COLUMN-Frackers learn by doing: John Kemp

LONDON, Dec 11 (Reuters) - Shale oil and gas production will become a lot cheaper as the industry absorbs and shares lessons of advanced fracturing techniques, accelerating along a learning curve that started in the 1940s.

The concept of a learning curve is most often used with clean energy technologies such as wind power and solar.

But it applies equally to petroleum technologies such as hydraulic fracturing, where it may be having an even greater impact on costs and energy prices.

The idea that costs fall as industries get more experience with producing goods and services dates back to at least the 1930s (“Factors affecting the cost of airplanes” 1936). Subsequent studies have found evidence of “learning by doing” in industries as diverse as shipbuilding, semiconductors and solar photovoltaic cells.

Now it may be having the same effect on oil and gas, as exploration and production companies accumulate more experience with advanced techniques like horizontal drilling, fracking and reservoir characterisation.

Many analysts have made the mistake of assuming that because shale started as a comparatively high-cost form of oil and gas production it will remain expensive, helping underpin oil and gas prices at a high level in the medium and long-term.

Most believe real oil prices will not decline below $80 or $90 per barrel in the long-run because of the fairly high production costs of marginal supplies like shale.

That largely ignores learning effects. Efficiency improvements have already begun to cut the cost of horizontal drilling and hydraulic fracturing for cutting-edge producers in mature shale plays like the Barnett and Bakken.

As the technology is refined further, and efficiency gains diffuse across the industry, the potential for further cost reductions is significant and could lower the projected floor for future oil and gas prices.


Learning curves, sometimes called progress functions, relate unit costs to the total quantity of the good or service produced to date. The learning effect is normally expressed in terms of the percentage cost reduction for each doubling of cumulative output.

A wide range of learning rates has been found in different industries, but the most common progress ratio is around 81-82 percent: ie for every doubling of cumulative output unit costs fall by 20 percent.

“Technical change is a gradual process that evolves through different stages,” Cambridge University researchers Tooraj Jamasb and Jonathan Kohler wrote in a study of learning curves in the energy sector (“Learning curves for energy technology” 2007).

Jamasb and Kohler explained that in Joseph Schumpeter’s classic invention-innovation-diffusion paradigm, the first stage “invention refers to the generation of new knowledge and ideas. In the innovation stage, inventions are further developed and transformed into new products while diffusion is the widespread adoption of the new products.”

The whole process takes time. New technologies may take years to become widespread and for efficiencies to be apparent fully.


Fracking is not new. The first oil well was fracked experimentally in Grant County, Kansas, in 1947. The first commercial application came at wells near Duncan (Oklahoma) and Holliday (Texas) in 1949.

In the first year, 332 wells were fracked, with an average production increase of 75 percent. Since then fracturing treatments have been employed more than 2 million times on oil and gas wells, according to a history written in 2011 for the National Petroleum Council (NPC), an industry body which advises the U.S. Department of Energy.

Fracturing is now used on 95 percent of all new wells, and the technology is being continuously refined and modified to optimise fracturing networking and maximise resource production, according to the NPC (“Hydraulic fracturing: technology and practices” working paper #2-29).

But while the technology has been available for over 60 years, its full potential has only been realised in the last two decades, when it was coupled with horizontal drilling to unlock oil and gas in rock formations such as shale.

It first revolutionised gas production, starting with the Barnett shale in Texas in the late 1990s and early 2000s, and more recently oil output, starting with North Dakota’s Bakken shale from 2005.


Basic horizontal drilling and hydraulic fracturing technologies have now moved all the way from invention (1940s-1980s) and innovation (1990s and early 2000s) to diffusion throughout the entire industry (2005-2012).

There have been corresponding cost reductions. Drilling company Helmerich and Payne claims its latest generation of rotary rigs can drill a horizontal well in just 15 days compared with an industry average using older equipment of 30 days.

Efficiency improvements are confirmed by the leading exploration and production company Continental Resources , one of the Bakken pioneers, which says the number of wells that can be drilled by a single rig each year has risen from eight in 2010 to 11 in 2012 and forecasts it will hit 12 in 2013.

Fracturing is also becoming cheaper. Services company Schlumberger has criticised the wasteful brute force approach employed in the past. It overused horsepower, fractured too many stages of a well with too little prospective production, and failed to target the formations with the most productive potential.

In future, Schlumberger predicts that “smart fracking” will be targeted at only those horizontal well sections which promise the best oil and gas yields .

There is still plenty of scope to squeeze much more efficiency out by spreading best practices from innovators like Helmerich, Continental and Schlumberger to the rest of the industry.

The goal for all companies in the exploration, production and oilfield services sectors is to create highly standardised “frac factories”, which would bring the assembly-line efficiencies of Henry Ford’s motor manufacturing to oil and gas production, raising output and squeezing costs.


The full extent of efficiency improvements has been masked by the inflation in drilling and fracturing costs over the past decade. The upsurge in oil and gas drilling activity has been so sudden it has caused acute shortages and massive cost escalation in everything from fracking sand and guar gum to experienced rig crews and petroleum geologists.

But those shortages should start to abate as the supply chain responds. The bull market in oil and other commodities is now well into its tenth year. The availability of everything from fracking sand to experienced staff has begun to increase as companies and workers respond to incentives to enter the oil and gas industry. Costs are already coming down in some areas.

Oil industry inflation is essentially a short-term cyclical phenomenon (albeit the short run can extend 3-5 years or more in complex capital-intensive businesses like petroleum). In the next few years, some of the run up in costs and wages should reverse. Hiring rates for onshore rigs and especially pressure-pumping equipment have already begun to soften.

As inflation abates, the full impact of learning by doing and efficiency improvements will become much clearer. It will probably force analysts to revise their assumptions about the industry’s long-run marginal costs and oil prices.