is off: more water than you thought, a wrong power rating, a warmer starting temperature. Your model just caught an error in your data — which is exactly what a falsifiable model is for.
Pushing on the model
A floor is only as good as the principles behind it. Here are some hard cases. Expect sharper thinking, not neat answers.
- The kettle company returns: “Our 30-second kettle has a built-in battery that charges slowly from the socket between uses.” Does our verdict still hold? What exactly did our sum forbid?
- A heat-pump water heater moves heat out of the surrounding air into the water, rather than making all of it from electricity. Could such a device warm a litre of water using less than 314 kJ of electricity? What does that do to our “floor”?
- Real aluminium smelters also burn up carbon blocks, turning them into carbon dioxide. Does that let them use less electricity than our 8.6 kWh? Does it let them beat the bond energy?
After you have written your answers — one way to refine the model
Energy is not power. Our sum showed two things: the water needs 314 kJ (an amount), and delivering it in 30 seconds takes 10.5 kW (a rate). A battery holding a little under a tenth of a kilowatt-hour, filled slowly from the socket, could release it fast. Physics no longer forbids the claim; it becomes a question of cost, safety, and size. Refined rule: an energy floor cannot be beaten; a power limit can sometimes be beaten by storage.
A missing principle can lower a floor. We assumed every joule of heat had to be made from electricity. A heat pump breaks that assumption: it uses electricity to move heat that is already in the air. The US Department of Energy reports heat-pump water heaters can be two to three times more energy-efficient than ordinary electric-element ones. The water still needs its 314 kJ of heat — that floor stands — but the electricity floor was only a floor for element heaters. Our model did not lie; it was incomplete. Why heat can be moved, and what it costs to move it, is the subject of Theme 06, Thermodynamics.
Count every energy input. Burning carbon releases energy of its own, so part of the bond-breaking bill is paid in carbon rather than electricity, and the electricity needed can be lower than our figure. But the total — electricity plus carbon — still cannot fall below what the bond demands. The refinement is to count all the energy flowing in, not just the bill you happen to see.
Why this matters now
Capable machines will soon draft plans, designs, and business cases for anyone who asks, fluently and in seconds. A fluent plan can still promise a 30-second kettle. Before you admire a plan, strip it: What is it made of? What must physically happen? What does physics charge? How far above that floor does this plan sit — and why?
And when wages stop being the centre of daily life, what remains is what you can actually do with matter and energy: grow, build, heat, cool, move, repair. Knowing the floor tells you which limits to accept, and which are only habits waiting for someone to change them.
Review & discussion questions
Answer these by explaining, in full sentences, as if teaching someone who missed the lesson. Try them alone first, then discuss with a friend or study group and compare your reasoning.
- Using the 30-second kettle, explain the difference between reasoning by convention, by analogy, and from first principles. Why can convention and analogy reach opposite verdicts by the same method?
- Without notes, rebuild the energy floor for boiling one litre of water from 25 °C. Then explain why a 2 kW element kettle can never boil it in much under two and a half minutes.
- Explain, using conservation of mass and energy, why recycled aluminium needs only a few per cent of the energy of new aluminium. What prediction did we write down first, and what observation tested it?
- Aluminium went from emperor’s tableware to kitchen foil. Explain what changed and what did not, using the words floor and gap.
- Go back to the “just expensive” or “just slow” thing you wrote at the start. Strip it: what is it made of, what must physically happen, and what is a rough floor? Is reality close to the floor or far above it — and is the limit nature’s or ours?
- Explain the difference between energy and power using the battery kettle. Give one more everyday example where storage beats a power limit.
- Explain how the heat pump showed our electricity floor to be incomplete. What does this teach about trusting a first-principles estimate?
- In energy terms, a full day of human muscle is worth less than one unit of electricity. Explain what follows from this fact — and what does not follow from it.
…you can explain, without looking, how to strip a problem down to what it is made of and what physics charges; work out an energy floor from conservation of energy; read the gap between floor and reality as nature’s limit or ours; and you have audited your own kettle.
Go further Once you have found the floor, how do you close the gap above it? Try the TRIZ refresher in Skills & resources.
Where the numbers come from. Every figure above is a well-established measurement; rounded values are used throughout.
- Water, 4.18 kJ per kg per °C (specific heat near room temperature): standard reference value, e.g. CRC Handbook of Chemistry and Physics; NIST.
- Household supply, 230 V; heavy-duty sockets rated 15/16 A: Indian nominal supply voltage and plug/socket ratings (Bureau of Indian Standards, IS 1293). 230 V × 16 A ≈ 3.7 kW.
- About 31 MJ per kg of aluminium: from the standard enthalpy of formation of alumina (corundum), −1,675.7 kJ/mol, NIST Chemistry WebBook. 1,675.7 ÷ 2 kJ per mole of aluminium ÷ 0.02698 kg per mole ≈ 31 MJ/kg ≈ 8.6 kWh/kg.
- Aluminium: specific heat ≈ 0.90 kJ per kg per °C; melts at 660 °C; heat of fusion ≈ 397 kJ/kg (10.71 kJ/mol): CRC Handbook. Careful value for heating from 25 °C to fully molten, ≈ 1.06 MJ/kg: NIST-JANAF thermochemical tables.
- Smelting electricity ≈ 14 kWh per kg: International Aluminium Institute statistics, primary aluminium smelting power consumption (world average about 14,300 kWh per tonne in 2019).
- Recycling ≈ 5% of primary energy (8.3 versus 186 GJ per tonne, 2019): International Aluminium Institute, “Aluminium recycling saves 95% of the energy needed for primary aluminium production”.
- Aluminium ≈ 8% of the crust by weight (8.2%, Rudnick and Gao 2003): US Geological Survey. Silver ≈ 0.075 ppm: CRC Handbook — roughly a million times scarcer.
- Napoleon III’s cutlery; 1884 production of 3.6 t aluminium versus 2,834 t silver; Hall and Héroult, 1886: Science History Institute, “Aluminum: Common Metal, Uncommon Past”. Washington Monument cap, 100 oz, installed 6 December 1884, shown at Tiffany’s: US National Park Service; US National Archives, Prologue (Summer 2014).
- About 75 W sustained for ~8 hours by a healthy non-athlete; ~50 W for a smaller, less well-fed person: NASA ergometer data as presented in Whitt & Wilson, Bicycling Science (MIT Press); MIT D-Lab pedal-power notes.
- Heat-pump water heaters two to three times more efficient than electric-element heaters: US Department of Energy, Energy Saver, “Heat Pump Water Heaters”.
Attribution. Lesson structure — Foundation, questions, observations and deductions, refined models — adapted from John S. Hutchinson, Concept Development Studies in Chemistry (Connexions / Rice University), licensed under Creative Commons Attribution 2.0 (CC BY 2.0). The topic, examples, and text of this lesson are original to learn.curiosta.com.