Updated study finds use of AI, machine learning and predictive analytics in well construction rids downtime and hazardous incidents

To date we have researched 90+ wells where eDrilling’s digital twin technologies has been used in i) preparations for well, ii) automated monitoring of well, and iii) real-time optimization of well – or a combination of these - and found that these wells has had no hazardous incidents and no sidetracks.

Time
flies, it is a year ago we published our first study.  We have added 30 more digital twins over the
past 12 months, and the analysis now include 90+ wells where a digital twin has
been used in i) preparations for well, ii) automated monitoring of well, and
iii) real-time optimization of well – or a combination of these,  And still, we find that

i)          0 –
zero – had to carry out a sidetrack

ii)         None
of the wells had hazardous incidents

The
software detects at a very early stage any anomalies, and provide crew with
warnings, notifications, and diagnostic messages in good time to make the
necessary adjustments.

We continue to think is the biggest opportunity available today for performance optimization, avoidance of NPT, and hazard prevention, and are grateful to our customers around the world for the continued cooperation and success.

Methodology

Analysis of 90+
wells where our technologies have been actively utilised in preparation and/or
operation, and where we have had correct and sufficient data to conclude on the
hypothesis.   Thus, 50+
of our
digital twins in the period have
been omitted, due
to availability (contractual, geographical,
policy) of data.

Sidetracks, or avoidance thereof, is the focal area
of the study, although the study have
also found that
there had been 0 hazardous incidents on the 90+ wells analysed.

«Geological surprises» as reason for having to
do a sidetrack are omitted.  Cuttings build up, collapse, washout, and packoff leading to plugging the well, are included.

We have had 5100+ participants in drill
well in simulator prior to operations, with 150+ well specific scenarios.

Automated Monitoring and Real time Optimisation is used in all drilling operations. This includes Tripping, Drilling, Casing/Liner operations and Cementing. It is also used in advanced operations like MPD and DGD.

Digital Twin

A Digital Twin is the digital representation of your actual well. Used in preparation it enables you to pre-drill your well. Used in real-time operation it does predictive analytics, as well as forward simulations and what-if simulations to detect mismatch and give input to change the drilling plan – to avoid problems, see and detect operational mismatches, and change parameters for optimal drilling (drilling the perfect well).

Findings

Of the 90+ wells analysed, 0 – zero
– has had to carry out a sidetrack.  As the software at a very
early stage detects anomalies, crews are provided diagnostics messages and
notifications in good time to make the necessary adjustments.

Not a prime objective of the study, but included anyways as it continues to be a high priority for us, and for the industry as a whole, was avoidance of hazardous incidents.  We are happy, and proud, to say that we have had a 100% success rate here too.

Performance Optimization

We have not included performance optimization results in the study.  Mostly because we, and our customers, want to have the study be as focused as possible, and not try to cover all aspects at the same time.  That said, we have had some considerable success stories here too, with a European IOC estimating 19 days saved while drilling a vertical exploration well in the North Sea, as the most prolific.

Products involved

wəllSim hiDRILL is our solution for drill own well in
simulator.  wəllSim hiDRILL
provides safe, effective, and efficient well delivery for any planned well by
giving crew skills and experience on the specific well.

wəllAhead is eDrilling’s software solution for automated
monitoring, real-time optimization, and live well support. It is designed for
use on rig, in Real Time Decision Support Centers, and for any other individual or teams
supporting live drilling operations. It is also the foundation of any future
rig-side automation.

Example scenario from preparation session.  Here crews pre-drill high-risk sections and likely scenarios
Example scenario from preparation session. Here crews pre-drill high-risk sections and likely scenarios
Example scenario from a live well. Here crews are alerted to early detections of anomalies, and warned/remided about upcoming risks

eDrilling is a world leading supplier of AI, machine learning, and predictive analytics solutions to the oil and gas industry.  We work closely with E&P companies, operators, and service companies to help them save cost, improve safety, and increase efficiency of drilling operations.

Our mission is to help energy professionals at E&P companies, operators and service companies reach their goals by providing drilling and well performance solutions needed to better plan, prepare for, perform and evaluate drilling operations. 

Our vision dates back to the late 1980s, when we wanted to create a downhole GPS, or Google Maps if you like, for drilling operations. Convinced artificial intelligence and machine learning technologies would represent a quantum leap in well construction, we set forth what has become a paradigm shift in the oil and gas industry. For over 25 years, leading E&P companies, service companies, and drilling contractors, have leveraged the power of these technologies to improve drilling performance, onshore and offshore, deep water and shallow water, complex wells and factory drilling. Globally.

Hazard prevention

Prior to operations, eDrilling provides you with software to
acquire skills and experience on you specific upcoming well, focusing on risk
avoidance and handling. During operations, eDrilling software at a very early
stage detects anomalies, crews are provided diagnostics messages and
notifications in good time to make the necessary adjustments.

Performance optimization

Our software provides forecasting, forward-looking and
what-if analysis, allowing for change in parameters for optimal drilling
(drilling the perfect well).

NPT avoidance

As above, eDrilling uses aritifical intelligence and
predictive analytics to foresee what will happen with your well. Software
provides diagnostics and first actions to the crew (or directly to the control
system), in order to change the drilling plan – to avoid problems

The post Updated study finds use of AI, machine learning and predictive analytics in well construction rids downtime and hazardous incidents appeared first on Digital Twin.