Drilling Digital Twin Predicts Drilling Dysfunctions in Real Time
- 1. 29-Sep-21 MindMesh Confidential © 2018 1 Drilling Digital Twin Predicts Drilling Dysfunctions in Real Time Raju Gandikota Chief of Innovations
- 2. 29-Sep-21 MindMesh Confidential © 2018 2 MindMesh Confidential © 2020 2 Drilling Performance Footage / Tool Life ROP Vibration Control Directional Effectiveness • Drilling programs continue to be pushedinto harder and more abrasive formations withsharper focus oneconomics. • Understanding whatis workingright, and howcan theybe improved are key to drilling performance. • Whatis not workingright, andhowcan they be eliminated or minimized are also key to drilling performance.
- 3. 29-Sep-21 MindMesh Confidential © 2018 3 MindMesh Confidential © 2020 3 • Computer based wellplanninganddrilling dynamics modeling is a standard practice for drillingperformance. • Standard BHA modeling software thoughfast havelimitations to include realistic downholebehavior. State of Digital Engineering Models CriticalSpeedMapof the BHA BHA deflection,BendingMoments & ContactForce
- 4. 29-Sep-21 MindMesh Confidential © 2018 4 MindMesh Confidential © 2020 4 Static vs. Dynamic • With Time Domain Analysis,complex behavior of BHA is clearly visible. • Large contact force at mud motor and MWD may cause failure • Thisis not predicted bystatic models • Time Domain is like getting live sensor data or video feed of yourBHA. • We drill over time! • We observe performance and dysfunctions overtime! • Why use static models to make dynamic decisions?? Traditional (snapshot) Time domain (video)
- 5. 29-Sep-21 MindMesh Confidential © 2018 5 MindMesh Confidential © 2020 5 • We drill over time! • We observe performance anddysfunctionsover time! Lateral Vibration Axial Vibration Torsional Vibration RPM DynamicDrilling Models
- 6. 29-Sep-21 MindMesh Confidential © 2018 6 MindMesh Confidential © 2020 6 • Model scalableto run from bit to top drive. • Multi-threaded with parallelcomputing capabilitiesto run multi-analysiswith same data feed • Physics model to handle most BHA’s and special tools like shock sub, agitators etc. • Integrate with real-timesystems • WITSMLfeed and run analysis continuously • Extensibleto handle OPC/UA Features of a Time Domain Model
- 7. 29-Sep-21 MindMesh Confidential © 2018 8 MindMesh Confidential © 2020 8 Time Domain Analysis – VirtualDrilling • Near Real Time drilling dysfunction prediction for optimization • Virtual Drill-off Test andsafe operating window
- 8. 29-Sep-21 MindMesh Confidential © 2018 9 MindMesh Confidential © 2020 9 RiMo Time Domain Analysis – Validation • Drilling with motor assembly • Downhole drilling dynamics captured with 50Hz memory sub
- 9. 29-Sep-21 MindMesh Confidential © 2018 10 MindMesh Confidential © 2020 10 RiMo Time Domain Analysis – Validation • Bit RPM evaluation shows stick-slip • RiMo captures the general stick-slip behavior with no additional models Stick-slip
- 10. 29-Sep-21 MindMesh Confidential © 2018 11 MindMesh Confidential © 2020 11 RiMo – Drilling Digital Twin • Real TimeModelswithReal TimeData – DrillingDigitalTwin • TheDigitalTwinuses real-timesurface data and drillstringdata to predict downhole drillingdynamics • Thedrillstring,BHA, and drillbit mechanics are utilizedfrom planning • Theworkflowis built to automatically recognizedrillingrigstates(rotary or slidedrilling)and connection makeupto start and stop the predictivemodel • Thetwin producesquantifiabledrillingeventslikeshock and vibration, downhole MSE,penetrationrate,etc.,in near real-time • Therealtimework flowcan also be used for post jobanalysis by streaming storedEDR data using theplayback function in RiMo
- 11. 29-Sep-21 MindMesh Confidential © 2018 12 MindMesh Confidential © 2020 12 Real TimeDrilling Dynamics • Predictdownhole response in new real time • Virtual drill-offtest & ImprovedrillingKPI’s
- 12. 29-Sep-21 MindMesh Confidential © 2018 13 MindMesh Confidential © 2020 13 Predict Performance Limiters • Lateral shocks were high at the bit & RSS • Lower drilling performance
- 13. 29-Sep-21 MindMesh Confidential © 2018 14 MindMesh Confidential © 2020 14 Improve Drilling Performance • Virtual step test with the goal of increasing ROP and reduced lateral vibration • Step 1, evaluate the shock & vibrations • Step 2, WOB increased by 10%, RPM reduced by 10% • Step 3, WOB increased by another 10% & RPM further reduced by 10% Lateral Shock Axial Shock ROP RPM WOB TOB BIT MSE BIT DEPTH
- 14. 29-Sep-21 MindMesh Confidential © 2018 15 MindMesh Confidential © 2020 15 Improve Drilling Performance- Virtual Drill-off Test Magnitude Bit RSS MWD >5G's 101 59 5 > 10G's 9 1 0 Shock count Observation Time Bit RSS MWD Initial 40 27 1 Optmization #1 37 10 1 Optmization #2 24 22 3 • Table #1 simulation shock counts > 5G’s & 10G’s are shown in table for entire simulation time • Table #2 shows shock count > 5G’s before and after optimization. The WOB & RPM changes do reduce the shock count and increase the ROP • Optimization runs do not affect axial vibrations Table #1 Table #2
- 15. 29-Sep-21 MindMesh Confidential © 2018 16 MindMesh Confidential © 2020 16 RiMo - Well Engineering Platform Planning – Real Time– Post JobAnalytics AdvancedPhysics Based Models Silos of Data Down Hole Vendor Topside Rig Materials DrillString Design DrillString Dynamics Digitally Enabling Your Enterprise Static Analysis Directional Tendencies Critical Speed Analysis Time Domain Analysis Soft String & T&D Stiff String & T&D Transient Hydraulics Silos of Data Wellbore Stability Steady State Hydraulics
- 16. 29-Sep-21 MindMesh Confidential © 2018 17 Building a new digital future www.mindmeshtech.com Contact: Raju Gandikota r.gandikota@mindmeshtech.com