Humanity has pulled roughly 1.5 trillion barrels of oil out of the ground since commercial drilling began in the mid-nineteenth century. That figure, based on cumulative production records maintained by energy agencies and industry databases, is both staggering and surprisingly imprecise. Definitions of what counts as “oil,” gaps in historical data, and the murky boundary between conventional crude and other petroleum liquids all make the true total a moving target with a margin of uncertainty that would itself be a massive number.
The Running Total Through History
By the end of 2007, global cumulative production of conventional petroleum liquids stood at about 1,119 billion barrels, according to data compiled by BP and analyzed alongside U.S. Geological Survey resource estimates.1GeoArabia. World production of conventional petroleum liquids to 2030: A comparative overview In the years since, the world has been extracting between roughly 80 and 100 million barrels per day, adding several hundred billion barrels more to the cumulative tally. For context, the United States alone had produced 239 billion barrels through 2021, a figure that itself exceeded the absolute upper limit that petroleum geologist M. King Hubbert once considered possible for the country.2AAPG Bulletin. M. King Hubbert and the rise and fall of peak oil theory
Production didn’t ramp up overnight. The first half-century of the oil age, from the 1860s through about 1910, contributed a relatively tiny sliver of that total. The real acceleration came after World War II, when industrialization spread globally and automobile ownership exploded. By the early 1970s, the world was already producing tens of millions of barrels a day. The trajectory has been relentlessly upward since, interrupted only briefly by price crashes, recessions, and the COVID-19 pandemic.
Why the Number Is Harder to Pin Down Than It Sounds
Asking how much oil has been extracted sounds like a straightforward bookkeeping question. In practice, the answer depends on what you choose to count. “Crude oil” in the strictest sense means the liquid hydrocarbons pumped directly from underground reservoirs. But global production statistics typically lump in lease condensates, which are very light liquids that condense from gas at the wellhead, as well as natural gas liquids (NGLs), which are separated from natural gas during processing. Some tallies also fold in synthetic crude from Canadian oil sands, Venezuelan extra-heavy oil that must be upgraded before refining, and even biofuels or gas-to-liquids output. Depending on which of these categories you include, the cumulative total can shift by hundreds of billions of barrels.
The USGS resource assessment referenced in early peak-oil analyses, for instance, set an ultimate recoverable resource of about 2,770 billion barrels at 95-percent confidence, then added 152 billion barrels for Canadian oil sands on top of that.1GeoArabia. World production of conventional petroleum liquids to 2030: A comparative overview The inclusion or exclusion of unconventional sources like oil sands and tight oil from shale formations reshapes not just the “how much is left” question but also the “how much have we already taken” question. This is partly why you will see different agencies and researchers quoting different cumulative production figures: they are sometimes counting different things.
Reporting quality adds another wrinkle. Some producing nations have historically treated reserve and production data as state secrets, and independent auditing of national oil company figures has been limited. The numbers we work with are the best available, but they carry more uncertainty than the tidy columns in a spreadsheet might suggest.
Most of the Oil Underground Stays Underground
One of the most counterintuitive facts about oil extraction is that most of the petroleum in a reservoir never comes out. On average, about two-thirds of the original oil in place remains trapped in the rock after a field has been produced using conventional methods.3Developments in Petroleum Science. Factors affecting oil recovery from carbonate reservoirs and prediction of recovery That residual oil is held by capillary forces in tiny pore spaces, stuck in low-permeability zones, or bypassed by the injected water or gas used to push oil toward wells.
This means the 1.5 trillion or so barrels that humanity has extracted represent only a fraction of what was geologically present. Enhanced recovery techniques, such as injecting steam, carbon dioxide, or specialized chemicals, can coax out some additional oil, but they are expensive and energy-intensive. The theoretical promise of “getting the rest out” keeps petroleum engineers busy, but even optimistic projections assume recovery factors climbing to perhaps 40 or 50 percent in favorable reservoirs, still leaving half or more of the original resource in the ground.
Giant Fields and the Shape of Decline
Much of the world’s cumulative production has come from a relatively small number of very large fields. These giant fields, defined as those holding at least 500 million barrels of recoverable oil, take time to reach their stride. On average, about five and a half years pass between discovery and the start of production, and another 17 years before a giant field hits its peak output, which averages around 250,000 barrels per day.4Elsevier. Investment and production dynamics of conventional oil and unconventional tight oil: Implications for oil markets and climate strategies After that plateau, decline sets in at a pace of roughly seven percent per year.
This matters for the cumulative extraction question because many of the world’s largest fields were discovered decades ago. Saudi Arabia’s Ghawar, Kuwait’s Burgan, Mexico’s Cantarell: these fields poured out enormous volumes during their peak decades, and their contributions dominate the historical total. As these aging giants decline, newer production increasingly comes from smaller, harder-to-reach deposits and from unconventional sources like tight oil, which have steep per-well decline rates even though the overall formation may be vast. The consequence is that maintaining, let alone growing, the global extraction rate requires constant drilling of new wells.
What Happens to All That Oil
Most people think of oil as fuel, and that is mostly correct. The great majority of extracted petroleum is refined into gasoline, diesel, jet fuel, and heating oil, then burned. But a meaningful slice never sees a combustion chamber. Petrochemicals derived from oil and natural gas serve as the raw feedstock for plastics, synthetic fibers, fertilizers, pharmaceuticals, and thousands of other products. Recent estimates suggest that about four percent of global oil and gas consumption goes specifically toward producing plastics, translating to roughly 10.9 to 16.7 exajoules of fossil-fuel energy diverted to plastic feedstock in 2019 alone.5Elsevier. Sustainable materials alternative to petrochemical plastics pollution: A review analysis
That four percent sounds modest, but it adds up to hundreds of millions of tons of plastic produced each year, most of which persists in landfills and the environment for centuries. So the question of how much oil has been extracted is not just about energy consumed and carbon emitted. It is also about a vast quantity of material now embedded in everything from medical devices to food packaging to the microplastics turning up in ocean sediment and human blood. The cumulative legacy of extraction lives on long after the barrel is pumped.
The Physical Traces Left Behind
Pulling that much fluid out of the earth leaves marks beyond the chemical ones. Every barrel of oil brought to the surface comes with water, and lots of it. Produced water is the single largest by-product of petroleum production, with a global average water-to-oil ratio of roughly four to one, and that ratio climbs as wells age.6Elsevier. The size and performance of offshore produced water oil-removal technologies for reinjection Over the lifetime of the industry, this means tens of trillions of barrels of salty, often chemically laden water have been separated from oil and either reinjected underground, treated and discharged, or in some cases simply dumped. Managing that water is one of the oil industry’s biggest operational costs and environmental headaches.
Removing fluid from underground reservoirs also causes the ground above to sink. In the Coalinga oil field in California, decades of extraction produced a measured subsidence rate of about 3.3 millimeters per year on Anticline Ridge, consistent with what geomechanical models predicted from the volume of fluid withdrawn.7Journal of Geophysical Research: Solid Earth. Stress and subsidence resulting from subsurface fluid withdrawal in the epicentral region of the 1983 Coalinga Earthquake A few millimeters per year sounds trivial, but sustained over decades and across large areas, subsidence can damage infrastructure, alter drainage patterns, and in some coastal regions accelerate the risk of flooding.
Then there are the wells themselves. A comprehensive global inventory identified roughly 4.5 million abandoned oil and gas wells across 127 countries as of 2022, about 90 percent of which had never been properly plugged.8Oxford Academic. A global inventory of methane emissions from abandoned oil and gas wells and possible mitigation pathways These orphaned wells collectively leaked an estimated 0.4 million metric tons of methane in 2022. Without intervention, cumulative methane emissions from these wells between 2023 and 2050 are projected to reach about 9.9 million metric tons, a volume comparable to Spain’s total annual greenhouse gas output. The infrastructure of extraction, in other words, keeps leaking long after the last barrel is pumped.
Why Hubbert’s Peak Didn’t Play Out as Predicted
For decades, the question of how much oil the world could extract was framed through the lens of “peak oil,” the idea that global production would follow a bell-shaped curve and inevitably decline once roughly half the recoverable resource had been produced. M. King Hubbert famously predicted in the 1950s that U.S. oil production would peak around 1970, and it did, briefly earning his model a reputation for prescience. But the longer-term story tells a different tale. Cumulative U.S. production through 2021 reached 239 billion barrels, 39 billion more than what Hubbert had considered the absolute ceiling.2AAPG Bulletin. M. King Hubbert and the rise and fall of peak oil theory
What went wrong with the prediction was not the geology but the economics and technology. Hubbert’s framework assumed a fixed, knowable resource base. In reality, horizontal drilling, hydraulic fracturing, deepwater production, and improved recovery techniques kept unlocking oil that his model treated as inaccessible. Unconventional tight oil from formations like the Permian Basin and Bakken Shale pushed U.S. production to record highs in the 2010s, long after it was supposed to be in terminal decline. The global peak-oil narrative has shifted accordingly: the constraint on how much oil gets extracted is increasingly seen as economic and political rather than geological.
The Energy It Takes to Get Energy
Not all barrels are created equal from an energy standpoint. In the early days of the oil industry, wells gushed freely and the energy required to extract a barrel was a tiny fraction of the energy that barrel contained. This ratio, often called the energy return on investment, peaked for global oil production in the 1930s and 1940s at roughly 50 to 1, meaning that for every unit of energy spent on extraction, 50 units came back as usable petroleum.9Ecological Economics. Long-Term Estimates of the Energy-Return-on-Investment (EROI) of Coal, Oil, and Gas Global Productions It has been declining since. Field-level analyses confirm that individual oil fields experience substantial drops in their energy return over their productive lifetimes as the easy-to-reach oil is depleted and more energy-intensive techniques are needed to maintain output.10PLOS ONE. Estimating decades-long trends in petroleum field energy return on investment (EROI) with an engineering-based model
This declining return matters for the cumulative extraction question because it means the last trillion barrels will cost far more energy to produce than the first trillion did. Tar sands, ultra-deepwater wells, and enhanced recovery from depleted conventional fields all demand enormous energy inputs. The net energy delivered to society per barrel is shrinking even as the gross extraction total grows. From a practical standpoint, there is a point at which spending a barrel’s worth of energy to extract a barrel’s worth of oil ceases to make sense, and that threshold is creeping closer with each passing decade.
How Much Has to Stay in the Ground
Climate science has introduced a hard constraint that geologists of an earlier era never had to consider. Keeping global warming within 1.5 degrees Celsius above pre-industrial levels requires that nearly 60 percent of known oil reserves remain unextracted by 2050, along with 90 percent of coal and a comparable share of fossil methane gas.11Nature. Unextractable fossil fuels in a 1.5 °C world Even under a less ambitious 2-degree target, a large share of proven reserves would need to stay underground. Separate modeling has estimated that between 500 and 600 billion barrels of existing oil reserves would need to go unproduced by 2035 alone, depending on whether carbon capture and storage technology is widely deployed.12Energy Policy. Un-burnable oil: An examination of oil resource utilisation in a decarbonised energy system
This creates a paradox. The oil industry has spent over a century getting better at finding and extracting petroleum, and vast quantities of technically recoverable oil remain. But burning that oil would push warming well past internationally agreed limits. Some scenario analyses have found that the extremely high production rates assumed in certain climate models’ worst-case pathways are actually physically implausible given geological and investment constraints.13Current Research in Environmental Sustainability. How much oil remains for the world to produce? Comparing assessment methods, and separating fact from fiction That is a strange kind of good news: even without aggressive policy intervention, the highest-emission oil scenarios may be hard to reach. But the more moderate scenarios still involve enough extraction to blow past climate targets, which means deliberate choices about leaving oil underground will shape the final cumulative total as much as geology ever did.
The Gap Between What Exists and What Gets Counted
It is worth stepping back to appreciate the scale of the gap between oil that exists in the earth’s crust and oil that makes it into a barrel. Geologists distinguish between the original oil in place, which includes every drop of petroleum sitting in reservoir rock regardless of whether it can be reached, and the recoverable resource, which is the portion that current or foreseeable technology could actually extract at a reasonable cost. Original oil in place globally has been estimated in the trillions of barrels, dwarfing both cumulative production and proven reserves combined. But as noted earlier, only about a third of the oil in a typical reservoir comes out, and some reservoirs yield far less.
This means the 1.5 trillion barrels extracted so far is the product of a long chain of filters. First, geology had to concentrate hydrocarbons in a porous, permeable rock capped by an impermeable seal. Then explorers had to find those accumulations. Then engineers had to devise ways to get the oil flowing. Then the economics had to work at the prevailing oil price. At every stage, vast quantities of oil were excluded. The cumulative total sounds enormous, and it is, but it represents only a thin slice of the petroleum that nature created over hundreds of millions of years of organic matter burial and thermal cooking.
The share of that remaining resource that eventually gets pumped will depend on technology, price, and increasingly on policy. What we can say with confidence is that the 1.5 trillion barrels already extracted have reshaped the planet’s atmosphere, oceans, and landscapes in ways that are easier to measure than the oil itself. The tally keeps climbing by about 35 billion barrels a year, and the consequences compound with every addition.