Assessment of resin systems as solution for well integrity challenge in carbonate reservoirs

Well integrity is defined as technical, operational, and organizational solutions specifically aimed at the reduction of the risk of formation fluids release throughout a well's life cycle. Cement integrity is one of the main aspects that comprises well integrity. Cement integrity must be ensured throughout a well's life cycle as well as after abandonment. If the cement was to lose its integrity, the consequences for the personnel, the equipment, and the environment could be severe. Cement failures might lead to leakages that can seep through the cement pathways; sealant materials are used to plug these pathways. The actual and real-time disputes are mainly dependent on the technical approaches currently available to limit the main causes of well integrity deficiency. These include cementing jobs, substantial design based on downhole conditions, material selection criticality, well construction performance, and technological resources. Additionally, in this specific context, well integrity also refers to the control of the flow inside the wellbore (between different horizons) and the flow from the well (especially in the annulus behind the casing). The worst-case scenario would be the loss of wellbore integrity, and it would be identified as the collapse of the well caused by failure of the construction material. The focal point of the research presented in this paper is the cement quality and its role in developing the most vigorous well casing possible. The focus was centered around resin additives such as microbond, latex, and crystal seal. Microbond is a white mineral cement-based liquid with exceptional performances when combined with cement. Latex is composed of rubber particles dispersed in water, usually found in white liquid form. Crystal seal is a yellow colloidal material highly regarded for sealing properties when mixed with concrete mixtures. These additives were individually incorporated into cement-based samples, in varying concentrations, for testing the changes in properties suitable for well integrity considerations. Highly specialized and sophisticated cement testing apparatuses, such as the ultrasonic cement analyzer (UCA), static gel Strength machine (SGSM), and curing chamber, have been used to perform qualitative and quantitative property gradient tests. Based on the performed experimental tests results, the optimal additive resin for a class G cement slurry is the microbond additive at a concentration of 7.5%. The results highlighted its ability to enhance the cement compressive strength by 52%. Additionally, the transit time showed a 26–42% range decrease in the travel period, indicating that the cement was stronger under increased pressure and temperature. This can clear up the permeable section of the well wherein the cement-resin slurry could efficiently close off wellbore intersecting faults and, therefore, prevent possible connections between the wellbore and water zones in the subsurface. As a result, microbond, latex, and crystal seal, respectively, show the effect of varying controlled conditions as functions of the increase in concentration, curing time, and temperature to its comparable properties such as breakload, compressive strength, and ultimate break force. This study can further support future analysis in improving the reliability of petroleum well constructions.


Introduction
In the petroleum industry, class G cement is usually used when it comes to cementing jobs. The primary cement of oil and gas wells is prone to fail under downhole conditions. Thus, remedial operations must be conducted to restore the wellbore integrity and provide zonal isolation (Alkhamis and Imqam 2021). Well integrity remains the main function of the well cement throughout the well life. Continuous developments in petroleum engineering have led to well life extension beyond conventional durations, and thus, cement properties have become very important (Teodoriu et al. 2013). If wellbore integrity cannot be guaranteed, production or drilling operations would be done under unsafe circumstances. In order to fix specific well leaks or cracks and to prevent any interaction between the wellbore fluids and the formation fluids, resin-based cement can be used (Amani et al. 2015). Because of its elasticity and high compressive strength resistance, the resin is able to fill in the fracture or space and seal it off, thus preventing any further issues. An example of this application can be found in the use of resin cement in the Utica shale formations. The issue faced was pressure buildup due to a fracture found in the casing. Following the resin cement application into the fracture, the pressure buildup issue completely disappeared (Jones et al. 2013).
A further issue faced was the requirement to fracture the formation in order to improve productivity. The resinbased cement increased the shear bond strength and bettered the mechanical properties of the casing, deemed to be instrumental in providing a reliable barrier to thwart any future consequences arising due to sustained casing pressure (Wagle et al. 2021). However, by fracking the formation, they would also damage the well and jeopardize its integrity. In order to preserve the wellbore integrity, sections in the well were repaired by injecting resin cement specifically in areas where the casing was not well cemented. The operator company was then able to fracture the formation with no further wellbore integrity damage (Jones et al. 2013. Resin technology has also been applied to permanently abandon a pilot hole and block the flow between horizontal production sections and vertical pilot holes (Pardeshi et al. 2016).
Resin cement has been used in the past numerous times. Applying this method to repair leakage problems under budget constraints is challenging due to its associated high costs. By applying resin cement to a fracture in a well, pressure build up problems can be eliminated. The way that this resin cement is delivered into the well is straightforward; by directly injecting it into the poorly cemented areas in the well or in correspondence of the largest fractures which will allow it to seep into the crack and eventually close off the unwanted leaks.
Similarly, resin cement can also be used for well abandonment. For any adequate well abandonment job, one of the most significant aspects is that no pressure should bleed off from the well. However, especially in the presence of gas, sometimes the abandoned well keeps leaking despite the presence of cement. In this instance is where resin cement comes in handy. Because of its unique elasticity and setting properties, the resin cement properties allow for a complete seal off the well, with no leaks. Examples of this can be seen in wells located in the Northern Colombian region, wherein the injection of cement with resin has been applied as a remedial cementing practice due to gas leaks. The well in these cases was wholly sealed off after implementing resin cement to stop gas leaks to the surface (Urdaneta et al. 2014). Besides resin cement being highly valued for its sealing properties, it also has a very high compressive and tensile strength resistance compared to the regular class G cement (Beharie et al. 2015).
Every well design has a specific barrier requirement, for risk mitigation, based on its entire life cycle of operation. At different stages, from low to high-pressure and ambient to high-temperature conditions, a well is expected to withstand significant causes of breakage, as shown in Fig. 1.
Due to resin cement's high elasticity and resistance to compressional forces, fractures that may occur throughout the cementing job can be easily sealed off, ultimately stopping any leak. When the challenge falls outside the capabilities of either conventional solution (Portland cement/ microfine cement), further methods are necessary to be implemented. Recent developments have shown that resins are an excellent alternative for these types of challenges (Perez et al. 2017). However, the objective of having the perfect balance between resin and class G cement is tough to achieve and costly. In order to ensure that this method is effective enough for field implementation, laboratory experiments must be conducted. The reason is due to the fact that on a larger scale, there is no window for error as it can have potential drastic financial consequences if not applied correctly. Lisigurski et al. 2009 discussed a carbonate formation that was extensively fractured and faulted because of domal uplift and regional tectonic events. Drilling long multilateral wells to intersect several fractures and faults is the ideal way to develop that reservoir. However, intersecting faults can then act as a pathway for both aquifer water and water injection to conduit to producer wells; therefore, the waterflood-sweep efficiency can be challenged. Figure 2 illustrates the location of the fault on a typical horizontal well. Figure 3 demonstrates fractures of a fault zone identified in a borehole image. Retnanto and Yamin (1999) and Retnanto et al. (2012) show that partially completed wells are easy for proper fluid placement. Figure 4 shows a top view with wells and fracture zone in a carbonate reservoir (Pedersen et al. 2009). Figures 2, 3, 4 depict common conditions in carbonate reservoirs that can reduce the effectiveness of waterfloods in such reservoirs.
This research aims to determine how resin cement and the mix of resins with class G cement can improve well integrity, as well as a combined solution for unwanted fluid shutoff specific for the Middle Eastern region, i.e., for carbonates 1 3 reservoir and under typical reservoir temperatures and pressures found in the Middle Eastern region. Through these results, it will be possible to assess the application that can be used for any remedial jobs independently of the issue and the severity of a seal. The main objectives of this research are to study the resin system and to assess its advantage on remedial jobs over the use of regular class G cement. In addition, this research utilizes the experimental results to develop the application of a resin system for both well integrity and unwanted fluid shut-off as well as to determine a relationship of a mix of class G cement with resin to cement behavior. All laboratory measurements such as viscosity, crush test, high-temperature high-pressure (HTHP) consistometer, and ultrasonic cement analyzer have been performed within the temperature and pressure ranges of Qatar fields to meet the objectives of this research. Evaluating the qualitative response to the overall test development starts with constructing the general scope. The methodology and experimental work were designed and exhibited primarily based on the desired apparatus, chemical reactivity, and researchers' involvement. Then, extensive training was given to all participants to gain full access to the operational conditions of the investigation. Next are the material preparations and formulation of samples and their variables. Following each test, the results were analyzed to determine any possible corrective action requirement.

Methodology and experimental work
The drilling laboratory at Petroleum Engineering Department, Texas A&M University at Qatar, has high standards for testing and measuring and can conduct the required experimental tests. Figure 5 shows the curing chamber equipment for well cement testing. Similarly, Fig. 6 depicts the ultrasonic cement analyzer, which is also equipped with a Static Gel Strength analyzer for thorough investigation of the effects of resin additives on the cement slurry. Three different types of resins were analyzed: microbond   Fig. 7), crystal seal (Fig. 8), and latex ( Fig. 9). For each resin, diverse additive samples with varying concentrations, ranging from 0 to 12.5% at 2.5% increments, were created (Fig. 10). The cement mixture strength varies as a function of the added resin additives concentrations.
The cement samples for actual breakload or crush tests were cured, in accordance with the typical downhole temperature and pressure conditions, in the curing chamber. Adhering to the API specification 10 and ASTM Standards C109 procedures and guidance, the cement sample could be developed following the minimum time requirements.
The ultrasonic cement analyzer or UCA determines the velocity gradient by measuring the acoustic signal (impedance) at 20,000 Hertz. It produces real-time, continuous, and non-damaging results for evaluating the sample properties such as compressive strength at elevated temperature and pressure at specific setting times.
The slurries were prepared at different resin concentrations and analyzed in the UCA. The equipment recorded the samples' compressive strength under Qatar's typical downhole pressure and temperature conditions. Breakload/Crush Test: 2-inch slurry mix cubes were prepared for different resin concentrations. Three cubes made from the same slurry were cured at varying reservoir temperatures, pressures and curing time intervals. The same samples were then crush tested (Fig. 11). Measurements of the cube's dimensions before and after completing the crush test were taken and analyzed.
All the results were seen simultaneously while the tests were being performed. Verification of any disruption could be straightforwardly detected. To address any safety concern, personal protective equipment (PPE) were used. Further, a poly-acrylic shield was also designed to isolate the cement debris inside the system during testing.   Figure 12 shows the flowchart process followed to perform the intended laboratory measurements.

Safety precautions
It is essential to understand that one of the most important priorities is developing a cost-effective solution. However, the most critical prerogative is to do so as safely as possible. Specific steps and responsibilities, aimed to ensure safety, were considered before proceeding with any experiment. Firstly, before using the high-pressure hightemperature (HPHT) equipment, curing chambers, and the ultrasonic cement analyzer (UCA), extensive training, provided by either the equipment vendors or the assigned technical laboratory coordinator, was required. Secondly, of crucial importance was an adequate training on the risks and effects of the chemicals used in the laboratory and experimentation. Safety data sheets (SDS) were the first line of information. Thirdly, the electrical components of the equipment were inspected before each usage to ensure safety compliance. Lastly, safety attire was enforced when performing any experiment, also known as personal protective equipment (PPE). Laboratory etiquettes required to wear were gloves, glasses, laboratory coats, and closedtoed shoes.

Experimental results
Individual resin additive samples at different concentrations were tested. The collected results are available in Tables 1, 2, 3. From the obtained results, analysis can be conducted on an individual resin additive in order to understand the impacts of the different concentrations on the cement properties. For each different sample, a number of parameters could have been used for analysis purposes. However, due to the scope of the research, the preferred parameter investigated was the compressive strength as it provides a simple yet efficient method to determine a specific additive's concentration to ensure the highest cement performance. The following Figs. 13, 14, 15 show the plotted results for each resin.
From the above-displayed figures, it is possible to adequately analyze the impact of different resin concentrations on the compressive strength of the mixed cement. From Fig. 13, it is possible to determine that the ideal concentration for increased cement compressive strength is around 7.5%. A decrease in compressive strength at higher or lower resin concentrations was recorded. From Fig. 14, latex additive optimum concentration is at 7.5%. However, compared with the microbond, the latex test results at different concentrations yield much closer results. On the other hand, higher resin concentration (at 12.5%) result in a cement with less compressive strength than the base case. From Fig. 15, the experiment conducted on the crystal seal mixtures highlighted a negative impact on the resulting compressive strength of the samples tested. Based on the results collected, it was inferred that crystal seal suffered interferences with the natural crystallization of cement when in contact with water. The progression of the mixture curing was incomplete. Therefore, it failed to produce a hardened material. Following the conducted experimental tests, results for each resin additive were collected and analyzed comparatively.
Tables 1, 2, and 3 display the collected results at different resins concentrations for ultimate compressive strengths, break-load force, and dimensions changes under crush tests. Of specific importance is the effects of increasing concentration of additives on the compressive strength represented by the breakload (lbf). The following Figs. 16, 17, 18 show the plotted data for ultimate break force (lbf), compressive strength (psi), and breakload (lbf), respectively. Figure 19 illustrates the comparative indication, as a direct proportion, on the effect of increasing concentration of additives in relationship to the distance (inch).

Limitations of the study
The overall scope of this research focused on the effect of the three additives, namely microbond, latex, and crystal seal, available for use in the industry. The formulation of each variable depends on the result of properties yielded. The outcomes are obtained from equipment used for 4-24 h. The concentration was determined concerning the overall contribution to enhancing the properties specified in the study's objectives. In addition, the level of parameters such as elevated temperature and pressure involved in this research is controlled for safety considerations. The series of data collection are also restricted to the output of the apparatus capacity. Similarly, the generated results can be used for future research proposals in order to enhance this experiment's conclusions and recommendations.

Summary and conclusions
This research aimed to investigate an innovative approach to the common practice of using Class G cement as a means to repair or close off of wells as well as long-term well integrity in the Oil and Gas Industry. With a specific focus on the Middle Eastern region's downhole conditions, the addition 1 3 of resin additives to the industry standard Class G cement proved ingenious, as highlighted by the experimental results.
From the individual additive analysis, conclusions on the most and least effective resin can be drawn. The research showed a distinct decrease in the cement hardness property when crystal seal was added to the slurry. Following the same analysis, it was noted that the best additive, in terms of properties enhancement, was microbond.
The microbond additive at 7.5% increased the cement compressive strength by 52%. Adding to it, a more general deformation resistance behavior and a much higher breakload force allow for the microbond resin to be a valuable additive for well integrity issues. Cement with microbond additive can withstand higher stresses preventing cracking and consequent leaks. Based on the results collected, crystal seal interferes with the natural crystallization of cement when it comes into contact with water. The progression of the mixture was incomplete. Therefore, it failed to produce a hardened material.
To produce the strongest cement, further investigations should be performed on the exact optimal resin concentration. As per the conducted experiment, increases in concentrations of 2.5% were implemented. Analysis of smaller concentration increases is suggested. More investigations on possible uses of less performing additives such as latex and crystal seal should be performed. The more elastic properties of the cement with such additives could prove helpful in other aspects of well integrity.
In conclusion, the findings of this research can be summarized as follows: • Microbond additive (at a 7.5% concentration) increased cement compressive strength by 52%, resulting in the cement having the highest general deformation resist- ance, with respect to the base samples as well as other concentrations and resins investigated, and greater breakload resistance. • The crystal seal mixture failed to produce a hardened material. It was inferred that, at any concentration, the resin tampered with the natural crystallization of cement when in contact with water. • Further and more detailed analysis on the applications of additive-rich types of cement should be performed with specific focus on the area of fault filling. Cement slurry with resin additives can efficiently fill wellbore intersecting faults and prevent possible connections between the wellbore and water zones in the subsurface.

Declarations
Ethical approval The submitted work is original and has not been published elsewhere in any form or language.

Conflict of interest
On behalf of all the co-authors, the corresponding author states no conflict of interest.
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/.