Lost circulation refers to the phenomenon that a large amount of drilling fluid leaks into the formation during the drilling process. It not only consumes a large volume of drilling fluid and prolongs the drilling cycle but also can cause complicated situations such as borehole collapse, blowout, and drill pipe sticking due to improper handling. It could even lead to the failure of the borehole and severe engineering accidents^[1]. The lost circulation rate (Q_DL) of the drilling fluid is the most intuitive and easily measurable essential parameter reflecting the degree of lost circulation of the drilling fluid. Depending on lost circulation rate, the lost circulation of drilling fluid can be roughly divided into four categories: slight lost circulation (Q_DFL≤5 m³/h), moderate lost circulation (5<Q_DFL≤15 m³/h), severe lost circulation (15<Q_DFL<30 m³/h) and lost circulation without mud return (Q_DFL≥30 m³/h). Lost circulation without mud return undoubtedly belongs to severe lost circulation. Furthermore, severe lost circulation occurring in the formation with developed fractures (fractures and cavities) can also be classified into severe lost circulation due to the difficulties in the lost circulation control process. As oil and gas exploration and development extend to deep and ultra-deep unconventional and low-grade oil and gas resources, drilling projects are facing increasingly harsh geological conditions and critical technical bottlenecks. Severe lost circulation in fractured (fracture-cavity) formation is one of the most common, difficult, and complicated accidents in drilling engineering and has become one of the key bottleneck problems restricting the further development of drilling engineering^[2-3].

According to statistics, the occurrence rate of lost circulation in the world is about 20% to 25% of the total drilling wells, and the annual cost for lost circulation control is as high as 40×10⁸ US dollars^[4]. For the carbonate rock and shale oil and gas reservoirs in North America (the United States and Canada), the wells with severe lost circulation during drilling account for about 40% of the total wells^[5]. The wells with severe lost circulation during drilling of carbonate fractured reservoirs in the Middle East account for more than 30%, and the lost circulation time accounts for more than 50% of the drilling time^[6]. China's oil and gas drilling engineers also face the lost circulation problem. According to the statistics of the CNPC Oilfield Service Company Limited, the downtime caused by lost circulation accounted for more than 70% of the total downtime caused by complicated drilling accidents in domestic and overseas oil fields of PetroChina Company Limited from 2017 to 2018, and the average annual economic loss caused by lost circulation amounted to more than 40×10⁸ Yuan. From 2017 to 2019, a total of 76 wells were drilled in the deep formations of the Tarim Kuqa Piedmont, and 354 lost circulation accidents occurred, of which 65% were severe.

Currently, the success rate of one-time lost circulation control in fractured formations with severe lost circulation is low at home and abroad. According to statistics, in the CNPC Sichuan-Chongqing region, the success rate of one-time plugging in the fault section of the second drilling section is smaller than 10%, and smaller than 30% in the fracture section of the third drilling section. The success rate of one-time plugging is smaller than 40% in the thick salt layer of the Tarim Kuqa Piedmont. The success rate of one-time plugging for the Permian Formation Khuff fractured formation in the Ghawar oil field, Saudi Arabia is smaller than 20%^[6]. Apparently, severe lost circulation occurring in the fractured formations has become a global problem restricting the development of oil and gas exploration. Enhancing the success rate of one-time plugging for severe thief formation is an urgent need of oil fields around the world to ensure safe, efficient, and economic drilling.

Through scientific research, currently, various lost circulation control materials represented by bridging material, high water-loss material, curable materials, etc., have been developed. Consequently, a series of pressure-bearing lost circulation control technologies have been developed, which can meet the requirements of lost circulation control in permeable formations and formations with small and micro-fractures^[7-8]. But currently there is no effective prevention and rapid handling method for severe lost circulation occurring in formations with large fractures/fracture-cavities. For such formations, the success rate of one-time plugging is low, and the plugging cycle is long; the plugging operation depends more on experience, making it difficult to borrow successful lost circulation control methods. In response to the above problems, this article systematically discusses the research status of drilling fluid loss circulation mechanisms, lost circulation control mechanisms, lost circulation materials, and lost circulation control technologies, analyzes the current problems existing in lost circulation control technology for formations with severe loss circulation, and points out the future development direction of lost circulation control technology.

The study on lost circulation mechanisms of drilling fluid mainly covers the relationships between the amount of lost circulation and parameters such as pressure, time, fracture characteristics, and rheological properties of drilling fluid. On the drilling site, real-time monitoring of the flow of drilling fluid is a common way to detect lost circulation of drilling fluid and analyze the laws of lost circulation. At present, researchers usually simulate distribution and scale of underground fractures by inversion and then study lost circulation mechanisms of drilling fluid by establishing mathematical and statistical models describing the lost circulation of drilling fluid in fractured formations.

The mathematical models describing the lost circulation of drilling fluid include mainly one-dimensional linear model, two-dimensional planar model, and one-dimensional radial flow model. The one-dimensional linear model assumes that there is only one fracture in the formation and the drilling fluid flows in laminar flow in one-dimension along the fracture's extension direction^[9-10]. Researches using one-dimensional linear model have gone through the development process from initially assuming drilling fluid as Newtonian fluid to assuming drilling fluid as Bingham fluid and then to power-law fluid. The one-dimensional fracture model has evolved from simple one not considering properties of the fracture to ones considering fracture's deformation, surface filtration, and roughness, etc. The two-dimensional planar model assumes that there is a rectangular fracture at any dip angle in the formation and the drilling fluid exhibits as two-dimensional laminar flow along the extension direction of rectangular fracture^[11-12]. The two-dimensional planar model mainly studies the two-dimensional flow of drilling fluid along the fracture plane, and is more complex and closer to the actual conditions than the one-dimensional linear model. The one-dimensional radial model assumes that there is an infinitely long radial fracture in the formation that intersects with the borehole; the drilling fluid exhibits as radial laminar flow along the fracture's extension direction^[13].

Since the one-dimensional linear model is too simple in assumptions, the one-dimensional radial model and the two-dimensional planar model can simulate the flow characteristics of drilling fluid in single fracture more accurately, and reflect the laws of drilling fluid loss in fractures more objectively. Compared with the two-dimensional planar model, the one-dimensional radial model is simpler in calculation process, so the one-dimensional radial model is the mathematical model most commonly used in describing the lost circulation in fractured formations.

These three mathematical models can be used to analyze the effects of drilling fluid's rheological properties, fracture roughness, fracture surface filtration, fracture width, and other factors on the lost circulation pattern of drilling fluid, and to calculate the lost circulation pressure and fracture width, etc., which provide a theoretical basis for drilling fluid lost circulation control technology. However, the above three models are mainly based on the study of single fracture, which is quite different from the actual situation of lost circulation formation, so they can not predict the lost circulation pressure and fracture parameters of formations with complex fractures accurately.

The statistical models for lost circulation of drilling fluid mainly include lost circulation models based on machine learning and lost circulation models based on data statistics. (1) Lost circulation models based on machine learning: These models train samples with given lost circulation parameters (such as formation lithology and physical property, lost circulation rate, rheological parameters, etc.) to estimate the dependence relationship between input and output data and then predict the effect of lost circulation control measures as accurately as possible. As it is difficult to describe complex non-linear relationships in drilling fluid loss with mathematical methods, machine learning models often predict the type of drilling fluid loss, the lost circulation rate, and the measures for preventing and controlling lost circulation by using non-linear functions. Based on algorithms such as artificial neural network (ANN), support vector machine (SVM), and decision tree, Sabah et al.^[14] applied machine learning to the diagnosis of drilling fluid loss, proving that machine learning-based lost circulation model could utilize big data to select prevention and control measures for lost circulation. (2) The lost circulation models based on data statistics. On the basis of a large number of on-site lost circulation data in drilling process, statistical analysis was performed on the relationships between the lost circulation rate of drilling fluid, the density of drilling fluid, viscosity, leak channels in the formation, and other parameters, to establish pressure calculation models for lost circulation of drilling fluid based on statistical principles. But the lost circulation models based on data statistics only use the lost circulation data of some production areas as the basic data and established the model through statistical analysis and data fitting. Therefore, the models are only suitable for some regions, and aren’t universally applicable.

From the above review, in the study on the lost circulation mechanism of drilling fluid, the establishment of an accurate and reliable model plays a key role. The present mathematical models for loss circulation are mainly based on single-fracture, which is quite different from the actual situation of the lost circulation layers. Lacking the research on calculation models for loss pressure and fracture width under complex three-dimensional fracture conditions. Furthermore, the present statistical mathematical models are based on the drilling data, lost circulation prevention and lost circulation control data of adjacent wells, belonging to empirical prediction models. These two kinds of models have low prediction accuracy on lost circulation pressure and fracture scale, so their calculation results have no strong guidance for plugging operation, so they are difficult to popularize^[15].

Research on the lost circulation control mechanisms is divided into research on theoretical lost circulation control mechanisms based on rock mechanics assumptions and research by physical modeling experiments on lost circulation control mechanisms, mainly including stress- cage theory, fracture plugging mechanism, and isolating plugging mechanism.

The stress-cage theory proposed by Aston et al.^[16] holds that regulating the tangential stress field around the well and the stress field at the fracture tip, a balance between the wellbore fluid column pressure and the in-situ stress field can be established to control loss circulation of the drilling fluid. After entering the fracture, the lost circulation aterial would form a plugging layer near the fracture entrance, blocking the transfer of drilling fluid pressure and fluid medium, increasing the circumferential stress, forming a stress cage, and then enhancing the pressure-bearing capacity of the formation. When the wellbore pressure increases, the increased circumferential stress can prevent fractures from opening or generation of new induced fractures. The key factor in strengthening the stress cage is to rapidly form a tight plugging layer with extremely low permeability in the fracture^[17].

(1) The mechanism of improving fracture closure stress: It is believed that the lost circulation material must deposit inside the fracture and simultaneously isolate the fracture tip to increases the fracture closure stress and then enhance the pressure-bearing capacity of the formation. After entering the fracture, the lost circulation control material quickly accumulates to form a plugging layer which enhances the closure stress of the fractures and isolates the fracture tip from the entrance, and the fluid behind the plugging layer filters into the pores of the surrounding rock, significantly reducing the pressure at the fracture tip^[18].

(2) The mechanism of controlling fracture tip extension: It refers to the use of lost circulation aterials to isolate the fracture tip and prevent the drilling fluid from transmitting pressure to the fracture tip, thus increase the fracture rupture pressure and stop the fracture from extension^[19]. When the drilling fluid is lost in the formation, the lost circulation material forms a tight plugging layer near the fracture tip to prevent the pressure wave from propagating and prevent the formation of induced fractures. The plugging layer at the fracture tip is easier to form in high-permeability formations but not easy in low-permeability formations. The lost circulation materials selected for fracture propagation control need to have a wide range of particle sizes.

The isolation plugging mechanism refers to the mechanism of isolating the wellbore and formation pressure systems by a high strength structure formed by lost circulation materials after physical and chemical reactions in the fracture under the induction of formation temperature, pressure, and fluid etc. to enhance the pressure-bearing capacity of the formation^[20]. The effect of improving formation pressure-bearing capacity by this mechanism depends on the self-adaptive capability in the environment, the reaction time and stability of the lost control materials, and the strength of the structure etc. The materials consist mainly of polymer gels, curable resin, etc., which are suitable for permeable-fractured formations with lost circulation.

Although some progress has been made in the study on lost circulation control mechanisms and field tests have been carried out with some satisfactory results achieved, the study on lost circulation control mechanisms for fractured formation is still not thorough in general: (1) Most theories are only qualitative understandings rather than quantitative calculations. (2) The stress-cage theory and fracture plugging mechanisms based on rock mechanics, and the isolation plugging mechanism based on experiment are suitable for lost circulation formations with high permeability and small/medium fractures, but have poor applicability in severe thief formations with abundant large fractures. (3) Previous researches mainly focused on lost circulation control mechanisms of rigid and granular materials, and few researches covered the lost circulation control mechanisms of flexible materials, and the migration, filling, and plugging mechanisms of lost circulation materials in three-dimensional fractures. So far, there is no in-depth study on the lost circulation control mechanisms of flexible or composite materials in three-dimensional fractures. Therefore, the current lost circulation control mechanisms have limitation in guiding current research and development of pressure-bearing lost circulation control technology, and on-site operation. The on-site control operations for severe lost circulation are not ideal in effect.

Lost circulation material is the foundation and key factor of lost circulation control technology. Researchers have successively developed several types of lost circulation materials, including bridging material, high water-loss material, liquid (water/oil)-absorbing swellable material, flexible gel, and curable material, etc. Lost circulation control mechanisms of different lost circulation materials for fracture have been explored. In recent years, smart materials have attracted increasing attention from researchers at home and abroad, and fundamental indoor research on smart lost circulation materials have been carried out.

The bridging lost circulation material is a composite lost circulation material prepared by inert granular, fibrous, flake materials according to a certain mass ratio and particle size grades. The commonly-used bridging materials include walnut shell, calcium carbonate, fiber, and mica flake. Bridging lost circulation materials mainly form a dense plugging layer by bridging, reinforcing, accumulating, and filling in the loss channels. With the advantages of small impact on the rheological properties of drilling fluid, low cost, and simple operation, this kind of material has been widely used in on-site operations and is suitable for permeable formations and formations with minor lost circulation. They play a crucial role in the lost circulation control. Amanullah^[21] developed a series of bridging granular lost circulation control materials (ARC) using date stone. Fractures 2 mm wide plugged with these materials had pressure-bearing capacity of greater than 8 MPa. Kang et al.^[22] evaluated the lost circulation control effects of different bridging materials, including rigid particles, elastic particles, and fibers, used individually and jointly on millimeter-level fractures. The experimental results showed that the composite bridging material of "rigid particle + elastic particle + fiber" had the optimal lost circulation control effect on millimeter-level fractures, which could increase the pressure-bearing capacity of 2-mm-wide fractures to 13 MPa.

But the commonly-used bridging lost circulation control materials have the following shortcomings: (1) The particle sizes of bridging materials poorly match with the sizes of the leaking channels in the formation. (2) Bridging materials may bridge and accumulate in the leaking channels, forming compacted accumulation under the action of pressure difference. The accumulation can reach high strength in small and medium-scale channels; but due to the influence of gravity settlement and erosion inside fracture etc., bridging materials are difficult to retain in large fractures with larger widths and higher longitudinal lengths, especially in karst caves, so the plugging layers in these fractures and caves are low in pressure-bearing capacity and prone to repeated lost circulation.

High water-loss lost circulation materials are the lost circulation materials prepared with diatomaceous earth, infiltrative material, and inert material according to a certain formulation. After entering fractures in the formation, this kind of material rapidly loses water under the pressure difference between the formation and the drilling fluid column, and its solid-phase components agglomerate, thicken, and form a film or filter cake rapidly, plugging the leaking fractures. The DiasealM lost circulation control material and the Diacel (DSL) lost circulation control material developed by Chevron Phillips are all typical high water-loss lost circulation control materials. Hou et al.^[23] developed a high water-loss, high pressure-bearing capacity, high acid-soluble lost circulation control material, FPA, by compositing the microporous and acicular natural mineral powders with different particle sizes as main materials, inorganic fibers as suspension materials, calcite particle as bridging material, fascircular mono-filament synthetic fiber and three-dimensional mesh synthetic fiber as fiber materials. When used alone, the plugging layer of FPA in the fracture of 1 mm wide reached the pressure-bearing capacity of 7 MPa. Easy to handle and quick to act, and high in success rate, high water-loss lost circulation control materials are suitable for plugging permeable and micro-fractured formations with low lost circulation rate, but need to be further improved to plug severe lost circulation effectively.

Liquid-absorbing swellable lost circulation control materials refer to water-absorbing/oil-absorbing materials used alone or in combination with other lost circulation materials, for example, lipophilic resin particles, pre-crosslinked gel particles, etc. After squeezed into the fractures under the pressure difference between the wellbore and the formation, the liquid-absorbing lost circulation control materials absorb water (oil) into their three-dimensional network structure via the intermolecular van der Waals force (or hydrogen bond) and the internal and external osmotic pressure difference, and expand in volume dramatically, forming a filling layer with good elasticity.

Wang et al.^[24] developed a super-absorbent gel particle with low cost, high strength, and high water absorption capacity. This material has a water absorption capacity of 137 g/g at room temperature, and a pressure-bearing capacity of over 3.0 MPa in high-permeability sand bed. Liu et al.^[25] developed a nano-micron deformable spherical oil-absorbing gel compatible with oil-based drilling fluid, which can plug micropores and micro-fractures while drilling to prevent or reduce the lost circulation of oil-based drilling fluid. Hashmat et al.^[26] found through research that compared with using swellable gel particles alone, using swellable gel particles in combination with rigid particles, fibers, and clay can significantly improve the pressure-bearing capacity of plugged fractures.

Liquid-absorbing swellable lost circulation materials have the advantages of expansion and good deformability, and hence are not affected by the shape and size of leaking channels. They can solve the self-adaptive lost circulation control issue that conventional bridging and high water-loss lost circulation control materials cannot, which are suitable for porous and fractured formations. But they have poor fluidabsorption delay and low tolerance to high temperature, so they don’t work so well in plugging formations with large fractures or karst caves.

Compared with other types of lost circulation control materials, the flexible gel materials with remarkable deformation by compression can self-adaptively enter leak channels with different sizes regardless of the shapes of the channels, and then form high-strength plugging layer in the loss channels. Therefore, this kind of material is suitable for loss channels of different sizes^[6], and can be roughly divided into crosslinked gel-forming and non-crosslinked gel-forming types.

The crosslinked gel-forming material refers to viscoelastic gel formed by polymer (or monomer), crosslinking agent, and initiator, etc. injected into the downhole loss channels in the form of water solution by crosslinking reaction, which plugs the leak channels. Crosslinked gels mainly are made by unhydrolyzed or partially hydrolyzed polyacrylamide and organic chromium, phenolic resin, and other crosslinking agents. With adjustable gelling time and gelling strength, the polymer gel system can be used to plug different types of thief formations. Non-crosslinked gels mainly form gel structures through entanglement or association between polymer chains with special functional groups. The ZND-type special gel developed by Nie et al.^[27] is the most representative non-crosslinked gel type lost circulation control material. In aqueous solution, the polymer chains spontaneously agglomerate through intermolecular hydrophobic association to form a reversible dynamic physical crosslinking network structure, filling leak fractures, forming a "gel slug" that can isolate the fluid inside the formation from the fluid in the wellbore, thereby achieving the purpose of lost circulation control.

Simultaneously with the research and development of polymer gel lost circulation control materials, researchers examined the lost circulation control mechanisms for fractures by polymer gels and proposed the "partition-type gel slug lost circulation control mechanism" for fracture-type severe lost circulation. The core idea is that the gel in the fractures should meet the requirements of "high fluidity, high viscosity, easy plugging, easy cleanup, easy filling, easy partitioning, and strong stability." Ivan et al.^[28] suggested that the crosslinked polymer gel can form a pressure-bearing plugging layer with sufficient strength in the fractures to prevent the propagation and induced expansion of fractures under the action of pressure, thereby isolating the high-pressure wellbore from the low-pressure formation. The study by Sweatman et al.^[29] showed that the pressure drop in gel slug from the wellbore to the formation reduces the pressure on the fracture tip and prevents expansion of the fracture. Therefore, enhancing the gel strength and the bonding strength between the gel and fracture surface can achieve the purpose of reducing pressure on the fracture tip and preventing expansion.

There are currently two main problems associated with polymer gel lost circulation control materials: (1) They are generally poor in high-temperature resistance and long-term stability under high-temperatures, resulting in ineffective plugging of fractures or high risk of recurring lost circulation after plugging. (2) Commonly-used polymer gel materials have poor internal drainage and poor retention in fracture; the laws of dynamic migration, spreading, and filling behaviors of gel in the three-dimensional fractures are not clear; and the relationship between gel migration/filling and the lost circulation control effect is not clear, and the lost circulation control mechanism isn’t clear.

The commonly-used curable materials refer to compound lost circulation materials made up of cement, slag, gypsum, lime, silicate, and other mixtures and activators. Cement is a typical curable material. The lost circulation control effect of cement can be enhanced by adding various additives into cement slurry and modifying squeezing process. Cement has the characteristics of strong pressure-bearing capacity and dramatic plugging effect for severe lost circulation. Its main lost circulation control principle is: after pumped into the downhole lost circulation formation, the cement slurry would thicken and cure after some time, forming a high-strength solid body cemented with the formation to plug the thief layer.

In recent years, researchers in China have developed a variety of new cement, such as rapid-drying cement, expansive cement, etc. In combination with high-efficiency cement accelerators or retarders, the various kinds of cement have a wider application scope, better performance, and higher success rate of plugging. Zhao^[30] develop a new thixotropic and curable lost circulation material by combining the advantages of bridging and cement materials. Its main plugging principle is that the highly-thixotropic plugging slurry increases in flow resistance after entering the fracture, so the lost circulation rate is overall reduced. The inert lost circulation materials bridge and accumulate at the fracture's mouth and temporarily plug the fracture, preventing further flow of the plugging slurry into the lost circulation formation. The plugging liquid entering the fracture cures after a certain period of time, bonding with the formation and enhancing the pressure-bearing capacity of the formation.

In addition to cement materials, Baker Hughes has developed a MAGNE-SET curable lost circulation control material, which can form a high strength solid by reaction between the mixture of magnesium and calcium oxides and water. Compared with other lost circulation materials, this curable lost circulation material has higher pressure-bearing plugging capacity, controllable curing time, low price, simple preparation and operation, so it is suitable for formations with severe lost circulation. But the material is easily diluted by formation water. When diluted, it drops in curing strength, and has curing time and rate difficult to control, so its safety risk in operation is high.

Smart materials are functional materials that can detect external stimuli and have judging and self-execution capability. Such materials can self-adapt to various complex formations and with excellent mechanical properties. Hence, they can significantly enhance the plugging efficiency and have broad application prospects in plugging drilling fluid loss. In recent years, both domestic and foreign researchers have successively developed materials such as smart shape memory polymers, smart gels, smart molecular membranes, and smart bionic materials. The smart shape memory lost circulation materials have the advantages of strong pressure-bearing capacity, self-adaptive bridging and sealing, and adjustable activation temperature, so they can be applied to fractured lost circulation zones. The smart gel lost circulation material has advantages of stronger self-adaptability, good compatibility, strong erosion resistance, and good degradability, and can be used to plug lost circulation formations with fractures and cavities. Both the smart molecular membrane and smart bionic plugging materials have good plugging performance for large pores and micro-fractures, and they have low damage to reservoirs due to unique biodegradability, and are suitable for plugging while drilling in lost circulation zones with high permeability and micro-fractures^[31].

Currently, most smart materials in the field of plugging drilling fluid lost circulation are still in the basic research stage, and field tests have not been performed in formations with severe lost circulation. Future study should focus on: (1) enhancing the pressure-bearing strength of smart shape-memory materials through solidifying technology; (2) enhancing the adaptability of self-healing and self-cementing smart gel materials under harsh conditions (i.e., high temperatures, etc.); (3) improving the thickness and strength of the smart film, and enhancing the applicability of the smart film to loss channels; (4) using bridging plugging, self-adaptive plugging, and smart bionic plugging jointly to synergistically intensify the plugging performance.

In conclusion, there is a wide variety of lost circulation materials developed. Typical lost circulation control materials and their main working mechanisms are shown in Table 1. To a certain extent, different types of lost circulation materials have achieved good application effects, but they still have some shortcomings: (1) The bridging lost circulation materials are complex in formula, and the bonding strength between the plugging layer and the fracture surface is weak, so recurrent lost circulation easily occurs due to pressure fluctuations and other factors. (2) The high-water loss lost circulation materials are complex in formula, with water loss rate difficult to control, and plugging difficult to remove, and it has poor applicability to water-sensitive formations. (3) The liquid-absorbing swellable lost circulation materials have liquid-absorption rate difficult to control, insufficient strength after liquid absorption, and poor high temperature resistance generally. (4) Conventional polymer gel lost circulation materials generally have poor resistance to high temperature and salinity, and thus poor long-term plugging stability. (5) Cement-based curable materials have poor resistance to dilution by drilling fluid and formation water; curing time difficult to control, so higher safety risk in operation. (6) Smart lost circulation materials are still in the basic research stage, with working mechanisms required further investigation, and need to be improved in adaptability to deep high temperature and complex formations.

Currently, cement, gel or composite lost circulation materials are used to handle fractured formations with severe lost circulation with low one-time plugging success rate generally. Efficient lost circulation materials suitable for severe lost circulation are still in deficiency. The commonly-used lost circulation materials for oil- based drilling fluid are mostly hydrophilic, and the types of special lost circulation materials are few. High pressure-bearing lost circulation materials for severe lost circulation of oil-based drilling fluid are scarce.

With the development of material science theory and technology, smart materials have been researched and used increasingly in several fields, especially in medicine and biology, which will give revolutionary promotion to the development of intelligent lost circulation control theory and technology of drilling fluid.

To deal with severe lost circulation in fractured formations, several types of lost circulation control technologies have been proposed and adopted. According to the type of lost circulation materials used, they are mainly divided into bridge plugging, gel plugging, cement plugging, composite material plugging, and mechanical plugging.

Bridging of lost circulation is currently one of the most widely used lost circulation control processes. Bridging materials are from a variety of sources, low in price, easy to prepare and operate, but the particle size of bridging materials must match with the fracture size. The phenomena of "entrance closure" or "loss inside the fracture" are likely to occur during operation, and recurrent lost circulation could occur during subsequent drilling.

In the Zhongguai-Manan area of the Xinjiang Oilfield, by optimizing and matching the particle shape, size, and strength of the bridging materials, a lost circulation control drilling fluid system for fractured formations with high pressure-bearing capacity has been developed, with a pressure-bearing capacity of over 7 MPa for fractures with the width of 1-3 mm^[32]. Aiming at the difficulty in matching the particle size of bridging material with the fracture size, Zhang^[33] proposed a “progressive bridging plugging process” in which the particle size of the bridging material varies from fine to coarse, and the concentration varies from low to high, to make it easy for the lost circulation materials to enter the fractures and form a bridge plug. This process can avoid the phenomenon of entrance closure and increase the success rate of plugging. The process has been used in six wells and seven well times on-site with a success rate of 100%.

It is difficult for bridging and high water-loss lost circulation materials to retain in lost circulation formations with large fractures and cavities to form a plugging layer. The polymer gel with strong deformation and retention characteristics can overcome the defects of granular materials and achieve ideal lost circulation control performance.

Based on the isolating plugging mechanism of gel slug, the ZND-type special gel with cement was tested in more than 30 wells in Changqing Oilfield and northeastern Sichuan area, achieving good effects. The technology has been also successful used in Well Shuangmiao 1 and Luojia 2 with blowout and lost circulation simultaneously. An important breakthrough was made in the use of the gel plugging technology to handle severe lost circulation^[34]. Based on the cross-linked polymer plugging agent (CACP) developed, combined with fiber and rigid particles, Lecolier et al.^[35] came up with the solid-phase particle reinforced gel slug plugging technology, which exhibited remarkable on-site performance in plugging severe lost circulation in wells in Louisiana, USA and Northern Iran. Baker Hughes developed MAX-LOCK, a magnesium oxide-based thixotropic inorganic gel, and optimized the gelling time of the MAX-LOCK gel based on factors such as the lost circulation rate, formation temperature and channel size to form the plugging technology of MAX-LOCK gel for lost circulation carbonate formation. This process has been successfully applied to handle severe lost circulation in carbonate formations in the Middle East, and has won the Best Oilfield Fluid and Chemicals Award at the 2019 World Petroleum Technology Awards and the 2019 MEA Best Drilling Fluid/Stimulation Technology Award.

In cement plugging process, the formulated cement slurry is pumped into the wellbore; the cement slurry enters the loss channels under the action of pressure difference, and solidifies under the action of the subsurface temperature to form a plugging body, thereby forming a high-strength plugging on the fractures. The plugging body of cement slurry has a high pressure-bearing capacity and strong bonding strength with the formation. Therefore, it is currently used more frequently in severe lost circulation situations than other plugging techniques.

Halliburton has developed a "one-bag" plugging technique. In this technique, plugging bags in different sizes filled with expandable cement are carried and placed into caves and large fractures, the cement expands, breaks the bag, and solidifies to plug the loss channel and control severe lost circulation. This technology can reach pressure-bearing capacity of 20 MPa, and plug fractures of 3-25 mm and has good temperature resistance and stability. X structure of the Yinggehai Basin in the western South China Sea has developed natural fractures, formation temperature of up to 204 °C, and formation pressure coefficient as high as 2.19. The formation has a low pressure-bearing capacity, and narrow safety window of drilling fluid density, so complex situations such as severe lost circulation are prone to occur during drilling. The intermittent cement squeezing process of "squeezing cleaning fluid + squeezing cement slurry" was used for high pressure-bearing plugging process for over 10 wells, and the success rate of plugging was 100%, which effectively ensured the safe and smooth drilling of the ultra-high temperature and high-pressure well section in the X structure of the Yinggehai Basin^[36].

The composite material plugging technology refers to the technology of plugging using multiple lost circulation materials or front and rear slugs. Particularly when high-density drilling fluid systems are adopted to drill into high-pressure formation with a narrow safe density window, induced fractures are likely to occur in the formation to cause lost circulation of drilling fluid. Composite materials can have the advantages of different types of lost circulation materials and achieve a satisfactory lost circulation control effect. At present, the main composite slug combinations are bridging material + high water-loss material, bridging material + curable material, gel + cement, etc.

Well Ming 1 in the Puguang Gas Field had severe lost circulation when drilled into the limestone formation. As the formation has low pressure-bearing capacity, firstly, bridge materials with high strength and immersion resistance were mixed with cross-linked consolidation agent to form a cross-linked film slurry; secondly, chemical consolidation slurry was prepared by using nanoscale and positively charged curable materials; finally, the cross-linked film slurry and the chemical consolidation slurry slug were successively injected into the formation to enhance the formation pressure-bearing capacity. This technology achieved favorable performance and no lost circulation occurred in the subsequent drilling process^[37]. Aiming at the formation in the Yingqiong Basin with high temperature, high pressure, and narrow safety drilling fluid density window (less than 0.10 g/cm³), Han et al.^[38] developed a high-density composite material of “bridging + high water-loss” plugging technology with a density of 2.30 g/cm³, increasing the safety drilling fluid density window of formation to 0.16 g/cm³ and the drilling efficiency greatly.

According to different plugging tools used, the commonly used mechanical plugging technologies mainly include expandable corrugated pipe and bypass valve.

Corrugated pipe is made by cold-pressing of round pipe to form plastic deformation in the radial direction and corrugated cross-sectional shape. The corrugated pipe has the characteristic of expanding and returning to a round pipe under the action of hydraulic pressure. Taking advantage of this characteristic, the corrugated pipe is lowered into the intended well section, and the pipe is expanded by hydraulic or mechanical methods in the well to form a round pipe with a larger cross-sectional diameter close to the well wall to achieve the purpose of plugging fractures in the formation. During the drilling of Well L7-71 in the Tuha Oilfield, more than 1000 m³ of drilling fluid was lost. Six times of plugging were performed in nearly 20 h, but failed. A 62.5 m long expandable corrugated pipe was run to 2451.0-2513.5 m, finally successfully plugging the lost circulation formation^[39].

When dealing with lost circulations in deep and ultra-deep wells, drill string has to be uplifted to replace plugging tools frequently, prolonging non-production operation time, and seriously affecting drilling efficiency. In order to solve this problem, DSI (Drilling Systems International), Baker Hughes, XDT (X Drilling Tools), and National Oilwell Varco have developed a variety of "wellhead ball-dropping type" and multiple activation-type bypass valve tools. With this kind of tool set in well, open-hole ball made of special material is used to plug the tool's inner hole, the build-up pressure in the tool pushes the internal sliding sleeve to complete the opening action of the hole; then another ball is dropped to close the bypass hole on the tool's side. This way, plugging operations can be completed without tripping out, reducing well control risks and shortening the drilling cycle. However, these tools have problems such as a limited number of switches in a single run, and high requirements on accuracy of parameters such as pumping rate and pressure^[40].

In short, lost circulation control technologies for formations with severe loss circulation have made breakthroughs in many aspects. In addition, field practices show that cement, gel, and composite lost circulation control technologies have a higher success rate in handling fractured thief formations and repeated plugging. But they have shortcomings like long plugging cycle; low success rate of one-time plugging; and non-replicability of successful lost circulation control technology.

Lost circulations (especially severe lost circulations occurring in fractured formations) are often sudden and complex. To enhance the success rate of one-time plugging of fractured formations by revealing the lost circulation mechanism of drilling fluid, developing special lost circulation materials and effective lost circulation control technologies is one of the key study contents in drilling engineering. A lot of experience have been made through lost circulation control technology research and practices at home and abroad. Lost circulation materials and methods have developed from single to multiple types. The formula design and operation process of lost circulation materials have gone from initial blind to empirical, and to preliminarily scientific. Exploration has also been made in the development of leakage point detection and mechanical plugging tools. Currently, the lost circulation control effect in fractured formations has improved significantly, basically meeting the needs of safe, efficient, and economical drilling in formations with slight and moderate loss circulation.

Although the progress in lost circulation control technologies in recent years has promoted further improvement in the success rate of one-time plugging, the lost circulation control effect of severe lost circulation isn’t ideal still, especially for formations with large-scale fractures and karst caves. Lost circulation mechanisms of drilling fluid and lost circulation control mechanisms, self-adaptive retention and filling performance of lost circulation control materials, high-temperature stability of lost circulation materials, and smart lost circulation materials and processes still need to be researched in depth.

Overall, the research of lost circulation control technology for fractured formation with severe lost circulation should focus on 5 directions in the future: (1) Research on the lost circulation mechanisms of drilling fluid and lost circulation control mechanisms: As the mechanisms of lost circulation and lost circulation control for various formations are not clear, giving little scientific guidance to lost circulation control technology. The main loss channels vary greatly under different geological conditions. Hence, it is necessary to select and optimize the lost circulation materials according to the different physical and chemical properties of the loss channels and the reservoir during plugging. The study on loss circulation and lost circulation control mechanisms is the foundation of lost circulation control technology. In the future, research on the mechanisms should be strengthened to clarify the lost circulation laws of drilling fluid in different formations and loss channels of different sizes find out the lost circulation control mechanisms and principles from the aspects of mechanical balance, loss channels, and lost circulation control methods, and provide a basis for the scientific selection of lost circulation materials, formulas, lost circulation control methods, and technologies. (2) Research and development of self-adaptive lost circulation materials: Conventional lost circulation materials have poor self-adaptive compatibility with the fractures at complex scales. In formations with developed natural (or induced) fractures, fractures are complex in scale, and sensitive to induction. The particle size of conventional bridging materials and the curing time of cement and gel lost circulation materials are difficult to match the various scales of fractures (especially dynamically-induced fractures). The "entrance closure" phenomenon is likely to occur on the well wall of the lost circulation layer or at the fracture's entrance. Consequently, recurrent lost circulation is common during subsequent drilling. In future work, it is necessary to determine the grading relationship between the lost circulation materials and the loss channels based on the lost circulation and lost circulation control mechanisms, and develop lost circulation materials with self-adaptive characteristics for the fracture space. (3) Research and development of lost circulation materials with strong retention and higher filling degree in three-dimensional fractures: In three-dimensional fractures, conventional lost circulation materials have weak retention and low degree of filling and plugging, especially in loss channels with large longitudinal scale such as large fractures, caves etc. Due to factors such as gravity and density, conventional lost circulation materials have weak retention in the longitudinal loss space, can not achieve effective full-filling of the leak space, and have poor lost circulation control effect. In the future, based on the parameters of the loss channel and physical and chemical characteristics of the lost circulation materials, high-efficiency lost circulation materials with strong retention and high degree of filling need to be developed. Self-adaptive distribution and filling laws of the materials should be figured out, to enhance their retention and filling degree in loss channels, improve the lost circulation control effect of loss channels, and increase the success rate of lost circulation control operations. (4) Research and development of high temperature-resistant lost circulation materials: Lost circulation materials have poor high-temperature resistance, and thus poor long-term plugging stability. Overall, their application is restricted in deep and high-temperature wells. But deep oil and gas are one of the future key targets for exploration and development of oil and gas resources in China. Compared with shallow to moderate depth formations, deep drilling generally faces high-temperature environment, which requires the lost circulation materials to have remarkable temperature resistance. The lost circulation materials commonly used have temperature resistance of less than 140 °C, and poor long-term stability at high temperatures, so they can not plug loss space of deep fractures firmly, and recurrent lost circulation is likely to occur later. Strengthening the research on the high-temperature resistance mechanisms of materials and developing lost circulation materials with high temperature resistance can ensure the long-term plugging performance of deep lost circulation formations with high-temperatures. (5) Development of big data and intelligent lost circulation control technologies: the current lost circulation control technologies for severe lost circulation are more dependent on experience, lack an expert system for analyzing and evaluating lost circulation control performance, and are not intelligent. Lost circulation control technologies for formations with severe lost circulation mainly rely on the lost circulation control experience of wells with similar conditions or adjacent wells, and there is no expert system for scientific optimization and evaluation, and a unified specification for the lost circulation control process. We should carry out research on integrated and intelligent lost circulation control technology, firstly, a plugging database for key regions needs to be established to form an applicable remote analysis and evaluation expert system. Secondly, basic research on intelligent lost circulation materials and methods needs to be strengthened while promoting the lost circulation control technologies to be digital and smart. In 2020, China National Petroleum Corporation formally established a Major Engineering Technology Field Test Project, "Field test of severe lost circulation control and high performance water-based drilling fluid technology", which will focus on developing intelligent lost circulation materials and tools, establishing a large comprehensive database of lost circulation, and constructing an expert system of remote judging and decision-making assistance for lost circulation prevention and control, with on-site application goals set. In the future, with the support from the National Oil and Gas Major Project of China, the National Key R&D Program of China, and several large oil corporations, intelligent lost circulation control technology and equipment is expected to achieve rapid development, to realize efficient and comprehensive management of severe lost circulation.

With the expansion of oil and gas exploration and development to deep and ultra-deep formations, unconventional, and low-grade oil and gas resources, severe lost circulation occurring in fractured (fractured-cavity) formations has become one of the most common downhole accidents and most difficult to handle during drilling. Some progress has been made in the research on drilling fluid loss mechanisms, lost circulation control mechanisms, lost circulation materials and lost circulation control technologies, but the existing technologies can not effectively handle the problem of severe lost circulation in fractured formations. In the future, the study on pressure-bearing lost circulation control technology for formations with severe lost circulation should integrate geology, engineering, material science and other disciplines; look deeper into the lost circulation and lost circulation control mechanisms; strengthen the adaptability of lost circulation materials and technologies for specific thief formations; and speed up research and development of intelligent lost circulation materials and expert database to establish an intelligent lost circulation material system. These researches can further improve the lost circulation control technical level for severe lost circulation, realize the “efficient, safe and economical” drilling for fractured formations with severe lost circulation, and accelerate the process of oil and gas exploration and development.