Utopian Dreams of Net-Zero

By: Tore Andøl, economist and author of ‘How dare you’ and ‘Green rotten apples’

Ambitious climate goals, phasing out oil and other fossil fuels, green transition, and net zero are discussed daily. At the same time, Europe in particular, but also other regions are struggling with an energy crisis. This article takes a closer look at today’s energy situation and tries to assess the ambitions toward 2050. How dependent are we on fossil fuels, and how realistic are the politicians’ ambitions? Politicians refuse to quantify cost of climate action, but have they ever tried to test the realism of their targets?

Current Situation

Many have written extensively about energy trends and analytical interpretations of the annual datasets from the Energy Institute’s Statistical Review of World Energy. Let me just summarize a few highlights:

• Fossil energy still accounts for 86.2% of all energy consumption, down 2.9 percentage points since 2015 (primary energy).
• In the same period, fossil consumption has increased by 13,500 TWh (e.g. six times the annual oil and gas production from the Norwegian continental shelf), with gas making up just over half of this growth.
• The main growth sectors within renewables – solar and wind – account for 1.7% and 1.6% of total consumption respectively, up from 0.2% and 0.6% in 2015.
• Solar and wind consumption growth is impressive in relative terms, with average annual increases (CAGR) of 27% and 12.8% during the period. Absolute growth over 10 years exceeds 3,500 TWh, nearly 1.5 times the annual production from the Norwegian shelf (corresponds to almost 6 million barrels of oil equivalent per day).
• Growth in installed solar capacity alone was about 500 GW in 2025, adding production of about 650 TWh. If this growth continues at the same pace, solar alone could cover the world’s total energy consumption by 2050—during the hours when panels operate at full capacity, typically when the sun shines. More on that later.
• Despite these impressive growth numbers within renewables, fossil fuels have grown more in absolute terms. 56% of the 2026 total 2300 TWh growth (8 EJ) came from hydrocarbons, with growth in all categories. However, there are signs coal may be about to peak (US and Indonesia main growth consumers in 2025).
• Therefore, CO₂ emissions from energy continue to rise—up 0.9% from the previous year and 9.2% since 2015 (Paris Agreement). This corresponds to increases of 300 million tons in 2025 and 3,000 million tons over 2015–25.

Source: Energy Institute Statistical Review 2026. Illustration E. Årstad

Future Energy Mix and Capacity Needs

Although the numbers clearly show that fossil fuels still grow more than renewables and nuclear—both last year and since 2015—the underlying growth in solar, and to some extent onshore wind, is formidable, both relative and absolute. It is therefore tempting to make projections toward stated vision of net zero by 2050. How realistic is that?

For such an analysis, I make some assumptions (much of this can be debated, but this is a “law of large numbers” approach):

• Annual energy demand growth: 1% per year (below historical average since 1990, assuming 2% annual efficiency improvement vs real GDP growth of 3% historically).
• Average end-user energy utilization from fossil primary energy: 40%.
• Solar utilization: 15–20% (currently 13% globally).
• Wind utilization (onshore): 35% (currently 24%, incl offshore).
• Nuclear utilization: 90%.
• No significant energy losses in electricity consumption (optimistic).
• Growth in hydropower and biofuel in line with 2015–25 averages.
• Elimination of biomass.
• 50% solar and wind (split evenly) and 50% nuclear by 2050, plus growth in hydro and biofuel.
• Land requirements: wind 0.4 km²/turbine, solar 15 km²/GW.

With these assumptions, energy demand drops from today’s 600 EJ to 370 EJ in 2050 (due to higher efficiency and 1% annual growth). This equals about 12 TW of constant average power.

To cover all needs with solar and wind when weather permits (production equals installed capacity), we must:

• Install about 210 GW each of solar and wind per year for the next 25 years—or just solar at roughly today’s pace (~500 GW).
• This means ~40,000 new wind turbines per year (@ 5 MW), or over 1 million units by 2050. Annual growth equals 25–30% of all installed wind capacity globally today.
• Solar land area: 3,000 km² per year, or 80,000 km² total, plus 415,000 km² for wind (about 1.5 times mainland Norway’s area—or 3.5 times if assuming offshore wind).
• The world, and China in particular, shows such construction levels are possible in a single year for solar. Whether it can be repeated for 25 years straight (and indefinitely) is questionable.

But here’s the catch: To ensure enough energy when weather doesn’t cooperate (at least 50% of the time), we need just as much nuclear:

• About 420 traditional large plants with average 1 GW reactors per year for the next 25 years—installing more new capacity annually than the entire global nuclear sector today. More than one new plant every day. 10,000 new plants by 2050. Today, 60–70 GW is under construction, with about 1 GW per plant.

In total, this means doubling the installed capacity compared to average demand, because solar and wind need 100% backup. Since electricity storage (BESS) is currently far more expensive[1] than building double capacity (hydropower reservoirs are negligible globally), this is the only option.

Which raises an uncomfortable question: If we must invest in nuclear for the entire need, why bother with solar and wind at all? Nuclear can run 24/7, even when it’s sunny or windy.

Some argue for synergies and geographical diversification, but that’s flawed because:

• Weather is often regionally homogeneous and correlated.
• This thinking requires everyone to overinvest to cater for local needs and neighbors’ needs when weather doesn’t play along.
• Which in turn requires massive transmission capacity regionally and across regions.

If this option (solar+wind only) was feasible, the installed solar and wind capacity mentioned above would need to triple to quintuple. Homogeneous weather and costly storage/transmission, however, kill that idea (and time zones don’t help, in a global – solar-driven – perspective).

Many people I discuss with, e.g. on social media dispute the claim that diversity and diversification don’t work. So let me give an example:

Three countries, with independent weather systems, each consume 100 units of energy—constantly. That’s a total of 300. Each country invests in wind power capacity equivalent to 100 units, which operates 8 out of 24 hours (33% yield on average, regardless of time horizon). They assume that “neighbors” will supply the remaining 16 hours when it’s windy abroad. But everyone has invested for their own consumption when it’s windy, so no one has any surplus to share. Therefore, all three must invest in 300 units, plus build transmission lines between all points.

Alternatively, instead of transmission lines, they could all invest in batteries—but they would still need to build production capacity of 300 units each plus 200 units of battery capacity. When it’s windy, 100 units are consumed and 200 are stored in batteries for the remaining 16 hours.

This effect is unavoidable, no matter how many players or weather zones you include in the calculation.

A somewhat interesting takeaway from the above numbers is that the total needed installed production capacity in 2050 would be almost unchanged compared to today’s mix (without economic growth), which is dominated by fossil fuels. This is completely contrary to what many claim, often referring to “the primary energy fallacy.” They correctly argue that demand will decrease since burning fossil fuels largely is wasting energy through heat loss and inefficient combustion.

But—because production capacity must be doubled (or more) due to weather dependency (if we rely on these technologies), this ends up balancing out with my assumed mix. If, on the other hand, one assumed only solar + wind (with storage or sufficient grid capacity to share production across weather zones), the need for installed capacity would increase compared to today. The utilization rate for solar and onshore wind equipment is on average lower than the energy utilization of fossil sources.

Many claim renewables are cheap because the energy, the wind and sunrays, come without a cost. But who said Mother Earth charges for oil, gas, and coal? Last I checked, these energy sources were also free. It “only” costs to extract them – just like solar and wind.

So, in terms of cost, this is a competition for economies of scale in production facilities per unit of productive (and reliable) end-user energy. This must include the need for backup and storage. Lifespan is also a factor. Solar and wind must be reinvested at a frequency so that, in a 2050 perspective, annual capacity additions must continue “forever”.

This analysis is, of course, simplified when it comes to, for example, the need for products for petrochemicals, fertilizers, etc. Green ammonia and fertilizer produced via electrolysis will, for instance, increase power demand by about 50% (compared to blue ammonia with CCS, and even more compared to today’s grey ammonia).

So, a net-zero society will probably increase demand beyond my calculations. Neither have I accounted for uneven demand, which increases capacity needs. This is partly covered through double systems, assuming demand peaks occur when solar and wind produce -something that may not be wise to plan for.

Costs

I’ve made some rough cost estimates. This won’t be cheap. Just on the supply side, costs for these production facilities are above $60 trillion, not far from the combined value of all U.S.-listed companies and around half of global market capitalization. Add grid needs, possible growth beyond 1% annually, and massive consumer-side investments (electrifying everything—cars, planes, tractors, trucks, ships, machines, industrial plants, trains, heating, cooking, your mother-in-law etc.), and we’re talking numbers over 25 years equivalent to many years of global GDP. And annual costs, if we choose large-scale solar and wind, must be repeated indefinitely due to lifespan.

I haven’t said it yet, but in short, this is utterly unattainable in 25 years – physically (not enough labor, raw materials, land, logistics, expertise) and economically.

This will take much longer. Therefore, more investment in fossil exploration and development is needed to offset natural decline from existing fields. Otherwise, by the 2030s and beyond, we risk an energy shortfall that could be life-threatening for the poorest and cause more geopolitical turmoil.

Summary

Key takeaways:

• Enough installed capacity to meet demand at all times is critical.
• Weather-dependent solutions need backup for when conditions fail.
• Overinvesting in weather-based solutions beyond peak demand is pointless without affordable and technical feasible storage or interregional sharing/transmission through costly new grids.
• Since backup—mainly nuclear (or gas plants with CCS) —must cover full demand when renewables don’t produce, large-scale solar and wind seem almost meaningless. They may serve local and remote needs on a small scale with short, affordable storage for variable demand.
• Physically, this transition is impossible in such a short time. Even nuclear alone at this scale would take far longer than 25 years to build.
• Cost estimates are secondary to physical constraints. The only “savior” would be a quantum leap in battery/storage costs. Solar is being installed at an impressive pace, but to rely on e.g. solar + batteries alone, deployment would need to increase 5–6 times (~2,300 GW/year) to reach net zero by 2050. Plus 8,000 GWh of new battery capacity annually (for 19 hours of consumption when panels don’t produce). Such annual battery growth is around 30 times higher than capacity added in 2025. Just to survive for 24h, given normal weather.
  
  At current prices, such annual batteries add-on would cost about $1-1.5 trillion for a day of normal solar conditions—or $30-40 trillion by 2050. Add another day of weather buffer? Throw in another $40 trillion. Completely utopian, taking into account weeks of buffer would be needed in a real-life system.
• Halting fossil exploration would be extremely risky and could cause a severe energy shortfall during the transition. Phasing out oil by 2040, as some advocate, would be catastrophic – hundreds of millions, if not billions, would die within months.

So, for now, “drill baby, drill” seems like the safest strategy. Whether one believes in the CO2 hypothesis or not, more fossils are existential. Both to ensure progress and growth, and to cope with climate change, manmade or not.

Appendix: Key Calculations and assumptions

• Current energy consumption: 600 EJ
• Fossil share: 86.2% (primary energy)
• Waste from fossil use: 60% average
• Energy need (today) with pure electricity: 600 – 600 × 86.2% × 60% = 290 EJ
• Energy need with 1% net growth (2050): 290 EJ × 1.01^25 = 370 EJ
• Required capacity in 2050: 370 EJ × 278 TWh/EJ ÷ 365 ÷ 24 = ~12 TW
• Assuming all installed solar and wind capacity today must be replaced by 2050:
  • Nuclear (minus hydro and biofuel):
    (12,000 GW – 1,500) ÷ 1 GW ÷ 25 years ≈ 420 per year
  • Solar and wind (50/50 split):
    (12,000 – 1,500) × 50% ÷ 25 = 210 GW per year (420 total)
• Battery cost: 100-150,000 usd/MWh

[1] When taking into account buffer needed for days or weeks of weather outages.