Durability of lightweight oil-well geopolymer system in sulfate environment

Sulfates when present in the formation water would attack and deteriorate the cementitious system. In the quest to investigate the possibility of using geopolymer systems in oil-well cementing, the durability of geopolymer in various corrosive environments has been simulated. Lightweight geopolymer systems exhibit different microstructural and macroscopic properties compared to the conventional geopolymer systems whose durability under sulfate attack has been widely investigated. It is therefore important to study the resistance of lightweight geopolymer to sulfate attack. A ternary geopolymer was formulated at 13 ppg (1.56 g/cm3) by admixing metakaolin, ground granulated blast furnace slag (GGBFS), and silica fume in an alkaline solution composed of sodium silicate and 10 M sodium hydroxide solution in a mass ratio 1:3. The geopolymer specimen was cured in a water bath at 163 °F for 72 h and subsequently submerged in a 50 g/L sodium sulfate solution for up to 2 days. The effect of the sulfate solution on the strength and the mechanism of the sulfate attack was analyzed using analytical techniques, pH, and ion exchange measurements. The compressive strength of the specimen at 72 h, having a value of 802 psi decreased by 19.8% and 26.2% after day 1 and day 2 in the sodium sulfate solution, respectively. Investigation of the mechanism indicated that the loss in strength was not a result of the formation of deleterious phases but rather the leaching of Na ions from the geopolymer indicated by the rise in the pH and amount of Na ions in the sodium sulfate solution after the geopolymer was submerged in a sulfate solution. Lightweight geopolymer has a relatively loose microstructure that reduces its tendency to inhibit the transport of alkalis during sulfate attack, making the effect of the sulfate environment more pronounced.


Introduction
When standard and high-density cement systems are used in shallow and weak intervals where the underlying formation has a lower fracture pressure gradient, the formation would break down due to a higher equivalent circulating density (ECD). Low-density systems are preferred in such scenarios as they present a lower hydrostatic column (Coker et al. 1992;Nelson and Guillot 2006). Ordinary Portland cement (OPC) has been the most used hydraulic cement for wellbore isolation. However, the use of the OPC has been marred by the huge amount of carbon dioxide (CO 2 ) released into the atmosphere during its production (Leung et al. 2014). Additionally, OPC-based systems used in petroleum wellbores face several technical issues. For example, these systems suffer strength retrogression when used at temperatures above 230 °F and also fracture easily due to their high brittleness during hydraulic fracturing (Khalifeh et al. 2017;Yan et al. 2020). Additionally, OPC-based systems are susceptible to acid attack (Bakharev et al. 2003;Nasvi et al. 2013). These factors and more have necessitated the development of low-carbon cementitious systems such as geopolymer (Wasim et al. 2021b). Geopolymer is an alternative cementitious binder formed when materials containing non-crystalline silica and alumina such as fly ash, slag, silica fume, and calcined kaolin are admixed in an alkaline solution (Davidovits 1991;Zhuang et al. 2016). Investigations on the behavior of different geopolymer systems under wellbore conditions have been conducted by several researchers. In a study performed by Adjei et al. (2022a) to investigate the effect of elevated temperatures on the microstructure of geopolymer, 70% metakaolin and 30% ground granulated blast furnace slag (GGBFS) were admixed in an alkaline solution having a modulus silicate (SiO 2 /Na 2 O) of 1.1. The modulus silicate is the parameter that controls the workability and strength of the geopolymer. The authors reported that the strength retrogression of geopolymer systems at elevated temperatures is due to the formation of crystalline phases which deteriorates the microstructure when they undergo thermal expansion. Giasuddin et al. (2013a) prepared and investigated the behavior of geopolymer in normal and saline water, and hence, its feasibility as a sealant in CO 2 sequestration wells by admixing using fly ash and slag in the ratio of 9:1, respectively, in an alkaline solution composed of sodium silicate and 8 M sodium hydroxide in the ratio of 2.5:1, respectively. In a similar study performed by the previous authors (Giasuddin et al. 2013b), the same alkaline system was prepared; however, the fly ash and slag were used in a ratio of 10:1. A ternary geopolymer system was developed by Khalifeh et al. (2017) through the dissolution of the silica and alumina from three aluminosilicate sources, namely silica fume, aplite, and slag, using a combination of 8 M NaOH and sodium silicate solution. Geopolymer has been touted as the future oil-well cement due to its comparatively cleaner production process, superior physical and mechanical properties, and higher chemical resistance (Sugumaran 2015;Zhuang et al. 2016;Kanesan et al. 2018;Adjei et al. 2021).
Various salts and acids create a corrosive environment which leads to the deterioration of the cement sheath. For instance, in carbon dioxide (CO 2 ) storage in geologic formations such as depleted oil wells, the CO 2 dissolves in the presence of moisture creating an acidic environment through the formation of carbonic acid (Jani and Imqam 2021), which corrodes tubular and subsequently converts the calcium compounds (tricalcium silicate and calcium hydroxide) in the OPC system into carbonate precipitates, adversely altering the physical and mechanical properties of cement sheath and the integrity of the well (Rimmelé et al. 2008;Brandvoll et al. 2009;Faqir et al. 2017). The durability of geopolymer in a CO 2 environment has been explored by several authors including (Barlet-Gouedard et al. 2010;Nasvi et al. 2013;Jani and Imqam 2021). In an acidic environment, Ridha and Yerikania (2015) reported that geopolymer systems developed with 70% fly ash and 27-30% silica fume exhibited higher resistance in comparison with the OPC system. Other authors have also shown that the strength of geopolymer systems is enhanced in saline environments (Lee and Van Deventer 2002;Giasuddin et al. 2013b).
Sulfates present a corrosive environment that leads to the deterioration of cementitious binders (Lauer 1990;Sancak and Özkan 2015). The durability of geopolymer under sulfate attack has been studied by some authors in the construction industry. However, the results indicated that the durability of various geopolymer systems varied in different sulfate environments. While some geopolymer systems exhibited a continuous decline in strength (Rajamane et al. 2012;Baščarević et al. 2014;Karakoç et al. 2016), other systems showed a continuous improvement in strength (Škvára et al. 2005;Bhutta et al. 2014) and another group displayed fluctuations in strength (Bakharev 2005;Baščarević et al. 2014). This is because the durability of geopolymer systems in a sulfate environment is controlled by several factors including sulfate type, sulfate concentration, type of aluminosilicate precursor, microstructural and macrostructural properties of the geopolymer, and exposure time.
Lightweight cementitious systems exhibit different microstructural and macroscopic properties compared to standard cement systems. The variations in properties would affect the durability of the cementitious systems under various corrosive environments, and hence, there is a need to investigate newly formulated systems (Wasim et al. 2020(Wasim et al. , 2021a. In this study, the durability of a lightweight geopolymer system under sulfate attack is investigated. A ternary geopolymer system is developed combining metakaolin, silica fume, and GGBFS, and the effect of 50 g/L sodium sulfate solution is studied. One major difference between this current study and other related studies is the use of the scratch test in determining the compressive strength of the specimen under sulfate attack. This device has been extensively used in the strength analysis of cementitious systems (Ulm and James 2011;Murtaza et al. 2019;Adjei et al. 2022a). As a nondestructive test, the same sample can be used multiple times by taking grooves at different sections, allowing for accurate monitoring of the changes in strength over time of the same sample.

Materials
The main mineral admixtures used in the production of the ternary geopolymer are metakaolin, GGBFS, and silica fume. The silica fume improves the microstructure and lowers free liquid and fluid loss, thereby enhancing the stability of the systems (Mueller and Dillenbeck 1991;Nelson and Guillot 2006;Paiva et al. 2018). The GGBFS improves the strength and the geopolymerization (Khalifeh et al. 2017;Elyamany et al. 2018). The metakaolin is an aluminosilicate material produced when kaolinitic clays are heated at 932° F to about 1562° F (Siddique and Klaus 2009;Adjei et al. 2022b). The GGBFS is a by-product from the blast furnace during the production of iron while the silica fume is a waste residue from the silicon and ferrosilicon industry (Siddique and Khan 2011). The results of the analysis of the chemical constituent of the mineral admixtures estimated using X-ray fluorescence equipment (Bruker) are given in Table 1. The mineral admixtures contain a considerable amount of SiO 2 and Al 2 O 3 needed for the geopolymerization. The metakaolin is dominated by SiO 2 and Al 2 O 3 . The GGBFS shows a high proportion of CaO, SiO 2 , and Al 2 O 3 , while the silica fume is mainly SiO 2 .
The mineralogical composition of these materials was measured with the Bruker XRD equipment using the following settings: Cu radiation (λ = 1.54184 Å), scanning range: 5°-80°, 2-theta, operating at 30 kV and 10 mA, Fig. 1. The silica fume and GGBFS are highly amorphous, containing small crystals of tridymite, bikitaite, and coesite, respectively. The tridymite and coesite are polymorphs of crystalline SiO 2 (Hunt et al. 2019;Payre et al. 2021). Quartz, another silica polymorph, and halloysite a part of the kaolin group are both present in the metakaolin (Keeling and Pasbakhsh 2015).
The sodium hydroxide of > 98% purity was purchased from Sigma-Aldrich. The Na 2 SiO 3 (SiO 2 /Na 2 O = 3.375, specific gravity = 1.390) was obtained from Loba Chemie, India. Distilled water was used in all the formulations. The sodium sulfate was supplied by AppliChem GmbH.

Design and optimization of geopolymer binder
The density of the system investigated is 13 ppg (1.56 g/ cm 3 ). Even though the objective is to study the durability of the geopolymer under sulfate attack, cementitious systems used in oil and gas wellbores should exhibit some properties like negligible free water, low fluid loss, and desirable rheology. Two ternary geopolymer systems G1 and G2 were designed with 4% and 8% fluid loss agent (FLA), respectively. The mixing, conditioning, rheological analysis, free water, and fluid loss tests followed the procedure given in (API RP 10B-2 2013). The rheology and free water (supernatant per 250 ml) were measured at ambient pressure and 80 °F. The American Petroleum Institute (API) filter press was used to estimate the fluid loss rate of the samples at the ambient temperature and 100 psi. The composition of the geopolymer systems, G1 and G2, is given in Table 2. The geopolymer systems were developed using sodium silicate and sodium hydroxide in the ratio of 1:3. Approximately 10 M sodium hydroxide solution was prepared 24 h before the experiments and the sodium silicate was admixed about 30 min before the addition of the mineral and chemical admixtures. Distilled water was used in all formulations. The components were added by weight of binder (BWOB), where the binder is the total amount of metakaolin, GGBFS, and silica fume.
The geopolymer specimens were cured at a static temperature of 163 °F and atmospheric pressure in a water bath. Compressive strength was measured using the scratch test method (Wombat device), Fig. 2. A diamond cutter is used to make a groove along the length of the cement core. The compressive strength along the specimen is obtained from the force acting on the cutter. The cross-sectional area of the groove and cutting velocity remains constant along the cut. At the end of the test, the log of the magnitude and inclination of the forces on the cutter is obtained. The magnitude of the force is dependent on factors such as rock mechanical properties, the groove geometry, and the geometrical properties of the cutter (Murtaza et al. 2019). The forces are correlated with unconfined compressive strength using the formula below: where Ft is the horizontal component of the cutting force, Et is the intrinsic specific energy, and A is the cross-sectional area. Microscopic analysis was performed using a scanning electron microscope (SEM) equipment equipped with energy-dispersive X-ray spectroscopy (EDS) (JEOL JCM 7000). The Fourier transform infrared spectroscopy (FTIR-BRUKER) was also used for microstructural analysis. Leaching analysis was done through pH (pH meter by (1) Ft = Et ⋅ A Metler Toledo) and ion exchange (ion chromatography by Metrohm) measurements.

Design and optimization of geopolymer
Both geopolymer systems have zero free water as indicated in Table 3. Additionally, Table 3 shows that the fluid loss rate in G2 is about 46.8% lower than G1. However, increasing the fluid loss agent (FLA) increases the shear stress and apparent viscosity as indicated in Figs. 3 and 4, respectively. The G1 was thus selected as the final geopolymer recipe due to its better rheology and acceptable fluid loss rate. The fluid loss rate should be lower than 50 mL/ 30 min for gas wells and 200 mL/30 min for oil wells (Nelson and Guillot 2006).

Scratch strength
The scratch test provides an indirect means of determining the compressive strength of hardened specimens by taking a groove along its length. As indicated, the use of this nondestructive approach in monitoring the effect of the sulfate solution on the strength of the geopolymer permits the use of the same specimens, which allows for the observation of even minute changes in strength. Figure 4 shows

Morphology and composition of microstructure
Images of the microstructure (left) taken in the secondary mode and elemental composition (right) of the dried and gold-coated samples after and before saturation in the sodium sulfate solution are shown in Fig. 5. The specimens have similar microscopic morphology. The calcium aluminate silicate gel (whitish and fibrous), produced from the geopolymerization of the GGBFS, is visible in the microstructure. This is confirmed in the energy-dispersive X-ray spectrum (right), where aluminum (Al) is present alongside calcium (Ca) and silicon (Si). Carbon (C), oxygen (O), sodium (Na), and potassium (K) are the principal elements present in all specimens. There is no sulfur present in the EDS analysis. This differs from the work of Baščarević et al. (2014) who observed the precipitation of sulfur but agrees with that of Sata et al. (2012). In general, the elemental analysis indicates that no new phases are formed after submerging the specimen in the sodium sulfate solution, suggesting that the mechanism of the sulfate attack which results in strength reduction is not due to the formation of deleterious phases.

Infrared (IR) analysis
The IR spectrum of the powdered samples is shown in Fig. 6. The FTIR can be used to study the degree of

Investigation of the mechanism of sulfate attack
The mechanism of sulfate attack on conventional systems is due to the formation of deleterious agents like ettringite and gypsum is well documented (Tian and Han 2017;Elyamany et al. 2018). Leaching analysis conducted through pH and ion exchange measurements can provide information on the mechanism that results in the loss of strength over time of the ternary geopolymer placed in the sodium sulfate solution.

pH Analysis
Results of the pH analysis are presented in Fig. 7. The pH of the system for the first 120 min is recorded every 30 min to capture the rate at which the pH changes within the early period. The pH at 1440 min and 2880 min is subsequently measured. The initial pH of the sodium sulfate solution is about 6.5. The pH rises rapidly to about 12.9 in the first 30 min. There is no significant rise in pH afterward. From 60 to 2880 min, the pH increases by only 3.8%. The rise in pH has been ascribed to the leaching of alkalis which weaken the microstructure. The rate of leaching is affected by the microstructure of the geopolymer. In a study conducted by Bakharev (2005), microscopic analysis showed that the specimen that experienced the highest leaching had microcracks. A compact and tight microstructure would inhibit the invasion of the sulfate solution and the subsequent transport of the alkalis from the system (Baščarević et al. 2014). However, the developed ternary geopolymer is a lightweight system with a loose microstructure compared to standard density geopolymer systems. This is observed in the secondary electron images in Fig. 5.

Conclusion
The resistance of lightweight geopolymer under sulfate attack has been investigated using strength and microstructural analysis, and pH and ion exchange measurements. Different geopolymer formulations would be characterized by distinct microstructural and macroscopic properties and hence would exhibit varying degrees of resistance under sulfate attack. It is therefore important to study how lightweight geopolymer would perform in a sulfate environment before using such a system for zonal isolation. Below are key observations from the study.
• Lightweight geopolymers are not durable in sulfate environments especially in oil-well cementing. • The loss in strength of the geopolymer is not due to the precipitation or formation of deleterious phases. • The main cause of the strength reduction is the leaching of the Na ions confirmed by pH and ion chromatography measurements.
Funding There is no external fund for this research.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.