An Investigation on the Function of Mud Cakes on the Inhibition of Low Molecular Inhibitor for Water-Based Drilling Fluids

Abstract

The major low molecular inhibitors showed inhibition in the hydration of clay in the laboratory for water-based drilling fluids, according to the principle of intercalation adsorption. However, inhibitors have failed and caused serious engineering accidents in drilling oil and natural gas. This paper investigated the transmission of several of drilling fluids to indicate whether low molecular inhibitor for drilling can effectively inhibit the wellbore hydration. The inhibition of drilling fluid with the plugging of mud cakes, was significantly weakened based on the hydration expansion of cores and cutting recoveries. The residual contents of inhibitors were determined with the precolumn derivation of high-performance liquid chromatography (HPLC) analysis and were chartered with Fourier transform infrared spectroscopy (FTIR) and nuclear magnetic resonance (NMR) analysis in the structure of the derivative. The clogging behavior of the mud cake was described by environmental scanning electron microscopy (ESEM). Experiments show that 40 wt% to 90 wt% by weight of the corrosion inhibitor cannot pass through the mud cake in the dynamic filtration of the drilling fluid. The mud cake can be further divided into a nanostructure layer, a homogeneous layer and an anisotropic layer with different permeability. Most inhibitors should be limited to the nanostructure layer and the homogeneous layer.

1. Introduction

Drilling fluid, sometimes called the blood of drilling, acts to carry cuttings, lubricate, and balance pressure. During drilling with water-based drilling fluid, wellbore instability is a serious problem in drilling shale oil or gas, and can lead to a series of problems such as breakage, fracture development, water-sensitivity, and leakage problems [1,2,3]. Shale is rich in hydrophilic clay minerals (illite, montmorillonite), excellent nanopores, and has a large specific surface area [4]. Shale is easily hydrated, which results in the expansion and dispersion of clay minerals and eventually leads to the collapsing of the shale [5,6]. Currently, based on the adsorption mechanism of hydration intercalation of clay lattice, various types of small molecular surface hydration inhibitors have been developed for water-based drilling fluid [7,8,9,10].

Suter et al. (2011) proposed the rule based design of clay swelling inhibitors and some reasonable molecules, such as 1,6-hexamethylenediamine (HMDA), tetramethyl ammonium (TMA), tetraethyl ammonium (TEA)and poly (ethylene glycol) (PEG) [11]. Xie et al. (2017) investigated the adsorption of 1,6-hexamethylenediamine (HMDA) with two primary amines adsorbed on sodium, using isothermal adsorption, adsorption kinetics, scanning electron microscopy (SEM), X-ray diffraction (XRD) and elemental analysis (EA) techniques [12]. As the result, the spacing of the interlayer of montmorillonite can be significantly decreased to 3.2 Å as large as the molecular diameter of the primary amine groups.

Currently, a series of poly(oxypropylene)-amidoamine (POAA) compounds were synthesized as potential shale inhibitors by condensation of a low molar mass of polyoxypropylene diamine POP230 with diacids [13,14]. The intercalation of POAA reduced the water content of the montmorillonite and made the clay mineral surface more hydrophobic.

In summary, they believed that low-molecular-weight inhibitor can be effectively inserted into the clay mineral interlayer to squeeze out the interlayer water molecules with a combined action of competitive adsorption and hydrogen bond interaction, so that the hydrated clay can be dehydrated. Therefore, active inhibitors that exhibit strong inhibition in the laboratory have been widely used in oil fields.

However, even with the increase of inhibitors in the production of oil and gas, wellbore stability cannot be effectively improved, and wellbore collapse also occurred [15,16,17,18,19]. Besides, the use of surface hydration inhibitors will increase drilling cost, the difficulty of drilling fluids treatment and harm to environment [20,21,22].

Therefore, this paper was adapted to investigate whether the low-molecular-weight inhibitor can interact with wellbore. As we know, the drilling fluid (mainly used to improve drilling efficiency and safety) can also lose water instantaneously, forming a nano-scale dense membrane on the wellbore with differential pressure. The membrane is mainly composed of nanometers, micrometer particles and polymers, and is an obstacle before the interaction of water molecules with the wellbore, also known as mud cake. However, in the past, we ignored the function of the mud cake on inhibitors.

Therefore, this paper investigates whether mud cakes can prevent the interaction of the low-molecular-weight inhibitor with wellbores, and provides reference for drilling engineers, oilfield chemistry scientists and clay technologists. A variety of drilling fluids were designed to study the effect of mud cakes on the entrance of the low- molecular-weight inhibitor and the formation of drilling fluids is shown in

Table 1. The preparation of different mud cakes with different drilling fluids.

ID	Bentonite/%	SMP/%	HMDA/%	LV-PAC/%	SiO2/%	KCl/%
1#	3.5	0.5	0	0	0	0
2#	3.5	0.5	1	0	0	0
3#	3.5	0.5	0	0.5	0	0
4#	3.5	0.5	1	0.5	0	0
5#	3.5	0.5	0	0.5	0.5	0
6#	3.5	0.5	1	0.5	0.5	0
7#	3.5	0.5	0	0.5	0.5	3
8#	3.5	0.5	1	0.5	0.5	3

The concentration of the agent was based on mass percent; the solvent of drilling fluids was deionized water.

.

2. Materials and Methods

2.1. Materials

In the preparation of drilling fluids, bentonite was obtained from the Nanocor Company [23]. The chemical composition of Na + Bent is 13.22% Al₂O₃, 71.30% SiO₂, 7.10% MgO, 4.79% Na₂O, and 3.59% Fe₂O₃. Bentonite is mainly composed of montmorillonite and is used to improve the cutting ability of water. Potassium chloride (KCl) with high chemical purity was obtained from Sigma-Aldrich (Darmstadt, Germany), which can provide ions to balance the charge of clay minerals and improve the stability of bentonite during drilling [24]. Sulfonated-phenol-formaldehyde resin (SMP) and low viscosity poly anionic cellulose (LV-PAC) for drilling are supplied by Dafang Synthetic Chemical Company (Chongqing, China) to meet industrial grade and to stabilize drilling fluids [25]. Nano-silica particles obtained from Aladdin (Shanghai, China) were satisfied with chemical analysis and were used to reduce the water loss of drilling fluids [26]. Further, the BET (Bentonite) specific surface area and size of nano-silica particles were 150 m²/g and 7–40 nm, respectively. Apart from that, a popular 1,6-hexamethylenediamine (HMDA) inhibitor was added to inhibit the hydration of clay minerals [27], which was obtained from Sigma-Aldrich and was satisfied with chemical analysis.

Methyl alcohol was obtained from Sigma-Aldrich, satisfied with high-performance liquid chromatography (HPLC)grade and was used to dilute the sample. The benzoyl chloride was obtained from Sigma-Aldrich and its quality reached the standard of chemical analysis. In order to determine the content of HMDA in the filtrate of drilling fluids, the benzoyl chloride was prepared for the precolumn derivation of HMDA. In the structure characterization of the derivation, the dimethyl sulfoxide (DMSO) and potassium bromide (KBr) were prepared by 1H nuclear magnetic resonance (NMR) analysis and infrared (IR) spectrum analysis, respectively. Both of them were quite satisfied with the standard of spectroscopy analysis and were obtained from Aladdin (Shanghai, China). Liquid nitrogen was supplied by Xinyuan Chemical Company (Chengdu, China), and was prepared to describe microscopic morphology of mud cakes.

2.2. Preparation of Drilling Fluids

The formulations of drilling fluids are showed in Table 1. In the first phase, the basal mud was prepared by using a mechanical stirrer. Each basal mud has 3.5 wt% bentonite and 0.5 wt% SMP polymer in deionized water. Stirring was applied for 24 h for full hydration of bentonite and SMP polymer [28]. In the second phase, the basal mud with 0.5 wt% LV-PAC reducer for 3 h reduced the water loss of the mud [29]. In the third phase, 0.5 wt% nano-silica was stirred with the mud for 2 h to further reduce the water loss of the mud by plugging the nano-pores of wellbores [30]. In the fourth phase, a large amount of ions were supplied by stirring 3 wt% potassium chloride with the mud for 2 h. By balancing the charge of clay and polymers, the stability of drilling fluids can further improve [31]. Four drilling fluids, such as 1#, 3#, 5# and 7#, were prepared in this way. In contrast, each of the previous four drilling fluids was stirred with 3 wt% HMDA inhibitor for 2 h to inhibit the hydration of wellbores during drilling [27]. The other four drilling fluids, 2#, 4#, 6# and 8#, were prepared in this way.

2.3. The Dynamic Filtration of Drilling Fluids

The dynamic filtration of drilling fluids was achieved according to the standard of American Petroleum Institute (API standard) by using GGS42-2A high-temperature and high-pressure water loss instrument which could promise the maximum temperature and pressure are 350 °C and 50 MPa respectively, and was obtained from Tongchun Petroleum Instrument Factory (Qingdao, China). Firstly, a filter paper was slotted on the bottom of the container of the instrument. The paper for testing was produced according to API standards and the maximum diameter of the paper holes was below 20 μm. After that, 270 mL drilling fluid was poured into the container and then nitrogen gas was gradually injected into the container until the pressure of the container increased to 3.5 MPa. It is worth mentioning that the valve in the bottom should be closed, otherwise the drilling fluid can flow out early. When the pressure of the container was stable at 3.5 MPa, the container was heated to 120 °C by the instrument. Until the temperature was constant, the valve in the bottom was opened for 30 min and the filtrate of drilling fluid was collected for linear expansion, roller recovery and liquid chromatography analysis. Finally, the residual drilling fluid in the container was poured and the mud cake on the bottom filter paper was collected for ESEM analysis.

2.4. Dynamic Linear Expansion Experiment

After the filtration of drilling fluids, the expansion height of the core affected by the filtrate was investigated according to API standards [23]. In order to show the changing of the hydration of the core, the core was made of dried bentonite. In the experiment, the bentonite was dried for 24 h by a drying oven at 220 °C, then 10 g dried bentonite was poured into a pressure tank. The pressure tank withstood 10 MPa pressure by hydraulic press for 5 min. Then 10 mL filtrate was poured into the pressure tank and the linear expansion height of the core with time was measured by CPZ—Ⅱautomatic linear expansion recorder, produced by Tongchun Petroleum Instrument Factory (Qingdao, China). The changing of the expansion was recorded 5 times per second and the accuracy of the recording of the expansion height was 0.1 mm. Based on the API standards, the linear expansion ratio was obtained using the below Equation (1).

$$\mathrm{V}/\mathrm{H}=({\mathrm{R}}_{\mathrm{t}}-{\mathrm{R}}_{\mathrm{o}})/\mathrm{H}\times 100\%$$

(1)

where V/H is the linear expansion ratio, %; H is the original height of the core, mm; R_t is the 16-h reading, mm; and R_o is the initial reading, mm.

2.5. Roller Recovery of Cuttings

The drilling cuttings were obtained from shale of Longmaxi formation (Yibin, China). Experiments have shown that 50 g cuttings through 10-mesh sieves were added into a high-temperature aging tank. Then, 350 mL filtrate was added at 120 °C and rolled together with the cuttings for 16 h. The residual cuttings in the high-temperature aging tank were carefully poured over 40-mesh sieves and washed with tap water. The cuttings left on the 40-mesh screen were collected and dried at 105 °C for 24 h until the amount remained unchanged. In this way, the first-roller recovery of the cuttings could be obtained with the following Equation (2) based on API standards [23]. Similarly, the residual cuttings were again rolled with 350 mL same filtrate and subjected to the same experiments. The second-roller recovery of drill cuttings was also obtained with Equation (2). Furthermore, the cuttings were subjected to a third roll under the same conditions. The third-roller recovery of drilling cuttings was also calculated from Equation (2).

$$\mathrm{R}=\mathrm{M}/50\times 100\%$$

(2)

where R is the roller recovery of cuttings, %; and M is the recycling quantity of the drill cuttings after hot-rolling.

2.6. The Precolumn Derivatization of HMDA

With HMDA lacking of ultraviolet absorption, the content of the HMDA cannot be determined directly using the chromatograph. Therefore, this paper referred to previous research about the derivatizing of amines and attempted to develop HMDA using benzoyl chloride [32]. The mechanism of the derivatization is shown in Figure 1.

The filtrate was centrifuged at 3000 rpm cycles for 10 min. After that, the solid impurities at the bottom were removed and the liquid at the top was collected. Then, the 10mL upper liquid was stirred with 0.4 g sodium hydroxide at 40 °C for 30 min. The product attached to the bottle side was washed with sodium hydroxide aqueous solution. It was then dried by rotary evaporation in 115 °C to give a white solid. The 200 mL of methanol was added to completely dissolve the residual solids and subjected to column chromatography on 0.5 UL.

2.7. The Characterization of the Derivative

The characterization of the derivative was conducted with Fourier transform infrared spectroscopy (FTIR) by using a WQF-520 FTIR spectrometer (Beijing Rui Li Analytical Instrument Co. Ltd., Beijing, China). The 1H NMR characterization of DMAA-14 in DMSO was performed with a Bruker AVIII 400 MHz spectrometer (Bruker, Germany).

2.8. Liquid Chromatography Analysis Experiment

The content of HMDA in the filter liquor can be quickly determined using Agilent 1290 reversed phase high performance liquid chromatographic (HPLC) which was supplied from Agilent. The chromatographic column for analysis was Eclipse Plus C18(2.1 × 50 mm, 1.8 um). In the experiment, the mobile phase consisted of methanol and water (V: V, 60: 40), the flow rate was 0.200 mL/min, the column temperature was 35 °C and the UV detection wavelength was 229 nm.

2.9. Environmental Scanning Electron Microscope

After the filtration, the fresh mud cakes were selected and quickly crystallized at −40 °C with continually added liquid nitrogen. Then the microstructure of the surface of the frozen cake was investigated under high vacuum environment, by using FEI Quanta 450 environmental scanning electron microscope (FEI, America).

3. Results and Discussion

3.1. Characterization of HMDA Derivative

The precolumn derivatization of HMDA with benzoyl chloride makes it possible to determine the change of HMDA content in drilling fluid. The IR analysis of the derivative is shown in Figure 2. A number of sharp absorption bands are observed near 833 cm⁻¹, which are the out-of-plane bending vibration of the benzene rings. The characteristic peak at 1536 cm⁻¹, located between 1560–1535 cm⁻¹, is the in-plane bending vibration of the N-H of secondary amide. This is the fundamental difference between parameters and primary amides. There are several absorption peaks near 1440 cm⁻¹, located between the 1610–1370 cm⁻¹, which is the stretching vibration of the carbon bond of aromatic ring. A moderate intensity absorption peak was observed at 3345 cm⁻¹, located between 3400 cm⁻¹ and 3300 cm⁻¹, which belongs to the stretching vibration of the N-H of aromatic secondary amide. The results indicated that the primary amine group of HMDA reacted effectively with benzoyl chloride. Besides, the characteristic absorption peaks of antisymmetric stretching vibration and symmetric stretching vibration of primary amine were not found in IR, indicating that the derivatives did not contain primary amine groups. Therefore, it was well-founded to speculate that the primary amine groups at both ends of HMDA reacted with benzoyl chloride.

The result of 1H NMR analysis development by HMDA is shown in Figure 3. The sharp peaks at 2.51 and 3.36 belong to solvent peaks. Through the hydrogen bonding of the solvent, there were multiple splitting peaks between 8 and 9 ppm, which were characteristic peaks of two benzene rings. The chemical shift of secondary amide was 8.46 ppm and turned to a low field. The single peaks at 3.24, 1.55 and 1.34 ppm are the characteristic peaks of 1H, H2, H3 in the carbon chain of the derivative, respectively. The results further determined that the development of HMDA followed the reaction mechanism.

3.2. Effect of Mud Cakes on the Content of HMDA in Drilling Fluid

Various drilling fluid pressures were in 3.5 MPa to simulate the physical process of the mud cake formation considering the bottom hole pressure. The blocking effect of mud cakes on the transmission of HMDA was analyzed by liquid chromatography. In HPLC analysis, molecules with different masses can be qualitatively distinguished from different retention time and are quantificationally distinguished from the peak area.

The HMDA content of the filtrate was indicated with the standard chromatograms of different concentrations of hexane diamine derivatives, as shown in Figure 4. In the figure, the curves only contained a single peak referring to the different concentrations of standard HMDA solutions, while 2#, 4#, 6#, 8# curves respectively referred to 2#, 4#, 6#, 8# drilling fluids after passing through mud cakes.

Based on the curves of standard HMDA solutions, the distribution of the characteristic peak of HMDA derivative can be ensured. It can be seen that the retention time of HMDA derivative is close to 2.03 min. Besides, the HMDA content can also be determined according to the integral area of characteristic absorption peak of HMDA, and the positive relationship between the peak area and the HMDA content can be obtained from Equation (3). Experimentally, the correlation coefficient is higher than 0.95, as shown in Figure 5.

S = 2671.26C + 138.63

(3)

where S refers to the peak area of hexane diamine derivatives, mAu*s; C refers to the HMDA concentration. Based on this equation, the residual HMDA concentrations after passing through the mud cake can be obtained.

The decreasing rate of HMDA content can be simply calculated with Equation (4), as shown in Figure 6.

$$\mathrm{X}=\frac{{\mathrm{C}}_0-{\mathrm{C}}_1}{{\mathrm{C}}_0}\times 100\%$$

(4)

where $\mathrm{X}$ refers to the decreasing rate of HMDA concentration; ${\mathrm{C}}_0$ refers to the initial HMDA concentration of drilling fluids before passing through the mud cake; ${\mathrm{C}}_1$ refers to the residual HMDA concentration of drilling fluids after passing through mud cakes.

Experiments show that 43.74% HMDA cannot pass through the mud cake composed with bentonite and SMP polymers. With the addition of reducer LV-PAC polymers, the decreasing rate of HMDA concentration was further increased to 66.47%, indicating that the plugging of the mud cake was significantly improved. In particular, with the adding of nano-silica particles, the decreasing rate of HMDA concentration can be dramatically increased to 85.29% which is nearly double than the decreasing rate of 2#. Although the plugging could be weakened in cake 8# with the adding of potassium chloride, the decreasing rate was still above 70%. Therefore, the plugging of mud cakes cannot be ignored, which is a fundamental reason for the low efficiency of the inhibitor.

3.3. Effect of Mud Cakes on the Inhibition of HMDA Drilling Fluid

3.3.1. Linear Expansion Experiment

The effect of different filtrates on the hydration and expansion of the core was investigated to analyze the influence of mud cakes on the inhibition of HMDA drilling fluid, as shown in Figure 7.

The experimental results show that the inhibition of pure HMDA aqueous solution is not obvious before and after the filter paper, and the core expansion rate is only reduced by 0.5%. Therefore, it is certain that most of the HMDA can pass through the paper. Besides, the core expansion also can be inhibited with the combination of HMDA and other treatment agents, indicating that some HMDA also can enter into the hydrated clay. However, by comparing the inhibitory effect of pure HMDA solution, the inhibition of hexane diamine through different types of mud cakes is obviously reduced, which indicates that most of the HMDA can be blocked by the mud cake. This is mainly because the pore structure of the mud cake composed of hydrophilic bentonite and polymers is significantly compact, which effectively improves the flow resistance of the inhibitor. Considering the chemical bond of the polar substance, there is a strong hydrogen bond and Van der Waals adsorption between mud cake and inhibitors, some small molecular inhibitors can be adsorbed into the cake.

3.3.2. Cuttings Roller Recovery

After the plugging of mud cakes, the cuttings of shale were rolled with filtrates at 120 °C for 16 h. As shown in Figure 8, the cutting recoveries in filtrates (filtrates 1 #, 3#, 5# and 7#) without HMDA were similar to that of deionized water. In the experiment, the first, second and third rolling recoveries of deionized water were 20.03%, 17.31% and 14.38%, respectively. Based on the rolling of deionized water, the cutting recoveries in filtrates 2# and 4# were significantly increased (see Figure 8). Even if the cuttings were rolled with the filtrate three times, the cutting recoveries in filtrates 2# and 4# were 32.45% and 29.93%, respectively. However, the cutting recoveries in filtrates 2# and 4# showed a great decrease in comparison with the cutting recovery in 1 wt% HMDA aqueous solution. This shows that only a few HMDA molecules can pass through mud cake and show an inhibitory effect on the hydration of cuttings. The rolling recoveries in filtrates 6# and 8# were similar with that in deionized water, owing to the adding of nano-silica particles. This indicated that the inhibition of HMDA drilling fluids could be faded with the plugging of mud cakes. Based on the dynamic linear expansion and rolling recovery, the plugging of mud cakes was closely related to the components of drilling fluids. Both LV-PAC polymers and nano-silica particles could significantly improve the plugging of mud cakes.

3.4. The Mechanism of the Generation of Mud Cake

Based on the dynamic filtration law of colloidal system (Equation 5), different sized particles will inevitably accumulate on the sidewall of the well [33,34,35]. Considering the forces on every clay particle, during dynamic filtration, under the condition of low Reynolds number, only particles with radius less than R^* can precipitate onto the mud cake surface.

$$\mathrm{R}*=\frac{3}{2}{(\frac{{\mathsf{\rho }}_{\mathrm{s}}}{\sqrt{3}{\mathsf{\rho }}_{\mathrm{f}}})}^{\frac{1}{\mathrm{n}}}·\frac{\mathrm{q}}{\mathrm{A}(1-{\mathrm{C}}_{\mathrm{m}}){\mathsf{\gamma }}_{\mathrm{m}}}$$

(5)

where ${\mathsf{\rho }}_{\mathrm{s}}$ refers to the density of solid phase; ${\mathsf{\rho }}_{\mathrm{f}}$ refers to the density of liquid phase; ${\mathrm{C}}_{\mathrm{m}}$ refers to the volume percentage of liquid in drilling fluid; A is the filtration area; n is a constant to describe the flow performance of the fluid; $\mathrm{q}$ is the rate of the dynamic filtration. ${\mathsf{\gamma }}_{\mathrm{m}}$ is the shear rate on the surface of the mud cake.

Therefore, the diameter of sedimentary particles is proportional to the dynamic filtration rate and inversely proportional to the shear rate on the mud cake surface. When the dynamic filtration rate of mud is high, both large and small particles can be deposited on the mud cake. However, the dynamic filtration rate will decrease gradually after the mud cake is formed, and only small particles can be deposited on the mud cake surface. Therefore, the resulting mud cake is heterogeneous, the bottom permeability of mud cake is large, but the upper permeability is small. As shown in Figure 9, a protective layer can be observed on the surface of mud cakes after the dynamic filtration. Similarly, the code symbol of mud cake referred to relevant drilling fluid. The pictures were obtained from the fresh cakes after filtration.

Obviously, the surface of mud cake changed from rough to smooth and a protective layer was successfully formed with the adding of polymers. Macroscopically, the protective layer was further improved by adding nano-silica particles (7–40 nm) and the surfaces of mud cakes appeared brighter and denser.

3.5. ESEM Characterization of Mud Cakes

The microscopic pore structure of mud cakes can be further caught with ESEM analysis, as shown in Figure 10. In order to save the fresh mud cakes, the cakes were frozen instantly at −40 °C with liquid nitrogen, and then ESEM analysis was carried out. In the figure, the yellow marking area referred to the developed pores on nanostructured layer which was significantly destroyed; the red marking area referred to the pore structures on homogeneous layer which was also visibly broken.

The clay particles mixed with cross-linked polymers significantly improved the glossiness and smoothness of the protective layer, which was due to the micron pores which could be further occupied by the polymers.

From pictures 1# and 3#, the pores on the surface of mud cakes were reduced obviously with the filtration reducer LV-PAC. Vividly, the microscopic pores were filled with agglomerate and amphiblestroid polymers, which may be due to the crosslinking of the LV-PAC and SMP polymers. By contrast, there were more pores on mud cakes with the adding of HMDA inhibitors from pictures 2# and 4#. Based on ESEM analysis, the obtained average pore size respectively increased from 2.62 to 4.81 μm and from 0.53 to 1.22 μm, as shown in Figure 11.

This indicated HMDA molecules could break and permeate into the surfaces of mud cakes. However, the surfaces of cakes were bright and smooth, and the pores were reduced further with the adding of nano-silica from pictures 5# and 6#, even if HMDA inhibitors were added. This was because the majority of nanopores were further occupied by nano-silica particles which effectively inhibited the permeation of HMDA molecules.

However, the nanolayer of the mud cake was slightly broken from picture 7#, due to the competitive adsorption of potassium ions to equalize the negative charge of clay minerals with exchangeable cations [36]. In particular, the size of pores on the surface of the nanostructured layer was improved by the synergistic reaction of potassium ions with HMDA molecules according to picture 8#. Besides, some pore structures on the homogeneous layer were developed to compare with picture 7#. However, 73.57% of HMDA molecules still could not measure residual HMDA concentration by high performance liquid chromatographic (HPLC) analysis, as shown in Figure 6.

In other words, although the pores of the mud cake developed well, most HMDA molecules were still restricted in the mud cake. This may be due to the strong adsorption between polymers and HMDA molecules with hydrogen bond and Van der Waals force. On the contrary, the increase in pore size may make it easier for water molecules to enter into formation and cause the hydration of wellbore.

3.6. The Behavior of the Plugging of the Mud Cake

Based on the mechanism of the generation of mud cake [33], the mud cake can be further divided into nanostructured layer, homogeneous layer and anisotropic layer depending on the difference of permeability, as shown in Figure 12. The high permeability anisotropic layer was formed at the beginning of dynamic filtration because large and small particles can be deposited with the high filtration rate of drilling fluid. With the forming of mud cake, only small particles can be deposited due to the reduced filtration rate of drilling fluids. In this way, the low permeability homogeneous layer can be formed. Furthermore, only nanoparticles can be deposited as the filtration rate further decreases, and eventually a nanostructured layer can be formed, as shown in ESEM analysis of mud cakes 3# and 5#.

With fluid infiltration, HMDA inhibitors and exchangeable cations can effectively break nanostructured layer but most of them cannot pass through the mud cake based on HPLC and ESEM analysis. This indicates that most HMDA inhibitors could be adsorbed and preferred to react with the clay mineral and polymers of the mud cake. Besides, most of them should be consumed by the low permeability homogeneous layers (nanostructure layer and homogeneous layer) which are rich in active polymers and clay particles.

4. Conclusions

With the low efficiency of inhibitors on the collapsing of wellbores, this paper investigated whether low molecular inhibitors for drilling fluid could effectively interact with wellbores. The inhibition of popular low molecular HMDA (Mr = 116) inhibitor was investigated with the plugging of mud cakes. After the plugging, the linear expansion of the core increased by 2–3 times with the filtrates of HMDA drilling fluids. Besides, the cutting recoveries of the filtrates were also significantly reduced to the same as that of pure water. The experiments indicated the failure of HMDA drilling fluids in inhibiting the hydration of wellbores.

The chemical structure of HMDA was modified with benzoyl chloride to determine the content of HMDA in filtrate. Based on HPLC analysis, 40 wt% to 90 wt% HMDA can be blocked by mud cakes before interacting with wellbores. It is found that the blockage of mud cakes is closely related to the composition of drilling fluids. Commonly used polymers (LV-PAC and SMP) and nano-silica particles can significantly improve the blockage of mud cake.

In addition, HMDA can obviously destroy mud cake surface and improve pore sizes with ESEM imaging, which can provide more opportunities for water molecules to interact with wellbores. Besides, the formation mechanism of mud cake and drilling fluid input are further discussed. The results show that the mud cake should be further divided into three parts according to the permeability (nanostructure layer, homogeneous layer and anisotropic layer). Most inhibitors should be consumed by nanostructure layer and homogeneous layer.