This study deals with the specific injection of grouts into granite bl
ocks which present cracks having an opening of less than 100 mu m. Two
conditions must be met injection must not cause the cracks to open du
e to the use of low pressure (< 0.1 MPa), in which case the dimensiona
l stress between the size of the solid elements of the grout and the s
ite of the cracks is increased; long-lasting impermeability must be at
tained; the hardened grout must be stable in a potentially aggressive
environment. A possible solution would be to inject grouts of very fin
e cement appropriate to the setting. The purpose of this paper is to p
resent some criteria to characterise the penetrability of such grouts
in a fairly impermeable milieu. The standardised ''sand column'' test
was used for the experiments. The optimal grout fluidity was sought (t
hirteen superplastifiers were tested and their optimal dosage determin
ed). The study shows that the criterium of fluidity alone is not suffi
cient to characterise the penetrability of grouts. The evaluation of t
heir stability particularly by a filtration test with a low pressure f
ilter press, provides complementary data for the characterisation of t
heir penetrability. The negative effect of the entrained air on penetr
ability runs also noted. Two superplasticisers, one from the condensed
sulfonated naphthalene formaldehyde group and the other from the cond
ensed sulfonated melamine formaldehyde group, proved to be the most ef
fective.