Investigating the physics of electromagnetics over steel cased wells

Presented at the Lawrence Berkeley National Lab Geophysics Seminar.

Economic and social motivations to understand the impact of activities such as hydraulic fracturing and carbon capture and storage continue to increase. Electromagnetic (EM) imaging techniques have the potential to be a valuable method for monitoring these activities. One particular challenge is the presence of steel-cased wells. Steel has both a high electrical conductivity (~106 S/m) and a significant magnetic permeability (~100𝜇0), thus, it can considerably (and non-intuitively) impact the behaviour of the currents, fields, and fluxes in an EM survey. Several authors have demonstrated that the presence of the casing can be beneficial because the casing acts as an “extended electrode” and can channel current to depth. However, there are still many open questions, particularly with respect to the role of magnetic permeability, about the finer details of the EM responses. To make progress on these questions we need to be able to simulate the responses from realistic wells. The geometry of steel-cased wells is a complicating factor as the wells are often a centimeter thick and may extend for several kilometers. We have developed software for simulating Maxwell’s equations using a 3D cylindrical discretization. The software is included within the open-source SimPEG ecosystem, which supports forward simulations and inversions across a range of geophysical methods including magnetics, gravity, direct current resistivity, induced polarization, electromagnetics and fluid flow. In this presentation, we will explore aspects of the physical responses in a time-domain EM experiment with a steel-cased well and I will provide context for how these developments fit into the wider open-source ecosystem of tools for geophysics.

Transcript

1. Investigating the physics of EM over steel cased wells Lindsey J. Heagy UC Berkeley LBL Geophysics Seminar March 27, 2019
2. Motivation Subsurface injections (hydraulic fracturing) 2
3. Motivation 3 Maxwell’s equations (quasi-static) time frequency constitutive relations Subsurface injections (hydraulic fracturing)
4. Casing: overview • Numerical modelling • DC Resistivity • Electromagnetics ◦ Conductivity ◦ Permeability 4
5. Modeling electromagnetics on cylindrical meshes • Finite volume discretization ◦ cylindrically symmetric ◦ 3D cylindrical meshes • DC, FDEM, TDEM • Open source • Implemented in SimPEG 5 Heagy & Oldenburg, 2018
6. Modeling electromagnetics on cylindrical meshes 6 Kaufman, 1990 Validating the physics • Kaufman (1990): Charges, currents, electric fields • Augustin (1989): magnetic permeability
7. DC Resistivity: fundamentals • Kaufman (1990), Kaufman & Whitman (1993), Schenkel & Morrison (1994) • Cross-sectional conductance (S · m) • Implications ◦ Survey design ◦ Approximating wells (to reduce computation) ◦ Sensitivity in an inversion Modeling electromagnetics on cylindrical meshes 7 currents charges short well long well
8. DC resistivity with steel-cased wells • Casing integrity ◦ Fundamental physics ◦ Feasibility ▪ Full break, partial flaw ◦ Factors influencing detectability ▪ Conductivity (casing, background) ▪ Depth of flaw 8 (Heagy & Oldenburg, 2019)
9. DC resistivity with steel-cased wells: casing integrity 9
10. DC resistivity with steel-cased wells: data at surface Intact well 10 primary Secondary (flawed - intact)
11. DC resistivity with steel-cased wells • Casing integrity • Survey design considerations • Detecting a target ◦ Conductor, resistor ◦ Electrical connection vs. not 11 Signal due to target resistor conductor
12. Direct current resistivity with steel-cased wells: impact points Modelling 12 Survey Design
13. Electromagnetics with steel cased wells: conductive 13
14. Electromagnetics with steel cased wells: permeable 14 t = 5ms t = 10ms t = 1ms conductive permeable
15. Electromagnetics with steel cased wells: permeable 15 conductive permeable
16. Electromagnetics with steel cased wells: permeable 16
17. Electromagnetics with steel cased wells: permeable 17 conductive permeable difference
18. EM with steel cased wells: impact points, conductivity 18 Survey design: currents = galvanic + image + channeled
19. EM with steel cased wells: impact points, permeability 19 t = 10ms permeable Forward simulation: • interplay between conductivity and permeability • poloidal currents in casing Survey design: • Important to include permeability • longer response through time conductive
20. Open questions 20 • Horizontal and deviated wells • Infrastructure • Scalability of numerical simulations • Inverse problem • ...
21. Tools for interactive geophysics 21
22. 22 Oldenburg, BIRS 2016
23. • Modular: building blocks ◦ organized in a framework ◦ pieces available to manipulation • Declarative: express intent ◦ write what you mean ◦ looks like the math • Extensible: new research ◦ quantitative communication ◦ built in feedback loops • Open: for the future ◦ reproducible ◦ opportunities for collaboration building for researchers
24. building for researchers • Modular: building blocks ◦ organized in a framework ◦ pieces available to manipulation • Declarative: express intent ◦ write what you mean ◦ looks like the math • Extensible: new research ◦ quantitative communication ◦ built in feedback loops • Open: for the future ◦ reproducible ◦ opportunities for collaboration
25. (Cockett et al., 2015)
26. Implemented in Python! framework
27. forward simulation Survey: Data collection and geometry Problem: Physics
28. finite volume
29. create a mesh
30. discretize & solve continuous discrete
31. Data Misfit Regularization Inverse Problem inversion elements
32. inversion as optimization
33. Modular, multi-physics 33 • Magnetics • Gravity • DC Resistivity • Induced Polarization (IP) • Electromagnetics ◦ Time ◦ Frequency ◦ With IP effects • Fluid Flow
34. part of an open-source ecosystem 34
35. http://em.geosci.xyz/apps.html
36. GeoSci.xyz 36 https://geosci.xyz EOSC 350 26 locations worldwide
37. thank you! @lheagy lheagy@berkeley.edu slack.simpeg.xyz
38. (the end) 38