• Rig-generated data provide a pathway to building better, lower cost wells.
  • Rig-generated data allow smaller E&P companies to participate in the Big Data revolution.

A drilling industry notion in the pre-rig class delineation days more than a decade ago postulated, “A rig is a rig is a rig.”

Adoption of alternating current variable frequency drive (AC-VFD) drilling technology ended the nostrum of the rig as a commodity service. Yet, the advent of Big Data analytics promises to revive that timeworn cliché in an unexpectedly positive way.

The Big Data movement in oil and gas is loosely bifurcated between production optimization, where benefits accrue to E&P companies from data mining archived well data, and the drilling process, which offers real-time data acquisition via the rig’s instrumentation package.

Those silos promise to converge into a holistic approach that integrates drilling, completion engineering and production optimization into a unified package that delivers better wells with lower lifetime costs.

Estimates on data volume from the drilling process are up to 7,000 items per second across 20 to 40 channels. But large volumes of data also carry the baggage of large volumes of noise. Efforts are underway to calibrate, normalize and identify the most important features of that rig-generated data stream. The process incorporates complex algorithms, development of key performance indicators and the ability to extract usable parameters for decision-making in real time for well construction via easily understood visualization.

In fact, the data, algorithms to parse the data and the ability to visualize and communicate the result exist now.

In February 2017 Exxon Mobil licensed its Drilling Advisory System and its machine learning algorithms to Pason Systems Inc., the largest provider of rental instrumentation packages for land-based drilling rigs. Pason recently completed a study that showed the licensed advisory system improved ROP by 35% over a 360-well dataset featuring similar properties.

In January 2017 Pason acquired Verdazo Analytics Inc., a software company that creates visual analytic tools, templates and customizable reports integrating public and proprietary data into machine learning processes. Verdazo is integrating wellbore porosity predictions captured via drilling to assist completion engineering.

The end game is to expand beyond drilling a faster lateral into building a better lateral, which leads to reduced costs and mechanical issues, not just in the completion process but compounded over a well’s lifecycle.

Service providers bring significant volumes of data to the table as every new well becomes a datapoint. The benefit from processes, such as machine learning, resides in extracting actionable options from large datasets curated in a way that provides quality information. The ability to capture, store and generate analytics via a simple interface is opening the path for machine learning in real time.

Instrumentation systems and analytics used in machine learning can fit modern AC-VFD rigs and be retrofitted for legacy diesel electric rigs. Technology is no longer the issue; the challenge is overcoming human foibles. Full adoption may require the creation of regional consortia of aggregated drilling data, which allow E&P companies to benefit while preserving individual well anonymity—a concept easier said than done.

Integrating real-time rig data to improve reservoir insight requires a holistic approach from an industry that is characterized by discrete silos of expertise, whether in the form of services or internally across the organization for E&P companies.

Doing so will create tangible benefits in hydrocarbon recovery and cost reduction via better wells by expanding communication between the ROP-focused drilling manager, the IP-focused completions engineer, the EUR-oriented production manager and the internal rate of return-oriented investor relations officer.

Comments