X Memo Comments on SSM s review of SKB s earthquake modelling Valda bilder, kapitelvis presenterade, som underlag för diskussion.
1 Introduktion
1 Respektavstånd kring mindre zoner En primärrörelse av 5 mm fordar ett skalv skalv av (minst) magnitud 5. Ingen sekundärrörelse i någon av de modeller som hittills analyserats har varit större än 1% av primärrörelsen, inte ens för fall där målsprickan har skurit genom primärzonen. I själva verket behövs därför ett skalv av minst magnitud 5.5, vilket svarar mot en rupture area av ca 2 km 2. Här finns dessutom marginaler: Lower bound regressionen Rupture area fault area
2 3DEC modelling approach
2 Perspectives on Conservativeness in 3DEC What if Approach Approximate impact of more realistic input assumptions: Examples X = largest target fracture slip (e.g. mm of slip per mm of target fracture diameter) Maximized Moment Magnitude Maximized Fault Slip Velocity Abrupt Rupture Arrest at Sharp Fault Edges Perfectly Planar Target Fractures Target Fractures in Elastic Continuum X Maximized Moment Magnitude Maximized Fault Slip Velocity Abrupt Rupture Arrest at Sharp Fault Edges Slightly Undulated Target Fractures Target Fractures in Elastic Continuum.5X (1) Moment Magnitude on Leonard Regression Maximized Fault Slip Velocity Abrupt Rupture Arrest at Sharp Fault Edges Perfectly Planar Target Fractures.5X (2) Target Fractures in Elastic Continuum Maximized Moment Magnitude Maximized Fault Slip Velocity Abrupt Rupture Arrest at Sharp Fault Edges.8X Perfectly Planar Target Fractures Target Fractures in DFN surrounding (3) (1) Lönnqvist M, Hökmark H, 215. Assessment of method to model slip of isolated, non planar fractures using 3DEC. ISRM Congress 215 Proceedings Int l Symposium on Rock Mechanics ISBN: 978 1 926872 25 4. (2) Fälth B, Hökmark H, 215. Effects of Hypothetical Large Earthquakes on Repository Host Rock Fractures. Posiva Working Report 215 18, Posiva Oy (3) Test model, not published
2 Planar fractures vs undulated fractures
2 Magnitude overestimates 8 8 7,5 7 Olkiluoto, extended BFZ214 Cases 1, 2, 6 Olkiluoto, extended BFZ214 Case 4 Olkiluoto, extended BFZ214 Case 5 Olkiluoto, extended BFZ214 Case 3 7,5 7 Olkiluoto, extended BFZ214 Cases 1, 2, 6 Olkiluoto, extended BFZ214 Case 4 Olkiluoto, extended BFZ214 Case 5 Olkiluoto, extended BFZ214 Case 3 6,5 6,5 Moment Magnitude 6 5,5 5 4,5 Example model, Fälth et al., 215 Olkiluoto, BFZ21, WR 212 8 Forsmark, ZFMA2, Fälth et al., 215 Forsmark, ZFMA2, Fälth and Hökmark, 213 (thermal) Olkiluoto, BFZ1, WR 212 8 WC Earthquake Catalogue, non SCR WC Earthquake Catalogue, SCR Synthetic 3DEC Earthquakes Leonard regression WC regression Moment Magnitude 6 5,5 5 4,5 Example model, Fälth et al., 215 Olkiluoto, BFZ21, WR 212 8 Forsmark, ZFMA2, Fälth et al., 215 Forsmark, ZFMA2, Fälth and Hökmark, 213 (thermal) Olkiluoto, BFZ1, WR 212 8 WC Earthquake Catalogue, non SCR WC Earthquake Catalogue, SCR Synthetic 3DEC Earthquakes 4,1 1 1 1 1 1 Rupture Area (km2) 4,1 1 1 1 1 1 Fault Area estimate (km2)
2 Magnitude overestimates Hanging wall Cumulative distribution (%) Cumulative distribution (%) 1 8 6 2 m 4 m 4 6 m 2 8 m 1 m 1 2 3 4 5 6 Induced displacement (mm) 1 8 6 2 m 4 m 4 6 m 2 8 m 1 m 1 2 3 4 5 6 Induced displacement (mm) Footwall Cumulative distribution (%) Cumulative distribution (%) 1 8 Moment Magnitude 7,3 7,1 6,9 6,7 6,5 6,3 6,1 5,9 5,7 5,5 5,3 Forsmark, ZFMA2, base case Base case (re) Olkiluoto, BFZ214, WR 212 8, case 2 Olkiluoto, BFZ214, WR 212 8, case 5 Case 5 (re) Case 2 (re) WC Earthquake Catalogue, non SCR WC Earthquake Catalogue, SCR Synthetic 3DEC Earthquakes Leonard regression WC regression Synthetic 3DEC Earthquakes (re) 1 1 1 Rupture Area (km2) 6 2 m 4 m 4 6 m 2 8 m 1 m 1 2 3 4 5 6 Induced displacement (mm) 1 8 6 2 m 4 m 4 6 m 2 8 m 1 m 1 2 3 4 5 6 Induced displacement (mm) Forsmark ZFMA2, endglacial Case 2 Uniform fault strength Case 5 Abrupt arrest Fault completely locked in high strength stability strip Cumulative distribution (%) Cumulative distribution (%) Fault trace 1 9 8 7 6 5 4 3 2 1 1 9 8 7 6 5 4 3 2 1 Olkiluoto BFZ214, Present day Dotted lines Full lines Region #1 Region #2 3 m 7 m 15 m 25 m 45 m dd/dip 9/8 dd/dip 18/8 dd/dip 195/77 dd/dip 15/3 dd/dip 3/3 dd/dip 9/3 1 2 3 4 Induced displacement (mm) 3 m 7 m 15 m 25 m 45 m dd/dip 9/8 dd/dip 18/8 dd/dip 195/77 dd/dip 15/3 dd/dip 3/3 dd/dip 9/3 2 4 6 8 1 Induced displacement (mm) Cumulative distribution (%) Cumulative distribution (%) 1 9 8 7 6 5 4 3 2 1 1 9 8 7 6 5 4 3 2 1 3 m 7 m 15 m 25 m 45 m dd/dip 9/8 dd/dip 18/8 dd/dip 195/77 dd/dip 15/3 dd/dip 3/3 dd/dip 9/3 1 2 3 4 Induced displacement (mm) 3 m 7 m 15 m 25 m 45 m dd/dip 9/8 dd/dip 18/8 dd/dip 195/77 dd/dip 15/3 dd/dip 3/3 dd/dip 9/3 1 2 3 4 Induced displacement (mm)
2 Target fractures in DFN surrounding 3 25 DFN fractures (282) Deterministic (dd15/dip3, along with DFN) Deterministic (dd15/dip3, base case) Shear displacement (mm) 2 15 1 5 2 4 6 8 Fracture radius (m)
2 Sharp fault edges (or edges of fault inhomogeneities)
3 Comments on Technical Note 214:59
3 2D 3D, coalescence Style and extent of coalescence?? Style and extent of coalescence?? Style and extent of coalescence?? Rock bridge between two neighbouring 3D fractures 2D representation of rock bridge 3D equivalent of 2D representation of rock bridge
3 2D 3D, thermal load Depth (m) 1 2 3 4 5 Major horizontal principal stress in PFC2D model after 25 years of heating Minor horizontal principal stress in PFC2D model after 25 years of heating Initial 5 yrs 5 yrs 1 yrs 5 yrs 1 yrs 5 yrs 1 yrs C B A Scanline B Repository depth 6 Horizontal stress in major horizontal in situ stress direction 7-55 -5-45 -4-35 -3-25 -2-15 -1-5 Stress (MPa) Horizontal stress in minor horizontal in situ stress direction Vertical stress
3 2D 3D, thermal load Shear stress (MPa) 2 15 1 5 c =.5 MPa σ v 2θ = 9 ϕ σ h ϕ = 35.8 5 1 15 2 25 3 35 4 45 5 σ H Effective normal stress (MPa) σh σh σh σv σh σv Blue Mohr circle shows stability of the vertical fractures considered in the horizontal 2D sections analyzed in the thermomechanical PFC models. Stability margins are 1 MPa and more (if pore pressure is accounted for as in this figure. If not, stability margins are even larger). Within the repository footprint, both horizontal stresses increase during heating, pushing the blue Mohr circle further to the right.
3 Impact and handling of outliers Mean displacements in most distance classes are non trusted outliers, i.e., : > 5 % of the smooth joint slip caused by numerical errors.
3 Impact and handling of outliers 5 mm slip on 195 m long fracture counts as trusted result. Same slip on 25 m long fracture counts as outlier, caused by numerical errors.
3 Unrealistic results 8 7,5 7 6,5 Kehetuohai E, China 1931, 14.6 m WC Earthquake Catalogue 2D representation of Singö fault zone produces 3 times larger fault slip than maximum slip documented in WC crustal earthquake database. Moment Magnitude 6 5,5 5 4,5 Forsmark, present day, SSM 214:59 Singö fault zone, DFN3h, 35.9 m Forsmark, forebulge SSM 214:59 ZFMWNW1, DFN6h, 48.9 m To match these slip results, one needs to compare with the largest subduction zone earthquakes ever instrumentally recorded, such as the Tohoku 211 earthquake with a 4 km by 2 km rupture area. 4 1 2 3 4 5 6 Max Displacement (m)
3 Unrealistic results Maximum displacement along large Class 3 and Class 4 fractures systematically significantly smaller than largest displacement along smaller Class 2 fractures subjected to same load. No explanation given.
3 Choice and outcome of modelling cases 11 modelling cases out of 38 regard the time of the stable glacial maximum. 8 cases regard the time of potential instability during ice retreat Stabilized ZFMA2 produces larger earthquake at time of glacial max than destabilized ZFMA2 at time of ice retreat. No explanations given.
4 As demonstrated and discussed in the previous chapter, there is a range of issues regarding the approaches and the results in Yoon et al. (214) that, according to our view, render the suggestions and recommendation based directly on results presented in that report highly questionable. Some of the recommendations are judged to be relevant, regardless of the PFC2D results. Analysing models with interacting neighbouring deformation zones, for instance, is a natural extension of the modelling efforts. It may also be necessary to ensure, specifically, that deformation zones with trace lengths smaller than 3 km will not require any respect distances. These issues are currently being addressed using the 3DEC modelling approach. We should also point out that we have no objections regarding the Synthetic Rock Mass approach in general or the PFC code in particular. However, to arrive at credible quantitative stress and slip estimates, it is essential that modelling of systems with 3 dimensional fracture networks, complicated stress gradients and seismic sources are modelled in three dimensions. It is also necessary to make systematic reality checks and to ensure that couplings between input assumptions and modelling results are logical and understandable.
Att tänka på för oss alla; kanske fortfarande håller? (Starfield and Cundall, 1988, Towards a Methodology for Rock Mechanics Modelling, Int J. Rock Mech. Min. Sci.& Geomech.Abstr. Vol 25, No3, pp 99 16 A model is a simplification of reality rather than an imitation of reality. It is an intellectual tool that has to be designed and chosen for a specific task. The design of the model should be driven by the questions that the model is supposed to answer rather than the details of the system that is being modelled. This helps to simplify and control the model.