North Dakota's rapidly rising oil output continues to defy the skeptics, who have predicted that production would stop growing as declining output from existing wells offsets extra production from new drilling.

Oil production soared to 911,000 barrels per day in August, up more than 200,000 bpd compared with the same month last year, the state's Department of Mineral Resources (DMR) said this week.

Production is on course to hit 1 million bpd by the end of the year or early 2014, according to the DMR.

By the end of August, 9,452 wells were in production. But another 450 had been drilled and were awaiting fracturing and completion.

Completions are running at about 1.5 times the threshold needed to maintain production, the DMR wrote in its monthly statement, which implies output will continue rising in the next few months as crews work through the backlog.

Shale skeptics have been confidently predicting since at least 2010, when output was below 300,000 bpd, that production would peak.

Only the DMR has struck a defiant and lonely optimistic note. In 2012 DMR projected output would plateau somewhere between 700,000 and 1.2 million bpd between 2015 and 2025, based on a total of up to 40,000 wells in the thermally mature part of the shale play.

Many out-of-state analysts, including leading energy consultancies, criticized those projections as overly positive. Now they appear conservative. So why did the skeptics get it so wrong?

Outrunning The Red Queen

Skeptics based their argument on the unusually rapid output decline from wells bored into shale formations compared with more conventional oil fields.

More and more holes would have to be drilled just to offset dwindling production from the existing stock of wells. More wells would require more drilling rigs and fracturing crews, which could not be increased indefinitely.

Eventually, production would reach an equilibrium based on some maximum feasible number of drilling rigs and crews.

"Yearly output must inevitably decline because the maintenance of a given output each year necessitates the drilling of an increasing number of wells," U.S. government geologist Carl Beal wrote in 1919, during an earlier panic about peaking domestic oil production.

More recently, The Oil Drum blog likened shale production to the Red Queen's Race in Lewis Carroll's book "Through the Looking-Glass."

"Here, you see, it takes all the running you can do, to keep in the same place," the Red Queen told Alice. "If you want to get somewhere else, you must run twice as fast."

Shale skeptics pointed to the tremendous variability in output from wells drilled into the most productive core areas of the Bakken compared with the less-prodigious outlying areas.

The core would be fully exploited quite quickly, they suggested, leaving only the less productive periphery, necessitating drilling even more wells with lower output.

Why Skeptics Were Wrong

In practice, the shale skeptics have proved wrong on every point, revealing a fundamental lack of understanding about the geology, economics and technology of shale production.

With more experience of horizontal drilling and fracturing and more knowledge about the play, drilling and pumping crews have been able to drill deeper wells and longer laterals, reaching total depth more quickly and applying more fracturing treatments per well as well as learning to target only the most productive parts of the formation.

"In 2007, the average treatment number, or stage count, in Bakken wells was three. By the end of 2011, that number was nearly 30, and some wells had more than 40 stages in a single lateral," according to oilfield-services specialist Schlumberger. ("Multistage Stimulation in Liquid-Rich Unconventional Formations" 2013)

Rather than increasing relentlessly, the number of drilling rigs operating in the Bakken has fallen from over 200 in early 2012 to just over 180 in October 2013.

Bakken accounts for just 10% of the rigs drilling for oil and gas in the United States, according to basin-by-basin data published by another oilfield services company, Baker Hughes. Rising Bakken output is not putting pressure on the domestic rig market. Rig rates remain soft.

More output with fewer rigs is a classic application of the learning curve effect, something shale skeptics overlooked.

Skeptics also drew the wrong conclusions about high decline rates. All oil and gas wells exhibit a sharp drop in output after the initial high rate of production in the first few months, as the natural pressure in the oil or gas field drops.

In general, the higher the initial output, the faster it will subsequently decline. Rapid decline rates are associated with unusually productive wells. Rapid declines also tend to be associated with high ultimate production over the lifetime of the well, as Beal demonstrated nearly a century ago. ("Decline and ultimate production of oil wells" 1919)

Skeptics often imply rapid decline rates are an unattractive feature of shale wells, ignoring the fact that production is declining from an unusually high initial rate.

The big upfront yield is what makes shale wells so economically attractive, resulting in a fast payback on investment and high rates of return. While skeptics worry about how much wells will be producing after 10 or 15 years, producers are more interested in how much they will produce within the first year or two.

The rates of return on shale wells are phenomenal. Continental Resources, the leading Bakken oil producer, claims it can achieve a 20 to 25% rate of return on shale wells even at oil prices as low as $60 per barrel, rising to 50 to 65% returns when oil prices are $100.

Manufacturing Oil and Gas

It is easy to overstate differences between conventional and unconventional production. Environmentalists hostile to shale oil and gas production highlight fears about groundwater contamination, seismicity, fugitive methane emissions and the chemicals used in fracking, without realizing the same issues apply to oil and gas produced from conventional fields as well.

How many realize that many conventional wells are also fracked, sometimes with acid rather than water, or that wells were being fractured in the 1950s and 1960s with diesel fuel and napalm?

In many ways, the differences between conventional and unconventional production are matters of degree rather than kind. But in one respect, the difference is profound and often underappreciated.

Conventional producers focus on finding highly concentrated accumulations of oil and gas and drilling a small number of highly productive holes. The process is akin to finding a needle in a haystack. It requires enormous investment in seismic and other surveys to try to improve the odds of drilling a successful hole, "de-risking" the process.

Producing shale oil and gas is much more like a manufacturing process. While some parts of the play will be more productive than others, the outcomes are far more predictable and less variable. The focus is on improving returns by reducing costs.

Standardization, assembly line techniques, reducing drilling and fracking time, and cutting the number of skilled workers needed while boosting the number of wells drilled and stages fracked lie at the heart of the business.

The differences between shale and conventional oil and gas plays help explain why shale specialists such as Continental and oilfield services companies such as Schlumberger, Halliburton and Baker Hughes have big winners in the Bakken and other shales, while oil majors such as Shell have struggled.

The manufacturing approach needed to succeed in shale is fundamentally different from the advanced engineering needed in deepwater exploration and on other complex megaprojects.

Shale's doubters failed to understand these differences. As a result, they treated shale plays as just another oilfield, rather than understanding the distinctive characteristics of the industry.