PROJECT

Final Report of the RockSealing Project -Identification of Zones Disturbed byBlasting and Stress ReleaseL. BörgessonR. PuschA. FredrikssonH. HökmarkO. KärnlandR. Sanden

Clay Technology AB, Lund, Sweden

January 1992

TECHNICAL REPORTAn OECD/NEA International project managed by:SWEDISH NUCLEAR FUEL AND WASTE MANAGEMENT CODivision of Research and Development

Mailing address:Box 5864, S-102 48 Stockholm. Telephone: 08-665 28 00

FINAL REPORT OF THEROCK SEALING PROJECT -

IDENTIFICATION OF ZONES DISTURBED BYBLASTING AND STRESS RELEASE

Lennart BörgessonRoland Pusch

Anders FredrikssonHarald HökmarkOla Kärnland

Torbjörn Sanden

Clay Technology ABIDEON, S-223 70 Lund, Sweden

January 1992

original containscolor illustrations

This report concerns a study which was conducted for theStripa Project. The conclusions and viewpoints presented inthe report are those of the authors and do not necessarilycoincide with those of the client.

A list of other reports published in this series is attachedat the end of the report. Information on previous reports isavailable through SKB.

11

ABSTRACT

Tests 2 and 3 of the Rock Sealing Project compriseddetermination of the hydraulic properties of thedisturbed rock around tunnels and drifts and thepossibilities of decreasing the hydraulic conducti-vity of the disturbed zones by an attempt to seal thevery fine fractures that are causing the increasedconductivity. This report deals with the hydraulictesting while the grouting procedures and theireffect are described in Volume III.

The BMT drift was used for the experiments whichbasically consisted of measuring the flow of waterfrom an inner pressurized slot and borehole curtainto an outer curtain with zero water pressure (MacroFlow Test). The 12 m long drift was sealed from waterinflow by a slurry that filled the entire drift. Theslurry was pressurized by a large water filledbladder for eliminating leakage along and through thedrift.

The flow tests were primarily evaluated by finiteelement modeling in which the rock was considered anequivalent porous medium. Two possible rock modelswere developed that satisfied not only the flow andwater pressure measurements during the Macro FlowTest but also 3 different flow situations precedingthe test. ;he models imply a shallow blast-disturbedzone with strongly increased hydraulic conducti-vity, an t stress disturbed zone with a decreasedradial 1 r-iulic conductivity. The main differencebetween c te models is the extension of the blast-damaged r c-.j e and the axial hydraulic conductivity ofthe stré ,s disturbed zone.

iii

TABLE OF CONTENTS

11.11.21.3

22.12.22.32.3.2.3.2.3.

3

3.13.23.33.2.3.2.3.2.3.2.3.33.3.3.3.3.4

44.14.24.34.44.5

55.15.25.35.45.55.5.5.5.5.65.7

123.

1234

12

12

ABSTRACT

SUMMARY

SCOPE OF TESTBACKGROUNDPURPOSE OF THE TESTOUTLINE OF THE TEST

MAIN CHARACTERISTICS OF THE ROCKGENERALFRACTURE AND INFLOW MAPPINGGENERALIZED ROCK STRUCTURE IN THE AREAGeneralApplication of general modelAgreement with actual rock structure

FORMER ACTIVITIES AND MEASUREMENTS INTHE TEST AREAGENERALACTIVITIESGENERALExcavation of the test driftMacropermeability Experiment MEBuffer Mass Test (BMT)Sealing Test 1MEASUREMENTSWater pressure in the rockHydraulic conductivityPOSSIBLE INFLUENCE OF THE PAST ACTIVITIESON THE ROCK STRUCTURE

PREPARATIVE WORKGENERALDRILLING OF SLOTS AND BOREHOLE CURTAINSCONSTRUCTION OF THE INNER WALLARRANGEMENTS AT THE BULK HEADREINFORCEMENT OF THE ROCK SURFACE

PREPARATIVE TESTS AND MEASUREMENTSGENERALINFLOW MEASUREMENTSLUGEON TESTING OF CURTAIN BOREHOLESDETAILED INFLOW IN 2 BOREHOLESWATER PRESSURE MEASUREMENTSLocation and description of the equipmentWater pressure measured before the testFRACTURE MAPPING OF BOREHOLESROCK STRESS MEASUREMENTS

Page

ii

vii

1124

667777

14

1616161616171819191923

24

262626282829

30303032374242454647

IV

6 ARRANGEMENTS FOR PRESSURIZING AND FLOW-MEASURING IN THE INNER SLOT AND BOREHOLECURTAIN 49

6.1 GENERAL 496.2 SLOT AND BOREHOLE ARRANGEMENTS 496.3 ARRANGEMENTS FOR PRESSURIZATION AND FLOW

MEASUREMENTS IN THE INNER SLOT AND BORE-HOLE SECTIONS 51

7 ARRANGEMENTS TO PRESSURIZE AND SURFACE

7.17.27.2.17.2.1.7.2.1.7.2.1.7.2.1.7.2.1.7.2.27.3

7.3.17.3.27.3.37.3.47.47.5

8

8.18.2

8.2.13.2.20.2.38.2.4d.2.5

12345

SEAL THE DRIFTGENERALLINING TECHNIQUEPre-testing of different lining techniquesGeneralAquata EpoxyEpoxy-PolyurethaneMeycoprenProcoatApplication of Procoat iiningTECHNIQUE WITH A BLADDER COMBINED WITHA BENTONITE SLURRYGeneralBladderCageSlurrySLURRY FILLING PROCEDURESEMPTYING PROCEDURES

PREPARATIVE FLOW MODELLING AND FLOWCALCULATIONSGENERALFEM CALCULATIONS OF THE FLOW TEST BEFOREGROUTINGElement model and basic dataStage 1 (without curtains)Stage 2 (with curtains)Stage 3 (after surface sealing)Stage 4 (after pressurization of theinner curtain)

56565656

5757575758

59596060626570

7171

7171737475

76fe.3 FEM CALCULATION OF THE FUNCTION OF THE

BOREHOLE CURTAIN 77•3.4 FEM CALCULATION OF LUGEON TESTING IN

SHALLOW BOREHOLES 78

9 MACROPERMEABILITY TEST FOR AXIAL FLOW 809.1 GENERAL 809.2 TEST SCHEDULE 803.3 RESULTS 829.3.1 General 829.3.2 Transient flow 829.3.3 Stationary flow measurements 849.3.4 Water pressure measurements 86

10 PERMEABILITY TESTING OF HEDGEHOG HOLES 9010.1 BACKGROUND 9010.2 MEASUREMENTS 9 010.3 TEST AREAS AND TEST RESULTS 90

1111.111.211.311.3.111.3.211.3.3

11.3.411.411.4.111.4.211.4.311.4.3.1

11.4.3.211.4.3.311.4.3.411.4.3.511.4.411.511.5.111.5.211.5.311.6

1212.112.212.312.412.4.112.4.212.4.212.512.5.112.5.1.112.5.1.212.5.1.312.5.1.412.5.212.5.2.112.5.2.212.5.2.312.5.2.412.5.312.612.6.112.6.212.712.812.8.112.8.2

EVALUATIONMETHODOLOGYSEPARATION OF BACKGROUND AND LEAKAGE FI.OWID EVALUATION OF THE MACRO-FLOW TESTGeneralAverage hydraulic conductivityHydraulic conductivity of the floor, wallsand roofRelevance of ID evalutation3D EVALUATION OF THE MACRO-FLOW TESTGeneralSlurry leakage calculationsAchieved rock aodel (Model A)Hydraulic conductivities and boundaryconditionsMacroperveability testInflow test after drilling of curtainsSlurry leakage testMacro-flow testAlternative rock model (Model B)EVALUATION OF SINGLE HOLE TESTGeneralHedgehog holesCurtain holesCONCLUSIONS

ROCK MECHANICAL CONSIDERATIONSGENERALMODELLING APPROACHSTRESS DATAMATERIAL MODELSIntact rock propertiesFracture properties, UDECFracture properties, 3DECEFFECTS OF TUNNEL EXCAVATIONUDEC calculationModel geometryIn-situ stresses and boundary conditionsCalculated casesResults3DEC calculationsModel geometryIn-situ stresses and boundary conditionsCalculated casesResultsImportance of fracture orientationsEFFECTS OF PRESSURIZATIONEffects of joint aperturesEffects on axial conductivityEFFECT OF BLASTINGCONCLUSIONSExcavationEffect of pressurization

969697999999

100104106106107109

109111113116118124127127129130134

135135135136136136136137138138138139139139143143144144145145149150150150153153154

13 CONCLUSIONS 15613.1 GENERAL 15613.2 CONCLUSIVE OBSERVATIONS 15613.3 DERIVED MODEL OF THE NEAR-FIELD ROCK

AROUND THE BMT DRIFT 157

ACKNOWLEDGEMENTS 160

REFERENCES 161

APPENDIX I

APPENDIX II

VII

SUMMARY

The purpose of Tests 2 and 3 was to measure the axialhydraulic conductivity of the zones disturbed byblasting and stress release around an excavatedtunnel or drift in granite rock and to investigatethe possibility to reduce the hydraulic conductivityby sealing the very fine fractures in these zones bycement grouting. The conductivity measurements andtheir evaluation are given in this report while theresults of the grouting are reported in Volume III.

The test, called the Macro-Flow Test, was performedin the innermost part of the drift used for theBuffer Mass Test and LBL's Macropenneability Experi-ment. A 70 cm deep slot was cut around the inner endof the drift and a borehole curtain, consisting of 72boreholes and extending 6.3 m into the rock from theslot, was drilled perpendicular to the drift axis. Asimilar slot and borehole curtain was made at thebulk head, 13 m from the inner curtain.

The idea of the Macro Flow Test was to pressurize thewater filled inner curtain and measure the flow ofwater to the outer curtain. The surface of the rockin the drift was sealed by filling the drift with 120m bentonite slurry, which was pressurized by a 100m3 water filled bladder. By keeping the slurry pres-sure higher than the curtain pressure, water was pre-vented from entering the drift.

The water pressure was measured at about 30 locationsin the surrounding r^ck and the water flow wasmeasured in 8 sections of the inner curtain and in 4sections of the outer curtain, before as well asduring the test.

The Macro-Flow Test was conducted in more than 10steps, with different relations between the slurrypressure and the curtain pressure, the maximum pres-sure being just below 1 MPa. These steps showed thatthe relation between applied pressure in the innercurtain and measured flow was linear and independentof the pressure on the rock surface from the slurry.

The test was primarily evaluated by finite elementcalculations of the flow and pressure, at which therock was simulated as an eqivalent porous medium. Thefollowing four different tests were simulated:

vm

- Before drilling of curtains (MacropermeabilityTest)

- After drilling of curtains (Curtain Inflow Test)- After pressurization of the slurry with zeropressure in the curtains (Slurry Leakage Test)

- Macro-Flow Test

The calculations showed that two different rockmodels could be defined, which both fitted themeasured water floj and pressure in the rock. Both•odels imply a highly permeable zone with a depth ofat least 0.8 m and an average hydraulic conductivityof 110'8 m/s which is more 100 times higher than theconductivity of the virgin rock and which mainlyoriginate from blast damage. The hydraulic conducti-vity is highest in the floor and lowest in the roof.Both models also imply a stress-disturbed zonereaching 3 m into the rock with a decreased radialhydraulic conductivity by about 4 times.

The main difference between the models concern theaxial hydraulic conductivity of the stress-disturbedzone and the depth of the blast disturbed zone asfollows:

Model A

Blast disturbed zone extending 0.8 m from surface.Stress disturbed zone extending 3.0 m from surfacewith an increased axial conductivity with a factor10.

Model B

Blast disturbed zone extending more than 0.8 m fromsurfaceStress disturbed zone with an axial conductivity thatmay range from that of Model A, i.e. 10 times higherthan the conductivity of virgin rock, to a figurecorresponding to the radial conductivity, i.e. 4times lower than the virgin rock

Individual measurements in the curtain holes and rockmechanical calculations support model A but thematter should be further studied.

SCOPE OF TEST

1.1 BACKGROUND

The question of a possibly existing excavation-disturbed zone with enhanced axial conductivityaround tunnels arose in the Buffer Mass Test in Phase1 and the Tunnel Plugging Experiment in Phase 2 ofthe Stripa Project. These tests clearly indicated theexistence of a disturbed zone, which will stronglyaffect the function of a repository. The depth andhydraulic conductivity of the zone were concluded tobe of primary importance, a major question being towhat extent they depend on the natural fracturepattern and the geometry of the drift and on the ex-cavation technique. A further question of greatimportance is naturally if such zones can be sealed.

All these questions cannot be completely answered byconductivity tests at only one test site, since thenatural variation in rock structure would hardly makeit possible to draw general conclusions. However,careful measurements in a well characterized rockmass combined with mathematical modelling wouldincrease the knowledge significantly.

Since the BMT drift in Stripa was well known from theprevious tests (BMT and before that the LBL ventila-tion test) and since the drift was already equippedwith some water pressure measuring systems and apartly intact bulkhead for separating the test areafrom the rest of the drift, it was decided to use itfor investigating the disturbed zone. It was assumedto consist of two parts, one dominated by blastingeffects (disturbed zone by blasting) and the otherdominated by stress release (zone disturbed by stressrelease).

The introductory study of the Rock Sealing Project,concerning "Sealing materials and techniques" wassucceeded by a series of pilot tests in Stage 1,which resulted in a decision to continue with Stage 2which was divided into 5 subprojects called Test 1 -Test 5. Test 2 dealt with the hydraulic properties ofthe disturbed zone by blasting and the possibilitiesto seal it, while Test 3 concerned the disturbed zoneby stress release.

The final reporting of Tests 2 and 3 will be made inthe following two reports:

- Final report of the rock sealing project - VolumeII: Identification of zones disturbed by blasting andstress release.

- Final report of the rock sealing project - VolumeIII: Grouting of zones disturbed by blasting andstress release.

This present report (Volume II) deals with the workdone to identify and measure the hydraulic propertiesof the disturbed zones, while the other report(Volume III) deals with the grouting and the hydrau-lic testing after grouting.

1.2 PURPOSE OF THE TEST

The initially defined purpose of these tests regar-ding the hydraulic properties was the following:

"to find out whether there is a disturbed zone aroundblasted tunnels" (Test 2) and "to check whether theaverage conductivity of the rock at larger distancethan 1.5 • from the drift is at all lower than thatof the ungrouted shallow zone and to find out whetherthe disturbed zone has a radial extension of »orethan about 1.5 m" (Test 3)

The intention was primarily to measure the axial con-ductivity and to find out whether it is continuousover longer distances. In order to be able to measurethe axial conductivity at a fairly long distance, itwas decided to use the whole 12 m long BMT drift,equipping it with a 0.7 m deep slot at both ends, fortesting the axial conductivity of the blasting dis-turbed zone and with 7 m deep borehole curtains atthe ends for testing the axial conductivity of thestress-disturbed zone.

Fig 1-1 shows a picture of the principles of thetests. A high water pressure was applied in the innerslot and borehole curtain and the axial conductivitymeasured by collecting the water flowing parallel tothe drift into the outer slot and curtain. This tech-nique required a lining on the rock surface for pre-venting water from flowing into and through thedrift.

After the hydraulic tests "hedgehog" drilling andgrouting of the blast-disturbed zone were done andthe hydraulic test then repeated, by which the effectof the hedgehog grouting was determined. In a thirdstep an additional borehole curtain was planned to bedrilled and grouted and the result investigated by anew hydraulic test for determining the possibility ofreducing it by grouting. This third step was nevertaken. The hedgehog grouting and their result will beaccounted for in Volume III, while the intended grou-ting of a new borehole curtain was not performed.

VIRGIN STATE

DB DISTURBED ZONE BY

BLASTING

DS DISTURBED ZONE BY

STRESS RELEASE

'HEDGEHOG' GROUTING

DB-ZONE SEALED

"CURTAIN" GROUTING

DS-ZONE CUT-OFF

Figure l-i Outline of tests 2 and 3. The axial conductivity ofthe zone disturbed by blasting DB and the zonedisturbed by stress release DS was measured asshown in the upper picture and after hedgehoggrouting the resulting conductivity was Measured inthe sane way, as shown in the middle picture. Thegrouting and subsequent testing of the DS zone werenot performed.

HG -

HG

HG

K

Figure 1-2 A vertical section of the BMT drift and the origi-nally planned arrangements for the tests. A water-tight lining was to be pressurized with water frominside.

1.3 OUTLINE OF THE TEST

Fig 1-2 shows a schematic section of the inner partof the BMT drift and the originally planned arrange-ments for the tests. The two heater holes, used inthe BMT study were filled with a compacted mixture ofsand and bentonite to minimize the influence of theseholes. The rock surface was sealed with a watertightlining and the drift filled with water under thepressure pi. In this way the drift was kept hydrauli-cally isolated from the surrounding rock. At theinner end of the drift a concrete wall was built,covering the pressurized inner slot and forming oneend of the test drift, while the bulkhead and theouter slot constituted its other end. The boreholecurtains, termed K in Fig. 1-2, were drilled from thebottom of the two slots. The borehole curtain inten-ded for grouting the stress disturbed zone but neveractually drilled, is called G. The figure also showsthe HG holes (5 holes) and the R holes (also 5) fromthe LBL measurements.

The inner slot and K borehole curtain were filledwith water and pressurized with the water pressurep2. pz consistently had to be lower than pi in orderto prevent water from entering the drift. Bymeasuring the inflow of water to the inner slot andborehole curtain and by also measuring the outflow ofwater to the outer slot and borehole curtain, theaxial hydraulic conductivity of the disturbed zonescould be estimated.

As will be described later in the report, it was notpossible to construct a watertight lining. Instead,the drift was filled with a thick slurry of bentonitewhich was pressurized by a large rubber bladder thatwas placed in the drift. By the low viscosity of thestiffened tixotropic bentonite slurry, leakagethrough the bulkhead/rock connections was eliminatedand a high slurry pressure maintained in the drift(Fig. 1-3).

' &'•.• • • •••+•.+'. '•• -.+ • •*:: •

Figure 1-3 The final arrangements for the tests. A bentoniteslurry (B) was pressurized through a bladder filledwith water from inside.

MAIN CHARACTERISTICS OF THE ROCK

2.1 GENERAL

The tests were performed in the inner part of the BMTdrift. The Extensometer drift was also used in thetests but only for storage of the bentonite slurrybetween the two test series. The location of thesedrifts in the Stripa Mine is shown in the overviewdrawing in Fig 2-1.

The rock in the test area was carefully investigatedduring LBL's "Macropermeability Test" as well asduring the Buffer Mass Test. The most importantfeatures will be described in this chapter as well asa generalized model of the rock structure in the BMTarea.

y850 y900 y950 y1000 yi050x450

-xAOO

-x350

-x300

-x250

x200

Figure 2-1 Location of the BMT drift, which was used for theflow tests and the Extensometer drift (Ext.), whichwas used for storage of the bentonite slurry.

2.2 FRACTURE AND INFLOW MAPPING

All the walls were carefully mapped by LBL during theventilation test. This napping, which is shown in Fig2-2, included all the fractures, which were numberedand described on data files. (Rouleau et al, 1981).The mapping did not specify whether the fractureswere carrying water or not.

During the buffer mass test the distribution of in-flow of water into the test area was estimated byidentifying the successive moistening of the roof andthe walls after stopping the ordinary ventilationduring one weekend. The result is interesting for theoverall picture of the inflow of water from the rockin the area although there are difficulties and un-certainties in evaluating the observations. Fig 2-3shows the inflow pattern in the roof and in thewalls. The floor could not be mapped in this way.

The figure shows that the eastern wall had thelargest water-bearing zones while the tunnel frontwas the driest part. The inflow from the western walland the roof were dominated by a few discontinuities.

2.3 GENERALIZED ROCK STRUCTURE IN THE AREA

2.3.1 General

Rock mechanical and conductivity calculations requiredefinition of the rock structure, which must be gene-ralized and simplified. The rock structure describedin this chapter was based on identification of majorlong-extending, water-bearing fractures, referring toa generalized rock structure scheme developed forrationalizing functional analyses of Swedish reposi-tory concepts. This scheme is described in detail inAppendix II.

2.3.2 Application of general model

Already the LBL study revealed the existence of amajor steep structure, which is hydraulically veryactive and which is termed "RP" in the present study(Fig.2-4). It is concluded to belong to the set of3rd order discontinuities that have a spacing of 50-150 m in the general structure model and is assumedto be a member of the set of almost parallel struc-tures termed J,K,M,RP, characterized by an averagespacing of around 75 m. The RP zone is intersected bythe flatlying B zone that passes just below the NWcorner of the drift and which is assumed to be amajor water supplier to the drift (Fig.2-5).

8

nemomurvn nucc *m mmu

nrtnKt »>••!» ;< ;̂

/ VIM or Kuunrc. o»r cnooTt AMC cm.omrc

• •• tmeon im O«.ODITC I O XMCIU US

•• wr-touuit IIIIH. neon un «o«t» w M <

y_- MUS COVC«. »LOO» OKLT

lufl^ACC. RMTH « M . L ONI.TO r »

Figure 2-2 Fracture mapping of the walls in the BMT drift

-I-

/.•A'.Y.Y.

1 1 i 11» lift. • f

• ' • ' • ' • ' • • >

".VAv.'v'.vXv

•'*'•:•:

ROOF

.£É? ! W-WALL

;.0:>^-;-:

Figure 2-4 Major long-extending fractures exposed in the BMTdrift. They are 4th order discontinuities while theRP zone is one of 3rd order.

The total number of major water-bearing long-extending fractures exposed in the inner 11 m longpart of the BMT drift is around 13, which correspondsto an average spacing of these major discontinuitiesof about 3 m. Applying the rock structure model tothe BMT rock, i.e. approximating the major disconti-nuities to form conformous orthogonal patterns, onewould get the simplified pattern of continuous long-extending (4th order) fractures and "activated",latent fractures (5th order) in the assumed disturbed

11

vA

Figure 2-5 Parallel perspective of the BMT drift showing thesteep RP zone intersected by the flatter HB-zone(the intersection marked by coarse broken line),the LBL holes Rl, R4, R5, HG3, and DBH2, and thetwo heater holes 1 and 2.

12

RP3rd ORDER

Figure 2-6 Rock structure model applied to the BMT drift

zone around the periphery as shown in Fig.2-6. Allthe steeply oriented discontinuities shown in thisfigure would be more or less parallel or perpendicu-lar tn the RP plane, i.e. striking WNW/ESE or NNE/-SSW, meaning that the first-mentioned dip slightlytowards NE, while the other steep set has a dip thatis expected to be in the range of 0-20° towards NW to0-20° towards SE if the general feature of one majorset being subhorizontal applies.

13

Assuming that the rock structure is sufficientlysimple to be characterized by only one orthogonalsystem of fractures, one would expect a verticalcross section fitting the pattern in Fig 2-6 to belike that shown in Fig. 2-7.

This would have the following implications withrespect to the axial conductivity and to thepossibility of applying grouting for successfulsealing:

The 3 m deep heater holes should be inter-sected by 1 or 2 flatlying fractures below1 m depth, and by 2-4 flatlying fracturesabove 1 m depth. Above 1 m depth theyshould also contain 4-6 steep fractures,while below this depth there should bemaximum 1 steep fracture. This is inreasonable agreement with actual observa-tions

The 7 m long holes drilled radially fromthe ir.ner and outer ends of the BMT driftin conjunction with Test 2 should have anaverage number of intersected water-bearing fractures of 3 per hole. The actualnumber of significantly water-bearingfractures identified by borehole inspectionwas found to be around 2 per hole, the dis-crepancy being due to the fact that the

SECTION A-A

Figure 2-7 Cross sections of the BMT drift. The disturbed zonewith activated 5th order discontinuities normal tothe periphery is taken to be 1 m

14

measurements were of short duration,leaving some less strongly water-bearingfractures undetected. In principle, therock structure model is hence supported bythese observations

3. The orientation of fractures intersectingthe radially drilled holes should be eithersteep or subhorizontal with relatively fewfractures in intermediate positions. Thisfits well with the borehole inspectionwhich showed that 35% of all fractureswere flatlying, 57% very steep, and 8%"erratic". Hence, these observations alsosupport the applicability of the model

4. The minimum angle between the drift andthat of the steep fractures is sufficientlyhigh (15-20°) to give only a moderate con-tribution to the axial conductivity sincethe fractures will offer tortuous flowpaths and the degree of continuity may besmall.

Considering the roof and floor, the corre-sponding angle is critical, however, and anumber of water-bearing fractures would beexpected to contribute considerably to theaxial conductivity of these parts.

5. The expected effects of blasting cannot beexplicitly evaluated, this matter istreated in chapter 12.

2.3.3 Agreement with actual rock structure

Visual inspection and measurement of the actual majorwater-bearing fractures revealed a rock structurethat was not very different from the simple rockstructure that one would expect by applying thegeneral model. However, it was concluded that thereare three superimposed systems of long-extendingfractures, which are accounted for by the generalizedsection in Fig. 2-8. The strike of the individualsets is not very different, while the dip is clearlydifferent of the steeply oriented sets. One findsthat the three sets combine to form wedges that areexpected to produce local large fracture widening andrisk of unstable conditions.

15

nr

Figure 2-8 Generalized rock structure of BMT as evaluated fromvisual inspection

16

FORMER ACTIVITIES AND MEASUREMENTS IN THE TEST AREA

3.1 GENERAL

Three large tests have been conducted in the BMTdrift prior to the described Tests 2 and 3. Since allthese tests have involved heating to differentlevels it is most likely that they have affected therock in some way. The tests will be roughly describedin this chapter as well as the technique for exca-vating the drift.

Several measurements in the rock have been nadeduring the preceding tests and those which may be ofimportance for the evaluation of Tests 2 and 3 willbe described in here.

3.2 ACTIVITIES

3.2.1 Excavation of the test drift

The BMT drift was excavated in 1977 for use in theMacropermeability test. A new type of smoth blastingwas used for the drift as well as for the entire 400m tunnel length that was excavated for the LBL/KBStests. The technique means that lower charging isused in the contour holes than in the rest of thedrilled holes.

The drilling and charging pattern is shown in Fig3-1. The two central large t 89 mm holes and all thesmaller 4 35 nun holes, except the outer ring of holesNos. 42-62, were charged with 0.8 kg ANFO per meterand blasted at first. Then the outer ring was chargedwith 0.3 kg Gurit per meter and blasted. By usingthis technique it was expected that the fracturescaused by blasting would not go deeper than 0.3*0.5meters in the walls according to old findings whichimply that the depth of the disturbance by blastingin meters will be about the same as the charge inkg/m. Core drilled holes in walls of tunnels blastedin this fashion were concluded to confirm this rule.

The length of the drilled holes and thus also thelength of each set of blast-holes was about 3.6meters. The technique is described in SAC-08(Andersson & Halén, 1978). It is implicite that moreexplosives were used in the blast-holes of the floorin the BMT drift.

17

OfOlim

•

• .ft ••*• \

.**

••»r. O •«

•1 ml*

\• * • *

••ZT V

3 -

3 7 -

4 7 -

nr

31

41

67

SOm

ANfO

- " -

» •»

0«

0.8

18

ventilated from the drift. Different air temperatureswere used in the drift, the highest temperature being45°C in a test with a duration of one month. The tem-perature was measured at different distances from thedrift and the temperature in July 1980 is shown inFig 3-2 as a function of the distance from the peri-phery. The temperature was increased up to 30°C onthe rock surface, moreothan 10°C to the depth 3meters and more than 1°C to 13 meters depth. The tem-perature gradient was thus about 10°C per meter overthe first 2 meter distance from the periphery.

3.2.3 Buffer Mass Test (BMT)

The Buffer Mass Test (BMT) was part of the firstphase of the Stripa Project, conducted in the years1980-1984. The drift inside the new bulk head,constructed for this test, formed the inner 12 m longpart of the 33 m drift used in the ME experiment. Thelower 3/4 of the drift was backfilled with a mixtureof 10% bentonite and 90% sand of high density whilethe upper 1/4 was filled with a mixture of 20% ben-tonite and 80% sand that was blown in. The swellingpressure exerted by the backfill was measured atseveral spots and formed to be in the range of 10-20kPa, occasionally as high as 100 kPa.

20

10

• Temperoiure hole doto• R 04 data

\ \ [ 1 I 1 I 1

0 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25Radioi distance from drift wall (m)

Figure 3-2 Temperature variation with distance from drift wallfor R04 and temperature hole on July 25, 1980

19

The two heater holes, with a diameter of 76 cm and adepth of about 3 m, that had been drilled in thefloor, were filled with highly compacted bentonitethat gave a pressure of up to about 10 MPa. Theheaters were powered at 600-1400 W for differentperiods of time, which resulted in a maximum tempera-ture in the rock of 70°C in hole No. 1 and 35°C inhole No. 2.

3.2.4 Sealing Test 1

In the present third phase of the Stripa Project,Tests 2 and 3 of the Rock Sealing project were pre-ceeded by Test 1 which is described in a separatereport. Since Test 1 was also conducted in the BMTdrift, a short description of the activities that mayhave affected Tests 2 and 3, will be given below.

In Test 1, the heater holes Nos. 1 and 2 were groutedwith bentonite slurry from inside the holes. A dyna-mic pressure with an average magniture of between 2and 3 MPa and 40-120 Hz pressure pulses with anamplitude of 200-400 kPa was applied during theinjection. After the grouting operations, the holeswere heated at a power of about 2.5 kW during 2-3months. The heating resulted in a temperatureincrease in the floor of up to 90°C 1-2 m below thesurface at the periphery of the holss. At that depththe temperature rose by 1O°C 2-3 meters from theholes which means that almost the entire floor in thedrift was heated byomore than 10°C. Heat calculationsalso showed that 10°C increased temperature wasreached as deep as 3-4 meters below the surface inmost of the floor.

3.3 MEASUREMENTS

Several hydraulic measurements were made during thepreceding tests. The entire Macropermeability Experi-ment was aimed at determining the hydraulic proper-ties of the surrounding rock and much valuable infor-mation was obtained in this test.

3.3.1 Water pressure in the rock

Piezometric recording was made at fairly largedistances in the surrounding rock in the ME whilesuch recording took place close to the drift in theBMT. The pressure was measured in 15 about 30 meterdeep, radically oriented holes around the drift inthe ME as indicated in Fig 3-3, which shows the loca-tion and orientation of the core drilled holes with76 mm diameter. The pressure was measured at every5th meter, which means that there were 60 recording

20

- 4 m

BACK OF VENTILATIONDRIFT

VENTILATIONDRIFT

SIDEVIEW

76 mm 030m LENGTH

30m

Figure 3-3 Location and numbering of the R and HG holes in theMacropermeability Experiment. The BMT bulk headwas placed 12 m from the end of the drift.

points altogether. The holes were packed off over alength of 3 m from the periphery, meaning that noreadings were taken close to the drift.

The measured pressure in the ME can be plotted as afunction of the distance to the centre of the driftas shown in Fig 3-4. The curve for the weightedaverage is a straight line in the semi-logarithmicdiagram and the hydraulic conductivity can be evalua-ted from the inclination of the line, if the inflowis known.

21

160

140

120

£ 100

oto

• 5o

22

Despite a high degree of water saturation of thebackfill no high pressures were recorded close to therock/backfill interface. Thus, the pressures in thefloor were between 30 and 40 kPa, while those 20-30cm into the walls varied between 10 and 35 kPa. At1.0 m depth in the walls the pressure reached 105 to290 kPa at the end of the test. Pressure contoursfrom the pressure in the very shallow rock are shownin Fig 3-5. The two sections through the two deposi-tion holes are shown. The high pressures 490 and 750kPa represent the pressure below the base of thedeposition holes.

Figure 3-5 Pressure contours at the rock/backfill interface atsections through heater holes Nos. 1 and 2 atbefore excavation of the backfill in late summer1984. The full line represents the section throughhole No. 1, while the broken line represents thesection through hole No. 2.

23

3.3.2 Hydraulic conductivity

The inclination of the average pressure drop line inFig 3-4 can be used for evaluating the hydraulic con-ductivity which was also the idea of the whole test.This evaluation gave

K=1.01(flom/s

for all tests (three different air temperatures).

The hydraulic conductivity of the rock was alsomeasured by single hole tests in the R and HG holesin altogether 146 sections using water flow tests.The authors' own compilation of conductivity datafrom SAC 46 gave the distribution shown in Fig 3-6.The geometric average of the values for the R-holesis fc=1.8-10~ °m/s which agrees very well with the MEmeasured value. In the preparative flow modellingrelated to Tests 2 and 3, ̂ l.O-lO"1 m/s has beenassumed to be the average hydraulic conductivity ofthe undisturbed rock.

25 .

20 -

10

5 -

0 -

1

1 1111 !• •

i11| 1 n . .o1.00E-12 1.00E-11 1.00E-10 1.00E-09

Hydraulic Conductivity, m/s

1.00E-08

Figure 3-6 Compilation of permeability measurements accountedfor in Figs 3.8-8.21 in SAC 46.

24

The axial hydraulic conductivity of the shallowblast-disturbed zone can be roughly estimated fromFig 3-5 if all water that entered the inner part ofthe BMT drift is assumed to flow out of the drift inthe disturbed, previous zone parallel to the drift,since the hydraulic gradient is known. With anaverage pressure drop of 1.2 mwh along 6m, a flowrate of 22 1/d and a cross section area of 20 m2

(about 1 m zone), the hydraulic conductivity of theblast-disturbed zone is found to be:

The average radial hydraulic conductivity of the zonebetween 1 and 3 meters from the surface of the driftcan also be roughly estimated by considering Fig 3-4.Thus, consider the average water head 70 m 3 m fromthe drift wall and assuming a very low pressure (0-10mwh) 1 m from the wall, the radial conductivity willbe about

k=4.010~nm/s

Thus the data from the previous tests indicate thatwe have:

- A disturbed zone (mostly from blast damage) 0.5-1.0m from the drift wall with k«510"8m/s and nopronounced anisotropy

- A zone reaching from 0.5-1 m to 3-4 m from thesurface with a decreased radial hydraulic conduc-tivity of fc«4-10 m/s

- A hydraulic conductivity of the surrounding un-disturbed rock more than 3-4 m from the peripheryof k=1.010~ m/s

3.4 POSSIBLE INFLUENCE OF THE PAST ACTIVITIES ON THE ROCKSTRUCTURE

The activities in the previous tests may have hadsome effect on the near field rock around the BMTdrift. Thus, it is possible that the heating, pressu-rization, and grouting in Test 1 and perhaps also thehole drilling may have affected the rock and thesematters will be discussed here.

Heating

Above all, the heating of the rock may influence thefracture system of the rock. As was shown at calcula-tions in connection with Test 1, the residual effect

25

of a heating/cooling cycle may be substantial due toplastic behaviour of the fractures. A rough estimateof how large the residual effect of the heat pulsesin the area may have been suggest that the averagechange in aperture is about 20% of the change causedby the tunnel excavation. The result is of coursevery much dependent on the fractrure geometry andproperties and it is difficult to give some averagefigure especially if the blasting damage is included.

Pressurization

The rock was exposed to pressures in Test 1 and inthe BMT. Only the heater holes were significantlypressurized, while the swelling pressure of the back-fill in the BMT was very low. The inside pressureduring BMT was about 10 MPa while it was only about 4MPa during the grouting in Test 1. No 3D calculationof this effect has been done but the influence can becompared to that of an increase in temperature. Anincreased temperature by 40-80°C will cause muchhigher tangential stresses in the vicinity of thehole than the internal pressure 10 MPa. The tensionstresses caused by the pressurization will be in theorder of 10 MPa and will only reduce the initial com-pressive stresses that are higher than 20 MPa. Thetangential stresses from the heating may have been 50MPa, and it is thus probable that the effect of theinternal pressurization is smaller than the effect ofthe heating, except, maybe, for the case when fairlyloose blocks are present in the floor.

Grouting of deposition holes

A local decrease in hydraulic conductivity of thefloor was achieved by the grouting of the depositionholes which may have affected the results of Test 2to some extent. The results from Test 1 showed thatthe heat pulse destroyed a large part of the sealingeffect, but it is most probable that the initial hyd-raulic conductivity of the floor was somewhat re-duced, at least in the upper part where very signifi-cant grout penetration into flatlying fractures wasobserved.

Hole drilling

The two heater holes were core drilled with verylittle effect on the surrounding rock. The extent ofpercussion drilling before Test 2 was very small andmay not have affected the rock significantly. Theonly drilling activity that may have affected therock is the comprehensive drilling of the slots andborehole curtains for Test 2. However, the effect ofthose activities is very local and has probably onlyimproved the function of the borehole curtain.

26

PREPARATIVE WORK

4.1 GENERAL

The test was preceded by reconstruction of the bulkhead, drilling of slots and borehole curtains, exca-vating of concrete for the pump well outside thebulkhead and casting of the inner wall. These testpreparations will be described in this chapter.

4.2 DRILLING OF SLOTS AND BOREHOLE CURTAINS

The borehole curtains consisted of boreholes with 7 mdepth and a separating drilled angle of 5° in a fan-shaped pattern as shown in Figs 4-1 and 4-2. Theywere numbered by the angle that they form with the

KA HOLES

33020

25 30

300 60

120

210 150

Figure 4-1 Numbering and location of the boreholes in theouter KB curtain

27

KB HOLES

20330 30

300 60

210 150

Figure 4-2 Numbering and location of the boreholes in theouter KB curtain

upper vertical hole, i.e. from 0 to 355, the totalnumber being 72 in each curtain. The distance betweenthe axes of neighboring holes was about 20 cm at therock surface and 90 cm at the inner end of the holes.As will be shown in chapter 6 the holes are closeenough to form a hydraulic slot with the depth 7 m.

The holes of the inner curtain are named KA-holes.They were percussion drilled with the diameter 48 mm.For practical reasons, the inner borehole curtain waslocated about 0.5 m from the end wall.

The holes in the outer slot are termed KB-holes. Theywere also percussion-drilled but with 35 mm diamter.The western buttress of the bulk head and concreteslab made the drilling of these hole quite difficult.For practical reasons the borehole curtain was placed1.2m outside the bulk head. Drilling through theheavily reinforced slab had to be made by comprehen-sive core drilling through the slab. For avoidingparticularly difficult parts some of the KB holes had

28

to be moved and redirected. The arrangements for flowmeasurements in the KA- and KB-holes will be descri-bed in chapter 6.

After completion of the borehole curtains, the slotswere drilled with percussion technique. The holes hada diameter of 100 mm and were drilled with a spacingof about 70 mm. By this överlapp, a continous slot ofsufficient width was created. The depth of the innerslot was 70 cm while the outer slot was only 40 cmdeep.

4.3 CONSTRUCTION OF THE INNER WALL

In order to separate the inner slot from the driftand simplify the application of the lining on the endwall of the drift, the rock at the inner end wascovered by a concrete wall (Fig 4-3). The wall whichwas about 75 cm thick and strongly reinforced , wasanchored in the rock by 20 steel rods. The slot wasfilled with mineral wool in order to prevent theconcrete from filling the slot. The installations formeasuring the water flow in the slot and in the bore-hole curtain will be described in chapter 6.

4.4 ARRANGEMENTS AT THE BULK HEAD

The excavation of the backfill in BMT requiredpartial removal of the bulk head, and the reconstruc-tion, which comprised extensive repair of the steelwall and steel buttresses, took several months. Likein the BMT, the wall was design to withstand a pres-sure from inside the drift of 3 MPa and it was alsoequipped with an 0,5 m opening with a strong lid forgiving access to the test drift.

6300

BOREHOLECURTAIN

DRILLEDSLOT• 100

— 'H

Figure 4-3 The inner concrete wall with the slot and boreholes

29

The BMT involved construction of a concrete slab of1,6 in thickness that covered the floor outside thebulkhead and part of this slab had to be removed inorder to arrange a well for collecting water in thefloor were no slot could be drilled. This requiredadditional support of the bulkhead which was made inthe way shown in Fig 4-4.

4.5 REINFORCEMENT OF THE ROCK SURFACE

To prevent rockfall during the hedgehog grouting theroof and walls in the drift were reinforced by 50bolts inserted in critical rock blocks. The bolts hadthe diameter 20 mm and were anchored 2 m into therock in 32 mm boreholes. The bolts were cast into theholes with the CEMBOLT system, i.e. without pre-stressing.

v

V

v

>j

y

I i_ / _ / •

/ /

' /

DOOR

\ \\ V\ \

\ \

STEEL WALL

BRACES

EXCAVATIONIN SLAB

BRACES

Figure 4-4 The arrangements at the outer bulk head withexception of the curtain

30

PREPARATIVE TESTS AND MEASUREMENTS

5.1 GENERAL

The flow measurements for determining the axial hyd-raulic conductivity of the rock between the curtainswere preceded by several tests and measurements. Theinflow into the drift before and after drilling thecurtains as well as the inflow into the curtains weremeasured. Every 5th borehole in the curtains were"Lugeon"-tested and 2 holes were very carefullytested with different methods. Borehole inspectionwith fracture mapping was made in about half thenumber of holes, and rock stress measurements werealso performed. Finally, additional water pressuregauges were installed in the rock and readings taken.

5.2 INFLOW MEASUREMENTS

Any measurements of the total inflow into the BMTdrift with the accuracy of a ventilation test werenot made. At the Macropermeability Test the inflowinto the entire drift, which was about 3 times longerthan the BMT drift, was 72 1/d.

Before drilling of borehole curtains

The total inflow of water into the BMT drift wasmeasured during the Buffer Mass Test and found to bebetween 20 and 35 1/d. 20 1/d was measured in thedrift and 15 1/d in holes Nos. 1 and 2. The influenceof the increased temperature during the test may haveaffected the inflow somewhat.

After drilling of borehole curtain

Each hole in the inner borehole curtain was equippedwith a small packer before the inner wall was built,which made it possible to determine the inflow intoeach hole, and this was also made in the outer holeswith the exception of the holes in the floor. Fig 5-1and Fig 5-2 show typical examples of inflow diagramsfor the KA and KB-holes. The figures show that in theinner KA-holes the inflow was very irregular withhardly no inflow into the western wall and a veryhigh inflow into three holes in the eastern wall. Inthe outer KB-holes the situation was the opposite,with a high total inflow into 4 holes in the westernwall and very little into the eastern wall.

31

Figure 5-1 Measured inflow into the inner KA hole

Figure 5-2 Measured inflow into the outer KB hole

32

The KB-boles in the floor were not individuallytreasured. Instead the total inflow was measured bycollecting the water in the pump well and in the BMTheater hole No 3, which is located a few meters fromthe bulk head. Since hole No 3 was very dry withhardly any recorded inflow at the measurements duringBMT, only the water overflow from the curtain holesand the pump well would flow into hole No 3. Theinflow into the outer slot was not separated fromthat into the boreholes in the floor

Figs 5-1 represent the conditions before the innerslot was sealed and the inner wall constructed. Afterthe construction it was only possible to measure thetotal inflow into 4 sections of the boreholes andfour sections of the slot (See chapter 6). The inflowinto these 8 sections was continuously recorded beforeos well as after application of the lining. Fig 5-3shows the average measured inflow into the inner andouter curtain before the start of the systematic flowtesting. The figures in brackets are the average valuesmeasured after application of the lining. A slightincrease in inflow especially in the floor was obser-ved. Table 5-1 summarizes the measurements.

Table 5-1 Measured inflow into curtains before test start

Occasion Inflow (1/d)Inner curtain Outer curtain

Before liningAfter lining

6062

2429

The inflow into the drift was not measured untilafter the application of the lining. In order to findout if the lining was tight, a simple inflow measure-ment was made by use of an air-drying unit connectedto a condense collector. In this way the approximateinflow could be measured. It was found that the in-flow was around 10 1/d in spite of the lining.

5.3 LUGEON TESTING OF CURTAIN BOREHOLES

The hydraulic conductivity of each 5th curtain holewas determined by "Lugeon" testing. The test procedurewas the following:

A packer with 0.5 m length was placed as close to therock surface as possible in the about 6.5 m longhole. The hole was filled with water and the pressurein the hole recorded after 15 minutes equilibrationtime. Then, an overpressure of 200 kPa was applied

MEASURED INFLOW INTO

OUTER SCREEN I = 2A l /d(29)

Figure 5-3

MEASURED INFLOW INTOINNER SCREEN I =60l/d

(62)

Measured inflow into the different sectors of theborehole curtains and slots before application ofthe lining and (within brackets) after.

34

for 5 minutes or until the pressure had stabilizedand the flow recorded. After the test, the inflowvalve was closed and the decrease in pressure recordedfor 5 minutes.

In order to avoid influence from the surrounding,closely located holes (0.2-0.9 m), two holes on eachside of the measured hole were sealed by 6.5 m longpackers. The measurements were made by Swedish Geolo-gical AB (SGAB).

Hence,the entire holes were at first measured and theevaluation made by use of the standard formula

where

Q= flow of water (m3/s)D= diameter of the hole (m)L= axial length of the measured zone (m)Ap= applied pressure over the ambient pressure (mwh)

If the hydraulic conductivity was higher than k=10"10

m/s, the hole was measured in three equally longsections, using double packers. Thus the permeableholes were studied more in detail. The lowest detect-able hydraulic conductivity of the full holes wasabout 6-10"1 m/s.

The results are summarized in Figs 5-4 and 5-5, whichshow the evaluated hydraulic conductivity of the fullholes. The figrres show that the hydraulic conducti-vity was quite low in most holes. In the inner cur-tain only 3 holes in the left wall yielded k>10~10 m/sand in the outer curtain 4 holes (two in each wall)gave k>10"1 m/s. Probably two highly permeable holeswere missed, i.e. those with the very high inflow(Figs 5-1 and 5-2). Still, the measured holes seem tobe fairly representative of the entire curtains sincethe average inflow into the KA holes was 0.019 1/h in65 holes, which was equal to the average inflow inthe measured holes. The average inflow into holes KBwas 0.014 1/h of 48 holes while the average inflowinto the measured holes was somewhat smaller, i.e.0.011 1/h.

Detailed measurements of the holes with k>10'10 m/sare reported in Table 5-2.

35

Table 5-2 Measured hydraulic conductivity in three sections ofcertain holes

Hole NoInner

k (m/s) for sectionMiddle Outer

KAKAKAKBKB

23025027575225

7.0-101.9101-7-104.8101.610

KB 275 9.4-10

-9-10-10-10

-10I

-10

3.2103.6-101.7-102.0-101.7-102 . 9 1 0

-9

-10

r1 0-10-10

i

-10

2.1.2.

1010

3-101.7-102.9-102.5-10

-10-10

I

-10-10

rio-10

INFLOW (l/h)

HYDRAULIC CONDUCTIVITY (m/s)

INNER SCREEN

q = 0.006-10

k

36

INFLOW (l/h)

HYDRAULIC CONDUCTIVITY (m/s)OUTER SCREEN

-10k

37

5.4 DETAILED MEASUREMENTS IN 2 BOREHOLES

The two inner holes KA O and KA 75 have been investi-gated in detail for the following purposes:

- To compare different methods of determining thehydraulic conductivity in low permeable rock

- To investigate if the simple water injection test("Lugeon" test) is accurate enough

- To investigate the connectivity between some of theholes in the curtain

Three different techniques were tested:

- Water injection at constant pressure- Fall-off test- Pressure pulse test

The tests were performed over a test section with thelength 54 cm, surrounded by two packers, the equipmentbeing shown in Figs 5-6 and 5-7. The two packers were

n.-rJf I

JM: IOOOOÖÖQOOOI(OS

Figure 5-6 Equipment for detailed measuring of the hydraulicconductivity in holes KA 0 and KA 75. The pump wasonly used for filling the hole with water. The GDSconnected to the computer was used for applying andmeasuring the pressure and the water flow in thetest zone while the digital pressure meters wereused for pressure measuring in the surroundingholes.

38

Figure 5-7 Detailed description of the test probe

inflated by pressurized water. The test section wasconnected to a so called GDS apparatur which is devicefor very accurate pressurization and volume changemonitoring with an accuracy of + 1 kPa and + 0.001 ml,respectively. The inner section i.e. between the innerpacker and the end of the hole and the outer sectionbetween the outer packer and the mechanical packerwere connected to digital pressure meters by steeltubes. All sections were also equipped with deairingtubes.

The 8 surrounding holes were also packed off andconnected to pressure gauges as shown in Fig 5-8.These holes and the test hole were filled with waterdeaired and closed. The tests were started when the

39

IQOOOOOOOOOl

PANEL Of DIGITAL WESi

MFTERSWMf> W«TFS SUfPLY

Figure 5-8 Arrangement for pressure measuring in thesurrounding holes.

pressure in the sections and the occessions holes hadreached a stable value. This pressure varied from 508kPa at 2.2 m from the surface to 715 kPa at 6.9 m fromthe surface in hole KA 0. In hole KA 75 it varied from741 kPa at 2.4 m from the surface to about 900 in theinner part, i.e. 6-7 m from the surface.

A detailed description of the applied technique fortesting and evaluationand the results obtained aredescribed in a separate report by Skantz & Åstedt(1990), the results being compiled in Figs 5-9 and5-10.

Fig 5-9 shows the hydraulic conductivity plotted as afunction of the distance from the rock surface. Thefigure shows that the agreement in good between thedifferent test methods except for the pressure-pulsetest. Values evaluated from the early part of thecurve were considerably lower than those obtained bythe methods in most sections.

Fig 5-10 shows the same trend. The pressure-pulsetests, evaluated in an early stage, seem to disagreefrom the other results especially at low conductivi-ties.

C.M0MKM1 _

40

A - Pressure-pulse, early stage of the curveD - Pressure-pulse, late stage of the curve• - Pressure-pulse, second measuringO - Injection at p=l«0 kPa- - Injection at p=280 kPa

2 ! 5

Figure 5-9 Hydraulic conductivity as a function of depth fromthe rock surface measured in hole KA 0.

0.00000001 -r

o.soaaaco:; —

B

!i J

A - Pressure-pulse, early stage of the curve• - Pressure-pulse, late stage of the curve• - Pressure-pulse, second meassuringO - Injection at p=140 kPaA - Fall-off

Figure 5-10 Hydraulic conductivity as a function of the depthfrom the rock surface measured in Hole KA 75.

41

The average hydraulic conductivity of the rocksurrounding the entire holes is peferably evaluated inthe following way. The geometric average of allresults is calculated for each level with the excep-tion of those evaluated from the first part of thepressure-pulse tests. These averages are then used tocalculate the arithmetic average for the entire hole.The reason for using the geometric average at eachlevel is to supress the influence from erratic highvalues, while the reason for using the arithmeticaverage for the entire hole is that the total flow isthe sum of the individual flow conditions (see chapter11.5).

The resulting hydraulic conductivity according to thismethodology is:

fc=2.8 10~nm/s for hole KA O and

k=1.9 10'lom/s for hole KA 75.

These values cannot be directly compared with themeasurements made over the entire hole length by SGABsince the measuring accuracy was too low at thosetests, which gave Jc

42

the water injection test is preferable at short holetesting due to the simple handling, less sensitivityto disturbances like air pockets and equipment pro-perties, and because of the short required testingtime

5.5 WATER PRESSURE MEASUREMENTS

5.5.1 Location and description of the equipment

The water pressure in the rock around the test sitewas measured at 45 spots by 3 different systems. Theoriginal number of 6 Glötzl piezometers was increasedto 13. Since the two old ones located below the depositionholes did not operate, the total number of usefulGlötzl cells was 17, the location of which is shown inFig 5-11. The cells are pneumatic and the accuracyabout +10 kPa. The new gauges are of the same type asthe old ones and the same operating equipment was usedas in the BMT test.

2

Figure 5-11 Location and numbering of the Glötzl piezometers

43

The BAT piezometers installed in the BMT project werereactivated. Fig 5-12 shows the location and numberingof all these gauges, the operation of which beingdescribed in Stripa Project Technical Report 85-11.

The R and HG holes from the ventilation test wereplugged off over 4-5 m from the rock surface afterremoval of the old packers. The pressure in the openpart of the holes was recorded by precision mano-meters. Fig 5-13 shows the location and numbering ofthe holes.

A-A

Figure 5-12 Location and numbering of the BAT piezometers

44

30 m 30m 30m 30m 30m

IIFigure 5-13

»H61

Location and numbering of the R and HG holes inthe inner part of the BMT drift.

The water pressure (or rather slurry pressure) insidethe drift was measured by 6 GlÖtzl earth pressurecells applied at the rock surface in the BMT. Thecells were used to check that the slurry pressure wasuniformly distributed. The cells are numbered accor-ding to Fig 5-14.

All measured piezometric data in this report refer tothe pressure at the measuring level.

45

A-A

A

B

B-B

21,22 ( 23

\ \

B

Figure 5-14 Location and numbering of the Glötzl earthpressure cells.

5.5.2 Water pressure measured before the test

The water pressure was measured during several weeksbefore the test start. The following fairly stablevalues were recorded (Table 5-3).

46

Table 5-3 W?ter pressure measured in the rock before test startin 1990

BATNo

BAT5BAT6BAT8BAT9BAT12BAT 13BAT 14BAT 16BAT17BAT18

PkPa(26/-

2683

1617470104413

GlötzlNo

GL1GL2GL3GL4CL5GL6GL7GL8GL9GL10GL11GL12GL13GL14GL15GL16GL17

PkPa(2/3)

1435

328147556

121181

84026165

6985

185

LBLNo

RlR2R3R4R5HG2HG3HG4HG5

PkPa(12/4)

12201220830410360

1060128011601240

The pressure© in the shallow BAT holes were very lowand only two reached about 50 kPa (BAT12 and BAT17),nor did the Glötzl piezometers give high values exceptfor the 3 m deep, freshly installed horizontal gaugesin the walls (GL3, GL8, GL13 and GL17) which togetherwith gauge GL9, gave pressures higher than 100 kPa.Most LBL holes showed considerably higher pressures.

5.6 FRACTURE MAPPING OF BOREHOLES

Careful fracture mapping was made by inspection ofabout 50% of the holes. The distribution of water-bearing fractures is given in Table 5-4.

Table 5-4 Distribution of water-bearing fractures in the innerborehole curtain

Number of fractures Fraction of holes, %

0123456

2525255

1055

47

5.7 ROCK STRESS MEASUREMENTS

Rock stresses in the vicinity of the test site havebeen measured at two locations as shown in Fig 5-15 byuse of overcovering technique. The results of themeasurements made by JAA (Carr, 1988) in 5 spots in a10 m long horizontal hole in the BMT drift and fromthe measurements made by Luft (Carlsson, 1988) in 19positions in a 20 m long horizontal hole in the KBSdrift formed the basis of the evaluation of the prin-cipal stresses. They were concluded to be oriented inthe following fashion:

- Major principal stress

48

The average measured stresses are shown in Table 5-5.

Table 5-5 Average measured principal stresses in the vicinity ofthe test site

Stress type Measured stress (MPa)KBS drift BMT drift

o-j 20 13

a2 1 0 6

°3 4 5

49

ARRANGEMENTS FOR PRESSURIZING AND FLOW-MEASURING INTHE INNER SLOT AND BOREHOLE CURTAIN

6.1 GENERAL

The original philosophy of the flow-measuring in theinner curtain was only to distinguish between the flowin the borehole screen and the slot. In order to tryto separate between the flow in the roof, walls andfloor it was decided, in the course of the planning,to divide the slot and the borehole curtain into foursections. The principle was to apply the same pressurein all 8 sections, for avoiding flow between thesections. It was decided to always keep the pressurein the curtain lower than the internal pressure in thedrift, in order to prevent water from penetrating thebentonite slurry in the drift. It was decided to keepthe pressure higher within drift than on the outsidefor avoiding destruction of the lining and foravoiding water from flowing into the drift.

In this chapter we will describe the arrangements forcollecting discharged water in the boreholes, thearrangements for measuring the flow to or from the 8sections, and the steps taken to pressurize the 8sections with a constant pressure.

6.2 SLOT AND BOREHOLE ARRANGEMENTS

Fig 6-1 shows the arrangements in the inner slot. Thefour sections were separated by concrete walls and thefloor filled with coarse sand with a thin layer ofmineral wool on top of the sand. The other slot sec-tions were filled with only mineral wool. Each slotsection was water filled through a tube leading to lowspot and deaired through a tube leading to the highestspot in the section. These tubes were used during thetests to pressurize the slot sections. The entire slotwas separated from the drift by the concrete wall.

The boreholes were connected to each other by 1/2"nylon tubes passing through a mechanical packer placedat the outer end of each hole. Two tubes led into eachhole, the inflow (filling) tube ending in the lowerpart of the hole, and the outflow (deairing) tubeending in the upper part (Fig 6-2). The filling proce-dure for each borehole section started by filling thefirst hole. When the water level reached the top ofthe hole the water began to fill the deairing tubewhich in its turn led to the next hole which then

50

Deairing of slot

Deairinq of slot

Fill ing of slot

Filling of slot

Filling of slot

Deainnq of slot

Mineral wool

Filling of slot

Deairinq of slot

Coarse sand

Figure 6-1 Arrangements in the inner slot

XL

Figure 6-2 Outline diagram for the arrangement at the fillingand pressurization of the borehole curtain

51

became "filling tube". In this way all holes in asection could be filled and deaired through two tubes.The pressurization was made using the inflow tube andit was checked in the outflow tube to ensure that allholes were completely water filled and equally pressu-rized.

Fig 6-3 shows two photos taken during the installationof the tubes before casting of the concrete wall. TheT-shaped tube-fittings leading into the holes throughthe packers can be seen as well as the tubes connec-ting the deairing part of one tube-fitting with thefilling part of the next tube-fitting.

6.3 ARRANGEMENTS FOR PRESSURIZATION AND FLOW MEASUREMENTSIN THE INNER SLOT AND BOREHOLE SECTIONS

Altogether 16 tubes led to the inner sections 8 ofthem for inflow and 8 of them for outflow of water.The outflow valves were closed during the test but theconnectivity was sometimes checked, especially afterincidents. The inflow tubes were connected to a recor-ding and pressurising system as shown in Fig 6-4.

The water for pressurizing the drift and inner curtainwas taken from a supply in the Tunnel Plug Drift whichwas kept filled with mine water (1). The water waspumped and pressurized by the three pumps used in theTunnel Plug Test (2). The water was transferredto thetest site by a regulating valve (NAF Micro PacMasoneilan; serie 29000) by which the pressure insidethe drift was controlled (3). Two tubes led from thisvalve, one into the drift and the other to a valvecontrolling the pressure in the inner curtain andborehole sections (5). In this way the pressure in theinner slot could never be higher than the pressureinside the drift. The drift and inner curtain werefilled with water from the large mine water supply atthe 260 m level. The valves were closed after thefilling operations (4).

The tube to the drift was connected to an electro-magnetic flowmeter (Fischer & Porter; Mini-mag) bywhich the inflow into the drift could be measured withan accuracy of + 1% in the flow range 0.07-7 1/min(6). The pressure was measured with a pressure gauge(NAF ETP-12-3241)(7) connected to the regulatingvalve 3.

The tube leading to the slot and boreholes was connec-ted to a Schlumberger type Epsy MJ regulating valve(5): The flow into these sections were measured sepa-rately by use of an electromagnetic flow meter of thetype Fischer and Porter Copa-XM (8) and the pressure

52

Figure 6-3 Photos of the arrangement for filling andpressurizing the inner curtain holes. The lowerpicture shows the end wall before casting with thereinforcement skeleton and the tubes leading intothe holes. The upper right picture shows a close-upof the T-shaped tube fittings in the slot

53

t t f f f t r

BOREHOLESECTORS

SLOTL SECTORS

DRIFT

(16)

Figure 6-4 Circuit diagram of the arrangements for pressuriza-tion and flow measurements in the inner curtain. Thefigures are referred to in the text

was measured by a pressure gauge Satt Control ETP-04(9) connected to the regulating valve 5. The lowestmeasurable flow with this flow meter was 1,2 1/h. Theborehole tube was split into 4 parallel tubes leadingto the respective borehole section (14). The flow intoor from each section was measured with a flowmeter oftype Brooks 1355 M-BR (11) with a measuring range of0.01-3.2 1/h. An identical system was used for theslot sections (12). The sections were filled throughsepararate tubes, each equipped with a valve (10) (13) .

Fig 6-5 shows the detailed flow recording systems (11and 12 in Fig 6-4). The flowmeter (1) could be connec-ted in two directions through a 4-tube valve (2). Forrecording, valve 4 was closed and valves 3 opened. The4-tube valve was turned such that the flow through theflow meter was always positive. This was necessarysince the flow meter could only measure one-way flow.

54

Figure 6-5 Circuit diagram of the arrangements for the flowmeasurements (11 and 12 in Fig 6-4)

1. Flow meter2. 4-tube valve3. and 4. Valves

Fig 6-6 shows photos of the regulating and measuringsystem. The upper picture is taken from the inside ofthe drift with the camera pointing outwards. TheBrooks flowtubes are seen in the foreground. The lowerpicture is taken in the other direction and it showsthe electrical pressure regulating and flow measuringsystem (3-9 in Fig 6-4) in the foreground.

The recording tools hence made it possible to measurethe flow into the drift with a minimum flow rate of4.2 1/h, while the flow into each of the innersections could be measured with a minimum flow rate of0.01 1/h. The sum of the flow into the boreholesections and the slot sections was checked with aminimum flow rate of 1.2 1/h.

Practically all tubes outside the drift were 19 mm(3/4") water pipes except for the arrangements withthe Brooks flow meters which were connected to 6 mmnylon tubes. All tubes inside the drift were 13 mmnylon tubes.

55

Figure 6-6 Photos of the pressure regulating and flowmeasuring arrangements

56

ARRANGEMENTS TO PRESSURIZE AND SURFACE SEAL THE DRIFT

7.1 GENERAL

It was clear at the planning stage that the exchangeof water between the pressurized drift and the testedrock had to be low enough not to disturb the measure-ments in the rock and the original idea was to apply awater-tight lining on the rock surface, to fill thedrift with water, and to pressurize the water. Itturned out, however, that it was impossible to producea lining that was sufficiently water-tight and analternative technique had to be developed. The diffe-rent lining tests and the developement of the alter-native technique will be described in this chapter.

7.2 LINING TECHNIQUE

7.2.1 Pre-testing of different lining techniques

7.2.1.1 General

Before the lining "was applied, a number of liningmaterials and techniques were investigated. The aimwas to find a material that fulfilled the followingcriteria.

1. The lining material must attach to a damp rocksurface.

2. The adhesion of the lining to the rock surfaceshould be strong enough to withstand a water pressurein the rock of 500 kPa when there is no water in thedrift.

3. The lining material must be sufficiently elastic toallow for rock strain of a few millimeters.

4. The lining material must be chemically stable forat least two years.

5. The unavoidable damage of the lining caused by thedrilling of holes for the hedgehog grouting, subse-quent to the flow tests, must be repairable.

The most promising materials were then subjected tofield tests at the Stripa Mine. The tests were per-formed on a 760 mm granite core, into which a 55 mmhole for water pressurizing, was drilled. The waterwas connected through a packer, making it possible to

57

produce a wet surface during the application of thelining. The system was also used for adhesiontests after the hardening process.

7.2.1.2 Aquata Epoxy

Aquata is an epoxy material that can be applied on awet surface and stick to it after hardening. Tests onthe 760 mm core confirmed that the adhesion to therock was good, and that the material was water tight.However, the hardened epoxy was brittle and rock move-ments were assumed to cause cracks in the lining. Thisrisk and the high cost disqualified the material.

7.2.1.3 Epoxy-Polyurethane

This lining material was suggested by Åke Calminder,Calminder Development Limited. It consists of an epoxyprimer which ensures good adhesion to the rock sur-face, and a sealing polyurethane layer. The poly-urethane was applied by spraying in two steps to athickness of 2.0 mm.

Two tests were made on a 760 mm core. In both casesthe polyurethane layer separated from the epoxyprimer. The failure was probably caused by mistakes atthe application but there was no time to wait for athird test. The relatively low elasticity, leading tobreakage at an elongation of about 400%, was alsoconsidered to be a drawback.

7.2.1.4 Meycopren

Meycopren consists of 80% bitumen and 20% syntheticlatex in a water emulsion. The emulsion is sprayedtogether with a nitrite solution, which acts as acatalyst during the gelling of the lining material.

The Meycopren was sprayed both on a 760 mm core, andon a test area in the mine. The material showed verygood elastic properties, but the adhesion to the rocksurface was bad. On the test area, small water filledbubbles were formed from the wet areas of the rock,indicating that the material was too weak to preventthe groundwater from penetrating the lining and thislining material was therefore not considered further.

7.2.1.5 Procoat

Procoat is a mix of bitumen as major component and apolyurethane base. No solvents are used. An epoxyprimer is used to ensure a good adhesion to the rock.Procoat should be applied by spraying to a thicknessof 3 to 5 millimeters.

58

The Procoat was sprayed both on a 760 mm core and on atest area in the mine. Both the elastic properties andadhesion to the rock surface were good and Procoatseemed to be the only tested material that fulfilledall the criteria mentioned above. Later it showed thatthe test area may probably not has been altogetherrepresentative of the conditions in the BMT driftsince the water inflow was considerably lower in thetest area than in the drift.

7.2.2 Application of Procoat lining

The rock surface in the drift was first quartz-blasted, cleaned and dried with an air dryer in orderto prepare a suitable surface for adhesion of thelining. The most water-leaking fractures were coveredwith rubber sheets from tires with tubes attached tothe valves for discharging water without causing sig-nificant water pressure at the application of thelining.

The rock surface was then coated with epoxy primer andthe Procoat sprayed on the primer in two layers to athickness of 3-5 mm. Fig 7-1 shows a picture takenduring the spraying.

Figure 7-1 Picture taken during the spraying operation when thelining is applied in the BMT drift

59

After finishing the application, water leakage wasobserved at »ore than a hundred spots. No offorts weresaved trying to repair these spots a last chance beingto apply rubber sheets that were glued to the liningover the leakage areas, and to spray an additionallayer of rubber on the lining and the sheets. Stillthe leakage into the drift could not be decreased toless than 5-10 1/d.

Two attemps of pressurizing the drift were made withthe expectation that back-valve effects Bight takeplace. At the first attempt on March 14, 1990 thedrift was pressurized to 220 kPa. The leakage wasseveral tens of liters per minute and significant out-flow of water was observed at the interface of thebulkhead and the roof. After the first attempt thewater in the drift was pumped out, the repair workcontinued and an extra layer of lining finallyapplied. A second pressurizing was made on Nay 2, 1990but only an insignificant reduction in leakage hadtaken place.

It vas concluded that this type of lining could nevercreate a surface seal that vas sufficiently vater-tight. An alternative technique therefore had to bedeveloped.

7.3 TECHNIQUE WITH A BLADDER CONBINED WITH A BENTONITESLURRY

7.3.1 General

When the lining failed to isolate the inside of thedrift from the surrounding rock it was decided to usethe back-up technique that was part of the technicalplanning of the project. This technique takes advan-tage of the sealing properties of bentonite. With aproperly composed mixture of water and bentonite, i.e.with a relatively high viscosity of the slurry in com-bination with its pronounced thixotropic propertiesthe slurry will be soft enough to be pumped into thedrift. On site, it will stiffen so that it can bepressurized without penetrating significantly into therock. Emptying the drift after the test will partly bepossible since pumping turns the mass into a morefluid state.

Since the volume of the drift was very large (>200 m3)and since it was not quite clear that the slurry wouldact as a liquid and produce the same pressure every-where, it was decided to install a rubber bladder inthe center of the drift. By filling the bladder withwater and pressurizing the water it would be possible

60

to reduce the slurry volume to about 100 m . An evenlydistributed pressure on the rock surface would also beachieved by such an arrangement.

A big advantage of using slurry was also that allthrough-connections for the cables and tubes goingfrom measuring pointr -n the rock into the drift andthrough the bulk head, would automatically be sealed.However, it was clear that all the sharp edges ofprotruding rock blocks and instruments would endangerthe tightness of the bladder and a large wooden cageequipped with deformable steel net had to be arrangedin the drift for preventing the bladder from approach-ing the rock surface (Fig 7.2).

7.3.2 Bladder

The rubber bladder was given a volume of about 100 m .It was made from a 1.5 mm thick film vulcanized toform a completely water-tight bladder. The rubber wasof the type Butyl, special quality 72704 manufacturedby Värnamo Isolerduk AB, Sweden.

The bladder had two inlets in the bottom for waterfilling and pressurization. It also had some inlets inthe upper part for air filling and deairing. Fig 7-3shows the bladder inside the cage when it was filledwith air to about 50%.

7.3.3 Cage

The cage consisted of a wooden framework and a steelnet fixed to the inside. The net was chosen to have amesh aperture that was sufficient for allotting thebladder to penetrate a few cm, which would account fora volume increase of the bladder caused by some minorslurry leakage or other unforeseen events.

Fig 7-4 shows the cage before the installation of thebladder. As can be seen from the photos, the insideof all beams to which the net was nailed was coveredwith a smooth plank in order not to damage thebladder. The bottom of the cage consisted of a bed oftimber. The flexibility of the cage in combinationwith the possibilty for the bladder to expand throughthe net, allowed for an expansion of the bladder byseveral m .

6 1

WATER FILLED

RUBBER BLADDER

Figure 7-2 Principle drawing of the water filled bladder, thecage, and the bentonite slurry

62

Figure 7-3 Picture of the bladder inside the cage. The bladderis partly filled with air

7.3.4 Slurry

The main idea was to use a slurry with a shearstrength high enough to yield penetration of theslurry o f0.3 m at maximum into a slot with the aper-ture of about 0.1 mm, or maximum penetration of 0.5meters into a slot of about 0.3 mm aperture. Back-calculations of the water leakage at the water pressu-rizing had indicated that the largest apertures werein that range (0.1-0.3 mm).

63

Figure 7-4 The cage photographed from the inside

64

The Bentonite of type Geko Q/I was the major clay sub-stance used for the preparation of the slurry, has thefollowing (cone) liquid limit:

I/L=255% if the test is made < 1 day after mixture

I/L=290% if the test is made > 2 days after mixture

If the second value is applied the measured shearstrength of a stirred slurry 2-3 days after mixing hasthe X-W/VL relation shown in Fig 7-5. Since the pene-tration (L) into a fracture can be estimated by usingthe simple expression

7:1

where p= slurry pressurea= slot apertureT= Shear strength

the required shear strength and thus also the requiredwater ratio can be estimated. The values shown inTable 7-1 give L as a function of the water ratio andslot aperture.

Material: Ceko Q,bOO

COCOOJ

en

ucuc

200

01.2 1.4 .6 .8 2

w/w

Figure 7-5 Influence of the relation between the water ratio vand the liquid limit w\. on the shear strength of thebentonite slurry

65

Table 7-1 Required Geko Q/I bentonite mixture for sealing slotsif the pressure p is 1000 kPa and the slot length L is0.5 m

amm

0.050.10.20.30.5

T

Pa

50100200300500

V/VL

2.01.91.71.51.2

V%

580550490440350

The table clearly shows that ratio V/VL of 1.4 to 1.9,corresponding to a water ratio of between 400 and 550,would be sufficient for resisting 1 MPa slurry pres-sure if the openings in the lining and rock have anaperture between 0.1 and 0.4 mm and a depth of about0.5 m.

In the real test the thixotropic properties of theslurry should be very pronounced in the sense that theshear strength should be at maximum after a period ofrest. Thus, in case of leakage of slurry, there wouldbe a possibility to decrease the pressure, wait forabout a week, and subsequently increase the pressureagain since the tixotropic properties would thenincrease the shear strength of the slurry by several100%.

As a result of these considerations it was determinedto use a slurry of Geko Q/I with a water ratio ofabout 500% in the whole drift except for in the upper10 m3 where 400% was chosen. The reason for using alower water ratio in the upper part was that the riskof leakage was judged to be higher at the interfacebetween the bulkhead and the roof. The reason for notusing v=400% in the entire drift was a fear of havinga too stiff slurry at the emptying of the drift.

7.4 SLURRY FILLING PROCEDURES

Preparations

The field work started with constructing the cage,after which the rubber bladder was placed inside.Problems with having folds in the bladder during thewater filling procedure were avoided by filling itwith air under low pressure. Fig 7-6 shows the bladderin the cage during the air-filling process. The highfriction of rubber to rubber in the folds made thefilling quite tricky.

66

Figure 7-6 Air filling of the bladder prior to the slurryfilling

Slurry filling

The filling of slurry in the drift and the filling ofwater in the bladder then started and continued simul-taneously. The idea was to maintain a water level inthe bladder that was one or two dm above the slurrylevel since the two liguids would then be in balance.The volumes were measured in order to keep control ofthe filling process after closing the manhole in thebulkhead, i.e. after having filled up about 50% of thespace.

In order to avoid air pockets at the roof, about 10deairing tubes were fixed to the roof and when theslurry finally appeared in these tubes, the valves attheir ends were closed. A good proof of the sealingability of the bentonite slurry was obtained when itturned out that several hundred kPa bladder pressurewas 'needed to force the slurry through the tubes.

67

When the drift was filled about 5 m extra slurry waspressed into the drift and the corresponding amount ofwater let out of the bladder. This was made in orderto further increase the expandability of the bladderin case of leakage. At the end, a total amount of 126JO3 slurry and 94 m3 water had been filled into thedrift.

The bentonite was delivered in 1000 kg big-bags andmanually poured into a colloid mixer after filling itwith water. After mixing, the slurry was pumped intothe drift by two high capacity pumps, leading to bothends of the drift through tubes passing through theupper end of the bulkhead.

Fig 7-7 shows a big-bag and the mixer as well as oneof the pumps. Before the door was closed, the slurryin the outer part of the drift was pumped through themanhole.

Fig 7-8 shows a photo from inside the drift when theslurry level was about one meter above the floor. Thethrough-connections of some of the cables and tubes inthe bulk head are seen in the photo as well as thecorner of the cage. The consistency of the slurry isindicated by the furrows in the surface.

Pressurization

Fig 7-9 shows a phenomenon associated with a firstpressurization of the slurry made by the pumpsdirectly after closing all deairing valves. Leakage ofsmall amount of slurry was observed at the interfacebetween the bulk head and the roof, but it soonstopped and no further slurry leakage took placeduring all the flow tests performed before the grou-ting. It was obvious that the technique worked welland that the bentonite efficiently sealed the leaks.Fig 7-9 also shows the inlet pipes for the slurry andthe tubes leading from some of the holes in the outerborehole curtain to the water collecters.

The manhole caused some problems during the tests. Thesealing construction turned out to be incorrectlydesigned and it leaked at a few instances. This causedsudden loss of about 1/2 m3 slurry, which causedexpansion of the bladder such that one of the woodenbeams close to the door was broken. The bladder wasnot damaged by this, but in order to prevent furtherincidents the manhole was sealed by welding.

68

Figure 7-7 Photos taken during the mixing and pumping of slurry

69

Figure 7-8 Photo taken inside the drift when the slurry isfilled to the level 1 m above the floor

Figure 7-9 Slurry leaking through the bulkhead close to theroof at the initial pressurization. The leakagestopped after a short while and no more leakageoccurred during the entire test

70

7.5 EMPTYING PROCEDURES

When the measuring program was finished, the slurryhad to be pumped out and the cage and bladder takenaway to give room for the hedge-hog drilling andgrouting. In order to save work and material a basinwas buijt in the Extensometer Drift to which theslurry wes pumped. About 80 m3 slurry could be savedin this way.

The bottom part of the slurry (about 40 m j was toostiff to be pumped due to the tixotropy and probablyalso some sedimentation and consolidation effects. Ithad to be softened by water mixing, vibrating andmanual digging, before pumping it, which caused muchproblems and considerable delay. The bladder had to becut in pieces and a new one ordered for use in theflow tests after grouting.

71

8 PREPARATIVE FLOW MODELLING AND FLOW CALCULATIONS

8.1 GENERAL

A good understanding of the relevance and function ofthe test has been achieved by modelling the flowsituation in the different stages of the test. Sincethe actual flow in rock is very complicated and takesplace in rather unknown fracture- and channel-systemsthe Equivalent Porous Media (EPM) model for flow hasbeen applied.

Beside modelling the entire flow test the followingtwo studies of detailed flow were made:

- The relevance of using curtains of boreholes forsimulating slots was investigated by an EPM model ofthe flow between two such curtains.

- The relevance of the hydraulic conductivitymeasured by the Lugeon test in the short hedgehogholes were also studied by an EPM model.

All the calculations accounted for in this chapterwere made with the finite element program GEOFEM-Gfrom the CHALMFEM package at Göteborgs Data Central,Sweden.

8.2 FEM CALCULATIONS OF THE FLOW TEST BEFORE GROUTING

8.2.1 Element model and basic data

Basically, the same element mesh has been used in allflow calculations. The mesh and the different zonesare shown in Fig 8-1. The mesh is rotational sym-metric around the drift axis and the boundaries arelocated 31 m from the rock surface. Fig 8-1 shows thefinal situation after grouting. The calculations weremade in several steps and the properties of thedisturbed and sealed zones, the curtain holes K, andthe inner part of the BMT shaft were changed but theboundary conditions were the same in all the prepara-tive calculations.

Right and lower boundary: Constant water pressurep=1.5 MPa

Left boundary: Isolated (no flow across)

72

SHAFTBULKHEAD BMT SHAFT

mizaz m aȣ5wwww

\ \ \ \

Disturbed zone (by blasting) A

Sealed zone A

Disturbed zone (by stress release) B

Sealed zone B

Screen of holes K

Figure 8-1 Element mesh and zones with different hydraulicproperties. The mesh is axi-symmetric around theupper boundary

The following basic hydraulic conductivities were usedin the calculations:

Undisturbed rock: k =1.0-10"10 m/s

Dist. zone by blasting (A): Kb=1.010*8 m/s

Dist. zone by stress release (B):

- axially

- radially /c"=5.010r

-11m/s

m/s

The extension of the different zones was not knownbut in these preliminary calculations, zone A wasassumed to reach 1.0 m into the rock and zone B 7.0 minto the rock. The axial extension of the zones atthe end of the drift is also unknown and was taken asshown in Fig 8-1.

A constant atmospheric pressure p=0 was assigned tothe rock surface in the drift except after surfacesealing of the BMT drift when an impermeable materialwas assumed to occupy the drift. After drilling theborehole curtains, a constant pressure of zero wasassigned to the nodes of these elements before pres-surization of the inner curtain.

73

The following stages of the test were calculated:

1. Before drilling the borehole curtain.2. After drilling the borehole curtain.3. After surface sealing of the BMT drift.4. After pressurization of the inner curtain.

The possible effects of grouting will be treated involume III of the final report. Since the flow testswere continued until equilibrium was reached all thecalculations refered to steady state conditions.However, one transient calculation was