Haitao Li ,Zhaolong Ge ,Na Wei ,Wantong Sun ,Yao Zhang ,Jin Xue ,Lin Jiang ,Jun Pei ,Bing Liao ,Hongliang Cao ,Ruixu Zhang ,Shuangli Li ,Qixia Lu ,Jinyan Luo

a State Key Laboratory of Coal Mine Disaster Dynamics and Control,Chongqing University,Chongqing,400030,China

b State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation,Southwest Petroleum University,Chengdu,610500,China

c State Key Laboratory of Natural Gas Hydrate,Beijing,100027,China

d Drilling Department of CNPC Offshore Engineering Company Limited,Tianjin,300451,China

ABSTRACT There are many emergency risks in the process of natural gas hydrate(NGH)drilling.In order to ensure the safe and efficient exploitation of NGH,it is urgent to establish an intelligent judgment method for the risks in the process of NGH drilling.In this paper,the response relationship between monitoring parameters and risk categories of NGH while drilling is established.Based on fuzzy analytic hierarchy process(FAHP),the comprehensive weights of 10 risk monitoring parameters are obtained,including gas production,wellbore instability,hydrate ice barrier,drill string fracture,sticking,bit balling,drilling tool piercement,gas seepage,seabed subsidence and seabed landslide.Besides,the comprehensive judgment weight matrix is constructed,and the reasonable fluctuation range of monitoring parameters is formed.Thus,the intelligent judgment method of NGH drilling risk is established.The intelligent judgment and alarm of NGH drilling risks can be realized quickly and accurately by this method,namely,it can monitor the risks in the process of operation and guarantee the construction safety of NGH drilling.

Keywords:Natural gas hydrate Monitoring parameter Risk Analytic hierarchy process Fuzzy judgement

1.Introduction

The total amount of globe NGH resource is about(1.8-2.1)×1016m3,and the carbon content is twice as much as the global carbon content of proven fossil energy[1].NGH is mainly distributed in the permafrost zone and seabed sediments of epicontinental periphery,and the amount of marine natural gas hydrate is about 100 times more than that in the permafrost zone[2,3].Facing such a huge amount of NGH resources,the drilling and production safety issue of marine gas hydrate reservoir has become a major problem restricting the development of drilling and production technology of marine gas hydrate reservoir[4-7].In order to ensure the safe and efficient exploitation of NGH,it is urgent to establish an intelligent judgment method for the risks in the process of NGH drilling.Since the publication of the Japan's Methane Hydrate Development Program in 2001,Saito et al.have proposed a monitoring system consisting of a three-component servo accelerometer and a gear mechanism in the field of strata deformation prediction technology[8],which can continuously measure the seabed displacement calculated by the dual integration of the acceleration data.Since 2002,researchers in the United States have carried out researches on marine hydrate monitoring,and completed the construction and testing of the monitoring model[9].In 2005,the National Institute of Seabed Science and Technology of the United States established a comprehensive seabed observation station for natural gas hydrate in Block 118 of the Mississippi Canyon in the Gulf of Mexico,and the hydrocarbon system status within the HSZ in the deep water area of the Gulf of Mexico was under a long-term monitoring[10].In 2013,in the pilot test of hydrate production in the sea area of Nankai Submarine Trough,for the first time,Japan set up a temperature monitoring system near the borehole of the expected settlement center to carry out long-term monitoring of temperature and pressure for three consecutive months,and successfully detected the decline of reservoir temperature caused by the hydrate dissociation front[11].In 2016,He Tao et al.designed a hydrate trial production process monitoring system,which can collect the basic data of the reservoir in the process of hydrate exploitation in the sea area,and provide scientific basis for the subsequent hydrate resource evaluation and seabed environmental protection[12].At present,most of the researches focus on the monitoring of environmental changes in NGH reservoirs,while few scholars carry out research on the risk monitoring of NGH drilling.Therefore,we studied the risk judgment during NGH drilling and production in this paper.According to the actual working conditions and risks of NGH drilling and production,the response relationship between NGH drilling and production monitoring parameters and risk categories is established.Based on the FAHP theory,the monitoring parameters are classified and stratified,the overall weight of risk monitoring parameters such as gas production,wellbore instability,hydrate ice barrier,drill string fracture,sticking,bit balling,drilling tool piercement,gas seepage,seabed subsidence and seabed landslide is analyzed,and the weight matrix of comprehensive risk judgment is constructed.By analyzing monitoring data from drilled wells and combining with expert's experience,the reasonable fluctuation range of monitoring parameters is formed,and the intelligent judgment method of NGH drilling risk is established.The intelligent judgment and alarm of NGH drilling risks can be realized quickly and accurately by this method,namely,it can monitor the risks in the process of operation and guarantee the construction safety of NGH drilling.

2.Intelligent judgment method for natural gas hydrate drilling risk based on FAHP

2.1.Relationship between gas hydrate monitoring parameters and risk response during drilling

According to the actual working conditions of NGH drilling and risks of conventional drilling,it is found that there are 10 risks in the process of NGH drilling,including gas production,wellbore instability,hydrate ice barrier,drill string fracture,sticking,bit balling,drilling tool piercement,gas seepage,seabed subsidence and seabed landslide.During the occurrence of each risk,there will be changes of relevant monitoring parameters.This paper studies the variation rule of monitoring parameters to determine the type of risk.The monitoring parameters in the drilling process of NGH include injection fluid pressure,injection fluid flow rate,hanging weight,drilling time,torque,rotation speed,total hydrocarbon value,return fluid flow rate,return fluid pressure,return fluid temperature,seabed methane concentration,methane concentration in the middle of seawater,methane concentration in the sea surface,carbon dioxide concentration in the sea surface,seabed longitudinal deformation,seabed lateral deformation,seabed water turbidity and seabed water density.Among all the parameters above,injection fluid pressure and injection fluid flow rate are called injection parameters.The hanging weight,drilling time,torque and ROP are called drilling parameters.The total hydrocarbon value,return fluid flow rate,return fluid pressure and return fluid temperature are called return parameters.The seabed methane concentration,methane concentration in the middle of seawater,methane concentration in the sea surface,carbon dioxide concentration in the sea surface,seabed longitudinal deformation,seabed lateral deformation,seabed water turbidity and seabed water density are called environmental monitoring parameters.Based on the variation trend of monitoring parameters,the risk prediction of NGH while drilling is carried out,and the response relationship between monitoring parameters and risks is established(as shown in Fig.1).

2.2.FAHP for NGH drilling risk

Although a qualitative response relationship has been established between monitoring parameters and risks in the process of NGH drilling,the variation of one monitoring parameter may respond to one or more risks in the process of risk judgment.The response degree of this monitoring parameter to each risk needs to be described quantitatively.Therefore,a mathematical method is needed to solve the quantitative description of the relationship between monitoring parameters and risk response during drilling.It is found that the FAHP proposed by Buckley[13-15]can solve this problem well.FAHP can solve the subjective and cognitive fuzziness in the evaluation scales of traditional analytic hierarchy process.Furthermore,FAHP can be used not only in the judgment of uncertainty and subjective information,but also in the decisionmaking judgment based on experts' rich experience,insight and intuition.

Based on the FAHP,the hierarchical structure model of NGH drilling risk and monitoring parameters is established.According to the actual working conditions of NGH drilling and the analysis results of conventional drilling data,experts can judge the influence degree of each risk and monitoring parameters corresponding to it.The evaluation matrix is established according to the judgment results,and then the comprehensive judgment matrix is formed according to the established evaluation matrix,and the weight of each monitoring parameter to the risk response is calculated.The comprehensive judgment weight matrix is established based on the weight results.According to the relative variation rate of the measured parameter value of the monitoring parameters,the monitoring parameter variation vector is constructed,and the product of the weight matrix and the monitoring parameter variation vector is the possibility of risk occurrence.

(1)Establish risk hierarchical structure model for NGH drilling

The FAHP method is used to divide the monitoring parameters and risks in the process of NGH drilling into different layers.The risk is called the target layer(i.e.the first layer),the classification of monitoring parameters(such as injection parameters and return parameters)is called the criterion layer(i.e.the second layer),and the monitoring parameters(such as drill pressure,torque and hanging weight)are called the description layer(i.e.the third layer).Each component of the criterion layer and the description layer is called a factor,and the upper factor forms a subregion with its lower factor.

(2)Construct the risk evaluation matrix of NGH while drilling

Using the evaluation scale table of FAHP method,first compare the evaluation factors of the second layer of the criterion layer(i.e.the second layer)to form the evaluation matrix of the evaluation factors of the second layer,and then compare the evaluation factors of the third layer of the description layer(i.e.the third layer)to form the evaluation matrix of the evaluation factors of the third layer.The comparison results among the evaluation factors are expressed quantitatively by triangular fuzzy numbersexy=(e1,e2,e3).

In the determined hierarchical structure model,each factor and the next factor dominated by it form a subregion,and an evaluation matrix is established for this subregion.In the questionnaire survey,the FAHP evaluation scale table is used to evaluate the relative importance of each factor in the sub-region.

Firstly,taking the target layer as the criterion,the FAHP evaluation scale is used to compare the main evaluation factors of the second layer.Next,determine the judgment value,establish the evaluation matrix of the main evaluation factor according to the judgment value,and then establish the evaluation matrix of the secondary evaluation factor of the third layer respectively based on the second layer.For example,when factorxis compared with factorx,triangular fuzzy numberexy=(e1,e2,e3)is used to quantitatively represent the evaluation results.The evaluation matrix can be expressed as follows:

Where:Eis a fuzzy positive reciprocal matrix.

(3)Establish the comprehensive evaluation matrix of NGH drilling risk and calculate the fuzzy weight

Assuming that the number of experts for evaluation isj,the comprehensive evaluation matrix can be obtained by the fuzzy averaging method:

The comprehensive fuzzy evaluation matrix of NGH drilling risk is established as follows:

Then,the relative weight of the comprehensive fuzzy evaluation matrix of NGH drilling risk is solved by using the geometric mean fuzzy weight calculation method.After transforming the triangular fuzzy weight into a clear value[16]and normalizing the clear weight,the comprehensive weight of connecting each layer is obtained:

Where:the overall weight of the secondary factor,is the main factor weight,G’2iis the secondary factor weight.

The final comprehensive evaluation weight matrix is as follows:

Where:ETis comprehensive evaluation weight matrix.

(4)Establish the variation vector of risk monitoring parameters of NGH while drilling

The monitoring parameters of NGH while drilling are constantly changing,and the two variation trends are:increasing and decreasing.The variation of monitoring parameters is described by the relative variation rate.The relative variation rate of monitoringparameter value at certain depth of a well is used as the constituent element of monitoring parameter variation vector.The relative variation rate of monitoring parameters reflects the response intensity of NGH drilling risk.The variation trend and variation rate of monitoring parameters in the process of NGH drilling are used to construct the variation vector of monitoring parameters,"+" indicating the increasing trend and"-"indicating the decreasing trend.It is defined as“0’’ when the monitoring parameters variation within a reasonable range and“1’’ when the variation range is greater than or equal to 100%.According to the field engineer experience and a large amount of monitoring data analysis,the reasonable range of NGH drilling risk is formed,as shown in Table 1.

Table 1 Rational fluctuation range of monitoring parameters for NGH drilling risk.

Table 2 Basic data of well A.

The calculation formula of the constituent elements of the variation vector of NGH drilling risk monitoring parameters is as follows:

Where:biis the relative variation rate of monitoring parameters during NGH drilling.ΔSis the variation of monitoring parameters during NGH drilling.Sjis the measured value of monitoring parameters during NGH drilling.Slis the theoretical value of monitoring parameters during NGH drilling.Δmis the rational range of monitoring parameters in the process of NGH drilling.

The initial value of monitoring parameters for NGH drilling risk can be divided into“0”and not“0”.Formula 6 is applied to the case that the initial value of monitoring parameters is not“0”.When the initial value is“0”,formula 7 is applied to monitor the increased measured value of monitoring parameters,and formula 8 is applied to decreased measured value of monitoring parameters.

The variation vector of monitoring parameters for NGH drilling is as follows:

(5)Obtain risk evaluation result of NGH drilling

The product of the weight matrix of comprehensive evaluation for NGH drilling risk and the monitoring parameter variation vector is the result of risk evaluation during drilling:

The value of Y indicates the possibility of each risk.Obviously,the greater the value of Y,the greater the possibility of the risk corresponding to this value,and the smaller the value of Y,the smaller the possibility of the risk corresponding to this value.

The FAHP method constructed above is applied to stratify the risk types,monitoring parameter types and monitoring parameters of NGH drilling.The risk types of NGH drilling is set as the target layer,the injection parameters,drilling parameters,return parameters and environmental monitoring parameters are set as the main evaluation factor layer,and the monitoring parameters are set as the secondary evaluation factor layer.The risk evaluation matrix of NGH drilling is thus established.The comprehensive evaluation matrix of NGH drilling risk is established and the fuzzy weight is calculated,and then the comprehensive weight of various monitoring parameters in response to risks is solved.Finally,the comprehensive evaluation weight matrix of NGH drilling risk is established as follows:

Each column of the comprehensive evaluation weight matrix for the NGH drilling risk represents a kind of risk.From left to right,the first column represents the risk of gas production,the second column represents the risk of wellbore instability,the third column represents the risk of hydrate ice barrier,the fourth column represents the risk of drill string fracture,the fifth column represents the risk of sticking,the sixth column represents the drilling bit balling,the seventh column represents the drilling tool piercement,the eighth column represents the gas seepage,the ninth column represents the seabed subsidence,and the tenth column represents the seabed landslide.Each row in each column of the comprehensive evaluation weight matrix of NGH drilling risk represents the comprehensive weight value of a monitoring parameter for the risk represented in this column.From top to bottom,it represents the comprehensive weight values of injection fluid pressure,injection fluid flow rate,hanging weight,drilling time,torque,rotation speed,total hydrocarbon value,return fluid flow rate,return fluid pressure,return fluid temperature,seabed methane concentration,methane concentration in the middle of seawater,methane concentration in the sea surface,carbon dioxide concentration in the sea surface,seabed longitudinal deformation,seabed lateral deformation,seabed water turbidity and seabed water density.The second column is taken as an example to illustrate that when the risk of wellbore instability occurs,the comprehensive weight value of the increased fluid pressure is 0.154,that of the decreased injection fluid flow is 0.035,the increased hanging weight 0.109,the increased drilling time 0.074,the increased torque 0.207,the decreased rotation speed 0.084,the decreased return fluid flow rate 0.121,and the decrease of return fluid pressure 0.216.Other monitoring parameters have no response to the risk of wellbore instability.Therefore,the comprehensive weight values of other monitoring parameters are 0.

Similarly,the increase of total hydrocarbon value has the greatest impact on the gas production risk,and its comprehensive weight value is 0.332.The increase of injection fluid pressure has the greatest influence on the risk of hydrate ice barrier,and its comprehensive weight value is 0.311.The decrease of hanging weight has the greatest influence on the risk of drill string fracture,and its comprehensive weight value is 0.637.The increase of toque has the greatest impact on the risk of sticking,and its comprehensive weight value is 0.407.The increase of drilling time has the greatest impact on the risk of bit balling,and its comprehensive weight value is 0.432.The decrease of injection fluid pressure has the greatest impact on the risk of drilling tool piercement,and its comprehensive weight value is 0.796.The increase of seabed methane concentration has the greatest impact on the risk of gas seepage,and its comprehensive weight value is 0.250.The increase of seabed longitudinal deformation has the greatest impact on the risk of seabed subsidence,and its comprehensive weight value is 0.900.The increase of seabed lateral deformation has the greatest impact on the risk of seabed landslide,and its comprehensive weight value is 0.900.

2.3.Calculation model of monitoring parameters

To construct the variation vector of monitoring parameters,it is necessary to calculate the model calculation values and field measurement values of each monitoring parameter.The field measurement value can be obtained by the field monitoring instrument,but the model calculation value needs to be calculated by the mathematical model of the corresponding parameters.Therefore,the mathematical model corresponding to each monitoring parameter is established as follows.

(1)Mathematical model of wellbore temperature and pressure distribution[17].The riser drilling system is divided into five sections for calculation,namely,seawater drilling string section,stratum drilling string section,stratum annulus section,seawater annulus section and stratum section.The temperature and pressure model of seawater drill string section is:

Where:Tsais the micro unit annular temperature,K.Tspis the outer surface temperature of micro unit drill string,K.qmis the unit mass of micro unit,kg/m.cis the specific heat capacity of drill string,J/(kg·K).zis the unit length,m.dpiis the inner diameter of drill string,mm.ρspis fluid density in drill string,kg/m3.Uis the comprehensive heat transfer coefficient between annulus fluid and drill string,W/m2·K.

Coupling model of temperature and pressure inside drill string in stratum:different working conditions of trip out,trip in,rotary drilling and sliding drilling,the mathematical model of drill string torque[18]is established as follows:

Straight section:

Oblique section:

Where:Tsais the micro unit annular temperature,K.Tpis the outer surface temperature of drill string,K.Ppis the pressure inside the drill string,Pa.

The model of temperature and pressure in stratum annulus is:

Where:hfis the formation thermal conductivity,W/(m·K).Tfis near-well stratum temperature,K.Tais annular temperature,K.qgis the unit mass of intrusive gas,kg/m.cgis specific heat capacity of gas,J/(kg·K).qmis the unit mass of stratum rock,kg/m.cmis the specific heat capacity of stratum rock,J/(kg·K).Pais annular pressure,Pa.dpois the outer diameter of drill string,m.dwis borehole diameter,m.Qris the heat generated by the friction of the micro unit bit,J.Qhis the heat generated by formation or decomposition of hydrate,J.

The model of temperature and pressure in seawater annulus section is:

Where:Tsis seawater temperature,K.Tsais the outer surface temperature of the riser,K.Psais pressure at the outer surface of the riser,Pa.dpois the outer diameter of drill string,m.dwis borehole diameter,m.Qris the heat generated by formation or decomposition of hydrate,J.

Heat transfer model in stratum section:

Where:Δris the diameter of formation micro unit,m.

(2)Model of drilling string torque

The model considers various working conditions in the drilling process:trip out,trip in,rotary drilling and sliding drilling.The load distribution under different conditions is also calculated.Yang Kuiyi,who has done a lot of research on the kinematics of cone bit at home and abroad,made an in-depth study on the dynamic characteristics of cone bit in Dynamic Analysis of the Relationship Between Torque and Axial Pressure of cone Bit.Through the analysis of the lateral force,axial force and torque distribution under

Steady section:

Where:MTis the torque on the drill bit,kN·m.Pzis the axial pressure on the bit,kN.δ is the resistance coefficient of rock to rolling cone bit,m.Ψ is half of the cone angle of the cone bit,rad.D is the micro unit diameter of drill string,m.μbis the micro unit circumferential friction coefficient of drill string.His the number of micro units divided from the bottom of oblique section to the bottom of straight section.Ti+1andTiare the axial forces on the upper and lower sections of the drill string,N.Ais the crosssectional area of drill string unit,m2.Lis the length of drill string unit,m.ρ is the density of drill string unit,kg/m3.Δθ is the center angle corresponding to the drill string unit,rad.θ is the deviation angle,rad.μbis the friction coefficient between drill string and borehole wall,the downward direction is positive.Rwis the curvature radius at the well depth where the drill string micro unit is located,m.

(3)Rotation speed model of drill string

Combining the principle of drill string torque and drive motor,the ROP model of top drive drilling rig is established.The working conditions of top drive motor can be divided into three categories during drilling:①normal working condition,②working conditions within overload range and ③working conditions beyond the overload range.The drill motors are set with a certain overload capacity.Under normal working conditions,the powerPis the rated powerPe,and the rotation speednremains constant,which is the set rotation speed.Under the working condition within the overload range,the powerPvariations between the rated powerPeand the maximum powerPmax,and the rotation speednremains constant,which is also the set rotation speed.

Under the working condition beyond the overload range,the powerPis the maximum powerPmax,and the rotation speed decreases with the increase of torque.Substituting the torque calculation formulas(17),(18)and(19)into the motor rotation speed calculation formulas,the rotation speed mathematical models in straight section,oblique section and horizontal interval are obtained as follows.

The rotation speed of the bit in Straight section is:

The rotation speed of the bit in oblique section is:

The rotation speed of the bit in steady section is:

Suspended weight is the difference between drill string weight and drill pressure.Considering the influence of drill string friction,when the drill string moves upward,the direction of friction resistance of the drill string is downward and the hanging weight increases.When the drill string moves downward,the direction of friction resistance of the drill string is upward and the hanging weight decreases.Based on this,the mathematical model of hanging weight is established as follows:

(4)Rate of penetration(ROP)mode

Considering the chip hold down effect of cuttings,Hareland obtained the following drilling rate equation[19]:

Where:Ris the ROP,m/h.Dis the bit diameter,m.Nis the rotation speed,rad/min.Sis the rock strength,kN/m2.Wis the drilling press,kN.aandbare the coefficients.μ is plastic viscosity,Pa·s.ρsis drilling fluid density,kg/m3.Imis drilling fluid impact force,kN.fe(Pe)is the chip hold down function.

(5)Gas component model

In the process of NGH drilling,the amount of reservoir fluid production and injection fluid is equal to the amount of return fluid.Based on this,the concentration of each component is obtained as follows:

Where:Qnris the total hydrocarbon production of the reservoir,m3/min.Qniis the amount of injection fluid,m3/min.Cnois the total hydrocarbon concentration,%.

(6)Hanging weight model

Where:GXis the hanging weight during drilling,kN.KBis the buoyancy coefficient.qpis the weight of drill pipe,kN/m.Lpis the length of drill string,m.qcis the line weight of drill collar,kN/m.Lcis the length of drill collar,m.PZis drilling pressure,kN.Fis the friction resistance(‘‘+”indicating that drill string moves downward,‘‘-”indicating that drill string moves upward),kN.ρtis the density of drill string,kg/m3.ρgis the density of circulation fluid,kg/m3.

Drill string friction is calculated by formula 26 when drill bit is in oblique section.When drill bit is in steady section,it is calculated by formula 27.

3.Application example of fuzzy intelligent judgment on NGH drilling risk

Well A is a deep well located in the South China Sea.Taking well A as an example,the basic data of the well are shown in Table 2.

When the well is drilled to the depth of 4833.7 m,downhole anomalies occur.The parameters of 4785m-4833.7 m well section are calculated by the model.During the drilling process,the field monitoringequipmentcollectedtheparametersof 4785m-4833.7 m well section.The comparison diagrams of theoretical values and measured values of different depths are drawn based on model calculated values,background values and measured values from the depths of 4785m-4833.7 m,as shown in Figs.2-19.In addition,Table 3 shows the model calculated values,background values and measured values corresponding to each monitoring parameter in depth of 4833.7 m.

According to the relevant monitoring data in the process of NGH drilling(Table 2),the variation vector of monitoring parameters at the depth of 4833.7 m in NGH drilling is established by FAHP method.Combining constructed comprehensive evaluation weight matrix of NGH drilling risk,the intelligent judgment results of risk when drilling to the depth of 4833.7 m are obtained as follows:

Table 3 Comparison between model calculated valves and measured valves at the depth of 4833.7 m

In order to facilitate comparative analysis,based on the above judgment results,a histogram of risk possibility when drilling to the depth of 4833.7 m is drawn(as shown in Fig.20).Through comparative analysis,it can be seen that the possibility of drill string fracture risk is the largest,with the possibility of 71.572%,followed by the probability of drill tool piercement risk,with the possibility of 27.824%,and the possibility of formation gas production is the smallest,with the possibility of 0.604%.Therefore,the judgment conclusion of the application of the intelligent judgment method of NGH drilling is that the risk of drill string fracture occurs at the depth of 4833.7 m.Compared with the actual drilling conditions,it is found that the drill string fracture does occur at this depth of the well in the actual drilling process,which further proves the accuracy of the intelligent judgment method of NGH drilling.The reason why drilling tool piercement risk and gas production risk also account for certain possibility is that the decrease of hanging weight not only responds to the risk of drill string fracture,but also responds to the risk of gas production.The comprehensive weight value of the decrease of hanging weight to the risk of drill string fracture is larger than that of gas production.Similarly,the decrease of injection fluid pressure and the increase of drilling time both respond to the risk of drill string fracture and the risk of drilling tool piercement.Due to different comprehensive weight values of the decrease of the injection pressure and increase of drilling time responding to drill string fracture and drilling tool piercement,it is also possible that gas production risk and the drilling tool piercement risk occur,yet the possibility of these two risks is smaller.

4.Conclusions

(1)According to the variation trend of monitoring parameters when various risks occur in the process of NGH drilling,the response relationship between monitoring parameters and NGH drilling risk is established.Based on FAHP method,the comprehensive weights of 10 risk monitoring parameters are obtained,including gas production,wellbore instability,hydrate ice barrier,drill string fracture,sticking,bit balling,drilling tool piercement,gas seepage,seabed subsidence and seabed landslide.Besides,the comprehensive judgment weight matrix is constructed.By analyzing monitoring parameters of drilled wells and combining with experts'experience,the reasonable fluctuation range of monitoringparameters is formed.Thus,the intelligent judgment method of NGH drilling risk is established.

(2)By analyzing the comprehensive judgment weight matrix of NGH drilling risk,it is found that:the increase of total hydrocarbon value has the greatest impact on the gas production risk,and its comprehensive weight value is 0.332.The decrease of return fluid pressure has the greatest influence on the wellbore instability,and its comprehensive weight value is 0.216.The increase of injection fluid pressure has the greatest influence on the risk of hydrate ice barrier,and its comprehensive weight value is 0.311.The decrease of hanging weight has the greatest influence on the risk of drill string fracture,and its comprehensive weight value is 0.637.The increase of toque has the greatest impact on the risk of sticking,and its comprehensive weight value is 0.407.The increase of drilling time has the greatest impact on the risk of bit balling,and its comprehensive weight value is 0.432.The decrease of injection fluid pressure has the greatest impact on the risk of drilling tool piercement,and its comprehensive weight value is 0.796.The increase of seabed methane concentration has the greatest impact on the risk of gas seepage,and its comprehensive weight value is 0.250.The increase of seabed longitudinal deformation has the greatest impact on the risk of seabed subsidence,and its comprehensive weight value is 0.900.The increase of seabed lateral deformation has the greatest impact on the risk of seabed landslide,and its comprehensive weight value is 0.900.

(3)The established intelligent judgment method of NGH drilling risk is applied to carry out the calculation combined with the actual drilling data.The final judgment result is that the occurrence of drill string fracture,and the possibility of drill string fracture risk is 71.572%,which was consistent with the actual drilling conditions,thus,verifying the correctness of the intelligent judgment method of NGH drilling risk.

Acknowledgements

The research was supported by the National Key Research and Development Program(2019YFC0312300),the 111 Project(D21025),National Natural Science Foundation of China Item of China(U20B6005,51874252 and 5177041544),Scientific Research Starting Project of SWPU(2018QHZ007),Open Fund Project of State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation(PLN2021-02 and PLN2021-03),Found of Southern Marine Science and Engineering Guangdong Laboratory(Zhanjing)(ZJW-2019-03).