June 08 Complete

Grout Line

OvertureHere we are again for the twelfth ap-pointment with our Grout Line.

For this issue we have 2 articles, anda short comment related to one of the ar-ticles. Both articles are related to previ-ous articles published in past editions ofthe Grout Line. I remind you here thatall the past articles of the Grout Line canbe downloaded at www.groutline.com/Articles.htm.

The first article is the answer of Dr.Lombardi (Lombardi SA Engineering

Limited, Minusio, Switzerland,[email protected]) to the article pub-lished on the Grout Line- GeotechnicalNews Volume 25 Number 3- on Sep-tember 2007 by Dr. Shuttle and others.This was consequent to a previousarticle of Dr. Lombardi published on theGrout Line- Geotechnical News Vol-ume 25 Number 3- on September 2007.

Dr. Shuttle also sent me a short finalcomment that you can find at the end ofDr. Lombardi’s article.

The controversy about the GIN

(Grouting Intensity Number) is contin-uing!

The second piece is Jim Warner’sclarification of the article published inthe March ’08 issue related to the differ-ence between Low Mobility Grout andCompaction Grout.

I wait for your grouting papers, arti-cles or comments. Please send them to:Paolo Gazzarrini, fax 604-913-0106 orpaolo @paologaz.com, [email protected] or [email protected].

Ciao!

Misunderstanding of GIN Confirmed

Giovanni Lombardi

IntroductionIn a paper published by “GeotechnicalNews” “Grout line” [1] titled “GINAgain Misunderstood” the writer pre-sented some comments to the paper“Penetrability Control of GIN Mixesduring Fractured Rock Grouting” byVafa Rombough, Grant Bonin andDawn Shuttle [2]. In response to thesecomments the paper “GIN Distilled” bythe same authors was published, also inGrout Line with date September 2007[3].

As this last paper contains againsome misunderstandings, the followingconsiderations will refer to them, but

will provide also some considerationsof more general nature.

GIN as LimitIn the intent to show the presumed “lim-its of GIN”, figure 2 from the AntaminaDam, was presented in paper [3]. Thesame is reported hereafter with its origi-nal legend, as Figure 1.

Unfortunately, its meaning wasagain misunderstood and was the occa-sion to invent the amazing tale of “mul-tiple GIN closures”. It was actuallyassumed that the so- called “closure”was reached 4 times at the points A, B,C and D, where the GIN-line was

touched by the pressure path.The question is thus: why only 4

times?In fact the GIN line could have been

easily reached and even crossed notonly 4 but 40 or more times1. Evenmore, if said GIN line would have beenfollowed continuously from a point of“high pressure - low volume” to a pointof “low pressure - high volume”, tosome final value, then the two lineswould have coincided an infinity oftimes.

Fortunately, the 4 mentioned pointsare not “closure points” at all, becausethe flow rate at these instants was about

Geotechnical News, June 2008 3

THE GROUT LINE

1 In fact the same grout path crosses easily 40 times a somewhat smaller GIN line as the one shown on figure 1.

20./12./9. and 6. L/min respectively.As one should know - and this is re-

peated again - the flow rate at the “clo-sure point” should and will be nil. Onlythen one can talk of “closure.”

From the graph of Figure 2 one may,in extrapolating the flow-rate tendency,easily conclude that the nil flow ratewould have been arrived at for a volumeinjected of about 400 l/m. Therefore theclosure would have occurred at thepoint E (or nearby), which point washowever not reached with the groutingworks, and which is the unique real“closure” point in spi te of thementioned tale.

In the case of figure 1, the groutingwas stopped too early, that is beforereaching the point of “closure” E. Thismeans simply that the concept of the

GIN-method was not really understood,already at the time of grouting. Only300 l/m were injected instead of theabout 400 l/m, which should have been.The real closure point was missed by33% of volume.

This point outlines again one of thesignificant aspects of the GIN value anddoesn’t need many comments, exceptthat it was overlooked both by the grout-ing engineer at site and the writers of thepaper from which figure 1 was taken.

By the way, this point E could havebeen reached in following the GIN line(which task requires however some skilland an adequate equipment, but whichcould be considered to be the mostlogical one). The point E could alsohaving been reached from below or

from above the GIN line (provided, inthis last case, that the volume takenwould not overpass the correspondingestimated value, e.g. 400 l/m) (see Fig-ure 3).

Indeed the GIN line is a target to bereached not a “wall” which could not becrossed. However, frequent cases aremet where certain GIN values representa physical limit, which can hardly becrossed. This happens when an impor-tant hydro-fracturing, hydro-splittingor hydro-jacking takes place.

One knows, from the theory and thegrouting practice, that when the pres-sure suddenly decreases at an almostconstant flow rate, a phenomenon ofthis kind does occur. Said decrease ofpressure, often at a constant GIN value,may be of a limited extent. This means

THE GROUT LINE

4 Geotechnical News, June 2008

a) = Pressure vs. Volume curveb) = Measured flow rate vs. injected volumeExample grout injection data from Antamina Dam showingmultiple GIN “closures “ [234BP 27.6-32.6 m] adaptedfrom Richie and others, 2003).

Figure 1. As presented in “GIN distilled” [3] as figure 2; withits original legend.

Figure 2. The real “closure” for the example shown onfigure 1 takes place at the point E. In fact, the unique closurewould have taken place at the point E, or near of it, at a flowrate nil. The grouting was stopped before reaching the realclosure.

that the hydro-fracturing discontinuityitself is of a limited length because cutby other more stable fractures or by dif-ferent rock layers.

During a grouting process a series ofsuch “local” limited hydro-fracturingscan therefore be observed on alwayshigher intensities, that is on higher GINvalues.

In some cases, however, a major hy-dro-fractur ing along a maindiscontinuity can take place.

The meaning of the GIN line as alimit for hydro-fracturing was de-

scribed in paper [7], from where Figure4 is taken.

The fact that hydro-jacking takesplace by increasing volume and de-creasing pressure is easily explained,because the GIN value - that is thegrouting intensity - represents roughlythe energy pumped into the rock mass,and because - like many other instabili-ties - the hydro-jacking event takesplace at a practically constant energylevel (see [7]).

It may also be mentioned that, asshown in Figure 5, when the pressure

path follows the GIN line two cases mayhappen:

Case 1: the flow rate tends to dimin-ish with increasing volume injected. Inthis case a normal grouting takes place.

Case 2: the flow rate remains more orless constant, then an important hy-dro-splitting is likely to occur.

In this last case the correspondingGIN line represents a barrier, whichcannot, or at least should not, beover-passed. Even more the continua-tion of the grouting can practically takeplace only along said GIN line. Increas-ing the power of the pump leads mainlyto an increase of the flow rate, while thepressure still decreases.

Reducing that power leads to a re-duction of the flow rate and soon to astoppage of the injection. Conse-quently, the GIN value has an additionalmeaning when hydro-jacking or hy-dro-splitting takes place.

It should thus be evident that the con-clusions drawn in paper [3] do not hold,and that an additional merit of the GINconcept has to be recognized, in thesense of a protection against excessivehydro-fracturing of the rock mass.


THE GROUT LINE

— . — . — . the quite usual path__________the possible theoretically best path— — — — —a possible but risky strategy, and++++++++a possible path if the final take can be esti-mated (see fig. 2)In nearing the point E it is always advantageous, in anycase, to overpass somewhat the GIN line so not to delay use-lessly the conclusion of the grouting works.

Figure 3. The ways or strategies to reach the unique, a prioriunknown, “closure point” E.

a) The “historical” grouting limits for pressure and take.b) The GIN limits for pressure, take and intensity. The main scope of limiting theintensity is to reduce the risk of hydrofracturing and to avoid useless grout losses.

Figure 4. The limitations of grouting.

Figure 5. Indication of hydro-splitting.Case 1 = closure to be expected (normal grouting)Case 2 = hydro-splitting quite likely to occur

GIN “Closure”It appears, further that in the back-ground of many statements about theGIN-method, the opinion prevails thatgrouting by that method must take placein using a constant, possibly very lowflow rate. It is thus stated in [3]: “hencein practice GIN becomes unreliable tocontrol practical grouting”. Fortu-nately, this is an additional confirmedmisunderstanding.

It was not seen by many that figure 3of paper [5] titled “Actual grouting pathand grouting intensity (GIN)” as well asthe concept of over-passing it by, for ex-ample 10%, have just the intent to ex-plain the question. (Possibly, theexplanation was not clear enough) (seeFigure 6).

It is said also, by the authors of paper[3], that “in a typical GIN grouting pro-ject you pump at a constant rate”2.

For sure, there are worldwide a num-ber of designers, who specify a constantand even a low flow rate, when referringto the so-called GIN grouting. (Hope-fully, they know the reason of!). But,such a specification and similar ones areindependent of the GIN principle and

are by far not always optimal.The right way to proceed is shown at

the reverse in paper “Grouting designand control using the GIN principle” [5]published already in 1993. It consists tosteer the grouting process in taking intoaccount the “penetrability” defined byq/p (flow rate divided by grouting pres-sure), which value has obviously to beshown to the operator at real time on thedisplay (see Figure 7).

In fact, the penetrability must be nilat the closure point; thus also the flowrate. As already shown, this principlewas not understood in the case ofAntamina shown (as well as in a num-ber of other jobs).

The principle to follow consists sim-ply in adapting progressively the pres-sure and thus the flow-rate, when theGIN value (at nil flow rate) accordinglyto the flow path, approaches the desiredvalue.

This way of doing allows a certainextrapolation in time of the decrease ofthe penetrability so there is no need toselect an “extremely low flow rate” longbefore stopping the process (see againfigures 2 and 3).

To avoid new misunderstandingssome additional explanations appear tobe necessary.

An example is shown on Figure 8 forthe case the pressure is maintained con-stant, which means that it is changedonly stepwise.

No matter how the grouting did pro-ceed until the point A. At that instant,the pressure p1 and the flow rate q1 werearrived at and the volume V1 was al-ready injected.

If the pressure were kept constant, itsline would cross the GIN line at thepoint B with a value VB for the take.However, a check shows that the de-crease of the flow rate would take placequite earlier and the extrapolation of itwould lead to a stop at the volume VC,which is smaller than VB. The two val-ues VB and VC should obviouslycoincide.

The pressure must thus be increasedat the point 2 so that the correspondingvolume, VB, will decrease while the ex-pected one, VC, would increase reduc-ing the gap (VB - VC) between them.

If necessary, a further adjustment ofthe pressure can be done a while later.

THE GROUT LINE


Figure 6. The grout process can be stopped at any final pres-sure prescribed or by reaching any required GIN value (there isno such situation as “refusal” by the rock).

________actual pressure path- - - - - - theoretical pressure path at flow rate nil (the viscos-ity related term of the pressure vanishes and only the cohe-sion related one remains)

Figure 7. Grouting process of a single borehole stage (typical)where:

1. limiting curve; pressure versus grout take2.actual grouting path; pressure versus grout take3. penetrability (q/p) versus grout takeF final point of groutingpF final grouting pressureVF actual (final) grout take

2 This was apparently not the case in Antamina according to figure 1!

At the reverse, if VC would appear tobe greater than VB, the pressure shouldbe reduced.

The procedure can be improved inreferring to the penetrability, (q/p), in-stead of the flow rate, q, itself. In doingso, at least certain irregularities aresmoothed and the process made moreprecise and easy to steer (refer to figure7).

A way to speed up the procedure, asalready presented in paper [7], consiststo target a somewhat higher (10%) GINand to stop when the flow rate falls be-low a certain small value.

It may be easily understood that thedescribed procedure can be automa-tized in using an adequate software.Similar procedure can also be devel-oped if different grouting methods wereadopted.

By the way, the following consider-ation is due.

The GIN value will be decided on thebase of some grouting tests in a rockmass similar to the one, which has to be

actually grouted; but, obviously not forany single borehole nor for any singlegrouting stage. Therefore it is not nec-essary to try to reach some theoreticalabsolute precision and therefore sometolerance has to be accepted. Othermeans, as for example additional bore-holes series, are available to compen-sate for too important scatters in therock mass properties.

This way of doing is by far not a“poor guidance” and the statement “inpractice GIN becomes unreliable tocontrol practical grouting” does nothold. In fact, the practical experience onmany sites confirms the good, if not ex-cellent, guidance it provides. GIN is ac-tually a target, which can be missed asmany times as one wishes. However,this is a problem of the shooter not ofthe target!

Also there are, for sure, many peoplewho don’t know how to steer correctlythe GIN grouting process.

At the end of the day, the actualphysical penetration distance reached

by the grout depends on the pattern ofthe discontinuities, on the grout volumeabsorbed, and on the cohesion-relatedterm of the pressure at the moment theprocedure is stopped. Indeed, when thepump is turned off the flow rate be-comes nil and the viscosity relatedterms disappear.

This is exactly the definition of theGIN value.

In fact, the actual pressure path, theflow rate and their variations during thefirst part of the grouting process are of aquite secondary importance, exceptthey may influence the duration of theprocedure.

In fact, in accepting the GIN princi-ple, higher grouting pressures, and thushigher flow-rates as usual can be usedas long as some distance from the tar-geted GIN value is kept. This allowsspeeding up the grouting process. Thesecond part of the grouting process wasalready discussed.

Definition of the GIN ValueIt should be clear that due to the ex-tremely high number of unknowns ex-isting in the actual rock mass, as typesof discontinuities, orientation, opening,deformability, rugosity, frequency,interconnections, infill, etc. no possibil-ity does exist to define a priori on a theo-retical base the parameters to be se-lected for the grouting process.

The only realistic way is to carry outtest fields for each zone of the rockmass, which may reasonably be consid-ered to be homogeneous, as explainedin paper [6].

The split-spacing method should beused in any case both for grout curtainsand for foundation consolidation. Thecriteria to judge the results are the de-crease of the take from series to series ofboreholes while keeping a constant GINvalue.

In matter of the grout reach it was notalways clearly distinguished betweenthe GIN value computed for a singlejoint of constant opening and the valueobtained by a grout test in an actual rockmass.

Therefore the statement “GIN over-estimates the grout reach” cannot haveany meaning when referred to a grouttest, in a real rock mass, nor when re-


THE GROUT LINE

Figure 8. Example of a way to steer the grouting process.p = grouting pressureq = flow rateVB = volume theoretically taken at a pressure p1 (VB = GIN/p1)VC = volume take estimated in extrapolating the decrease of the flow rate(better the decrease of the penetrability = q/p)If VC<VB then increase the pressure, so VC increases and VB decreases, reduc-ing the gap (VB-VC).

ferred to a single joint as used for theo-retical considerations. In fact, the for-mulae of paper [3] confirm the reachgiven by paper [4] in the theoreticalcase.

The figure 4 of paper [3] showing thegrouting of two parallel joints of differ-ent openings with different reaches,which was already included as figure 12in paper [4] refers simply to thefundamental rule that the grout reach (atnil flow rate) depends linearly on theopening of the discontinuity. Said fig-ure is thus meaningless in the presentdiscussion and just confirms an addi-tional misunderstanding.

It should be known that in the case ofan actual treatment of a rock mass;• the primary holes will fill mainly the

widest joints to the largest distance;• the intermediate secondary holes

will fill the widest of the remainingjoints to quite a distance because,due to the minor take, the same GINvalue will lead to a higher groutingpressure;

• the same happens successively withthe intermediate tertiary, quaternaryand possibly further grouting series;

• the success of the grouting is shownby the decreasing take from bore-holes series to borehole series;

• if such a decrease, from borehole se-ries to borehole series, doesn’t takeplace, an escape of grout, e.g. to aCarstic cavern, or a very importanthydro-jacking (or fracturing), isquite likely to have occurred.

Radial against Strip FlowComputationsThe “channelling” of grout or wateralong the rock discontinuities is by farnot a new concept. It was, for example,analysed in the paper “The FES RockMass Model” 1992 [8]. See Figure 9.

However, this effect cannot be calledin to justify the use of a strip of constantwidth to analyse the grouting process,except the “channels” were all paralleland independent each from anotherwithout any bifurcation or crossing.

The experience shows that, as a rule,

the joints and fissures are filled by thegrout in all directions. The model of aradial flow is thus a bit more realisticthan the one of a strip flow3.

The weak point of the strip model isthat the “specific flow rate “, that is theflow by unit width is assumed to be con-stant all along the discontinuities fromthe borehole to the border of the groutedzone, while it is obvious that this spe-cific flow-rate decreases strongly withthe distance from the borehole. (Ap-proximately with 1/r in function of theactual reach.) As consequence of thestrip-model the pressure gradient ap-pears to be constant and thus the pres-sure to decrease linearly from theborehole, while in fact the pressuredrops very strongly near the boreholeand is very significantly reduced atdistance.

The linear distribution is reachedonly when the process is stopped andthe flow rate is nil everywhere.

For this reason, at this moment, thetheoretical final reach and the pressuredistribution are the same, whether com-puted with the strip or with the 2-Dmodel.

In the reality, the grouting process issometimes more a 3-D, than a 2-D prob-lem; but for sure not a 1-D one.

In any case the statement “the GINmethod does overestimate the reach ofthe grouting” is the consequence of thewrong assumption that the method im-plies a low; near nil flow rate, that is alow grouting pressure and thus a longgrouting time. The contrary is true be-cause in keeping a sufficient distancefrom the “dangerous-zone”, accord-ingly to figure 4, higher grouting pres-sures can be used at the beginning of theprocess, which will be successively re-duced, as already explained.

One of the consequences of the as-sumption made in using the strip modelis that the splitting force produced bythe grout is wrongly computed, that isoverestimated until the end of the grout-ing, where the flow rate nil corresponds

THE GROUT LINE


Figure 9. FES model of a fissure along its plane (for hydraulic mathematical analy-sis).

a) = pattern of channels and “islands “b) = cross section of a channelQ1 = flow through a channel1. = contact zones2. = flow channel3. = semi-rigid core (in the case of a cohesive fluid)V = velocity in the case of a Bingham’s grout

3. To consider the effect of “channelling”, a factor k1 is introduced in the formula for GIN so to take care of the longer path to reach a given pointthan the radial distance from the borehole. To take into account also the difference between the volumetric averaged and the hydraulic averagedopening of a discontinuity a factor k2 is also considered. Neverthless, as already said, the optimal value of GIN must be defined on the base ofgrout test fields.

to the value obtained by the radialflow-model.

The strip model is thus misleading inevaluating the risk of hydro-fracturingthe rock. See [1].

Flow Rate ComputationIn the paper by Dawn Shuttle and othersit is claimed that the analytical formulafor a radial flow does not exist. It couldat the best be developed under very re-strictive assumptions (e.g. constantpressure or flow rate). In fact, such aformula is not necessary at all becausethe following steps can be easily carriedout• at a given grouting flow rate the

“specific flow” slows down from theborehole with the ratio (1/r), pro-vided the possible opening of thediscontinuity due to the pressure it-self is disregarded (no hydro-jackingconsidered);

• consequently, the gradient of thepressure drop can be computed atany point using the properties of thegrout that are cohesion and viscosityof the Bingham body;

• in integrating that gradient from themomentaneous border reached by

the grout towards the borehole thecorresponding grout pressure can beobtained;

• the flow rate can then be changed asrequired, for example in order to re-spect the characteristics of the grout-ing pump or any other requiredsteering procedure, in particularwhen the grouting process isstopped;

• as the integration is carried out nu-merically, any kind of grouting pro-cedure can be considered, in contrastto any analytical formula, whichcould, at the best, take into accountonly a predefined one.It is agreed, that the procedure is less

simple that an analytical formula, buthas a much more general meaning andis significantly more flexible.

So, the example shown in paper [4],1985 and represented in Figure 10 wascomputed accordingly.

It is thus quite difficult to imaginehow “to give Bingham a try” as paper[3] proposes in its conclusion, becausein any case the computation of the pres-sures and of the flow rate is based on theproperties of the Bingham body.

Of course, there is nothing against

any proposal in order to improve theGIN procedure. However a realisticway to do so has not been suggested yet.

For sure, the improvement can notconsist to come back to historical crite-ria as, for example, to reach any a priori,arbitrarily specified constant pressurein ignoring the risk of hydro-fracturingthe rock mass and thus to waste anyamount of grout if such a phenomenondoes occur.

Also, any arbitrarily given maxi-mum take can not be of any help if notput in relation with the pressure re-quired to reach it, that is to the groutingintensity.

ConclusionsIt is well known, that there are world-wide quite a number of people who donot like and do not use the GIN methodof grouting. They have, of course, theright to do so, even if this is not in theinterest of the job, nor in that of theowner. The reasons of this fact are in-deed numerous.

A first one is that GIN disturbs cer-tain, sometimes century old, habits andways of thinking.

Very interesting historical cases arealso referred to, without taking into ac-count how normal or exceptional theywere.

Procedures and ways of thinking de-veloped for unstable mixes continue tobe applied when using stable ones.

Some operator prefers indeed to fol-low the principle “don’t think, justclick” or introduces new rules so to get a“simplified procedure”, as for exampleto define a GIN value on the designboard instead of carrying out thenecessary grouting tests on the site.

A list of frequent mistakes can befound in paper [7].

It is also quite usual to make any kindof errors during the grouting operationand then to criticise the GIN method in-stead of recognising the mistakes made.The case of Antamina is clearly one ofthem.

Often also, the GIN method is usedto grout carstic or similar formations forwhich it is not intended nor applicable.It should be used only at a second stage,when the carstic cavities have alreadybeen filled, e.g. with mortar.


THE GROUT LINE

Figure 10. Pressure distribution along the joint at different time intervals duringgrouting.

On a given dam the lack of commu-nication between the pumping stationand the borehole mouth led to frequentexcessive grouting in over-passing theGIN line because the pump was turnedoff too late. As usual, not the malfunc-tioning of the communications but theGIN method itself was considered re-sponsible for the mishaps and thiswrong conclusion was largelypublished.

Finally, people, even very special-ized in some particular field, try to leada grouting campaign in spite of a lack ofan overall knowledge of this kind ofwork in the theoretical field or in thepractical operation.

Indeed, grouting is a quite special ac-tivity, which requires both theory andpractice, as well as some thinking.

The overall conclusion is thus thecounsel to try to apply seriously GIN;one will finally like it. Indeed, the onlyproblem with GIN is that it requires tobe understood.

Bibliography1. G. Lombardi, “GIN again misunder-

stood”, Geotechnical News-GroutLine, Vol 25, Nr. 2, June 2007.

2. V. Rombough, G. Bonin and D. Shut-tle, “Penetrability Control of GINMixes during Fractured RockGrouting”, Proceeding of Sea to SkyGeotechnique, 59th CanadianGeotechnical Society ConferenceOctober 2006-pp 528-535.

3. D. Shuttle, V. Rombough and G.Bonin “GIN distilled”, GeotechnicalNews-Grout Line, Vol.26 Nr.3,September 2007-GeotechnicalNews.

4. G. Lombardi, “The Role of Cohesionin Cement Grouting of Rock”,ICOLD. 15th Congress on LargeDams, Lausanne, 1985, Q.58 R.13,pp. 235-261

5. G. Lombardi and D. Deere, “Grout-ing Design and Control using theGIN Principle”, Water Power &Dam Construction, June 1993, pp.15-22

6. G. Lombardi, “Selecting the Grout-ing Intensity”, The InternationalJournal on Hydropower & Dams,

Volume Three, Issue Four, 1996, pp.62-66

7. G. Lombardi, “Grouting of RockMasses”, 3rd International Confer-ence on Grouting and Ground Treat-ment, New Orleans, February 9-12,2003

8. G. Lombardi, “The F.E.S. Rock MassModel - Part 1”, Dam Engineering,Vol. III, Issue 1, February 1992, pp.49-76, “Part 2”, Dam Engineering,Vol. III, Issue 3, August 1992, pp.201-221Dr. Lombardi’s article references

several articles published on the GroutLine and from himself. You can down-load these articles (of course for free!)from: www.groutline.com/Articles.htm(article published) - www.lombardi.ch(publications).

Dr. Eng. Giovanni Lombardi, LombardiSA, Via R. Simen 19, 6648 Minusio (TI),Switzerland, Tel: +41 91 744 60 30,Fax: + 41 91 743 9737email: [email protected]

THE GROUT LINE


An Alternative Viewpoint on GIN

Dawn ShuttleVafa RomboughGrant Bonin

We would like to commend GiovanniLombardi [1] on this more complete ex-planation of the GIN method. A prob-lem with GIN is that it is extremely dif-ficult to separate the writings ofLombardi (covering a range of grouttopics) with an actual GIN methodol-ogy that can be applied by a groutingengineer in the field. This new presenta-tion of ideas by the author contains themost complete summary of GIN wehave seen, including the effects of groutinjection rate which was poorly clari-fied before. We are sure it will be of useto the grouting community.

We are pleased to find that Lombardinow agrees with the primary technicalpoint raised in our original conferencepapers [2, 3] and subsequent Grout Linediscussion [4] – specifically that groutinjection rate does matter. Contrary tothe misleading tone of [1], there is notechnical point on which Lombardi dis-

agrees with our previous work. Thatsaid, we do continue to disagree withGiovanni Lombardi on some funda-mental issues.

Grout Penetration GeometryReal fracture networks are neither idealradial flow nor perfect linear flow (al-though both of these simplifications areequally useful for illustrating the effectof flow rate on grout penetration).Discussion on the relative merits of ra-dial flow versus strip flow is irrelevant.The way forward is to realistically rep-resent the true fracture network. This isnow possible using discrete fracturenetwork software (e.g. reference [5]).This approach, applied to a fracturedrock grouting project, is illustrated inFigure 1.

Misleading Guidance on GINIf applying the GIN approach, the meth-odology proposed to obtain the ‘nil’flow rate required by GIN in Section 3of Lombardi’s article [1] is misleading.Specifically the suggestion, illustratedin Figure 6, that by overshooting GINby 10% and then reducing the pressureyou will easily intersect the specifiedGIN curve at ‘nil’ flow rate. The effectof the 10% overshoot was previouslydiscussed in Shuttle et al. [4] using theAntamina grout data of Ritchie et al [5],and the same figure is also included inLombardi’s article as Figure 2 (and alsoreproduced here as Figure 2). If the 10%lower GIN is set as our target GIN it isclearly seen that this GIN is exceeded atA, B, C and D, and as confirmed by


THE GROUT LINE

Figure 1. Cross section of transmissivity distribution throughgrouted discrete fracture network (grout curtain shown as darkgrey) (from [5]).

Figure 2. Example grout injection data from Antamina Damshowing multiple GIN “closures” [234BP 27.6-32.6m](adapted from Ritchie et al, 2003).

Administrator

Cross-Out

Administrator

Replacement Text

6

Lombardi [1], even at D this is not at‘nil’ flow rate and hence not at GIN.Therefore the ‘practical’suggestion thatby overshooting GIN by 10% you canquickly obtain the required ‘nil’ flowrate is erroneous.

Similarly, Lombardi’s Figure 7 ofthe “grouting process of single boreholestage (typical)” is grossly simplistic.The figure suggests that it is (a) typicalto intersect the GIN curve at zero flowrate, and (b) that all penetrability (q/p)versus grout take curves show a gradualdecay with increasing take. This is un-true. As previously shown in Figure 2,and stated at length in [1], most first in-tersections of the GIN curve do not co-incide with ‘nil’ flow. Figure 3 shows apenetrability curve plotted for anAntamina grout injection stage. Twomixes were used. The first mix was astandard low viscosity grout typicallyused in GIN grouting. Negligible, ifany, reduction in penetrability was ob-served and the grout ‘flowed away’.The second mix was a higher viscositygrout, and a reduction in penetrabilitywith continuing injection is seen. Ex-cept, unlike the idealized version in Fig-ure 7 of [1], the penetrability suddenlydrops to zero as refusal is met. There isno gradual reduction, and the recom-mendation to extrapolate the penetra-bility curve to zero flow rate iserroneous.

We would also note that GIN recom-mends a single grout mix [7]. For thegrout stage shown in Figure 3, mix ‘A’(the lower viscosity grout) ‘flowedaway’ and continued injection of thistoo thin grout would have been costly.We believe a more economic solutionwould have been achieved by modify-ing the mix and volume injected basedon water pressure test results.

Lombardi’s Figure 8 builds uponFigures 6 and 7, so although the engi-neer knows that GIN and a ‘nil’ flowrate must be reached coincidentally,there is no confidence that the method-ology proposed for reaching this state isachievable.

A Misunderstanding of GIN? OrDoes GIN Mislead?Lombardi’s reply to our publicationssuggests that the lack of success of theGIN method is solely down to a misun-derstanding of GIN, and that any com-petent engineer should be able to under-stand and apply GIN successfully. Wedo not agree.The purpose of rock grouting is to fillapertures within a pre-defined radius ofthe injection grout hole, the radius be-ing based on the chosen grout hole spac-ing. For a given radius and a set of frac-ture apertures, this equates to a groutvolume to be injected. Flow rate andpressure are merely the means to getthis volume in place. The faster this can

be achieved (without hydrofracture) themore economic the project – hencethere is significant advantage in mini-mizing the use of low flow rates. Thereis really no reason to obtain a conditionof “nil flow rate” at the end of injection.

The question for the grouting engi-neer is the connectivity of the fracturenetwork intersected and the fracture ap-ertures. Because each grout stage inter-sects a specific part of the fracturenetwork, with particular rather than av-erage properties, an optimum groutingapproach is to relate the unit take crite-rion to the measured penetrability andgrout rheology. This is a very differentstrategy to the GIN approach to grout-ing. It is not a question of misunder-standing GIN.

Despite the new clarifications ofwhat the GIN method comprises, thebasic limitation of GIN remains. Zeroflow rate is confused with the desiredradius of influence, often leading to ex-tended injection periods at very lowflow rates to no effect. In contrast, bas-ing a grouting protocol on the measuredpenetrability of each grout stage pro-vides the grouting engineer with aclearly defined basis to bring a curtainto closure. Necessarily this requiressome understanding of the fracture net-work being grouted, but we do not con-sider this a bad thing. Thus we assertbasing an entire grouting program onGIN has the potential to mislead thegrouting engineer. It is not a misunder-standing of GIN – GIN can indeed mis-lead.

References1. Lombardi, G. (2008), “Misunder-

standing of GIN confirmed”, TheGrout Line, Geotechnical News,June 2008.

2. Rombough, V., G. Bonin, and D.A.Shuttle (2006). ‘Penetrability con-trol of GIN mixes during fracturedrock grouting’. Proceeding ofSea-to-Sky Geotechnique: 59th Ca-nadian Geotechnical Society Con-ference, Vancouver, Canada, 1-4October 2006, pp. 528-535.

3. Shuttle, D.A., V. Rombough, and G.Bonin (2007), “Impact Of GroutRheology on GIN”. Proceedings ofthe mini-symposium on grouting,

THE GROUT LINE


Figure 3. Example penetrability (q/p) versus grout take record from Antamina.

Administrator

Inserted Text

with increasing take

Administrator

Cross-Out

Geo-Denver 2007: New Peaks inGeotechnics, 18-21 February, Den-ver, CO.

4. Shuttle, D.A., V. Rombough and G.Bonin, (2007) “GIN Distilled”. TheGrout Line, Geotechnical News,September 2007.

5. Shuttle, D.A., W. Dershowitz, E.Glynn, S. Burch, and T. Novak(2000) “Discrete Fracture NetworkAnalysis of Foundation Grouting”.in Girard, Liebman, Breeds and Doe(eds) Proceedings of the FourthNorth American Rock MechanicsSymposium, Pacific Rocks 2000. 31July-3 August. Balkema, Rotterdam,pp 1369-1376.

6. Ritchie, D.G., Garcia, J. P., Amaya, Fand Jefferies, M.G. (2003) ‘CurtainGrouting for the Antamina Dam,Peru: Part 2—Implementation andField Modifications’. Proceedingsof the third international conference

on Grouting and Ground Treatment -Geotechnical Grouting and DeepMixing. Edited by L.F. Johnson,D.A. Bruce and M.J. Byle, pp.929-940, 10th to 12th February 2003in New Orleans, Louisiana.

7. Lombardi, G. (2003), “Grouting ofRock Masses”, Proceedings of thethird international conference onGrouting and Ground Treatment -Geotechnical Grouting and DeepMixing. Edited by L.F. Johnson,D.A. Bruce and M.J. Byle, pp.164-197, 10th to 12th February 2003in New Orleans, Louisiana.

Dawn Shuttle (corresponding author),Consultant, White Rock, Canada V4B1R2, [email protected]

Vafa Rombough, Junior Engineer,Golder Associates Ltd, 500 – 4260 StillCreek Drive, Burnaby B.C, Canada

V5C 6C6, [email protected]

Grant Bonin, Intermediate Engineer,Golder Associates Ltd, 500 – 4260 StillCreek Drive, Burnaby B.C, CanadaV5C 6C6

I have a final short comment about thiscontroversy about the GIN method.

The paper and further articles/com-ments from Dawn Shuttle and others aredirected only to criticize drastically theGIN method. I have a short question forthem: what is their proposal for proce-dures, pressures, grout mixes, flow etcfor grouting in rock? Also this can be agreat contribution to the grouting com-munity, and specifically to the groutingengineer on the field!

It is easy to criticize, but what alter-nate do they have to offer?

Low Mobility Grout (LMG); CompactionGrout (CG) Clearing the Confusion

Jim Warner

As so clearly illustrated by Vanderpool,Norris, and Elfass in the March, 2008Grout Line (Warning: Low MobilityGrout (LMG) May Not Equal Compac-tion Grout (CG), not all low mobilitygrouts can be effective for soildensification by compaction grouting.The ASCE Geo-Institute Glossary ofGrouting Terminology (2005) defineslow mobility grout (LMG) as “Lowslump grout, such as compaction-typegrout, that does not travel freely andthat becomes immobile when injectionpressure ceases”. Whereas compactiongrout is a low mobility grout with spe-cial properties, and is typically sitebatched and mixed, much low mobilitygrout is centrally batched and transitmixed, being delivered to the jobsite inready mix trucks. Although this offersthe advantage of easy availability andreasonable cost, it limits the grout com-position to ingredients normally

stocked and used in the production ofconcrete, which requires propertiesquite different from those typically de-sired for grout. None the less, it is quiteeasy to compound a low mobility groutusing available constituents, includingviscosity modifying and concretepumping aid admixtures. Though oflow mobility, most such mixtures willbehave as a fluid if injected in theground however, resulting in hydraulicfracturing of the soil, and precludingcontrolled compaction. Thus, such eas-ily obtained and economical low mobil-ity grouts are usually not satisfactoryfor compaction grouting.

Compaction grouting involves injec-tion of a special stiff mortar-like groutinto soil. It does not fill voids but ratherexpands as a globular mass under pres-sure so as to uniformly displace anddensify the adjacent soil. Should thegrout behave as a fluid, hydraulic frac-

turing of the soil will likely occur, pre-cluding controlled compaction. In loosesoils, some densification will occureven with fluid behavior of the groutand resulting hydraulic fracturing, butthe improvement will be only a smallportion of that obtainable by propercompaction grouting, as so vividlydemonstrated in the cited work. Toachieve the compactive mechanismthus requires a much more preciselycompounded low mobility grout.

This does not mean that plain lowmobility grouts should never be in-jected in soil. As pointed out by Gomez,Robinson, and Cadden in the June 2006Grout Line (Use of Limited MobilityGrout for Shallow Foundations inKarst) “The LMG technique, often er-roneously referred to as compactiongrouting, can help meet the needs of theproject and the engineer in a cost-effec-tive manner” and improvement can be


THE GROUT LINE

achieved “through densification or dis-placement of the very soft soils (empha-sis added) above and near the rocksurface” (in Karst terrain). LMG supplyand injection is simpler and less costlythan compaction grouting and while itwill not provide near the degree of im-provement to the soil, it is a valid appli-cation where very loose materials existand a high standard of improvement isnot needed.

Production of compaction grout thatwill behave as a growing solid in theground severely limits the mix constitu-ents, and any that might contribute tofluid behavior must be avoided. Unfor-tunately, fluid behavior is the very prop-erty that is desired in concrete pumpingand thus provided by most viscositymodifiers and concrete pumping aid ad-mixtures. The Geo-Institute Glossary ofGrouting Terminology (2005) definesCompaction Grout as “Grout injectedwith less than 1” (25 mm) slump. Nor-mally a soil-cement with sufficient siltsizes to provide mobility together withsufficient sand and gravel sizes to de-velop sufficient internal friction tocause the grout to act as a growing massas injection continues under pressure.The grout generally does not enter soilpores (except , perhaps, whereopen-work boulder gravels are present)but remains a homogeneous mass thatgives controlled displacement to com-pact loose non-plastic soils, gives con-trolled displacement for li f t ingstructures, or both”.

Extensive research and experiencehas found that merely providing a lowslump will not provide a suitable com-paction grout, though many practitio-ners have failed to recognize this. Thecase described by Vanderpool, Norris,and Elfass is unfortunately a commonexample of performance, and had theresearchers not recognized the inappro-priate grout material used in the first in-stance, and insisted in retaining acompetent contractor to re-grout thepile tip, it would no doubt have beenincorrectly concluded that compactiongrouting was not capable of improvingshaft capacity.

History of Low Mobility GroutUseAs a young contractor in 1952, I hadconstructed a small structure in a floodcontrol channel. Unfortunately, somesand filter material eroded from underthe structure in an early storm. The re-sulting void had to be remediated, and itseemed reasonable to fill it with a ce-ment-sand mortar. Neither concrete normortar pumps had yet been developedso a crude hand powered piston pumpwas fabricated and the void was filledusing a typical masonry mortar. Thisearly experience led to development ofa powered piston pump and furtherwork filling sub-surface voids. As expe-rience was gained, it was found thatstructures as well as pavements could beaccurately lifted or jacked to propergrade by such injection. The stiff mortarwould tend to stay near its point of in-jection allowing jacking of even differ-entially settled and heavy structures to aclose tolerance.

An ever expanding range of applica-tions for pressure injected mortar (nowLMG) evolved, all the way from fillingsmall voids in concrete and masonry, tofilling of large tanks, abandoned pipe-lines, and annular spaces, Figure 1.With the expanding range of applica-tion, a variety of different mix designswere made, but always of a low mobility

plastic consistency. Fortunately, I wasunaware of pressure grouting as thenpracticed and most importantly the wellestablished rule that “if you can’t pourit, you can’t pump it”.

In 1957, our now established grout-ing firm was engaged to raise one end ofa swimming pool in a courtyard com-pletely surrounded by an apartmentbuilding. The site had formerly beenhome to a hospital and the pool was lo-cated in what was previously a base-ment boiler room. The plans called forthe bottom slab to be perforated and theentire basement to be filled with com-pacted granular soil, however, the slabhad apparently not been perforated andthe fill was neither granular nor com-pacted, but a low cohesion clayey silt.The deep end of the pool was situated atabout the old floor elevation, while theshallow end was founded on several feetof the fill. The shallow end settled caus-ing cracking of the pool shell, allowingpool water to saturate the fill which wascontained in the old structure.

Grout holes were placed on a grid ofabout four feet throughout the bottomof the pool, Figure 2. The geotechnicalengineer had requested holes in theshallow end be extended to the old floorslab. Recognizing the then prevailingopinion that fine grained soils couldn’tbe grouted, he reasoned a number of

THE GROUT LINE


Figure 1. Concrete pumps had yet to be invented in 1960 so four custom pumpswere used to place read.

small holes filled with grout might pro-vide enough support to prevent furthersettlement. Water ran from the lowergrout holes into the pool as soon as in-jection began, and the inflow continuedfor several hours. The job required anamount of grout equal to more than tentimes the calculated volume, before liftoccurred. A horrible mess was made bythe removal of the muddy water withbuckets which had been spilled all theway from the pool to an outside catchbasin. Everyone involved was obvi-ously upset and not pleased except forthe geotechnical engineer, who ex-claimed “you have squeezed the waterout of the soil….this is not supposed to bepossible….it’s wonderful”. The only ex-planation was that the stiff grout hadsomehow expanded in masses thatforced the soil grains into tighter pack-ing, expelling the water that wasotherwise trapped in the old basement.

To better understand the phenomenaa test program was conducted whereintrial injections were made and extractedfor evaluation, Figures 3. Several hun-dred holes were injected using a varietyof different grout mixtures and injectionparameters. The grout mix compositionwas found to have a major influence onthe shape and volume of the resultinginjected mass. In general, the best re-sults were achieved with the stiffest,

harshest consistency grout that could bepumped. Conversely, poor results wereexperienced with easy to pump groutscontaining clay, entrained air, and/or

pozzolanic materials such as fly ash,with frequent fracturing of the soil andleakage to the surface.

Because the nature, and gradation of


THE GROUT LINE

Figure 2. Water squeezed from the soil ran into pool as fast asgrout was injected (1957).

Figure 3. Extraction of injected grout masses.

Figure 4. Original Brown and Warner envelope of acceptable gradation forcompaction grout aggregate.

the aggregate was found to be a signifi-cant contributor to the shape and vol-ume of the injected mass, trial injectionswere made using aggregates and aggre-gate combinations from many differentsources. A particle size distribution en-velope of gradation that provided goodperformance was prepared and is pro-vided here as Figure 4. Clay or otherpumping aids which were found tocause fluid behavior were prohibited.The results of the test evaluations in-cluding the acceptable envelope of ag-gregate gradation were reported byBrown and Warner (1973).

Several research efforts of DenverGrouting Services in the 1990s also in-cluded retrieval of grout masses result-

ing from different aggregates, additives,and injection procedures (Figure 5).These generally confirmed the datafrom the earlier Brown and Warnerwork but also determined that injectioncontrol could be improved and the dele-terious effect of clay lessened by ex-tending the top aggregate fraction,including gravel s ize par t ic les(Bandimere, 1977). The upgraded en-velope is provided here as Figure 6.Many thousands of projects have beensuccessfully completed with grout con-taining aggregate falling within the lim-its of both gradation envelopes. Whilesome of the most experienced contrac-tors continue to use the original Brownand Warner aggregate, the more recent

gradation envelope containing gravel,has proven to be even more suitable andis recommended in sensitive situationswhere high performance is required.Examples would be when grouting in ornear unrestrained downslopes, or in wa-ter retaining embankments where hy-draulic fracturing of the soil cannot betolerated. Most compaction grout is asimple combination of the appropriateaggregate combined

with 8% to 12% common cement.No other constituents or admixtures areneeded or should be used. Cement is notan absolute requirement, but if elimi-nated, a larger fraction of silt size parti-cles will be required.

In addition to the above formal re-search, many hundreds of injectedmasses have been extracted or exposedin a variety of different soils and usinggrout of varying composition. The re-sults have been amazingly consistent;grouts containing the appropriate ag-gregate create massive inclusions in theground as illustrated in Figure 7.

Conversely, inclusion of clay orpumping aids nearly always results inhydraulic fracturing of the soil with thinwings of grout extending out from thepoint of injection as shown in Figure 8.

It is unfortunate that the reputationof compaction grouting has been badlymaligned due to poor work fromgrouters that refuse to follow the wellestablished material requirements anduse an appropriate grout mix. As in thecase reported by Vanderpool, Norrisand Elfass, many contractors, espe-cially large “geotechnical contractors”claim to be experts in compactiongrouting but once on a project, resort toready mixed grout with inappropriateaggregate, high levels of round grainedfly ash, and a variety of additives suchas bentonite, chemical viscosity modifi-ers and concrete pumping aids. The re-sult is all too often, insufficientimprovement of the treated soil and abad name for compaction grouting.Though some commonly claime thatconforming aggregate is not available,responsible compaction grouters havemanaged to procure acceptablematerials for projects around the world.

THE GROUT LINE


Figure 5. Denver Grouting Services performed many large scale demonstrations inthe 1990s to evaluate grout performance with varying aggregates, additives, and in-jection methods.

Figure 6. Preferable gradation envelope for compaction grout aggregate.

So How Do We Find AppropriateAggregate?Silty sand which is the basic material re-quired, will virtually never be found at aready mix plant because concrete con-stituents must be clean and free of fines.Occasionally, producers will have siltysand overlying cleaner aggregates in aquarry. Perhaps the best place to locateappropriate material is top soil suppliersto the landscaping industry. They notonly often stock satisfactory silty sandsbut nearly always have portable screen-ing plants that can be used to removeoversize particles from natural deposits.On many projects, I have found the na-

tive site soil appropriate by simplyscreening out the oversize fraction.Whereas locating satisfactory siltysands free of clay and conforming to theBrown and Warner envelope is usuallypossible, finding such deposits alsocontaining pea gravel is much more dif-ficult.

Gravel is nearly always stocked byaggregate producers and ready mixplants however, so it is readily available,and can be easily blended with the basicsilty sand. In those rather rare instanceswhere silty materials free of clay are notavailable, it might be necessary to sepa-rately mix silt into common sand which

is nearly always obtainable. In this re-gard, I have hauled silt considerable dis-tances, but as it comprises only 15-30%of the total aggregate mix, the cost isstill manageable. Acquiring the appro-priate aggregate does require more ef-fort than simply purchasing a readymixed compound but it is requisite toproper performance.

Summary and ConclusionsLow mobility grout is any compositionthat “does not travel freely and that be-comes immobile when injection pres-sure ceases”. It is often batched andsupplied from ready mix plants. Com-paction grout is a special type of lowmobility grout that is compounded so asto not behave as a liquid when injectedinto soil. There is no question that ac-quiring proper aggregate for compac-tion grouting frequently requires moreeffort and cost, however failure to do sowill result in a high probability of unsat-isfactory performance, and should notbe allowed.

References and FurtherReadingBandimere, Samson, (1997) “Compac-

tion Grouting; State of Practice”Grouting: Compaction, Remed-iation and Testing, GeotechnicalSpecial Publication No. 66, ASCE,New York, pp. 18-31.

Brown, D. and Warner, J., (1973)“Compaction Grouting”, Journal ofSoil Mechanics and Foundations Di-vision, ASCE, Vol. 99, No. SM8,Proc. Paper 9908, pp. 589-601.

Garner, S., Warner, J., Jefferies, M., andMorrison, N., (2000), “A ControlledApproach to Deep CompactionGrouting at WAC Bennett Dam”,Proceedings, 53rd CanadianGeotechnical Conference, Montreal,Quebec, Oct. 15-18, 2000.

Gomez, J., Hobinson, H., and Cadden,A., (2008) “Use of Limited MobilityGrout for Shallow Foundations inKarst”, Geotechnical News, pp.51-56.

Grouting Committee of the Geo-Insti-tute (2005) “Glossary of GroutingTerminology”, Journal ofGeotechnical and Geoenvir-onmental Engineering, Vol. 131, Is-


THE GROUT LINE

Figure 7. A variety of massive grout shapes resulting from aggregate conformingto the gradation envelopes in Figures 4 and 6.

Figure 8. Hydraulic fractures result from grouts with clay or pumping aids.

sue 12, pp 1534-1542.Warner, J., “Compaction Grouting

Mechanism – What Do We Know?”(1997), Proceedings, Grouting –Compaction, Remediation, Testing,Geotechnical Special PublicationNo. 66, ASCE.

Warner, J., Schmidt, N., Reed, J.,

Shepardson, D., Lamb, R., andWong, S., (1992), “Recent Advancesin Compaction Grouting Technol-ogy”, Proceedings, Grouting, SoilImprovement and Geosynthetics,ASCE, New Orleans, GeotechnicalSpecial Publication No. 30, pp.252-264.

Warner, J., (1992), “Compaction Grout;Rheology vs. Effectiveness”, Pro-ceedings, Grouting, Soil Improve-ment and Geosynthetics, ASCE,New Orleans, Geotechnical SpecialPublication No. 30, pp. 229-239.

THE GROUT LINE


Documents

June 08 Complete