REVUE FRANCAISE DE GEOTECHNIOUE \o 23 PARIS 1983

Gerieral report topic 3 / Rapport général thème 3

HYDRAULICAL AND HYDROGEOLOGICAL TESTING TO DETERMINE THE PERMEABILITY

OF THE PORES OR TRANSMISSIVITY OF THE FRACTURES

ESSAIS HYDRAULIQUES ET HYDROGEOLOGIQUES EN VUE DE LA DETERMINATION DE

LA PERMEABILITE bn ponBs oU DES FRACTURES DE LA TRANSMISSIVITE

MANFREDINI G.

L. Introduction

The field of investigation covered by Session 3 is verybroad ; it rcgards in situ permeability measurements in soiland rock uiing widely differing techniques according tothe different nature of the soils, the depth to investigate,and according to the particular problems of the case inquestion (fig. l).A significant aspect of permeability measurements is thatonly very rarely-are laboratory tests representative of actualfielâ conditions. It is well-know that the values obtaine d tnsitu differ by several orders of magnitude from the labora-tory results, because of the presence in natural soils of"structural features". In the case of rock massifs fissures

and joints arc present (accounting for mo-st-of the waterflow), and in the case of soils, very fine lithologig or grain-

size variations and thin layering are a characteristic trait.It is therefore indispensable to carry out in situ tests tocharac terrze the penneability. The very structural charac-

teristics make it generally necessary to consider pelmea-bility as a "tensor". [n most cases the planning and carryingout of in situ tests has to take this aspect into account.

The variety of testing equipment and methodologies vary

Papers dealing mainlY with:Soil mechanics problems ooooooooooooo

Rock mechaniCs problems oo oo oo oooooooooo

Depth in volved bY the test :

0 +-10

10+ 200

200+200 0

oooooooaoooooooooooooooooooo

Tvpe of equipment used to determinetr

permeability :

Packer test of different types o o o' o o ' oo o ooo

Pumping wells, open hole tests o o o o o o o o

Trace rs

Self boring or Pushing-in oooooinstruments

Large scale testFlowmeterGeophysical methodsFig. 1. Distribution of papers according to nature of soil, depth

involved, equiPment used.

ooaoo

o

thoroughly as a function of the depth of investigationwhich ituy U. on the order of a metre when only the per-

meability'of the superficial soil layer is to be determined(e.g. fof agriculturaÎ purposes), or of thousands of metres

wh-tn the lermeability of formations to be used as waste

disposals Ë to be known, of of_ formations from whichnuia is to be drawn, or finally, when the stability of broad

subsident areas is to be examined.

Usual investigations for civil engineering - problems rarely

exceed a Aepth of 15G200 metres ; problems related topermeability measurements ut greater depths generally lie

but side the day-today engineering experience and musttherefore be dealt with with gteat caution.permeability site investigations are performed in differenttypes of sôils and rockJ i th9 planning 1nd interpretationof these investigations, furtheimore quite often involve,

as demonstrated in the papers presented, not 94y-know-tedge in the fields of Soil Mechânics and Rock Mechanics,

but also in other domains.

An interesting aspect which is emphalqgd by the papers

presented at ihis sèssion i.s thqt,^ permeability.measurementsàrr certainly amongst the diffèrent investigations, those

whose planning andlnterpretation require a rather accurate

6i;gifal knowledge. of -the

arca under examination. [nsome cases, where" the permeability of whole geological

formations arc required- in order to correctly .interpretthe data, an exhauitive geological knowledge of the region

is imperative.

In recent years there has been much progress in the area

of in siti permeability measurementi, esPe_cially in t4.development of new testing equipment, and tlis very topic

hut beàn discussed in special seisiôns at recent lnternationalConferences. The interest aroused by the topic, and the

advancement in knowledge that has been achieved is fullywitnessed by the 28 papers presented at this Session, p.ro"

duced by i t EuropêaÀ and 5 non-European countries.

It is interesting to note that in spite of developments inthe tn situ perti.rbility measurement techli-tues, there stillare many uïcertaintiei as to the meaningfulness of measu-

rements bbtained through the different methods in relationto the specific problems under examination. Indeed, it mustbe rec.Îled thât even today it is not uncommon to find acoefficient of variation of the permeability on the order

of 200%.The papers presented at Session 3 of this Syr.npgsium T?ybe UioâOty ïivided into 5 groups of topics dealing mainlywith:In situ determinations for charac tetuing a "regional"permeability ;

Modern "in borehole" testing techniques, or meaningfulupgradin of routine techniques in the field of civil engi-

neering;

*ENEL - Italian State Electricity Board, Dept. of Construction Rome (Italy)

32

Measurements of permeability at great depths, or in relationto non-traditional problems ;Theoretical and general analyses with a view to more cor-rectly interpreting permeability tests ;Case histories, and special in situ tests.

It is evident that each paper does not strictly focus ononly one of the topics of the outline, but the latter canbe used for ease of exposition. Most of the papers howeverdeal with the second topic thus emphasizing the progressthat has at present been achieved in this iector. tn thereview, sPecial attention is paid to the comparison betweenresults of different in situ testing methods, and betweenthese and the results obtained from full-scale tn situ testsor from back-analyses. This comparison will certainly leadto a better understanding of the field tests.

2. In situ determinations for charac tefizing a"regional" permeability

Qu ite often, for large-scale engineering probleffis, it isnecessary to know the permeability of formations extend-ing out over kilometres ; a typical problem is the construc-tion of artificial reservoirs ; furthermore, perfireabilitycoupled with deformability govern the subsidence of largeareas (Miyabe, 1969 ; Poland, 1969; Croce , 1973; Colombô,1973) which is related to the lowering of the water tablecaused by the pumping of fluids from deep wells.

Large scale permeability is firoreover important for takinginto account changes in the hydrogeological conditionsbrought about by the urbannation of large areas (Louis,1e81).

In relation to "regional permeability, particular importancehas always been attache d - as regards carbonatic rotks - tokarst phenomena. The presence of this, may cause perméa-bility to rise by various orders of magnitude.

The "large-cale" charact eraation of permeability has deve-loped essentially by trying to attache (at least to a regionalextent) general validity to permeability tests carried ôut insome typical geological formations.

A recent and interesting review of many in situ permeabi-lity tests in differing types of rock formations is given byBrace ( 1980).

The different permeability values referred to vary withina range of 20 orders of magnitude, and at the same investigation depth and for the same rock formation, within therange of 4 orders of magnitude.

Besides the classical in situ tests, that in any case wouldhave to be carried out in large numbers and over extensiveareas in order to be meaningful on the regional scale , ins-itu penneability may be calculated by analysing variouslarge--scale phenomena whose developmènt depends on per-meability. It is possible to consider such phenômena as fuil-scale "in siru tests", carried out on large volumes of rockor soil, but on the other hand, this approach has the draw-back that the phenomena under investigation (delaysbetween fluid injection and seismic response, flow-raieincreases following seismic events, examination of hydro-thelqal phenomena, etc.) are generally extremely complex,and their functional relation with penneability not easyto solve. Even if the values obtained àre unlikeli to be useâfor most civil engineering works, they may prove to beuseful for characteruing the permeability of deeper soils.

An interesting application of the analyses of large-scalephenomena for solving typical civil engineering problemshas been provided by Uriel Romero and Aoiz (1981), whodeveloped a method for determining the permeability of

the terrain affected by large-scale excavation near the coast,on the basis of the damping effects with which the oscilla-tion of the tide is transmitted througn the porous media.

A recent example of regional characteraation of the Britishnon-carbonatic mudrocks in hydrogeological terms hasbeen given by Tellam and Lloyd (1981) on the basis of awide variety of in situ and laboratory tests, by utilizingmany uncollected data, or data scattered throughout thegeological, hydrogeological, soil rnechanics, mining andagricu ltural literatures.

Also the paper presented at this Session by KRAPP, whichis very interesting from a methodological standpoint, refersto the characteraation of geological formations for prcblems related to the civil engineering activity. This pàperexamines and evaluates a series of data referring to the per-meability of Pale ozaic and Mesozoic rock in the wesiernRenish mountain areas of Germany. The evaluation wasbased essentially on a critical re-examination of numerousin situ borehole permeability tests, on observations of waterinflow into tunnels and of water tables in monitored holesand wells in the area. Further information on permeabilityhas been obtained by means of tracers (especially in karsticareas), and finally, by utilizing hydrogeological balance datafollowing run-off during dry season in small catchmentareas.

For a regional characteruation of the superficial layer ofrock formations (where most of the flow occurs) thedominant factor for the permeability is the "effectiveopen crack volume".

on the basis of this parameter (which may vary from0.1 % in the case of phyllitic rocks, up to 1 a$ % forkarstic limestones and up to 50 % for travertine) a usefulclassification scheme is presented which could be used inpermeability maps.

The paper stresses that in evaluating the "regional" charac-teristic of geological formations, it is usually insuffîcientto util:r,e a single type of test, but rather, all the possibleelements available for a synthetic judgement shôuld betaken into consideration.

GANGOPADHYAY prsents a practical case where thepermeability of a large atea was to be defined for the cons-truction of a hydroelectric plant in a highly karstified area.The investigation followed the lines mentioned in the fore-going, that is: a) survey of the piezometric levels and of therelating pressure gradients; b) extensive surveys usingchemical and radioactive tracers in order to determineshort- and long- distance interrelations due to karst pheno-mena; c) hydrological balance for the different areas, Theseinvestigations positively showed karstic interconnectionsin the different areas of the plant.

A different approach to permeability charact eruation ofgeological formations is provided by Alessandrello andLemoine's paper who point out, on the basis of an exten-sive series of experimental results, the possibility of corre-lating resistivity (rneasured by means of geophysicalmethods) with permeability in the case of soms sandy-gravelly alluvial deposits in France. The need to know thepore water resistivity to have better correlations and thususe geophysical data for making forecasts is pointed out,and the influence of clayey soils on the global resistivityof the formations is also discussed. The collection of similardata for the alluvial formations of other countries couldprove to be greatly useful, thus allowing to amplify andverify the validity of such correlations for application pur-poses, and possibly to reach more general semi-empiricalrelations between the different parameters.

Technique

Permeabilitytests

Packertests(Lugeon test)

Falling headtests

Constant headtests

Pumpingtest

Principles of Technique

Water is pumped out fromor injected into a boreholein the ground

A section of a borehole isclosed with packers at topand bottom and water ispumped in under pressure ;

water losses are recorded.Alternatively a single packeris used to test the part ofthe borehole below it.

A test section in a boreholeis isolated by casing. Theborehole is filled with wa-ter and the rate of fall ofwaterlevel is observed.

A test section in a boreholeis isolated by casing. Theborehole is filled with wa-ter; water is added to theborehole to keep the waterlevel constant.

Water is pumped out of a

borehole, observations aremade of waterlevel in thehole and in nearby holes.

33

Table 1

Data Obtained

Permeability

Permeability of the investi-gated rock mass expressedin standard units of per-rneability or in lugeon units.

Permeability of the test sec-

tion.

Permeability of the test sec-tion.

Application

Ground masses of differenttypes.

Used in rock mass of rela-tively low permeability.

Comment

A precise knowledge of thenatural groundwater levelis required.

The value of the results willbe related to the frequencyand aperture of the jointsystems in relation to thesize of the borehole.

Permeability of the inves-tigated rock or soil mass.

In-situ Testing - Permeability Testing

(after "Report of the IAEG Commission on site Investigations, 1981).

3. Modern 'oin borehole" testing rechniques ormeaningful upgrading of routine techniques incivil engineering

The in situ permeability of soils and rock formations isgenerally determined, for civil engineering purposes, bythe application of a hydraulic pressure head differencebetween the water in the borehole and that in the surround-ing ground, and to measure the resulting flow. A numberof analyses and tests were developed since the last centuryfor different geometrical and hydrogeological conditions,which take into account the transient or permanent flowin order to obtain the mean permeability of a "porousmedium".

The testing techniques and the interpretation theories werelater "extended" or "adjusted" to non-granular media, suchas fractured rock masses and very low permeability soils.soils.

In the following table (after the Report of the IAEGCommission on 'Site lnvestigations', 1 981) the in situpermeability tests routinely used are outlied*.

Besides these widely used types, innovations have cometo add with the aim of obtaining indications also for low-

* Field test relatrng to permeability scheduled to be published as

"suggested Methods" by ISRM are the following (see Franklin,reT 9).

- Permeability / transmissivity (5 methods)

- Flow velocity logs

- Flow velocity - tracer dilution- Flow paths using tracem (4 methods)

Mostly in soils of low per-meability.

Mostly in rocks of mode-rate or high permeability.

Used in rock and soil mass. If sufficient observationholes are made, some infor-mation on the anisotropyof the groundmass maybe deduced.

penneability soils by running water tests in piezometers,for obtaining continuous permeability logs with depth(flowmeters, tracers, dipole probes) or for measuring localtransmissivity of fissures in rock masses (harmonic probe).

A detailed description of routine in situ permeability testsand a presentation of the different methods for inter-preting results as well as technological details involved,is found in "Ground water Manual" edited by U.S.B.R.(1977). Useful suggestions are also presented in manyrecent Site Investigation Manuals or Standards.

An extensive in situ investigation making use of most ofthe available routine methods, with some innovations, hasbeen carried out for the Ok Tedi Project (Papua NewGuinea) by Sutcliffe and Mostyn. A description is given ofthe advantages and liabilities for each type of test carriedout (constant head test, falling head test, packer test), thecosts involved are computed. The considerable advantagesin terms of time and coit offered by the wireline pneumaticpacker system against the mechanical or hydraulic packersare described.

Also the paper presented by Foyo Marcos deals withborehole permeability tests. Referring to experience gainedwith dam foundation investigations in Spain, he suggeststhat by producing hydraulic fracturing, information aboutthe mechanical characteristics of the rock mass may beinferred. tt is however known that the pressure at whichhydraulic fracturing phenomena initiate is essentially de-pendent upon the hoùzontal stresses os max and o, min,form the tensile strength of the rock mass, and from pre-existing pore pressures (Haimson, 1978 and at this Sympo-sium) ; only a more detailed analysis (with the measure of

34

the shut in pressure and of the secondary breakdownpressure) would, ideally, allow an evaluation of the tensilestrength of the rock mass.

It is often pointed out that the interpretation of packerperrneability tests may be strongly biased when some parti-cular aspects are not taken into account, namely, actualpore pressure in the surrounding soil, water losses throughthe packer, and head losses in the drop pipe ; indeed, somehead loss may occur (especially in the case of high flowrates) also at the perforated pipe section and inlet section(Moore and Balbis, 1979). Much effort has been put intoimproving the reliability of such tests; Maini et al. (1972)suggest to use a directional triple cell packer ; to use trans-ducers to measure the effective pressure in the test cell'to use two

--trstilg'ïiruÀïàrî loi source-sink tests, etc]Along the same research lines, &fl experimental programmehas been carried out by Pearson and Money (1977) with themain objective of distinguishing between anomalies due totest method faults (such as packer leakage, head loss inthe equipment, etc.) and non-equilibrium effects due tothe hydraulic properties of the rock mass, which maydepend on the duration of the test and on the sequence ofthe operations. An extensive revision of the theoretical andpractical background of this standard test has also beencarried out by the German Geological Survey and at theTechnical university of Aachen (Heitfeld and Krapp,1e81).

The paper by Blinde et al. presents useful suggestions to im-prove the reliability of the test, amongst which is therecording of the water head inside the cell and of the flowthroughout the test. To better interpret the results, opticalbore-hole soundings and dye tracings were performéd. Acornparison was made with the experimental grouting ina test area with bentonite-cement suspensions. Subsequentexcavations showed that the grout had propagated aroundthe hole between 1.50 and 2.00 m.

As in many other cases reported in the literature (Capozza,1977) geophysical methods (and in particular scori wavevelocity) proved to be appropriate for inspecting the effec-tiveness of the grouting.

The paper by Haimson focuses on the need to carry out ina coordinated manner, and virtually with one set of tools,me asurements of the state of fracturing, of the state ofstress and of permeability inside each bore-hole. Indeed,it is well-known that, even if the parameter "fracturing"is not univocally related to permeability, in principalthe latter does depend on the nature, and on the openingof the o'hy draulically active" discontinuities, that are incorrununication with each other to form a single "net-work". The state of stress greatly influences permeability(bv reducing space openings) as discussed in ôther papeispresented at this session. It should therefore be measuredwhere possible jointly with the measurement of permea-bility.

The proposed equipment consists essentially of a straddlepacker appropriately instrumented with pressure transdu-cers, by means of which hydraulic fracturing is achieved,and an "impression packer" which is used in the fracfuringzone, so as to check the real orientation of the fractureswhich develop, and the original fissuring of the rock mass.Before fracturing (or even afterwards) the same equipmentcan be used to measure the permeability according to the"Pulse test" method (that is, by applying a pressure pulseon the order of 0.5 - I MPa and by recording the decaycharacteristics of the pulse), or by the more classicalmeasurement methods such as the 'oconstant head tech-nique".

The full orientation of the entire core (in order to get a

global view of the discontinuities arrangement) is obtainedby means of a television camera or by occasional impressionpacker tests in the critical stretches.

Daw and Scott's paper examines permeability tests appro-priate for obtaining information about the possibility oflarge water inflow during the excavation of deep under-ground works: the possibility of foreseeing such accidentsmay lead to localised preliminary reclamation measures.

In these conditions the simpler tests do not provide ade-quate results especially because of the limited area of in-fluence of the test. The "drill stem test", that has been usedin the oil industry for some time now, consists essentiallyof educting water from an aquifer over a specified flowperiod by utluing the volume of the empty drill stem, andin recording the recovery of pressure following the shuttingoff of the flow. In this manner, volumes of rock massesfrom tens to hundreds of metres around the hole may beaffected. By plotting appropriate diagrams, the transmissi-vity of the rock mass is determined. [n strong aquifers atshallow depths, however, the drill stem fills up so rapidlythat the' effective flow rute during the test becomes un-certain. The solution worked out to overcome this diffi-culty includes the use of submersible pumps capable ofoperating down to depths of 200 m.

Continuous monitoring of the pressure with appropriatetransducers, the control of the effective complete expan-sion of the packers by monitoring the compressor used forinflation, further improves the utilization of such data.

A very interesting development for recognning hydrauli-cally active fractures is described in Crosnier et alii's paper.The paper discusses the utilaation of the "harmonic probe"(already used in the laboratory tests, and for in situ testsat great depths) in geotechnical investigations. The basicprinciple of the equipment consists in exciting the fiszuresof the rock mass filled with fluid, by a sinusoidal signalconsisting of a pulsing flow, and in recording the corres-ponding pressures as output signals.

The complex relationship between the output signal (inpressure) and the output signal (flow), which is a functionof the frequencv of the sisnal. characterizes the testedfissuring sy'stem.- The fissuiing parameters are identifiedby comparing them with a set of characteristic signalsobtained through numerical modelling, which may keepaccount of the forces of inertia, of those due to viscosityand of the compressibility of the rock and fluid. Suchequipment allows permeabilities of the order of evenl0-12 m/s to be meazured.

The laboratory calibration of the prototype, carried outon a physical plexiglass model proved a substantial agree-ment between the physical and the mathematical models.It is pointed out that the use in practice of this equipmentwill entail the need to study different interpretationalmodels according to the geological and geotechnical charac-teristics of each site to be investigated.

The permeability of rock masses is in general influencedby the state of stress, and for this reason the state of flowand the state of stress in the rock mass cannot be evaluatedseparately ; the two problems must be considered stronglycoupled.

The relationship between permeability and state of stresshas been investigated by different Authors (Morgensternand Guther, 1972; J ouanna , 1972; Bernaix, 1969 ; Rosen-gren and Jaeger, 1967). It appears that permeability canvary by rnore than one order of magnifude for the modifi-cations of the state of stress norrnally induced by engineer-ing structures. This is particularly important in evaluatingdam behaviour (Manfredini et al., 1975). As to the type of

35

functional relationship between permeability and stresstensor, a simplified hypothesis is to assume that the per-meability of the rock mass is always isotropic and that itsvalue depends only on the mean stress. On this assumptiondifferent relâtionships were proposed by Morgenstern andGuther (1972) and by Louis and Perrot (1972).

Laboratory tests were carried out by Rosengren and Jaeger(1967) on the geomechanical models of fractured rocksobtained by thermal disgregation of a marble ; the resultsshowed that the permeability in a given direction does notdepend on the stress acting along that same direction, butonly on the stresses acting on a plane normal to it.A more adequate model of the stress-dependent permea-bility in which such behaviour is taken into account can beworked out by assuming that the principal directions of thestress and permeability tensor coincide and that each principal permeability is independently influenced by the twoprincipal stresses in the other principal directions (Manfre-dini et al., 197 5).

The validity of such constitutive laws can be checked onlyby in situ tests on significant volumes or, where possible,on the basis of the behaviour of life-size structures.

A useful contribution to this study is provided by the paperpresented by Carlsson and Olsson, where an in situ test isdescribed which consisted in measuring the permeabilityof a significant volume of a rock mass (= 6 m3) at variablestates of stress induced in the rock mass by a series ofhydraulic jacks acting on steel bars. The permeability testsconsisted essentially of water injection tests by means ofdouble packers.

It must be mentioned that this type of test has the advan-tage, as compared with tests on isolated blocks, of avoidingloosening in the sample and therefore a fictitious increasein the penneability.

The tests showed that at increasing vertical loads there weredecreasing horizontal permeabilities, but, as a general trend,an increase in conductivity through steeply dipping frac-fures..

The latter result does not appear to be immediately under-standable as the increase in vertical stress should also giverise to an increase in horizontal stress and therefore to a

closing of the vertical fissures. As can be expected, therise in the hydraulic test pressure, by reducing the effectivepressures, causes an increase in permeability.

The measurement of the dilution over time of tracersintroduced into the holes, sometimes provides useful indications as to the permeability of the soils and especiallyabout the type of flow of water inside the aquifer. [t canfurthermore provide indications on the effectiveness ofgrouting measures or, in some cases, give indications on thetrend of seepage in earth dams. Three papers of the Sessionwere devoted to this topic. Radioactive tracers for geophy-drological investigations are discussed in detail in Gautierand Rousselin's paper. Bromium 82 and indium 113 m wereused as markers. The half-life of the former is 36 hours (itthus allows to carry out tests lasting 5-6 days), and of thelatter it is about 100 minutes, which means that testslasting one day may be carried out and that the same testor other tests may be repeated on the siune place. Bothhave the advantage of being lowly toxic.

Some useful technological measures (for example a direc-tional probe) make it easy to determine the direction ofthe horuontaJ, component of the water motion.

The use of the radioactive tracers is discussed also in thepaper presented by Huller, et al. ; measurements of thewater flow according to the dilution of the tracer over time

and measurements of the soil permeability on the basis ofthe vertical displacements of the tracer inside the holewere carried out.

The paper also discusses the use of radioactive tracers forcontrolling a sealing wall built with self-hardening suspen-sions, and for studying the geometry of a landslide.

In Malkki and Wihuri's paper a similar problem is discussed.To determine the permeability of fluvioglacial deposits,colourimetric or electrolytic tracers were used instead ofthe radioactive ones, so as to avoid drawbacks due to toxi-city. Dilution is measured either by means of a light sourceand photocells, or, where electrolytes are used, by measur-ing the conductivity of the medium between two elec-trodes. In order to obtain greater reliability, an experimen-tal calibration was used to turn dilution velocity into actualvelocity.

The flowmeter allows a recording of permeability valueswith depth for soils having permeabilities in the range of10-4 l0-7 m/s. This technique has the advantage ofbeing very rapid and inexpensive.

The paper presented by Bruzzi et al., discusses the resultsobtained by using such equipment in an alluvial plain.The testing techniques are described. The results obtainedunder conditions of nafural flow, or "dynamic" flow follo-wing an artificial rise in level are compared with Lefranc'stests carried out in the same holes.

Notwithstanding the uncertainties and approximationwhich charact enze, in general, such determinations, it wasverified that the agreement between the two sets of measu-rements deviated by only l0 %.

Other comparisons were made with laboratory tests onundisturbed samples, or on the basis of empiiical corre-lations based on grain- size analyses.

Emphasis is given on the importance of a very carefulinstallation of the perforated casing and of the filterbetween casing and soil to get reliable information aboutthe permeability.

The determination of permeability of fine soils (silts andclays) was carried out almost exclusively in the laboratoryup to 15 years âgo, by means of time-settlement curvesin oedometer tests. The forecasting of the settlements overtime for engineering structures, has always therefore provento be rather uncertain due to the fact that laboratory testscannot adequatly represent the structural conditions of thein situ material (presence of thin more permeable intercala-tions in soft clays, the presence of joints and fissures in theoverconsolidated clays, e tc.). In order to improve theknowledge about the actual in situ permeability of silty-clayey soils, different research lines are being pursued, mostof which are complementary.

An exhaustive review on this topic was presented by Jamiol-kowski et al. (1979). A possible line is that of taking andtesting samples of such a sue as to include also the o'stnrc-

ture" of the clayey deposits being examined; samples ofFiumicino soft clays having a diameter of 0.49 m andheight = 0.20 rn, were tested in ædometric conditions(Burghignoli and Calabresi, 1977); a second possibilityis that of trying to get a more detailed as possible repre-sentation, at least in quantitative terms, of the strucfuralcharacteristics (sandy intercalations, etc.) of the depositby using the piezometric probe or the piezocone conti-nuous test, which is highly sensitive to the detection oflayering (even if slight) having different permeability insidea clay deposit. A third way is that of directly evaluatingthe consolidation coefficient starting from the curve ofpore pressure dissipation generated tn situ by stress varia-tions. This may be attained for example with the same

piezocone, stopping the penetration and allowing porepressure dissipation; some drawbacks, however, consistin high remoulding and disturbance in the structure of thedeposit around the point, and in the geometric conditions,not duly represented by the theory of cavity expansion.

A very promising tool from this standpoint is the SelfBoring Pressuremeter (Clarke èt al., 1979) that minimizesdistrubance to the soil and is capable of inducing undrainedstresses in the soil at a prefixed strain value, and to measuresubsequent pore pressure and the dissipation of these withtirne.

The cylindrical geometry allows a simple interpretationof the results.

A different approach for improving the reliability of theforecasts regarding the behaviour over time of the settle-ments of structures, starts from the consideration that itis usually possible to estimate to a close approximationthe compressibility characteristics of the material (as isshown by the fact that the estimation of final settlementsof engineering structures proves in general to be quitereliable), and therefore it may be sufficient to improve thereliability of the permeability measurement to obtain a

better estimation of the consolidation coefficient. Threepapers presented at this Session have followed thisapproach.

Regarding the direct in situ measurement of permeabilityin very fine soils by means of piezometers, a rather exten-

sive bibliography is already available (Gibson, 197 0 ;Bjerrum et 0.1., 1972; Wilkinson, 1968; Wilkes, 1974;Massarsch, 1978; Mieussens and Ducasse, 1977).

In general, the piezometers to be used for measuringpermeability can be installed in boreholes or directlyplaced inside the ground. In both cases it is necessarythat drilling operations or the pushing in of the piezo-meter should not destroy the soil "structure" ; moreoever,in the former case, the effective geometry of the sand filterinto which the piezometer is placed (difticult to be knownin detail) may considerably affect results ; in the latter casethe pushing of the piezometer may lead to the remouldingof the soil (up to 3 to 5 times the diameter of the piezo-meter) and also to hydraulic fracturing if it is pushed invery rapidly ; furthermore, tlre piezometer filter may getclogged with clayey material.

The paper by Tavenas et ol. described a very promising toolthat should eliminate most of the above-mentioned draw-backs.

The equipment proposed is a selÊboring permeameterwhich is introduced at depth inside the soil in a similar wayas the self-boring pressuremeter (Baguelin et al., 1974).Inthis v/ay the effects of the pushing or the unknown geo-metry of the filter are eliminated.

Another innovation is the replacing of porous filter with aperforated metal lamina, whose holes are closed by a rubbermembrane pressed against them during the penetration of

50 60rc80timc , min.

ôo70-ætimc, m in.

€%loIô

4

2

20 30 40mcmbrone

Pressu r€

B

obrlJaooL-

20 30

t20o a, kN/mP

80

.sob+t@

ûaooL-+.o

-(u+)o+.

oL.oCL

ooooxo

50Q

Spr inRu bber

FiC. 2. The Cambridge self boring prcssuremeter ; holding test (after Clarke er al., 1979).

t imc,

37

the tool. Inevitably, some remoulding due to the sliding ofthe tool against the soil still doe s occur, but reduced inextent. As regards the different modes of carrying out, thepaper discu sses their advantages and disadvantages ; _thevaiiable head test (which is very easy to carry out) does

however entail effective stress variations during the testand thus sorne errors in measuring the permeability whichis difficult to evaluate ; it furthermore refers to a ratherlimited soil volume.

As regards the constant head test, it is possible to keepaccount of the swelling of the material and thus definethe steady flow and as a consequence the permeabilityof the soil; the volume involved by the test is substantiallylarger, but the testing equipment used so far is extremelycomplex and inadequate for routine testing.

The Authors suggest a very simplified equipment forcarrying out constant head tests which are reasonablyreliable and accurate. Comparisons are made between valuesobtained tn situ with this equipment and the laboratoryvalues for different soils. Interesting comparisons have alsobeen made with the determinations obtained by introduc-ing the permeameter without the self-boring device. Theremoulding effect of the structure of the deposit is in thelatter case quite evident.

'For deterrnining the permeability of a weathered andfissured clay layer, Lafleur and Giroux have develop a

"reaming pelrneameter" which was to allow the cleaningof the walls of the hole (thus avoiding that the fissuresmight be filled with material) ; the possibility of carryingout measurements practically throughout the length of thehole ; the possibility of measuring permeability- valuesdiffering grêatly amông each other (even at a 10s ratio)according

- to the presence or not of open fissures in theclay. The equipment used, which is introduced into an

already bored guide hole, consists of a cutting reamer, twoinflatable packers and an internal injection cell ; it is intro-duced intô the ground by means of hydraulic jacks. T4.permeability detérminations were carried out at variableloads for the sake of simplicity in spite of the uncertaintieslinked to the interpretation of such measures. The hydrau-tric and the inflation pressures of the packer are propor-tioned so as to avoid hydraulic fracturing phenomena,yielding of the soil near the packer, and water leakagethrough thern.

lt must be naturally borne in mind that in such a weatheredclays, the opening of fissures and thus permeability mayhave large seasonal variation. [n case they are located abovethe watér table, during the permeability test there may alsobe some swelling of the clay and closing of the fractureswhich considerably complicate the interpretation of testresults.

The more correct way for determining the effective impro-vement in knowledge that can be attained by measuringpeffneabllity in situ instead of util:zing the classical labo-iatory tests, is the back analysis of the behaviour of struc-tures (or better, a comparison between design forecastsand real behaviour of structures) as regards the settlement-time behaviour.

An interesting paper by Mieussens is devoted to this topic :

during perrneability tests the increase in effective stresses(for Àegative head changes) or the decrease of effectivestresses (for positive head changes) are small. The resultsof permeability obtained may then be considered as refer-ring to the soil in the preconsolidated state.

For the calculation it is assumed that the value of cv

should not vary within the two soil domains (which are

respectively the normally consolidated and the over-consolidated soils), and that the ratio of the two corres-

ponding values (c,,r/cu) is equal to the reciprocal ratio ofthe compressibility indices. The findings of the in situ per-meability tests obtained by means of the "PerméamètreAutoforeur", applied in two practical cases, yielded cvvalues which v/ere much closer to those obtained by theback analysis than the laboratory results, but still notcompletely satisfactory (about 4.5 - 2 times greater than theactual values). Furthermore, the permeability values obtain-ed in situ can be essentially referred to the horuontal per-meability, whilst in the examined consolidation process itis mainly the vertical permeability which is involved, andthis means a probably larger discrepancy between field testsand actual behaviour.

As an example, and to give an idea of the global usefulnessof these deierminations, the following fig. 3 (after Jamiol-kowski et al., I979) summarizes the results obtained by an

extensive field and laboratory investigation carried out atPorto Tolle (Italy) for the construction of a thermoelectricplant. The findings are compared with a back analyses ofirial embankments with or without vertical drains (whichmeans involving mainly vertical or horizontal in situ per-

meability re spectively).

Fig.3. Comparison betrveen laboratory and in-situ c., and ch

values at Porto Tolle (Italy) ; (data after Jamiolkowskiet aI., 197 9) .

A look at the data shows that, first of all the differe nt insitu tests for determining the horizontal permeability, orthe horizontal c onsolidation coefficient c6 (piezometerprobe dissipation test, constant head permeability test andielfboring permeameter tests) provided comparable values;such values were higher than those obtained in the labora-tory, but lower than those inferred from a back analysesof the development of settlement over time ; moreover, as

one might expèct, the largest differences between the actualbehaviour and the laboratory tests were found for the horizontal consolidation coefficient, that in some cases is mostimportant (for example for an accurate design of the verti-cal drains).

The existing discrepancies are probably due to the differentscale of the probe and of the macrostructure of the soil.

The Delft Soil Mechanics Laboratory has recently deve-

loped an electrical porosity probe capable of measullngthè porosity of saturated sands below water tables. Thisdeteimination may be advantageously used for properlyreconstructing, itl the laboratofY, sand samples with thesame field density, thus obtaining reliable permeability

lL oed. testcon t. oed. testback anahsiswit hout dra i n s

lL oed. testpie zometer probe

constant head testse lf bo ring.permea mete rback analysiswith drains

38

values for homogeneous sand deposits (for the same objec-tive the nuclear density probe, developed by the samelaboratory, combined with a CPT test may be used).

Given the formal analogy between the laws governinghydraulic and electric fields, a probe similar to that usedfor rneasuring resistivity was built for measuring soil per-meability (dipole hydraulic probe). (Riestsema

-and Vier-

gever, 1979). The article presented by De Leeuw andSilence, discusses this equipment. The probe, which has adiameter of 36 ffiffi, consists of four porous filters; throughthe first of these water is pressed into the soil;by a second.filter at the other end water is sucked out by means of areciprocating pump with a constant but adjustable rate ofdischarge. The potential difference between two otherintermediate filters is measured with a differential pressureg.auge. It can be shown that, unlike what one might think,lhir e quipment measures e ssentially horizontaf permea-bility. Vertical permeability may also be evaluated, onceone knows the horizontal value, by placing two parallelprobes, spaced 0.5 m, which can measure the pôtentialdifference produced by adopting different schemes ofwater inflow and discharge.

It is yery likely that, as occurs with piezometers pushedinto the ground, there may be a certain disturbance in thesoil around the dipole probe, due to packing of sand, oralso due to localized failure in individual- sand grains.

Numerical verifications carried out with the SEEP pro-gramrne have shown that, due to the type of measurementsand of the type of materials generally investigated withthis test, this distrubance does not give rise to large in-fluences on the measured permeability value.

4. Measuring permeability at high depths

It is only in recent years that the permeability of soilsand of rocks, situated at great depths, has become more andmore important. The fields of application which need thisknowledge aîe geothermal problems, production problemsrelated to oil-fields, in-situ coal gasification, nuclear wastedisposals, and also problems related to earthquake mecha-nics.

The reduction in frequency and opening of fissures withdepth, and the corresponding decrease in permeabilityis reported by a large number of Authors. This decrease,for the first 100 m, is generally very sharp and is found inalmost all permeability tests and grouting measures involv-ing rock foundations in dams (Snow, 1968 ; Davis andTurk, 1964 ; B onal di et al., I 983).

A recent review on the distribution of permeability at greatdepths for different rocks was made by Brace (1980). Ingeneral, there still does appear to be a decrease in permeabi-lity with depth, but the diversity of the rocks examined,and quite likely also the different testing methods used donot help in finding a general defined trend (fig.4).Sorne intere sting overall indications on the presence ofhydraulically active fissures even at great depths are pro-vide d by the re sults of a study on the structural andgroundwater conditions in about 20 mines located in theCanadian Shield (Raven and Gale , I971, also reported, byWitherspoon and Gale, 1977) fig. 5. In examining the datait must be borne in mind (as pointed out by the Authors),that the mining activities at different levels may mutuallyaffect the water conditions and that ventilation can reduceseepage, which can be noticed in the walls.

The recent problem of the disposal of nuclear waste at gratdepths (generally below 500 m) in low-permeability soilshas furthered knowledge in this fields.

I n-s it u vs Dept h

H

I t]

lln;lgl

lltlll

[il'r7

-8

-9

<50 s0 - 500

D E PTH

Fig. 4. In situ permeability values at different depth reviewed byBrace (after Brace, 1980).

tooo

ril[r,l

-l

O

al -?E \''

o)

-3 -5-6

v,L(u

(l)

E

IFùu.Jo

Fig. 5. Variation of groundwater inflowsselected underground mines (after

as a function of depth inRaven and Gale, L977).

0tlllIrllllrlll

u

I

ril

il

TIMATED GROUNDWATER INFLOWLITERS PER MINUTE

ro 50 too . to3

o Sserved or reportedzonc of conslonlsecçrfoig

--< Zonc of sccpogc wittrnolwol voriotion offlo with timc

+-r-lq1s of scepogc withinôrced vaiotion offlor with tirnc duc togrouting, plugging etc

39

Most of the investigations related to the examination ofpossible site s for waste disposal involve the determinationof fracturing and of rock permeability, as the most likelyway of transport of the radionuclides in the biosphere isvia the water flow through the fissures in the rocks. Thewater flow in the rock mass around a waste disposal isessentially controlled by the heat which develops in thewaste itself ; the complex phenomena governing the waterflow and heat transport (variations in the fluid viscosityor state variation of the fluid, variations in fissure openingsor opening of new fissures due to thermal stresses etc.) areto be taken into account when dealing with these problems.

Three papers presented at this Session refer on special pro-jects that a;te currently underway for clarifying some ofthese problerns.

The first of these is the "stripa Project", sponsored byseven countries, which examines the rock mass inside whichlies an abandoned mine in Sweden; a second project exa-mines the characteristics of the granite and the gabbroplutons present in the Canadian Shield.

The paper by Witherspoon and Gale refer about a detailedsite investigation focusing on the chancterization of per-meability in the Stripa granit massif ; the programme con-sists of a detailed reconstruction of fracturing, tests in bore-holes, and a large-scale permeability test along a stretch ofdrift which \ilas appropriately equipped.

The injection tests in the boreholes showed, in spite ofsome scatte r, a general trend of decreasing permeabilitywith depth. Extremely interesting results have been ob-tained from a large-scale permeability test in a drift 5 m x5 m x 33 m at el. 338. Water inflow was measured as thenet moisture pickup of the ventilation system in the drift,and water pressure was measured in isolated stretches inseveral boreholes departing from the drift. Water flow andpressures were monitored during 11 months. The findingsof the various tests carried out at different temperaturesof the ventilation air, provided practically constant valuesof hydraulic conductivity of about l.1g-to rn/s.

It is interesting to note that, by improving some of themeasuring techniques and for longer testing stretches of thedrift, one rnight even obtain permeability values on theorder of 10-12 m/s, and this means that this type of large-scale test could be used in the future as an acceptabilitytest of individual waste storage drifts.

Besides the great specific interest of this determination,such large-scale tests provide extremely useful data as tothe actual rneaningfulness of borehole tests, that are morefrequently adopted and easier to carry out.

In this specific case, the average permeability finding( 10- 13 cm2) proves compatible with the results obtainedfrom water injections in boreholes starting from the wallsof the same test room (average value 8.9 . 10 - t 3- cm2).Keeping account of the different testing modalities of thetwo types of test (fluid injection in boreholes and fluidwithdrawal in the adit), but the comparison stresses also theimportance of the "scale-effect" on this measure.

The examination of existing fractures (coupled with thewater absorption test in theboreholes) may finally allowto determine the fracture porosity of the rock mass, towhich the transport of the radionuclides is linked.

Carlsson et aL present the results of special hydraulic testscarried out wiihin the framework ôf tne iame "stripaProject" ; the paper discusses the usefulness of the cross-hole technique for defining the hydraulic parameters ofthe formation and for identifying the geometrical characte-ristics of a band of highly fractured material.

The hydraulic interference test was conducted after a

buildup phase of over 3 months. The packed-off sectionin the first borehole was changed into a free-flowing state,with a continued pressure recording in the second borehole.The response showed a hydraulic interconnection betweenthe two holes. It was possible on the basis of the hydraulicproperties obtained, to evaluate the specific storage coeffi-cient, and to indicate, in a preliminary wây, the directionof the fractured zone.

The use of geophysical measurements (o'mise à la massemethod") was also used in order to locate the structure.The comparison of the permeability results obtained for thedifferent test methods used is quite interesting. Such acomparison confirms the need not to restrict the investi-gation to a single method for measuring permeability, andthe need to define the permeability characteristics of therock mass on the basis of different tests which often com-plement each other.

The paper by Lee et al. refers about the investigations con-ducted to charact erize the permeability of rock masseslocated in the Canadian Shield. Also in this case, differenttypes of tests were carried out (at depths from 200 m downto over 1200 m), basically according to three types ofprocedures: a) instantaneous withdrawal tests (slug tests,pulse tests, drill stem tests) ; b) instantaneous injection tests(slug tests and pulse test) ; c) constant pressure injectiontests (step type).

The comparison among values obtained by the constanthead tests, and by the transient tests, shows a ratio betweenthe perrneability values measured by the transient head,and the permeability values measured by the constant headwhich vary from 1.1 to 7.6.

Furthermore, both in the "withdrawal" and in the "injec-tion" cases, the "pulse" test provides permeability valueswhich are much lower than in the "slag test".

It is also interesting to examine the decrease of permea-bility with depth; also in this case such a decrease is sharpfor fhe first dOçZSO m of depth(from 10-7 m/s to 10-11m/s),and it becomes less sizeable at greater depths. In thedeep tests in granite with no fractures permeability valuesof 10-14 m/s were measured, that is values at the extremelower range of the field testing equipment used.

5. Theoretical and general analyses for a betterinterpretation of the permeabiiity tests

It is well-known that the study of water flow inside frac-tured media may be conducted from a theoretical pointof view by following two different approaches : in the first,the rock mass is considered as consisting of impermeablematerial divided by discontinuities more or less inter-cofirmunicating, and the water flow through these discontinuities is examined in detail; according to the secondapproach, the "impermeable rock-discontinuity" unit isrêplaced with a côntinuous medium, which is generallyanisotropic and which on average has the same hydrauliccharacteristics. The latter method may be adopted whenthe rock mass may be taken as being homogeneous, thatis when the spacing between the "hydraulically active".fractures is sufficiently close in relation to the scale ofthe problems to be studied. This schematization helps indealing with water flow problems inside a rock mass,especialy because these problems can be tackled withnumerical calculation methods.

ln order to define an "equivalent permeability tensor",it is necessary that it be independent from the limit condi-

40

Indeed, it should be borne in mind thatof structures are present, the equivalentto be defined for each of such different

tions imposed on the reference voluffie, that it be symme-tric, and that it should tend to a limit value as the referencevolurne increases. This however, does not always occur atthe scale of the in situ tests carried out in test boreholes.The paper presented by Feuga deals with this problem.

| .purticular test array was studied to evaluate the possi-bility of obtaining information on the permeability tènsorof a fissured granite rock. It consists of injecting water intoa borehole (with a double-packer unit) and then measuringthe pressure rise in three holes drilled according to a tefti-hedric pattern, so as to involve a rock mass of about 12 m3(the fracturing density found for each borehole was onave rage 3 fractures/m).

The test pointed out that in order to be able to considerthe rock mass being investigated as equivalent to a porousmgdium, the borehole lengths, the lengths of the injectioncell and of the pressure measurement cèll, should have beengre-atly, increased with the risk, however, of involving notgnly the level of diffused fractures, but also that of largediscontinuities.

be unwise. In this wày, it would be possible to overcomethe uncertainties of the design forecasts related to the eva-luation of the permeability characteristics (trial embank-ments can be considered "at limit" large-scale in situ tests).

Sigismond and Doucerain report about an experimentallargescale test for the design of the dewatering-system ofthe excavations for the Nogent-sur-Seine nuclear powerplant.

For the foundations, the water table had to be loweredabout 3-5 m, the environmental impact of such a measurewas not to be too large and the amount of water removednot too high.

Four experimental curtain walls were built at depths from14 to 26.50 rn, and the size in plant being 15 x20 m.

Pumping tests were carried out before and after thebuilding of such curtain walls. An analysis of the result prcvided by the piezometers installed at various elevationsinside and outside the wall allowed to determine the per-meability of the various soil levels ; an analysis of the wàterflow as a function of the depth of the watertight wall,allowed to make correct choices for the final construction.An evaluation of the water flow to be removed during theworks made on the basis of the findings of this test, wasconfirmed after the construction stage.

A type of test which in many respects is quite similar tothat described is reported in the paper by Manfredini andSilvestri. In this case the permeability was to be determinednot for dewatering purposes, but on the contrary to eva-luate the losses there might be through the bottom of anartificial hydroelectric re servoir.

The heterogeneity of the deposit (and the large costs in-volved in an artificial lining of about 7 50.000 m3) suggestedthat large-scale permeability tests should be carried out.A vertical soil prism 5.2m wide and 32 m deep, ,was hydrau-lically isolated by means of a concrete diaphragm wall. Asteel cap was secured to the top and the gap between soiland cap was filled with water at a pre-established hydraulichead. The corresponding increase in pore pressure was mea-sured in the soil and a flow analysis allowed a reasonablepicture of soil permeability. In this case, as in the previousone it will turn out to be very useful for future develop-ments to make a thorough review of all the permeabilityroutine tests carried out and compare, where possible theforecasts that would have been made on the basis of theseelements with the findings of the large-scale permeabilitytests.

In their papers, Haq and Hashmi refer about a broad seriesof "in borehole" permeability tests (essentially Lugeontest) conducted for charucterizing the foundation rockof the Simly Dam. Preliminary tests showed large waterleakage along the first 20-25 m, and so an elaborate inves-tigation grouting programme was carried out.

As in many valleys, âs a consequence of different states ofstress, higher permeability values are found at the two abut-ments, whilst smaller values are obtained along the bed.

Permeability tests were extremely useful in determiningdesign criteria for the grout curtain, the depth, the spacing,the inclination and the injection pressure of ths groutholes.

The paper by Loudière and Fatton is the only one of theSession that deals with the permeability of very super-ficial soils (from I to 3 metres). In particular, a seriei oftests for charactefizing the soils at the bottom of lagoonsor shallow artificial ponds were carried out according to the"Porchet" method, which consists in measuring as a func-tion of time, the rate of water flow into a shallow hole.

if different kindsporous medium is

sentative volume charact eraing a fractureinclude larger discontinuities.

levels. The repre-level should not

It rnust also be recalled that the larger the volume investi-gated for determining the permeability, the greater theperrneability value obtained. This depends on the fact thatwith larger volumes different levels of fracturing are invol-ved, which arc generally characterued by larger openings,and thus much higher the hydraulic conductivity.

A detailed and accurate "geostatistical" investigation ofthe fracturing of the rock mass is to be carried out beforedeciding on how the in sttu permeability tests are to beconducted.

In the paper presented by Mironenko the processes occur-ring during the pumping tests and other permeability testsarc discussed. The influence of the state of stress of therock massif on the re sults of the permeability tests ispointed out, as well as the possibility of evaluating com-pre ssibility by experirnental pumping tests, and the utihza-tion of field water tests for evaluating the state of stress ofrocks.

In particular, the possibility of monitoring the area whereintense pumping is to occur so as to evaluate the deformabi-Iity characteristics of the soils is discussed, as well as thepossibility of measuring the state of stress by means ofhydraulic fracturing techniques during mining operations,starting from holes with different orientations.

6. Case histories and special in situ tests

Well-documente d case-histories illustrating special te stsfor the construction of engineering works are importantfor verifying the applicability of advanced research techno-logies, and for comparing experiences amongst techniciansoperating in this sector.

In the specific case of investigations for determining per-meability, these verifications are very interesting in consider-ation of widely different methodologies used. Even a merecomparison of results obtained by different tests carriedout for the same soils may yield some useful indicationsfor advancing knowledge in this rield.

For works involving important technological and economicdecisions, permeability tests at a scale that allow to keepaccount of the complexity of the soils may not prove to

Causes for possible mistakes inherent in the measurementsarc discussed and suggestions are made to avoid sucherrors.

Finally, a series of experimental findings referring to differ-ent soils are shown. The comparison with laboratory testsshows that the latter generally provide lower permeabilityvalues. It is very likely that these may be due, beside thefactors mentioned by the Authors, to the in situ effect ofa meaningful "structure" and to some permeability aniso-tropy of the soil.

The paper by Krupoderov describes the results of a seriesof in situ investigations on debris resistance to washingout on natural slopes at various flow rates.

The influence of the grain-size of the soils, of the density,of the water content on debris resistance to wash out arealso examined.

This topic is borderline in terms of the determination ofin sttu perneability, but it is important for predictingthe development of erosional and mudflow processes inmountain and hilly regions.

7 . Final rernafks and points for further discussion

Some general comments are to be made at the conclusionof this review.

As in many other fields of soil and rock mechanics, there is

a growth of interest in the development of new techniquesand nev/ methods for the determination of permeabilityby means of in situ tests.

The main developments in this area may be summaizedas follorws:

Rock mnss permeability

The drawbacks and the limits of the pack tests are wellrecogn ned: it is therefore important that technologicalmeasures airping at limiting the uncertainties of the testsbe adopted also for routine tests. [n relation to this, thepossibilities offered by the wide range of cu.rrently aval-Iable instrumentation do not appear to be fully exploited.

Perrneability tests, typical of oil prospection arc beingapplied successfully with appropriate changes in the ateaoî-rock mechanics. On the basis of this example it appearsto be quite important for geotechnical researchers to use

in the future, tô a much greater extent than has been doneso far, the experience of technicians working in other fileds(oil prospection and oil production, mining, irrigation andagricultural research and so on).

The "harrnonic probe" may prove to be very useful inmeasuring the transmissivity of individual fractures; thismay be taken as a typical example of a joint applicationof modern measurement techniques and of calculationmethods for interpreting the tests.

A valid aid in interpreting permeability test results forrock masses is provided by fracturing surveys which are

now carried ouf on a routine basis, by means of boreholetelevision cameras, with continuous data registration, orby impression packers in some particular stretches.

Soil permeability

The most promising results have been obtained by usingpressuremeters or self-boring permeameters which do notèxcessively disturb the soil. A more extensive use of suchequipment, also for routine tests will undoubtedly lead toa

-méaningful improvement in the design forecasts of the

behaviour of engineering structures.

Some points may deserve further discussion, for instance :

- In situ measurement of the permeability of soils androcks as L function of the variation of the state of stressinduced by the engineering structure ;

- Meaningfulness of the different "in borehole" tests indefining the permeability of large rock volumes;

- Meaningfulness and limits of the "in borehole" tests fordefining an "equivalent porous mediurn" for a rock mass;

- Possible improvements in test methods and in the waythey arc carried out so as to make pressuremeter of self-boring permeameter tests more meaningful from the"scale" standpoint in relation to the macrostructure of the

- Possible future innovations in selÊboring instrumentsor in operation modalities, to evaluate also the vertical per-meability of soils.

Acknowledgements

This study was supported by a C.N.R. Research Contract(Working

-Group on Rock Mechanics at the University of

Rome).

References

BAGUELIN F., JEZEQUEL J.F., LE MEHAUTE A. (1974): Selfboring placement method of soil characteristics measure-ments. Proc. Eng. Found. Conf. on subsurface exploration.Henniker. New Hampshire.

BERNAIX J: (1969): New laboratory methods of studying themechanical properties of rocks. Int. J. Rock Mech. Min.Sci., 6.

BJERRUM L., NASH J.K.T.L., KENNARD R.M., GIBSON R.E.(197 2): Hydraulic fracturing in field permeability testing.Geotechnique, 22, No. 2.

BONALDI P., MANFREDINI G., MARTINETTI S., RIBACCHI R.SILVESTRI T. (1983): Foundation rock behaviour of thePassante dam (Italy). V Int. Conf. ISRM. Melbourne.

BRACE W.F. (1980): Permeability of crystalline and argillaceousrocks. Int. J. Rock Mech. Min. Sci., 17.

BURGHIGNOLI A., CALABRESI G. (1967): A large sampler forthe evaluation of soft clays behaviour. Spec. Sess., Proc.IX ICSMFE. Tokyo.

CAPOZZA F. (1977): Panel discussion on Session 3. Int. Symp.on the Geotechnics of structurally complex formations.Capri.

CLARKE B.G., CARTER J.P., WROTH C.P. (L979): In situ deter-mination of the consolidation characteristics of saturatedclays. Design parameters in Geotechnical Engineering, 2.

Proc. VII ECSMFE. Brighton (U.K.).

COLOMBO P. (1973): Discussion on theme 2. Effetti meccanicidel moto dell'acqua nelle rocce sciolte e lapidee. XI Con-vegno ltaliano di Geotecnica. Milano.

CROCE A. (L973): Discussion on theme 2. Effetti meccanici delmoto dell'acqua nelle rocce sciolte e lapide. XI ConvegnoItaliano di Geotecnica. Milano.

DAVIS S.N., TURK L.S. (L964): Optimum depth of wells incrystalline rocks. Groundwater, 2, No. 2.

FRANKLIN J.A. (I979): Geneml report. Use of tests and monitor-ing in the design and construction of rock structures. IVInt. Conf. ISRM. Montreux.

GIBSON R.E. (1970): An extension to the theory of the constant-head in situ permeability test. Geotechnique, 20, No. 2.

HAMSON B.C. (1978): The hydrofracturing stress measuringmethod and recent field results. Int. J. Rock Mech. Min.Sci., 15. ,

42

HEITF'ELD K.H., KRAPP L. (1981): The problem of water perme-ability in dam geology. Bull. IAEG, 23.

IAEG Int. Ass. Eng.Geol. (1981): Report of the Commission onsite investigations.

JAMIOLKOWSKI M., LANCELLOTTA R., PASQUALINI 8.,MARCHETTI S., NOVA R. (L979): General report. Designparameters for soft clays. vII ECSMFE' Brighton (u.K.).

JOUANNA P. (I97 2): Essais dn situ de percolation sous contrainte.Proc. Int. Symp. Percolation through fissured rock. Stuttgart.

LOUIS C. (1981): General report on Session 3. Eau souterraine etproblèmes d'infiltration. Proc. X ICSMFE. Stockholm.

LOUIS C, PERROT M. (1972): Analyse tridimensionnelle desécoulements dans le massif de fondation du barrage deGrand Maison. Proc. Int. Symp. on Percolation throughfissured rock. Stuttgart.

MAINI Y.N.R., NOORISHAD J., SHARP J. (I97 2): Theoretical andfield considerations on the determination of in siru hydraulicpararneters in fractured rock. Proc. Int. Symp.Percolationthrough fissured rock. Stuttgart.

MANFREDINI G., MARTINETTI S., RIBACCHI R. (1975):Mutual influence of water flow and state of stress in theanalysis o f dam foundations. Proc. Symp. Criteria andassumptions for numerical analysis of dams. Swansea.

MASSARSCH K.R. (1978): New aspects of soil fracturing in clay.Proc. ASCE, 104, No. GT8.

MIEUSSENS c., DUCASSE P. (1977): Mesure en place des coeffi-cients de permeabilité et des coefficients de consolidationhorizontaux et verticaux. Revue Canadienne de Geotech-nique, 14.

MIYABE N. (1969) : Directions of research on land subsidence.Proc. of the Symposium on Land Subsidence.

MOORE P.T., BALBIS R.E. (L979): Head losses in packer tests.Proc. ASCE, 105, No. GT9.

MORGENSTERN N.R., GUTHER H. (L972): Seepage into anexcavation in a medium possessing stress-dependent per-

meability. Proc. Int. Symp. on Percolation through fissuredrock. Stuttgart.

PEARSON R., MONEY M.S. (1977): lmprovements in the Lugeonor packer permeability test. Q.J. Eng. Geol., 10,

POLAND J.F. ( 196 9) : Status of present knowledge and needsfor additional lesearch on compaction of aquifer systems.Proc. of the Symposium on Land Subsidence.

RAVEN K.G., GALE J.E. (1977): Subsurface containment of solidradioactive waste: A study of the surface and subsurfacestructural and groundwater conditions at selected under-glound rnines and excavations. Internal report, Library ofGeological Survey of Canada.

RIETSEMA R.A., VIERGEVER M.A. (L979): In situ measurementof permeability. Proc. VII ECSMFE. Brighton (U.K.).

ROSEGREN K.J., JAEGER J.C. (1968): The mechanical propertiesof an interlocked low-porosity aggregate. Geotechnique, 18.

SNOW D.T. (196 8): Rock fracture spaciogs, openings and porosities.J. Soil Mech. Found. Div., Proc. ASCE, 94.

TELI-AM J.H., LIOYD I.W. (1981): A review of the hydrogeologyof British onshore non-carbonate mudrocks. Q.J. Eng. Geol.

U.S. Department of the Interior. Bureau of reclamation (1977)Grand Water Manual.

URIEL ROMERO S., AQIZ L.U. (1981): Coastal soils permeabilitybased on tidal damping. Proc. x ICSMFE. Stockholm.

WILKES P.F. (L974): Permeability tests in alluvial deposits andthe determination of Kg. Geotechniqve,24, No. 1.

WILKINSON W.B. (1968): Constant head in situ permeabilitytests in clay strata. Geotechnique, 18.

WITHERSPOON P.A., GALE J.E., cooK N.G.w. (L977): Radio-active waste storage in argillaceous and crystalline rockmasses. I Int. Symp. on Storage in ewcavated rock caverns.Rockstore, 77. Stockholm.

REVUEFRANCAISEDEGEOTECHNIOUE No23 PARIS 1983

General report topic 4 I Rapport général thème 4

ESSAIS DE CHARGEMENT, DE DECHARGEMENT ET DE CISAILLEMENT REALISES A LA SUR-FACE DES SOLS ET DES ROCHES

LOADING. UNLOADING AND SHEARING TESTS ON SOIL AND ROCKS

GARNEAU R.I

Introduction

Le nombre de communications techniques présentées à ce

symposium sous le thème "Essais de chargement, de déchar-gemént et de cisaillement réalisés à la surface des sols etdes roches", soit environ une trentaine, témoigne du rÔle

important que la communauté internationale des géotech-niciens accorde à ces méthodes d'essais et d'observationsd'ouvrages en vraie grandeur ou de grandeur réduite.

Dans la plupart des cas, des essais de chargement ou de cisaillement sont effectués sur place lorsque la nature du solou de Ia roche, sa stratigraphie ou son état ne permettentpas d'extrapoler les résultats d'essais de laboratoire oumême d'effectuer les essais requis sur des échantillons re-présentatifs, parfois très difficiles à obtenir. Aussi, en vuede pouvoir établir avec suffisamment de confiance lecomportement d'un ouvrage à construire ou à modifier,ou les pararnètres gouvernant ce comportement,'le géotech-nicien fera souvent appel à des méthodes d'essais en place.Il choisira alors, suivant I'ampleur du problème, suivant les

moyens financiers disponibles et en fonction du degré de

risque acceptable, d'effectuer des essais sur un modèle rê-duit de fondation, sur une partie de I'ouvrage à construireet dans certains cas, sur l'ouvrage lui-même terminé.

Les communications techniques faisant I'objet de ce rap-port général couvrent ces différents aspects des essais de

chargement à la surface des sols et des roches ; les sous-thèmes suivants découlent des sujets traités :

Sous-thème 4.I : Essais de chargement sur plaquesSous-thème 4.2: Essais de chargement en vraie grandeurSous-thème 4.3 : Essais de cisaillement in situSous-thème 4.4 : Essais ponctuels de surfaceSous-thème 4.5 : Mesures des contraintes à I'intérieur des

massifs

Le nombre de communications se rattachant à chacun dessous-thèmes est comme suit :

Sous-thème Nombre de communications

l18

3 + 1 du sous-thème 1

23

Le tableau ci-dessous présente la liste des communicationsgroupées sous chaque sous-thème, incluant les noms desauteurs et quelques mots clés permettant de préciser le sujettraité. Certaines communications se sont avérées plus va-lables que d'autres et quelques-unes ont pu échapper à labonne compréhension du rapporteur à cause de difficultésde rédaction de nature linguistique et, dans certains cas, del'absence de figures accompagnant les textes.

I2345

TABLEAU DES COMMUNICATIONS

Sous- Auteursthème (PaYs)

1 CHIN F.K.(Malaisie)

Titre

"Bilateral plate bearing test"

"Site investigation and foundationdesign in stratified soil"

"A geotechnical apparisal of in situdeformation tests at Jakham Dam,India"

Mots-clés

Chargement sur plaques I'une contreI'autre, effet de grandeur, relationhyperbolique, dimensionnement defondations, sols résiduels

Profondeur d'influence de fonda-tions, sols bi-couches, reconnais-sance géotechnique

Chargement sur plaque carrée, rocheà degré d'altération et de fissurationvariable, modules de déformationsécant et tangent, effet de satura-tion, relation avec géologie locale

HANNA(Canada)

KASLI\ryAL(Inde)

JORDEN E.

* Chef du Service géotechnique Rousseau, Sauvé, IVarren [nc. Montréal, Québec, Canada.