A Review of Recent Research on Contamination of Oil Well Cement with Oil ‐ based Drilling Fluid and the Need of New and Accurate Correlations

: Drilling fluids and oil well cement are important well barriers. Their compatibility affects the long ‐ term integrity of the well. The mixing of drilling fluid with the oil well cement causes contamination of oil well cement. If the contamination is due to diesel/oil ‐ based drilling fluid (OBF) it adversely affects the rheological and mechanical properties of oil well cement—in other words, the long ‐ term integrity of the well. An initial study on OBF contamination of oil well cement was carried out two decades ago. In recent years, several research projects were carried out on the same topic to understand the reason for changes in the properties of oil well cement with OBF contamination. This literature review shows that using OBF eliminates several drilling problems, as the long ‐ term integrity of the well depends on the amount of OBF contamination in the cement slurry. This paper compares the experiments performed, results and conclusions drawn from selected research studies on OBF contamination of oil well cement. Their shortcomings and a way forward are discussed in detail. A critical review of these research studies highlights the need for new and accurate correlations for OBF ‐ contaminated oil well cement to predict the long ‐ term integrity of wells. have a catastrophic impact on well integrity. This paper shows that there is a large inconsistency in the way the data is reported and, in particular, the sample preparation. Reference values are hard to find among the studied references, which makes it difficult to accurately compare the results of various authors. Moreover, the curing time for which the mechanical properties have been reported varies largely from paper to paper, which also makes comparative studies more difficult. It can be concluded that laboratory testing at expected bottom hole conditions is necessary for approximation of the reduction in mechanical properties of OBF ‐ contaminated cement slurries in order to understand OBF’s effect on well cement integrity.


Introduction
The use of a specific type of drilling fluid for a well depends on various factors such as the geological formation to be drilled, the temperature, pressure, depth and formation evaluation procedure to be used, the environmental and ecological impact, costs, etc. Similarly, the type of oil well cement used depends on the depth range, rheological properties required, wellbore conditions, costs and so on. Oil-based drilling fluid (OBF) consists of oil/diesel in the continuous phase with a percentage of water in the dispersed phase. Additives are added to achieve the desired drilling fluid properties. The base of the OBF is usually diesel or mineral oil, with the former being more toxic than mineral oil systems. The toxicity of OBF is reduced by lowering the aromatics in diesel/mineral oil. Emulsifiers help in maintaining a stable water-in-oil emulsion under downhole conditions. Using OBF instead of water-based drilling mud (WBM) has several pros and cons associated with it [1][2][3].
Case histories [4][5][6][7][8][9][10][11] justify the use of OBF instead of WBM and helps to eliminate several drilling problems. The productivity index of long, horizontal open-hole gravel packed wells in West Africa improved three times when drilled with OBF compared to those drilled with WBM [4]. A multilateral well was drilled in the Aasgard field (a high-temperature reservoir in Norway) using low-solid OBF which saved 37 days of budget time [5]. Special OBF was designed and used for drilling exploration and appraisal wells for a major operator in the North Sea, where the expected reservoir pressure and temperature were 1700 psi and 400˚F respectively. The designed mud system provided the required thermal stability, consistency in properties and compatibility with the wireline programs [6]. Laboratory experiments on shale oil core samples from the Eagle Ford field were carried out to understand the effect of OBF and WBM on shale oil properties and the swelling properties of the formation [7]. Laboratory and field results (Gudrun Field) shows OBF can be used as a cost-effective and less-damaging perforation fluid for fields with High Pressure and High Temperature (HPHT) conditions [8]. Severe drilling problems, like lost circulation into weak zones and wellbore stability issues, can be eliminated by using OBF and Managed Pressure Drilling techniques [9]. The case history of Southeast Kuwait fields shows a successful application of OBF with a 60:40 oil-water ratio reduces the environmental impact compared to previously drilled wells with 80:20 oil-water ratio OBF [10]. An economic analysis of large fields (approx. 500 wells) using the holistic approach [11] proves that using OBF is better than using WBM.
The success of any drilling project depends on the compatibility of drilling fluid with the spacer and oil well cement [3,[12][13][14][15][16][17]. Due to the oil-wetting characteristics of OBF, displacing OBF becomes a critical operation before cementing. The spacer must be uniquely designed to displace the drilling fluid from the annulus and leave it water-wet [14]. It is highly recommended to test the compatibility of the drilling fluid with the spacer and oil well cement before field application. It helps to overcome the challenges and prevent remedial cementing operations [15][16][17]. It is difficult to displace 100% mud from the annulus using the spacer. The drilling fluid left behind mixes with the cement and contaminates it.
The effects of oil-based mud have been investigated in the past, especially to understand how the composition and chemistry of OBF affect cement performance [18]. One of the most common pieces of equipment used for cement mechanical properties is the Ultrasonic Cement Analyzer (UCA), which is a great instrument for describing the cement strength evolution [19][20][21][22][23][24]. For example, in 1993, Harder et al. [18] carried out laboratory experiments on 17 ppg density Class H Portland cement consisting of fluid loss additives and friction reducers designed for 200˚F. The cement slurry was contaminated (10%, 20%, 30%) with four different types of OBF, which were prepared in the lab with combinations of two base oils (diesel and mineral oil) and two emulsifiers (standard fatty acid and alkanolamide) as shown in Table 1.  Figure 1 shows the one-day compressive strength results obtained by performing a nondestructive test on the contaminated cement samples measured using the Ultrasonic Cement Analyzer (UCA).   Based on the two base oils used in the study, diesel oil had a more adverse effect on the compressive strength compared to mineral oil. On comparing the two-primary emulsifiers, the presence of alkanolamide showed better strength development compared to standard fatty acid (calcium soap) [18].
It is evident that the contamination of oil well cement with OBF causes well integrity issues and there is a need to better understand the effect of this contamination. Research studies [25][26][27][28][29][30] have been carried out in recent years to evaluate and quantify the effect of OBF contamination of oil well cement on the long-term integrity of the well. A summary of these recent studies is presented in the following sections, followed by critical analyses and discussions of the same. This paper aims to highlight the results of these studies and to point towards new and better investigations into cement contamination with oil-based mud.
The paper will describe in the Materials and Methods chapter all of the found case studies (in historical order) about cement contamination, which will later be used for discussions and data dissemination.

Contamination of Oil Well Cement with OBF
In recent years, research was carried out to discover the mechanism behind changes in mechanical and rheological properties of OBF-contaminated oil well cement. Modern research methodologies and equipment have allowed scientists to look in great detail at the OBF and cement interaction. A decade ago, research was mainly focused on optimizing the spacer fluid program to reduce the OBF contamination and/or look for additives to improve the compatibility between OBF and oil well cement. This section summarizes selected experimental studies performed in recent years to understand the phenomenon of contamination of oil well cement with OBF. A summary of the case studies considered in this paper will later presented in form of a table.

Case Study 1
The objective of study carried out by Aughenbaugh et al. [25] was to quantify the effects of contamination of various cement slurries with synthetic-based mud (SBM) and look for additives to reduce the effect of contamination. This research was/is divided into multiple phases and these were the objectives of the first phase. API RP 10A standard recommendations [31] were followed for preparing and mixing cement slurries in this study. The composition of cement slurries tested in this study are shown in Table 2.

Slurry Name
Composition H-1 API Class H-1 and tap water H-2 API Class H-2 and tap water C-1 API Class C and tap water L-1 Lightweight cement and tap water S-1 Blast furnace slag and alkaline activating solution DW-H-2 API Class H-2 and Tap water and Additives The above cement slurries were contaminated (5%, 10% and 15% by volume) by replacing part of the cement slurries with field SBM (11.6 ppg; 70/30 invert emulsion-oil/CaCl2), laboratory-formulated SBM (Lab-SBM) and silica sand. Slurries were contaminated with silica sand to test the effect of a reduction in cement contents. A drill press and a paint stirrer were used to mix the contaminants and the cement slurries. Samples were cured for 48 h and destructive as well as non-destructive tests (UCA) were performed to obtain the compressive strength values (curing at 170 °F and 3000 psi). Figure 3 shows the percentage reduction in compressive strength of field SBM-contaminated cement slurries with respect to neat cement slurries and the 48 h compressive strength of silica sand contaminated cement slurries (0% to 15%). The results obtained from silica sand contamination tests proved that the decrease in compressive strength due to contamination with field SBM is because of chemical interaction and not due to the dilution of cement content. It was noted that the time required for strength development was constant, irrespective of the percentage of contamination.
Lab-SBMs were prepared (in the laboratory) in two different ways to detect which component was responsible for the decrease in compressive strength of contaminated cement slurries: Lab-SBM with the same composition as the field SBM and Lab-SBM (no brine) where brine was replaced with an equal volume of freshwater. Lab-SBM and field SBM showed similar results of compressive strength, which were less than the Lab-SBM (no brine) compressive strength values. This test proved that brine affects the compressive strength negatively and the reason for lower compressive strength values for SBMcontaminated cement slurries could be due to the osmosis of the water from cement slurries to SBM.
Compressive strength values obtained for the slag-based cement slurry contaminated with SBM were least affected compared to other cement slurries tested in this study, see figure 4. Several additives were added to SBM-contaminated cement slurries to compensate for the reduction in strength. The only additive that improved the strength was alkali when added at 10% of the weight of SBM. These were the findings from Aughenbaugh et al., 2014.

Case Study 2
Vipulanandan et al.'s [26] study tried to find out the correlation between the piezoelectric properties, rheological properties and mechanical properties of modified API Class H cement. The sensing properties of cement slurries were improved by adding conductive fillers (0.1% by the weight of cement). The modified cement slurries were contaminated (0.1%, 1% and 3% by the weight of cement) with vegetable oil-based drilling fluid (75/25 invert emulsion). Cylindrical samples (2" diameter and 4" height) with two conductive wires 5 cm apart were cured for 28 days at room temperature. The densities of the modified cement slurries were measured using a standard mud balance cup; rheological properties were tested using the rotational viscometer (ambient pressure and temperature for a rotational range from 3 to 600 rpm); a standard API Resistive meter was used to measure the electrical resistivity and destructive test for measurements of compressive strength, which were performed using a hydraulic compression machine for 1, 7 and 28 days cured samples.
The author proposed a hyperbole model to predict the shear strain rate vs. shear stress. This proposed model was fitted with laboratory results and produced better results compared to Herschel-Bulkley model and Bingham plastic model. It was also observed ( Figure 5) that the initial electrical resistivity of the modified cement slurries increased with the increase in contamination. The waiting on cement time could be calculated by monitoring the changes in the electrical resistivity of the cement slurries.  The correlation between the electrical resistivity and compressive strength of samples tested in this study for different curing ages was found to be linear in nature.

Case Study 3
An extensive study was carried out by Li et al. [28] at the microscopic level to understand the mechanism of OBF contamination of oil well cement. The hydration process of contaminated cement slurries was studied using X-ray diffraction (XRD), Scanning Electron Microscope (SEM), Environmental Scanning Electron Microscope (ESEM), Thermogravimetry (TG) and Energy Dispersive Spectrometer (EDS). The changes in rheological properties and mechanical properties of contaminated cement slurries were quantified first and then the mechanisms behind them were studied.
The cement slurries used in this study were mixed based on API recommendations, which consist of API Class G cement, free water control additives, water, dispersant, etc. These slurries were contaminated (0%, 5%, 25% and 50% by weight or volume of cement) with UDM-2 system diesel-based drilling fluid (85/15 invert emulsion). Compressive strength was measured by performing destructive tests on contaminated cement slurries cured in a water bath at 93 °C for 1, 3 and 7 days. Microstructure analyses of 5% and 25% contaminated cement slurries were discussed in the paper.
The results obtained for the compressive strength and bonding strength for different contaminations (Figure 7) showed a 100% reduction in strength when the contamination was 50%. XRD test results confirmed the reason for the decrease in strength. OBF hinders the hydration reaction without interacting chemically. Incomplete hydration of the contaminated samples leads to the formation of a honeycomb structure. Figure 8 shows the general process of hydration of contaminated cement slurries. Demulsification and osmotic pressure changed the rheological properties of contaminated cement slurries. Figure 9 shows the process of water migration in OBF-contaminated cement slurries. This study also proved that the addition of surfactants to contaminated cement slurries improves the mechanical and rheological properties.

Case Study 4
An extensive study was performed by Li et al. [29] to find out the effect of OBF and its components on the rheological properties, mechanical properties, porosity and permeability of cement slurries. An Ultrasonic Cement Analyzer (UCA), X-ray diffraction (XRD), Thermogravimetry (TG), a Scanning Electronic Microscope (SEM) and Fourier Transform Infrared Spectroscopy (FTIR) were used in this study.
The cement slurry used in this study consisted of API Class G cement, 2% anti-gas migration agent, 25% silica powder, 5% filtrate reducer, 1% dispersant, 2% retarder and 0.2% defoaming agent. The above cement slurry was contaminated (0%, 5%, 25% and 50% by weight or volume of cement) with diesel-based drilling fluid. The cement samples were cured for 2 days at 135 °C and 20.7 MPa. With the increase in contamination, the compressive strength and bonding strength decreased, while porosity and permeability increased (Table 3). An increase in porosity and permeability was confirmed by SEM tests (Figure 10).  Furthermore, the effects of contamination of cement slurry with different components of OBF was also studied. The compressive strength was reduced to zero when cement slurries were contaminated with 50% emulsion and 50% diesel, respectively. The reduction in compressive strength values for different contaminations of primary emulsifier, secondary emulsifier, and organic clay was much less compared to the effect seen with diesel and emulsion contaminations.
Li et al. [29] also contradicted their previous work [28] and concluded that OBF does not hinder the hydration process of contaminated cement slurries. An increase in contamination of OBF causes an increase in lubrication and porosity of the contaminated cement slurries, thereby decreasing the strength of the hydrated samples.

Case Study 5
Soares et al. [27] conducted a study on contaminated cement samples to determine the rheological properties, mechanical properties and slurry sedimentation testing, and evaluated the hydrated samples using XRD and SEM. The reference cement slurry (RS) consists of API Class G cement, water, antifoam, dispersant, fluid loss control and retarder which weighed 15 ppg. Two different OBFs (10 ppg, 63/37 invert emulsion) were formulated-one with a wetting agent (DF) and another without a wetting agent (DF *). The sample names and their corresponding contaminations are presented in Table  4. The samples were cured for 24 h at 52 °C. They were demolded 45 min earlier followed by 30 min cooling under flowing water and destructive tests were carried out to determine the compressive strength values.

Sample Name RS/DF (%) Sample Name RS/DF * (%)
S95 Samples with 50% contamination were still in the slurry phase even after curing time of 24h, which is consistent with the values published in the previous literature. A reduction in compressive strength was more pronounced in the presence of the wetting agent compared to without the wetting agent, see Figure 11. The yield point and plastic viscosity increased with the increase in contamination. Microcavities in the hydrated samples increased with an increase in OBF contamination, causing the compressive strength to decrease.

Case Study 6
Performing the non-destructive tests to measure the compressive strength of cement has gained popularity in recent years. To simulate the poor-quality wellbore cleaning, cement slurries were contaminated with OBF and ultrasonic pulse velocity was measured. Olteanu et al.'s [30] study aimed to check the trustworthiness of ultrasonic measurements in the presence of OBF. API Class C cement was contaminated with 40 mL OBF and cured at room temperature (20 °C). The results obtained for over 200 tests are shown in Figure 12. The contaminated cement slurries behave better than uncontaminated cement slurries during the initial hours of curing, as shown in Figure 13. This can mislead the engineer and consider the poorquality cement job as a success. The authors also presented the correlations of Unconfined Compressive Satrength (UCS) vs. Ultrasonic Pulse Velocity (UPV) for contaminated, uncontaminated and uncontaminated-thermal cycles, see Table 5. The thermal cycle tests were carried out in a pre-heated water bath at 60 °C for 8 h/day.  Table 6 summarizes important details from the above case studies.  * this information is not clearly specified by the cited work, but is assumed.

Discussion
Our review study selected the abovementioned case studies since their objective was to understand the effect on the mechanical and rheological properties of cement slurries contaminated with OBF and/or to understand the mechanism behind the reduction in mechanical and rheological properties of cement slurries contaminated with OBF. Upon comparing these studies, differences in the sample preparation methods and testing procedure of the samples are evident. The type of API cement, additives, type and amount of contamination, curing time, curing temperature and pressure differs from one group to another.
Inadequate information on sample preparations is evident in the literature-many research groups have not mentioned whether the OBF contamination is by weight of cement or by volume of cement. Moreover, many groups have not mentioned if the OBF is added to standard cement slurry compositions or if OBF is replaced by equal volumes of cement slurry. Few groups follow the API 10D recommendations to prepare the 2″ × 2″ samples for measuring the UCS, others have not specified the dimensions of the samples. Also, many groups have not specified the number of samples prepared and tested to prove the accuracy of their UCS results.
It can be seen from studies carried out two decades ago that the OBF composition hinders the mechanical properties of cement slurries. Inadequate information about the OBF used in the study makes the results obtained from the study invalid for comparison. It is seen that the reduction in the mechanical properties depend on the testing temperature and pressure for a given class of cement and curing time. Differences in the composition of OBF used along with the differences in curing time, temperature and pressure makes it difficult to compare and validate the results obtained by different research groups.
In study performed by Aughenbaugh et al. [25], it is also seen that for the same class of cement (H-1 and H-2) under the same testing conditions and the same OBF contamination, different results were obtained and the reason for this is unclear. Furthermore, the long-term effect on the mechanical properties of contaminated slurries is examined by few research groups. The limitation with longterm tests is to maintain the same pressure and (elevated) temperature over a longer duration. Research has been done on the rheological properties, but due to inadequate information provided in the literature, it becomes difficult to draw conclusions.
Romanowski et al. [32] have presented destructive and non-destructive tests carried out to determine the relationship between the unconfined compressive strength (UCS) and the ultrasonic pulse velocity (UPV) in the presence of additives. Three cement compositions tested in this study were API Class G cement, API Class G cement and 4% Bentonite, and API Class G cement and 10% Bentonite. The prepared samples were cured at atmospheric pressure and temperature for 1, 3, 7, 21, 30, 40, 70 and 150 days. Figure 14 shows the results obtained in this study and Figure 15 shows the comparison of the correlations obtained in this study with the previous work done on the same topic. Similar to the findings of Olteanu et al. [30], Romanowski et al. [32] specifically indicated that additives may change the UPV vs. UCS response and, thus, if the correlation equation is not known, the results of various researchers cannot be compared accurately.  A lot of effort has been undertaken recently to understand the mechanism of changes in mechanical and rheological properties of OBF-contaminated cement slurries. Shortcomings like the lack of standardization in testing methods for OBF-contaminated cement slurries and inadequate information provided in the literature made it difficult to compare the results. It is evident that the mechanical properties of cement slurries decrease with an increase in contamination. However, the reduction in mechanical properties is different for the same classes of cement in similar conditions.

Conclusions
This paper reviewed the work done by several researchers and tried to compare their results. The data found in the literature show that oil contamination may alter the cement mechanical properties by up to 50% if even a small amount of contaminant is trapped in the cement. This could have a catastrophic impact on well integrity. This paper shows that there is a large inconsistency in the way the data is reported and, in particular, the sample preparation. Reference values are hard to find among the studied references, which makes it difficult to accurately compare the results of various authors. Moreover, the curing time for which the mechanical properties have been reported varies largely from paper to paper, which also makes comparative studies more difficult. It can be concluded that laboratory testing at expected bottom hole conditions is necessary for approximation of the reduction in mechanical properties of OBF-contaminated cement slurries in order to understand OBF's effect on well cement integrity.

Conflicts of Interest:
The authors declare no conflict of interest.