The Energy Cost of Energy

Why the data shows wind and solar — especially when paired with battery storage — are the most energetically expensive ways to power a civilization

There is a metric that energy debates almost never mention, and its omission distorts nearly every conversation about the future of power. It is called Energy Return on Energy Invested (EROEI), and it measures something fundamental: how much energy you get back for every unit of energy you spend producing it. The higher the ratio, the more surplus energy flows to hospitals, schools, manufacturing, and the standard of living that those things represent. The lower the ratio, the more of society’s resources get consumed simply keeping the lights on.

When this metric is applied rigorously to every major power source — and when wind and solar are evaluated not just as they perform on a sunny or breezy day but as they must perform to power a modern economy around the clock — the data reaches an uncomfortable conclusion: wind and solar are the most energetically expensive ways to produce reliable power, and pairing them with lithium-ion battery storage makes them more expensive still. Firmed solar, on the best available data, delivers a net energy return that sits below the estimated minimum required to sustain a modern industrial society. Firmed wind barely clears that floor. Meanwhile nuclear and hydropower — sources that politics and policy have often marginalized — outperform their renewable counterparts by a factor of 8 to 12 on a like-for-like, dispatchable basis.

This paper explains the EROEI metric, walks through the evidence source by source, and makes the case that any serious conversation about energy policy needs to reckon with what EROEI actually says — not what the industry brochures imply.

What Is EROEI?

Every energy source requires energy to produce. You have to mine coal, drill for oil, smelt steel for wind turbines, manufacture silicon wafers for solar panels, and build concrete dams for hydropower. Energy Return on Energy Invested (EROEI) is the ratio that captures this fundamental trade-off: how much energy do you get back for every unit of energy you spend to produce it?

The formula is simple:

EROEI = Energy Delivered ÷ Energy Invested

An EROEI of 30:1 means you invest one unit of energy and get thirty back — a net gain of twenty-nine. An EROEI of 3:1 means you invest one and get three — a net gain of just two. The higher the number, the more surplus energy available to power everything else in society.

Think of it like a business gross margin. A company with 95% gross margins has enormous resources to reinvest in growth, pay employees, and fund R&D. A company with 30% gross margins is still viable, but every dollar matters. A company with 5% gross margins is on life support. Energy systems work the same way.

“Higher EROEI = more surplus energy for everything civilization does beyond energy production itself.”

Why the Number Matters More Than You Think

Modern civilization runs on surplus energy — the energy left over after the energy sector feeds itself. Hospitals, universities, highways, manufacturing, agriculture, and the internet all run on this surplus. When the energy sector is highly efficient (high EROEI), a small fraction of society’s effort keeps the lights on and enormous resources flow to everything else. When the energy sector is inefficient (low EROEI), more and more of the economy gets consumed just maintaining energy supply.

Researchers have identified critical threshold values below which industrial civilization struggles to sustain itself:

• ~3:1 — Absolute floor. Barely enough to run the energy system. (Hall et al., 2009)

• ~7:1 — Minimum for a modern industrial economy. (Weissbach et al., 2013)

• ~10–14:1 — Needed for high-complexity society with healthcare, education, and infrastructure. (Hall)

• ~20–30:1 — Associated with a high standard of living. (Lambert et al., 2014)

The amber band on the chart below marks the 7–10 range — the zone where researchers flag the “net energy cliff,” a point at which an increasing share of total economic output gets consumed just keeping the energy system running.

Figure 1. EROEI by energy source. Blue bars = unbuffered (no storage). Red bars = firmed with 4–8 hour lithium-ion battery storage. Error bars show full literature range. Amber band = estimated societal minimum threshold (7–10:1).

Table 1. Summary of EROEI estimates by source.

Source by Source: What the Numbers Mean

Hydropower: The Gold Standard (EROEI: 40–110, midpoint 75)

Hydropower is the most energetically efficient power source ever deployed at scale. A concrete dam and steel turbines built once, operating for 60–100 years, require almost no ongoing energy inputs beyond maintenance. The fuel — flowing water — is free and requires zero processing. The result is an EROEI that dwarfs every alternative, with midpoint estimates around 75:1.

The wide range (40–110) reflects genuine site variability: a large dam on a high-gradient river in favorable geology is orders of magnitude more productive than a run-of-river installation in a flat landscape. But even at the low end, hydro delivers 5–10 times the net energy return of firmed wind or solar.

Nuclear: Dispatchable and Deeply Misunderstood (EROEI: 50–75, midpoint 63)

Nuclear power carries political baggage that its energy economics do not justify. An EROEI of 50–75 places it firmly in the top tier alongside hydro — and unlike hydro, nuclear is not constrained by geography. It can be built anywhere.

A critical methodological note: some studies report much lower nuclear EROEI figures by assuming that uranium enrichment is powered by coal-fired electricity via outdated gaseous diffusion technology. No such plants currently operate. Modern centrifuge enrichment is far less energy-intensive, and the EROEI figures in this analysis reflect current technology.

Nuclear’s decisive advantage in this analysis: it is fully dispatchable. It generates power on demand, 24 hours a day, regardless of weather, season, or time of day. No storage penalty applies. Its EROEI figure is its real-world delivered EROEI.

Coal: Declining But Still Competitive (EROEI: 26–45, midpoint 32)

Coal’s EROEI of roughly 32 reflects modern operational reality. Historical literature cited figures as high as 80, but those numbers come from 20th-century Appalachian surface mining — shallow, dense, easily accessible seams that no longer represent the global coal supply. Today’s operations involve deeper mines, mountaintop removal, longer transport distances, and lower-grade deposits.

Even so, 32:1 is a strong number. Coal is fully dispatchable, requires no storage, and its energy density makes transportation efficient. Its liability is not energetic but environmental: carbon dioxide emissions at scale that carry climate costs not captured in EROEI.

Natural Gas: Flexible, Efficient, Fading (EROEI: 20–40, midpoint 30)

Natural gas sits close to coal in net energy terms, with a midpoint around 30:1. The range is wide because the energy cost of extraction varies dramatically: conventional reservoirs drilled in the 1970s and 1980s had very high EROEIs; modern shale gas wells, which require hydraulic fracturing and rapid redrilling, are at the lower end.

Gas earns its place in most energy systems through flexibility. A gas peaker plant can go from cold to full power in minutes, making it the de facto balancing resource for grids with high renewable penetration. That role is energetically valuable — though it also means natural gas’s real-world EROEI should arguably be credited with some of the burden currently attributed to storage.

Wind (Unbuffered): Genuinely Competitive — Without Storage (EROEI: 18–34, midpoint 26)

Onshore wind’s unbuffered EROEI of roughly 26:1 is legitimately competitive with coal and gas. A modern wind turbine requires significant energy to manufacture — steel towers, fiberglass blades, copper wiring, rare earth magnets — but amortized over a 25-year operating life with no fuel cost, the return is respectable.

The critical word is unbuffered. These figures assume wind can deliver power whenever it generates it and that the grid absorbs or discards the rest. In a world where wind serves a small fraction of generation, that assumption roughly holds. As wind’s share grows, so does the imperative to store excess generation and dispatch it on demand. That is where the economics collapse.

Solar (Unbuffered): Location Is Everything (EROEI: 5–25, midpoint 15)

Solar’s unbuffered EROEI has the widest range of any source on this chart — and for good reason. A solar panel in the Arizona desert, where the sun shines 300 days a year, generates roughly three times the annual output of the same panel installed in Germany. The energy required to manufacture the panel is identical. The return is not.

At midpoint (15:1), solar is below coal, gas, nuclear, and hydro but above the societal minimum. In optimal locations, the upper end (25:1) is genuinely competitive. In low-irradiance markets, the lower end (5:1) is marginal before any storage costs are considered.

The Storage Problem: Why Firming Wind and Solar Is Economically Unattractive

The most consequential data in this analysis is not the height of the blue bars but the collapse of the red ones. When wind and solar are ‘firmed’ — converted from intermittent sources into dispatchable power using lithium-ion battery storage — their EROEI plummets:

• Wind: from 26 (unbuffered) to 7.5 (firmed) — a 71% reduction

• Solar: from 15 (unbuffered) to 5.5 (firmed) — a 63% reduction

Firmed solar’s midpoint of 5.5:1 sits below the societal minimum of 7:1. Firmed wind’s midpoint of 7.5:1 sits barely above it. Both assume only 4–8 hours of battery storage — enough to smooth daily variation, but not nearly enough to provide the seasonal and multi-day reliability that baseload power plants deliver. Achieving true baseload equivalence would require weeks of storage, collapsing these EROEI figures further still.

“The battery does not just store energy — it consumes it. Every electron that passes through a lithium-ion cell is taxed twice: once by the embodied energy cost of manufacturing the cell, and once by the round-trip efficiency loss of using it.”

Why Batteries Are So Energetically Expensive

To understand the storage penalty, consider the energy life cycle of a lithium-ion battery pack:

1. Manufacturing energy: Mining lithium, cobalt, nickel, and manganese; refining to battery-grade purity; synthesizing cathode and anode materials; assembling cells in dry rooms; forming and testing packs. This is enormous — studies estimate 60–150 kWh of energy consumed per kWh of battery capacity produced.

2. Short service life: Solar panels last 25–30 years. Wind turbines last 20–25 years. Lithium-ion grid batteries last 10–15 years. You may need to replace the battery system twice over the life of the generation asset, doubling the manufacturing energy burden.

3. Round-trip efficiency losses: Every charge-discharge cycle loses 10–15% of stored energy as heat. Over thousands of cycles and decades, this compounds into a substantial tax on every unit of electricity dispatched.

4. Shallow firming only: The 4–8 hour storage duration typical in today’s grid deployments addresses the daily cycle mismatch. It does nothing for multi-day wind droughts, week-long cloudy periods, or the fundamental seasonal mismatch where solar generates surplus in summer and deficits in winter. Addressing those would require 30–90 days of storage — a scale at which lithium-ion becomes energetically and economically absurd.

Researchers quantify this through a metric called ESOIe (Energy Stored on Electrical Energy Invested): the ratio of lifetime energy a storage device delivers to the energy required to build it. Pumped hydro scores ~210. Lithium-ion scores ~30–35. This 7x gap in storage energy efficiency is why replacing pumped hydro assumptions with lithium-ion assumptions actually produces worse firmed EROEI numbers than the already-pessimistic Weissbach estimates.

The Comparison That Matters

The practical question for energy planners is not “what is the unbuffered EROEI of wind?” but “what is the all-in EROEI of a system that delivers reliable power?” On that basis, the comparison is stark:

• Nuclear delivers 63:1 with no storage required, no weather dependence, and no geographic constraint.

• Hydro delivers 75:1 with no storage required, though geography limits deployment.

• Natural gas delivers 30:1 with no storage required and maximum operational flexibility.

• Wind + Li-ion delivers 7.5:1 — just above the societal minimum, assuming only short-duration firming.

• Solar + Li-ion delivers 5.5:1 — below the societal minimum even before seasonal storage is considered.

To put it plainly: a civilization that runs primarily on firmed wind and solar is operating its energy system at 5–8 times lower net energy efficiency than one running on nuclear or hydro. That gap does not just translate into higher electricity prices. It represents a structural reduction in the surplus energy available to fund everything else — healthcare, education, manufacturing, research, and the standard of living that those activities support.

Conclusion

EROEI is an inconvenient metric for a policy environment that has largely committed to a wind-and-solar-dominated future. It does not determine whether the energy transition is possible — human ingenuity and economic investment can overcome low EROEI, as they have before. But it does determine the cost of that transition: the fraction of societal resources that must be devoted to energy production rather than to the things energy enables.

The data in this analysis points to a hierarchy that politics often obscures. Hydro and nuclear are, by this measure, the most productive energy sources ever developed. They are dispatchable, high-returning, and require no storage penalty. Coal and gas are strong performers that carry environmental liabilities. Unbuffered wind and solar are viable minority contributors to a diversified grid. Firmed wind and solar, as a primary baseload strategy, carry a net energy burden that sits at or below the floor historically associated with modern industrial civilization.

None of this is an argument against wind and solar deployment. It is an argument for honesty about what they can and cannot do — and for not dismissing nuclear and hydro from the conversation on grounds that have nothing to do with their energy performance.

“The question is not whether we can build a civilization on 5:1 energy returns. The question is what we give up to do so and its total cost.”

Sources and Methodology Notes

• Weissbach, D. et al. (2013). “Energy intensities, EROIs, and energy payback times of electricity generating power plants.” Energy 52: 210–221.

• Barnhart, C.J. & Benson, S.M. (2013). “On the importance of reducing the energetic and material demands of electrical energy storage.” Energy & Environmental Science 6: 1083–1092.

• Hall, C.A.S., Balogh, S. & Murphy, D.J.R. (2009). “What is the minimum EROI that a sustainable society must have?” Energies 2: 25–47.

• Lambert, J.G., Hall, C.A.S. et al. (2014). “Energy, EROI and quality of life.” Energy Policy 64: 153–167.

• Raugei, M. et al. (2019). “Net energy analysis must not compare apples and oranges.” RSC Sustainable Energy & Fuels 3: 2047–2049.

• Fizaine, F. & Court, V. (2016). “Energy expenditure, economic growth, and the minimum EROI of society.” Energy Policy 95: 172–186.

• World Nuclear Association (2025). Energy Analysis of Power Systems. wna.org.

• Wikipedia contributors. “Energy return on investment.” Wikipedia, The Free Encyclopedia (accessed May 2026).