Alfredo García Fernández (@OperadorNuclear) is a science and technology communicator and one of the leading advocates of the role of nuclear energy in the fight against climate change. A nuclear operator and supervisor at the Ascó power plant since 2001, he combines his technical experience in the control room with intensive outreach work, which has been recognized with the Spanish Nuclear Society's Communication Award. He is the author of La energía nuclear salvará el mundo (Nuclear Energy Will Save the World) and Geoestrategia de la Bombilla (Geostrategy of the Light Bulb), and a regular voice in the energy debate in the media.

The title of your book, Nuclear Energy Will Save the World, is quite bold. Beyond the impact, what is the central technical argument behind this assertion?
Books, especially popular science books, should have a title that is eye-catching and provocative, but at the same time truthful and justified by the content. Aware of this, I justified the title of my first book in four different ways. First, nuclear energy is an essential tool for mitigating global warming and reducing air pollution, along with other technologies such as renewable energies. Second, nuclear medicine and nuclear diagnostic techniques save millions of lives every year. Third, when we achieve nuclear fusion, we will have an inexhaustible source of energy, with no risk of accidents and virtually no radioactive waste. And finally, I will leave the fourth justification for the reader to discover.

Nuclear energy continues to be perceived by society and some institutions as dangerous or inefficient. Are we facing a real risk problem or an ill-founded political and cultural veto?
Zero risk does not exist, as in any human activity, but the perception is completely distorted. If we analyze mortality per TWh generated, nuclear energy is one of the safest sources ever used by humanity, well below coal, oil, and even biomass. The problem is not technical, it is cultural. Radiation generates a different psychological reaction than smoke or air pollution, even though the latter kills many more people. In addition, energy decisions have become politicized. Nuclear energy is not evaluated on the basis of objective data, but rather on historical symbols. It is a sociopolitical veto built on three extraordinary accidents, two of which did not result in any fatalities due to radioactivity, rather than on the actual daily operation of hundreds of reactors.

You defend nuclear power as an essential tool alongside renewables. Why do you believe that renewables alone cannot guarantee decarbonization, supply, and industrial competitiveness?
Because electricity is not just energy, it is power available at every moment. Variable renewables depend on weather conditions and day and night cycles. A 100% renewable system requires massive storage or thermal backup, such as natural gas or coal. Today, massive battery storage simply does not exist on a national scale, and hydrogen is still inefficient and expensive. When there is no wind or sun for several days, something must sustain the grid. If that something is gas, there are more emissions and price volatility. Industry needs constant, not probabilistic, energy. Nuclear power eliminates that uncertainty. It allows large amounts of renewables to be integrated without making gas the real backbone of the system.

How should the model of complementarity between nuclear energy and renewables work in practice to achieve a stable and decarbonized electricity system?
It is a very simple physical model. Nuclear energy covers the stable base of the system, the permanent demand that always exists: hospitals, industry, digital networks, electrified transport. Renewables cover the variable part when the resource is available. This minimizes the use of gas. Nuclear energy does not compete with wind or solar energy, it competes with gas. In fact, the more nuclear energy there is, the more renewables you can install without destabilizing the grid or causing prices to skyrocket. The system ceases to depend on the weather and becomes dependent on controllable infrastructure.

From a technical point of view, what does nuclear energy bring to the electricity system that other technologies cannot offer today?
Three things simultaneously: firm power, grid stability, and independence from atmospheric phenomena. A nuclear reactor produces energy continuously for 18 months without interruption, except in the event of a breakdown. It also provides synchronous inertia and frequency control, which are essential for electrical stability, something that renewable power electronics cannot yet completely replace. And it does so with an extremely dense fuel: a few trucks per year compared to millions of tons of gas. It is an energy source designed to sustain complex electrical systems, not just to generate kilowatts.

The growth of data centers and artificial intelligence is driving up energy demand. Can nuclear energy become a strategic pillar for the digital infrastructure sector?
Without a doubt. AI doesn't just need cheap energy; it needs predictable energy. Servers can't shut down when the wind drops or night falls. Unlike other industries, a data center is severely penalized by interruptions. Nuclear power offers continuous, predictable, programmable, and stable production for decades. That makes electricity a reliable infrastructure, not a resource conditioned by the weather. In fact, the growth of intensive computing is reintroducing the need for firm generation into the global energy agenda.

Data centers need continuous, predictable, and competitively priced energy. Is nuclear one of the few technologies capable of meeting these requirements on a large scale?
Exactly. Nuclear has very low and predictable variable costs because fuel represents a small fraction of the total cost. That means its price does not depend on volatile international markets such as gas. It also operates with availability factors above 90%. For a data center, that's gold: constant, plannable electricity with little exposure to energy crises. It's not the fastest to build, but it is the most reliable for 60, 80 years, or even longer.

There is increasing talk of direct energy supply contracts for data centers. Do you see a model in which nuclear energy directly powers critical digital infrastructure as viable?
It is perfectly viable and is already beginning to be studied. Essentially, it is a PPA, but with one key feature: firmness of supply. A long-term contract with a nuclear power plant allows a digital operator to know its actual electricity cost for decades, something that is impossible with gas-linked electricity. For critical infrastructure, energy security is as important as cybersecurity. Nuclear power offers both: operational stability and economic stability.

From your experience, how is nuclear energy efficiency actually measured when analyzing its use in electricity-intensive sectors such as data centers?
It is not measured in thermal performance, it is measured in availability. For a data center, efficiency is continuity. A reactor operates at nominal power practically all year round. This means that the actual use of the electrical infrastructure is much higher than with other intermittent technologies. Less backup, less redundancy, less auxiliary storage, and less need to oversize the grid connection. Systemic efficiency is higher, even though people only look at the cost per MWh.

Can nuclear energy help stabilize electricity prices and reduce the volatility that currently concerns industries and data center operators?
Yes, because it decouples electricity prices from gas prices. In Europe, the marginal price is almost always set by combined cycle gas power plants. When nuclear production is high, the system needs less gas and price spikes disappear or are greatly reduced. It does not make electricity free, but it does make it predictable. For any electricity-intensive industry, predictability is almost more valuable than low prices.

The cost of nuclear power is often criticized. Does it make sense to evaluate its price without taking into account factors such as continuity of supply, useful life, and climate externalities?
No. It is a common methodological error. The cost of building a nuclear power plant is compared with the instantaneous cost of producing one MWh of renewable energy, ignoring backup, storage, grid, and avoided emissions. In other words, the costs of the system, which we customers also pay for. A nuclear power plant operates for six or seven decades. It spreads its investment over thousands of days of continuous production. If you include the cost of guaranteeing supply with renewables and gas, the comparison changes completely. Evaluating nuclear power solely on the basis of its CAPEX is like judging a hospital solely on the price of the building.

Cases such as Chernobyl and Fukushima continue to shape the debate. What has changed in terms of design, regulation, and operation to prevent something like this from happening again?
Practically everything. Reactors currently being built incorporate passive safety systems: they can be cooled for days without human intervention or external electricity. In addition, safety culture is now central to the industry. International oversight, periodic reviews, mandatory simulators, and extremely conservative operating protocols. Fukushima also changed the management of external risks: floods, total power loss, and extreme events are now assessed much more rigorously.

From your experience as an operator and supervisor, what is the real risk of nuclear energy today and how is it managed technically?
The real risk is not a Hollywood-style explosion, it is the loss of fuel cooling. All nuclear engineering revolves around preventing this. That is why there are multiple redundant, independent, and physically separate systems. Furthermore, the operator is not a passive watchman: we continuously train for accident scenarios in full-scope simulators controlled by powerful computers. Nuclear power plants are probably the most monitored industrial facilities in the world.

To conclude: in a scenario of massive electrification, AI, and growth in digital infrastructure, do you think Europe can afford to do without nuclear energy?
Not without serious economic and climate consequences. If you electrify transport, heating, and industry, and also power the digital economy, electricity demand will grow significantly. Giving up nuclear energy means replacing it with imported gas or accepting structurally high prices. Europe can decide not to use nuclear power, but it cannot decide the physical laws of the electricity system. And those laws say that a decarbonized system needs firm generation. And that firm, low-carbon generation is called nuclear power.