IHS Markit has used artificial intelligence (AI) to assess, automate, and forecast future production for each of the nearly 1 million currently producing oil and gas wells in its North American databases—the first time this technology has been leveraged to forecast well-by-well future production.
The AI-based tool uses smart technology—it “learns” from adjacent wells in the same reservoir to improve its accuracy of forecasting. Another capability that mimics human behavior in terms of well analysis is the critical process of partitioning a well’s history over time to more accurately assess its future performance.
The evaluation assessed the entire North American production system, including both conventional and unconventional wells across the system. The results have yielded important insights into the challenges facing E&P companies.
Fast-moving treadmill
According to the results, North American onshore base oil-production decline rates are extremely rapid. Existing onshore producing wells will decline by 35% during the next 12 months. This compares with base decline rates of 5% to 14% for most petroleum systems globally during the same period, and a base decline rate of less than 15% a decade ago in the US. So, it is going to be costly for companies just to keep production flat.
In addition, the North American onshore base production-decline rates are accelerating in both absolute and percentage terms. Comparing 2017 with 2019, IHS Markit calculates that the onshore oil base decline nearly doubled (measured by the production drop from January through December of a given year, without accounting for wells added by new capital). Base production declined by 1.8 million b/d or 28% in 2017 but will fall by 3.5 million b/d or 35% in 2019.
“The treadmill that producers are fighting is moving very fast,” said Raoul LeBlanc, vice-president of North American unconventional oil and gas at IHS Markit. “As producers come under pressure to restrain investment, this decline rate is becoming the main factor that promises to slow the explosive US production growth we’ve witnessed the past few years.”
Predicting well productivity and estimating reserves has long been the lifeblood of oil and gas companies, LeBlanc said, since reserves estimates are a key metric in assessing the company’s value and future performance. Operators use detailed internal data to model their own wells, but engineers face limited information and arduous workflows when trying to quickly benchmark against competitors or screen for acquisitions.
“Most engineers don’t have the time they once did to conduct in-depth, well-level, decline-curve analysis on a handful of wells, so the prospect of rapidly analyzing thousands of wells is attractive, and we’ve delivered that analysis for nearly one million wells,” said Russell Roundtree, vice-president of upstream data analytics at IHS Markit. “This derived dataset becomes an invaluable addition to the oil and gas community and to financial investors who need to assess a company’s risk and future performance.”
Other findings
Despite the breathtaking plunge of 65% to 85% in first-year production declines that most young wells take, in a steady state, closed system, the dynamic (of steep declines) is sustainable because high-initial productivity offsets steep absolute declines.
Individual onshore US wells drilled today, with a few exceptions, follow a similar decline path as neighboring wells drilled in the past. In other words, companies are not “pulling” wells harder, and much more intensive hydraulic fracturing is not generally leading to faster decline rates.
The base-decline rates for North American assets vary directly with the production-weighted age of the well base. In other words, the overall base decline will vary significantly depending on how rapidly production expanded in the previous 1-2 years.
“Due to their unique producing portfolios, companies today are in very different positions relative to their current base-decline rates,” LeBlanc said. “The danger of high well production decline rates is twofold—namely vulnerability and degradation in production efficiency. Vulnerability involves anything leading to materially reduced completion activity, such as a price drop, weather, or a financing crisis. Volatility and quality degradation are the real threats,” LeBlanc said.
