Spacer Fluids for Use With Rock-Based Geopolymer

Over the past few decades, various substitutes for ordinary Portland cement (OPC) have been extensively researched. This exploration has been driven by the goals of enhancing performance, reducing costs, and minimizing environmental effects by curbing greenhouse gas emissions [1]. Among the proposed alternatives, geopolymers stand out as a notable option with low chemical shrinkage, permeability, fluid loss, and high durability in corrosive downhole environments [2–5]. In primary cementing, drilling fluid needs to be removed from the annulus and replaced by cement, which then transitions from a fluid to a solid state through the hydration process. Drilling fluid contamination, which can manifest as intermixing, channeling of cement toward the wider side of an eccentric annulus, and residual mud layers on the walls, significantly impacts the final quality of the hardened cement behind the casing. Geopolymer, as an alternative material to cement, is not an exception in this regard. In the liquid state, the presence of water-based drilling fluid (WBDF) lowers the flow curve, while the presence of oil-based drilling fluid (OBDF) increases the gel strength of the geopolymer. Contamination with WBDF is more detrimental to the mechanical properties of the geopolymer compared to OBDF. When contaminated with 10% WBDF, geopolymer slurry does not harden after 1 day [6–8]. The presence of OBDF on the casing wall can decrease the bonding strength of the geopolymer to the casing by inverting the wettability of the surface to oil-wet. This implies the importance of using proper spacer to efficiently displace the mud in the annulus and avoid direct contact between mud and geopolymer. In the selection process of spacer, there are some criteria that should be considered: (1) adjusting rheological properties and density of the spacer to meet the design rules for complete mud removal. For instance, the displacing fluid must have density at least 10% higher than that of the displaced fluid and the frictional pressure gradient produced by the displacing fluid should exceed that of the displaced fluid by at least 20%. (2) Improving bonding of geopolymer by washing the formation or steel interface from mud and making it water-wet. (3) Being compatible with both drilling fluid and geopolymer, meaning it should not significantly affect their properties in case of intermixing. The first criterion is paramount since even a compatible spacer can fail if its viscosity and density are not properly designed. Poorly designed spacers increase the likelihood of channeling and direct contact between the geopolymer and mud, which can compromise the cementing job.

With the constant evolution in the needs of the energy industry and an increased emphasis on the durability of zonal isolation, new formulations and technologies for spacer fluids have been introduced. Recently, spacers with unique structures, such as microemulsions and nano-emulsions, have demonstrated acceptable potential for the removal of OBDF [9–11]. However, there are drawbacks linked to this type of spacer, including the high cost of preparation, having a negative influence on cement hydration [12], and concerns regarding potential environmental impacts. Another type of spacer is hardening spacer which is based on the geopolymerization concept and the ability to solidify the remnant drilling fluid [13]. Rheological compatibility of such fluid and OBDF has been previously studied [14]. The hardening spacer investigated by Li et al. [15] also shows curing compatibility with OPC and it aids the bonding of cement to the formation surface. Nevertheless, the compatibility of such fluids with geopolymer has not been studied.

Geopolymer is currently evaluated for oil and gas applications. Recent efforts have been directed toward adapting geopolymer as a suitable barrier material for well cementing [3,16,17]. This involves adjusting its pumpability with retarders [18] and enhancing its zonal isolation capabilities by introducing expansive agents [19]. To unlock the substantial benefits of geopolymers as an alternative to OPC in primary cementing, it becomes imperative to formulate specialized spacer fluids tailored for geopolymers to minimize drilling fluid contamination.

In this study, we aim to develop a spacer that hardens with time and could improve well integrity if left in the wellbore. First, we investigate the impact of different rheology modifiers on the flow behavior of the spacer. This would pave the way for tuning the viscosity profile with respect to the geopolymer profile. Afterward, we adjust the surfactant content to ensure that the spacer performs efficiently in the presence of OBDF. Finally, contamination of geopolymer by designed spacer is studied. The design process of the spacer for geopolymer is presented in Fig. 1. The spacer in this study is designed for cementing jobs in intermediate and production casing with a bottom-hole circulation temperature (BHCT) of 50 °C. The hardening ability of the spacer, although beneficial for wellbore integrity if designed carefully, comes with challenges, such as being sensitive to temperature. Examining the impact of the temperature variation on the performance of spacer is out of the scope of this work.

Table 1 shows the composition of the base spacer slurry. Ground granulated blast-furnace slag (GGBFS) and granite were used as precursors. Table 2 presents the oxide composition of precursors as determined by X-ray fluorescence (XRF) analysis. Na₂CO₃, KOH (12 M), and potassium silicate solution with SiO₂/K₂O molar ratio of 0.98 were used as activators. The properties of the hardening spacer are listed in Table 3.

A rock-based geopolymer with properties suitable for well cementing at 50 °C of BHCT and 70 °C of bottom-hole static temperature (BHST) was used. For more information regarding mix design, refer to Ref. [18]. The slurries were prepared in accordance with the API 10B-2 [20]. A liquid hardener was incorporated into a solid precursor that is rich in aluminosilicate. Before conducting each set of tests, the slurry was conditioned in an atmospheric consistometer for a time duration of 30 min, once BHCT was attained.

The mix design of the tested OBDF is shown in Table 4. The oil/water ratio for this lab-formulated emulsion was 74/26 and the density measured with pressurized mud balance was 1.16 s g.

The spacer design process had three stages.

In this stage, the effect of viscosifiers such as xanthan gum (XG), polyanionic cellulose (PAC), and bentonite on the rheological properties of spacer was investigated. These components were added to the base spacer in varying concentrations, as detailed in Table 5. In the case of bentonite, it was prehydrated in water for 30 min. The concentration of each viscosifier in the spacer mix design was adjusted to maintain the viscosity and yield stress hierarchy with geopolymer.

In the second stage, different concentration (by wt% of water) of the surfactant package was added to the base spacer to find the most effective concentration that could clean the surface of casing and invert the wettability of surface. The selected concentration was the one that achieved the highest cleaning efficiency in the rotor cleaning test while minimizing the contact angle of water on the steel coupon after washing. Fatty alcohol surfactant and 2-butoxyethanol solvent were used as surfactant package.

In the last stage, geopolymer was mixed with the optimized spacer mix design from stages 1 and 2 with the mixture ratio recommended by API (95/5, 75/25, 50/50, 25/75, and 5/95 geopolymer/spacer), and rheological properties were measured at 50 °C and atmospheric pressure. Geopolymer was also contaminated by 10% and 20% by volume with the optimized spacer and cured for 1 day and 7 days at 70 °C and 13.8 MPa. Afterward, the mechanical properties of the cured geopolymer were measured. Mix identification (Id) and concentration of admixtures in each spacer sample are presented in Table 5.

The thickening time of the slurry was measured using an atmospheric consistometer according to API 10B-2 [20]. The pumpability of slurries is measured in Bearden Consistency (BC) Units at elevated temperatures. In oil well cementing operation, according to operator's criteria, 30–40 BC is regarded as a value beyond which the slurry is not capable of being pumped [22]. In start–stop mode, the consistency of the slurry is monitored after a period of rotation stoppage which mimics any interruption in the pumping job.

This test evaluates the effectiveness of spacer in water-wetting the surface of casing. First, steel coupon was dipped in the OBDF for 10 min. Following that, steel coupons were placed in the container that is filled with spacer fluid and the fluid mixed at 100 rpm for 30 min. Afterward, the contact angle of water and the steel coupons that are washed with spacer fluid are measured using Kruss drop shape analyzer.

X-ray diffraction (XRD) of the cured geopolymer when contaminated with spacer were analyzed using a Bruker-AXS Micro-diffractometer D8 Advance. The analysis was performed over an angle 2θ which range from 5 to 70 deg.

An isothermal calorimeter was used to evaluate the impact of extra water on the geopolymer thermal behavior. After mixing of geopolymers, 5 g of the slurry was poured into a plastic ampoule and placed into calorimeter and the heat evolution was monitored for 60 h at 50 °C.

The rheological properties of a spacer fluid are a major factor determining the quality of fluid displacement in the annulus. The ratio of the yield stress between displaced and displacing fluid is an indication of whether there will be a static mud layer on the wall [24] and viscosity hierarchy plays a positive role in increasing the displacement efficiency in the inclined sections [25]. Viscosity hierarchy can be defined as higher shear stress of the displacing fluid compared to displaced fluid at shear rate that is common during primary cementing. Thus, it is important to optimize rheology of the spacer with respect to the rheology of drilling fluid and geopolymer. Figure 2 shows the impact of different concentrations of xanthan gum as a viscosifier on base spacer fluid. It is clear that by increasing the concentration of xanthan gum, the viscosity of the spacer increases. The viscosity of all fluids decreases sharply as the shear rate increases, indicating that these systems have a non-Newtonian shear-thinning behavior. This behavior is caused by reduced intermolecular interactions and polymer network disentanglement, leading to reduced friction forces and partial alignment of macromolecules in the direction of shear flow [26]. Upon shear removal, the network structure rapidly reforms by Brownian motion, recovering the viscosity almost instantaneously. This explains the similarity between the ramp up (filled symbols in the plot) and ramp down shear stress measurements. The high shear viscosity at low shear rates or at rest results from hydrogen bonding and polymer entanglement of xanthan gum's large molecules [27]. The fluids that had low viscosities experienced flow instabilities at high shear rates during rheometry mainly due to creation of Taylor vortex flow [28]. Consequently, the shear stress measurements are not shown here for shear rates higher than 500 l/s for some of the fluids.

Figure 3 shows the flow curve of the base spacer with different concentrations of PAC added. There was an increase in the flow curve values with a higher concentration of PAC and a behavior close to that of a Bingham fluid was observed. Furthermore, Table 5 shows the viscosity model parameters calculated from fitting the flow curve of spacer fluids with Bingham and H–B (Herschel–Bulkley). Base spacer fluid had behavior close to Bingham fluid with low yield stress calculated with different models. By increasing the concentration of XG in the spacer, n decreased drastically which means the degree of shear-thinning of the fluid increased. Yield stress was also increased with the addition of XG. Notably, Bingham model predicted higher yield stress compared to H–B. Table 5 also shows the viscosity model parameters for the spacer with PAC. Compared to XG with the same concentration, PAC decreased n to a less extent which means behavior closer to Bingham is expected. This could be an advantage for the spacer fluid as it may improve its displacement performance, based on the study by Nguyen et al. [29]. The addition of PAC also resulted in lower yield stress compared to XG with equal concentration (Table 7).

Figure 4 shows the flow curve of the base spacer with different concentrations of bentonite added under temperature of 50 °C. A peak in shear stress between the shear rates 0 and 100 l/s was observed for the fluids containing bentonite which is attributed to viscoelasticity. Bentonite has a shape of platelets that tends to form a house of cards structure due to electrostatic charges on the surface and edges of these platelets [30]. This results in a continuous gel structure at conditions close to rest that can break at a certain shear rate (peak in shear stress) which increases as the concentration of the clay increases. Due to viscoelasticity, Bingham and H–B models could not fit the flow curves accurately and the resulting R² was low (not shown here). Thus, it was decided to find yield stress with the oscillatory method although knowing that it can result in different approximations compared to the rotational method. Figure 5 shows the shear strain amplitude sweep of the base spacer and spacer with added bentonite at 50 °C measured with the oscillatory method. The storage modulus (G′) of a material represents the energy it stores when it is deformed and later uses to return to its initial structure, indicating its elastic behavior. In contrast, the loss modulus (G″) is an indication of the material's viscous behavior during deformation, representing the energy lost due to internal friction between particles and molecules. The addition of bentonite increased G′ from approximately 100 Pa with no bentonite to close to 5500 Pa with 6 wt% bentonite. G″ decreased with the addition of bentonite up to 2 wt% while 4 and 6 wt% bentonite increased the G″. When the loss modulus becomes equal to or higher than the storage modulus, the material structure undergoes a change to such an extent that it begins to flow (flow point). Measured flow points for spacer with different concentration of bentonite are shown in Table 8. The addition of bentonite increased the flow point of the spacer.

Generally, in vertical sections with larger annular spaces, gravity is the dominant force, and having a higher density contrast between fluids (geopolymer/spacer/drilling fluid) can prevent channeling. In contrast, in lower inclined sections with smaller annular gaps, viscous forces play a more critical role, requiring more frictional pressure to effectively displace the mud [31]. Design rules for primary cementing suggest that the spacer density should be 10% lower than the cement and 10% higher than the drilling fluid, with the viscosity 20% lower than the cement and 20% higher than the drilling fluid [32]. While these guidelines are sufficiently accurate for many cases, they can lead to false predictions for near-horizontal and horizontal wells, where a positive density difference can cause displacing fluid slumping downward in the wellbore [33]. However, in this research work, the recommended range of density and viscosity contrast was not the objective to propose. Therefore, it is recommended to run a displacement simulation for the specific well, considering its geometry and operational information, to get an overview of the density and viscosity of the spacer before performing the actual cementing job.

The ability of spacer fluid to clean the surface of casing was determined by rotor cleaning test. Figure 6 shows the viscometer sleeve covered with OBDF and after washing with spacer that is surfactant-free. Some traces of mud on the surface of sleeve were notable. Surfactant in the spacer design resulted in better cleaning of viscometer sleeve (Fig. 7). For comparison, same test was performed with water as spacer and Fig. 8 shows that remarkable OBDF was left on the surface. It can be concluded that particles in the spacer are beneficial for cleaning of the casing wall. Cleaning efficiency calculated based on the formula is also provided in Fig. 9. It can be seen that spacer without surfactant had cleaning efficiency of 72% and addition of 0.83 wt% (by water) surfactant improved the cleaning efficiency to 94%. Once a certain concentration of the surfactant package was reached, the cleaning efficiency reached a plateau, meaning that increasing the concentration beyond this point does not result in any further improvement in cleaning efficiency. Using water as spacer fluid resulted in 37.5% cleaning efficiency indicating that using water as preflush is far enough in cementing operation.

Figures 10 and 11 show the contact angle of water on the surface of steel when treated with different fluids. The clean surface of steel coupon had a contact angle of 72 deg which is considered as intermediate wettability [34]. After immersing the steel in the OBDF, the contact angle remained almost same. By washing the mud-covered steel with the spacer that has surfactant in the design (SP-S1), the contact angle reduced significantly, resulting in a strong water-wet surface. The lipophilic tail of the surfactant molecules attaches to the steel surface, causing it to become water-wet and enabling water droplets to spread almost completely. Further increasing the concentration of surfactant reduced the contact angle to 5 deg.

The spacer fluid is in direct contact with the geopolymer as a cementitious material in the wellbore. Thus, the impact of spacer fluid on the rheological and mechanical properties should be investigated.

The flow curve of geopolymer/spacer with mixture ratio recommended by API (95/5, 75/25, 5/95, 25/75, and 50/50) was measured at 50 °C with rotational viscometer (Fig. 12). For instance, 95/5 stands for 95% (vol%) geopolymer and 5% spacer in the mixture and 5/95 stands for 5% spacer and 95% geopolymer in the mixture. It can be noticed that spacer had a higher shear-thinning behavior than the geopolymer slurry. As the spacer was added to geopolymer, the shear stress reduced. This is normal since the spacer has higher water content and dilutes the geopolymer resulting in a lower particle fraction and viscosity. R-index calculated based on the formula is shown in Table 9, expressing the degree of compatibility of the two fluids. Based on the criteria, fluids are rheologically compatible in the investigated shear rate region. In other words, the rheological characteristics of the mixtures maintain a relatively consistent profile compared to the rheological behavior of the individual fluids. Gel strength of same mixtures was measured following API 10B-2 [20]. Figure 13 shows that 25/75 mixture ratio had unexpectedly higher 10 s and 10 min gel strength. It is believed that high pH of geopolymer accelerates the gelation of spacer which contains notable amount of GGBFS. Gelation can make the pumping job harder or lead to high friction pressure, which is undesirable as it may cause damage to the weak formation. The pumpability of specific mixture ratio that gelation happened was investigated with the start–stop method. The result shows (Fig. 14) that recorded consistency was low and jumped back to initial values after restarting the consistometer following stop period.

Figure 15 shows the compressive strength of geopolymer contaminated with 10% and 20% (by volume) spacer. Three mix designs of spacer with different solid to water and granite to GGBFS ratios were tested (Table 5). The results show that spacer that had higher content of GGBFS (SP-O2) had slightly lower contamination impact on geopolymer compared to SP-O1. Contamination with 10% SP-O2 reduced the compressive strength of geopolymer from approximately 6 MPa to less than 3 MPa after 1 day of curing. Contamination with SP-O1 with same percentage reduced the strength to approximately 2 MPa. GGBFS which is a reactive calcium source can provide a chemical binding of water which reduces the permeability of geopolymer matrix [35,36].

The impact of spacer contamination on the tensile strength of spacer is shown in Fig. 16. Tensile strength of geopolymer after 1 day reduced from approximately 0.5 MPa to less than 0.35 and 0.25 MPa with 10% contamination by SP-O2 and SP-O1, respectively. Water content plays significant role in strength of geopolymer and having water to solid higher than threshold may deteriorate the strength completely [37,38]. Water primarily acts as a reaction medium and the geopolymerization process may generate additional water, which, along with the initial mixing water, is present in the form of evaporable water and is located within small pores [39]. Hence, the water content of spacer fluid should carefully be monitored. The comparison between two different spacer designs (SP-O1 and SP-O3) demonstrated that an increase in the solid/water ratio resulted in a significant improvement in the compressive and tensile strength of the contaminated geopolymer after 1 and 7 days.

Figure 17 shows the XRD patterns of the cured samples of neat geopolymer and geopolymer contaminated with spacer. The main phase in both neat and contaminated geopolymer was Quartz (Qz) due to using granite as solid main precursor. Minor phases of albite (Alb), microcline (Mic), and biotite (Bio) were also present. The analysis of the contaminated samples did not reveal the formation of any new phases, indicating that there was no reaction between the spacer and geopolymer that forms new minerals. However, the intensity of peaks changed with contamination. Addition of 10% and 20% SP-O1 to geopolymer increased the intensity of quartz phase. The changes in intensity were less noticeable for the SP-O3 contamination case. The highly alkaline environment causes the dissolution of the crystalline phases in the precursors, which are then consumed in the reaction during the geopolymerization [40].

To further investigate the role of water in geopolymerization, 2.5% and 5% water (by volume of geopolymer) were added to geopolymer slurry and heat evolution of slurries were monitored. Figures 18(a) and 18(b) show the normalized heat flow in early period and total duration, respectively. The first exothermic peak occurs in less than 10 min and is attributed to dissolution of solid precursor in alkali activator. When the amount of water was increased, this peak decreased, meaning that reaction heat evolution rate hindered. When additional water is added to geopolymer, the OH⁻ concentration in activator solution is dropped because of dilution effect resulting in impaired ability of breakage of Si–O and Al–O bonds [41–43]. Neat geopolymer exhibited second peak approximately after 17 h. This peak is related to polymerization of alumina and silica species which releases energy. When 2.5% and 5% water was added to geopolymer, the time to reach the peak increased to 23 and 32 h, respectively, which means the additional water delayed polymerization. The intensity of the peaks was also diminished by increasing the water content. During polymerization phase, water functions as a byproduct, and an excessive amount of it has the potential to impede the kinetics of reaction [42].

Development of spacer fluid for geopolymer as cementitious material was discussed. The base spacer had low viscosity and ability to harden. The design procedure had three stages of tuning the rheology, optimizing the surfactant package, and contamination study. The main conclusions were:

1. The viscous parameters, flow behavior index, and yield stress, which are important parameters for displacement, exhibit higher sensitivity to XG than PAC content. The presence of static mud layer is dependent on the ratio of the yield stress between displaced and displacing fluid.

2. Rotor cleaning test enables assessment of the ability of surfactant or spacer in cleaning the surface of casing.

3. When non-ionic surfactant (commonly used in conventional spacers) is employed in the design, the hardening spacer could successfully clean the surface of viscometer sleeve in rotor cleaning test even in low concentrations. This shows the surfactant survives the high pH of the spacer.

4. The settable spacer without surfactant has better cleaning efficiency than water which highlights the role of particles in the spacer for cleaning purposes.

5. Contact angle of the water and steel coupon is reduced significantly when it is washed with the spacer containing surfactant.

6. Flow curve of the mixtures of spacer and geopolymer was measured and resulting R-index was calculated which indicates the compatibility of these two fluids in the flowing state.

7. API gel strength measurements showed high gel strength for 25/75 geopolymer/spacer mixture, while the thickening time of the mixture showed that the consistency curve stays stable and low after stop period.

8. Solid to water and GGBFS to granite ratio of hardening spacer seem to be an important factor in the degree of mechanical strength contamination of geopolymer by spacer. The higher the solid to water of the hardening spacer results in higher compressive and tensile strength of contaminated geopolymer. This is also in line with the heat evolution of geopolymer measured with iso-calorimeter where added water caused reduction in heat flowrate in dissolution and polymerization stages. Therefore, the intermixing of geopolymer with water-based drilling fluid should be avoided as much as possible.

Authors gratefully acknowledge TotalEnergies, Aker BP, ConocoPhillips, and Research Council of Norway for financially supporting the SafeRock KPN Project (RCN #319014-New Cementitious Material for Oil Well Cementing Applications—SafeRock) at the University of Stavanger, Norway.

There are no conflicts of interest. This article does not include research in which human participants were involved. Informed consent not applicable. This article does not include any research in which animal participants were involved.

The authors attest that all data for this study are included in the paper.

$k$ =

consistency index, Pa sⁿ

n =

flow behavior index, dimensionless

G′ =

storage modulus, Pa

G″ =

loss modulus, Pa

R² =

coefficient of determination

rpm =

revolutions per minute

SSE =

sum of squared of error, Pa²

$γ˙$ =

shear rate, 1/s

θ_m =

highest dial reading among different mixtures

θ_p =

highest dial reading among pure fluids

$μp$ =

Bingham plastic viscosity, Pa s

$τ$ =

shear stress, Pa

$τy$ =

yield stress, Pa