EntryUpdated 2026/08/08

If We Switch to Clean Power, Does the Leverage Go Away?

No — the leverage moves. In the oil era it sat in sea lanes and spare production capacity; after electrification it sits in mineral processing, equipment production lines, and grid components. What those have in common is that you cannot stockpile them.

Read this first: When Hormuz Tightens, Why Does East Asia Feel It First?

Four common measures, each with something it cannot see

Measure	What it asks	Does a high score mean secure?	What it cannot see
Self-sufficiencyDomestic output as a share of consumption	How much energy need not be bought abroad. The most quoted measure, and the one most readily read as “secure”.	No. It sums different energies into one number without asking whether they substitute under stress — domestic coal does not cover missing gas.	Structure. The same ratio can mean five sources sharing the load or a single one carrying all of it.
Source diversityHow many countries and suppliers you buy from	Whether purchasing is concentrated. It corrects what self-sufficiency misses on structure, and is therefore often treated as the fix.	Not necessarily. The diversity is on the vendor list while the routes may overlap: five LNG suppliers whose cargoes all transit the same chokepoint are one route.	Overlap. The question is where the routes converge, and that does not appear in procurement statistics.
Switching speedFrom deciding to switch to actually running on it	How long it takes to run on another fuel or another route once supply stops. The only variable still movable during a crisis.	Yes — this one is hard to game, because it is not a statistic but a measurement: retrofit, procurement, and commissioning times cannot be hidden.	It is rarely computed. There is no ready statistical definition, so it means auditing your own equipment — which is why most reports have no such row.
AffordabilityEnergy as a share of household and industrial cost	Beyond availability: can it be afforded. Most energy crises are not outages but prices high enough that some uses stop first.	Depends who pays. Subsidy lowers the number on the bill without removing the cost — it moves who bears it, and detaches demand from price.	Distribution. An average that looks affordable can still be squeezing out an entire industry or an entire group first.

The four have to be read together, because each alone can be made to look good. Self-sufficiency sums unlike energies into one number, diversity counts suppliers rather than routes, and affordability can be held down by subsidy — all three can be improved on paper without improving anything real.

The one that resists flattery is switching speed, because it is a measurement rather than a statistic: retrofit, procurement, and commissioning times cannot be hidden. It is also the least likely to appear in an official report, for an understandable reason — there is no ready statistical definition, and computing it means auditing equipment one item at a time.

Source: Impactful Creative, compiled from the criteria described on this page and the glossary entries on energy security and chokepoints

Is energy security just about how much you produce yourself?

No. Countries with high self-sufficiency still lose power, and countries with very little of it can be extremely stable. The difference is not the ratio but how many options remain when supply is interrupted.

Energy security asks four things at once: can you buy it, can you move it, can you afford it, and can you switch. The first three are intuitive; the fourth is the one most often skipped, and it is the only variable still movable during a crisis — when you cannot buy, can you buy from someone else, burn something else, or route it another way.

Thinking of it as insurance rather than inventory is closer to the truth: what matters is not how much you stored in normal times but how many routes exist on the day something breaks.

So which measures should you look at?

Self-sufficiency, source diversity, switching speed, and affordability. They have to be read together, because each one alone can be made to look good.

Self-sufficiency has the most obvious flaw: it counts total energy without asking whether those sources substitute for one another under stress. A country can have ample domestic coal and still ration electricity for want of gas — the aggregate number will not show it.

Source diversity is often nominal. Buying LNG from five countries looks diversified, but if all five cargoes pass the same chokepoint there is really only one route. The question worth asking is where the routes overlap, not how many suppliers are on the list.

Switching speed is the hardest of the four to improve and the least likely to appear in a report: between deciding to burn something else and the equipment actually burning it lie procurement, retrofitting, and commissioning — measured in years.

Does switching to renewables make those dependencies go away?

They do not vanish; they change shape. Dependence on fuel becomes dependence on equipment and materials: turbine gearboxes and bearings, battery cathode material, transformers, high-voltage switchgear and cable — and the critical minerals processing that sits behind all of them.

One difference makes this dependence easy to miss: fuel is bought daily, equipment once a decade. What happens every day gets written into risk reports; what happens every ten years does not — until the year you need to replace something and discover lead times are measured in years.

The new dependence also sits in processing rather than geology. Where oil is depends on what is underground; the critical steps of electrification are mostly wherever someone was willing to invest and to carry the environmental cost. That matters because it can in principle be changed — but what it takes is not one subsidy, it is a decade of committed demand.

Why not simply stockpile more?

Because far less of an energy system can actually be stored than people assume. Crude and refined products keep for weeks in tank farms; gas can be stored at high cost and limited volume; electricity essentially cannot be — a grid must balance supply and demand every second, which is physics, not policy.

Equipment is harder still. Nobody holds a decade of transformers and high-voltage switchgear: they are expensive, bulky, and specified to a particular grid. Processing capacity is not an inventory question at all — you can stockpile ore, but not the plant that turns ore into material.

So supply chain resilience means considerably more in energy than “strategic reserve”. A reserve buys time, and time is only worth something if there is a second route to take — ninety days of stock with no alternative simply moves the same crisis ninety days later.

Only the top two layers can be stockpiled — and dependence is moving downward

Crude and refined productsWeeks
Ambient liquids, storable for long periods, fungible in specification — the category strategic reserves actually work for. Most national energy security policy is built on this layer, which is why it is easy to assume the whole system behaves the same way.
Natural gasDays
Storable, but only cryogenically or under pressure, in tanks that are expensive, limited in volume, and losing gas to boil-off every day. Also a fuel, yet the storable horizon is an order of magnitude shorter than oil’s.
ElectricityHours
Essentially unstorable. A grid must balance supply and demand every second; batteries cover hours of fluctuation, not weeks of shortfall. This is physics rather than underinvestment — the more electrified a system, the smaller the storable share.
Equipment and componentsLead times in years
Transformers, high-voltage switchgear, cable, gearboxes. Expensive, bulky, and specified to a particular grid, so nobody holds enough. They are bought once a decade and therefore rarely appear in risk reports — until the year one has to be replaced.
Processing and separation capacityNot an inventory question
Cannot be stored at all, because it is a capability rather than a good. You can stockpile ore but not the plant that turns ore into material — and this is precisely where dependence moves after electrification.

Storable — reserves work hereLimited — cost and volume both cap itNot storable — only alternative routes help

Strategic reserves were designed for the top layer: ambient, long-storable, fungible. Each layer below shortens the storable horizon by an order of magnitude, and by the time you reach equipment and capacity it is not an inventory question at all — those are capabilities, not goods.

The point of the figure is the direction of travel: electrification pushes dependence from the first layer down into the fourth and fifth — the side that cannot be stockpiled. So “store more” is not a wrong answer but one that covers a shrinking share of the problem. A reserve buys time, and time is only worth something if there is a second route to take.

Source: Impactful Creative, compiled from the storage conditions described on this page and the glossary entries on energy security and supply chain resilience

What does this mean for Taiwan?

Taiwan imports nearly all its energy and runs an isolated island grid. A European country short of power can import from a neighbour; Taiwan cannot — which makes the fourth test, can you switch, unusually important, because there is no external supply to fill a gap.

Taiwan is also taking on the new dependencies electrification brings: data centres and advanced-node fabs are both heavy electricity users, and the transformers, switchgear, and battery storage behind them all fall into the category that cannot be stockpiled.

For companies the useful move is not forecasting oil prices but answering two questions precisely: if this critical piece of equipment failed today, what is the lead time on a replacement; and do the suppliers on my list converge on the same firm three tiers up. Neither is an energy-policy question, and together they decide what happens on the day supply stops.

Terms used here

Energy Security — The ability to maintain energy supply at tolerable cost, resting on source diversification, alternative transport routes, and reserve depth. Energy security and decarbonization are not naturally aligned: securing near-term supply often means retaining high-carbon options, while the long-term transition creates new critical-mineral dependencies. Actual policy is usually a compromise between those two clocks.
Chokepoint — A narrow passage through which global trade or energy transport must pass, with no equivalent alternative — the Strait of Hormuz, Malacca, Suez, Panama. The risk does not require closure: rising insurance rates, added voyage days from rerouting, and delay costs typically reshape trade economics well before any actual interruption occurs.
Critical Minerals — Minerals that are irreplaceable for particular industries and whose supply is concentrated in a handful of countries — rare earths, gallium, germanium, graphite, lithium. Criticality usually stems not from where deposits lie but from where refining and processing sit: extraction can be diversified, whereas spec-compliant refining capacity takes years and extensive permitting to establish. That lag is the hardest element for policy to compress.
Supply Chain Resilience — The capacity of a supply chain to keep operating through disruption and recover afterward, measured by the number of qualified alternate sources, inventory buffer depth, and switching time. Resilience trades directly against efficiency: every redundant supplier is another qualification and management cost. The real question is never whether to have resilience but which specific links justify paying the premium.