International Energy Agency. 2025. Electricity 2025. Paris: IEA. Published February 14, 2025.
International Energy Agency. 2025. Energy and AI. Paris: IEA. Published April 10, 2025.
U.S. Energy Information Administration. 2025. ¡°Amid Regional Conflict, the Strait of Hormuz Remains Critical Oil Chokepoint.¡± Today in Energy. Washington, DC: EIA. Published June 16, 2025.
U.S. Energy Information Administration. 2025. World Oil Transit Chokepoints. Washington, DC: EIA. Updated 2025.
International Atomic Energy Agency. 2025. Nuclear Technology Review 2025. Vienna: IAEA. Published August 27, 2025.
The Simultaneous Return of Gas and Power
- The Counterattack of Stability in an Age of Volatility
[Key Message]
* The growth of renewable energy does not automatically make the power system stable. As solar and wind expand rapidly, their output variability is making grid operation more complex, not less.
* Energy is once again becoming a matter of security. Tensions in the Middle East and risks to maritime transport routes show that electricity and fuel supplies can still be deeply shaken by geopolitics.
* Gas and nuclear power are not fading away but reemerging as strategic assets. Gas is drawing attention as a flexible backup source, while nuclear power is being reevaluated as a low-carbon and stable source that can supply electricity around the clock.
* The standard of the electricity market is shifting from price competition to reliability competition. What matters now is not simply how cheaply power can be produced, but how steadily and uninterruptedly it can be delivered.
* In the AI era, industrial competitiveness is ultimately tied to power infrastructure competitiveness. Because data centers require massive amounts of electricity around the clock, regions and companies that can secure stable power supply are likely to hold a stronger advantage going forward.
***
For a long time, the energy transition was described as a straight line. As solar and wind power increased, coal and gas would naturally be pushed out, and the power system would move in an ever cleaner and more efficient direction. But real power grids do not move so simply. Renewable energy is certainly growing fast, yet power systems are actually becoming more sensitive and more complex. The sun and wind are the central axis of the future, but they are also forms of electricity that do not move precisely in step with the moments humans need them. As military tensions surrounding the Middle East, risks to maritime shipping routes, and the enormous constant electricity demand driven by AI and data centers are added on top of this, the world of electricity is once again returning to the old question of ¡°stability.¡± As a result, gas and conventional power, which once seemed to be on a path toward phase-out, are being called back as core resources. This is less a failure of decarbonization than a sign that reality has begun adjusting the ideal once again.
The Shadow of Clean Energy
The biggest change renewable energy has brought to power systems is that it has changed the very principle of generation. The power systems of the past were basically structured to follow demand. If people used more electricity, power plants generated more; if they used less, output was lowered. The rhythm of generation was matched to human consumption patterns. But solar and wind power overturn this order. Now electricity is not produced when it is needed, but when sunlight and wind allow it to be produced. The rhythm of supply has begun to be tied not to demand, but to natural conditions.
This change is not merely a technical characteristic. It is the starting point of systemic risk. Solar generation stops completely at night, and when clouds appear, output can fluctuate sharply even within a short period of time. Wind power also declines immediately when the wind weakens. The problem lies in the fact that electricity is unlike other goods. Because it is difficult to store electricity on a large scale for long periods, production and consumption must almost coincide in real time. When supply fluctuates, the entire grid enters a state of tension.
These characteristics are already becoming clearly visible in many countries. During the daytime, solar generation floods the system and electricity becomes excessive, but when the sun sets or the wind weakens, power is urgently needed again. In other words, a structure is taking hold in which surplus and shortage appear one after another within the same day. On the surface, this may look like a positive scene in which ¡°a lot of electricity is being produced,¡± but in actual operations the opposite is true. Grids have to balance themselves at much greater speed, and operators have to respond far more often and on a much larger scale.
California¡¯s duck curve is a symbolic example of this change. At midday, solar power pushes electricity supply upward, but by evening other power sources must ramp up output sharply within a short time. This means that the more renewable energy increases, the less the system stabilizes automatically; instead, it requires more sophisticated backup mechanisms and more flexible resources. In the end, renewable energy has become something that is both an answer and a source of new problems at the same time. The fact that the expansion of clean energy does not automatically mean an expansion of stability is now being confirmed across power systems around the world.
The Switch Shaken by War
Even the physical volatility of renewable energy alone has been enough to make the system highly complex, but recently the global energy market has taken on one more variable: geopolitics. In particular, military clashes surrounding Iran and the risks around the Strait of Hormuz are once again showing how quickly energy can shift from being an economic issue to a security issue. For energy-importing countries, what is most frightening is not simply price increases, but the sense that the supply chain itself can be shaken.
The Strait of Hormuz is a symbolic space. Tension in this narrow maritime corridor immediately spreads into crude oil and LNG prices, shipping costs, insurance premiums, procurement strategies, generation costs, and electricity rates. Energy prices do not fluctuate merely because of market psychology. They move this way because the world¡¯s electricity systems are still deeply tied to traditional infrastructure such as shipping lanes, tankers, terminals, long-term contracts, and geopolitical lines of control. The increase in renewable energy has not erased these chains. On the contrary, the larger electricity demand becomes, and the longer fossil fuels remain as backup power sources, the more strategic value these chains acquire.
Europe¡¯s shock during the Russian gas crisis shows this reality well. For a time, Europe expected that expanding renewable energy could achieve energy security and decarbonization at the same time. But in an actual crisis, classic tasks such as securing LNG, expanding reserves, building terminals, restarting existing power plants, and protecting industrial electricity supply all returned to the forefront. The energy transition was not a project that erased geopolitics; it was a project laid once again on top of a new geopolitics.
The message coming from recent tensions in the Middle East is the same. War does not eliminate the need for renewable energy. But it makes the importance of stable power sources and supply chains that can support those renewables far greater. There are clearly moments when what matters more than designing an ideal fuel mix is having facilities that can be run immediately and fuels that can be brought in immediately. And at precisely those moments, gas and conventional power shift from being ¡°uncomfortable but necessary¡± to being ¡°indispensable.¡±
Dispatchable Power Called Back Again
Under these conditions, the resource that has been restored most quickly is natural gas. The strengths of gas-fired power are clear. It can be turned on quickly, reduced quickly, and respond relatively nimbly to changes in demand and fluctuations in renewable output. The more unstable the power system becomes, the more the value of gas is assessed not simply by its fuel cost, but by its ¡°ability to adjust.¡± In the past, it was mainly criticized on the basis of its carbon emissions, but now its nature as a device that keeps the system standing is becoming more prominent.
The United States is one of the clearest places showing this trend. At the same time that renewable energy continues to expand, gas remains an important power source. In regions with high shares of wind power in particular, gas generation fills the gap when wind weakens, and steps back when wind is strong. This means gas and renewables are no longer in a simple competitive relationship, but are increasingly moving into a relationship of shared roles. If renewable energy is the direction of the system, gas is closer to a safety mechanism that keeps that direction from wavering.
In Europe as well, the position of gas has changed beyond that of a simple transitional resource. In the long run, batteries, hydrogen, grid interconnection, and demand response may replace part of the role of gas, but at this very moment the answer that is already commercialized at scale and can be used immediately is still gas. That is why gas remains an uncomfortable presence within clean-energy discourse, yet an increasingly important asset in actual operational reality. The ¡°return of gas¡± is not someone¡¯s political slogan; it is closer to the result demanded by the system itself.
Nuclear power is also drawing renewed attention for similar reasons. Nuclear is not as fast to start as gas, but it is a representative low-carbon power source capable of supplying electricity stably around the clock. Small modular reactors in particular are being discussed as candidates that could partially reduce the weaknesses of large nuclear plants, such as high upfront investment costs and long construction periods, and even as possible future partners for industrial complexes or data centers. Many challenges remain, but the very fact that power systems have begun once again to search for ¡°electricity that is always on¡± is enough to bring nuclear gradually back to the center of policy.
In the end, today¡¯s electricity systems cannot be explained by the simple confrontation of renewables versus fossil fuels. Rather, they are shifting into a multilayered structure in which renewables and gas, renewables and nuclear, and storage and demand response are all interlocked. What matters here is not completely driving out any one source, but precisely combining which source can support the system at which time of day and under which circumstances.
More Than Cheap, It Must Not Go Out
The standards of electricity markets are also changing quietly but fundamentally. In the past, the core market question was simple: who can generate electricity more cheaply? But now a completely different question is rising to the center of the market: who can guarantee electricity at the moment it is needed? Electricity is no longer just an energy commodity. It is becoming a service with timing and stability.
Even the same amount of electricity has different value depending on when it is produced. Power generated during hours of oversupply at midday may be low in value or even burdensome, while the same amount of electricity during the evening peak can be far more valuable. This change reveals that electricity is not just a matter of quantity, but a matter of how and when the balance of the system is sustained. That is why markets are increasingly assigning higher prices not only to the amount of electricity actually produced, but also to preparedness, reserve capacity, immediate response capability, and resilience.
At this point, capacity markets and flexibility compensation structures become important. Power plants have meaning even during hours when they are not selling electricity, simply because they remain ready to supply in a crisis. In the past, this kind of cost may have looked inefficient, but in an age of rising volatility, it is closer to an insurance premium. When the grid shakes, the damage to the entire system goes far beyond the profit and loss of one or two generating units. In the end, reserve capacity becomes not waste, but the price of stability.
Battery storage technology is the field drawing the greatest expectations in this transition. In fact, batteries are becoming powerful tools for peak shaving and short-term volatility response. But they are not yet a master key capable of solving every problem. They still face limits in dealing with prolonged low-wind conditions lasting several days, seasonal variations, fuel price spikes caused by war, and sudden surges in large-scale demand all at once. This is why the market cannot let go of gas and conventional stable power even while expanding batteries. This contradiction exists not only because technology is insufficient, but because the size of the safety margin the power system requires is itself that large.
In the end, electricity markets are moving from ¡°lowest-price competition¡± to ¡°nonstop-power competition.¡± In the future, what matters more may not be how cheaply electricity can be produced, but how stably it can be supplied even in times of crisis. This change reshapes not only the power mix, but also investment criteria, rate systems, corporate location strategies, and national industrial policy. Stability is no longer a hidden cost. It has become a central item in the economics of electricity.
The Nature of Demand Changed by AI
The change on the demand side is even more radical. AI and data centers are not simply industries that consume a lot of electricity. They are industries for which electricity ¡°must never be interrupted even once.¡± Traditional manufacturing cared about power costs, but today¡¯s digital infrastructure is far more sensitive to the quality and continuity of electricity itself. A single outage, a few seconds of voltage fluctuation, can lead to service disruptions and enormous losses.
For this reason, electricity demand in the AI era is changing not only in quantity, but in quality. The key is no longer simply how much electricity is used, but how uninterruptedly it can be received. Data centers must run 24 hours a day, and cooling, server operations, and network equipment are all designed on the assumption of stable power. That is why, in regions where data centers cluster, generation volume alone is not enough; transmission lines, substations, reserve capacity, and grid connection capability become even more important competitive advantages.
In some regions, this change has already become a social and political issue. When data center demand grows beyond what the grid can handle, new connection permits are delayed, and conflicts arise over priorities between industrial and household electricity use. Digital industry may appear to be an intangible industry, but in reality it is a highly material industry that requires enormous amounts of electricity, physical grid infrastructure, cooling systems, land, and water. That is why competition in AI is at the same time competition in power infrastructure.
In this flow, large technology companies are beginning to judge that simple renewable power purchase agreements are not enough. Long-term electricity procurement contracts, securing dedicated power sources, linking with nuclear plants, investment in grid strengthening, and cooperation with local power systems are all being considered at the same time. On the surface, this looks like a story about the AI industry, but in essence it is the opening of an era in which companies seek to procure their own energy security. The more this structure hardens, the more likely it becomes that companies with strong procurement capabilities will secure more stable electricity, while companies and regions without such capabilities will bear higher prices and greater instability. AI is not just a new industry. It is a variable that changes the very order of electricity allocation.
Five Scenes That Will Divide the Future
First, gas is likely to remain longer than expected, and in a more strategic way. Over the next decade, gas may not disappear quickly simply because it is a carbon-intensive fuel. On the contrary, the more renewable energy expands and the faster AI-driven demand grows, the more gas may acquire stronger meaning as a backup power source and flexibility resource. Some countries will likely try to redefine gas not as ¡°fuel that must be reduced,¡± but as ¡°manageable stable power,¡± by combining it with carbon capture technology. If that happens, competition over the international LNG market, shipping routes, and supply contracts may continue far longer than expected.
Second, nuclear power is likely to return not in the language of climate, but in the language of stability. Until now, the nuclear debate has largely been framed as a conflict between decarbonization and safety, but going forward, the value of nuclear as ¡°low-carbon power that can supply electricity around the clock¡± may become much more prominent. Small modular reactors, in particular, may be pursued in forms linked with large demand centers, industrial complexes, or data centers. Commercialization may proceed more slowly than hoped, but the direction of policy and finance is already moving toward placing greater emphasis on stable power sources.
Third, electricity is likely to become a more expensive and more complex product. The increase of renewable energy facilities does not automatically lower electricity rates. Costs are no longer simply about the construction cost of a single generator. They are expanding to include storage, reserve capacity, transmission grids, grid connections, backup facilities, operational flexibility, and supply-chain security. Consumers are likely to end up paying not only for the unit price of electricity, but for the cost of stability across the entire system. In this process, regions that secure cheap and stable electricity may become new industrial centers, while regions that fail to do so may be placed at a disadvantage in attracting investment.
Fourth, energy security is likely to return to the very center of national strategy. Even if renewable energy expands, geopolitics will not disappear. On the contrary, LNG, uranium, transmission equipment, transformers, battery raw materials, semiconductors, and cooling infrastructure may all be drawn more deeply into a security framework across the entire power system. The structure in which war in one region or tension in a strait shakes electricity costs and industrial activity across the world is likely to recur. Countries will be forced to redesign their energy strategies around not what they can buy cheaply, but what they can still procure in a crisis.
Fifth, in the end, the countries that move ahead are likely to be not the most environmentally pure ones, but the ones that design the most stable systems. The outcome of the energy transition will not be determined by installation volume alone. More important than how many solar panels or wind turbines are added may be the ability to make that electricity available whenever needed. Countries that expand transmission grids on time, compensate flexibility resources, connect AI and manufacturing demand with grid planning, and maintain backup power even in crises are likely to gain industrial competitiveness in the end. Clean energy is the direction; stability is the capacity to execute. Countries with direction alone will be shaken. Countries with execution capacity will endure.
An Age of Designing Stability
What is happening now is not a minor adjustment in which gas remains a little longer and renewable energy grows a little more slowly. The very perspective through which power systems are viewed is changing. Renewable energy is still the central axis of the future, but that alone does not automatically complete the system. On the contrary, the more renewable energy expands, and the more the digital economy demands electricity, the more the grid needs more stability devices and more long-range design. That is why gas and nuclear, storage, demand response, transmission grids, capacity markets, and energy security all matter at the same time. What matters is not choosing only one of them, but how to combine different power sources and institutions within a single system.
The simultaneous return of gas and power is not a scene of going back to the past. It is an expression of the fact that the future demands more backup and a higher level of stability than expected. The energy transition ahead is certainly a matter of what to expand and how quickly, but at the same time it is a matter of what will keep the system standing. Not only the speed of clean energy, but the ability to keep electricity from being interrupted even amid uncertainty. From now on, the real contest in energy is likely to be decided by who can design that ability with the greatest precision.
Reference
International Energy Agency. 2025. Electricity 2025. Paris: IEA. Published February 14, 2025.
International Energy Agency. 2025. Energy and AI. Paris: IEA. Published April 10, 2025.
U.S. Energy Information Administration. 2025. ¡°Amid Regional Conflict, the Strait of Hormuz Remains Critical Oil Chokepoint.¡± Today in Energy. Washington, DC: EIA. Published June 16, 2025.
U.S. Energy Information Administration. 2025. World Oil Transit Chokepoints. Washington, DC: EIA. Updated 2025.
International Atomic Energy Agency. 2025. Nuclear Technology Review 2025. Vienna: IAEA. Published August 27, 2025.