176 14 The Stability Graph method 14.1 Introduction Potvin (1988), Potvin and Milne (1992) and Nickson (1992), fol- lowing earlier work by Mathews et al. (1981), developed the Stabili- ty Graph Method for cablebolt design. The current version of the method, based on the analysis of more than 350 case histories cal- lected from Canadian underground mines, accounts for the key fac- tors influencing open stope design. Information about the rock mass strength and structure, the stresses around the opening and the size, shape and orientation of the opening is used to determine whether the stope will be stable without support, stable with support, or unstable even if supported, The method also suggests ranges of cablebolt den- sity when the design is in the realm of ‘stable with support’. 14.2 The Stability Graph method The design procedure is based upon the calculation of two factors, N', the modified stability number which represents the ability of the ~ rock mass to stand up under a given stress condition, and S, the shape factor or hydraulic radius which accounts for the stope size and shape. 14.2.1 The stability number, N' The stability number, N’’, is defined as N'=QAxBxC (4. where Q' is the modified @ Tunnelling Quality Index A is the rock stress factor B is the joint orientation adjustment factor C is the gravity adjustment factor ‘The modified Tunnelling Quality Index, Q, is calculated from the results of structural mapping of the rock mass in exactly the same way as the standard NG? rock mass classification, except that the stress reduction factor SRF is set to 1.00, The system has not been applied in conditions with significant groundwater, so the joint water reduction factor J, is commonly 1.0. This process is identical to that used earlier in this book for estimating the strength of jointed rock masses (see Equation 8.18 on page 97). The rock stress factor, A, reflects the stresses acting on the free surfaces of open stopes at depth. This factor is determined from the unconfined compressive strength of the intact rock and the stress acting parallel to the exposed face of the stope under consideration. The intact rock strength can be determined from laboratory testing of the rock or from estimates such as those discussed in Chapter 8. The induced compressive stress is found from numerical modelling otThe Stability Graph method estimated from published stress distributions such as those in Hoek and Brown (1980a), using measured or assumed in situ stress values, The rock stress factor, A, is then determined from 6,/c,, the ratio of the intact rock strength to the induced compressive stress on the opening boundary: for aia, <2:A= 01 for2 < a/6, < 10: A = 0.1125 (ade) - 0.125 (14.2) and for o.f6, > 10:A = 1.0 A plot of the rock stress factor A, for different values 0/6; is given in Figure 14.1. The joint orientation adjustment factor, B, accounts for the influ- ence of joints on the stability of the stope faces. Most cases of struc- turally controlled failure occur along critical joints which form a shallow angle with the free surface. The shallower the angle between the discontinuity and the surface, the easier it is for the bridge of in- tact rock, shown in Figure 14.2, to be broken by blasting, stress or by another joint set. When the angle @ approaches zero, a slight strength increase occurs since the jointed rock blocks act as a beam. The in- fluence of the critical joint on the stability of the excavation surface is highest when the strike is parallel to the free surface, and smallest when the planes are at right angles to one another, The factor, B, which depends on the difference between the orientation of the criti- cal joint and each face of the stope, can be found from the chart re- produced in Figure 14,3. The final factor, C, is an adjustment for the effects of gravity. Failure can occur from the roof by gravity induced falls or, from the stope walls, by slabbing or sliding. 0.8 06: Rock stress factor A 02 o $ 10 15 Ratio of uniaxial strength to induced stress 2 Figure 14.1; Rock stress factor A for different values of 6/0, i77178 Support of underground excavations in hard rock e Rock bridge Opea. Figure 14.2: Critical joint orientation with respect to the opening surface (Afier Potvin, 1988). Joint orientation adjustment factor 0 19 2 30 40 50 60 70 8 8690 Relative difference in dip between the critical joint and the slope surface Figure 14.3: Adjustment factor, B, accounting for the orientation of the joint with respect to the stope surface (After Potvin, 1988).The Stability Graph method 8 Falling 7 Stope surface a ‘Stabbing’ Gravity adjustment factor C (gravity falls and slabbing) @ 10 20 30 40 $0 60 7 8 Inclination of stope surface a Figure 14.4: Gravity adjustment factor C for gravity falls and stabbing. After Pocvin (1988). was Gravity adjustmate factor C (Sliding) Rowe Oo 10 20 30 40 50 60 70 80 90 Inclindtion of critical joint 8 Figure 14,5: Gravity adjustment factor C for sliding failure modes. After Potvin (6988), Potvin (1988) suggested that both gravity induced failure and stabbing failure depend on the inclination of the stope surface a. The factor C for these cases can be calculated from the relationship, C=8-6cosc,, or determined from the chart plotied in Figure 14.4. This factor has a maximum value of 8 for vertical walls and a mini- mum value of 2 for horizontal stope backs. Sliding failure will depend on the inclination B of the critical joint, and the adjustment factor C is given in Figure 14.5. 179180 Support of underground excavations in hard rock Stability number N” 1000 300 — 200 100 50 Figure 14.6: Stability graph showing zones of stable ground, caving ground and ground requiring support. After Potvin (1988), modified by Nickson (1992). 14.2.2 The shapé factor, § The hydraulic radius, or shape factor, for the stope surface under consideration, is calculated as follows: _ Cross sectional area of surface analysed Ss 14.3) Perimeter of surface analysed 14.2.3 The stability graph Using the values of MV, the stability number, and S, the hydraulic radius, the stability of the stope can be estimated from Figure 14.6. This figure represents the performance of open stopes observed in many Canadian mines, as tabulated and analysed by Potvin (1988) and updated by Nickson (1992),The Stability Graph method 14.3 Cablebolt design Where the stability analysis indicates that the stope requires support, the chart given in Figure 14.7 can be used as a preliminary guide for the cablebolt density. In this chart, the cablebolt density is related to the frequency of jointing through the block size (parameters ROD/J,) and the hydraulic radius of the opening; both must be considered to get an idea of the relative size of the blocks. Of the three design envelopes shown in this figure, the one used should be based both on the use of the opening and on experience with cablebolt support at the site. At the start of a project, the designer should consider using the more conservative envelopes. Potvin et al. (1989) noted that there is a great deal of scatter in the data used in deriving Figure 14.7, reflecting the trial and error nature of current cablebolt design. They also stated that cable bolts are not likely to be effective when the relative block size factor, (RODI,) Hydraulic radius, is less than 0.75 and when the cable bolt density is less than 1 bolt per 10 square metres at the opening boundary. The length of the cablebolts must be such that they are anchored far enough into undisturbed ground for the anchor to be effective. Potvin et al. suggested that 2 rough guideline for design is that the length of the cablebolt should be approximately equal to the span of the opening, ‘They found that cablebolts are generatly not successful in stabilising very large stopes. Cablebolt density (cablebolts/an?) Inadequate cablebott density RODIN Hydraulic radius S (m) Figure 14.7: Cablebolt density design chart, After Potvin and Milne (1992). 181182 Support of underground excavations in hard rock Potvin et al. (1989) suggested that the design of the cablebolts must include consideration of the potential failure mechanism. Where the failuze is predicted to be by sliding, the cablebolt should be inclined at 17° to 27° to the plane on which sliding is likely to occur. The most favourable orientation of cable bolts supporting a slabbing failure is perpendicular to the foliation. 14.4 Discussion of the method Potvin and Milne (1992) warn that the use of the design charts must be limited to the conditions similar to those encountered in the mines used as case histories in the development of the empirical data base. Anomalous geological conditions such as faults, shear zones, dykes or waste inclusions, the creation of a slot or brow within the stope and poor cablebolt installation can all! lead to inaccurate results. In addition the cablebolts must cover the excavation surface fully, since the support design is based upon the assumption that the cables form a continuous zone of reinforced rock surrounding the opening. Practical observations suggest that the main area of uncertainty in using the method lies in the density of jointing in the rock mass. ‘Where the number of joints and other discontinuities per unit volume of rock is highly variable, the value of Q' will be open to question. ‘Under these conditions, the design derived from the stability graph method should be regarded as a first step in the design process'and local adjustments to the design will have to be made, depending upon the conditions observed in the stope, The quality of the cable bolt installation is another variable which has to be recognised when using this method. Where uncertainty on the effectiveness of grouting exists, a conservative approach has to be adopted. In addition, the use of modifying elements such as plates or birdeaged cable bolts has not been included in the design method, perhaps because these items were not used a great deal at the time of e development of the charts, With time and increasing experience, it is likely that these shortcomings will be addressed in this empirical design method, 14.5 Worked stability graph example A 15 m thick (hangingwall to footwall) orebody is located at a depth of 500 m below surface and is to be mined by open stoping methods. Access is from the hangingwall and the option exists to fan cablebolt support into the hangingwall from cable stub drifts. Details of the structural geology of the rock mass and the Q' classification are given in the following sections. Stope design, using the stability graph method, is an iterative process. To start with, reasonable stope dimensions depending upon drilling access, practical mining considerations, and economics, should be proposed, as illustrated in the margin sketch. In this example, the full width of the orebody (15 metres) will be mined in a single stope, and the drilling sublevels are planned for every 25 metre interval of depth, with over and undercuts every 100 metres. The stability graph design procedure is then carried out for these di-The Stability Graph method 183 mensions. This analysis indicates the stability of the proposed stope, and if the dimensions have to be altered, further analyses should be carried on the new dimensions. The procedure is iterated until a satis- factorily stable desiga is achieved. In this case, the geometry of the orebody suggests that the support should take the form of rings of cable bolts, installed from the access drifts, These rings should extend up from cach drift to support the back of each stope, and out from the end of the drift into the hanging wall rock mass, thus creating a stable “butiress’ at each drift. Also, if the design indicates the need for additional support, cables should be installed from hangingwall access drifts to provide adequate cover- age. 145.1 Structural geology The orebody strikes east-west and dips at 65° to the north. Extensive borehole core logging and underground mapping have been carried cout and a total of 1250 features have been recorded. Analysis of this structural geology information by means of the program DIPS indi- cates thal the rock mass contains 5 joint sets which are described in Table 14.1. 145.2 Q" classification The data collected from the geological mapping is used to calculate the modified Tunnelling Quality index, Q' as defined by Equation 8.18 on page 97. Hence Q'= ROD IJ, x I, 4 J, The average ROD value for the rock mass was found to be 78, with a range from 70 to 86. Based upon an inspection of the rock mass in the shaft and devel- opment excavations, it was decided that not all five joint sets oc- curred at all locations and that a reasonable description of the jointing is ‘three sets plus one random set’. Table 4.6.2G, gives the ‘value of the joint set number for this description as J,= 12. Table 14.1; Structural geology for the example mine. Set_Dip® _Dipdirection? Description “ Set_Bip? _Dip direction” _Desenprin A 64410 009420 Planar, smooth to medium roughness. Mica or calcite infilling with some gouge zones and seric- ite. Spacing 10-30 cm, Joints parallel the orebody hhangingwall B 84kT 098424 Slightly rough to ough with no infilling. Spacing. 35-45 em. Joints are perpeudicular to the orebody. Cc 1589 180840 Poorly developed but continuous over several me- tres in places. Rough with calcite or gouge infill- ing. Widely spaced. D 479 09529 ‘Suiking parallel ta B joint, but with shallower dip, Planar, smooth to medium rough, with no infilling. Spacing 50 em, EB 4588271413 Not evident very often; classed as random, Smooth, to medium rough with title or no infiting. ism Overcut Sublevel 1 Sublevel 2 Sublevel 3 Undereut Cablebolts installed from access drifts, 15 m ‘Overcut Sublevel 1 Sublevel 2 ‘Sublevel 3 ‘Undereut ‘Additional cablebolts installed from hanging- wall access drifts if required184 Support of underground excavations in hard rock The joint roughness number J, was found to vary between 1 (smooth planar in Table 4.6.3F) and 2 (smooth undulating in Table 4,6.3C). Similarly, the joint alteration number J, was found to vary between | (unaltered joint walls, surface staining only in Table 4.6, 4B) and 2 (slightly altered joint walls with discontinuous coatings of mica or sericite in Table 4.6.4C). The values chosen for inclusion in the evaluation of Q' were dependent upon the location of the stope being designed and the joint set or sets considered to be most im- portant at that location, Values of Q', together with laboratory evaluations of the intact rock strength 6,, Young's modulus E, and Poisson's ratio v, are: Location Q ©, E v (MPa) (GPa) Hangingwall 24 ® 40 28 ‘Ore zone (Stope back) 63 too 53 10 Footwall Sa 195, 55 2i 145.3. Preliminary stope design The preliminary stope design will be based upon stope dimensions of a stope back span of 15 m and a 25 m stope height. The assessment of the stability and the third stope dimension (strike width in this case) then depends on the estimates of the factors A, B and C, in- cluded in Equation 14.1. Factor A, the influence of the mining induced stresses, is found from Equation 14.2 from the ratio of the intact rock strength to the induced compressive stress, 6/6). The intact rock strength is dis- cussed above, and the induced compressive stress can be estimated from a consideration of the in situ stresses and the proposed stope geometry, The in situ stresses are listed in the following table, and the orientatious plotted on the lower hemisphere projection shown in the margin sketch. Trend? Plunge? Magnitude Magnitude at 500m depth (MPaftn depth) (MPa) 3 35810 0.0437 29 o, 093 28 0.0299 15 e250 60 0.0214 107 A preliminary estimate of the induced compressive stress on each portion of the stope boundary can be obtained from simple elastic numerical modelling. As discussed above, the stope back and hangingwali dimensions have been established from practical mining considerations. The stability graph can then be used to determine a reasonable value for the stope width. A PHASES analysis of a 15 m span, 25 m high stope gives the contours of the major principal stress reproduced in Figure 14.8. From this plot, the induced compressive stress on the back of the stope is found to be about 30 MPa, and on the hangingwall it is less than 5 MPa. An unconfined compressive strength of o, = 100 MPa is assumed for the ore in the stope back and a. = 70 MPa is assumed for the hangingwall rock. Therefore, the respective ratios of o./0) areThe Stability Graph method Figure 14.8: Contours of maximum principal stress oh induced in the rock surround- ing a 15 m span, 25 m high stope. The in situ stresses acting on the stope are as- sumed to be 22 MPa (inclined at 10°) and 12 MPa as showa in the figure. approximately 3,3 and 14. Using these values, the rock stress factor can be calculated frora Equation 14,2, giving A = 0.25 for the stope back and A = I for the hangingwall ‘The factor B is used to account for the influence of the joint orientation on the stope stability, The most critical joint, influencing the stability of the stope boundary, is generally the one closest to parallel to the boundary. For this example, the critical joint sets for the various components of the stope boundary are listed below, to- gether with the values of B, found from Figure 14.9, Joint Set Difference in strike” Difference in dip” Factor B Stope hangingwall A 0 0 03 Stope ends B 0 0 03 Stope back c 0 Is 02 Factor C accounts for the influence of the stope wall orientation. A comparison of the geometry of the example mine with the sketches shown in Figures 14.4 and 14.5 suggests that the dominant modes of failure will be gravity falls from the stope back and buckling failure from the stope hangingwall and ends, The gravity adjustment factor is obtained from Figure 14.10 which gives C = 2, for the stope back, €=5.5, for the hanging wall, and C'= 8, for the stope ends. 185186 Support of underground excavations in hard rock gor Difference in strike Joint orientation adjustment factor B 0 10 2% 30 40 50 6 7 8 9 Relative difference in dip between the critical joint and the slope surface Figure 14.9: The correction factor B for the example mine, u 6 a a ul 3 © Stope ends 2 @ = Stope hangingwal! WS Stope back o@ 10 2 30 40 50 6 7 80 90 Inclination of stope surface a Figure £4.10; Gravity adjustment factors for example mineThe Stability Graph method 187 The stability number, NV’, for the stope back and hangingwail can now be calculated ftom Equation 14.1 and plotted on the stability graph as shown in Figure 14.11. a oA B c Ny Stope back 63 025 02 20 063 Stope hangingwall 24 1035S AO The stability graph gives the hydraulic radii of the stope that will be stable with and without support. The values of the hydraulic radii and associated stope widths are as follows: Baek Hagin Known dimension LS m spon 27.6 m height Hydraulic radius Stable <3 <45 Unsupported tran: 3045 451065 Stable with support 451075 6S 1010 Supported transition 7.5109 10 to 12 Catewiated stepe width Stable <0 «134 Unsupported transition 1010 22.5 13.4 1024.6 Stable with support 22.50 00 24.6 1072.6 Supported tran 72.6 to 184 100 ———_———— 500 Stability number N” 8 Ot Hydraulic radius $ ~ metres Figure 14.11: Stability of stope back and hangingwall for example mine,188 Support of underground excavations in hard rack ‘The analysis indicates that the back is generally more critical than the hangingwall unless both are supported. For a 15 m span stope with a vertical height of 25 m, the width of the stope along strike should be fess than 10 m for the stope to be stable without support. This strike distance is too short to allow for economical and safe development of drawpoints. Therefore, both the back and the hang- ingwall of the stope will have to be supported. ‘The maximum safe strike length of a supported stope is controlled by the stability of the hanging wall and is about 75 m. The decision on a reasonable strike length should be made on the basis of consideration of mining practicalities (overall orebody length, stope sequencing, drawpoint design etc.). If, for example, a reasonable strike length of the stope is determined to be 60 metres, a check calculation, using the same procedure as described above, will show that this stope is stable with support, The preliminary design for the non-entry stope support can be carried out using Figure 14.7. The input data required for this ana- lysis are given in the following table, and are plotted in Figure 14.12. 8 RQD/S Bolt density Bolt spacing {boltsAm® {m) Back 6 1d 19-33, 17023 Hangingwall = 9.45. 0.69 B. 17021 Ends 486 134 16-3 18025 Cablebolt density (cablebolts/m") RQDIn Hydraulic radius S (mm) Figure 14.12: Cablebolt density for the pretiminary design of the open stope at the example imine,The Stability Graph method This analysis indicates that the cablebolts should be placed on regular patterns of about 2m x 2m spacing. In order to provide this denity of support, cablebolts will have to be installed from both the sublevels and the hangingwail cablebolt stub drifts, as indicated in the lower margin sketch on page 183. Note that the hangingwall data plots close to the ‘support ineffec- tive’ zone in Figure 14.12. As the design progresses, it would be ad- visable to find means of reducing the hydraulic radius of the stope, eg., by early backfilling. If this can be achieved reliably during regu- lar mining cycles, the cableboit spacing may be increased to at least 2.5 m x 2.5 m, saving about one third of the cablebolts required without such a reduction. Following the preliminary analysis, the design of the stope di- mensions and support spacing should be refined as more information about the rock mass characteristics and operational constraints be- come available, and as the economics of mining the ore and the cost of the support are evaluated. 189