FR2673620A1 - Composition for cementing oil wells at low temperature - Google Patents

Abstract

Slag composition for cementing oil, gas, water, geothermal and similar wells. The slag includes essentially plaster, cement and microsilica. Applications to the control of gas at low temperatures and in situations requiring a fast setting of the cement slag.

Description

COMPOSITION FOR THE CEMENT OF PETROLEUM WELLS
LOW TEMPERATURE
The present invention relates to the field of petroleum services and oil services, and in particular that of the cementation of the annular of an oil well, gas, water, geothermal and the like.

 Such cementing operations, as well as their importance for the life and the production of the well, are known to the skilled person for several decades, it is useless to recall the principle and details.

 For a good understanding of the text and the invention, it will be simply stated that it is a matter of injecting a cement slurry into the casing of the well. This cement slurry descends at the bottom of the well through the casing and, under the pumping pressure, back into the annulus between the borehole and the casing, to the surface of the well. Methods quite conventional exist so that, at the end of pumping, there is inside the casing an inert fluid and not the slag. The cement slurry is then allowed to take.

 The general purpose of the operation is to maintain the casing in the well, to isolate the underground zones crossed, and to consolidate the well itself.

 The cement slurry is pumped under pressure, and goes from the surface preparation conditions (pressure and ambient temperature) to the conditions of the bottom of the well (higher pressure and temperature) to return to the annulus, for a part of the slag, to conditions close to the surface. Under certain conditions, Arctic zones or submarine wells in the North Sea for example, the surface conditions can be very cold and the correct setting of the slag poses very particular problems.

 Despite these pressure and temperature cycles, the slag must naturally remain pumpable, it must perfectly move the drilling mud (critical factor of a good cementing operation), it must not lose fluid in the areas crossed, and take as quickly as possible but exactly at the intended place ie on the height of the ring and, naturally, especially not inside the casing. Finally, the hardened cement must have excellent mechanical properties and in particular good compressive strength, which must develop as quickly as possible after setting.

 When the well passes through a gas zone it is extremely important that during cementation, the hydrostatic pressure exerted by the fluid column in the annulus remains greater than the gas pressure in the formation. The density of the drilling fluids as well as that of the cement are adjusted accordingly.

 Once the slag is in place, he will start taking it. This intake is first of all expressed by the gelling of the slag. Links are created between the cement particles as well as between the cement and the walls of the well. Because of these connections, the cement column self-supports which results in a decrease in the hydrostatic pressure exerted in front of the gas zone. During this period, the viscosity of the cement is not yet sufficient to oppose the flow of gas bubbles. The end of this period corresponds to the hydration, exothermic reaction between the cement and the water which results in the rise of temperature, the development of the resistance to the compression as well as the chemical withdrawal of the system. During this phase, the pressure in front of the gas zone decreases even more and room is created in the form of porosity in the cement matrix where the gas can then migrate.

 During the entire setting, the cement must be able to resist a gas invasion that would have serious consequences ranging from poor cementation to the catastrophic surface gas explosion.

 The only enumeration of the required properties, which are often antagonistic or contradictory to each other, temperature and pressure conditions, and serious or catastrophic consequences of non-respect of the properties, makes it possible to measure the difficulties of developing a slag exhibiting the properties in question.

 The invention proposes a cement slag which has all of the properties listed above, and which is therefore particularly suitable for the most difficult wells, in particular those passing through gas zones, and at a low surface temperature. tell the wells that have the greatest difficulties.

 The cement slurry compositions according to the invention are characterized in that they essentially contain, in addition to the mixing water: - alpha plaster, - ordinary Portland cement (API class A or C), - microsilica, and, to the extent necessary, a retarding agent and a dispersing agent.

 The cementing of a well at low temperature (in particular <30 ° C.) presents quite special difficulties, in particular it is difficult for such wells to obtain a reasonable setting time of the cement slurry after a few hours. and a compressive strength of the order of 3500 kPa after eight hours It is recalled that setting time is the time required for the consistency of the slag measured at the consistometer according to API N2 10 to reach 100 Bc (Bc consistency Bearden ).

 Known accelerators such as calcium chloride can provide a solution but with a slow increase in consistency to 100 Bc and, when setting begins, it lasts several hours before the cement is hardened. During this time there is a rheumatism withdrawal and a possibility of gas migration. In addition, below 152C acceleration is very difficult to obtain.

 US Pat. No. 3,891,454 has already described the use of plaster (calcium sulphate semi-hydrate, CaSO4.1 / 2H2O) in a cement slurry in order to reduce the setting time. It is a question of replacing the cement with a mixture of plaster and cement in a mass ratio ranging from about 50:50 to 75:25. However, the addition of plaster in a cement slurry can cause serious disadvantages.

 In the first place, the setting time is extremely dependent on the type of plaster. In particular, most plasters are much too reactive in the presence of cement, and we often observe an acceleration of setting at low temperature, which represents in the oil field a rhédibitoire disadvantage (see above). In some cases it is possible to compensate for taking too fast a large amount of set retarder but then the taking can take days.

 Second, plaster / cement systems have virtually no filtrate control. Finally, plaster is known to be a thixotropic agent for cement, whereas according to the invention a fluid slag is desired.

The microsilica has already been used in a cement slurry (GB 2,179,933;
GB 2,212,150; GB 2,212,489; GB 2,587,988). It is a by-product derived from the manufacture of ferrosilicucts. The microsilica particles have an average size of about 0.15 microns and a specific surface area of the order of 20 m 2 / g.

Due to the difficulty in handling fine particles in the crude state, it is preferred to use the microsilica suspended in water.

 However, at low temperatures and in the concentration range where it is to be used, the microsilica lengthens the setting of the cement and retards the development of compressive strength thereby becoming one of the essential properties of gas control.

 Thus it was encountered in the prior art to the rhédibitoires disadvantages.

 The invention presents a system that solves for the first time the cumulative problems posed by a cementation at low temperature (0-30lC approximately) and the need for a system that can ensure a good control of gas migration.

 It is known that there are two main categories of plaster, calcium sulphate semi-hydrate type alpha and type beta, hereinafter alpha plaster and beta plaster. Their difference lies mainly in the preparation process. Alpha plasters are prepared under controlled vapor pressure while beta plasters are produced by a continuous fluidized bed process. Alpha plasters have an apparent density of about 1000 kg / m3, with spherical particles, while beta plasters have an apparent density of about 600 kg / m3 and are in the form of platelets.

 Figure 1 below represents a revolution in consistency as a function of time for different types of plaster. The operating conditions are as follows: 60% plaster, 40% API class A cement, density 1.89 kg / l, test at 209C.

 Curves A, B and C correspond to alpha plasters and have the much sought-after characteristic of right angle taps. Two beta-type plasters D and E have led to an immediate and uncontrolled setting of the mixture. Curve F corresponds to a beta plaster for which it has been possible to obtain a reasonable setting time by the use of retarder, without however reaching a right angle grip, nor a good resistance to compression in the short term.

The right-angle grip is a very fast increase in consistency at the time of setting, only a few minutes to go from 30
Bc at 100 Bc.

 It will be remembered that a right angle tap is one of the parameters involved in gas control. These results show that the use of alpha plaster in the present invention is quite critical.

 Figure 2 shows the influence of temperature on setting time, for a 60% alpha plaster system, 40% API class A cement and 1.89 kg / l density. Unlike cement whose setting time is reduced when the temperature increases, the opposite is observed in the case of a plaster / cement system. The foregoing reveals the complex nature of the physicochemical mechanisms involved in the invention.

Table 1 below summarizes the results obtained with respect to TT (Thickening Time) and Compressive Strength (CS) after 8 or 48 hours. We used an alpha plaster
LAMBERT, an ordinary Portland cement (CPA-HP), slag density 1.89 kg / l, 20go temperature, 0.3% by weight of dispersant (sodium polynaphthalene sulfonate formaldehyde or "PNS") and the retarder is a lignosulfonate sodium or calcium.

TABLE 1

<tb><SEP> Plaster <SEP> Cement <SEP> Reactor <SEP><SEP> Tr <SEP> 8hCS <SEP> 48 <SEP> h <SEP> CS <SEP>
<tb>% <SEP> in <SEP> weight <SEP>% <SEP> in <SEP> weight <SEP>% <SEP> in <SEP> weight <SEP> h: min <SEP> kPa <SEP> kPa
<tb><SEP> 60 <SEP> 40 <SEP> 0.6 <SEP> 3:50 <SEP> 8015 <SEQ> 13405
<tb><SEP> 50 <SEP> 50 <SEP> 0.6 <SEP> 3:35 <SEP> 7700 <SEP> 15750
<tb><SEP> 40 <SEP> 60 <SEP> 0.7 <SEP> 3:17 <SEP> 6160 <SEP> 13335
<tb><SEP> 30 <SEP> 70 <SEP> 0.8 <SEP> 3:52 <SEP> 1575 <SEP> 12320
<Tb>
The greater the amount of cement, the shorter the setting time of the plaster. However, a retarder such as lignosulfonate makes it possible, under the test conditions of Table 1 above, to easily adjust the setting time.

 Furthermore, it is very clearly observed that the compressive strength after 8 hours is ensured by the setting of the plaster and that the useful range of alpha plaster to obtain a compressive strength of the order of 3500 kPa at 8 hours lies above 30% by weight of mixture.

 According to the invention, it will necessarily incorporate the plaster / cement system a significant amount of microsilica especially for applications where the control of gas is required. The proportion of microsilica will preferably be between 15 and 25% by weight of the plaster / cement mixture. This amount is dictated by the need to provide the system with filtrate control below 100 cc / 30 min (Table 2 below). The use of microsilica in such proportions has, surprisingly, only a minor effect on the increase of the rheology of the system which can easily be diminished by the use of a conventional "PNS" type dispersant. It is further observed that the presence of microsilica provides the slag a particular stability which results in the complete absence of free water and sedimentation.

 During the development of the invention we encountered serious difficulties related to gelling problems during the use of certain class G oil cements in particular. This disadvantage is eliminated if only Portland cement (API class A or C) is used. In addition, ordinary Portland cements will be preferred.

 To obtain sufficient compressive strength in a short period of time (8 hours) and this in the presence of high concentrations of microsilica, it is necessary to use in the system a sufficient amount of plaster. On the other hand, it is observed that an excessively large amount impairs the long-term mechanical properties of the system. Consequently, in the present invention, only plaster / cement mixtures in a weight ratio ranging from 40:60 to 60:40 will be limited. The 60:40 ratio is generally the most appropriate (Table 3 below).

It is observed that the nature of alpha plaster as well as the variety of cement
Ordinary Portland have little influence on the properties of the system (Tables 3 and 4 below). Likewise, the desired property of right angle grip is not affected by the plaster / cement ratio nor by the presence of microsilica in the proportions recommended in the invention.

 Most of the systems studied during the development of the invention have a density of 1.89 kg / l. They contain a 60:40 plaster / cement mixture to which is added 20% microsilica by weight of mixture. Under these conditions, the amount of mixing water is 40% by weight of plaster / cement mixture. It is possible to lower the density of the slag without affecting either the stability or the rheology, however the filtrate control and compressive strength at 8 hours will be considerably reduced. To remain within the scope of the invention, where a slag having good anti-gas migration properties is sought, the lower limit in density is 1.80 kg / l, which corresponds to a quantity of mixing water of 50% by weight of plaster / cement mix.

 In reading the present description, Tables 1 to 4 and Figures 1, 2, 3, 4, 5, 6 and 7, those skilled in the art will understand that a representative composition of the invention is characterized in that it contains essentially: - cement, - 66 to 150% of plaster alpha relative to the weight of the cement, - 15 to 25% by weight of microsilica with respect to the weight of the mixture cement + plaster, - 30 to 50% of water by relative to the weight of cement + plaster mix, and possibly a retarder and a dispersant.

 On the other hand, higher densities are easily conceivable, but often without any practical interest for the operations specifically targeted by the invention.

 The plaster / cement / microsilica systems described above therefore have some of the essential characteristics of gas control: stability, low rheology, good filtrate control, right angle grip and rapid development of compressive strength.

 There is no standard API test for measuring the ability of a cement slurry to retain gas. The method used for this invention is briefly described below.

 A cell 33 cm high (Figure 3) closed at the top end is filled with cement slurry and placed in a thermostatically controlled chamber. The particular shape of the cell was developed to avoid experimental artefacts and force the gas to migrate inside the cement and not at the cement / wall interface. The bottom of the cell is connected to a nitrogen line under 17.5 bar pressure. During the experiment, the pressure at the top of the cell, the temperature of the slag and the flow of nitrogen entering the cell are measured.

 Throughout the freezing period, see above, the slag behaves like a liquid that transmits the entire gas pressure. During this period, the pressure measured at the top of the cell remains constant as well as the gas flow rate and the temperature. Then comes the exothermic part of the hydration.

This results in a rise in temperature, the chemical shrinkage of the cement and the development of compressive strength. The pore pressure in the cement matrix drops due to the place created by chemical shrinkage. Depending on the system studied, the gas will enter more or less inside the cement. A system will be said to be resistant to gas invasion if the pressure at the top of the cell decreases and if the gas inlet flow rate remains low, or better still zero.

The typical behavior of a plaster / cement system is represented on the
Figure 4. This is a system containing 60% by weight of alpha plaster and 40% API class A cement (CPA-HP). The first peak of temperature at about 3 hours corresponds to the setting of the plaster and at the same time a peak of gas inlet is observed to fill the porosity created by the chemical shrinkage of the system. Between 8 and 20 hours, we observe a second gas inlet which corresponds to the setting of the cement. Throughout the experiment, the gas pressure is transmitted to the top of the cell, there has been gas migration in the system and permanent communication between the bottom and the top of the cell.

 If during a gas migration experiment there is permanent communication between the bottom and the top of the cell, then the total amount of gas entering the cell is equal to the amount of chemical shrinkage. During the experiment shown in FIG. 4, the total amount of gas corresponds to 7% of the volume of the cell whereas for a cement only only 4% shrinkage is observed during this phase. The internal shrinkage of a plaster / cement system is therefore clearly superior to the removal of the cement alone, which favors the invasion of the gas, which was naturally considered by the person skilled in the art as a serious disadvantage.

The effect of the microsilica is shown in Figure 5 where 20% microsilica was added to the previous system by weight of plaster plus cement. At the beginning of
In the test, the pressure drops rapidly to the top of the cell until it reaches about 6 bar. Because of this differential pressure (11.5 bar on 33 cm), a very slight gas entry occurs during setting plaster. The pressure at the top of the cell rises to 11.5 bar and then drops back to around 28 hours. Throughout the experiment, the pressure measured at the top of the cell remained below the applied gas pressure. There has therefore never been gas communication between the bottom and the top of the cell despite the differential pressures involved.

 We can observe (Figure 6) the same type of behavior at 59C with another type of cement (LAFARGE class A).

 Taking a right angle in a plaster / cement system is not enough to prevent the migration of gas. The presence of microsilica is therefore essential, in synergy with cement and plaster, for the particular applications envisaged.

 The results presented above demonstrate the remarkable properties of the alpha plaster / ordinary Portland cement / microsilica systems for the control of gas during a cementing operation.

 Once the system is taken, it is also important that the integrity of the cement column is maintained throughout the life of the well (several years). However plaster / cement systems are known to have longevity problems. In fact, the sulphate ions contained in the plaster react chemically with the aluminate phases (C3A) contained in the cement. The reaction product (ettringite) has a larger volume than the sum of the initial component volumes which causes an expansion of the system. Too much expansion may cause microcracks in the material.

 Figure 7 shows the expansion curve of a plaster / cement system where 2.5% sample expansion is observed after about five months. The slope of the curve suggests that expansion may continue. On the contrary, the system that contains microsilica has a limited expansion to 0.4% which seems not to progress beyond 8 days. This particularity was not predictable.

TABLE 2
API class A cement (40% by weight)
LAMBERT alpha plaster (60% by weight)
Microsilice MICROBLOCK
Density of slag 1.89 kg / l

Rheology <SEP> after <SEP> mixing <SEP> Rheology <SEP> to <SEP> 15 C
<tb> Microsilica <SEP> Retardant <SEP> Dispersant <SEP> PVii <SEP> YPiii <SEP> PVii <SEP> YPiii <SEP> 10 <SEP> min <SEP> freeze <SEP> TTiv <SEP> FLv
<tb><SEP>%<SEP> weight <SEP>% <SEP> weight <SEP>% <SEP> weight <SEP> Pa.s <SEP> Pa <SEP> Pa.s <SEP> Pa <SEP> Pa <SEP> hr: min <SEP> cm3 / 30 <SEP> min
<tb> - <SEP> 0.6 <SEP> 0.3 <SEP> 0.111 <SEP> 0.5 <SEP> 0.072 <SEP> 1.4 <SEP> 04.8 <SEP> 2:30 <SEP > 426
<tb> 10 <SEP> 0.6 <SEP> 0.3 <SEP> 0.048 <SEP> 3.6 <SEP> 0.066 <SEP> 5.1 <SEP> 17.2 <SEP> 0:55 <SEP > 10 <SEP> 0.8 <SEP> 0.3 <SEP> 0.045 <SEP> 2.0 <SEP> 0.060 <SEP> 2.4 <SEP> 12.9 <SEP> 2 :: 30 <SEP> 146
<tb> 15 <SEP> 0.9 <SEP> 0.4 <SEP> 0.046 <SEP> 2.4 <SEP> 0.059 <SEP> 3.2 <SEP> 16.8 <SEP> 2:15 <SEP > 92
<tb> 20 <SEP> 1.0 <SEP> 0.5 <SEP> 0.047 <SEP> 3.4 <SEP> 0.080 <SEP> 3.7 <SEP> 20.1 <SEP> 1:40 <SEP > 66
<tb> 20 <SEP> 1.2 <SEP> 0.6 <SEP> 0.040 <SEP> 2.3 <SEP> 0.049 <SEP> 1.6 <SEP> 10.5 <SEP> 4 :: 50 <SEP> 64
<tb> 25 <SEP> 1.4 <SEP> 0.7 <SEP> 0.041 <SEP> 2.1 <SEP> 0.050 <SEP> 1.4 <SEP> 09.1 <SEP> - <SEP> 50
<tb> ii PV Plastic Viscosity iii YP Shear Threshold iv TT Setting Time v FL Loss of Fluid TABLE 3
API class A cement (40% by weight)
LAMBERT alpha plaster (60% by weight)
Microsilice MICROBLOCK (20% by weight of plaster / cement mix)
Density of slag 1.89 kg / l

CSvi
<tb> Cement <SEP> Retardant <SEP> Dispersant <SEP> BHCTi <SEP> PVii <SEP> YPiii <SEP> Gel <SEP> to <SEP> 10 <SEP> min <SEP> TTiv <SEP> FLv <SEP > 8 <SEP> h <SEP> 72 <SEP> h
<tb> type <SEP>% <SEP> weight <SEP>% <SEP> weight <SEP> C <SEP> Pa.s <SEP> Pa <SEP> Pa <SEP> hr: mn <SEP> cm3 / 30 <SEP> min <SEP> kPa <SEP> kPa
<tb> CPA-HP <SEP> 1.0 <SEP> 0.6 <SEP> 5 <SEP> 0.053 <SEP> 2.3 <SEP> 15.3 <SEP> 3 :: 50 <SEP> 68 <SEP> 3437 <SEP> 10402
<tb> CPA-HP <SEP> 1.1 <SEP> 0.6 <SEP> 5 <SEP> - <SEP> - <SEP> - <SEP> 5:20 <SEP> - <SEP> - <SEP >
CPA-HP <SEP> 1.0 <SEP> 0.6 <SEP> 15 <SEP> 0.047 <SEP> 1.4 <SEP> 10.1 <SEP> 3:10 <SEP> 56 <SEP> 3108 <SEP> 14924
<tb> CPA-HP <SEP> 1.0 <SEP> 0.6 <SEP> 25 <SEP> 0.040 <SEP> 1.7 <SEP> 09.6 <SEP> 4:30 <SEP> - <SEP > 3318 <SEP> 15421
<tb> LONESTAR <SEP> 1.0 <SEP> 0.6 <SEP> 15 <SEP> 0.052 <SEP> 3.1 <SEP> 37.3 <SEK> 2:45 <SEK> 68 <SEP> - <September>
LAFARGE <SEP> 1.0 <SEP> 0.8 <SEP> 5 <SEP> 0.068 <SEP> 2.7 <SEP> 22.5 <SEP> 3:20 <SEP> 72 <SEP> 3374 <SEP> 7651
<tb> LAFARGE <SEP> 1,2 <SEP> 1,0 <SEP> 5 <SEP> - <SEP> - <SEP> - <SEP> 4:30 <SEP> - <SEP> - <SEP>
LAFARGE <SEP> 1.0 <SEP> 0.6 <SEP> 15 <SEP> 0.057 <SEP> 2.9 <SEP> 33.5 <SEP> 1:55 <SE> 64 <SE> 3948 <SEP> 16177
<tb> LAFARGE <SEP> 1.0 <SEP> 0.8 <SEP> 25 <SEP> 0.054 <SEP> 2.3 <SEP> 27.3 <SEP> 2 :: 00 <SEP> - <SEP> 3185 <SEP> 13790
<tb> i BHCT Bottom Flow Temperature ii PV Plastic Viscosity iii YP Shear Threshold iv TT Setting Time v FL Fluid Loss vi CS Compressive Strength TABLE 4
API class A cement (40% by weight)
Alpha plaster GEORGIA PACIFIC (60% by weight)
Microsilice MICROBLOCK (20% by weight of plaster / cement mix)
Density of slag 1.89 kg / l

CSvi
<tb> Cement <SEP> Retardant <SEP> Dispersant <SEP> BHCTi <SEP> PVii <SEP> YPiii <SEP> Gel <SEP> to <SEP> 10 <SEP> min <SEP> TTiv <SEP> Flv <SEP > 8 <SEP> h <SEP> 72 <SEP> h
<tb> type <SEP>% <SEP> weight <SEP>% <SEP> weight <SEP> C <SEP> Pa.s <SEP> Pa <SEP> Pa <SEP> hr: mn <SEP> cm3 / 30 <SEP> min <SEP> kPa <SEP> kPa
<tb> LONESTAR <SEP> 1,2 <SEP> 1.0 <SEP> 15 <SEP> 0.080 <SEP> 2.4 <SEP> 28.2 <SEP> 2 :: 20 <SEP> 58 <SEP> - <SEP>
LONESTAR <SEP> 1.3 <SEP> 1.0 <SEP> 15 <SEP> 0.082 <SEP> 2.4 <SEP> 35.0 <SEP> 3:15 <SE> 56 <SE> 2058 <SEP> 8792
<tb> LONESTAR <SEP> 1.4 <SEP> 1.0 <SEP> 15 <SEP> 0.084 <SEP> 1.4 <SEP> 23.0 <SEP> 4:20 <SE> 54 <SEP> - <September>
LAFARCE <SEP> 1.3 <SEP> 1.0 <SEP> 15 <SEP> 0.089 <SEP> 2.6 <SEP> 36.9 <SEP> 2:20 <SEP> 60 <SEP> - <SEP>
LAFARGE <SEP> 1.3 <SEP> 1.2 <SEP> 15 <SEP> 0.099 <SEP> 2.1 <SEP> 24.9 <SEP> 1:50 <SEP> 56 <SEP> - <SEP>
LAFARGE <SEP> 1.4 <SEP> 1.0 <SEP> 15 <SEP> 0.112 <SEP> 3.6 <SEP> 36.9 <SEP> 2:10 <SEP> 66 <SEP> - <SEP>
LAFARGE <SEP> 1.5 <SEP> 1.2 <SEP> 15 <SEP> 0.092 <SEP> 2.0 <SEP> 22.5 <SEP> 3:35 <SEP> 52 <SEP> - <SEP>
LAFARGE <SEP> 1.4 <SEP> 1.0 <SEP> 25 <SEP> 0.081 <SEP> 2.4 <SEP> 27.3 <SEP> 2:45 <SEP> 56 <SEP> 3094 <SEP> 3423
<tb> LAFARGE <SEP> 1.5 <SEP> 1.0 <SEP> 25 <SEP> 0.086 <SEP> 1.8 <SEP> 13.4 <SEP> 3:: 30 <SEP> 54 <SEP> - <SEP> i BHCT Bottom Flow Temperature ii PV Plastic Viscosity iii YP Shear Threshold iv TT Setting Time v FL Fluid Loss vi CS Compressive Strength

Claims (8)

 1. A composition of cement slurry for oil, gas, water, geothermal and similar wells, characterized in that it essentially contains: cement, -66 to 150% of alpha plaster relative to the weight of the cement, 15 to 25% by weight of microsilica with respect to the weight of the cement + plaster mixture, 30 to 50% of water relative to the weight of cement + plaster mixture. 2. Composition according to claim 1 characterized in that the cement is an ordinary PORTLAND cement (API class A or C). 3. Composition according to claims 1 or 2 characterized in that it contains a set retarder. 4. Composition according to claim 3, characterized in that the retarder is a calcium or sodium lignosulfonate. 5. Composition according to claims 1 to 4 characterized in that it contains a dispersant, especially of the sodium polynaphthalene sulfonate / formaldehyde type. 6. Use of the compositions according to claims 1 to 5 for the cementing of oil wells, gas, water, geothermal and the like, having a temperature of 0 to 30g. 7. Use according to claim 6 for cementing wells with a risk of gas invasion. 8. Use according to claims 6 or 7 for the cementing wells at low temperatures and with a risk of gas invasion, especially for cementing surface casings.