System and Method for Activity Identification and Problem Prediction During Oil and Gas Well Drilling
Field of the Invention
[0001] The present invention relates generally to drilling activities in oil and gas well-sites and more particularly, the present invention relates to a system and a method for optimized activity identification and problem prediction during oil and gas well drilling.
Background of the invention
[0002] Extraction of oil and gas requires a mechanical framework i.e. a drilling rig, to drill deep formations. Nowadays, modern drilling apparatus and sensor-equipped monitoring infrastructures are deployed, for instance, oil well rigs are deployed that accommodates various sensors integrated in different working units of the rig during an ongoing drilling process. It has been observed that optimized and safe execution of the drilling process needs constant supervision of several drilling parameters like hydraulic, mechanical and operational parameters measured through the integrated sensors.
[0003] For instance, typically, measured multivariate sensor data is gathered in a three-tier database known as Supervisory Control and Data Acquisition (SCADA). Supervision of ongoing drilling process is achieved by continuous monitoring of drilling parameters like hook load , Weight on Bit (WOB) , Rotation per Minute (RPM) , torque, Rate of Penetration (ROP) , mud properties, Total Depth (TD) , Bit Depth (BD) , Block Position (BP) , inlet and outlet mud densities which are collected in SCADA systems. SCADA systems continuously monitor drilling parameters and generate alerts whenever drilling process is identified as problematic and complications such as stuck pipe, hole deviation, loss

circulation, drill pipe failure, washout, mud contamination, hole cleaning and deterioration of drilling equipment arise during oil well drilling process. However, it has been observed that exhaustive monitoring performed by the existing SCADA systems tend to cause a high false alarm rate and fails to predict or detect drilling problems.
[0004] Further, it has been observed that conventional Artificial Intelligence (AI) and Machine Learning (ML) techniques used to develop high dimensional models using oil well drilling data for processing drilling activities fails to provide procedures necessary for data assimilation and assessment. Furthermore, typically, existing systems for classifying drilling activities process incoming data in windows (or data frames) and therefore accuracy of such systems is sensitive towards selection of an appropriate size of the window, which is fraught with inaccurate results. Yet further, it has been observed that existing systems are trained over small data sets for identifying and processing drilling activities that fail to capture different situations that may arise during drilling. As the existing drilling system's activity detection methodology are unreliable, real¬time triggering for problem prediction may not be performed by the existing systems. Most of existing methods which are followed to predict the drilling problems are not suitable for real-time application.
[0005] Moreover, it has been observed that most of the existing systems that process drilling activities for predicting hole cleanliness require certain parameters which may never be maintained in real drilling operations like constant hole diameter, constant friction factor between formation and cutting across all type of formation encountered and steady state hole cleaning process. Also, it has been observed that a 'no flow condition' occurs while performing drill pipe connection that affects a cutting transport efficiency, and as such prevents accurate processing outcome. Further, most of the existing experimental systems take into consideration of a drilling depth of not more than 100 feet,

whereas the actual drilling depth may go up to a few kilometers.
[0006] Furthermore, conventionally, hole cleaning at well-sites is carried out in two circumstances viz: while drilling where the rate of penetration (ROP) >0, and complete circulation before tripping where the rate of penetration
(ROP) =0. It has been observed that the change in velocity of drilling fluid along the length affects cutting transport of drilling systems. Most of the current models are based on cutting weight balance that depends on measurement of cuttings weight outlet at a shale shaker. However, in a place where the real-time cutting weighing facility is not integrated with data acquisition systems, cleanliness of a wellbore is to be estimated with some other available data, which the existing systems have been found to not take into consideration.
[0007] Yet further, typically existing systems fail to provide any report to predict right time to carry out tripping during the drilling operation. Existing systems also fail to provide real time optimization of drilling parameters to improve cutting transport by integrating developed software with real-time data acquisition system. Additionally, it has been observed that directional drilling is a complex process of controlling applied force on drill string subject to fluctuating external forces and usually require evaluation of drilling process and its derived parameters. For example, when the hydrostatic pressure of the wellbore is lower than the formation pressure, the formation fluid may enter into the wellbore with high force, which is referred as a kick. The kick may be regarded as the onset of blowout. Similarly, if the pressure applied is greater than pore pressure, lost circulation may take place.
[0008] In light of the aforementioned drawbacks, there is a need for a system and a method for efficiently and effectively detecting and classifying the oil well drilling activities. There is a need for a system and a method for

efficiently predicting drilling problems including anomalies and complications during drilling activities and providing real-time feedback to drillers. There is a need for a system and a method for optimizing drilling process parameters in real¬time for efficient drilling.
Summary of the Invention
[0009] In various embodiments of the present invention, a system for predicting problems during drilling operation is provided. The system comprises a memory for storing program instructions and a processor for executing program instructions stored in the memory. The system comprises a drilling problem prediction engine executed by the processor and configured to determine drilling parameters from sensor data collected from multiple sensors where the drilling parameters are associated with a drilling activity. The drilling problem prediction engine is configured to classify the drilling parameters using one or more fuzzy rule based classifiers where the classification is performed by assigning a class label to each of the determined drilling parameters to obtain the classified parameters. The class label represent pre-defined classes of activities during drilling operation. The drilling problem prediction engine is configured to sub-classify the classified parameters using a random forest classifier where the sub-classification is performed by assigning a sub-class label to the classified drilling parameters to obtain the sub-classified parameters. The sub¬class label represents attributes associated with the pre¬defined classes of activities during the drilling operation. The drilling problem prediction engine is configured to predict data based on the drilling parameters and the sub-classified drilling parameters and generate an alarm by identifying a deviation of the predicted data with a threshold value.
[0010] In various embodiment of the present invention, a method for predicting problems during drilling operation is

provided. The method comprises determining drilling parameters from sensor data collected from multiple sensors where the drilling parameters are associated with a drilling activity. The method comprises classifying the drilling parameters using one or more fuzzy rule based classifiers where the classification is performed by assigning a class label to each of the determined drilling parameters to obtain the classified parameters. The class label represents pre-defined classes of activities during drilling operation. The method comprises sub-classifying the classified parameters using a random forest classifier where the sub-classification is performed by assigning a sub-class label to the classified drilling parameters to obtain the sub-classified parameters. The sub-class labels represents attributes associated with the pre-defined classes of activities during the drilling operation. The method comprises predicting data based on the drilling parameters and the sub-classified drilling parameters and generating an alarm by identifying a deviation of the predicted data with a threshold value.
Brief description of the accompanying drawings
[0011] The present invention is described by way of embodiments illustrated in the accompanying drawings wherein:
[0012] FIG. 1 is a detailed block diagram of a drilling problem prediction system, in accordance with an embodiment of the present invention;
[0013] FIG. 2 is a flowchart illustrating drilling problem prediction, in accordance with an embodiment of the present invention;
[0014] Fig. 3 is a dataflow diagram illustrating prediction of the cutting transport and flow hydraulics, in accordance with an embodiment of the present invention;
[0015] FIG. 4 is a flowchart illustrating optimization of the cutting transport, in accordance with an embodiment of

the present invention;
[0016] FIG. 5 illustrates prediction of kick and lost circulation, in accordance with an embodiment of the present invention;
[0017] FIG. 6 illustrates determination of energy consumption and carbon dioxide emission; in accordance with an embodiment of the present invention and
[0018] FIG. 7 illustrates an exemplary computer system in which various embodiments of the present invention may be implemented
Detailed description of the invention
[0019] Exemplary embodiments herein are provided only for illustrative purposes and various modifications will be readily apparent to persons skilled in the art. The general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the invention. The terminology and phraseology used herein is for the purpose of describing exemplary embodiments and should not be considered limiting. Thus, the present invention is to be accorded the widest scope encompassing numerous alternatives, modifications and equivalents consistent with the principles and parameters disclosed herein. For purposes of clarity, details relating to technical material that is known in the technical fields related to the invention have been briefly described or omitted so as not to unnecessarily obscure the present invention.
[0020] FIG. 1 is a block diagram of a drilling problem prediction system 100 for performing classification and prediction of critical drilling data and identifying a deviation from pre-determined outcome of activities during drilling operation in an oil and well rig, in accordance with various embodiments of the present invention. In an embodiment of the present invention, the drilling problem prediction system 100

comprises a sensor data collection unit 104, a data storage unit 106, a drilling problem prediction subsystem 102 and an alarm generation unit 120.
[0021] In an embodiment of the present invention, the data storage unit 106 of the drilling problem prediction system 100 comprises data associated with laboratory experiment results, geo-technical data, manual data and test results, which are conducted on an oil and gas drilling site. In another embodiment of the present invention, a rheometer (not shown) is configured to transmit dial readings obtained at various rotational speeds to the data storage unit 106 which stores data associated with the dial readings . In an exemplary embodiment of the present invention, the rheometer readings are manually entered into the data storage unit 106 via a graphical user interface (not shown) . In an embodiment of the present invention, the sensor data collection unit 104 comprises sensor data collected from multiple sensors arranged in the drilling rig at the oil and gas drilling site. The sensor data comprises drilling data associated with ongoing drilling activity taking place in the oil and well rig. Further, the sensor data collection unit 104 includes non-drilling data such as invalid data, sensor errors and data related to unidentified operations and activities. The sensor data collection unit 104 is configured to determine drilling parameters from the drilling data. In an exemplary embodiment of the present invention, the drilling parameters include Hookload (HL) , Weight on Bit (WOB), Rotation Per Minute (RPM), Torque, Rate of Penetration (ROP), Mud Properties, Total depth (TD), Bit depth (BD), Block Position
(BP) , inlet and outlet densities. Table 1 below illustrates units of different parameters obtained from the sensor data.

S.No. Drilling Parameters Unit
1 Hookload tons-force
2 Stand Pipe Pressure (SPP) Kgf/cma
3 Strokes Per Minute (SPM) spm
4 Weight on Bit (WOB) tons-force
5 Rotation Per Minute (RPM) rpm
6 Flowout (FO) rel %
7 Rate of Penetration (ROP) metre/hour

8 Total Depth (TD) metre
9 Bit Depth (BD) metre
10 Inlet Density (DI) g/cm3
11 Outlet Density (DO) g/cm3
Table 1: Units of different drilling parameters
[0022] In an embodiment of the present invention, the drilling parameters and the non-drilling data is transmitted from the sensor data collection unit 104 to the data storage unit 106. In an embodiment of the present invention, the data storage unit 106 is connected to a graphical user interface (GUI) (not shown) for continuous monitoring and supervision of drilling parameters associated with the drilling data.
[0023] In an embodiment of the present invention, the drilling problem prediction subsystem 102 communicates with the data storage unit 106 and with the alarm generation unit 120 via a communication channel (not shown). The communication channel
(not shown) may include, but is not limited to, a physical transmission medium, such as, a wire, or a logical connection over a multiplexed medium, such as, a radio channel in telecommunications and computer networking. The examples of radio channel in telecommunications and computer networking may include, but are not limited to, a local area network (LAN) , a metropolitan area network (MAN) and a wide area network (WAN) . In various embodiments of the present invention, the drilling problem prediction subsystem 102 processes the drilling parameters by classifying the drilling parameters and predicting data. In various embodiments of the present invention, the drilling problem prediction subsystem 102 performs classification and prediction functionalities in real time and invokes the alarm generation unit 120 to generate an alarm to alert drillers of complications during the drilling activity in the event of identifying a deviation from pre-determined outcome of drilling activities. In another embodiment, a notification is sent to the driller to alert the driller of complications.

[0024] In an embodiment of the present invention, the drilling problem prediction subsystem 102 comprises a drilling problem prediction engine 114, a processor 116 and a memory 118. The drilling problem prediction engine 114 comprises a database 108, a rule based classification unit 110, a random forest classification unit 112 and a prediction unit 122. In an exemplary embodiment of the present invention, the drilling problem prediction engine 114 may be implemented in a cloud computing architecture in which data, applications, services, and other resources are stored and delivered through shared data-centers. The various units (110,112 and 122) of the drilling problem prediction engine 114 work in conjunction with each other and the various units of the drilling problem prediction engine 114 are operated via the processor 116 specifically programmed to execute instructions stored in the memory 118.
[0025] In an embodiment of the present invention, the database 108 of the drilling problem prediction subsystem 102 stores the data received from the data storage unit 106. In an embodiment of the present invention, the database 108 is linked to a graphical user interface (GUI) (not shown). In an embodiment of the present invention, the GUI may accept inputs from users. In an embodiment of the present invention, the rule based classification unit 110 of the drilling problem prediction subsystem 102 is configured to pre-process the drilling parameters by removing errors from the drilling parameters received from the database 108. In an embodiment of the present invention, the rule based classification unit 110 is configured to apply one or more fuzzy rule based classifiers on the drilling parameters received from database 108 to obtain classified drilling parameters. The drilling parameters are classified by assigning a class label to the drilling parameters that correspond to the ongoing drilling activity at the drilling rig.
[0026] In an embodiment of the present invention, the rule based classification unit 110 uses fuzzy rule based classifiers to create a plurality of fuzzy sets for assigning class labels to the drilling parameters. In an exemplary embodiment of the present invention, the drilling parameters include, but are not

limited to, Hookload (HL), Flowout (FO), Total Depth (TD) and Bit Depth (BD) . The fuzzy sets are created by processing the drilling parameters using the non-drilling data and data stored in the database 108. In an exemplary embodiment of the present invention, the drilling parameters are divided into fuzzy sets based on expert knowledge and statistical analysis of the drilling parameters. In an example, Table 2 illustrates the fuzzy sets that are created from the drilling parameters.

S.No Drilling Parameters Fuzzy Sets
1 Hookload (HL) High HL, Medium HM, Low HL
2 Flowout (FO) High FO, Medium FO, Low FO, Very Low FO
3 Total Depth (TD) High , Medium and Very Low difference of TD and BD
4 Bit Depth (BD) High , Medium and Very Low difference of TD and BD
Table 2: Fuzzy Sets created from the drilling parameters
[0027] In an exemplary embodiment of the present invention, the fuzzy sets with a gaussian membership function are used to create fuzzy rules for the fuzzy rule based classifiers. The values of the drilling parameters of the gaussian membership function of the fuzzy sets are computed using subtractive clustering technique. The rule based classification unit 110 is configured to train the fuzzy rule based classifiers based on the created fuzzy rules. The rule based classification unit 110 applies the trained fuzzy rule based classifiers on the drilling parameters for classifying the drilling parameters.
[0028] In an exemplary embodiment of the present invention, the fuzzy rules are created based on a Takagi-Sugeno-Kang (TSK) rule in the following form:
Rule1: IF Xi is Ai1 AND AND xn is An1 THEN
ajlXJ AND AND y\n = ) ajmxj
j = 0 *—'j = 0
where xi, j = 1,2, ..., n is the j th input variable, Ai1 is an

antecedent fuzzy set of rule i, yki, k = 1,2, ... , m in the output rule i for each class and aij k is the jth consequence parameter of the output k of the rule i. In an exemplary embodiment of the present invention, the fuzzy rule based classifiers are a collection of IF-THEN rules.
[0029] In an exemplary embodiment of the present invention, the drilling parameters are classified by assigning class labels into one or more predefined classes viz. ^drilling' and ''tripping' activity. The classification of the drilling parameters into ^drilling' and ''tripping' activity is illustrated in Table 3, in accordance with an exemplary embodiment of the present invention. In an embodiment of the present invention, the rule based classification unit 110 applies the trained fuzzy rule based classifiers on the drilling parameters for classifying the drilling parameters by assigning class labels after normalization of firing strength of all the fuzzy rules.

S. No. Hookload Flowout Difference of Total Depth and Bit depth Class Label
1 High Low Very Low Drilling
2 High Medium Very Low Drilling
3 High High Very Low Drilling
4 Medium Low Very Low Drilling
5 Medium Medium Very Low Drilling
6 Medium Low Very Low Drilling
7 High Very Low High Tripping
8 High Very Low Very High Tripping
9 Medium Very Low High Tripping
10 Medium Very Low Very High Tripping
11 Low Very Low High Tripping
12 Low Very Low Very High Tripping
Table 3: Classification into drilling and tripping activities
[0030] In an exemplary embodiment of the present invention, the fuzzy rules based classifier may include, but is not limited to, a rule viz. 'If hook load is low and RPM is low and bit depth is low then class label is ^drilling'. In another embodiment of the present invention, a plurality of fuzzy rule based

classifiers are considered and an aggregate output of all the fuzzy rule based classifiers are considered for classifying the drilling parameters into one or more predefined classes.
[0031] In an embodiment of the present invention, the random forest classification unit 112 of the drilling problem prediction engine 114 fetches the classified drilling parameters from the rule based classification unit 110 and further sub-classifies the classified drilling parameters. In an embodiment of the present invention, the random forest classification unit 112 is configured to create a subset of classified drilling parameters and select best parameter amongst the classified drilling parameter. The random forest classification unit 112 then creates a plurality of decision trees by applying a random forest classifier to perform sub-classification of the selected best parameters from amongst the classified drilling parameter. The random forest classification unit 112 is configured to assign an activity tag to the selected classified drilling parameters corresponding to the drilling activity carried out in the drilling rig during the sub-classification process. In an embodiment of the present invention, the activity tags are sub¬class labels that represent specific attributes associated with the ongoing activities. Examples of attributes associated with the ongoing activities include, but are not limited to, 'drilling with rotation', 'drilling without rotation', 'rotation on bottom', 'rotation off bottom' 'tripping with rotation' and 'tripping without rotation', 'circulation' and 'reciprocation'.
[0032] In operation, in an exemplary embodiment of the present invention, the random forest classifier is an ensemble model that ensembles decisions of various decision trees where each node of the decision tree represents a value of the classified drilling parameters. In an exemplary embodiment of the present invention, a predefined number of decision trees are generated during training of the random forest classifier. In an exemplary embodiment of the present invention, 70 decision trees are generated during training of the random forest classifier. The random forest based classifier is trained by learning weights

of each of the decision trees. The nodes of the decision trees are further divided using a splitting criterion for computation of a gini index. The set of classified drilling parameters with highest value of the gini index (best parameters) is selected to form the decision trees which are used for sub-classification of the selected classified drilling parameters. Each leaf node of the decision trees provides sub-class labels for the selected classified drilling parameters. Weighted sum of the decision trees which are formed is finally used to sub-classify the selected classified drilling parameters by assigning sub-class labels representing specific attributes of activities to the classified drilling parameters.
[0033] In an embodiment of the present invention, the random forest classification unit 112 analyzes data associated with the sub-classified drilling parameters with respect to a pre-defined data. The pre-defined data includes data representing desired outcome of the specific attributes associated with activities at the drilling site. In an embodiment of the present invention, the random forest classification unit 112 transfers the sub-classified drilling parameters to the database 108 for storage.
[0034] In an embodiment of the present invention, the prediction unit 122 of the drilling problem prediction subsystem 102 fetches drilling parameters and sub-classified drilling parameters from the database 108 and predicts torque and drag data of the ongoing drilling activity at the drilling rig. In an exemplary embodiment of the present invention, the prediction unit 122 predicts torque and drag data along the wellbore at the oil and gas well rig using mathematical model such as a 3D friction model. In an exemplary embodiment of the present invention, the prediction unit 122 performs a three dimensional mathematical modelling to determine the torque and drag data by modelling a sequence of torque and drag logs based on frictional forces acting along the wellbore at the drilling rig by considering change in sheave efficiency, global friction factor, and one or more factors affecting the hookload measurement associated with quantifiable

parameters of block and tackle system in the drilling rig. In an embodiment of the present invention, the predicted torque and drag data is stored in the database 108.
[0035] In another exemplary embodiment of the present invention, the prediction unit 122 performs a three dimensional mathematical modelling to determine prediction data associated with the torque and drag by modelling the determined sequence of torque and drag logs based on determination of a rig uncertainty compensation (RUC) factor using load variation due to change in position of block factors affecting the hookload measurement due to quantifiable and non-quantifiable parameters associated with a block and tackle system in the drilling rig. In an exemplary embodiment of the present invention, the prediction unit 122 is configured to determine a neutral point of drill string in the drilling rig using tuned parameters. During drilling activity, the neutral point may vary due to changes in frictional drag, pressure losses, weight on the bit and other relevant factors. Assuming that sheave efficiency, global friction factor and RUC factor does not change, incremental tensile load from the neutral point is cumulatively summed up to predict the hookload at deadline. Similarly, the prediction unit 122 predicts torque acting along the drilling rig using tuned parameters.
[0036] In an embodiment of the present invention, the alarm generation unit 120 fetches the predicted torque and drag data from the prediction unit 122 and compares the predicted torque and drag data with measured torque and drag data. In the event predicted torque and drag data crosses a permissible threshold limit, the prediction unit 122 identifies a deviation and invokes the alarm generation unit 120. The alarm generation unit 120 then raises an alarm which is used to alert drillers of complication during the drilling activity.
[0037] In another embodiment of the present invention, the prediction unit 122 fetches the drilling parameters and sub-classified drilling parameters from the database 108 and predicts transport capacity and flow hydraulics data of drilling fluid in the wellbore. In an embodiment of the

present invention, the transport capacity and the flow hydraulics data is predicted for monitoring and optimization of the drilling activity. In an exemplary embodiment of the present invention, the prediction unit 122 is configured to predict data associated with flow hydraulics and transport capacity by determining a flow behaviour index (n) and a flow consistence index (k) of drilling mud from dial reading of the rheometer stored in the database 108. In an embodiment of the present invention, the predicted transport capacity and flow hydraulics is used for optimization of the drilling activity by determining optimal flow rates, drill string Rotation per Minute (RPM), Rate of Penetration (ROP) and mud property while maintaining an Equivalent Circulation Density (ECD) in its permissible operating window.
[0038] In another exemplary embodiment of the present invention, the prediction unit 122 is configured to predict the transport capacity data based on a cutting concentration variation in an annulus region along depth of the wellbore using the drilling parameter with respect to time as a function of rate of penetration, drill bit diameter, drill string outside diameter and borehole diameter, mud flowrate, slip velocity and wellbore inclination. In yet another exemplary embodiment of the present invention, the prediction unit 122 is configured to predict cutting concentration, transport efficiency, annular transport time, effective density, stand pipe pressure and equivalent circulation density that are relevant for monitoring the drilling activity. Fig. 3 illustrates a dataflow diagram illustrating prediction of the cutting transport and flow hydraulics, in accordance with an exemplary embodiment of the present invention.
[0039] In an embodiment of the present invention, the alarm generation unit 120 fetches the predicted cutting transport and flow hydraulics data from the prediction unit 122 and compares the predicted cutting transport and flow hydraulics data with a measured cutting transport and flow hydraulics data. In the event of the values of predicted cutting transport and flow hydraulics data crosses a permissible threshold limit, the prediction unit 122 identifies a deviation and invokes the alarm generation unit

120. The alarm generation unit 120 then raises an alarm which is used to alert drillers of complication during the drilling activity.
[0040] In an embodiment of the present invention, the prediction involves monitoring of the data, validating a model and optimizing of the drilling parameters in real time. In an exemplary embodiment of the present invention, the prediction unit 122 predicts transport capacity and flow hydraulics of a drilling fluid during the ongoing drilling activity in the well-site based on a data-driven model and/or knowledge-driven model in real-time to monitor and optimize the drilling activity. In an embodiment of the present invention, a plurality of empirical constants defined in a knowledge-driven model are calibrated within the prediction unit 122 using data stored in a database 108 in real-time to monitor and predict future behaviour of the wellbore.
[0041] In an embodiment of the present invention, the prediction unit 122 fetches the drilling parameters and the sub-classified drilling parameters from the database 108 and determines a D-exponent parameter data that is used to predict a pressure gradient while drilling in the wellbore . In an exemplary embodiment of the present invention, the drilling parameters used to determine the D-exponent parameter are drilling rate (ROP), rotary speed, weight on bit, bit diameter and mud weight along with the sub-classified drilling parameters. The D-exponent parameter is used to find the possibility of kick and mud loss possibility by calculating the formation pressure. The D exponent parameter is determined by using the formulae:
D-exponent = log ( (ROP/(C1*RPM) ) )/log ( (C2*WOB)/(DBD*C3) )
where ROP - Rate of penetration, RPM - rotations per minute, WOB - Weight on Bit, DBD - Drill bit diameter, C1,C2,C3 are constants.
[0042] In an embodiment of the present invention, the alarm generation unit 120 fetches the determined D-exponent parameter data from the prediction unit (122) and compares the determined D-exponent parameter data with one or more

threshold values for identifying kick and lost circulation by determining a formation pressure. In the event the determined D-exponent parameter data exceeds the threshold value, an alarm is generated by the alarm generation unit (120) to alert the drillers.
[0043] In another embodiment of the present invention, the prediction unit 122 fetches the drilling parameters and the sub-classified drilling parameters from the database 108 and processes the drilling parameters to determine energy consumption data and carbon emission data during the drilling activity of the drilling rig. The determined values may be used to analyze the possibility of reducing energy consumption and carbon emission. In an exemplary embodiment of the present invention, the prediction unit 122 determines energy consumptions in the form of electrical, hydraulic and mechanical energy. In an exemplary embodiment of the present invention, thermal energy changes may be considered based on the drilling rig and formation type. In another exemplary embodiment of the present invention, the energy consumption of an equipment in the drilling rig may be determined based on direct parameters such as power rating and usage time and indirect parameters obtained based on work done by the equipment and its efficiency. In yet another embodiment of the present invention, an activity wise and equipment wise energy consumption with respect to different sections of the wellbore may be determined which may be used for optimizing drilling strategy for energy saving and other benefits. . In an embodiment of the present invention, the prediction unit 122 optimizes drilling parameters for efficient operation.
[0044] In an embodiment of the present invention, the energy consumption is determined for different drilling actives such as drilling (with and without rotation), tripping (with and without rotation). In an embodiment of the present invention, the energy consumption is determined in all the important equipment of the drilling rig and quantified with respect to different drilling activities. In another embodiment of the present invention, the energy consumption is compared at different depth and phase of drilling. In another embodiment of the present invention, major energy

consuming activities and change in their trend with respect to different depth and phase of a wellbore is determined.
[0045] In an embodiment of the present invention, the alarm generation unit 120 fetches the energy consumption data and carbon emission data from the prediction unit (122) and compares the determined energy consumption data and carbon emission data with one or more threshold energy consumption value and carbon emission value. In the event the determined energy consumption data and carbon emission data exceeds the threshold value, an alarm is generated by the alarm generation unit (120) to alert the drillers.
[0046] In an embodiment of the present invention, the prediction unit 122 fetches sub-classified drilling parameters to determine one or more key performance indicators (KPI) that aids high level decisions. In an exemplary embodiment of the present invention, before starting the drilling, based on geological survey, previous rig well-site drilling history and expert knowledge, threshold values for most of the key performance indicators is pre-determined. A violation and percentage deviation of KPI's from threshold values may be used for decision making. Examples of KPI's include drilling time, standpipe pressure, drill bit life, ROP, Weight on Bit (WOB) , KPI for pump discharge rate, mud weight, non-drilling time ratio, over balanced drilling percentage. In an embodiment of the present invention, drilling parameters used for determining KPI includes, but are not limited to, standpipe pressure, drill bit life, ROP, WOB, mud weight and non-drilling time ratio.
[0047] In an embodiment of the present invention, the alarm generation unit 120 fetches the determined key performance indicators (KPI) data from the prediction unit 122 and compares the determined key performance indicators
(KPI) data with one or more threshold value. In an embodiment of the present invention, threshold value is ascertained for each KPIs based on well design, wellbore lithological properties, equipment and tool specifications. In the event the determined key performance indicators (KPI) exceeds the threshold value, an alarm is generated by the alarm generation unit 120 to alert the drillers that the drilling

rig is underperforming with respect to a drilling plan for optimizing the drilling parameters. The determined KPI also aids in critical parameter benchmarking for evaluating performance of the drilling rig. In an embodiment of the present invention, the optimization can be done with multiple objectives like energy minimization, increasing the cutting transport and increasing efficiency of the drilling rig. Further, optimization can be done at multilayers with different objectives and execution time.
[0048] FIG. 2 is an exemplary flowchart of a drilling problem prediction system illustrating classification and prediction of drilling parameters, in accordance with an embodiment of the present invention.
[0049] At step 202, drilling parameters are determined from sensor data. In an embodiment of the present invention, the sensor data is collected from multiple sensors arranged in the drilling rig at an oil and gas drilling site. Further, non-drilling data such as invalid data, sensor errors and data related to unidentified operations and activities are also collected. A plurality of drilling parameters associated with the ongoing activity taking place in the oil and well rig are determined from the sensor data. In an exemplary embodiment of the present invention, the drilling parameters include, but are not limited to, Hookload (HL), Weight on Bit (WOB), Rotation Per Minute (RPM), Torque, Rate of Penetration (ROP), Mud Properties, Total depth (TD), Bit depth (BD), Block Position (BP), Inlet and Outlet densities.
[0050] At step 204, the drilling parameters are classified by applying fuzzy rule based classifiers. In an embodiment of the present invention, the drilling parameters are classified by applying a plurality of fuzzy rules using fuzzy rule based classifiers on the drilling parameters. In an exemplary embodiment of the present invention, the drilling parameters are classified using a class label that correspond to ongoing drilling activity taking place in the drilling rig. In an embodiment of the present invention, the drilling parameters such as Hookload (HL) , Flowout (FO) , Total Depth (TD) and Bit Depth

(BD) are used for assigning the class label to the drilling parameters. For instance, the drilling parameters are assigned the class labels of 'drilling activity' and 'tripping activity'.
[0051] In an exemplary embodiment of the present invention, the fuzzy rule-based classifier uses a Takagi-Sugeno-Kang (TSK) rule in the following form:
Rule1: IF x± is Ai1 AND AND xn is An1 THEN
ajlXJ AND AND ym = > ajmXJ
j = 0 *—'j = 0
where xi, j = 1,2, ..., n is the j th input variable, Ai1 is an antecedent fuzzy set of rule i, yki, k = 1,2, ... , m in the output rule i for each class and ai j k is the jth consequence parameter of the output k of the rule i. In an exemplary embodiment of the present invention, the fuzzy rule based classifier is a collection of IF-THEN rules.
[0052] In an embodiment of the present invention, the drilling parameters are used to create a plurality of fuzzy sets for assigning class labels to the drilling parameters. The fuzzy sets are used for training the fuzzy rule based classifier. In an exemplary embodiment of the present invention, the fuzzy sets with a gaussian membership function are used to create fuzzy rules of the fuzzy rule based classifier. In an embodiment of the present invention, the fuzzy rule based classifier classifies the drilling parameters by assigning class labels after normalization of firing strength of all the fuzzy rules.
[0053] In an embodiment of the present invention, the fuzzy rules are designed from recommendations/suggestions given by drilling experts. In an exemplary embodiment of the present invention, the fuzzy rules include a rule 'If hook load is low and RPM is low and bit depth is low then class is 'drilling activity'. In another embodiment of the present invention, a plurality of fuzzy rules are taken and an aggregate output of

all the fuzzy rules are taken for classifying drilling parameters as per predefined classes.
[0054] At step 206, the classified drilling parameters are sub-classified by applying random forest based classifiers. In an embodiment of the present invention, a plurality of trees in the form of a random forest classifier are created to perform sub-classification of the classified drilling parameters. A subset of classified drilling parameters is created and a best parameters is chosen amongst the classified drilling parameters. An activity tag is assigned to the classified drilling parameters corresponding to the drilling activity going on in the drilling rig. In an exemplary embodiment of the present invention, the activity tags are class labels assigned to classified drilling activities. Examples of activity tags include, but are not limited to, 'drilling with rotation', 'drilling without rotation' , 'rotation on bottom', 'tripping with rotation' and 'tripping without rotation'. In an embodiment of the present invention, the random forest classifier uses drilling parameter like hook-load, strokes per minute, total depth, bit depth and Rotation Per Minute (RPM) to form parameters to train the random forest classifier.
[0055] In an exemplary embodiment of the present invention, the drilling parameters classified into drilling or tripping activity is forwarded to the random forest classifier where the random forest classifier uses drilling parameters such as strokes per minute (SPM), rotation per minute (RPM), total depth (TD), bit depth (BD) and hook load (HL) parameter to form a new parameter set. The new set parameter set is used by the random forest based classifier to identify a detailed drilling activity such as, but is not limited to, 'drilling with rotation', 'drilling without rotation', 'circulation', and 'reciprocation' associated with the drilling data.
[0056] In an embodiment of the present invention, the random forest classifier is an ensemble model that ensembles decisions of various decision trees where each node of the tree represents

value of the parameter. The nodes are further divided using a splitting criterion that requires computation of a gini index and the highest value of the gini index is selected to split the node. In an embodiment of the present invention, a predefined number of decision trees are generated during training of the random forest classifier. In an exemplary embodiment of the present invention, 70 decision threes are used to train the random forest classifier. In an embodiment of the present invention, the training procedure of the random forest based classifier aims to learn weights for each of the decision trees. The weighted sum of decision of each classifier is finally used to sub-classify the drilling activity of the drilling parameter.
[0057] At step 208, data is predicted and determined using the sub-classified drilling parameters. In an embodiment of the present invention, the sub-classified drilling parameters are fetched and torque and drag data is predicted related to the ongoing drilling activity at the drilling rig. The torque and drag data is predicted along the wellbore at the oil and well rig using mathematical modelling such as a 3D friction model. In an exemplary embodiment of the present invention, the mathematical modelling is performed based on frictional forces acting along the wellbore at the drilling rig under a sheave efficiency and a global friction factor.
[0058] In an embodiment of the present invention, a neutral point of drill string is determined in the drilling rig using tuned parameters. During drilling activity, the neutral point may vary due to changes in frictional drag, pressure losses, weight on the bit and other relevant factors. Similarly, the torque is predicted that acts along the wellbore drilling rig using tuned parameters.
[0059] In an embodiment of the present invention, the predicted torque and drag data is compared with measured torque and drag. In the event of the predicted torque and drag data crossing a permissible threshold limit, an alarm is raised at step 210 for alerting driller of complications during the drilling activity.

[0060] In another embodiment of the present invention, the drilling parameters and the sub-classified drilling parameters are fetched for predicting transport capacity data and flow hydraulics data of drilling fluid circulating in the drilling rig. The transport capacity and the flow hydraulics data are predicted for monitoring and optimization of the drilling activity. In an exemplary embodiment of the present invention, flow hydraulics data and transport capacity data are predicted using rheological properties of the drilling mud. The rheology property is a strong function of the fluid flow behaviour index (n) and flow consistence index (k) which are determined from dial reading data of the rheometer. In yet another embodiment of the present invention, cutting concentration variation is predicted in an annulus region along depth of the drilling rig using the drilling parameters such as rate of penetration, drill bit diameter, drill string outside diameter and borehole diameter, mud flowrate, slip velocity and wellbore inclination. In an embodiment of the present invention, cutting concentration, transport efficiency, annular transport time, effective density, stand pipe pressure, hydrostatic pressure and equivalent circulation density are predicted that are relevant for monitoring the drilling activity. Fig 3 illustrates the prediction of cutting transport and flow hydraulics using the drilling parameters, in accordance with an embodiment of the present invention. Fig. 4 illustrates optimization of the cutting transport, in accordance with an embodiment of the present invention.
[0061] In an embodiment of the present invention, the predicted cutting transport and flow hydraulics data is compared with measured cutting transport and flow hydraulics data. In the event of predicted cutting transport and flow hydraulics data crossing a permissible threshold limit, an alarm is raised at step 210 for alerting driller of complications during the drilling activity.
[0062] In an embodiment of the present invention, the drilling parameters and sub-classified drilling parameters are fetched and a D-exponent parameter data is determined that is used to predict a pressure gradient while drilling

in the drilling rig. In an exemplary embodiment of the present invention, the drilling parameters used to determine the D-exponent parameter are drilling rate (ROP), rotary speed, weight on bit, bit diameter and mud weight. The predicted D-exponent parameter is fetched and compared with threshold value for identifying kick and lost circulation. In the event, the determined D-exponent parameter data exceeds the threshold value, an alarm is generated to alert the drillers at step 208. Fig. 5 illustrates flow diagram of the calculation of the D-exponent variable and subsequent triggering of the alarm, in accordance with an embodiment of the present invention.
[0063] In another embodiment of the present invention, the drilling parameters and the sub-classified drilling parameters are fetched and energy consumption data and carbon emission data is determined during the drilling activity of the drilling rig. The determined data may be used to analyze the possibility of reducing energy consumption and carbon emission. In an exemplary embodiment of the present invention, determination of energy consumption across various operations and equipment are carried out. In another exemplary embodiment of the present invention, energy consumptions in the form of electrical, hydraulic and mechanical energy, neglecting thermal energy changes is carried out. In yet another exemplary embodiment of the present invention, the energy consumption of an equipment in the drilling rig may be determined based on direct parameters such as power rating and usage time and indirect parameters obtained based on work done by the equipment and its efficiency. In another embodiment of the present invention, an activity wise and equipment wise energy consumption with respect to different sections of the wellbore may be determined which may be used for optimizing drilling strategy for energy saving and other benefits. Fig 6 illustrates determination of the energy consumption data and carbon emission data, in accordance with an embodiment of the present invention. In the event of determined energy consumption and carbon emission data crossing a permissible threshold limit, an alarm is raised at step 210 for alerting driller of complications during the drilling activity.

[0064] In another embodiment of the present invention, a key parameter indicator is determined based on the drilling parameters and the sub-classified drilling parameters that aids in making high level decisions with respect to a drilling plan. Based on the determined key performance indicator an alarm is generated at step 210 to alert the driller that the drilling rig is underperforming with respect to a drilling plan. The determined KPI also aids in critical parameter benchmarking for evaluating performance of the drilling rig. In an exemplary embodiment of the present invention, before starting the drilling, based on geological survey, previous rig well-site drilling history and expert knowledge, benchmarking values for most of the key performance indicators is predetermined. A violation and percentage deviation of KPI' s from benchmarking data may be used for decision making. Examples of KPI's include drilling time, standpipe pressure, drill bit life, ROP, weight on bit (WOB), KPI for pump discharge rate, mud weight, non-drilling time ratio, over balanced drilling percentage.
[0065] FIG. 7 illustrates an exemplary computer system in which various embodiments of the present invention may be implemented. The computer system 702 comprises a processor 704 and a memory 706. The processor 704 executes program instructions and is a real processor. The computer system 702 is not intended to suggest any limitation as to scope of use or functionality of described embodiments. For example, the computer system 702 may include, but not limited to, a programmed microprocessor, a micro-controller, a peripheral integrated circuit element, and other devices or arrangements of devices that are capable of implementing the steps that constitute the method of the present invention. In an embodiment of the present invention, the memory 706 may store software for implementing various embodiments of the present invention. The computer system 702 may have additional components. For example, the computer system 702 includes one or more communication channels 708, one or more input devices 710, one or more output devices 712, and storage 714. An interconnection mechanism (not shown) such as a bus, controller, or network, interconnects the components of the

computer system 702. In various embodiments of the present invention, operating system software (not shown) provides an operating environment for various software executing in the computer system 702, and manages different functionalities of the components of the computer system 702.
[0066] The communication channel(s) 708 allow communication over a communication medium to various other computing entities. The communication medium provides information such as program instructions, or other data in a communication media. The communication media includes, but not limited to, wired or wireless methodologies implemented with an electrical, optical, RF, infrared, acoustic, microwave, Bluetooth or other transmission media.
[0067] The input device(s) 710 may include, but not limited to, a keyboard, mouse, pen, joystick, trackball, a voice device, a scanning device, touch screen or any another device that is capable of providing input to the computer system 702. In an embodiment of the present invention, the input device(s) 710 may be a sound card or similar device that accepts audio input in analog or digital form. The output device(s) 712 may include, but not limited to, a user interface on CRT or LCD, printer, speaker, CD/DVD writer, or any other device that provides output from the computer system 702.
[0068] The storage 714 may include, but not limited to, magnetic disks, magnetic tapes, CD-ROMs, CD-RWs, DVDs, flash drives or any other medium which can be used to store information and can be accessed by the computer system 702. In various embodiments of the present invention, the storage 714 contains program instructions for implementing the described embodiments.
[0069] The present invention may suitably be embodied as a computer program product for use with the computer system 802. The method described herein is typically implemented as a computer program product, comprising a set of program instructions which is executed by the computer system 702 or any

other similar device. The set of program instructions may be a series of computer readable codes stored on a tangible medium, such as a computer readable storage medium (storage 714), for example, diskette, CD-ROM, ROM, flash drives or hard disk, or transmittable to the computer system 702, via a modem or other interface device, over either a tangible medium, including but not limited to optical or analogue communications channel(s) 708. The implementation of the invention as a computer program product may be in an intangible form using wireless techniques, including but not limited to microwave, infrared, Bluetooth or other transmission techniques. These instructions can be preloaded into a system or recorded on a storage medium such as a CD-ROM, or made available for downloading over a network such as the internet or a mobile telephone network. The series of computer readable instructions may embody all or part of the functionality previously described herein.
[0070] The present invention may be implemented in numerous ways including as a system, a method, or a computer program product such as a computer readable storage medium or a computer network wherein programming instructions are communicated from a remote location.
[0071] While the exemplary embodiments of the present invention are described and illustrated herein, it will be appreciated that they are merely illustrative. It will be understood by those skilled in the art that various modifications in form and detail may be made therein without departing from or offending the scope of the invention.

We claim:
1) A system (102) for predicting problems during drilling operation, wherein the system comprises:
a memory (118) storing program instructions;
a processor (116) executing program instructions stored in the memory; and
a drilling problem prediction engine (114) executed by the processor and configured to:
determine drilling parameters from sensor data collected from multiple sensors, wherein the drilling parameters are associated with a drilling activity;
classify the drilling parameters using one or more fuzzy rule based classifiers, wherein the classification is performed by assigning a class label to each of the determined drilling parameters to obtain the classified parameters, the class label representing pre-defined classes of activities during drilling operation;
sub-classify the classified parameters using a random forest classifier, wherein the sub-classification is performed by assigning a sub-class label to the classified drilling parameters to obtain the sub-classified parameters, the sub-class label representing attributes associated with the pre¬defined classes of activities during the drilling operation;
predict data based on the drilling parameters and the sub-classified drilling parameters; and
generate an alarm by identifying a deviation of the predicted data with a threshold value.
2) The system as claimed in claim 1, wherein the drilling problem prediction engine (114) comprises a rule based classification unit (110) configured to perform classification of the drilling parameters by:
creating a plurality of fuzzy sets for the drilling parameters by processing the drilling parameters using non-drilling data including invalid data, sensor errors and data related to unidentified operations and activities and data associated with laboratory experiment results, geo-technical

data, manual data and test results which are conducted on an oil and gas drilling site; and
creating fuzzy rules for the fuzzy rule based classifiers, wherein the fuzzy sets with a gaussian membership function are used to create the fuzzy rules.
3) The system as claimed in claim 1, wherein the drilling problem prediction engine (114) comprises a rule based classification unit (110) configured to performs the classification of parameters using a fuzzy rule based classifier that uses plurality of fuzzy rule created based on a Takagi-Sugeno-Kang rule.
4) The system as claimed in claim 1, wherein the drilling problem prediction engine (114) comprises a random forest classification unit (120) configured to perform sub-classification by:
creating a subset of the classified drilling parameters and selecting best parameter amongst the classified drilling parameter;
creating a plurality of decision trees by applying the random forest classifier to perform sub-classification of the selected best parameter amongst the classified drilling parameter; and
sub-classifying the selected best parameters amongst the classified drilling parameters based on a weighted sum of the decision trees.
5) The system as claimed in claim 1, wherein the drilling parameters comprise hookload, weight on bit, stand pipe pressure, strokes per minute, rotation per minute, torque, rate of penetration, mud properties, total depth, bit depth, block position, inlet and outlet densities.
6) The system as claimed in claim 1, wherein the class-labels representing the pre-defined classes of activities during drilling operation include ^drilling activity' and ''tripping activity' .

7) The system as claimed in claim 1, wherein the sub-class labels representing the attributes associated with the pre¬defined classes of activities during the drilling operation include 'drilling with rotation', 'drilling without rotation', 'rotation on bottom', 'rotation off bottom', 'tripping with rotation' and 'tripping without rotation', 'circulation' and 'reciprocation'.
8) The system as claimed in claim 1, wherein the drilling problem prediction engine (114)comprises a prediction unit
(122) configured to:
predict data associated with torque and drag related to the drilling activity, wherein the data associated with torque and drag is predicted by determining a sequence of torque and drag logs; and
generate an alarm by identifying a deviation based on a comparison of the predicted torque and drag data with a measured torque and drag data.
9) The system as claimed in claim 8, wherein the prediction
unit (122) is configured to:
determine the data associated with the torque and drag by modelling the determined sequence of torque and drag logs based on a plurality of parameters including frictional forces acting along a well-bore under a sheave efficiency, a global friction factor and one or more factors affecting a hookload measurement associated with quantifiable and non-quantifiable parameters of a block and tackle system in a drilling rig.
10) The system as claimed in claim 9, wherein the prediction
unit (122) is configured to:
determine the data associated with the torque and drag by modelling the determined sequence of torque and drag logs based on a determination of a Rig Uncertainty Compensation (RUC) factor.

11) The system as claimed in claim 10, wherein the
quantifiable parameters include tuned parameters which are
utilized to determine a neutral point of a drill string, wherein
the neutral point may change due to changes in frictional drag,
pressure losses, weight on the drill string and other relevant
forces, the neutral point is used to predict the hookload.
12) The system as claimed in claim 10, wherein the
prediction unit (122) is configured to:
predict data associated with transport capacity and flow hydraulics of a drilling fluid during the drilling activity, wherein the data associated with transport capacity and flow hydraulics is predicted by determining a flow behaviour index and a flow consistence index of a drilling mud from dial reading of a rheometer at various rotational speeds; and
generate an alarm by identifying a deviation in the predicted data from a predetermined data for optimizing the drilling parameters.
13) The system as claimed in claim 12, wherein the
prediction unit (122) is configured to:
determine the transport capacity by determining cutting transport mechanism based on a cutting concentration variation in an annulus along the depth of a wellbore with respect to time, wherein the transport capacity is a function of drilling parameters that includes rate of penetration, drill bit diameter, drill string outside diameter, borehole diameter, mud flowrate, slip velocity and a wellbore inclination.
14) The system as claimed in claim 13, wherein the prediction
unit (122) is configured to:
predict the data associated with transport capacity and flow hydraulics based on a data-driven model, wherein a pressure drop is computed as the prediction data based on variables defined in the data driven model at various sections of a wellbore.

15) The system as claimed in claim 14, wherein the prediction
unit (122) is configured to:
predict the data associated with transport capacity and flow hydraulics based on a knowledge-driven model, wherein a plurality of empirical constants defined in the knowledge-driven model are calibrated in a model validation unit using data stored in a database in real-time to monitor and predict future behaviour of a wellbore.
16) The system as claimed in claim 15, wherein the prediction
unit (122) is configured to:
optimize the drilling operation based on the determined transport capacity and flow hydraulics, the optimization is performed by determining optimal flow rates, drill string, revolutions per minute (RPM), rate of penetration (ROP) and mud property while maintaining an equivalent circulation density (ECD) in its permissible operating window.
17) The system as claimed in claim 8, wherein the prediction
unit (122) is configured to:
determine a D-exponent parameter associated with the classified parameters, wherein the D-exponent parameter facilitates predicting a pressure gradient during a drilling activity in a drilling rig; and
generate an alarm by identifying a deviation of the D-exponent parameter over one or more threshold values for optimizing the drilling parameters.
18) The system as claimed in claim 17, wherein the drilling parameters used to determine the D-exponent parameter compriseing drilling rate, ROP, rotary speed, weight on bit, bit diameter and mud weight.
19) The system as claimed in claim 18, wherein the identification of the deviation between the D-exponent parameter and the one or more thresholds includes identifying kick and lost circulation by determining a formation pressure of a wellbore.

20) The system as claimed in claim 8, wherein the prediction
unit (122) is configured to:
determine energy consumption and carbon emission associated with the classified drilling parameters; and
reduce the energy consumption and the carbon emission for optimizing the parameters associated with the drilling activity.
21) The system as claimed in claim 20, wherein the prediction unit (122) is configured to determine the energy consumption based on direct parameters including power rating and usage time and indirect parameters associated with performance and efficiency of one or more equipment with respect to different sections of a wellbore.
22) The system as claimed in claim 21, wherein the prediction unit (122) is configured to determine energy consumption in the form of electrical, hydraulic, mechanical and/or thermal energy, and wherein the energy consumption is determined in one or more equipment of a drilling rig and quantified with respect to different drilling activities, and wherein the energy consumption is compared at different depth and phase of drilling, and wherein change in energy consumption activities and trend with respect to different depth and phase of the wellbore is determined.
23) The system as claimed in claim 22, wherein the prediction unit (122) is configured to:
determine one or more key performance indicators (KPIs) associated with the classified drilling parameters; and
generate an alarm by identifying a deviation of the determined KPIs in relation to pre-stored threshold values for optimizing the drilling parameters.
24) The system as claimed in claim 23, wherein the KPIs
include drilling time, standpipe pressure, drill bit life,
ROP, weight on bit, KPI for pump discharge rate, mud weight,

non-drilling time ratio, and over balanced drilling percentage .
25) The system as claimed in claim 24, wherein the threshold values for the KPIs are ascertained based on well design, wellbore lithological properties, equipment and tool specifications.
26) A method for predicting problems during drilling operation, the method comprising:
determining drilling parameters from sensor data collected from multiple sensors, wherein the drilling parameters are associated with a drilling activity;
classifying the drilling parameters using one or more fuzzy rule based classifiers, wherein the classification is performed by assigning a class label to each of the determined drilling parameters to obtain the classified parameters, the class label representing pre-defined classes of activities during drilling operation;
sub-classifying the classified parameters using a random forest classifier, wherein the sub-classification is performed by assigning a sub-class label to the classified drilling parameters to obtain the sub-classified parameters, the sub-class label representing attributes associated with the pre¬defined classes of activities during the drilling operation;
predicting data based on the drilling parameters and the sub-classified drilling parameters; and
generating an alarm by identifying a deviation of the predicted data with a threshold value.
27) The method as claimed in claim 26, wherein classification
of the drilling parameters is performed by:
creating a plurality of fuzzy sets for the drilling parameters by processing the drilling parameters using non-drilling data including invalid data, sensor errors and data related to unidentified operations and activities and data associated with laboratory experiment results, geo-technical

data, manual data and test results which are conducted on an oil and gas drilling site; and
creating fuzzy rules for the fuzzy rule based classifiers, wherein the fuzzy sets with a gaussian membership function are used to create the fuzzy rules.
28) The method as claimed in claim 27, wherein the fuzzy rule based classifier performs the classification of parameters using the fuzzy rule created based on a Takagi-Sugeno-Kang rule.
29) The method as claimed in claim 26, wherein the step of sub-classifying the classified parameters comprises:
creating a subset of the classified drilling parameters and selecting best parameters amongst the classified drilling parameters;
creating a plurality of decision trees by applying the random forest classifier to perform sub-classification of the selected best parameters amongst the classified drilling parameters; and
sub-classifying the selected best parameters amongst the classified drilling parameters based on a weighted sum of the decision trees.
30) The method as claimed in claim 26, wherein the drilling parameters comprise hookload, weight on bit, stand pipe pressure, strokes per minute, rotation per minute, torque, rate of penetration, mud properties, total depth, bit depth, block position, inlet and outlet densities.
31) The method as claimed in claim 26, wherein the class-labels representing the pre-defined classes of activities during drilling operation include 'drilling activity' and 'tripping activity'.

32) The method as claimed in claim 26, wherein the sub-class labels representing the attributes associated with the pre¬defined classes of activities during the drilling operation include 'drilling with rotation', 'drilling without rotation', 'rotation on bottom', 'rotation off bottom', 'tripping with rotation' and 'tripping without rotation' , 'circulation' and 'reciprocation'.
33) The method as claimed in claim 26, wherein the method comprises:
predicting data associated with torque and drag related to the drilling activity, wherein the data associated with torque and drag is predicted by determining a sequence of torque and drag logs, the determined sequence of torque and drag logs are labelled for measurements; and
generating an alarm by identifying a deviation based on a comparison of the predicted torque and drag data with a measured torque and drag data.
34) The method as claimed in claim 33, wherein the step of determining the data associated with the torque and drag comprises modelling the determined sequence of torque and drag logs based on a plurality of parameters including frictional forces acting along a wellbore under a sheave efficiency, a global friction factor, a load variation with respect to block position used in the drilling activity due to a position of a drill string at the well-site, and one or more factors affecting a hookload measurement associated with quantifiable parameters of a block and tackle system in a drilling rig.
35) The method as claimed in claim 34, wherein the step of determining the data associated with the torque and drag comprises modelling the determined sequence of torque and drag logs based on a determination of a Rig Uncertainty Compensation (RUC) factor.

36) The method as claimed in claim 35, wherein the quantifiable parameters include tuned parameters which are utilized to determine a neutral point of a drill string, wherein the neutral point may change due to changes in frictional drag, pressure losses, weight on the drill string and other relevant forces, the neutral point is used to predict the hookload.
37) The method as claimed in claim 36, wherein the method comprises:
predicting data associated with transport capacity and flow hydraulics of a drilling fluid during the drilling activity, wherein data associated with transport capacity and flow hydraulics is predicted by determining a flow behaviour index and a flow consistence index of a drilling mud from dial reading of a rheometer at various rotational speeds; and
generating an alarm by identifying a deviation in the predicted data from a predetermined data for optimizing the drilling parameters.
38) The method as claimed in claim 37, wherein determining the transport capacity comprises determining cutting transport mechanism based on a cutting concentration variation in an annulus along the depth of a wellbore of a well-site, during the drilling operation, with respect to time as a function of rate of penetration, drill bit diameter, drill string outside diameter, borehole diameter, mud flowrate, slip velocity and a wellbore inclination.
39) The method as claimed in claim 38, wherein the step of predicting the data associated with transport capacity and flow hydraulics is carried out based on a data-driven model, wherein a pressure drop is computed as the prediction data based on variables defined in the data driven model at various sections of a wellbore.
40) The method as claimed in claim 39, wherein the step of predicting the data associated with transport capacity and flow hydraulics is carried out based on a knowledge-driven model,

wherein a plurality of empirical constants defined in the knowledge-driven model are calibrated in a model validation unit using data stored in a database in real-time to monitor and predict future behaviour of a wellbore.
41) The method as claimed in claim 40, wherein the method
comprises the step of optimizing the drilling activity based
on the determined transport capacity and flow hydraulics, the
optimization is performed by determining optimal flow rates,
drill string,revolutions per minute (RPM), rate of penetration
(ROP) and mud property while maintaining an equivalent circulation density (ECD) in its permissible operating window.
42) The method as claimed in claim 26, wherein the method
comprises:
determining a D-exponent parameter associated with the classified parameters, wherein the D-exponent parameter facilitates predicting a pressure gradient during a drilling activity in a drilling rig; and
generating an alarm by identifying a deviation of the D-exponent parameter over one or more threshold values for optimizing the drilling parameters.
43) The method as claimed in claim 42, wherein the drilling parameters used to determine the D-exponent parameter are drilling rate (ROP), rotary speed, weight on bit, bit diameter and mud weight.
44) The method as claimed in claim 43, wherein the identification of the deviation between the D-exponent parameter and the one or more thresholds includes identifying kick and lost circulation by determining a formation pressure of a wellbore .
45) The method as claimed in claim 26, wherein the method comprises:
determining energy consumption and carbon emission associated with the classified drilling parameters; and

reducing the energy consumption and the carbon emission for optimizing the parameters associated with the drilling activity.
46) The method as claimed in claim 45, wherein the method
comprises determining the energy consumption based on direct
parameters including power rating and usage time and indirect
parameters associated with performance and efficiency of one
or more equipments with respect to different sections of a
well- bore.
47) The method as claimed in claim 46, wherein the method comprises determining the energy consumption in the form of electrical, hydraulic and mechanical energy, and wherein the energy consumption is determined in one or more equipments of a drilling rig and quantified with respect to different drilling activities, and wherein the energy consumption is compared at different depth and phase of drilling, and wherein change in energy consumption activities and trend with respect to different depth and phase of the wellbore is determined.
48) The method as claimed in claim 26, wherein the method comprises:
determining one or more key performance indicators
(KPIs) associated with the classified drilling parameters; and
generating an alarm by identifying a deviation of the
determined KPIs in relation to pre-stored threshold values for
optimizing the drilling parameters.
49) The method as claimed in claim 48, wherein the KPIs include drilling time, standpipe pressure, drill bit life, ROP, weight on bit, KPI for pump discharge rate, mud weight, non-drilling time ratio, and over balanced drilling percentage.
50) The method as claimed in claim 49, wherein the threshold values for the KPIs are ascertained based on well

design, wellbore lithological properties, equipment and tool specifications.