Engineering Geology 121 (2011) 18–27

Contents lists available at ScienceDirect

Engineering Geology

j ourna l homepage: www.e lsev ie r.com/ locate /enggeo

Hangingwall surface subsidence at the Kiirunavaara Mine, Sweden

Tomás Villegas a,b,⁎, Erling Nordlund a,1, Christina Dahnér-Lindqvist c

a Division of Mining and Geotechnical Engineering, Luleå University of Technology, 971 87 Campus University, Luleå, Swedenb Department of Civil and Mining Engineering, University of Sonora, 83000, Rosales y Navarrete s/n, Hermosillo, Sonora, Mexicoc Luossavaara–Kiirunavaara AB (LKAB), LKAB, Box 952, 971 28 Luleå, Sweden

⁎ Corresponding author at: Division of Mining and GUniversity of Technology, 971 87 Campus University,493212, +52 662 2592183.

E-mail addresses: tomvil@ltu.se (T. Villegas), erling.nchristina.dahner.lindqvist@lkab.com (C. Dahnér-Lindqv

1 Tel.: +46 920 491335.

0013-7952/$ – see front matter © 2011 Elsevier B.V. Adoi:10.1016/j.enggeo.2011.04.010

a b s t r a c t

a r t i c l e i n f o

Article history:Received 16 September 2010Received in revised form 8 April 2011Accepted 15 April 2011Available online 22 April 2011

Keywords:Kiirunavaara mineSublevel cavingHangingwallSubsidence

Large scale surface subsidence has been experienced at the Kiirunavaara mine since sublevel caving wasimplemented as a mining method. Surface disturbances are affecting part of the city of Kiruna, the railway,and the power station. Continuous and discontinuous subsidences characterize the hangingwall deformation,which is periodically monitored using surveying techniques and mapping of surface cracks. A historic reviewof subsidence prognoses has been carried out and the results compared with the actual condition of thehangingwall. The review showed discrepancies between different prognoses. In addition, limit equilibriumanalyses indicated that break angles flatten while the mining depth increases. However, this tendency is notclear in the field where break angles show a large dispersion of values. By using surveying data, two differentanalyses have been performed. The time-dependent movements of the hangingwall have been describedusing time–displacement curves and strain analysis has been performed for different sections of thehangingwall. Three different stages of the time–displacement behavior have been identified and described. Ithas been concluded that extension strain can reach values which may damage civil structures before surfacecrack can be observed.

eotechnical Engineering, LuleåLuleå, Sweden. Tel.: +46 920

ordlund@ltu.se (E. Nordlund),ist).

ll rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

The Kiirunavaara iron ore mine is located to the west of Kiruna cityin northern Sweden about 180 km north of the Arctic Circle. The mineis owned and operated by Loussavaara–Kiirunavaara AB, LKAB. At thebeginning of the last century, the ore was extracted by open pitmining. The transition to underground mining was carried out duringthe1950's. Today, the entire orebody is mined using sublevel caving.

As shown in Fig. 1, the orebody is developed in sublevels at aregular vertical spacing of 28.5 m, with a footwall drift (parallel to theorebody) and cross-cuts (perpendicular to the orebody). Fans ofproduction holes are drilled in the ore from the cross-cuts with aburden of 3.0–3.5 per ring. The broken ore, mobilized by the effect ofgravity flow, is drawn at the brow of the cross-cut. Load–haul–dumpmachines carry the run-of-mine ore to the nearest ore pass. Using oneof the shuttle trains operating on the 1045 m level the ore istransported to one of four crushing stations. Then the ore is skiphoisted in two stages to the 775 m level and then to surface.

The orebody is divided into ten production blocks. All the accessand mine infrastructure are located in the footwall (see Figure 2)

because the hangingwall fails, caves, and subsides by the effect of thegravity and induced stresses when the broken ore is drawn.

Since sublevel caving was implemented in the mine, the hanging-wall has experienced large scale subsidence. In general two types ofdeformation zones, shown in Fig. 3, can be observed on the groundsurface of the hangingwall. They are discontinuous and continuousdeformation zones, described as follows:

• Discontinuous deformation zone: this zone is characterized by largesurface disturbances affecting limited regions. Features such astension cracks, steps and chimney caves normally appear in thiszone. Tension cracks are formed by extension strain with a tendencyto develop along subvertical discontinuities striking parallel to theorebody (which is the most predominant joint set). Steps, whichstarted as tension cracks, define the extension of failed wedges orblocks. Chimneys form in the failed rock mass under stressrelaxation where individual rocks move vertically through flowchannels after ore draw.

• Continuous deformation zone: in this area, there is a smoothlowering of the ground surface detected only by periodic surveying.This zone, which extends to around 150–200 m beyond the limit ofthe discontinuous deformation zone (Lupo, 1996), shows onlyelastic deformation or continuous non-elastic strain (Singh et al.,1993).

The block of rock formed between the outermost fracture and thecrater limit can be considered to consist of a “semi-intact” rock mass

Fig. 1. Sublevel caving mining.After Atlas Copco Rock Drills AB, 2000.

Fig. 2. Mining blocks in the Kiirunavaara mine (LKAB, 2001).

19T. Villegas et al. / Engineering Geology 121 (2011) 18–27

Hangingwall

Ore

Cavedrock

Crater

Mining level

Tension Crack

Continuous deformation zone Discontinuous deformation zone

Footwall

Step Chimney

Break angle

Limit angle

Fig. 3. Surface subsidence zones of the hanging wall at Kiirunavaara mine.

20 T. Villegas et al. / Engineering Geology 121 (2011) 18–27

which is thereafter progressively fragmented by the effect of caving(Herdocia, 1991).

Today, as mining advances downwards depth, surface andsubsurface subsidence extend in the hangingwall towards the city ofKiruna, the railway, and the power station. Several studies during the

Fig. 4. Geological map of tModified from Parák, 1973

past years have developed predictions of the subsidence area. Thepresent paper contains a historical review of previous prognoses ofthe hangingwall failure and the result is compared with the currentcondition at the mine. Thereafter the ground surface movement isanalyzed based on monitoring data.

he central Kiruna area..

Table 1Identified structures (Magnor and Mattsson, 1999).

Strike Dip Observations

WNW–ESE 60°–70° SW The most dominant brittle structuresN–S 60°–70° E Plastic deformation zonesNW–SE 30°–40° NENNE–SSW 70°–90° SE The most dominant plastic structuresNE–SW 18°–36° SEE–W 40°–50° S Brittle faults

21T. Villegas et al. / Engineering Geology 121 (2011) 18–27

2. Geology and structural geology

The Kiirunavaara orebody is tabular, 4 km long and up to 100 mthick. It extends at least down to 1300 m below the ground surface(Bergman et al., 2001). The orebody strikes nearly north–south and isdipping 50° to 60° to the east. The orebody lies between a thicksequence of trachyandesitic lavas (syenite porphyry) on the footwallside and pyroclastic rhyodacites (quartz porphyry) on the hanging-wall side as is shown in Fig. 4. In Fig. 4 some brittle deformation zonesare indicated as dashed lines. Along the contact between the orebodyand the hangingwall, magnetite–actinolite breccia is developed. Veinsof magnetite and actinolite extend tens of meters up in the hanging-wall, and they are locally wider and spaced more closely together,forming rich ore breccia or lenses of massive ore. In addition, someareas are affected by biotite–chlorite alteration (Bergman et al., 2001).

Major geological structures have been located by interpretation ofgeophysical measurements and field mapping. These structures havesimilar orientations as the main directions of small-scale discontinu-ities (Magnor andMattsson, 1999). Two types of structures have beenidentified by Magnor and Mattsson (1999) — plastic deformationzones and brittle deformation zones. The largest systems of structuresare presented in Table 1.

Although previous studies made comments on the effect thatgeological structures have on surface subsidence, its effect was notdocumented. However, the observed surface cracks often follow theorientation of the major structures and joints.

3. Historical review

Even though deformation of the hangingwall was expected byintroducing the sublevel caving method in 1957 (Gustafsson, 1981),survey stations were not installed to monitor surface deformationbefore 1976. Consequently, during the 1960's and part of the 1970'sonly surface cracks were correlated to the working levels in the mineusing the term of break angle. The break angle is defined as the anglebetween the horizontal and an imaginary line from outermost

Fig. 5. Similarities between (a) the physical model, (Stephansson et al., 1978) and (b

observed surface crack to the active mining level (see Figure 3). Aconstant break angle of 60° normally fitted the observed data. Thisangle has been used as a rule of thumb in Swedish sublevel cavingmines to predict the extent of surface disturbances (Lupo, 1996).

A similar result was obtained by Stephansson et al. (1978) after acomprehensive study of the hangingwall in the vicinity of the Zenobiaarea of the orebody (north of themine). The prediction of break anglesvaried between 60 and 50°. This study, which consisted of fracturemapping, finite element modeling and two and three-dimensionalphysical modeling, concluded that the failure mode is a combinationof toppling and shear failure. The two-dimensional physical modelsindicated that during the caving process tensile fractures develop firston the surface and are followed by shear failure along a sliding plane.In the case of three-dimensional physical models, a complex type ofstep failure was observed where fractures propagate along sub-vertical joint sets striking oblique to the axis of the orebody. Then ajoint set parallel to the orebody connects these fractures (Figure 5).This type of failure has been observed and mapped in the field in thenorthern part of the hangingwall.

After 10 years of measuring surface deformation, it was possible tocorrelate surface strain and cracking. It was estimated that surfacecracks appear for a critical extension of 2.3 mm/m at a constant breakangle of 60° (Chizari, 1988). In addition, the field data was fitted to anexponential function developed for trough subsidence to estimate themagnitude and extent of surface subsidence. The results obtained bythis method indicate that the break angle become steeper withincreasing mining depth.

During the 1990's new analytical approaches were used not onlyfor the prediction of the break angle but also to explain the failuremechanism. The hangingwall in the sublevel caving method wasenvisioned byHoek (1974) as a rock slope inwhich the toe is undercut.Due to the undercutting the hangingwall fails in a progressive mannerwhen themining progresses down-dip. Assuming that a plane of shearfailure is formed from the undercut and connected to a tension crack,Hoek (1974) developed a limit equilibrium method for stabilityanalysis adding the effect of the caved material over the walls duringstatic conditions. This method was extended by Lupo (1996) takinginto consideration tractions generated by the caved rock during drawalong the face of the hangingwall and footwall. Herdocia (1991), on theother hand, did not consider tension cracks and the effect of the cavedmaterial in his analysis. Additionally, Lundman and Vollen (1991) andDahnér-Lindqvist (1992) back calculated the rock mass strength byconducting limit equilibrium analyses considering circular failure onthe hangingwall and footwall, respectively. Using this approach, theypredicted new limits for the discontinuous subsidence zone for thenorthern part of the mine.

Fig. 6 shows that there is no agreement among the break-anglepredictions. Although Herdocia's and Lupo's predictions show

) an aerial photograph of the northern portion of the Kiirunavaara mine (2005).

200

300

400

500

600

700

800

900

1000

40 50 60 70 80 90

Break Angle (degrees)

Min

e L

evel

(m

)

Herdocia, 1991Lupo, 1996Chizari, 1988Stephansson et al, 1978

Fig. 6. Break angle prognoses of the hangingwall at the Kiirunavaara mine.

22 T. Villegas et al. / Engineering Geology 121 (2011) 18–27

decreasing angles when the depth of the mining level increases, theformer predicts significantly smaller angles. Lupo's model introduceda tension crack that reduces the area of the plane of failure resulting inhigher values of the back-calculated rock mass shear strength.Furthermore, Lupo's model included the effect of the caving materialthat provides support during stable conditions but increases tractionswhenmoving downwards. Themodels in both studies were calibratedfor different sections of the mine which could indicate that rock massstrength was different. Chizari's prediction shows increasing breakangles with deeper mining levels. However, observed break angles inthe mine, presented in Fig. 7, do not show any tendency to changewith mining level, only a large scatter for mining levels between350 m to 550 m. Herdocia (1991) pointed out that the degree ofscatter in the observed break angles depends, among other factors, onthe scatter in the properties of the rock mass and geologicalstructures. Jointed rock masses show a more uniform caving process,while strong, less jointed rock masses show a larger scatter of thebreak angles and a tendency to accommodate displacement alonggeological structures.

200

300

400

500

600

700

800

900

40 45 50 55 60

Break An

Min

ing

Lev

el (

m)

Fig. 7. Break angles me

In 2007, using the concept of break angle and limit equilibriumtechniques LKAB estimates new limits of deformation zones (Figure 8).By definition the break angle is determined from the outermostsurface crack on the hangingwall. However, numerical analysis hasshown that this fracture is the result of tensile strain (Villegas andNordlund, 2008a) that does not necessarily mean the total failure ofthe hangingwall as is assumed in the limit equilibrium analysisemployed during the 1990's. Therefore, if the wrong crack wasselected during the analysis, the back-calculated rock mass strengthmay have been underestimated and, as a result, lower break anglesobtained in forward predictions. Today, two dimensional numericalmodels are been tested to analyze the hangingwall failure anddevelop new prognoses (Villegas and Nordlund, 2008b). The minecross section Y1500, which is shown in Figs. 9 and 10, has been usedfor this purpose.

4. Monitoring

At the beginning of the underground operation only visualinspection and surface mapping was carried out. A formal surveymonitoring system was initiated in 1975 covering only some sectionsof the hangingwall. Surveying methods such as transit, theodolite andcloth measuring tapes with an estimated resolution of 0.01 m wereused (Lupo, 1996). The low resolution was attributed to therudimentary construction of surveying stations (unmarked barspounded into the soil or waste rock). Currently there are ten lines ofsurvey stations in almost perpendicular directions to the pit as isshown in Fig. 11 with a total of 236 stations. New points have beenadded and some old points, which are too close to the caved area, arenot in use anymore for safety reasons. The new stations, which weresurveyed annually with total station, are marked bars mounted incylindrical concrete bases. The measurements with total station havea precision of about 2 mm under perfect conditions. However, sincesome surveying stations are mounted in soil and waste dumps,movements less than one or 2 cm cannot be detected according toHenry and Dahnér-Lindqvist (2000). From 2003 and on themeasurements are carried out using GPS. The accuracy in thehorizontal direction is 5 mm and 20 mm in the vertical direction(Villegas and Nordlund, 2010).

Interferometric synthetic aperture radar, InSAR, has been tested inthe area with relative success but at the moment it has some

65 70 75 80 85 90

gle (degrees)

Lupo, 1996

Herdocia, 1991

2005

asured in the field.

Fig. 8. Estimated limit of the deformation zone (LKAB, 2007).

23T. Villegas et al. / Engineering Geology 121 (2011) 18–27

limitations related to adverse environmental conditions such as snowand vegetation. With this technique it is possible to measuredeformation in areas where there is no access for safety reasons.

Surveying data are updated every 6 months and surface mappingis carried out annually. The reports consist of updated surface profilesand surface crack description. In addition, the new limits of thesubsidence zones are marked in a map that covers the area betweenthe mine and the city.

Analysis of surveying data of the hangingwall identified two largezones that move at different rates (Mäkitaavola and Taipalensuu,2005). In addition, the InSAR interpretation identified a third blocklocated in the southern part of the hangingwall (Henry et al., 2004)where there were no stations. Thus it was not identified by thesurveying analysis.

Fig. 9. Cross section Y1500 showing the total surface

5. Time-dependent movements of the hangingwall

Calibrating the models for failure analysis with limit equilibriummethods and numerical methods requires a criterion to determine thesurface crack that is part of the plane of failure. The time-dependentbehavior of the hangingwall at the Kiirunavaara mine resembles thebehavior of open-pit slopes classified as transitional type by Broadbentand Zavodni (1982). The hangingwall deformation experiences threedifferent phases of time-dependent deformation behavior— regressive,progressive and steady state; respectively (Figure 12). During theregressive phase the movement starts with dilation and relaxation ofthe rock mass due to changes in stresses after the previous failure.Following the initial response, the time–displacement curve showsshort-term decelerating displacement cycles. Apparently, an increase of

subsidence and the location of tension cracks.

Fig. 10. Fracture mapping.

24 T. Villegas et al. / Engineering Geology 121 (2011) 18–27

mineral extraction or the exploitation of a deeper sublevel initiates acycle and the velocity of movement decay when there is a reductionin the rate of extraction. For instance, Fig. 13 shows the time–displacement curve of the surveying station S1 (located between themine coordinates X6484 and Y2445) representing a regressive phase.The decay of the last cycle is coincident with a reduction of 50% ofmine extraction from block 25 during the years 2004 to 2005. Thenext cycle was initiated with an increase of mine extraction from the

Y0400

Y2800

Y5200

X7200 X8800X5600

Kiruna City

Pit

Railway

N

Kiir

un

avaa

raM

ine

D

T

SNH

ML

FBC

Fig. 11. Hangingwall map showing the mine coordinate system and survey stations.

year 2005 to the year 2006. Thus, there is a direct correlation betweendisplacement cycles and variations in the production rate. Theprogressive phase is characterized by an accelerating displacement.During this phase extension strain is concentrated along tensioncracks. This change of phase in the time–displacement curve isidentified by an inflection point in the curve. The total verticaldisplacement of this point is defined as the critical vertical dis-placement, CVD. Finite element analyses of the mine sections Y1500and Y2300 showed that the failure surface was coincident with thecontour of the CVD (Villegas and Nordlund, 2008a,b). This findingcould indicate that, at this time, the failure surface is well defined.Continuation of mining produces a creeping motion or “steady state”where the hangingwall moves at constant velocity. Themagnitude ofthe motion is influenced by the draw rate. Adding the fact that thehangingwall undergoes large displacements, it can be assumed thatthe surface along which failure takes place reaches the residualstrength state during this stage, which is the case for rock slopes(Savely, 1993). Finally, the steady-state behavior is interrupted bythe proximity of caving, starting the progressive phase where thehangingwall displaces at an accelerating rate to the point of collapsein a chimney.

When the scale of graphs is incremented the cycle effect cannot bedetected but the point of inflection in the curve is clear as shown inFig. 12. This figure shows the time–displacement curve of station T5located between the mine coordinates Y2950 and X6550. The CVD is0.6 m for the year 2003. Fitting straight lines to each phase made itpossible to estimate the rate of movement. The results of this arepresented in Table 2.

Previous subsidence data analyses showed that the first surfacefractures appear at about 15 cm to 20 cm of vertical displacement(Chizari, 1988; Lupo, 1996). If the range of CVD values, based on Table 2,is 0.6–1.0 m, then the first cracks appear during the regressive phase.These cracks were used before 2007 by LKAB and the municipality toindicate the fracture zone which was the limit of the hazardous zonewhere buildings and constructionswere not allowed. Different criterionwas used for the subsidence limit of surface subsidence determined byusing a horizontal strain cut-off of 0.7 mm/m (Lundman and Vollen,1991).

0

1

2

3

4

5

6

1997 1999 2001 2003 2005

Year

To

tal v

erti

cal d

isp

lace

men

t (m

)

CVD

Regressive phase

Steady state

Progressive phase

Fig. 12. Time–displacement curve of the surveying station T5.

0

1

2

1997 1999 2001 2003 2005 2007

Year

To

tal v

erti

cal d

isp

lace

men

t (m

)

Fig. 13. Time–displacement curve of the surveying station S1 showing a regressive phase.

25T. Villegas et al. / Engineering Geology 121 (2011) 18–27

6. Strain analysis

It is a normal practice in caving mines to estimate the limit of thediscontinuous deformation zone on the ground surface (Brown, 2003)because mine and civil infrastructure can be seriously damage in thiszone. Thus, less effort is done to investigate the zone of continuousdeformation. However, strain analyses carried out for the groundabove mines where longwall mining is used and continuousdeformation occurs; demonstrate that buildings and civil infrastruc-

Table 2Estimated rate of movements.

Station CVD Regressive phase Steady stateMovement rate(mm/yr)

Magnitude(m)

Period(year)

Movement rate(mm/yr)

B4 0.67 1997–2005 63 1136F4 1.0 1994–2003 127 2434L6 0.85 1996–2004 223 992M4 – 1996– 177 –

H22 – 1993– 64 –

N29 – 1996– 149 –

S1 – 1997– 178 –

T5 0.8 1995–2003 108 1668D8 0.64 1997–2002 105 608Average 0.8 7.6 133 1368

ture have been damage (Bell and Donnelly, 2006). The components ofthe ground deformation for continuous subsidence have beenclassified by Kratzsch (1983) as vertical components (subsidence,tilt, radius of curvature) and horizontal components (horizontaldisplacement and extension or compression). Analysis of structuraldamage caused by continuous subsidence concluded that uniformdisplacement and uniform subsidence will not cause any damage; it isthe horizontal strain and curvature that are the major factors causingstructural damage (Singh, 1992).

The extension strain inKiirunavaaraminewas calculated bydividingthe differential horizontal displacement between two surveyingstations by the horizontal distance between them. Fig. 14 shows theaccumulated tensile strain along survey lines L, D and H, respectively.The mine coordinate X is perpendicular to the orebody and decreasefrom the hangingwall towards the orebody (see Figure 11). As expectedthe tensile strain increases with decreasing distance to the mining area.

Line D, in Fig. 14, shows a similar trend to line L though itundulates. On the other hand, the undulation is amplified along line Hbetweenmine coordinates X6550 and X6600. This means that there isa localized zone of strain concentration between two surveyingstations. This effect has been documented as abnormal deformation intrough subsidence where surface steps are formed over contrastinglithological contacts and faults (Donnelly and Reddish, 1994;Donnelly, 2006). Using a lineament map shown in Fig. 15, the zonesof strain concentrations were indicated. The lineaments have been

0

1

2

3

4

5

6

7

8

9

10

6400 6500 6600 6700 6800 6900 7000 7100 7200

Mine coordinate X

Acc

um

ula

ted

Ten

sile

str

ain

(m

m/m

)

LDH

Fig. 14. Accumulated tensile strain for line L, D and H.

26 T. Villegas et al. / Engineering Geology 121 (2011) 18–27

correlated with geological structures in the field (Magnor andMattsson, 1999).

The lower value of extension strain correlated with the presence ofa new surface crack was 2.3 mm/m in June 2006. Several damagecriteria of large rock mass can be found in the literature from differentcountries (Singh, 1992; Bell and Donnelly, 2006). For a horizontalstrain in the range of 0.5 to 2.0 mm/m only very light damage can beexpected, depending on the type of building or civil structure and

Fig. 15. Lineaments and points of strain concentrationModified from Magnor and Mattsson, 1999.

their physical condition. This level of strain can be reached before acrack could be observed on the ground.

7. Discussion

Herdocia's and Lupo's break angle prognoses for the Kiirunavaaramine showed the same tendency with decreasing break angle forincreasing mining depth. However, Lupo's prediction showed higherbreak angles. Their analyses were conducted for different crosssections of the mine and it is likely that the geomechanical conditionswere different in these two studies. On the other hand, break anglesobserved in the mine did not appear to follow any tendency. It seemsthat there are several factors that have not been accounted for in themodels.

The time–displacement behavior of the hangingwall is similar tothat of high-wall rock slopes. The creep response can be attributed tothe resisting force provided by the caved rock. Thus, the rate ofmovement of the hangingwall depends on themovement of the cavedrock which is linked to the mine production rate.

The CVD (critical vertical displacement) found in the time–displacement curves can be used to define the tension crack thatbelongs to the failure on the surface. In previous analyses itwasnot clearhow the location of the tension crack of the failure was determined. Theoutermost fracture on the surface is generated by extension strainbefore the failure surface is formed. If one of these fractures wasselected, lower break angles were calculated and lower estimated rockmass strength obtained from back-analysis of previous failures.

Table 2 provides valuable information that can be used for a roughestimation of the rate of movement of the ground surface. Forinstance, from themoment that a point on the surface experiences thefirst signs of subsidence it will take around 8 years to complete theregressive phase or to reach the CVD.

The limit of the discontinuous deformation zone has been used byLKAB to define a critical damage limit for buildings and civilstructures. Thereby, this limit is defined by tracking and mappingthe surface cracks along the hangingwall. However, the strain analysisindicates that the magnitude of tensile strain can reach critical levelsbefore a crack is observed. Regarding this fact the municipality of theCity of Kiruna and LKAB defined a maximum allowed extension strainof 3.0 mm/m. Buildings or houses that reach the strain limit will berelocated out of the affected zone.

There is a strain concentration along lineaments that areassociated with geological structures. These structures alter subsi-dence profile and perhaps a step will form in this area.

27T. Villegas et al. / Engineering Geology 121 (2011) 18–27

8. Conclusions

Previous break-angle prognoses showed well-defined tendenciesthat the break angles measured in the field have not shown. Modelsused for these predictions were generated with limited input dataregarding rock mass strength, in situ stress and structural geology.

The time-dependent response of the hangingwall to excavationresembles the behavior of large-scale open pit slopes. Three phases ofdeformation have been identified for the hangingwall of theKiirunavaara mine — regressive, steady state, and progressive.Through time–displacement curves it has been possible to determinethe movement rate of the hangingwall at different cross-sections ofthe mine.

The critical vertical displacement, CVD, has been defined andestimated for different places of the mine using the time–displace-ment curves recorded from surveying stations. The CVD can be used todetermine a potential location of the tension crack of the failure on thesurface.

The critical damage limit defined by outermost fractures in thehangingwall does not define the limit for damage since sensitive civilstructures can be affected by lower values of horizontal strain than the2.3 mm/m that was found to correlate with small cracks in the field.

Acknowledgments

The authors would like to thank the Hjalmar Lundbohm ResearchCentre (HLRC) and LKAB for support of this research work and for thepermission to publish the results. Thanks also go to Dr. Jonny Sjöbergfor his review of this paper.

References

Bell, F.G., Donnelly, L.J., 2006. Mining and Its Impact on the Environment. Taylor &Francis, London.

Bergman, S., Kübler, L., Martinsson, O., 2001. Description of Regional Geological andGeophysical Maps of Northern Norrbotten County (east of Caledonian orogen).Geological Survey of Sweden.

Broadbent, C.D., Zavodni, Z.M., 1982. Influence of rock structure on stability. Stability inSurface Mining, Volume 3. Society of Mining Engineers. Chap. 2.

Brown, E.T., 2003. Block caving geomechanics. JKMRC Monograph Series in Mining andMineral Processing 3. The University of Queensland.

Chizari, H., 1988. Kiirunavaaragruvans hänagvägg, En analys av sprickkartering ochdeformationsmätning i Zenobia, Master's Thesis. Luleå University of Technology(in Swedish).

Dahnér-Lindqvist, Ch., 1992. Footwall stability in KirunaMine. Swedish RockMechanicsDay, pp. 37–52. Stockholm, Sweden, SveBeFo.

Donnelly, L.J., February 2006. A review of coal mining induced fault reactivation in GreatBritain No.1 Q. J. Eng. Geology Hydrogeology 39, 5–50 Geological Society of London.

Donnelly, L.J., Reddish, D.J., 1994. The development of surface steps during miningsubsidence: “not due to fault reactivation”. Eng. Geology 36, 243–255 Elsevier.

Gustafsson, H.E., 1981. Field test of sublevel shrinkage caving (MSTM), at LKAB Kiruna.In: Stewart, D.R. (Ed.), Design and Operation of Caving and Sublevel Stoping Mines.AIME, New York, pp. 419–423.

Henry, E., Dahnér-Lindqvist, C., 2000. Footwall stability at the LKAB's Kiruna sublevelcaving operation, Sweden. In: Chitombo, G. (Ed.), Proceedings MassMin 2000,Brisbane. Australasian Institute of Mining and Metallurgy, pp. 527–533.

Henry, E., Mayer, C., Rott, H., 2004. Mappingmining-induced subsidence from space in ahard rock mine: example of SAR interferometry application at Kiruna mine. CIMBulletin, Vol. 97http://meteo9.uibk.ac.at/pub_listen/Rott_Henry_etal.pdf. No. 1083.Internet consult on March 2006.

Herdocia, A., 1991. Hanging wall stability of sublevel caving mines in Sweden, DoctoralThesis, Luleå University of Technology.

Hoek, E., 1974. Progressive caving induced by mining an inclined orebody. Trans. InstnMin. Metall. 83, A133–A139.

Kratzsch, H., 1983. Mining Subsidence Engineering. Springer-Verlag, Berlin.LKAB, 2001. Archive image. http://www.lkab.com2001.LKAB, 2007. Archive image. http://www.lkab.com2007.Lundman, P., Vollen, J., 1991. Bergmekanisk utvärdering av Sjömalmen och Zenobia

LKAB Kiruna, Master's Thesis, Luleå University of Technology (in Swedish).Lupo, J.F., 1996. Evaluation of deformations resulting from mass mining of an inclined

orebody, Doctoral Thesis, Colorado School of Mines.Magnor, B., Mattsson, H., 1999. Strukturgeologisk Model over Kiirunavaara, CTMG

report 00001. Luleå University of Technology. (in Swedish).Mäkitaavola, K., Taipalensuu, S., 2005. Rörelsemönster inom deformationsområdet

på Kiirunavaaras hängvägg. BSc Thesis, Luleå University of Technology, Sweden,(in Swedish).

Parák, T., 1973. Kirunamalmernas bildning. Doctoral Thesis (in Swedish). LuleåUniversity of Technology.

Savely, J.P., 1993. Slope management strategies for successful mining. In InnovativeMine Design for the 21st Century. Balkema, Rotterdam. pp. 25–34.

Singh, M.M., 1992. Mine subsidence, In: Hartman, H.L. (Ed.), 2nd Edition. SMEMiningEngineering Handbook, Vol. 1. Society of Mining Engineers, Lettleton, Colorado,pp. 938–971. Society of Mining Engineers.

Singh, U.K., Stephannsson, O., Herdocia, A., 1993. Simulation of progressive failure inhanging-wall and footwall for mining with sublevel caving. Trans. Instn. Min.Metall. 102, A188–A194 (Sec. A:Min. industry).

Stephansson, O., Borg, T., Bäckblom, G., 1978. Fracture development in hanging wall ofNorth Kiruna Mine. Technical Report 1978:51 T, Division of Rock Mechanics. LuleåUniversity of Technology, Sweden.

Villegas, T., Nordlund, E., 2008a. Numerical analysis of the hangingwall failure at theKiirunavaara mine. In: Shunnesson, H., Nordlund, E. (Eds.), Proceedings MassMin2008, Luleå, Sweden. Luleå University of Technology, pp. 867–876.

Villegas, T., Nordlund, E., 2008b. Numerical simulation of the hangingwall subsidenceusing PFC2D. In: Shunnesson, H., Nordlund, E. (Eds.), Proceedings MassMin 2008,Luleå, Sweden. Luleå University of Technology, pp. 907–916.

Villegas, T., Nordlund, E., 2010. Investigation of surface subsidence due to sublevelcaving. In: Potvin, Y. (Ed.), Proceedings of the Second International Seminar onBlock and Sublevel Caving, Perth, Australia. Australian Centre for Geomechanics,pp. 181–188.