TECHNICAL FIELD

[0001] The invention relates to a method for fault detection in plunger lift operation in hydrocarbon wells.

BACKGROUND

[0002] A well consists of concentric tubes: an inner portion called tubing and an outer portion called casing. Liquid loading is a phenomenon occurring in gas producing wells where liquids (oil or water) accumulate at the bottom of the well. The accumulated liquid can be thought of a standing liquid column, which creates pressure barrier that restricts gas flow from reservoir to the surface of the well. Due to liquid loading the well is not able to produce either gas or liquid to the surface, and thus requires artificial means of lifting the accumulated liquid. Plunger lift is used to de-liquefy gas wells in such cases of liquid loading.

[0003] Plunger lift is a common form of artificial lift mechanism used for de- liquefaction of gas or oil wells. A plunger is a specifically designed metal rod with diameter slightly less than the tubing diameter. The plunger travels from top to bottom of the well, when the production valve is closed. The accumulated liquids collect above the plunger. When the production valve is opened, the plunger moves the liquids from bottom of the well to the surface. The aim of plunger lift is to produce liquids to the well-head, minimize liquid levels at the bottom of the well and thus ensure greater hydrocarbon production.

[0004] In a typical shale gas field, thousands of wells are spread across several miles.

Each well has a remote terminal unit (RTU) for online control, which communicates with a central control room using radio signals. Typical measurements available in the plunger operated wells are casing pressure, tubing pressure, sales-line pressure, gas flow rate, and plunger arrival time at surface. Some wells also have liquid flow measured at surface. Since each operator manages almost hundred wells or more, problems and faults during well operation go undetected. Moreover, it is time consuming activity for operators to drive to each well and check for possible issues. Many times such problems continue to exist undetected and adversely affect well production.

[0005] Equipment such as sensors, actuators, plunger, well construction, controller etc. used for obtaining the measurements in a plunger lift system can malfunction/fail during a course of operation. Poor operation practices also cause problems since liquids are not removed efficiently from the well. In case of any such malfunction typically production is adversely affected. Unless the failure initiates a safety system or stops production, causing to shut down the plunger lift well, the well continues to produce at low production rates. Such faults/malfunctions are important to identify so that corrective measures can be initiated restoring wells at full production capacity and avoid extended shut-in. In order to address the operating challenges, the industry needs an automated fault detection system for plunger lift operation.

[0006] Quantitative model based approaches need either a dynamic or static model of the process. The error from predicted and measured values is calculated by comparing the model prediction with actual measurements. Further, deviation in state and/or parameters values can also be reported as faults. On the other hand, in qualitative model based approaches, either a causal models or abstraction hierarchy is established using the process know how. The drawback of quantitative model based approach is that it requires a model to be available to perform analysis, which is often unlikely. In qualitative model based approaches process knowledge, process fault diagrams, flow network etc. are needed. Though such information may be readily available for a given process, however it changes from plant to plant. Therefore qualitative model based approaches present challenges to be applied in a generic manner.

[0007] Thus there is a need to develop a system and method for fault detection and diagnosis for the plunger operation in hydrocarbon wells that addresses the challenges of the prior art method as stated above. BRIEF DESCRIPTION

[0008] In one aspect a system for plunger fault monitoring of a plunger lift operation in a hydrocarbon well is disclosed. The system includes sensors to obtain plunger operational measurements. An operational measurement module receives the plunger operational measurements. A plunger reference module is configured to classify a plurality of plunger faults and corresponding fault signatures associated with the plunger operation cycle into at least one of three categories of a sensor fault, a structural fault, or a behavioral fault, for generating a plurality of derived features that are derived from plunger operation measurements. The plunger reference module is further configured for storing reference operation data comprising an upper bound and a lower bound for each of sensor operation measurements, one or more of comparative data, standard co-relational data or predefined model, estimated operational data based on at least one of historic plunger operation cycles, a pre-defined plunger operation model. A plunger fault processor module is configured to execute a plurality of rules that operate on the derived features and the plunger operation measurements to indicate an abnormality in the plunger operation cycle; and a plunger fault diagnosis module is configured to analyze the abnormality based on a decision-classification tree structure to decide an occurrence of at least one of the sensor fault, the structural fault, or the behavioral fault in an increasing hierarchy level, where the sensor fault is at first hierarchy level, structural fault is at second hierarchy level and the behavioral fault is at third hierarchy level. The operational measurement module, the plunger reference module, the plunger fault processor module and the plunger diagnosis module are configured on a computer processor.

[0009] In another aspect, a method for plunger fault monitoring for a plunger operation cycle is disclosed. The method includes steps for classifying a plurality of plunger faults and corresponding fault signatures associated with the plunger operation cycle into at least one of three categories of a sensor fault, a structural fault, or a behavioral fault; generating a plurality of derived features that are derived from plunger operation measurements; generating a plurality of rules that operate on the derived features and the plunger operation measurements to indicate an abnormality in the plunger operation cycle; and analyzing the abnormality based on a decision- classification tree structure to decide an occurrence of at least one of the sensor fault, the structural fault, or the behavioral fault in an increasing hierarchy level, wherein the sensor fault is at first hierarchy level, structural fault is at second hierarchy level and the behavioral fault is at third hierarchy level, where the plunger operation measurements are obtained from sensor measurements

DRAWINGS

[0010] These and other features, aspects, and advantages of the present invention will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:

[0011] FIG. 1 is a schematic representation of an exemplary plunger lift installation;

[0012] FIG. 2 is an exemplary plunger lift cycle;

[0013] FIG. 3 is a graphical representation of typical trends of exemplary measured values in a plunger lift cycle;

[0014] FIG. 4 is a flowchart representation of exemplary steps in the method of the invention;

[0015] FIG. 5 is a further flowchart representation of exemplary steps in the method of the invention; and

[0016] FIG. 6 is a block diagram representation of a system for plunger fault monitoring of a plunger lift operation in a hydrocarbon well

DETAILED DESCRIPTION

[0017] As used herein and in the claims, the singular forms "a," "an," and "the" include the plural reference unless the context clearly indicates otherwise.

[0018] As noted herein, in one aspect the invention provides a method for plunger fault monitoring in a plunger operation cycle. Turning to the figures, FIG. 1 shows a plunger lift system, which consists of the following typical components: an outer tube called casing 100, an inner tube called tubing 102, which is connected to a sales line (via 112), a control valve 110 that can be opened or closed to allow the well to flow or shut-in, a plunger 104 that can move up or down the tubing. During a typical plunger operation cycle, the valve is opened or closed. When the valve is opened, the plunger is intended to eventually come to rest in the catcher/lubricator 108 located at the wellhead. When the valve is closed, the plunger falls and eventually comes to rest at the bottom seat 106. Perforations 114 in the casing 100 are provided to allow flow of fluids from the reservoir / formation 116. A controller 130 is provided to collect measurements, determine control actions and communicate data with a SCADA.

[0019] Some exemplary sensors that are connected to the controller 130 include, but not limited to: Casing pressure sensor 120 to detect casing pressure (CP); tubing pressure sensor 122 to detect tubing pressure (TP); line pressure sensor 124 to detect line pressure (LP); gas flow rate sensor 128 to detect gas flow rate and an arrival sensor 126 to detect arrival of the plunger in the catcher and record the arrival time.

[0020] Plunger lift is a cyclic operation that consists of the following stages, a schematic of which is shown in FIG. 2: (i) the valve is closed 200 and the plunger falls; (ii) plunger arrives at the well-bottom 210; (iii) pressure in the well builds up with plunger at the bottom 222; (iv) the valve is opened 202 and the plunger rises with the accumulated liquid slug 224; (v) liquid slug reaches the surface 212; (vi) the plunger arrives at the surface 214; and (vii) plunger is held in the catcher while the well flows 228. After some time, the valve is closed again and the cycle repeats.

[0021] FIG. 3 shows typical trends of the measured values of casing pressure 300, tubing pressure 302 and flow rate 304. The area under the flow rate curve, indicated by the shaded region 306 is used to calculate the daily production rate in Mcfd (thousand cubic feet per day). The various events during a plunger cycle (denoted in FIG. 2 as 200, 210, 202, 214 and 200) are indicated at the bottom of the time axis (X-axis) in FIG. 3. On the time axis, reference numerals 308, 310, 312 and 314 correspond to the plunger fall time 220, buildup time 222, plunger rise time 224 and 226 and after-flow time 228, respectively.

During the plunger lift operation as explained in relation with FIG. 2, a variety of faults can occur. The method of invention is used for plunger fault monitoring of possible faults during a plunger operation cycle. These faults are classified according to the method of the invention as shown in flowchart 316 in FIG. 4 into sensor/actuator faults, structural faults and behavioral faults at step 318. An exemplary list of different faults in each category is shown in Table 1 below:

TABLE 1

The classification includes corresponding fault signatures in the sensor data obtained from historic plunger operation cycles, that are associated with each of the faults in the different categories. It may be noted here that the faults are not necessarily detected in the order written here. In an exemplary embodiment, first, pressure/flow sensor or communication fault is detected. If fault is found, it is reported. Next, valve leaking fault is checked. Next, arrival sensor fault is checked. If detected, missed arrival fault (from "behavioral fault" list) is checked. Note that "arrival sensor" and "no arrival info" are the only sensor malfunction faults that trigger the method steps to proceed.

Next, at step 320, the method includes generating a plurality of derived features that are derived from plunger operation measurements that are obtained from sensor measurements such as those mentioned in relation with FIG. 2 and FIG. 3. Some exemplary non limiting derived features include the following: (i) actual measurements of casing, tubing and line pressures (CP, TP and LP, respectively); (ii) actual measurement of flow rate (FR); (iii) arrival time, i.e., the time at which plunger arrives at the surface after the valve has been opened; (iv) average plunger speed; (v) difference (CP - TP); (vi) difference (CP - LP); (vi) difference (TP - LP); (vii) differential of CP with respect to time (denoted as 5(CP)); (viii) differential δ(ΤΡ); (ix) differential 5(LP); (x) differential 5(CP-TP); (xi) power of (TP-LP) ⁿ , where n is a number between 0 and 1 ; (xii) integral of gas flow rate with respect to time, i.e., J (FR) and (xiii) difference between time-shifted data (x(t)-x(t-i)), where, x represents any of CP, TP, LP and FR. In addition to these, individual cycles are identified and time series data is divided into cycles, yielding the following features: (xiv) cycle data (data divided into cycles); (xv) closed cycle data (i.e., data during closed cycle); and (xvi) open cycle data (i.e., data during open cycle). In other words, differentials with respect to time are calculated for each closed cycle, and for each open cycle. Once the cycle-wise data are identified, the following non limiting derived features are calculated (xvii) maximum value of CP, TP, LP and FR during each cycle; (xviii) derived features i to vi calculated at start and end of each cycle; and (xix) derived features vii to x calculated in each closed cycle and open cycle. In a specific implementation these derived features are calculated only if there is no "sensor fault" detected.

[0024] Next, at step 322, the method includes generating different rules that operate on the derived features and the plunger operation measurements to indicate an abnormality in the plunger operation cycle. These rules are explained in more detail in reference to flowchart of FIG. 5.

[0025] Next at step 324, the method includes analyzing the abnormality based on a decision- classification tree structure to decide an occurrence of at least one of the sensor fault, the structural fault, or the behavioral fault in an increasing hierarchy level, wherein the sensor fault is at first hierarchy level (i.e., these are detected first), structural fault is at second hierarchy level (these are detected next after sensor fault detection is completed) and the behavioral fault (these are detected after structural faults are detected) is at third hierarchy level. This tree structure enables higher confidence in detecting faults as per the classification.

[0026] The method steps 322 and 324 of the invention are explained in more detail in reference to the flowchart in FIG. 5, and is generally represented by numeral 328. The method steps of generating and applying rules, includes providing an upper bound and a lower bound for each of sensor operation measurements. This step is depicted by numeral 330 in FIG. 5. The upper bound and lower bound may be obtained from a variety of sources, or may be derived from various possible estimations, and calculation. In some embodiments, the upper and lower bound are based on historical data; in another embodiment, they are obtained from a previous sensor measurement during the plunger operation cycle.

[0027] The method includes providing dynamic change capturing for the sensor measurements using one or more of comparative data, standard co-relational data or predefined model, represented by numeral 332 in FIG. 5. The method also comprises providing estimated operational data based on at least one of historic plunger operation cycles, or a pre-defined plunger operation model, shown in in FIG. 5 by numeral 334.

[0028] As already noted herein, during a typical plunger operation cycle, plunger operation measurements are obtained, depicted by method step with reference numeral 336 in FIG. 5. The plunger operation measurements are then compared with the corresponding upper bound and the lower bound for each sensor operation to determine a sensor condition in the method of the invention, which step is represented by numeral 338 in FIG. 5. Based on the comparison, the sensor condition may be designated as at least one of a sensor normal condition, or a sensor fault condition, shown in FIG. 5 by numeral 340. In absence of redundant measurement of any sensor, the correlation of one measurement may be done with itself (from the same cycle, as well as consecutive cycles) and with other measurements. For example during the valve close the 'flow sensor' should show zero reading. In case of positive reading during valve close indicates either the valve close error (actuator fault) or flow sensor error (sensor fault). Further, these two faults can be distinguished by matching flow patterns in the signal to regular flow pattern, or using a model based dynamic analysis. Other such correlating measurements will become obvious to one skilled in the art, and is contemplated to be within the scope of the invention. In this manner, any error associated with faulty measurement equipment, sensors, actuators, valves etc., may be determined in a facile manner.

[0029] If the sensors have been found to be functioning in a normal condition based on the method step as described herein, then the plunger operation measurements are compared with the one or more of (a) comparative data, (b) standard co-relational data or (c) predefined model for dynamic change capturing to determine a status of structural condition. This step is represented by numeral 342 in FIG. 5. In this manner, the status of structural condition is determined as being at least one of a normal structural condition, or a damaged structural condition. Exemplary damaged structural condition include, but not limited to, problems in tubing, casing, plunger, elbows, tees, connectors, and the like, and combinations thereof.

[0030] If a structural damage has happened, irrespective of the control mechanism used the system is bound to underperfbrm. Such faults are identified after ensuring that the sensors are working properly. Any process dynamic change can be captured either by correlating the available measurements with each other and comparing them with standard correlation or by use of model based approach where the case of poor model fitting indicate fault as described herein in 342 in FIG. 5.

[0031] If the structural condition is determined to be normal or damaged structural condition based on the method of the invention described herein so far, then the plunger operation measurements are compared with estimated operational data, which step is depicted by numeral 344 in FIG. 5 to determine a behavioral condition. The estimated operational data may be obtained based on at least one of historic plunger operation cycles, or a pre-defined plunger operation model. By comparing the plunger operation measurements with estimated operational data, the behavioral condition may be determined as being normal behavior condition or abnormal behavior condition. [0032] Typically, behavioral faults arise from improper operation of the well. In a plunger lift operation, well shut-in time and open time are often selected by an operator. Incorrect or poor choice of theses settings can result in operational problems, which can cause significant damage to well, reduced life of equipment and loss of production, if undetected. These are detected as behavioral condition in this invention by comparing the current operation cycle with the expected qualitative and quantitative trends for a normal cycle. In one exemplary embodiment, arrival sensor data is correlated with a signature in the flow-rate data to identify plunger arrival. Non-arrival of plunger triggers diagnostic mode where the potential cause(s) are identified. This is done by using estimates of down-hole liquid loading and energy stored in the casing during well shut-in. In a similar manner, correlation analysis from the measured data and estimates of unmeasured operating parameters are used to identify the behavioral condition.

[0033] The method of the invention includes a step of reporting at least one of the sensor condition, structural condition, and behavior condition, depicted by numeral 346 in FIG. 5.

[0034] The method of the invention as mentioned in relation to FIG. 4 and FIG. 5 includes diagnosing a fault condition wherein the fault condition is at least one of the sensor fault condition, the structural damaged condition or the abnormal behavior condition to generate diagnostic fault parameters. If faults are detected (either single fault or multiple faults), the identification procedure for the fault condition is analyzed and a root-cause/ fault data leading to the fault is identified. The method for fault detection includes, but not limited to a structured fault tree, data reconciliation, model based redundant measurement generation, and the like, and combinations thereof. Such root-cause identification for a number of fault conditions may be pre-programmed and provided for rapid identification of the cause of fault. Thus back-tracking of identification is enabled by use of the structured approach during fault detection delineated in the method of the invention. The fault condition obtained by the method of the invention may be accepted or rejected by an operator, or in an automated manner based on the diagnostic fault parameters. [0035] The method also includes storing at least one of the plunger operation measurements, sensor condition, structural condition, behavior condition, and the fault condition, and combinations thereof for further future analysis. The stored values may also be used to update the upper bound and the lower bound, the one or more of comparative data, standard co-relational data, predefined model, the historic plunger operation cycles, the pre-defined plunger operation model.

[0036] The method of the invention provides for detecting possible faults occurring during operation of plunger lift system while relying only on typical measurements (for example, casing pressure, tubing pressure, line pressure, gas flow, plunger arrival and battery voltage).

[0037] The output of the method may include highlighting a sequence of faulty data in the full data set, and provides diagnostic information on why this fault is flagged. This information enables operator to accept or reject the generated fault and take appropriate corrective action. In this manner, prioritizing of faults when more than one fault condition is identified may also be done based on a number of factors that will be obvious to one skilled in the art.

[0038] In another embodiment, the method of the invention enables identification of a golden cycle, i.e. the optimal plunger lift cycle for highest hydrocarbon production. The objective of golden cycle is to identify best control setting/ conditions for a given well and compare it with operating data to identify any control related issues.

[0039] The description below explains exemplary implementation of the method of the invention for select non-limiting fault conditions. As explained herein above, sensor faults in any of the pressure or flow sensors, or valve leak are classified as major faults according to the method of the invention. When any of these faults are detected, the method terminates so that these faults can be resolved on priority. An exemplary procedure used for identifying sensor faults is given as follows:

(a) Pressure sensor faulty (negative value): This is a major sensor malfunction fault on either of the casing, tubing or line pressures. If any of the CP, TP or LP values are outside normal operation range, then the method flags this as a fault and exits, and checks are performed by operators on casing, tubing and line pressures to fix the sensor fault.

(b) Large variation in pressure sensor: Statistics for each of the pressures (CP, TP and LP), such as mean, standard deviation and median are calculated. If pressure fluctuations are beyond a certain threshold calculated using these statistical quantities, then this is flagged as a pressure sensor/transducer error.

(c) Pressure sensor, transmitter or communication error: This is calculated from the difference between time-shifted data. An exemplary derived feature is for example a feature variable "dsig" for CP is calculated as: dsig(t) = [CP(t)-CP(t-l); CP(t)-CP(t-2); CP(t)-CP(t-n)] for t spanning all the data from the sensor measurements for CP for a plunger operation cycle. The decision variable is then obtained to flag the sensor condition as sensor fault condition when if multiple consecutive data have the same value. This is repeated for TP and LP as well.

(d) The method is also able to detect and report direct measurement related faults that are based on high unacceptable variation in data or biased data. These aren't an indicator of a faulty sensor or transducer, but are flagged due to possible calibration or operational issues under sensor faults. Such exemplary and non-limiting fault conditions are:

(i) Line pressure unstable: This is a not a sensor fault. It only informs that the line pressure is varying too much adversely affecting the plunger operation. The statistical quantities, median and spread of predefined set of data is calculated. A fault condition is identified

(ii) Casing or tubing sensor calibration: This is a sensor bias fault. This derived feature captures the variation in casing and tubing pressure above and below its median, by calculating the median and spread of middle 50% of the data. [0041] In the exemplary method the sensor fault detection is based on time-trend data.

If the sensor/valve failure is not detected in the time-trend data then the method continues, with the time-trend data split into cycle-wise trends. The derived quantities based on cycle-wise trends were described in paragraph [0023].

[0042] The cycle-wise data is checked for errors in the arrival sensor. These include sensor fault (in arrival sensor), non-arrival and no arrival information. First, the cycle-wise data is checked for sensor arrival information (i.e., arrival time). It would be appreciated by those skilled in the art that such data indicates arrival sensor is functioning or the condition that plunger failed to timely surface (i.e., slow arrival, which is verified by comparing average plunger speed with a pre-defined value).

[0043] The following exemplary rules were applied on the cycle data: For cycles where arrival sensor information is missing; Flow or Tubing pressure data is analyzed for presence of peak during arrival. If no peak is detected the cycle is identified for "missed arrival" otherwise it is detected as "arrival sensor error". Thus using multiple overlay of rules, the correct abnormality for the sensor fault is detected which is often missed in manual detection methods, and any other current techniques for detecting faults in plunger operation cycle.

[0044] Next, for structural fault detection, the derived features calculated from closed cycle data were used. For example, for detecting hole in tubing as a structural fault following rules were applied on the derived features. First, based on statistical analysis of historic data from a large number of wells and/or using a representative numerical model of a well, a derived signal threshold value is calculated. Next, the derived quantity (for example CP-TP in closed cycle data) is compared with derived threshold. When more than a predefined percentage of this value lies between the derived threshold, then this is identified as "hole in tubing" fault.

If hole in tubing is detected, then the method directly moved to detect behavioral faults (the remaining structural faults are not checked). [0045] As another example, for detecting "Liquid prefers casing" that can happen when perforations are clogged or when running a plunger in a packer well, the derived features described in (xix) in paragraph [0023] were calculated for closed cycle data and open cycle data. If these derived features met one or more conditions (for example, negative values, decreasing trend, etc.), then the condition is detected as "liquid prefers casing" fault.

Next the method moves to determining behavioral faults.

[0046] Yet another example is for detecting low plunger efficiency: plunger efficiency is defined as amount of liquid removed by plunger in a cycle to the amount of liquid entered the vertical section of well from reservoir. A threshold value was used as derived feature, for example a threshold of 15% for dual pad plunger was chosen and an efficiency value less than 15% was determined as a fault condition. This threshold varies with the type of plunger used.

[0047] Next, for behavioral fault detection, the derived features (xvi) - (xviii) in paragraph [0023] were used. First, the historical operating data and/or representative mathematical model of well behavior was used to determine range of conditions for normal operation. If the derived feature is in open cycle data lies within this range, no fault is detected; if it lies below the minimum threshold, "well flowing too long" is detected; if it lies above the maximum threshold, "well flowing too short" is detected.

[0048] In another aspect, the invention provides a system for plunger fault monitoring of a plunger lift operation in a hydrocarbon well. The system 400 is shown in the block diagram representation in FIG. 6. The system includes different sensors to obtain plunger operational measurements, as indicated by reference numeral 410. Then, the system includes an operational measurement module 412 to receive the plunger operational measurements from the different sensors.

[0049] The system 400 also includes a plunger reference module 414 configured to classify a plurality of plunger faults and corresponding fault signatures associated with the plunger operation cycle into at least one of three categories of a sensor fault, a structural fault, or a behavioral fault, for generating a plurality of derived features that are derived from plunger operation measurements and for storing reference operation data that include an upper bound and a lower bound for each of sensor operation measurements, one or more of comparative data, standard co-relational data or predefined model, estimated operational data based on at least one of historic plunger operation cycles, a pre-defined plunger operation model;

[0050] The system 400 further includes a plunger fault processor module 416 configured to execute a plurality of rules that operate on the derived features and the plunger operation measurements to indicate an abnormality in the plunger operation cycle. For example the plunger fault processor module includes rules to compare the plunger operational measurements with the reference operation data to determine a fault condition as a sensor fault condition, a structural damage condition, or a behavior fault condition based on a pre-existing logic associated with each form of comparison to arrive at the final status of the sensor, structure and/or behavior.

[0051] The system also includes a plunger fault diagnosis module 418 to analyze and diagnose the abnormality based on a decision-classification tree structure to decide an occurrence of at least one of the sensor fault, the structural fault, or the behavioral fault in an increasing hierarchy, wherein the sensor fault is at lowest hierarchy and the behavioral fault is at highest hierarchy, and to also generate diagnostic fault parameters as described herein.

[0052] The system of the invention also includes a user interface 420 for displaying an outcome from the plunger fault diagnosis module and for receiving user inputs for interacting with data in the plunger reference module. The system of the invention also includes other such modules, such as an output module 422, a display module (not shown separately), and the like, which are used for generating reports and communicating the outcome from the plunger diagnosis module and user interface data to another communication device as will become obvious to one skilled in the art, and is contemplated to be within the scope of the invention. It would be understood by those skilled in the art that the different modules referred herein are configured on a computer processor.

[0053] One skilled in the art will also immediately recognize that the method and system of plunger fault monitoring may be effected through a suitable software tool that is implemented using a computer processor. Thus, in other aspects, the invention provides a tool for implementing the method for plunger fault monitoring, as a series of independent tools, or as a combined tool that can provide all solutions needed for plunger operation.

[0054] The software tool may be made available through any known formats, such as a downloadable file from a suitable location, or in a storage medium such as CD or DVD or a flash drive, or alternately in an EPROM that is integrated into the existing control system of the process plant. One skilled in the art will also immediately recognize that the historical data, the models, dynamic change capturing, and all other aspects of the comparisons may be effected through a suitable software tool that can implement the models developed for this purpose. This tool may be made available as an independent additional software or an application that is capable of being integrated with an existing plunger operation tool that may already be in place.

[0055] Other hardware and software devices needed for smooth and efficient operation of the plunger are also contemplated to be within the scope of the invention. For example, the sensors or actuators may be connected to the software tool via a processor through wired means or wireless means. Other such variations and possibilities will become obvious to one skilled in the art, or may be arrived at without undue experimentation.

[0056] In the exemplary implementation MATLAB was used to formulate the rules, and for generating * .m files, as would be understood by those skilled in the art. MATLAB files can then be converted into C# .net executable files as would be understood that were delivered to the plunger fault diagnosis module and then the output from the plunger fault diagnosis module was displayed on the user interface. [0057] The method and system for plunger fault monitoring of a plunger lift operation in a hydrocarbon well described herein is provides higher level of accuracy in fault detection, and removes the complexity of both manual detection and existing complex quantitative and qualitative fault detection methods a the system and method of the invention only use typical surface measurements (i.e. one or many of, casing pressure, tubing pressure, line pressure, gas flow, plunger arrival and battery voltage) to identify possible faults in the operation cycle. These surface measurements are easily obtained at a central location using wireless communication such as radio signals or wifi for processing by the method and system of the invention. It is further advantageous as it does not depend on redundant measurements, as the lack of redundant measurements is addressed by inferring unmeasured variables and/or correlating multiple measurements. Further the method and system of the invention enable prioritizing the identified faults in a structured way in order to correctly isolate and identify multiple faults.

[0058] While only certain features of the invention have been illustrated and described herein, many modifications and changes will occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.