In the opening entry in this series, Solarchy: Our Sun-Worshipping PV Future, I laid out some of the likely utopian consequences of photovoltaic solar power over the long-term. However, in the interests of a positive tone, I kind of glossed over some of the transitional steps between here and there - an interim that, in grimmer scenarios, could involve near-total short-term collapse of the global water, food, and political infrastructure before new equilibrium is achieved. But rather than wallow in apocalypticism, which I loathe, I will instead illustrate the pitfalls and what steps would avoid them.
First of all, humanity will not go extinct. Ever. You need to accept this. It is conceptually more difficult, and personally very humbling, to open your mind to the fact that we are individually just links in an infinite chain going both backward and forward into deeper time than we can imagine. Even if you reject the anthropic philosophy that consciousness is self-selecting on an ongoing, moment-to-moment basis, it would be far too easy to underestimate the survival advantages of technological intelligence. There is a lot of fragility in current technology because it has not really been tested under pressure, but don't be deluded by surfaces - humanity is the single most robust organism on Earth, and grows more so over time.
We will not go away. Ever. Technological intelligence is a form of meta-evolution that both reflects and transcends the underlying imperatives of living organisms, abstracting them into an information ecosystem that can survive (or even benefit from) extreme population bottlenecks. The events that sent the dinosaurs into extinction would not have had the same effect on us. Even in an apocalypse so scrupulous that everything down to the core philosophic principles of the Enlightenment and the practical forms of the scientific method are eliminated, human DNA would still contain the same long-term collective potential, and the physical geography would still carry the same inescapable lessons leading to convergent social evolution (albeit from an alternate direction).
Some people may actually find this fact appalling, for various reasons: For one thing, it means the decisions people make have real and unbounded consequences - it's not all going to be wiped away in a convenient slate-cleaning. Certainly, most of what we do does not change the overall trajectory of events, and simply blends into the background noise of time, but the choice is always there to do something Significant - and that choice is often bothersome if we would rather be egocentric and pretend our problems are insurmountable, and that future generations must be intellectually and morally beneath us.
Now, this is not to say that a given civilization will not end - that kind of extinction is unfortunately a direct consequence of our social intelligence, as it is always shifting and forming new alignments. So it is a possibility - albeit the most extreme one - that civilization as we know it could end relatively abruptly and give us a poverty-stricken, desperate, violent interregnum before new liveliness overcame the downward cascade. All it took to spark the end of the European Dark Ages was the reintroduction of a tiny fraction of the literature and philosophy of antiquity to a handful of cities on the Adriatic, so I wouldn't be worried that concepts like democracy, steam engines, solar power, and the transistor would be irrevocably lost even in the worst plausible scenario. But a Dark Age is a Dark Age, and looks like an infinite abyss to the people who live in it, so let's talk about avoiding it.
I. Priority 1: Make the supply chain for renewable energy systems self-sustaining.
In other words, the first step is to make renewable energy actually renewable. Right now it isn't, and isn't yet at a point where it can be: The mining equipment that extracts and processes the metals used in solar panels, electronic circuitry, mirrors, and both steam and wind turbines is overwhelmingly still powered by fossil fuels. Ditto the vehicles that transport the feedstocks to the factories, the production machinery itself, the physical buildings that house the machinery, and the transports that move the output to consumers and utilities.
This presents a keen danger in the short-term if fossil fuels become radically scarce in a short period of time, because they would then have to be reserved for food and water systems while the transition to renewability would be de-prioritized: An unavoidable move that would all but guarantee a Dark Age of indeterminate length. With fossil fuels nearly all directed to meeting immediate needs, their depletion and its destructive effects on civilization would accelerate while less and less surplus energy is available to build a sustainable system. Even the existing renewable infrastructure would fail after a few years because its operation, maintenance, and replacement would still depend on mostly non-renewable systems.
In the worst scenario - one that I don't consider very probable, mind you - the supply chain would completely unravel to the point where even highly localized renewable manufacturing would become untenable. This would require pretty much every decision made between here and there by every potentially self-sufficient economy to be the wrong one - there could be no pockets of regional stability for such a scenario to occur, because any such economic islands would be rapidly catalytic to the rest of the world (much like the role played by US manufacturing in the aftermath of WW2).
All that would be needed to stop a cascading collapse would be a single economy - large enough to defend itself from the military predations of more desperate neighbors - making an intelligent decision to stay focused on renewability despite short-term hardships. If the availability of fossil fuels dropped through the floor tomorrow, there would still probably be several such economies. Nevertheless, the human toll of even a mild collapse would be extreme: Mass migrations would depopulate the global megacities, totally annihilate some national borders (and even destroy some nation-states altogether), and cause general chaos in psychological terms even where the economics are relatively tolerable.
So the first step in avoiding that is to make mining equipment directly used in extracting renewable feedstocks (e.g., silicon, gallium, indium, etc.) itself renewably-powered, and from there move up the supply chain as well as horizontally to parallel industries. The further one extends the renewability of the supply chain, the shallower a collapse would be. Once you get to the point where a complete, operational renewable-energy module can be produced at volume using only that type of energy production, a collapse would probably not occur: It would simply be a very deep global Depression with resources being radically reinvested into extending the production capacity of renewables.
Now, it doesn't have to be PV solar that we're talking about - you could do the same thing if the entire supply chain that supports a solar thermal (ST) plant or wind farm were itself renewably powered - but as you may guess, not all supply chains are created equal. It would take a depth of quantitative analysis beyond my capability to argue this point conclusively, but we can say that PV is an information-intensive technology based more on subtle electronic knowledge than on large quantities of material input, and would intuitively seem to require much lower material throughput (i.e., lower marginal input energy) to sustain than ST or wind. Granted, I'm thinking about PV as it would be once it's already operating with significant economies of scale, so I wouldn't hazard to guess what the case may be while it's still sucking up the bulk of its investment just figuring out how to scale up.
A rewewable renewable energy infrastructure is the kernel that makes all subsequent progress possible, so hopefully - and I realize such rationality is perhaps too much to ask of business in general, let alone a fragile young industry that's still finding its footing - companies involved in renewable energy are forward-thinking enough to be putting their own product to use in their production lines, if not subsidizing their use upstream in their own supply chains. Vertical integration is probably a good idea, although there is always that compromise between present efficiency and future innovation.
II. Priority 2: Make water infrastructure renewably-powered.
Once you've secured the ability to sustainably make renewable energy capacity, the next logical step is to begin securing the most immediate necessity, water. You may argue this should be priority 1, but again, if you can't maintain the energy infrastructure that supports the renewable extraction/capture, desalination, treatment, pumping, and reclamation of water, you're going to run out of it anyway. Although the priorities would have to be served simultaneously, the renewability of the energy-generation systems must be a higher priority than anything the energy is used for or else you're subject to the kind of cascading economic collapses mentioned earlier.
As it is, water is the second-most-immediate economic need, the most immediate individual need, and a very energy-intensive system because (among other processes) it moves large masses significant distances and removes metals and salts from a powerful solvent. The absolute amount of energy used by water systems is only going to increase over time as economies shift to seawater and more rigorous recycling systems, so it becomes even more critical that those energy needs be met renewably to avoid a sudden fossil fuel collapse being translated into potentially lethal regional water shortages or, in milder cases, the widespread destruction of agriculture from lack of irrigation leading to dangerous spikes in food prices.
If Priority 1 is fulfilled but Priority 2 neglected, whole nations could disappear from the face of the Earth while the water system is reorganized - entire regions turned into large, generalized failed states with only nominal governance and uniform desperation across multiple ethnic boundaries causing endless, pointless strife. Parts of Africa may already be characterized in these terms, but the economics of that issue have more to do with politics than with energy.
Water crisis would not, however, lead directly to depopulation of global urban centers - they are overwhelmingly already located next to significant bodies of water, and crash programs operated on a military footing could build the infrastructure needed to serve their metropolitan regions. The immediate disaster in this case would occur everywhere other than the megacities and coasts, turning inland regions into impoverished, lawless, Dust Bowl environments that would drive people toward the cities and cause destabilization by virtue of increased density rather than diffusion. That could secondarily cause ethnic tensions, ghettoization, food riots (as the agricultural input from the countryside wanes), refugee crises, and political crises potentially resulting in wars and massacres to stop people from flooding across borders from one city into another, or from rural regions in one country into another country's cities.
The nightmare scenarios from neglect of Priority 2 would be more like Soylent Green than Mad Max - urban, crowded, requiring people to sacrifice, scrimp, and cooperate in the most petty and soul-crushing ways to maintain access to resources. Mild scenarios, however, would simply feature an acceleration of what we already see: The decline of suburbia, greater urban densities, desertification of the countryside, etc.
III. Priority 3: Make farm equipment, railways, trucks, and cargo ships renewably-powered, in that order.
This step progressively secures the chain of food supply, first securing the equipment directly involved in farming, then the backbone food-transportation systems, and finally the global trade systems that provide backup food supplies in case of regional devastation. Short of this point, there is still a great deal of vulnerability, although the probability of global castastrophe is significantly smaller.
Even if you have a substantially renewable supply chain for renewable energy and sustainable, renewably-powered water infrastructure, the sudden collapse of fossil fuel availability could still cause massive upheaval if farm and bulk transportation systems still largely depend on them. If a significant proportion of farm equipment and the transportation used to bring food to market becomes uneconomical more quickly than it can be replaced, the result is in many ways the same as a water shortage: Fields laying fallow while people go hungry, turning the countryside into impoverished dust bowls and causing huge price spikes in cities.
However, unlike a water collapse, an exclusively food-related crisis would have a diffusing effect on population because the economics would favor relocation closer to wherever the food is still being produced, and those areas would not be impoverishd: They would have plenty of water, and decentralized renewable energy - although there would still be problems with a relatively sudden population influx.
Cities would not be depopulated, mind you - they would still have huge energy and water infrastructure that would be of mutual benefit to the farming regions, but we could expect them to be grimmer, poorer, and lacking in vital energy as long as the crisis lasted. In milder cases, you wouldn't see much migration, but food would be much more localized, leading people to eat more seasonally and regionally-available staples rather than as wide-open consumers buying packaged foods from agribusiness. Net-food-importer countries would bear the worst of the damage, and the poorest of the poor could face large-scale deaths and regional instability.
IV. Priority IV: Create strategic reserves of emergency inputs
First among these, create a strategic electricity reserve: I.e., a reserve of already-generated electrical energy, rather than merely an energy reserve that would then have to be translated into useable form. An electricity reserve could take the form of whole mountains full of long-life batteries, ultracapacitors, a variety of liquid or solid fuels (reasonably stable, of course), or other approaches, and they could be as geographically distributed as you need to ensure their safety and the access of regional populations in the event of a crisis. If people haven't made enough progress on Priorities I-III by the time a crisis hits, the strategic electricity reserves would give them leeway to make that progress and begin moving forward. The Strategic Petroleum Reserve would also help, but may in some ways end up being used counterproductively - e.g., "Holy shit, we're running out of oil! Quick, let's gas up our Abrams tanks and go take other people's oil have before it's too late!"
In addition, of course, you'll want a couple of common-sense reserves: Strategic water reserve, and food reserves (currently our food reserves are just what other countries export, but in the event that it became uneconomical to haul food in bulk globally, we would need a local reserve). That can be done in any number of trivially imaginable ways, but once you've actually gotten to that point, you're really starting to build an economy robust against a potentially rocky transition to renewable energy.
Of course, there is also the possibility - one I don't discount at all - that the transition may be more or less smooth sailing even without these special precautions: There may be periodic spikes in prices here and there accompanied by new investment in renewables, and then the whole thing slowly cycles toward an equilibrium point exactly like what I've mentioned with a robust system. Brief shortages will be caused by market speculation, but that can only sort itself out. Now, this optimistic scenario is if fossil fuels ramp down at a manageable rate, rather than just going kaput, and I think there is reasonable basis for expecting that. Either way, it's good to think ahead. We'll be in a much better position after the transition if we've planned ahead for it, and won't have to wait as long to see some of the utopian possibilities of solarchy realized.