Energy featured

Brace for Peak Impact

June 19, 2024

Guilty as charged: my recent postings have been all about human population and when it might peak. I don’t mean to be a bore, but it’s an important topic connected to planetary limits, human impacts, ecological health, and the appealing prospect that a near-term peak may offer an earlier off-ramp for modernity. In the past, I have stressed the point (in a blog post from 2013 and later in a textbook chapter) that population per se isn’t the phenomenon of greatest concern, but its multiplication by resource usage. It’s the combination that launches us over the ecological cliff edge, commonly expressed by the I=PAT formula for impact on the planet.

In this post, I belatedly take my own advice and re-frame the population investigation in resource terms. Now that I have a demographic tool, I can ask questions relating to when we might hit peak power as a civilization. I use power (rate of energy use) as a proxy for all manner of resource dependencies, as energy usage correlates strongly with materials use and ecological impact. Plus, it is a readily-available measure.

So, given various assumptions about how fertility rates evolve regionally, and factoring in different models for regional survival rates and migration, when might we expect global resource use to peak and begin a decline? In tandem with this event, we might correspondingly expect peak industrial output, and peak rate of (accumulating) damage to ecological health—which includes our own health. In the U.N.’s standard demographic model, population does not peak until 2086 at 10.4 billion—largely bolstered by population growth in Africa, which the U.N. parameters indicate will climb to 4 billion by 2130 (we’ll see…). But, since Africa is by far the region with the lowest per-capita consumption, declines elsewhere could more than offset Africa’s population increases in terms of resource burden.

Enough speculation: let’s unleash the model and see what happens.

Regional Power

Our exercise starts with the following table that parallels Table 3.5 in Chapter 3 of my textbook:

REGION POP (B) KW/CAP POWER (TW)
Africa 1.49 0.5 0.8
Asia 4.79 1.8 8.6
Europe 0.74 4.9 3.6
Latin 0.67 2.2 1.5
N. A. 0.38 9.9 3.8
Oceania 0.05 5.4 0.3
World 8.12 2.3 18.5

Multiplying population (in billions) by the power per capita (in kW) gives terawatts (TW) of power. For the whole world, it comes to about 18.5 TW. [Note that the American regions in this table are defined differently than the continental divide employed in the textbook, in that “Latin” stands for Latin America (including Mexico) and the Caribbean, while Northern America is primarily the U.S. and Canada (see the Geographic Regions tab here for exact definitions).]

For simplicity, I hold the per-capita power usage at these values for the entire simulation. This isn’t insane, as the plot below shows global per-capita fossil fuel use being relatively flat since 1970 (right-hand plot; suggestive red curve), in contrast to the ever-climbing total power, at left. In truth, per-capita energy will more likely decline as fossil fuel availability wanes and other (associated) disruptions arise, possibly pulling the moment of peak power even sooner than the following projections suggest. Additionally, net energy peaks sooner than total energy when extraction/processing efficiency declines as easy resources are consumed first.


Fossil fuel use history and per-capita rate. It’s stacked, so that gas is 0.5 kW per person, not 2 kW. The red curve at right shows a relatively stable per-capita power since 1970.  The surge around 2005 is largely due to a major coal increase in China.

Anyway, holding per capita power flat is fine for the purposes of this exploration, since accuracy is unobtainable in such exercises.  We just want to provide some window into the situation, differentiating various outcomes at the decade scale.

U.N. Case

Let’s start with the vanilla U.N. assumptions and resulting projection. This is the one that peaks in 2086 at 10.4 billion people, with Africa still on the rise until about 2130. In the tables that follow, I list for each region (and the whole world) the year of peak power consumption and the associated value in terawatts.

AFRICA ASIA EUROPE LATIN N.A. OCEANIA WORLD
2130 2056 2022 2055 2094 2109 2061
2.0 TW 9.6 TW 3.7 TW 1.6 TW 4.4 TW 0.4 TW 20.5 TW

So, while global population peaks in 2086 in this scenario, energy consumption peaks 25 years earlier. This is a convincing illustration that the two things have substantially different timing. Incidentally, for each region (but not the world as a whole), the year listed for peak power also coincides with peak population, as a single per-capita number is applied for the whole region.

U.N. with Flat Survival

Just as we freeze power per capita, we can freeze survival rates at today’s level. This is equivalent to saying that regional life expectancy neither rises nor falls into the future (keeping in mind that post-peak, the model becomes garbage as lots of things change). For reference, pre-COVID life expectancy in the U.S. was remarkably flat from 2010–2019 (peak was 2014; see plot below).


Data from the CDC: https://www.cdc.gov/nchs/data/nvsr/nvsr72/nvsr72-12.pdf

COVID contributed to a significant dip, and the 2022 number is on a par with what it was 18 years earlier in 2004. In other words, a monotonic increase in life expectancy through the year 2100—as modeled by the U.N.—already seems broken and is not guaranteed to materialize. In perhaps yet another disconnect, the 2022 UN demographic projection model expects U.S. life expectancy to break 80 in 2025—somehow. Anyway, here’s what happens in the flat-survival scenario:

AFRICA ASIA EUROPE LATIN N.A. OCEANIA WORLD
2100 2047 2021 2045 2050 2134 2048
1.6 TW 9.2 TW 3.7 TW 1.6 TW 4.0 TW 0.4 TW 19.7 TW

Now peak power pulls back another 13 years to occur before mid-century: not far away.

My TFR and Flat Survival

I have serious qualms about the U.N.’s apparent denial of nearly-ubiquitous downward trends in total fertility rate (TFR) over the last decade, as I first expressed here and further dissected here. The basic disconnect is that many countries around the world are exhibiting low and declining TFRs, but the U.N. model repeatedly fails to capture or anticipate this trend, imagining instead a quick rebound to higher (historically “normal”) TFR values that persistently fail to materialize in projection after projection as the years roll by and the declines stubbornly continue. Using the alternate TFR model I developed (introduced here), and keeping the “flat” survival model (locking at 2023 rates), this is the emerging story:

AFRICA ASIA EUROPE LATIN N.A. OCEANIA WORLD
2075 2036 2021 2039 2036 2071 2036
1.2 TW 8.8 TW 3.7 TW 1.5 TW 3.9 TW 0.3 TW 19.0 TW

Now we’re just 12 years out from the global peak! Africa peaks late-century, still, but at a small total power that does little to shift the larger trend. Incidentally, using the U.N. model for monotonically increasing survival rates in combination with my alternate TFR model produces a peak in 2044.

My TFR and Reverse Survival

In a declining energy scenario and the attendant economic convulsions, it could make more sense to imagine a reversal of life expectancy as advanced medical care becomes less prevalent. The slow reversal model I employ is a mirror image, so that ten years from now it’s as it was ten years ago, etc.  Hey—as displayed above, life expectancy peaked in the U.S. in 2014 and is now where it was in 2004, so we’re already experiencing something qualitatively similar to this model.  Reflecting the U.N. survival models around the year 2023, we get this development:

AFRICA ASIA EUROPE LATIN N.A. OCEANIA WORLD
2064 2034 2021 2035 2037 2055 2034
1.1 TW 8.8 TW 3.7 TW 1.5 TW 3.9 TW 0.3 TW 19.0 TW

The change from the previous case is not dramatic, because peak power is so close-at-hand that medical conditions don’t change drastically from the “flat” model in ten years’ time. Still, we’re down to a decade away from now!  The peak power in the last two cases is only a few percent higher than what we pump out today, so that life right now could pretty-much be what the peak feels like.  It’s that odd moment near the crest of the roller coaster, when the basic contours of the experience are about to change entirely.

As an aside, if you noticed and are concerned by the strange result that Northern America moves one year later by reversing the trend (and other counter-intuitive changes in earlier cases), it’s due to the fact that the “survival” model I extracted from U.N. projections simultaneously accounts for migration, so that the reversal in this case reflects an uptick in (past) immigration.

Although plots of these model results aren’t ridiculously instructive, let’s have one anyway, including regional contributions.  Dominant Asia largely sets the pace, as can also be appreciated in some of the tables above.

Global and regional energy demand in the model projection.

Actually, this plot just gave me a bit of Déjà Vu. Within the first month of starting Do the Math, I put up a post pointing out that data on U.S. energy usage was starting to roll under the exponential curve, but that it was too early to tell whether it was heading for leveling-out or powering-down. The curve above (for the world) is firmly in the power-down camp based on demographic projections. This isn’t a post on peak oil, but in some ways amounts to the same phenomenon of declining power, by a different mechanism: not one of the scenarios I had considered for bringing about the peak of fossil fuel use.  The real world will surprise you like that!

Africa Rising

While most of the world has TFR falling faster than the U.N. models reflect, this may not be true for Africa. What if Africa is recalcitrant in its fertility decline, dropping more sluggishly than the U.N. projections? Setting aside biophysically-rooted skepticism I expressed last week, let’s see how the story changes if we keep the alternate TFR model for most of the world, but allow fertility decline in Africa to be slower than the UN case, as in the following plot.


Modified TFR with slow Africa: solid is data; dashed used for the model; dotted is U.N. projections.

AFRICA ASIA EUROPE LATIN N.A. OCEANIA WORLD
2082 2034 2021 2035 2037 2055 2034
1.5 TW 8.8 TW 3.7 TW 1.5 TW 3.9 TW 0.3 TW 19.0 TW

This change results in Africa’s population peak moving out 18 years from the previous scenario and adding another 0.75 billion people to its peak (to 2.9 billion in this reversing-survival case). Meanwhile, the global population peak only moves from 2040 to 2043 (not shown in tables).  But the global peak in power did not move in year or magnitude. That’s because Africa has the smallest per-capita resource footprint of all the regions—by a significant amount—so that prolonged growth on its part does not hold much sway over the global story in terms of resource peak.

Power, not the People!

The main point of all this is that concern about population growth tends to be motivated by concern over the impacts on Earth and/or resource availability—not over the raw number of people. In this context, the date of peak-people-number may turn out to be less significant than the date of peak rate of resource exploitation. Some of the scenarios above put this within 10 years! I’m surprised. Are you surprised?

Because energy is a decent proxy for the scale of the global economy: extractive activities, industrial output, transportation, agricultural production, construction—and all the other things we do with energy—a global peak in power output likely corresponds to a global peak in the industrial scale of modernity and its economy. Wow. And that may happen in the next decade. Hold on to your hats!

Let’s throw up again a plot of the doubled-resources run from the Limits to Growth 30-year update (best match to data thus far):

World3 model from the Limits to Growth 30 year update.

I wasn’t at all trying to match this result, but damn! Looks like a peak in industrial output and population around 2040. Food per capita is at an all-time high about now, in their model. By 2050, we should know if the model has anything to it. That’s a “short” 25 years from now, but it must have seemed a distant eternity beyond comprehension in 1972. An important distinction between Limits to Growth and demographic projections: LtG is a systems model, tracking a whole host of fundamental drivers.  Demographic models tend to be narrow extrapolations based on past trends and some guesswork (idealization) as to future evolution of trends, but not attempting to dynamically model the entire interconnected system.

Is Post-Peak Bad News?

John Michael Greer offers context on what declining population could mean in a recent essay titled An Unfamiliar World. Economic growth has heretofore been entirely in the context of growing population, and not by coincidence. More people translates to more demand, more jobs, more houses, more retail, more cars, more energy, and all the rest. Fewer people translates to…

This must seem like dire news to an economist, but it may be just what the living world (to which we belong!) needs to move closer to balance and functionality. Bad news for Team Cancer, but good news for the whole organism—Team Life. And to be clear, I’m not calling humans the cancer and advocating eradication. Before being afflicted with cancerous cultural norms, humans operated within an ecological context. Pancreatic cancer does not mean that the pancreas is inherently bad—just that an unfortunate development changed its essential character and how it interacts with the whole organism. Note that a successful cancer is bad news even for itself, as it dies upon killing its host, without ever passing its genius to a future version of itself. The more “successful” it is, the worse the outcome—just as for modernity. Let’s rediscover what it is to be non-cancerous humans as integral actors in “right relationship” with the rest of the community of life.  Everyone would be healthier for it.

Our present cancerous economy is not geared to handle the switch to a decline phase gracefully. Investment relies on the notion that tomorrow will be bigger than today. What does a generations-long contraction scenario look like to investors? The mattress starts to look pretty good as a place to stash cash. The question—whose answer we can’t know yet—is: how quickly and turbulently will the economic chemotherapy proceed, and can the organism survive the treatment or is the cancer too invasively suffused throughout? One ray of hope here is that each newborn is not yet infected with cultural maladies, so that real change can begin as soon as the culture recognizes its folly and tries something different—allowing the irredeemably afflicted members to fade out in a natural, slow progression.

The Stadium

At the risk of inducing metaphorical whiplash, I have also used some form of the following metaphor to paint the picture as I previously imagined it developing: Crowds flock to fill a stadium to witness a spectacular event. They aren’t sure exactly what it is they’re going to see, but word-of-mouth promises that it will be amazing. It turns out that the grand spectacle is the collapse of the stadium—entirely due to the size of the gathered crowd.

That’s how I perceived the tragic irony of having the most-ever people on the planet suffer the worst-ever event: an event brought on precisely because of so many people: a self-fulfilling tragedy that maximizes total suffering as if by dastardly design. I saw it as an unwitting and inevitable trap. Population would heedlessly grow until it hit a wall and broke its nose. And the wall.

In light of the potential for an early peak due to rapid fertility declines underway in the resource-hogging (“developed”) world, I wish to modify the stadium metaphor: fewer people came than expected, and began thinning out before triggering utter collapse.

Now, the stadium (modernity) was never built on principles of sustainability, so it inevitably crumbles anyway, and it still has substantial crowds putting a load on it, at great risk to themselves and others. But instead of wholesale catastrophic collapse that takes almost everyone down with it, smaller-scale asynchronous crumblings are confined to regions. Questions: Is it still tragic and does it still involve human suffering of an unprecedented scale? Very probably: that can’t be helped at this point, unfortunately. Is it as bad as a full-scale collapse brought on by unrelenting growth of the crowd? Surely not. I’ll take whatever sliver of good news I can get!

I should point out that the mere presence of the stadium has crushed and extinguished many species who did nothing wrong, but didn’t stand a chance. Going forward, the stadium will continue to harm others (and humans) both by its existence and during the uncooperative process of its crumbling. In other words, humans are not the only ones to suffer for its being built and for its self-destruction. Moreover, it makes no sense to pretend that harm to the community of life does not also harm humans—being that we are a part of the community, not separate from it.

Let’s face it: the stadium was never a good idea. It’s been really amazing and cool in a number of ways, but never “of this world” in the sense that it had no contextual support from ecological, biophysical, or evolutionary history, and thus was never vetted for sustainability (long-term success). It’s a short-lived stunt.

The fortunate development could be that ticket sales wane as youngsters already brought into the stadium see the cracks and decide not to invite more youngsters—ignoring the panicked old-guard who hold unshakable faith that this grand event is going to turn out to be amazing in a spectacularly good way if only the young folks would share their faith and invite more people.  Meanwhile, perhaps some within the stadium will quietly find the exits and others will manage to survive the Great Crumbling to swear off stadium life forever more.

Bonus: Video!

I made a video on the topic!

Tom Murphy

Tom Murphy is a professor of physics at the University of California, San Diego. An amateur astronomer in high school, physics major at Georgia Tech, and PhD student in physics at Caltech, Murphy has spent decades reveling in the study of astrophysics. He currently leads a project to test General Relativity by bouncing laser pulses off of the reflectors left on the Moon by the Apollo astronauts, achieving one-millimeter range precision. Murphy’s keen interest in energy topics began with his teaching a course on energy and the environment for non-science majors at UCSD. Motivated by the unprecedented challenges we face, he has applied his instrumentation skills to exploring alternative energy and associated measurement schemes. Following his natural instincts to educate, Murphy is eager to get people thinking about the quantitatively convincing case that our pursuit of an ever-bigger scale of life faces gigantic challenges and carries significant risks.

Note from Tom: To learn more about my personal perspective and whether you should dismiss some of my views as alarmist, read my Chicken Little page.