Maybe we’re approaching the digital oil field from the wrong angle.

Instead of moonshot efforts addressing complex data-intensive drilling or completion iterations, the narrative is really a “War of the Worlds” analog where the smallest things matter most.

In this case, the smallest of things attention-wise involves production optimization, the forgotten stepchild of actually generating the EUR promised by the IP.

The technology exists to optimize wellsite production, regardless of whether you are a multinational or large independent playing in the unconventional space or a private operator nursing production out of legacy conventional wells. The challenge is the adoption cycle.

Despite the focus on drilling and completion digitalization, the immediate low-hanging fruit may be found in production where personnel deal daily with wellbore construction problems originating from decreased drilling days and dogleg horizontals and completion practices that emphasize short-term IP versus long-term EUR.

The production team expends significant effort correcting unscheduled problems using an antiquated process involving pumpers and attendants making rounds in trucks to monitor production, oversee equipment and fend off relentless decline rates. It is a system based on trust and intuition, mostly involving repetitive tasks. It is a system ripe for optimization.

Two vendors of production optimization services outlined the possibilities at Hart Energy’s DUG Permian conference last spring.

San Antonio-based WellAware spoke specifically on the chemical support side of wellsite operations where data show the margin of error using tankbased monitoring is “plus or minus” 20%, according to company COO Blake Carlson. When sensors are placed on flowlines, the data are more accurate but show chronic under-delivery of product.

The key is high-resolution data, which prompted WellAware to pursue a full stack approach involving a hardware data sensor and communications network, app-based software for management using desktop or mobile systems and a backshop analytics engine.

Separately, Houston-based Ambyint focuses on automating artificial lift. Data from field deployment show almost 60% of rod lift systems are overpumping while 25% are underpumping. The former uses excess electricity, increases wear and tear and creates service cyclicality that leads to workover costs. The latter leaves money on the table.

Ambyint is developing an autonomous lift model, sort of a self-driving car of wellsite production optimization, combining sensors, software and an enormous North American well database derived from sampling data at 5 milliseconds versus 15- to 30-minute polling on traditional SCADA and PLC systems.

Field experiments in the Bakken Shale indicate the current platform, which adjusts set points for rod lift, shaved 15% out of lease operating expense. Once the system becomes fully autonomous, which Ambyint expects to demonstrate before year-end, total lease operating cost savings for artificial lift may exceed 30%.

Ambyint is developing a model to address field anomalies via machine learning that detect and characterize anomalies and autonomously corrects issues. Ambyint CEO Alex Robart foresees ultimate benefits accruing from a combination of production uplift and a reduction in lease operating cost.

So, the technology exists. Getting industry buy-in is another issue altogether.

“How do we transfer the oil field to an autonomous operating model away from the standard simple exception model?” Robart asked.

WellAware’s Carlson echoed the challenge. “This isn’t just a tech problem,” he said. “It’s a business model issue.”

Richard Mason’s Market Intelligence column originally appeared in the October 2018 issue of E&P.