Four years into the shale revolution, the U.S. is on track to pass Russia and Saudi Arabia as the world’s largest producer of crude oil, most analysts agree. When that happens and by how much, though, has produced disparate estimates that depend on uncertain factors ranging from progress in drilling technology to the availability of financing and the price of oil itself.

Forecasts for U.S. shale oil production vary from an increase of 7.5 million barrels per day by 2020—almost doubling current domestic output of 8.5 bpd—to a gain of 1.5 million bpd, or less than half of what Iraq now produces.

The disparities are a function of the novelty of the shale boom, which has consistently confounded forecasts. In 2012, the U.S. Energy Information Administration (EIA) estimated that production from eight selected shale oil fields would range from 700,000 bpd of so-called tight oil to 2.8 million bpd by 2035. A year later, those predictions had been surpassed.

“The key issue is not whether production grows, it’s by how much,” said Ed Morse, global head of commodities research at Citigroup in New York. “We’re only at the beginning of the first inning and this is a nine-inning game.”

The stakes couldn’t be bigger, ranging from the multibillion-dollar investments needed to explore and drill to oil supply issues that go to the heart of U.S. foreign policy. Relations with countries ranging from Iraq and Iran to Russia, Ukraine, Libya and Venezuela are colored to one degree or another by the question of energy.

The U.S., a nation transformed by the 1973 Arab oil embargo, could become energy independent by 2035, according to bullish forecasts from BP Plc and the International Energy Agency. Coupled with growing output from oil-rich neighbors, the continent has a growing shield from supply shocks.

“Looking at North America, including Canada and Mexico, we’re much more politically stable,” said Lisa Viscidi, program director of the Inter-American Dialog in Washington.

Still, many drillers have found that healthy forecasts of oil in the ground don’t guarantee it can be economically extracted.

For example, based on the promise of free-flowing oil, Chesapeake Energy’s then-top executive Aubrey McClendon bought up land in Ohio’s Utica shale oil field and touted it in 2011 as a $500-billion opportunity. State geologists estimated the shale play could hold as much as 5.5 billion barrels of reserves.

But last year, after months of drilling, Chesapeake’s average output per well per day was just 80 barrels. Competitor BP wrote off $521 million and exited the Utica just two years after leasing 85,000 acres.

Six Estimates

Shale production from the oldest shale patch, the Bakken of Montana and North Dakota, alone may rise to as much as 1.74 million barrels per day in the second half of this decade, according to the highest of six estimates compiled by Reuters. The lowest was 1 million bpd. Even that range belies disagreement over just how fast output will grow — and when it may peak. (Graphic: http://link.reuters.com/ref32w)

The EIA, the U.S. agency responsible for energy forecasts, predicts that tight oil output will rise 37 percent from about 3.5 million bpd in 2013 to 4.79 million barrels per day by 2020. The forecast includes the Bakken, Three Forks and Sanish, Eagle Ford, Woodford, Austin Chalk, Spraberry, Niobrara, Avalon/Bone Springs and Monterey.

“There are other forecasts that are much more optimistic than this one,” said agency administrator Adam Sieminski, speaking at a conference in New York. “We’re still a little concerned about what the geology looks like for crude oil production. As technology moves, these numbers could grow.”

The agency has already made some big adjustments to previous estimates. It recently slashed its forecast recoverable reserves for California’s Monterey shale to just 600 million barrels, 96 percent less than the total amount of oil in place, citing the difficulty in pumping it out economically.

IHS Energy’s projections are higher, with an estimated 6 million bpd from the Bakken, Eagle Ford and sections of the Permian and Niobrara by the end of 2020.

At the low end, Energy Aspects Ltd sees production of 3.5 million barrels a day from shale by 2017, a 1.5-million bpd increase from its current output estimate of 2 million bpd.

“In order to keep production going, you have to maintain your drilling and therefore, capex investments need to be in a continuous cycle,” said Virendra Chauhan, an oil analyst at Energy Aspects in London.

McKinsey & Co.’s forecasts illustrate the uncertainty. While the consulting firm uses a reference case that puts tight oil production at the equivalent of 7.1 million bpd by 2020, it said the number could range from 5 million to 9 million bpd.

In its annual outlook released last month, BP estimated that U.S. tight oil production will increase to 4.5 million bpd in 2035. Exxon Mobil Corp. says global tight oil production, driven by North America, will rise 11-fold from 2010 to 2040, when it will account for 5 percent of global liquids output. Exxon added that in 2015, North American tight oil supply in 2015 will likely surpass any other OPEC nation’s current oil production, with the exception of Saudi Arabia. Iran is the second largest OPEC producer, with about 4.2 million bpd.

Tricky Forecasts

Production forecasts are inherently problematic, especially years in the future, as they fail to anticipate major new discoveries or abrupt depletion rates.

Even so, the industry’s reliance on multi-year mega-projects such as those off the coast of Angola or in Brazil’s sub-salt region—which progress along generally predictable time frames and produce stable volumes of oil for years afterward—made it relatively simpler to anticipate new oil coming onto the market.

The shale oil industry is more complicated.

For instance, the rapid development of reserves in places like China and Russia could force prices lower, curtailing U.S. drilling. New technology may render development cheaper and more efficient, speeding it up. A change in current domestic policy, particularly an easing of the ban on crude exports, would also shape the forecasts.

Add to that growth the pipelines connecting Canadian producers to U.S. refiners, including TransCanada Corp’s 830,000 bpd Keystone XL pipeline, whose approval has been delayed by the U.S. government for more than five years.

Never mind the vagaries of the credit cycle, which has also become a larger part of the puzzle. Companies face high levels of reinvestment to ensure the same levels of return for drilling oil, meaning companies have to take on additional amounts of debt.

Consultancy Wood Mackenzie estimates that it would require capital spending of $9.58 to $32.97 a barrel to drill in the Eagle Ford basin, depending upon which part of the formation was targeted.

“We’re operating at present in a low interest rate environment, but a risk is what happens if the cost of credit rises,” Energy Aspects’ Chauhan said.

(Reporting by Catherine Ngai, editing by Jessica Resnick-Ault and John Pickering)