The study rests on a lineage of systems thinking that stretches back to Howard T. Odum, whose lifetime work established that nature obeys the basic thermodynamic rules at all times regardless of human opinion about it. Odum laid the foundations for ecosystem theory and made clear how ecosystems function as feedback systems linked with the rest of the physical world. Mass and energy balance always rules sovereign, and any prosperity gained from harvesting above sustainability limits will with certainty decline when the resource stock gets depleted. The study also builds on the pioneering World3 model developed by Donella and Dennis Meadows and their team at MIT in 1970–1972, which first demonstrated that the world’s resources are finite. Sverdrup and colleagues have since developed the model through WORLD4, WORLD5, WORLD6, and now WORLD7, a fully detailed Integrated Assessment model that links natural resources, population, energy, pollution, climate change, health, society, industrial dynamics, and the economy into one structure.

The WORLD7 model operates on a global scale with no countries and no regions—everything is a global average. It applies mass balance and energy balances to global resource flows, and it generates commodity and resource prices endogenously from first principles of supply and demand within the market arbitration mechanism over price. The model was operated in the STELLA Architect version 3.5 modeling environment and was used to estimate alternative energy capacities and resource use. The team ran simulations from 1850 to 2600, a 750-year span determined by the need to cover three times the longest delays in the system. System delays are of the order of 60 to 120 years—the lifetime for a human is 80 years, the lifetime for a major building is 100 to 200 years, and the turnover of carbon in soil ranges from 40 to 100 years. A minimum simulation time of at least 400 years is required, making normal political cycles of 2 to 4 years or budget plans of 5 to 10 years irrelevant and short-term by comparison.

The methodology involved estimating the available resource and reserves size, distributing this over cost of extraction categories, estimating realistic extraction costs for each resource quality grade in dollars per ton, using WORLD7 to simulate resource prices over time matched with historical data where available, estimating the extractable fraction of each deposit, and calculating the total revenue, running extraction costs, and market profit flowing to society. The team assessed 48 metals, minerals, and energy commodities, all published in peer-reviewed articles between 2011 and 2026. For each commodity, the model was tested against observed data for extraction rate, ore grade, and market price. The model is not subject to much calibration because it is based on physical properties and mass and energy balance constraints. All model parameters have physical significance and can in theory be measured. When the model runs, all metals are run simultaneously, and the model recreates the observed extraction rates and price trajectories for all 48 commodities at once. The team found that uncertainties do not build up with time in the model, unlike statistically based econometric models of the Leontief type, which build up errors exponentially and are limited to about a 5-year forward time-span.

The results reveal that the quality of all resources in the model—fossil fuels, metals, and materials—are all going down and have reached a quality between one-third in the best cases to more frequently one-thirtieth of what the quality was 120 years ago. This is a verified fact on field data: humanity has exhausted all high-quality natural resources, and only the poorest grades remain for current and future generations. The simulation outputs show that GDP will continue to grow until 2060, then turn down in a period of ups and downs until 2180, when a final decline and contraction sets in. After 2300, the final decline is fast, down to a level corresponding to the GDP in about 2005. From then, global energy is mainly derived from thorium. Without thorium energy, the plunge comes earlier, in 2130. The golden age of the Global Industrial Age runs from 1950 to 2250, or about 300 years, followed by a period of about 100 years of decline and turbulence, and then a hard landing with a parallel in what happened at the end of the Roman Empire—society simplified, fewer administrative levels, cancellation of large bureaucracies, collapse of complex organizations, and reduction in communication efficiency.

The team estimated the stock values and annual revenues of the world’s natural resources in monetary terms. The geosphere—stony material, metals, and soils—holds a stock value exceeding 5.7 quadrillion dollars with annual revenues of about 6.7 trillion dollars. Energy resources hold a combined stock value of about 167.7 quadrillion dollars with annual revenues of about 6.9 trillion dollars. The biosphere—land ecosystems, marine ecosystems, and forests—holds a combined stock value of about 133.7 quadrillion dollars with annual revenues of about 53.5 trillion dollars. The total across all three sectors amounts to about 307 quadrillion dollars in stock value with combined annual revenues of about 67 trillion dollars, compared to an observed global GDP of about 85.9 trillion dollars. The team notes that when ecosystem services are included for 2020, their fraction would be 61 percent of GDP, and if manufacturing to consumer goods is included, the total would amount to about 65 percent. Adding these together with direct sales value gives 156 percent, implying that GDP does not properly measure the value generation for society—it misses very substantial contributions, especially in the informal sector and environmental services.

Among all resources assessed, thorium emerges as the single most valuable energy resource on the list, provided thorium-based power plants are adopted. The stock profit difference between thorium and uranium comes from the high expenses of final waste disposal and nuclear arms proliferation associated with uranium. Thorium molten salt reactor technology has a fuel energy yield approaching 95 percent, creates no weaponizable material, and produces waste amounts one-fiftieth of conventional nuclear power with less severe radioactivity. The burn-out of thorium-related waste is completed in 250 to 300 years. To cover the future global energy deficit of 6,945 million tons of oil equivalents per year in 2060, the thorium needed by molten salt reactors has been estimated at 3,400 tons per year. With a total global resource of minimum 25 million tons of thorium, that would last 7,300 years. If the energy content is considered, the resource is more likely to be 100 million tons, resulting in about 30,000 years. If the population declines to 3 billion people, thorium could provide nearly all energy for about 50,000 years—equivalent to more than 1,000 generations. The team notes that China fully embraces thorium technology and already has operational reactors, India is running experimental reactors, and Copenhagen Atomics in Denmark has two commercial thorium molten salt reactor units standing ready, with one unit scheduled for technical verification and certification at the Paul Scherrer Institute in Switzerland in late 2027 or 2028.

The study also reveals critical dependencies between energy resources and technology metals. Wind power uses copper and Rare Earth Elements in significant amounts for generators and grid connections. Electric vehicles require copper, silver, aluminum, Rare Earth Elements, and cobalt for motors, and lithium, nickel, cobalt, vanadium, and graphite for batteries. High-performance turbine blades for jet engines and power plants use rhodium, palladium, tantalum, rhenium, zirconium, hafnium, niobium, molybdenum, cobalt, and nickel. Computer electronics use gallium, germanium, silver, copper, cobalt, gold, ruthenium, tantalum, Rare Earth Elements, tellurium, and selenium. The recycling rates for these critical metals are currently between 0 and 20 percent, which is making renewable energy technologies unsustainable in their current form. The team argues that decoupling and substitution are dead ends. Substitution on industrial scale only works when the element being substituted is significantly smaller in consumption volume than the resource used to substitute. Using aluminum to substitute for copper implies that the capacity used for substitution must be made available by canceling an existing use. The team concludes that a fundamental system change in material circularity is required before 2060.

The team combined their estimates with earlier work by Robert Costanza and colleagues on the value of the world’s ecosystem services. Costanza’s 1997 study estimated the annual value of all production and services at 33 trillion dollars per year, with a range from 16 to 54 trillion dollars. Updated valuations considering changes in land use and present ecosystem use increase this to 125 to 145 trillion dollars per year. The team applied an adjustment factor of 4 from 1997 to 2026 to lift these values to present time, compensating for inflation and changed utilization of ecosystems. Both this study and Costanza’s work indicate that natural resource values dwarf the annual GDP. The team also drew on the concept of Energy Return on Investment (EROI) developed by Charles Hall, Professor emeritus at SUNY, who studied with Odum. Hall’s concept is generalizable to a Resource Utility Return on Investment (RUROI) applied to the extraction of utility from any type of natural resource, finite or renewable. The team notes that the underlying assumption of traditional economic theory—that increased physical consumption of mass and energy could occur with a simultaneous decrease in negative sustainability impact—has been discovered to be flawed. What appeared to be observed as decoupling was just a manifestation of the classical system archetype of shifting the burden, where impacts were moved beyond the surveillance horizon through globalization.

The study presents four scenarios. Under Business-as-usual (BAU), the model suggests a period of significant decline after 2150–2200 with major social unrest, giving a prosperous period of about 250 years from 1900. Under BAU modified with better recycling, more circularity, and better efficiencies, the decline is delayed by 50 to 100 years, giving a prosperous period of about 400 years. Under the Ecosystem World scenario, with broad rules for metal conservation, renewable energies, and a global population scaled down from 8 to 2 billion by 2130, the decline is delayed to about 2350–2400, giving prosperity for about 450 to 500 years or 17 generations. Under the Thorium scenario, combining modular thorium molten salt reactor technology with critical metal conservation and renewable energies, the model suggests power for 8 billion people for at least 7,000 years, or 2 billion people for at least 40,000 years—equivalent to 1,700 generations. When the Ecosystem World scenario is combined with the Thorium scenario, the team’s simulations show that world civilization can be sustained for more than 1,000 generations. The team concludes that adjusting the settings of the present system is not sufficient for a fully sustainable society, and that natural resources are the root currency of human civilization, equivalent to what we call money—a conclusion concurrent with Odum’s 1996 seminal book on environmental accounting.

Subject of Research: The final count-down: estimating the total value of all the world's natural resources and an estimate of how long they may last

Article Title: The final count-down: estimating the total value of all the world's natural resources and an estimate of how long they may last

Article References: The final count-down: estimating the total value of all the world's natural resources and an estimate of how long they may last. (n.d.). https://doi.org/10.3389/fsrma.2026.1841357

Image Credits: AI Generated

DOI: 10.3389/fsrma.2026.1841357

Keywords: final, count-down, estimating, total, value, world's, natural, resources, estimate, long, they, last