Ultrafine Cementitious Grout

1. Field of the Invention (Technical Field)

The invention relates to ultrafine cementitious grouts, particularly ultrafine cementitious grouts suitable for sealing microfractures in rock.

2. Background Art

Grouts are widely used in the construction and mining industries. The most commonly encountered grout is the thin mortar used to fill cracks and crevices in masonry. Grouts used in common residential and commercial construction consist generally of mortars of sand and cement. Where finer grouting is required, such as in decorative or waterproof tile installation, grouts composed of finer sands and plasters are utilized.

Grouts are used in the mining industry to seal fractured rock. The grouts utilized in ordinary mining are similar to those used in masonry construction, and sometimes include sand. As mining becomes more sophisticated, and as artificial subterranean repositories become increasingly attractive alternatives to aboveground storage of nuclear and other hazardous wastes, the importance of sealing microfractures (cracks whose aperture is 100 microns (100 .mu.m) or less) has increased. Extremely small fissures and cracks in repository walls must be reliably closed against the transmission or migration of hazardous fluids and gases.

A principal obstacle to reliable sealing of microfractures with cementitious grouts is the production of grouts with sufficiently small particles. Conventional understanding is that the successful permeation of a fracture with a particle-containing grout requires that the maximum particle dimension should not exceed one-third of the fracture aperture. As a result, ordinary cementitious grouts known in the art contain unacceptably large particle diameters to effectively enter and seal microfractures.

The need to seal microfractures and silty (fine grained) soil, has lead to the use of chemical grouts, such as polyacrylamide and polyacrylate, which are injected as liquids containing no particles. Because they are low-viscosity liquids when applied, chemical grouts adequately penetrate and seal microfractures. The chemical or so-called solution grouts, however, have become undesirable from environmental and worker safety standpoints. Polyacrylamides are known neurotoxins and suspected carcinogens. Polyacrylates can cause serious burns to the skin and mucous membranes. The use of polyacrylamides has been banned, and many manufacturers are phasing out the production of polyacrylates. Moreover, the degradation testing required to demonstrate the longevity of chemical grouts is time-consuming and imprecise; long-term reliability, however, is essential for grout sealing hazardous or nuclear waste repositories.

Interest has recently returned to the use of cementitious grouts to seal microfractures. The most important parameter for evaluating the quality and effectiveness of cementitious grouts is particle size. Portland cements with 90% of the particles smaller than about 10 .mu.ms, and slag microfines with 90% of the particles smaller than 8.4 .mu.ms, have been developed, with the slag composed of the fine grained "dust" which accumulates on precipitation units in the exhaust stacks of steel mills. The existing microfine grouts, however, have mixing instructions which universally specify large quantities of water (high water to cementitious materials ratios). The amount of water required in the existing microfine cementitious grout mixes causes undesirable results such as bleed (water separating from the grout as it hardens), reduced unconfined uniaxial compressive strength, and increased permeability.

U.S. Pat. No. 5,346,012 to Heathman and Crook for Fine Particle Size Cement Compositions and Methods describes a microfine grout comprising fine grained Portland cement and fly ash, both with a maximum particle size of about 10 microns, mixed to a water to cementitious material (W/CM) ratio of about 0.7 to about 2.8, preferably 0.9 to 1.3. The fly ash was chiefly SiO.sub.2 (54%) and Al.sub.2 O.sub.3 (30%) with minor amounts of other minerals. Recommended additives include dispersing agent (up to about 3%), fluid loss control additive (up to about 3%), and a set retarding additive. This reference suggests that the relatively high water content of its preferred mixture can produce grout compositions of low density and low viscosity without unsatisfactory loss in strength. No mention is made of the permeability or bleed characteristics of the grout mixture or how it can be prepared.

Unfortunately, neither the slag based grouts or the fly ash based grouts have exhibited the performance necessary for many applications. These performance requirements include the ability to seal fractures in the 10 .mu.m size range, essentially zero bleed, essentially zero porosity and permeability, controlled set time without additives, and reasonable cost.

The grouts that use slag as the pozzolanic component exhibit less compressive strength, have relatively large particle size, and require additional lime additives. The grout disclosed in U.S. Pat. No. 5,346,012 also has relatively large particle size (no greater than 30 microns as specified in the examples), relatively high water content to injectability which result in lower compressive strength, water `bleed` during crystallization that leaves channels of interconnected porosity ("wormholes"), and an undesirable lower density. Although it is clear that grouts with smaller particle sizes are better able to seal the smallest microfractures, no reference provides guidance as to which constituents may be successfully employed in the ultrafine particle size regime (less than about 6 .mu.m), in what ratios one to another, optimal W/CM ratios, or even how the ultrafine grout mixtures can be manufactured.

Consequently, an unmet need remains for a microfine, cementitious grout for use in sealing microfractures, which is easily worked when wet and results in a grout seal having high strength, low permeability and zero bleed.

The invention relates to ultrafine cementitious grout compositions and methods of making and using them. A preferred composition of the invention consists of a mixture of cementitious materials, the mixture being made by combining from about 40 to about 50 weight percent of a Portland cement, from about 50 to about 60 weight percent of pumice pozzolanic material, and up to about 1.5 weight percent superplasticizer, with 90 percent of the composition particles being smaller than 6 .mu.m in diameter and, preferably, less than 5 .mu.m. The combination of cementitious materials is mixed with water in water-to-cementitious materials ratio between about 0.40/1 and 0.60/1. Various specific embodiments of the invention include the preferred use of ASTM Type V Portland cement, and particular superplasticizers.

The methods of making the foregoing compositions of the invention, whereby high-speed attritor apparatuses and specialized pulverization media are used to produce a grout composition having 90% of particle diameters less than 6 .mu.m are of primary importance and also disclosed herein. The method of making the grout includes both dry and wet pulverization processes, where the grout may be pulverized dry for later wetting, or can simultaneously be pulverized and mixed with water for immediate use. The dry grout mix, or wet grout paste, have practical use especially for sealing microfractures in rock.

The accompanying drawings, which are incorporated into and form a part of the specification, illustrate several embodiments of the present invention and, together with the description, serve to explain the principles of the invention. The drawings are only for the purpose of illustrating a preferred embodiment of the invention and are not to be construed as limiting the invention. In the drawings:

FIG. 1 is a cross-sectional side view of a high-speed continuous feed attritor utilized in the invention;

FIG. 2 is a cross-sectional side view of a batch attritor utilized in an alternative embodiment of the invention;

FIG. 3 is a schematic diagram depicting the process and apparatus for production of dry pulverized grout according to the invention; and

FIG. 4 is a schematic view of the grout injection system according to the invention.

This invention relates to ultrafine grouts. Most particularly, the invention relates to an ultrafine cementitious grout suitable for reliably sealing microfractures in rock walls of subterranean mines and chambers against the leakage of acid-, metal-, or other waste-bearing water and/or gases. Additional potential applications of the invention include the strengthening of weak soils, whereby the composition of the invention is used to increase the bearing capacity of soils at construction sites (and to prevent liquefaction), to facilitate the construction of very low permeability subsurface grout curtains and walls, and to repair cracked dams, as well as reinforcing defective grout curtains below the centerline of dams.

The grout of this invention has a unique combination of components at particle sizes below those disclosed in the art and exhibits properties that are a significant improvement over prior grouts. This grout has successfully entered and sealed 6 micron fractures and, in one measured instance, has sealed a 3 micron fracture. Penetration in microfractures as small as 6 microns was found to exceed 3 meters. The unconfined, uniaxial compressive strength is 6.2 MPa (899 psi) one day after injection, increasing to 38 MPa (5510 psi) after 28 days. The grout has essentially no permeability and is self healing in the event of a fracture. No water separates from the grout during hardening (zero bleed). The setting time of the grout is adjustable from about 2 hours to about 6.5 hours with a minimum of one hour of good workability. Further development may increase these boundaries.

The ultrafine particle sizes and unique combination of Portland cement and amorphous silicon dioxide (pumice) produce these superior operating parameters. The ultrafine particle sizes enable the superior penetration characteristics of this grout and also enhance the beneficial chemical interaction between the Portland cement and the pumice due to the increased surface area of the ultrafine particles. The chemical interactions between the two main constituents successfully combat the main degradation mechanism for cementitious grouts. With ordinary cementitious grouts there is always some permeability. Water from external sources or bleed water produced during hardening collects on the exterior surfaces of the grout and in the interconnected porous spaces in the grout and begins to dissolve the Portlandite phase (calcium hydroxide) of the grout. This produces an excess of calcium and hydroxyl ions in solution. Mass transfer drives these ions from areas of high concentration to areas of low concentration. This predominately means that the ions exit the external interface between the grout and the host rock, degrading this interface and creating a fluid pathway through the grouted rock. The grout of this invention employs the high silicon dioxide content pumice to prevent this chemical reaction from occurring. The silicon dioxide reacts with the free calcium ions immediately to form calcium trisilicate which has very low solubility. The relatively high fraction of pumice pozzolan decreases the relative amount of Portland cement, thereby reducing hydrational heat during hardening. The ultrafine pozzolan also results in smaller pores in the grout (average diameter is one micron) and causes the pores to be more disconnected as evidenced by the extremely low hydraulic conductivity exhibited by this grout.

It is generally accepted in the art that thorough grouting of a microfracture requires the use of a grout whose maximum particle size does not exceed approximately one-third the size of the fracture aperture. Larger particles in any significant quantity interlock ("bridge") at the entrance point and this can prevent the grout from permeating and sealing the fracture.

The permeability and strength of hardened grout depend in part upon the degree of hydration, the temperature reached during hydration reactions, the degree to which hydrated calcium silicate phases replace the most soluble phase of the hardened grout, which is calcium hydroxide, and the ability of hydration of the grout to consume all of the water. These factors also affect the chemical stability and permeability of the hardened product (the latter is partially a function of the grout's degree of microcracking, which in part results from thermo-mechanical strain induced during curing). For a grout to meet the physical and chemical requirements for a particular application, the mix is optimized by examining the effects upon material properties of variations in type of cement, the weight ratio of water to cementitious materials (W/CM), and the types and quantity of additives. For purposes of this disclosure and claims, "cementitious materials" means Portland cement powders, pumice pozzolanic material, and optional cement additives such as plasticizers. A single adjustment in the quantity and/or type of any one of these ingredients tends to affect more than one mechanical property in the final grout.

The present invention provides an ultrafine cementitious grout that is suitable for sealing microfractures in substances, particularly hard rock. For purposes of this disclosure, "ultrafine" means a grout mix having 90% of particle diameters less than 6 .mu.m and preferably less than 5 .mu.m. The composition of the invention meets the criteria, such as suitable rheology when wet and compressive strength and low permeability when hard, required for reliable long-term sealing of subterranean rock, e.g., of hazardous or nuclear waste repositories or sealing of tunnel walls.

The pulverization of cementitious material to achieve the required maximum particle size in the grout mix is a complex process. Pulverization may be performed either "wet" or "dry." Dry pulverization is the preferred process and involves continuous processing of the various quantities of cement, pozzolanic material, and superplasticizer. The dry mix may then be bagged and stored for later transportation to the site of use. On site, the dry grout is mixed with a prescribed quantity of water for use. Wet pulverization involves the simultaneous batch pulverization and mixing, at the job site, of the complete grout mix, wherein the pulverization process is performed in presence of water (in the amount prescribed for the complete grout mixture). The wet process is a batch process with relatively limited capacity (100 liters) with existing machines and seems limited to particle sizes at the 90% less than 8 .mu.m range. The proportional quantities of grout mix components are substantially the same for dry and wet pulverization, but pulverization times and apparati differ between the two methods. Attention to the type of milling apparatus (including pulverization media) employed, the grinding rate and duration, and the precise compositional formula, results in a suitable ultrafine grout composition. Effective limits and optimums have been determined for, among other variables: the type and weight percent of cement; the ratio, by weight, of water to cement; the weight percent of pumice pozzolanic material; the type and weight percent of superplasticizer; the diameter and composition of the satellites (small grinding balls) in the attritor; the rotational speed of the attritor and the pulverization retention time in the attritor.

The ultrafine grout components of the invention are pulverized in attritors. The wet pulverization process preferably utilizes a "batch" attritor, while dry grout preferably is pulverized in a continuous feed high-speed attritor (HSA). Attritors function to produce powders, primarily by agitating pulverization media together with the materials to be pulverized in a stationary vertical tank. The use of an HSA permits the achievement of the very small maximum particle dimensions required for effective grout sealing of microfractures. Attritors are used in the food, paint, rubber, pharmaceutical and ceramic industries, but are not known to have been used previously to manufacture grout. Preferably, the method of the invention is performed using a SZEGVARI.RTM. Attritor Model HSA-100, manufactured by Union Process, Inc. Suitable alternatives may be engaged to accommodate batch sizes, power availability, and the like. Attritors such as the SZEGVARI.RTM. Attritor Model HSA-100 reduce particle size much faster than standard ball mills and are much more energy efficient.

A typical continuous-feed attritor apparatus is depicted in FIG. 1. The attritor 30 comprises a vertical steel (e.g., 316 stainless steel) tank 10, open or closed at the top. Extending downward into tank 10 is vertical shaft 12 equipped with a plurality (e.g., six) of cross-arms 16, of durable composition such as tungsten carbide. Pulverization media 20 (i.e., very hard spheres, partial number shown) are densely stacked in tank 10 to level just covering the upper cross-arm 16 when the machine is at rest. An electric or other motor rotates shaft 12, causing cross-arms 16 to move through pulverization media 20, imparting random motion to the media spheres. Pulverization of material within tank 10 results from impact and shear effects between individual media spheres. Materials may be added to tank 10 via materials inlet 22, and extracted from tank 10 via outlet 24, using pumps or augers (not shown) known in the art, or otherwise.

Acceptable pulverization media 20 include 2 mm-diameter (preferred), 3 mm-diameter, and 6 mm-diameter zirconium oxide spheres (stabilized with rare earths), as well as 3 mm-diameter carbon steel spheres and 3 mm-diameter, 440 C, 316 stainless steel spheres. For dry pulverization, the pulverization media 20 preferably comprises 2 mm-diameter zirconium silicate spheres, based upon demonstrated milling performance (including the need to avoid mixture sticking, caking, and agglomeration), specific gravity, cost, and longevity. Wet pulverization preferably employs 3 mm, 316 stainless steel spheres.

For dry pulverization, the various dry ingredients of the composition are pulverized and mixed in an HSA 100 attritor 30, from which they are removed after pulverization to an average particle size of 4 microns (4.0 .mu.m). Preferably, the retention time of the mix within a SVENGARI.RTM. continuous feed 100-HSA attritor 30 is between approximately four minutes and six minutes, at 390 to 400 RPM. The throughput under these preferred conditions is approximately 1500 lb./hr. The dry ingredients then pass through an air classification unit where the product (particle size 90% smaller than 6 .mu.m) is split off and packaged. The oversize is returned to the HSA 100 attritor in a closed circuit. Attrition of dry components preferably is performed in a very-low humidity environment, as ultrafine cementitious materials are highly hygroscopic and absorption of ambient water vapor may impair the quality and performance of the dry mixture. During the pulverization, the HSA attritor operates at approximately 250.degree. F. to 350.degree., which drives off moisture. For the same reason, the final dry mix is hermetically sealed as produced in waterproof containers for storage and transportation.

To perform the dry pulverization process of the invention, a continuous feed high speed attritor HSA-100 30 is used to pulverize the solid grout components. Pumice (preferably minus 325 mesh) is utilized as a grinding aid and pozzolan. An inadequate amount of pumice in the dry mix undesirably results in the generated fines agglomerating and sticking to the pulverization media 20 and to the interior of the attritor tank 10 itself The preferred pulverization media 20 for dry pulverization is 2 mm-diameter zirconium silicate spheres.

Reference is made to FIG. 3, showing the use of the attritor 30 to make ultrafine cementitious grout according to the dry pulverization method of the invention. Commercially available electronically controlled, loss-in-weight feeders 52,52' and 52" are used to feed the dry mix components into hopper 56, in the precise weight percents prescribed. Transportation of dry mix components from materials feeders 52,52' and 52" to hopper 56 is via screw or auger conveyor 58 or acceptable equivalent. Dry mix components are then dropped from hopper 56 into attritor 30. Pulverization occurs in the attritor 30, where mechanically derived heat (250.degree.-350.degree. F.) counteracts the hygroscopic character of the dry mix.

Dry pulverization involves the use of an air classification unit 54 for reliably consistent size separation of dry mix particles. Air classification unit 54 functions to evaluate and physically separate ultrafine dry mixes according to particle size. An acceptable air classification unit for practicing the invention is the Model MS-5 air classification unit manufactured by Progressive Industries, Inc., Sylacauga, Ala.

Dry grout mix components are added to the attritor 30 in the proportions hereinafter specified, preferably using loss-in-weight materials feeders 52,52' and 52". The dry mix components are proportioned, by weight: Between about 40 weight percent to about 50 weight percent ASTM Type V sulphate-resistant Portland cement; between about 50 weight percent to about 60 weight percent pumice; and up to about 1.5 weight percent superplasticizer. The preferred proportional amounts are: 45 (.+-.3) weight percent ASTM Type V sulphate-resistant Portland cement; 55 (.+-.3) weight percent pumice (minus 325 mesh); and 1.0 (.+-.1) weight percent superplasticizer (Dowell Schlumberger D-65c or equivalent).

The dry pulverized output of the attritor 30 is conveyed by auger or other conveyor 62 to air classification unit 54. The pulverized output is passed through air classification unit 54 where the portion (about 40%) of the pulverized and mixed output from the attritor that is smaller than 6 .mu.m is segregated and physically removed and immediately hermetically packaged as finished product. A Laser Particle Size Analyzer 42, such as the Malvern Model 3200E MASTERSIZER.RTM. laser particle size analyzer from Malvern Instruments, Ltd., may be used for monitoring the particle size and percent distribution thereof of the output from the classification unit 54. The removed finished product may be conveyed to a container 64, from which it is removed for packaged storage or for further immediate on-site use. Further on-site use may involve conveying the product via conveyor 66 to a grout mixer (FIG. 4). Oversized material (e.g., >6 .mu.m) is not removed by air classification unit 54, but rather, is returned, by screw or other conveyor 68 in a closed-circuit configuration, as the preferential feed to attritor 30 for further pulverization. Passage, or multiple passages, of the mix through attritor 30 and the air classification unit 54 reliably produces a dry grout mix composition containing particles 90% smaller than 6 .mu.m. The final product may be immediately mixed with water and used, or may be hermetically packaged against ambient humidity for transportation to the mixing site.

The forgoing process creates a dry grout mix with 90% of particles smaller than 6 .mu.m. When mixed with clean tap water in the proper water-to-cementitious materials ratio, the grout mix produced according to this process results in a grout with zero bleed, a uniaxial compressive strength greater than 5,800 psi at 28 days, and a hardened permeability of 10.sup.-21 meters squared. The W/CM weight ratio is less than or equal to 0.70/1, preferably between 0.40/1 to 0.60/1, inclusive, and most preferably 0.5/1. These lower W/CM ratios reduce workability somewhat by shortening the set time, but increase strength and decrease permeability.

Reference is made to FIG. 4. The on-site mixing of the dry mix and the prescribed amount of water is performed in a mixer 70, which is a high-speed, high-shear "colloidal" type grout mixer. Machines sold under the COLCRETE name work well. The feed for mixer 70 may include the on-site output from the container 64, or bagged dry product imported from off-site, typically handled as 50 pound bags. Water is added in the prescribed proportion to the dry components within mixer 70, the wet ingredients are thoroughly mixed (usually this takes less than one minute), and then transferred by circulation pump 71 to an agitation tank 40 from which the wet grout is dispensed.

The thixotropic grout of the invention must be constantly kept in motion to remain fluid. After the wet grout ingredients are mixed in mixer 70, the grout mix is transferred, via low pressure hose 72 and circulation pump 71, from mixer 70 to agitation tank 40. Grout mix is cooled and constantly stirred (e.g., at approximately 60 rpm) in agitation tank 40. Agitation tank 40 may be fashioned from 316 stainless steel, and preferably is capable of variable stirring speeds with motor-driven paddles. Like the "batch" attritor tank 10 used for simultaneous wet pulverization and mixing, agitation tank 40 can be surrounded with a cooling jacket (not shown); tank cooling for both the agitation tank 40 and the attritor tank 10 is accomplished by circulating water at 42 (.+-.5) degrees Fahrenheit through their respective cooling jackets.

FIG. 2 depicts a "batch" attritor 31 usable for performing wet pulverization. Batch attritor 31 operates similarly and on the same general principles as a continuous feed attritor, with the difference between the two types evident from their titles. Like high speed attritor 30, a batch attritor 31 includes pulverization media 20 contained within tank 10 and disposed around a plurality of cross-arms 16 (preferably tungsten carbide). Each cross-arm 16 is attached to and driven by vertical shaft 12.

When wet pulverization is performed in a "batch" attritor, tank 10 can be surrounded by cooling jacket 32. Cooling jacket 32 defines a space envelope 34 around tank 10. Chilled water is circulated through envelope 34 via cooling water inlet 36 and cooling water outlet 38, to maintain a reduced temperature of the grout mixture to increase the period during which the grout remains injectable. Also featured are recirculation line 11, recirculation pump, and discharge 19.

Combined reference is made to all the figures. A grout "wet" pulverization system according to the present invention utilizes essentially the same components as the dry pulverization system. For wet pulverization, colloidal mixer 70 is not required, since pulverization and mixing are concurrently accomplished within the batch attritor 31. Attritor 31 instead is used in lieu of a mixer and in direct conjunction with an agitation tank 40.

An important aspect of quality control during the practice of the invention is the assessment of grout particle size and percent distribution thereof during pulverization/mixing and injection. Grout particle size and percent distribution by size preferably is determined periodically (e.g., every 15 minutes) using a laser particle size analyzer 42, such as the Malvern Model 3200E MASTERSIZER.RTM. laser particle size analyzer from Malvern Instruments, Ltd. Use of a particle size analyzer is indicated to assure that maximum grout particle size is within the ultrafine range (90%<8 .mu.m for the wet process and 90%<6 .mu.m for the dry process) prior to grout injection. Grout mix samples for particle size analysis are removed from the wet system during pulverization/mixing via attritor sampling port 48, and from sample port as during injection. If the dry pulverization method is used, samples are taken from sample port 85.

In either the wet pulverization or dry pulverization method, the completed wet grout mixture is pumped by high-pressure circulation pump 74 through high-pressure hose 75 to a desired or appropriate injection system. An injection system will include the various valves, gauges, and fittings known in the art for grout injection. A suitable injection system is depicted in FIG. 4. Wet grout mix is pumped under pressure to an injection tee 76, where a portion of the mix is directed through injector 78 and into injection bore 80. As illustrated in FIG. 4, injector 78 and bore 80 are disposed through rock or concrete slab 82 overlying the stratum to be sealed. The passage between bore 80 and injector 78 is sealed with injection bore packer 83.

Injection tee 76 may contain pressure-relief and other desirable elements and features, such as a second sampling port 85, and also allows non-injected mix to be returned to agitator 40 via circulation line 86. Pressure in injector 78 may be monitored by a pressure transducer 88 disposed therein.

Accordingly, wet grout mixture may be pumped under pressure from agitator 40, through hose 75 to injector tee 76, then under high pressure through injector 78 into the formation to be sealed. Excess grout is recirculated to agitator 40 for re-use. Pressure transducer 88 may be engaged to monitor pressure and actuate safety and shut-off valves.

The grout mixes pulverized in attritor 30 (dry pulverization) or 31 (wet pulverization) comprise cement, superplasticizer and pozzolanic material. Each constituent of the grout mix to be pulverized in the appropriate attritor 30 or 31 is added in the prescribed proportions to the attritor. In the case of wet pulverization/mixing, the measured amount of water to be added to the mix is introduced first to the attritor and circulated continuously, followed by the superplasticizer, pozzolan, and Portland cement.

The specific types and relative amounts of the grout mix constituents are of paramount importance in the successful practice of the invention, as is the achievement of sufficiently small particle size. Attempting to obtain an extremely low maximum particle dimension requires relatively longer pulverization times and is limited by the setting time of the wetted mixture, since finer materials react faster. Efforts to treat the grout mixture with admixtures or extra water to delay setting, however, contribute to undesirable shrinkage and bleed, and jeopardize the mechanical properties of the hardened grout, most particularly the unconfined, uniaxial, compressive strength and permeability.

An illustrative problem in the present art is the challenge of preparing, using simultaneous pulverization and mixing, of cement-based grouts with 90% of the particles smaller than 8 .mu.m composed of conventionally ground cements and silica fume or fly ash as pozzolanic materials, and incorporating superplasticizers. Increasing the water-to-cementitious materials ratio (.mu./CM), or the superplasticizer content, increases the fluidity of the grouts and increases the time the grout remains injectable. But prior art ultrafine grouting methodologies generally specify W/CM ratios in excess of 1.0/1. Such high water and superplasticizer contents are required to decrease the viscosity of the mixes during grinding and to delay setting sufficiently to allow wet pulverization and/or injection of the mixes for adequate periods of time.

A need to increase the W/CM ratio to greater than 1.0/1, and grout with maximum particle size of, say, 90% less than 8 .mu.m conflicts with other desirable grout properties. The mechanical properties of grouts prepared with such high contents of water and superplasticizer are not always certain, but it is clear that the use of higher water content in a grout tends to decrease the compressive strength and increase the permeability. Both effects are undesirable.

Specifically, grout mixtures composed of common Portland cement (American Society for Testing and Material (ASTM) Type V), silica fume or fly ash, superplasticizer (2%<wt %<8%), and W/CM ratio in excess of 0.9/1, with 90% of the particles smaller than 8 .mu.m, are found to have unacceptably low density (between 1.14 and 1.47 g/cm.sup.3) and high permeability, as well as unacceptably low unconfined compressive strength (less than 18 MPa) (2600 psi) as result of the higher water content. "Bleed" (separation of water when the grout "sets" or hardens) can also occur. Increased amounts of superplasticizer can result in increased linear shrinkage when the grout hardens.

Grinding efficiency can be controlled by varying the rotational speed of the attritor and by using different pulverization media. It is determined that for wet pulverization there is an optimum pulverization time beyond which any increase in grinding time does not significantly decrease particle size.

Thus, the present invention consists in part of a determination that when using the wet pulverization process, it is not feasible (particularly within economic limits) to produce grout mixes with a 90% smaller than 7 .mu.m without overly increasing the W/CM ratio and plasticizer content and consequently jeopardizing the final properties (permeability and strength) of the hardened grout. Further, the component proportions and pulverization conditions yielding a grout paste with 90% of the particles smaller than 8 .mu.m (wet pulverization) or less than 5 .mu.m (dry pulverization), with reduced water and superplasticizer contents, and commensurate with economics, has been determined. Grout mixes with such particle size limits are adequate to seal microfractures in most commonly encountered subterranean materials. The invention accommodates the ultrafine particle size required, while meeting other performance criteria (compressive strength, shrinkage, zero "bleed" and low permeability).

Principal workability and performance criteria satisfied by the grout composition of the invention, and believed unsatisfied by compositions known in the art, include: the reduction of agglomeration (to less than 5% of grout mass) for at least one hour subsequent to pulverization; acceptable rheology (<0.040 Pa.s at shear rate=1208/s) (necessary for wet grout working, transportation, and penetration); initial Vicat set (at 80.degree. F.) between two and about six hours subsequent to mixing; zero bleed; unconfined uniaxial compressive strength greater than at least 40 MPa; permeability of 10.sup.-21 meters squared or less; and less than 1% linear shrinkage.

The preferred embodiment of the inventive composition is a grout mix incorporating ASTM Type V Portland cement, pumice and superplasticizer. Specifically, the preferred composition is: 45 (.+-.3) weight percent ASTM Type V sulphate-resistant Portland cement; 55 (.+-.3) weight percent pumice; plus 1.0 (.+-.1) weight percent superplasticizer, such as DOWELL-SCHLUMBERGER D-65.RTM. (a sulfonated sodium naphthalene formaldehyde condensate) or acceptable alternative. The pozzolanic pumice preferably is 325 mesh pumice (available from Hess Pumice Company of Malad, Idaho). This preferred dry grout mix formula is mixed with ordinary tap water in a W/CM ratio equal to or less than 0.70/1, preferably between approximately 0.4 0/1 to 0.6 0/1, inclusive, and most preferably 0.5/1.

It is known in general that the presence of a superplasticizer in a grout mix enables the use of a lower W/CM ratio which desirably increases the density and durability of the resulting hardened grout while retaining good workability and reducing bleed. The setting time is also increased somewhat. However, the intense grinding process employed to produce the ultrafine particle sizes herein most probably severs a significant portion of the long chain molecules that make up the superplasticizer and thereby reduces the effectiveness of the material by an undetermined amount. In the case of continuous grouting during the boring of tunnels, short set times are preferred, indicating that less superplasticizer should be used.

The inventive composition preferably includes ASTM Type V Portland cement as a principal component. Other common Portland cements (ASTM Types I, II, III, and IV) may be suitable in certain applications. However, ASTM Type II may generate undesirably high heat of hydration, and ASTM Type III may display too rapid a setting time for grout injection. ASTM Type V is preferred for its overall strength, low heat of hydration hardening time, and sulfate resistance. Sulfate resistance is indicated in subsurface grouting in crystalline rocks where groundwater may contain cement-attacking sulfates.

As previously discussed, the proportional amounts of the component ingredients of grout can dramatically impact the physio-mechanical properties of both the wet mix and the hardened grout. The preferred embodiment of the inventive composition is determined best to meet the principal criteria set forth above. Generally, if the proportional amount of pozzolanic material (e.g., pumice) exceeds or falls short of that specified, the resulting mix may set too quickly to be workable for a sufficient time, or may be too porous, permeable, or weak when hardened. Similarly, excessive quantities of superplasticizer will cause excessive shrinkage, while insufficient plasticizer will unacceptably shorten set time. Excessive W/CM ratios unacceptably reduce strength, increase permeability and result in "bleed."

Specific Embodiments:

The invention is further illustrated by the following non-limiting examples.

After preliminary screening, three candidate grouts were selected for further investigations. Mixes 94D, 94J and 94L (Table 1) were selected because they exhibited the best combination of low viscosity, adequate strength and relatively low shrinkage.

The viscosities and the particle size distributions of selected mixes following the mixing in the laboratory scale colloidal mixer, and being stirred at 60 rpm in the agitation tank for 2 h, are shown in Tables 2, 3 and 4. Other data, not shown, indicate that the particle size distribution can be further reduced to 90% less than about 5 .mu.m with average size less than about 2 .mu.m. This is accomplished by adjusting the classifier to produce this smaller size range and slightly increases the processing time. This smaller size has even better injectability, similar mechanical characteristics and is preferred.

The results indicate that the rheological behaviours of the mixtures are similar (Tables 2, 3 and 4). The data show, as expected, that the viscosity tends to slowly increase with stirring time in all cases. The in all cases. The study showed that the amounts of pumice and water used in the preparation of the mixtures were the critical determinants of rheology. The relatively high viscosity of mixes 94J and 94L was attributed to the high specific surface area imposed by the addition of pulverized pumice.

The selected grout mixes maintained good workability for longer than 2 h, which may be attributed to two possible causes. One is that the addition of superplasticizer prevents flocculation and sedimentation of the grout particles, as reflected in the particle size distribution. Only very limited changes in the particle size were observed during the 2 h of stirring. Alternatively, the stirring energy is sufficiently high to prevent agglomeration and setting. In any case, the selected grout mixes stirred easily and retained their fluidity for more than 2 h after mixing in the laboratory scale colloidal mixer.

Penetrability

The process of grouting involves the injection of cement paste into fractured rock, with limited control of flow direction and travel distance. In fractured rock the grout encounters a wide variety of potential pathways. A grout mixture should have low viscosity to increase the penetration distance.

Generally, penetrability can be defined as the ability of a grout paste to penetrate fractures. Laboratory tests have been performed to evaluate the injectability and penetration of unset grout pastes in small-aperture fractures. The penetrability of the selected grout mixes prepared in the laboratory scale colloidal mixer was determined according to the French standard NFP 18-891. The procedure consisted of injecting a grout paste under constant pressure (0.075 MPa.+-.0.0025 MPa) into a series of vertical plastic columns filled with sand of predetermined particle size distribution. The grout is injected from the bottom of the column and the time to reach different heights marked on the column was recorded. In this test, four sand mixes with different particle size distributions were used to stimulate fissures with different apertures. Table 5 summarizes the particle size distribution of the sand used in each column.

The injectability of the unset grout pastes was evaluated under dry and wet conditions. The penetrability of selected grout mixes 94D2, 94J2 and 94L2 was found to be different, although all the grout mixes were prepared using the same raw materials and mixing technique. Under the experimental conditions, the data obtained in these tests indicate that all the mixes investigated exhibited some penetrability under both dry and wet conditions. Limited penetrability was observed under dry conditions. This may reflect an increase in the viscosity of the paste during the injection due to the loss of some of the mixing water. The dry condition of the sand may have caused additional absorption of some of the mixing water, thus reducing the fluidity of the grout.

Under constant pressure, penetrability appears to strongly depend on grout composition, particularly on the water content; thus, the viscosity of the unset grout paste. In the early stage of grout penetration, the water facilitates placement of the grout. The penetration distance increased with the increase in W/CM ratio and superplasticizer content. The more fluid nature and the better penetration in all four columns of the grout 94D paste compared to grout 94J and grout 94L is, however, probably entirely due to extra water and superplasticizer used in the preparation of the mix.

In spite of the higher superplasticizer content (2.0%), the 94L grout paste could not penetrate the sand (0.112/0.106 mm) in column 4. Furthermore, grout 94L penetrated only relatively short distances in the other columns. It does appear that increasing the content of pozzolanic material in the mix 94J improved the penetrability characteristics of the grout.

The observed effect may be attributed to the decrease in the amount of cement and, therefore, to the decrease in the rate of hydration.

Column 1, containing sand with 0.850/0.425 mm particle size distribution, was selected to evaluate the injectability of the selected mixes. As determined by the particle size analyzer, the freshly prepared mixes contained an average particle size below 2.5 .mu.m and .about.90% of the grain sizes were finer than 6 .mu.m.

Under the experimental conditions used in these tests (i.e., constant pressure of 0.075 MPa), penetrability is characterized by the relation h=f(t), the penetration height reached by a grout in a column was a function of time. Overall, the trend of the penetrability is to increase with injection time.

The results show that under dry conditions, with the exception of the 94D grout paste, neither 94J nor 94L grouts totally penetrated the sand in column 1 (35 cm); the grouts penetrated the column up to 15 cm and 10 cm, respectively. Under wet conditions both mixes 94D and 94J penetrated the sand in column 1 (35 cm). The penetration of grout 94L increased slightly; however, it penetrated only 15 cm. The results also show that the time required for the grouts to reach a certain height in the sand column decreased in the tests performed with wet sand. Therefore, it is reasonable to assume that the wet sand surfaces influenced the rate of grout penetration. This may reflect a decrease in the viscosity of the paste during the injection. The wet condition of the sand may have eliminated absorption of some of the mixing water, thus maintaining or increasing the fluidity of the grout.

Microscopic examination of specimens sliced from sand columns grouted with 94J and 94L grout mixes after curing for 28 d, confirmed the ability of the selected grouts to penetrate openings between sand grains with a wide range of apertures. The range of openings filled with grout in the sand columns as measured with the scanning electron microscope are listed in Table 6. Difficulties were encountered during the measurements of the very fine openings between sand grains. In most cases it was difficult to distinguish the boundaries between the grout and sand grains. Microscopic examination of grouted specimens revealed that the sand grains were totally incorporated in the grout mass.

The examination also showed that the grout totally conformed to the irregularities of the sand grains. No dislocation of sand grains during cutting of the specimens was observed, suggesting that a very good bond was established between grout and sand grains.

Hydraulic Conductivity

The water permeability of the three selected grout mixes (Table 1), produced by mixing in the laboratory scale colloidal mixer was measured according to the procedure described below.

The grout mixes utilized as a starting point are Type K grout consisting of 75% (by mass) ordinary sulphate-resistant Portland cement (Canadian Type 50), 25% (by mass) pumice, and 2.5% superplasticizer, all with an average particle size below 2.5 .mu.m and .about.90% of grain size less than or equal to 6.5 .mu.m. The pumice was predominantly amorphous SiO.sub.2 (72%) and Al.sub.2 O.sub.3 (12%) with minor amounts of other minerals. The three mixes were each prepared by measuring the masses of the individual components, dry Type K grout, dry pulverized pumice, superplasticizer (if needed), and tap water necessary to make 10L of grout. The water and superplasticizer were added to the colloidal mixer first, then the dry pulverized pumice, then the dry Type K grout mix, followed by mixing.

The procedure continued by preparing the selected grout mixes, casting them into plastic cylinders 150 mm long by 75 mm in diameter, and hardening for 24 h before demoulding. The specimens were cured at room temperature in a brine lime solution for 28 d. After curing, the specimens were axially cored through their centre to produce a hollow cylinder with a 25 mm wall thickness. All cylinders cured in NaCl solution cracked during the coring process. Attempts to produce intact hollow cylinders were unsuccessful. This was unexpected since all specimens cured in NaCl solutions exhibited the highest unconfined compressive strength. It is possible that due to curing in brine the material became brittle.

The hydraulic conductivity tests were carried out in specially constructed constant-head hydraulic conductivity cells capable of sustaining hydraulic pressures up to 10 MPa. Water flowed radially through the specimen, with the exterior pressure on the specimen higher than the interior. The flow into or out of the cell was measured to a precision of 0.01 mL.

All the tests were carried out under circumferential compression, with the direction of flow from outside the specimens toward the inside. The flow was monitored under cell pressure up to 8700 kPa with the grout core pressure kept at 0 psi. These relatively high hydraulic pressure differences were needed for all three grouts to transmit the water in the direction of the hydraulic gradient and to obtain a measurable through-flow of water. However, even at this hydraulic pressure difference, the grouts stored some water and the inflow was higher than the outflow.

The data show that the hydraulic conductivity ranged between 10.sup.-14 and 10.sup.-16 m/s. The hydraulic conductivities for grout mixes 94D and 94J (K=10.sup.-16 m/s) were two order of magnitude lower than the hydraulic conductivity of grout mix 94L (K=10.sup.-16 m/s). The observed difference in the hydraulic conductivity measured was attributed to the difference in microstructure of the hardened grouts.

The data also show that the hydraulic conductivity of the grout mixes decreased as the permeating time increased. The observed changes in the hydraulic conductivity were significant for grout mixes 94D and 94J. The data for grout mix 94L suggest that the hydraulic conductivity of the materials is approaching equilibrium. A number of factors account for the decreasing hydraulic conductivity with percolation time. These may include the presence of unhydrated material in the hardened grouts which, when exposed to water, hydrates to form solids with increased volume, consequently changing the microstructure of the hardened grout.

The results also show that the equilibrium flow condition has not been achieved (i.e., the inflow rate equal to the outflow rate) for the 55 d duration of the tests. The outflow rate, Q.sub.out, decreased with time and was lower than the inflow Q.sub.in. It appears that the specimens may take a very long time to reach an equilibrium flow condition. The latter is characterized by a constant discharge rate equal to the inflow rate. This may be related to the complex pore structure of the grouts. It may take some time for the water to enter the previously empty pore spaces under pressure. The deviation from equilibrium may also be attributed to the changes in the pore structure as the result of continued hydration.

The results from pore structure characterization using mercury intrusion porosimetry (MIP) indicate differences in the total pore volume and pore size distribution in the grouts investigated. The use of high water content (W/CM=0.7) in the preparation of grout mix 94D in the development of hardened grouts with a total pore volume of 0.3530 cm.sup.3 /g. The decrease in water content (W/CM=0.6) and the increase in the pumice content in the preparation of grouts 94J and 94L resulted in the development of a hardened grout with a low pore volume that ranged between 0.2983 cm.sup.3 /g and 0.3616 cm.sup.3 /g. The differences in the permeabilities of the grouts investigated are associated with the microstructural differences. The increase in the pumice content as well as the use of materials with a very fine particle size distribution resulted in the development of a microstructure with a lower, less connected pore volume.

The results from the pore structure characterization using MIP indicate changes occurred in the structure of the grout containing pumice during the permeability test. The grout samples showed a decrease in the total volume and shift of pore radii toward smaller values. The effects are considered to be mostly caused by the continued dehydration reaction as well as the continuation of the pozzolonic reactions and, therefore, progressing densification of the microstructure of the grout specimens when in contact with water. The absence of pores with diameters larger than 1 .mu.m (as measured by MIP) is taken as a somewhat arbitrary dividing point between "large" pores, which contribute most to permeability, and "small" pores, which are much less significant.

Change in porosity during permeability tests was related mainly to changes in the volume of solids caused by the formation of new hydration products as the result of an increase in the degree of hydration as well as the continuation of pozzolanic reaction. Microscopic analysis of the grout specimen, after exposure to the percolating water in the permeability tests, provides evidence to support this finding. Microscopic examination showed a much denser microstructure and large pores filled with solid hydration products.

Microstructure

Hardened cement paste is a heterogeneous assemblage of amorphous and crystalline phases bonded together in a porous mass. The microstructure of the hardened grout will vary greatly with W/CM ratio, grout composition, curing time and curing conditions. The microstructure will change very rapidly during the early stage, slowing down as the contact time between the cement and water increases. Therefore, the hardened grout will generate not a single microstructure, but a successive set of microstructures as the paste ages.

The grout mixes investigated in this study contained pulverized Type K grout and pulverized pumice as the pozzolanic material. The increase in the fineness of the materials used in the development of the new grouts was expected to affect the development of the material microstructure. Additionally, the large amount of pozzolan material used in the grout mixes, in an attempt to decrease the water required for enhancing workability of the unset pastes, was expected to react with developing hydration products, and thus change the microstructure of the hardened grouts.

Scanning electron microscopy (SEM) and MIP were used to investigate the microstructure of these novel materials, produced by mixing pulverized Type K grout and pulverized pumice as the pozzolanic material.

The observed microstructure for the three mixes differed from the developed microstructure in the reference Type K grout. The use of materials with a very fine particle size distribution (90% sub 5 .mu.m and average particle size below 3 .mu.m) and excess of pozzolanic materials led to the development of a homogenous, dense microstructure with very few isolated small pores dispersed in a precipitated matrix. It was clear from the examinations that the individual hydrated phases (i.e., CSH gel) are not independently dispersed but are cemented or fused together within the matrix.

The MIP results show that all grouts contained only gel pores representing the inherent structure and porosity of cement hydration products (FIG. 26). The gel pores play a major role in shrinkage, while the capillary pores are more important for the strength, permeability and fracture properties of the hardened grout.

The properties of a material are recognized to originate from its internal microstructure, and these properties can be modified by making suitable changes in the structure of the material. The low hydraulic conductivity (<10.sup.-14 m/s) and good mechanical strength of the developed grouts can be related to the grout's dense microstructure and lack of pores with diameters larger than 0.1 .mu.m.

Although the invention has been described in detail with particular reference to these preferred embodiments, other embodiments can achieve the same results. Variations and modifications of the present invention will be obvious to those skilled in the art and it is intended to cover in the appended claims all such modifications and equivalents. The entire disclosures of all references, applications, patents, and publications cited above, are hereby incorporated by reference.