• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    computer system readable volatile medium, system and method for assessing water injection into a subsurface hydrocarbon reservoir in which one or more producer wells and one or more injector wells have been drilled and method for detecting flow rate change events of said wells in said hydrocarbon reservoir method and computer system to derive a statistical reservoir model from the associations between injection wells and production wells. potential injector events where changes in the 1 æ injection rate are interactively identified from time series flow rate measurement data in the wells in a production field, with confirmation that some response to these Gun events appear in producing wells. Gradient analysis is applied cumulative time series of production volume from production wells to identify points in time at which the cumulative production gradient changes by more than one threshold value. Potential identified producer events are scattered over time again against thresholds. An automated association program classifies injector-producer associations according to the strength of the association. The reservoir capacitance-resistivity model is evaluated using the flow rate measurement data for the highest-rated injector-producer associations. Additional associations are added for subsequent iterations of the reservoir model until the improvement in uncertainty in the parameters of the evaluated model is not statistically significant.
    公开号:BR112014010661A2
    申请号:R112014010661
    申请日:2011-11-18
    公开日:2019-08-27
    发明作者:Shirzadi Shahryar；Bailey Richard；Ziegel Eric
    申请人:Bp Corporation North America Inc；Bp Exploration Operating Company Limited；
    IPC主号:
    专利说明:

    DESCRIPTIVE REPORT
    LEGIBLE VOLATILE MEDIA BY COMPUTERIZED SYSTEM, SYSTEM AND METHOD FOR ASSESSING INJECTION OF WATER IN A SUBSUPERFICIAL HYDROCARBONET RESERVOIR, IN WHICH ONE OR MORE PRODUCING WELLS AND ONE OR MORE INJECTOR WELLS HAVE BEEN PERFORMED BY EVERYTHING OF MURDERS AND MORE OF Said Wells in Said Hydrocarbon Reservoir
    CROSS REFERENCE TO RELATED PATENT APPLICATIONS
    Not applicable.
    DECLARATION ON RESEARCH OR DEVELOPMENT SPONSORED BY THE FEDERAL GOVERNMENT
    Not applicable.
    BACKGROUND D7i INVENTION
    This invention belongs to the field of oil and gas production. Embodiments of this invention are more specifically aimed at analyzing secondary recovery actions to maximize oil and gas production.
    The current economic climate emphasizes the need to optimize hydrocarbon production. This optimization is especially important considering that the costs of drilling new wells and operating existing wells are high by historical standards, largely because of the extreme depths at which new producing wells must be drilled and because of other physical barriers to discovery and exploration of reservoirs; those reservoirs that are easily accessible have already been developed and produced. These major economic challenges require operators to dedicate
    2/82 substantial resources for the effective management of oil and gas reservoirs, and the effective management of individual wells in production fields.
    As is known in the art, an important secondary recovery operation injects water, gas, or other fluids into the reservoir of one or more injection wells, commonly referred to as water injection. In theory, this injection increases the pressure in production wells that are connected to the injection wells, through the reservoir, thus producing oil and gas at high flow rates. In planning and managing secondary recovery operations, the operator is faced with decisions about whether to start or stop such operations, and also how many wells will serve as injection wells and their locations in the field, to maximize production at minimal cost.
    As is known in the art, the optimization of a production field is a complex problem, involving many variables and presenting many options, exacerbated by the complexity and unfathomability of the subsurface architecture of the current production reservoirs. Especially for reservoirs at extreme depths, or located in difficult or inaccessible land or sea locations, the precision and accuracy of the indirect methods necessarily used to characterize the structure and location of hydrocarbon-carrying reservoirs are necessarily limited. In addition, the subsurface structure of many reservoirs has complexities, such as porosity and variable permeability of the rock; fractures and failures that compartmentalize formations may also be present
    3/82 in the reservoir, which further complicates the subsurface fluid flow. Numerical models and techniques to estimate and analyze the effect of injection in a well, on the flow rates in one or more producer wells, are desirable tools to solve this complex problem of production optimization.
    A class of models for analyzing the effects of water injection is known in the art as capacitance models, or capacitance-resistivity models. Examples of these models are described in Liang et al, Optimization of Oil Production Based on a Capacitance Model of Production and Injection Rates, SPE 107713, presented in 2007 at the SPE Hydrocarbon Economics and Evaluation Symposium (2007); Sayarpour et al, The Use of Capacitance-resistivity Models for Rapid Estimation of Waterflood Performance and Optimization, SPE 110081, presented at the 2007 SPE Annual Technical Conference and Exhibition (2007); and Kaviani et al., Estimation of Interwell Connectivity in the Case of Fluctuating Bottomhole Pressures, SPE 117856, presented at the 2008 Abu Dhabi International Exhibition and Conference (2008). In a general sense, the capacitance-resistivity model (CRM) is the result of a regression (for example, multivariate linear regression) applied to the flow rate of the injector well and the flow rate of the producing well, to express the production rate accumulated in a production well over time as the sum of a primary production term (typically an exponential from an initial production rate value), a term that expresses the effect of changes in pressure well bottom (BMP) of the producing well itself, and an
    4/82 third term corresponding to the flow rate in an injector multiplied by a coefficient of interconnectivity for the trajectory between the injector well and the producer in question, added up over all relevant injectors in the field. This model allows the evaluation of changes in the production of a producing well, in response to changes in the injection rate in one or more injectors.
    Of course, modern production fields generally involve more than one production well, each responding to the injection in one or more injection wells. In other words, the flow of a given injector will not be uniformly distributed across the formation to the various production wells; in addition, the producer-producer effects may also be present, in which the increase in production from one producing well affects the production from another producing well (for example, by the local reduction of the reservoir pressure in the affected well). These mechanisms prohibit CRM evaluation in each well individually - instead, the definition and evaluation of the model requires regression to be performed simultaneously in all producing wells in relation to all injector wells. Considering that conventional capacitance-resistivity models use three parameters for each combination of producer injector wells, even a field of modest size will require convergence of the model for a relatively large number of parameters. As a result, CRM is necessarily over-parameterized, often resulting in an inability to achieve a reasonable solution when applied to real production fields. Even with modern computing resources, this operation is, at best, quite time-consuming and inefficient.
    5/82
    For mature production fields, the well flow rate over time provides a significant source of data useful for deriving a connectivity model. In some cases, flow rates over time, for both producer and injector wells are readily available; in other cases, downhole or wellhead temperature and pressure measurements, from which flow rates can be inferred, are available. Again, even for a modestly sized production field, the volume of such data can quickly become overwhelming. Rigorous numerical analysis of these data in defining and evaluating a connectivity or response model (eg, CRM) consumes substantial computing time and resources. These 15 large data sets and the complex interaction of flows between injectors and producers make it difficult for a human user or an automated numerical system to identify the causal relationships between injection events and produced fluids.
    As a relevant record, US Patent No. 8,800,200, issued on February 15, 2011, entitled Process-Related Systems and Methods, often mentioned in this document and incorporated by reference in its entirety into this patent application, 25 describes the system and method for monitoring the values of multiple process variables over time, and identification of causal relationships between process variables, including the identification of cause events in one process variable and response events in another corresponding process variable . According to this
    6/82 patent, the system and method also associate confidence levels with the events identified.
    SUMMARY OF THE INVENTION
    According to several embodiments, current studies provide an automated method and system that can efficiently derive a statistical model for injector-producer behavior in an oil and gas field from historical production data.
    According to several embodiments, current studies provide an easily scalable method and system capable of efficiently analyzing a large number of events over long periods of time, in a non-interventionist manner from the point of view of reservoir engineering personnel.
    According to several embodiments, current studies provide a method and system that provide statistical insight into the parameters of the model, in order to be useful in optimizing field production.
    According to various embodiments, current studies provide a method and system that can easily identify correlated causal events in production data automatically.
    According to several embodiments, current studies provide a method and system that can facilitate user input and selection in identifying causal events and relationships in production data.
    According to various embodiments, current studies provide method and operable system for flow measurements over time and also for indirect signals for flow rates.
    7/82
    According to several embodiments, current studies provide a method and system that can filter intrapace events, such as changes in gas elevation or obstruction position, by detecting causal events in production data.
    According to several embodiments, current studies provide a method and system that can identify injection response events that could be masked by an intrapace event in the production well.
    According to several embodiments, current studies provide a method and system that can take into account the correlation of injection events that occur in several injection wells simultaneously.
    According to several embodiments, current studies provide a method and system that can assess the economic benefit of injection in specific wells.
    According to several embodiments, current studies provide a method and system that can use unstructured data in the origin and evaluation of the statistical model.
    Other objects and advantages of the exemplary embodiments in this patent application will be evident to those of ordinary skill in the art who use the following specification and drawings as a reference.
    This invention provides a computer system and method to evaluate the effect of potential secondary recovery actions by water injection to be applied to oil and gas reservoirs in which there are several production wells and several injection wells. Measurement data, such as well flow rates and downhole pressures, are acquired over time. These measurement data are analyzed to
    8/82 to identify associations of cause and effect between injectors and producers. Associations are classified according to levels of trust, for example, into subsets of strong association, moderate association, weak association, and no association. The injector-producer interconnections corresponding to the highest classification associations are applied to a reservoir capacitance-resistance model. The resistivity capacitance model is evaluated according to the measurement data, to obtain an estimate of the error. One or more of the next best-classified interconnections is applied to the model, which is again evaluated according to the measurement data. Additional associations are applied to the model, and the evaluation is repeated, until the incremental change in the adjustment to the measurement data resulting from an added interconnection has no statistical significance. Other important exclusions, for example based on geography or geology, can also be applied. The resulting convergence model is then used to optimize water injection and production.
    The exemplary system and method provides quick feedback in assessing the potential of water injection actions. By the iterative application of interconnections in order of their confidence levels in the identification process, the number of interconnections applied to the capacitance-resistivity model is limited to only those necessary to adjust the measurement data. Interconnections that have little or no effect are not involved in the construction and evaluation of the reservoir model. This results in a lean and efficient reservoir model that can quickly assess candidates for
    9/82 secondary recovery. The system and method is also easily scalable for production fields, including a large number of injection and production wells, and flow history data obtained over relatively long periods of time.
    The exemplary system and method is capable of performing standard error and confidence calculations in the capacitance-resistivity model by the iterative elimination of parameters with high standard error and, thus, increasing confidence around the remaining parameters. As a result, the system and method can achieve a high degree of confidence in its analysis.
    exemplary system and method is able to estimate the average response time for the production field by means of capacitance-resistivity modeling of the reservoir level, and enable the linking of these estimates to the cause-response analysis to better estimate injector-producer associations.
    exemplary system and method is capable of estimating the value of water (that is, the volume of oil produced in relation to the volume of water injected in each injector) to assign injection priority among the injectors in the production field in optimizing injection performance of water. BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
    Figure la is a schematic representation of an oil and gas production area to which exemplary embodiments described in this patent application can be applied.
    Figures 1b and 1c are examples of representations of injection and production flow time series,
    10/82 respectively, which correspond to wells in the production field of Figure la.
    Figure 2 is an electrical diagram, in the form of blocks, of a computer system built in accordance with exemplary embodiments described in the present patent application.
    Figure 3 is a flow diagram illustrating the operation of the computer system of Figure 2 according to exemplary embodiments described in this patent application. ' :
    Figures 4a and 4b are flow diagrams that illustrate the operation of the system of Figure 2 in the identification of injector events in the operational flow of Figure 3, according to an exemplary embodiment in this patent application.
    Figures 5a to 5d show ; several graphs of examples of injector measurement data and injector events identified, as they can be generated in the identification of injector events, according to the embodiment shown in figures 4a and 4b.
    Figure 6 is a flow diagram illustrating the operation of the system of Figure 2 in identifying producer events in the operational flow of Figure 3, according to an exemplary embodiment 'in this patent application.
    Figure 7 : is a flow diagram that illustrates the gradient analysis process to detect producer events, according to the embodiment shown in Figure 6.
    Figures 8 to 8c are graphs of cumulative production measurement data for an example of a well
    11/82 in production, which illustrates the gradient analysis according to the embodiment of Figure 7.
    Figures 9 through 9c illustrate an example of averaging and temporal smoothing applied to potential producer events detected in accordance with the embodiment shown in Figure 7.
    Figure 10 is a flow diagram that illustrates a method for detecting causal relationships between injector and producer events, according to the embodiment shown in Figure 6.
    Figures' 11a and 11b are visualizations of an example of detected causal events resulting from the method of Figure 10, according to that embodiment.
    Figure 12 is a flow diagram that illustrates a method of classifying detected injector-producer pairs, according to the embodiment shown in Figure 6.
    Figure 13 is a flow diagram that illustrates a method of evaluating a capacitance-resistivity model (CRM) with a subset of the injector-producer associations identified, according to the embodiment shown in Figure 6.
    Figures 14a and 14b illustrate examples of classified lists of injector-producer associations, as results from the method of Figure 12, according to that embodiment.
    Figure 15 is a flow diagram illustrating the operation of the computer system of Figure 2 according to an alternative embodiment.
    DETAILED DESCRIPTION OF THE INVENTION
    12/82
    The present invention will be described in conjunction with one or more of its embodiments. More specifically, the present description refers to embodiments of the present invention that are implemented in a computer system programmed to carry out various stages and method processes for the optimization of production, through secondary recovery actions, specifically water injection, since it is understood that the present invention is especially beneficial when used in such an application. However, it is also considered 'that the present invention can be advantageously applied to other systems' and processes. Therefore, it should be understood that the following description is given by way of example only, and is not intended to limit the true scope of this invention, as claimed.
    In order to establish the context for the present description, figure 1 illustrates, in plan view, an example of a small proportion production field with which : embodiments of the present invention can be used. In this example, several wells PI through P7 and II through 15 are implanted in various locations within the production field 6, and in the conventional way they extend into the earth through one or more layers of the subsurface. Typically, each of the wells from Pl to P7 and II to 15 is in communication with one or more production formations by means of drilling, in the conventional way. In this example, wells Pl to P7 are production wells (producers), so that hydrocarbons from one or more subsurface formations flow out through those wells. On the other hand, in this example, wells II through 15 are injection wells (injectors), through which gas, water, or other fluids are pumped into the
    13/82 formations in order to increase the production of production wells from Pl to P7.
    As is known in the art, modern oil and gas wells are implanted with various sensors through which various operational parameters can be measured, or otherwise deduced. From the point of view of entry and exit, the most direct flow measurement is performed by a flow meter implanted in each well Pl a P7 and II to 15. In production fields where the flow of several producing wells is mixed in one valve piano, a flow meter can be implanted in the collector and measure the combined flow of these wells; the flow rate from the individual wells is then normally deducted by other means, such as flow tests. Many modern wells are deployed with downhole pressure and temperature sensors, wellhead pressure and temperature sensors, or a combination of both. Modern computational techniques, for example, based on predictive well models, can be used to derive flow rates from these pressure and temperature measurements. The publication of US patent application No. 2008/0234939, made on September 25, 2008, entitled Determining Fluid Rate and Phase Information for a Hydrocarbon Well Using Predictive Models, often mentioned and incorporated by reference, in its entirety, into 'this application patent, describes systems and methods for deriving flow rates from pressure and temperature measurements in the well, for use in connection with embodiments of the present invention. Other measurements that can be obtained from oil and gas wells include the measurement of parameters such as
    14/82 temperature, pressure, valve settings, oil / gas ratio, and the like. Measurements other than well measurements can also be performed, examples of which include process measurements carried out on the surface, results of laboratory analyzes of production samples, and also estimates of various computational models based on the measured parameters. These measurements and estimates can be useful in analyzing measured or deducted flow rates, or they can also be useful in managing the production field.
    Even for the relatively simple production field 6, as shown in Figure 1, the subsurface connectivity between wells Pl to P7 and II to 15 can be quite complex, as far as the real flow behavior is concerned. of oil, gas and water. The porosity and permeability of the rock can vary at different locations in the soil on the subsurface of the periphery of the production field. : In addition, geological structures, such as defects, passages, barriers and preferential orientation of fluid-permeable paths, can complicate the flow of subsurface fluid. Understanding the movement of the fluid within a hydrocarbon reservoir can therefore become quite complicated, even in the presence of relatively few resources in a relatively small domain.
    As is well known in the art and as mentioned above, secondary recovery techniques are useful in maximizing oil and gas production in typical reservoirs. In the context of the embodiments of the present invention, secondary recovery efforts that are of interest involve the injection of gas, water, or other
    15/82 fluids in injection wells, such as injectors II to 15 in the production field. 6 of Figure la. As is known in the art, due to cost considerations and also due to the possibility of undesirable consequences for the reservoir, such water injection is not generally constant over time, but is applied to one or more injection wells at specific times, for specific time intervals. Often, the injection is applied simultaneously to more than one injection well in the field, but not necessarily to all available injection wells. ·: Χλ ··
    As discussed above, however, the relationship between injection into a given injection well and the resulting increase in production from a production well is not simple, as it depends on the complex architecture and connectivity of the subsurface formations and interfaces. In addition to considering : 5 simply the overall flow rates, fluid flow rates of different phases (ie oil, gas, water) must be considered. For example, subsurface short circuits can occur, in which injected water flows disproportionately to a nearby production well, causing an increase in water flow from that nearby well with little effect on oil production. These 5 and other complexities complicate the design and optimization of secondary recovery by injection.
    As mentioned above, the metering capability deployed in modern production fields provides useful temporal intelligence regarding flow rates over time from each of the wells in the production area. These measures provide an important source of
    16/82 measurement data useful for designing, evaluating and optimizing secondary recovery efforts. However, the complexity of the production field mentioned above, together with the somewhat known response of formations to injection efforts, make it difficult to readily identify the ideal injection stimulus to maximize the hydrocarbon output response.
    Figure 1b illustrates an example of an injection flow time series, which could be measured in injection wells II to 15 do ; production field 6 of Figure la. As is evident from figure 1b, the injection flow rates in injection wells II to 15 are different over time, but in certain periods they show some correlation. For example, at time tl of Figure 1b, the injection flow rate in injection well II drops sharply, while the injection flow rate in injector 12 increases sharply. From time t2 in Figure 1b, the injection flow rates in injectors II, 14, 15 begin to increase slowly over time. Other modifications' : correlated and uncorrelated in 'injection flow rates are present over the period of time illustrated in Figure 1, which can extend over a relatively long period of time (for example, by seasons' measured in years).
    Figure 1c illustrates an example of a typical time series of production flow rates, for one or more phases, which can be measured in the production wells PI to P7 of the production field 6 of Figure la, for a period of time during the which · secondary recovery efforts, such as injection, 'shown in Figure 1b, can be applied. These flow rates include the typical decline
    17/82 in production over time, such as pressure drops in the reservoir, but whose fundamental effect is usually masked by various actions carried out in the wells themselves. For example, as is evident in Figure 1c, several closing events occur over the measurement period (which, again, can span months or years). Changes in the position of the blocking valve in the wellhead of each of the production wells P1 to P7 may also be involved as a cause of several changes in the production flow rate. As shown in Fiqura 1c, wells P6 and P7 are (closed (or perhaps not even exist) until very late in the period of time illustrated. In addition, the secondary recovery action by injection in injectors II to 15 is also superimposed on production rates and other events, in the time series in Figure 1c.
    During the water irrigation, other secondary recovery actions can also be carried out in the production wells themselves. An example of such other techniques Retrieve action middle is the lifting gas, which gas is injected 1 River annular space between the production tubing and the well casing production, which results in the aeration of the oil well River formation production. The consequent reduction in oil density allows the pressure of the formation to lift the oil column to the surface and increase the production volume. Lifting gas can be injected either continuously or intermittently, depending on the production characteristics of the well and the arrangement of the gas lifting equipment. The effects of these anti-trafficking stimuli are also reflected in the time series of production flow rates, as shown in Figure 1c. :
    18/82
    It must therefore be evident from the discussion above that the tasks of designing, evaluating and optimizing secondary recovery actions that involve water injection, based on the large database of flow measurements over time or calculations, involve analysis complicated and costly. Computerized system
    Embodiments of this invention are directed to a computerized method and system of analysis of measurements or calculations of rates '' injection and production flow to accurately design 1 and efficiency, evaluate and optimize oil production ; and gas from one or more wells in a production field through water injection. Figure 2 illustrates, according to an exemplary embodiment, the construction of the analysis system (system) 20, which performs the operations, described in this specification efficiently to derive a statistical model of the association between the injectors and the producers of a production field, based on measurements or flow calculations or other response variables acquired over the time of implanted wells. In this example, system 20 can be realized by means of a computer system, including workstations 21 connected to server 30 by means of a network. Of course, the architecture and construction of a computer system useful for the particular invention in particular can vary widely. 'For example, system 20 can be realized by a single physical computer, such as a conventional workstation or personal computer, or, alternatively, by a computer system implemented' distributed across multiple physical computers. In this way, the architecture
    19/82 generalized illustrated in Figure 2 is provided merely as an example.
    As shown in Figure 2 and, as mentioned above, system 20 includes workstation 21 and server 30. Workstation 21 includes central processing unit 25, coupled to the BUS system bus. Also connected to the BUS system bus is the input / output interface 22, which represents the interface resources by means of which peripheral I / O functions (for example, keyboard, mouse, monitor, etc.) interface. with other components of the workstation 21. The central processing unit 25 represents the data processing capacity of the workstation 21, and, as such, can be implemented by one or more cores, the coprocessing circuitry, and the like . The construction and specific capacity of the central processing unit 25 is selected according to the needs of the workstation applications 21, these needs, which include, at a minimum, the performance of the functions described in this specification, also include all the other functions that can be performed by the system 20. In the system architecture 20 according to this example, the system memory 24 is coupled to the system BUS bus and provides memory resources of the desired type useful as data memory for storing data input and the results of the processing performed by the central processing unit 25, as well as the program memory to store the computer instructions to be performed by the central processing unit 25 in carrying out these functions. Of course, this arrangement of memory is just one example,
    20/82 with the understanding that system memory 24 can implement such data memory and program memory in separate physical memory resources, or distributed, in whole or in part, outside the workstation 21. In addition, as shown in Figure 2, the measurement inputs 28 acquired from surface equipment and along wells such as: flow meters, pressure and temperature transducers, valve configurations, and the like implanted in both injection wells and wells in the production field, are introduced via the input / output function 22, and stored in a memory resource accessible to workstation 21, either locally or via the network interface 26. These measurement inputs 28 may also include process measurements obtained in the processing of the produced output, and results of laboratory analysis of production samples, etc .; in addition, measurement inputs 28 may include estimates from computer models (whether run on workstation 21 or elsewhere within system 20) based on measurement inputs 28 or other extrinsic information.
    The network interface 26 of workstation 21 is a conventional interface or adapter through which workstation 21 accesses network resources on a network. As shown in Figure 2, the network resources to which workstation 21 has access via the network interface 26 include server 30, which resides on a local network, or geographic network, such as an intranet, a virtual private network, or on the Internet, and that is accessible to workstation 21 through one of these network arrangements and corresponding communication facilities with
    21/82 or wireless (or both). In this embodiment, server 30 is a computer system, of conventional architecture similar, in general, to that of workstation 21, and, as such, includes one or more central processing units, system buses, and memory, network interface functions, and the like. In accordance with this embodiment of the invention, the server 30 is coupled to the programming memory 34, which is a computer-readable medium that stores executable computer program instructions, according to which the operations described in this specification are performed by the system analysis 20. In this embodiment of the invention, the computer program instructions are executed by the server 30, for example in the form of an interactive application, on the input data transmitted from the workstation 21, to create the data ' output and results that are communicated to workstation 21, for display or output by I / O peripherals in a form useful for the human user of workstation 21. In addition, library 32 is also available for server 30 (and perhaps from workstation 21 over the local network, or a geographic network), and stores such file or file information references that may be useful in the system 20. Library 32 may be on another local network, or alternatively, accessible via the Internet or any other geographic network. It is contemplated that library 32 may also be accessible to other associated computers on the global network.
    Of course, the private or local memory resource where the measurements, library 32, and program memory 34 are physically located can be implemented in several
    22/82 locations accessible to the system 20. For example, this data and program instructions can be stored in local memory resources on workstation 21, on server 30, or in memory resources accessible over the network by these functions. In addition, each of these data and program memory resources can, by itself, be distributed among several locations, as is known in the art. It is considered that those skilled in the art will be readily able to implement the storage and retrieval of measurements, models and other useful information applicable in connection with this embodiment of the invention, in a manner suitable for each specific application.
    According to this embodiment of the invention, by way of example, system memory 24 and program memory 34 store computer instructions executable by central processing unit 25 and server 30, respectively, to perform the functions described in the present specification , through which a computer model of the causal interrelationships between wells in the production field can be 'generated from the actual measurements obtained from the' wells, and through which the model is evaluated and analyzed * · for finally determine the effects of proposed secondary recovery activities on production. These computer instructions may be in the form of one or more executable programs, or in the form of source code or high-level code from which one or more executable programs are derived, assembled, interpreted or compiled. Any of the variety of computer languages or protocols can be used, depending on how the desired operations are to be performed. For example, these
    23/82 computer instructions can be written in a conventional high-level language, either as a conventional linear computer program or arranged for execution in object-oriented form. These instructions can also be incorporated into a top-level application. For example, a web-based executable application can reside in program memory 34, accessible to server 30 and client computer systems, such as workstation 21, receive inputs from the client system, in the form of a spreadsheet, execute modules of algorithms : on a web server, and provide an output to the client's system on some convenient monitor or printed form. It is considered that those skilled in the art, having reference to this description, will be easily able to carry out, without undue experimentation, this embodiment of the invention in a manner appropriate for the desired installations. Alternatively, these computer executable software instructions can reside '' elsewhere on the local network or geographic network, or to '' download from higher or local level servers, via information encoded in an electromagnetic carrier signal via some network interface ; i or input / output device. Computer executable software instructions may have originally been stored on removable media or other non-volatile, computer-readable storage (for example, a DVD disc, flash memory, or the like), or downloadable as information encoded in an electromagnetic carrier signal, in the form of a software package from which the software instructions
    24/82 computer executable are installed by system 20 in the conventional way of installing software. Computerized system operation
    Figure 3 illustrates the generalized operation of system 20 in performing the analysis and statistical functions involved in evaluating the effect of potential secondary recovery actions by water injection, according to embodiments of the invention. As discussed immediately above, it is considered that the various 10 steps and functions in this process can be performed by one or more of the computational resources of the system 20 of executing computer program instructions residing in the available program memory, together with user entries, 'as appropriate. While the following description will present an example of this operation performed on workstation 21 of the system network arrangement 20, shown in Figure 2, it should, of course, be understood that the specific computing component used to perform particular operations may vary 20 widely, depending on the implementation of the system. Therefore, the description that follows is not intended to be limiting, particularly in identifying the components involved in a specific operation. Therefore, it is considered that those skilled in the art 25 will easily understand, from this specification, the manner in which these operations can be performed by these computing resources in various implementations and realizations. Therefore, it is considered that the benchmark for the performance of certain operations by the system 20 will be sufficient to allow these readers
    25/82 specialists readily implement embodiments of the present invention, without undue experimentation.
    In the high level flow diagram of Figure 3, the procedure starts with process 40, in which the measurement data related to the flow rates of the wells in the production field 6 under investigation are obtained and processed. As shown in the detailed flow diagram of Figure 4a, process 40 can be performed by initially importing this measurement data from the appropriate data source, from : process 50. In system example 20 shown in- Figure 2-; process 50 can be performed by obtaining data values corresponding to measurements obtained directly from flow meters and other sensors in the field through measurement inputs 28, and by obtaining historical measurement data stored in data library 32 and available for workstation '21 via network interface 28 and server 30. These measurement data obtained in process 50 can thus include historical flow measurements (including measurements for different phases of the multiphase flow) of each injector II to 15 and producer Pl to P7 of production field 6, the flow rates for these wells as calculated from indirect measurements in the wells (for example, from pressure and temperature measurements), as well as other well measurements referring to flow, such as downhole pressure (BHP) '. over time. It is considered that the length of time during which these measurements are obtained can be relatively long, covering months or even years. As is known in the art, changes in well counting (either just one type or both injectors or producers) in a production field often
    26/82 relations between the wells in the field, and alters the response capacity of existing and previously existing producers to the injection activity; as such, the measurement data acquired in process 50 and analyzed according to the embodiments of the present invention can be restricted to a specific time when the counting of injector and producer wells is constant. Unstructured or non-periodic data, such as data from fluid samples; well tests and chemical analysis, can also be incorporated into the specific time series recovered in the process : s-ó '50. The data obtained in process 50 will be recovered, or otherwise examined, as a time series of measurements according to embodiments of this invention.
    Process 40 also includes various filtering and processing of this measurement data, as convenient for analysis according to embodiments of the present invention, as performed in data filtering process 52 (Figure 4a). According to this embodiment of the invention, process 52 can be performed by the user of workstation 21, interactively by selecting certain data streams for analysis, such data streams that include one or more measurements (specific flow rates, BHP, etc.) of one or more of the injectors II to 15 and producers P1 to P7 of the production field 6. For the selected data streams, system 20 preferably processes the data to remove invalid values from the data streams (for example , measurements obtained by defective sensors, values for the days when the sensors were disabled, physically impossible measurement values, such as negative pressures, etc.), and filters
    27/82 the data to remove statistical anomalies. Such invalid values or statistical anomalies can be replaced, in the data filtering process 52, by interpolated values, calculated from data values surrounding the time series. This statistical filtering can be performed interactively via workstation 21, with the user selecting specific statistical criteria for the exclusion of anomalies, for example, observation : of histograms and visualizations of J series of measurement data as soon as processed. In addition, the filtering process 52 preferably adjusts or filters the measurement data on a regular and periodic basis, for example, with one measurement per day; for example, measurements corresponding to partial days can be adjusted to values corresponding to the output of an entire day. Reservoir barrel corrections, or some other normalization to a single base for manipulation of diopho '· can also be implemented in process 52, for example,' to compensate for substantial differences in fluid compatibility (for example, between water and gas in an alternative system between water and gas), and other minor but influential changes due to salinity treatment (for example, LoSal treatments).
    Back : to Figure 3, after obtaining and processing the : measurement data, in process 40, system 20 executes : process 42 is followed, in which injector events are identified in the processed measurement data. Generally speaking, the injector events identified in process 42 are changes in the flow rate of the injected fluid (gas, water, chemicals, or other
    28/82 fluids, or mixtures thereof) in injectors II to 15 of the production field 6 under investigation, and especially those changes in the injection flow rate, which can cause a response in the flow rates in one or more producers Pl to P7 in that production area 6. Other events, such as the initiation of alternating injection of water and gas into injectors, or changes in an output measurement, such as gas production or the gas / oil ratio (GOR) in one or a set of 'producers, can also be analyzed in this association. As will be described in detail below, for situations in which interpox effects (ie, action in a well that affects other wells) are of special interest, certain embodiments of the invention are able to filter out intrabo effects (for example, the effect of gas elevation or changes in the obstruction valve configurations in a producing well at its flow rate) that can mask the: interposing effects that one seeks to understand.
    Figures 4á and 4b illustrate the operation of process 42, in greater detail, according to an embodiment of the invention. In particular, process 42 involves identifying events from injectors II through 15 that are likely to be related to a response in one or more producers P1 to P7 in the production field 6. In this embodiment of the invention, process 42 begins with process 54 (Figure 4 a), in which the correlation of cross plots of the injector flow rate and the producer flow rate are shown on the workstation 21, which allows the visualization of the general relationship of the daily flow rate in a selected 1 Ij injector plotted against the daily flow rate in a selected producer P k , over
    29/82 days within a time interval selected interactively by the user. The way of selecting the producer P k and the relevant time interval is thought to be within the user's judgment, clarified by the measurement data obtained in process 40. For example, Figure 5a shows an example of a cross plot of the flow rate of the base fluid (i.e., the flow rate of the entire fluid) in the PI producer versus the flow rate of the base fluid in the switch II over a selected period of time. In this figure 5a, each data point corresponds to a specific day within the period of time in which the flow rate of the base fluid in both injector II and producer PI is different from zero. Workstation 21 or another computing resource in system 20 can further calculate a correlation coefficient, in the conventional way, to give the user a more in-depth view of the overall flow rate relationship. In the example in Figure 5a, the user can conclude that flow rates in injector II and producer PI are generally correlated, and that producer P1 is then a candidate for further investigation to identify injector injector events in II in the present case 42. Other injector-producer pairs can then also be investigated in case 54, as a result of which the user can include and exclude several pairs in further investigation. Other data streams, such as downhole pressure (BHP), downhole temperature, wellhead temperature, on both injectors and producers, can also be used in the analysis.
    30/82
    Process 42 then continues with process 56, in which system 20 performs an automated interactive process for identifying injector events. It is considered that these diverse approaches to the identification of the injector event can be applied according to this invention. A particularly beneficial approach in the injector event identification process 56, according to an embodiment of the invention, will be described below-with reference to Figure 4b.
    The process 56 of identifying begins with process 60, in which '21 workstation displays to the user a time series of measurements (as processed by process 52 des: scribed above) corresponding to the flow rate for a gun selected Ij. According to this embodiment of the invention, this time series displayed in process 60 is a time series of the injection flow rate over time. Alternatively, the time series displayed in process 60 may correspond to a different measurement, for example, downhole pressure over time. Figure 5b illustrates an example of such a time series of injection flow rate in window 61 of a monitor at workstation 21, acquired over a historical period of time. In this example, a certain amount of leveling by the mean was applied by system 20, to smooth individual data points at the illustrated flow rate for this selected injector Ij. Additional visualization tools can also be provided as a result of process 60, including, for example, a histogram tool illustrated in window 63, through which the user can
    31/82 view the distribution of the flow rates of the time series shown in window 61.
    As shown in Figure 5b, interactive tools are also provided to the user by workstation 21 in window 65, through which process 62 can be performed by system 20 to identify potential injector events in the currently selected time series. In window 65, the user can define various criteria by means of which '20 system identifies potential events in this process 62. For example, as shown in Figure 5b, the user can select the sampling period (interval) between 'moments of time in the presented time series in which 5 retrospective and prospective instant gradients are 'calculated, together with the duration (shelf) over which each of these gradients has to be calculated. The threshold values by which events are identified are also shown in window 65. For example, as shown in Figure 5b, a high threshold value of about 250 is operative; times when a change between retrospective and prospective gradients exceeds this value will be identified as potential events in response to the user's action on the Find events like this button in window 65. Alternatively, the user can enter a series of events to be identified in the time series displayed in window 61 (for example, events 20, as shown in Figure 5b); after the user clicks the Find threshold button, the threshold values will be calculated. In both cases, the potential injector events are shown in window 61 as vertical lines at specific points that cover the flow time series over time. It is considered that
    32/82 the user can interact with system 20 in this way to identify potential injector events for subsequent analysis. Of course, other approaches to carrying out the event identification process 62 may alternatively be implemented. A particularly useful approach for identifying significant changes in the gradient in time series representations will be described in detail below, in association with the identification of producer events; this approach can also be used in process 62 to identify potential injector events.
    Returning to Figure 4b, system 20 then executes process 64 to allow the user to view the selected injector events as identified in process 62, and to view the possible responses to these injector events by producers Pl to P7 in the same field production process 6. This process 64 allows the user to determine whether the identified potential injector events can ! cause a corresponding response in the flow rate produced. According to this embodiment of the invention, visualization process 64 presents a focused view (in time) of a selected injection flow rate, in combination with flow rates corresponding to uif or more producers from Pl to P7 more or less at the same time ·, to assist in this verification.
    Figure 5c shows an example of a flow rate time series, displayed on workstation 21, which includes potential events identified by process 62, in that time series. As in Figure 5b, potential events are indicated by vertical lines. The flow rates illustrated in Figure 5c correspond to the
    33/82 specific sampling as identified in process 62, for example at a time every 31 days, selected in table 65 in the monitor of the example in this figure. In this example in Figure 5c, the user interactively selected the event at time t k for viewing. Also at this point in the interactive process, the user could have selected one or more time series to investigate possible responses to this potential injector event at time t k of the available time series.
    The visualization process 64 according to this embodiment then generates a display of the flow of the selected injector (for example, the injector Ij in this example), together with one or more response time series selected by the user. For example, the selected response series can be one previously found, in the cross-plot correlation process 54, to have a reasonable correlation with the injector Ij. Process 64 generates a visualization of the selected time series so that the user can easily compare the forms of the potential response time series with the shape of the selected potential injector event, to determine whether sufficient plausible correlation is present and to investigate the case of the injector by subsequent processing (described below). To perform this visualization, system 20 considers a relatively short period of time on both sides in the time of the selected event t k (such period of time is user selectable), normalizes the amplitude of the selected time series within the period of time in question, and also normalizes the times when corresponding changes in the gradient of each of the selected responses occur. The figure
    34/82
    5d illustrates an example of a visualization generated in that process 64, according to an embodiment of the present invention, for the event at time tk of the chosen potential injector, as shown in Figure 5c. As is evident, in this plot overlay of Figure 5d, the average for each of the plots in the selected time series is averaged over the same sampling period of the Ij injector flow rate, · the normalization of time shifts forward the responses presented by plots P x , in order to coincide with the change in the gradient of the flow rate of the injector Ij rio time t k (time 0 of Figure 5d). Of course, in reality, there will be a certain finite delay (usually in days) between the potential injector event at time t k and any effective response. In this example, the visualization of figure 5d extends from 60 days before time t k to about 120 days after time tk- As shown in Figure 5d, a response curve closely follows the rate series curve injector flow rate, · others vary in their fidelity with the injector flow rate.
    After the user has finished analyzing a potential injector event via process 64, as shown in Figure 5d, system 20 operates in order to receive input from the user that indicates whether the potential injector event has been verified (ie , seems to invoke a response from one or more of the producers) or rejected (that is, it does not present a response in a producer, therefore, it does not correspond to a real injector event or corresponds to an event that does not need to be further studied), in the process of Figure 4b. This interaction with the user in processes 64, 66 is repeated for each of the potential events of
    35/82 injector identified by process 62 for the current injector Ij, to the extent desired by the user. Upon completion of the analysis of potential injector events in an injector, decision 67 is executed to consult the user if there are additional injectors for analysis. If so (decision 67 is yes), then another injector Ij is selected in process 68, and the process is repeated for that injector Ij from process 60.
    Returning to Figure 4a, after all the desired injectors have been analyzed by process 56 (decision 67 is no), the process 42 for identifying the injector event is completed by exporting the data that indicates the various injector events verified. These exported data will include the injector identification and the time that the verified event occurs, as well as the magnitude of the event. More specifically, the magnitude of the event is an indication of the size of the event relative, in a functional sense, to the change in the cumulative injection flow rate over a period of time selected by the user (i.e., a shelf period). The inclusion of a measure of event magnitude can serve as a basis for the selection of subsets of the complete set of injection events. In addition to being based simply on the magnitude of the event, this selection can consider the consistency of the magnitudes of the events in each producer in response to injection events; those producers who do not respond consistently to major injection events can be considered to have a less reliable connection than those who respond consistently to these events. Other data, such as the time delays for the corresponding responses (known to the
    36/82 normalization carried out in connection with the 64 process), and other attributes of the corresponding responses, can be included in the exported data. These data are exported in a format suitable for use by system 20 in process 44 (Figure 3), to detect producer events and the association of producer events with injector events, as described below. For example, the format of the exported data can be a spreadsheet.
    The particular implementation of processes 40, 42 in identifying potential injector events may vary from that described above in connection with Figures 4a and 4b. For example, importing and filtering process data 50, 52 can be performed by time series of individual injectors flow rate after selection by the user (that is, after selection process 68 at each pass through process 56), if applicable; alternatively, as suggested by the description above, data import and filtering can be performed for all injectors of interest prior to identification by process 42. These and other variations in the implementation of processes 40, 42 will be evident to those skilled in the art, having this specification as a reference.
    In this respect, such a variation in the implementation of processes 40, 42, more specifically, as a preparatory step for the analysis of the injector event, is to identify isolated events in the time sequence of the injector population. Since injectors are often subject to simultaneous changes under the control of the operator (human or automated), or as a result of mechanical, electrical or other interruptions that cause the loss of injection in all or part of the injectors, it can
    37/82 it is difficult to resolve which of the injectors is potentially responsible for a change in a production well. On the other hand, isolated events from individual injection wells are not subject to this uncertainty, and are therefore relatively more revealing about the connection pathways in the reservoir. Therefore, the automatic detection of isolated injector events, unlike events common to some or all of the injectors, can be very useful in assisting research among producing wells that are plausible to respond, and can be performed in the system and method of achieving the invention, as will be described below.
    In an approach according to the concept of the invention, the search for isolated events from injection wells extends the marking of an isolated event to individual wells, taking into account the directions of changes. As the expected physical behavior of injection fluids is an increase in production along with an increase in injection rates and a decrease in production with the decline in injection rates, an increase in isolated injection in one injector with a simultaneous decrease in injection in several others injectors can be considered as an isolated event, and saved for comparison patterns with the production variation (either visually, as described above, or through numerical scores as will be discussed in more detail below). In another variation, the compensation for the travel time between wells, allows differences in distance between producers and injectors, to be applied in tests of perceived simultaneity in each of the target producers. This travel time compensation is considered to be especially useful when applied to resolved data
    38/82 more often than on a daily basis (for example, every 3-6 hours).
    Another improvement in the isolation of injector events identifies the periods during which no injector activity occurs, particularly after truly isolated or pseudo-isolation (that is, only when the other contemporary injector events are all in the opposite direction to another isolated injector event. ). As these periods are devoid of various other 'masking' events, suggestions for plausible connections from injector-producer well pairs can be detected more easily during these quiet periods. Although the numerical scores of these isolated events are considered to be likely to be weak, due to the low incidence rate of such events, these isolated events are likely to generate useful links that can guide the research path.
    Back to Figure 3, after the completion of the identification of injector events in process 42, system 20 then analyzes the measurement data pertinent to the production flow rate of production wells from Pl to P7 in the production field 6 (Figure 1) in process 44. According to embodiments of this invention, the measurement data analyzed in process 44 can include direct measurement of flow rates at each of the producers P1 to P7 of interest, flow rates allocated to the individual producers calculated from mixed flow measurements, calculated or estimated flow rates for each phase of interest from the measurement of a multiphase flow, or flow rates calculated based on temperature, pressure or other indirect measurements (proxy)
    39/82 at the bottom of the well or at the wellhead of each producer. In addition, process analysis 44 can be performed on other measured or calculated flow rate values, such as downhole pressure (BHP). In addition, as will become apparent from the description below, the measurement data regarding the flow rates of injectors II to 15 in the production field 6 can also be analyzed by process 44, as well as information derived from the process 42, in which injector events were identified, and further characterized if desired. Measurement data can be corrected for reservoir barrels to consistently normalize analyzes, both within the flow characteristics of an individual well despite changes in GOR and water cutoff, and in relation to other producer and injector wells . This measurement data of greater frequency, compared to reconciled and allocated well flows, allows the resolution of intrapace events with high precision over time. In doing so, entire days of allocated production flow do not need to be masked (that is, removed) from analysis in order to eliminate the intro-effects. As a result, measurement data for a larger global proportion of the time period under analysis can remain available for the identification and development of interpoort associative connections and relationships.
    As is known in the art, wells are subject to many and varied changes resulting from changes in the independent variables of the well, as typically made by a human operator. However, the intervention of automated actions, whether initiated by control systems or
    40/82 safety or by human operators, causes frequent variations in production and other dependent variables (for example, pressures and temperatures), for reasons not primarily due to the interaction with the injection wells.
    So, another step of useful preparation corrects The production allocated to these effects before analysis in effects interventions. How simple example, if a well operated for 12 hours in a certain day, its flow allocated would probably be about half gives
    whole day operation. Multivariable linear regression can be used to correct all changes in the independent variables, and the file resulting from corrected flows sent to the data filtering and anomalous removal steps, according to embodiments of the invention. Anomalies that can distort linear regression, for example, zero hour production or zero obstruction openings, cannot be usefully corrected to 24 hour values and therefore must be treated accordingly. Values that are physically unrealistic or used as error codes (for example, negative valve openings) can be excluded.
    As is known in the art, wells that have been in a non-flow condition for a period of time will regain pressure on recovery, after which their flow will tend to be higher than the expected values over a period of time. Multiple linear regression can correct production to modal values, or expected these independent variables, for example, using an exponential correction for periods between zero days online since restoration and an appropriate number of days for the well to return to normal state
    41/82 differential pressure between the reservoir and the well. The additional parameters that describe the closing period can further improve this correction.
    Referring now to Figure 6, the operation of system 20 in carrying out process 44 will now be described in detail. With respect to producer measurement data, process 44 begins with process 70, in which system 20 retrieves measurement data in the form of, or suitable to be arranged as, one or more time series for each producer Pl to P7 of interest. This measurement data is obtained from the appropriate data source, including acquiring recent measurements obtained directly from flow meters and other sensors in the field via measurement inputs 28, and by retrieving historical measurement data stored in the data library 32 and available on workstation 21, via network interface 26 and server 30. As mentioned above, measurement data obtained in process 70 may include historical flow rate measurements (including measurements for separate multiphase flow phases) of each producer Pl to P7 of production field 6, flow rates for those wells as determined by indirect measurements in the wells (for example, from temperature and pressure measurements), as well as other well measurements such as downhole (BHP).
    It has been observed, in connection with this invention, that the time series representations of cumulative production from producing wells form a particularly useful set of measurement data for the purpose of evaluating secondary recovery actions, in accordance with embodiments of the present invention. Accumulated data
    42/82 of production are useful in this regard, since such data naturally reflect the reduction of pressure in the reservoir of a production field over time, and the corresponding typical decline in flow rate. Thus, for the purposes of this description, the time series measurement data retrieved in process 70 will be referred to as cumulative production data. Of course, as described above, other measurement data and calculated values, as the case may be, alternatively or additionally, can be retrieved and analyzed according to embodiments of the present invention.
    As in the case of obtaining measurement data for injectors II to 15, it is considered that the period of time in which measurements are obtained can be relatively long, up to months or years. As mentioned above, since changes in the well count normally alter the injector-producer relationships in the field, the measurement data obtained in the process 70 and analyzed according to embodiments of the present invention can be restricted to a specific period in which the injector and producer wells is constant, and repeated for each well counting period over the period of time of interest. Process 70 preferably also includes various filtering and processing of these measurement data, in a way that is suitable for analysis according to embodiments of the present invention, as described above. In addition, the recovery process 70 may correspond, in whole or in part, to the processes 40, 42 described above in connection with the initial recovery of measurement data before the identification of injector events; alternatively, process 70 can apply
    43/82 different or additional selection or filtering criteria, as desired. Other pre-processing of the retrieved measurement data can also be applied within process 70. For example, the measurement data for a given well can be normalized to modal values of the well's independent operational parameters, so that the intro-flow effects during production are automatically compensated before the establishment of indicative events for interpointe communication. More specifically, the performance of each well can be linearly regressed against its own variables, such as, among others, the position of the obstruction, gas or other elevation parameters (eg flow, pump speed, etc.), and hours in production. When selecting an input from each pair of correlated inputs (for example, inputs with correlation> 0.8), the measurement of the well flow can be corrected to the expected value in the absence of variation in the intrapace parameters in relation to its modal value.
    In this embodiment of the invention, the time series data obtained in process 70 for one of the producers P1 to P7 is analyzed to detect potential events of producers through a gradient analysis in process 72. In general, this process 72 of analysis Gradient Analysis analyzes the rates of change over time over a period of time at a selected point of interest, to determine whether a statistically significant change in the gradient of measurement values occurred at that point in time. Such significant changes in the measurement data gradient (for example, reflecting changes in the production well flow)
    44/82 may indicate an event that is of interest in assessing the effects of injection in one or more injectors in the field. More specifically, as is known in the art, significant variations in the rate of change in the outflow rate of a producing well will occur in response to changes in the injection speed in the injector of the same production field, if significant connectivity between the injector and producer is present in the subsurface. As discussed above, it is these interposing effects that arouse interest in connection with this invention, because knowledge of the interaction between injectors and producers is important to optimize the management of the reservoir through secondary recovery actions. On the other hand, the intrapoço effects of the gas lift, restriction valve configurations, and similar actions in the producing well, as reflected in the changes in the flow of that well, are of less interest for the purposes of this invention; in fact, in some cases, these intro-effects can impair the visibility of the injector-producer interaction that is being optimized.
    With reference to Figure 7, the operation of the system 20 in carrying out the analysis process 72 according to an embodiment of the present invention will now be described in detail. As will become apparent to those skilled in the art, having this specification as a reference, the way in which the process 72 and performed according to this embodiment of the invention has greater sensitivity for detecting interposing effects (such as injector-producer relationships) in combination with reduced sensitivity to intrapace effects that are of less interest in secondary recovery.
    45/82
    In accordance with this embodiment of the invention, ο process 72 of gradient analysis is initialized in process 86 with the selected values of a gradient of duration kl, a duration of calculation of the mean k2 and threshold values τΐ, 12 for use in the operation of the process 72. It is considered that these initial values will be selected based on injector event attributes, as indicated by the injector event identification process 42. Alternatively, these initial values can be based on results of past optimization, characterization of this or similar production fields, or based on theory. Alternatively, it is contemplated that one or more of these values can be varied over 72 iterations of the process, to improve the robustness of the statistical optimization values for group m. In process 88, the measurement data time series for a given producer P k is selected, as well as a point in time t 0 along that time series at which the analysis should begin.
    In process 90, system 20 assesses a regressive gradient in the time series of measurement data at time t 0 selected from the kl samples prior to that time. Some criteria can be applied to this regressive gradient calculation, which includes a minimum number of valid data points within these kl samples. For example, if kl is initialized to seven days, then a minimum of four samples valid in the previous seven days may be required. Process 90 is performed by system 20 according to a conventional best fit or curve curve algorithm, such as least squares, and a correlation coefficient (eg R 2 ), or another measure of data fit to the line
    46/82 regression from which the gradient is determined, is calculated to quantify the degree to which the data points fit into the regression line. An alternative statistical test suitable for process 90 is a two-tailed t-test, for which a user-selected criterion p is used to determine whether a genuine slope change has occurred.
    In decision 91, system 20 assesses whether the adjustment of the regression line at the moment to is significantly poorer, in the statistical sense, than the adjustment of the data to the regression line calculated at the time of the previous sample. If not (decision 91 returns no), decision 95 determines whether the time series analysis is complete or whether additional points remain in the time series to be analyzed. If the 95 decision determines that such additional points remain (its result is no), the time of interest t 0 is advanced (process 96) and process 90 is repeated. For the first pass through process 90, decision 91 will, of course, be useless, and process 90 will be repeated at the next point in time throughout the historical series. If, however, the measurement data adjustment, including the current data point at time t 0 degrades significantly from the adjustment at the previous point at time ti, this poorer adjustment may indicate a response in producer P k to an event injection.
    According to this embodiment of the invention, therefore, decision 91 determines whether the adjustment measure (for example, the correlation coefficient) of the measurement data (for example, cumulative production) for the regression line is poorer in the moment as of the previous point in time tj to a significant degree. Per
    47/82 example, decision criterion 91 can determine whether the correlation coefficient R 2 (t0) <0.97R 2 (t-2). If so (decision 91 is yes), system 20 then executes process 92 to calculate a cumulative production gradient (or other attribute of the measurement data under analysis) over sampling points kl ahead of time t 0 . The number of sampling points ahead of time, over which the progressive gradient is calculated, may differ from the number of sampling points over which the regressive gradient is calculated in process 90, if desired (and, depending on the data available for that sampling period).
    Figures 8a to 8c illustrate an example of the operation of processes 90, 92, for a cumulative production sample data set from producer P1 over an interval of several days. In Figure 8a, the result of a previous example from process 90 is illustrated by means of a regression line for the regressive gradient of the six data points, including time t- 2 and the previous five samples. As shown in Figure 8a, this previous occurrence of process 90 performed a best-fit least-squares regression for a line with gradient gradient gradient Aback (ti) · A correlation coefficient R 2 (tJ was also calculated in this case of process 90 for time t_ 2 and its previous samples In Figure 8b, the result of process 90 at time t 0 is illustrated, with a regression line shown for time t 0 and the previous five data points. slope of this regression line is the regressive gradient A BA cK (to), and the data fit for this line
    48/82 regression is indicated by the correlation coefficient r 2 (t 0 ). As is evident from Figure 8b, a significant increase in cumulative production at producer PI occurred at time t 0 . For the purposes of this example, this instantaneous increase in accumulated production at time worsens the adjustment of the regression line for time t 0 of that made at time ti, by a value that meets decision threshold 91 (ie decision 91 Yes it is). As a result, process 92 is run for the data at the tor moment to derive a better regression fit for cumulative production at the time to and over the next five samples in time, to help determine whether this instant increase in time to can constitute a producer event Pl. The result of process 92 is illustrated in Figure 8c, by the regression line that extends forward beyond the time to · This regression line has a progressive gradient slope A FWD (t 0 ). As is evident from Figure 8c, the progressive gradient A FWD (t 0 ) at time to has a steeper slope than the regressive gradient Aback (to) at that time.
    Referring to Figure 7 again, once system 20 has calculated a progressive gradient over the next kl samples of the current analysis time to process 92, decision 93 is then executed to determine if the difference between the progressive and gradients regressive at the moment to exceed the useful threshold (defined in process 86). For example, the threshold τΐ may correspond to the average increase in accumulated production over the respective time periods kl, divided by five. If the change in slope between the progressive and regressive gradients exceeds this limit τ 1 (for example, if IA F wd
    49/82
    Δβαοκ I> tl) <a yes decision 93 returns a process 94 , and calculates a normalized value abnorm differential gradient (to) <normalized and stores this value in the memory associated with the time t 0. For example, the normalized value of the differential gradient A in rm can correspond to an algebraic value (the sign indicating the direction of the gradient change at time to), with an amplitude corresponding to the ratio between the difference between the progressive and regressive gradients for the threshold τΐ. For example, process 94 can simply calculate:
    _ A fi y O (t 0 ) - A gzÍCfÍ - (t 0 )
    This value can be rounded up to the nearest whole number, if desired, for easier storage and calculation. This value allows events to be detected on a normalized basis in relation to the threshold τΐ. The control then proceeds to decision 95 to determine whether the time series has been fully evaluated. Decision 95 is also executed if the change in slope does not exceed the τΐ threshold (decision 93 is no), as the change in slope is considered not to correspond to a potential injector-producing event.
    After completing the analysis of the time series for the producer P k (decision 95 is yes), system 20 then smooths the event over time, starting with process 100. According to embodiments of the present invention, this smoothing over over time converts significant gradient changes in time series measurement data (for example, significant changes in the rate of change in cumulative production) from a large-scale change representation to a change with great effect over time.
    50/82
    It was discovered, according to this invention, that this spreading over time makes it easier to distinguish between large and small events, and also improves the system 20's ability to detect events, given the uncertainties in the delays between the injector and producer events typically observed in real production fields. In addition, it has been discovered, in accordance with this invention, that the approach described above in identifying potential producer events by analyzing changes in the gradient, especially in combination with process time spreading 100 and following to be described below , tends to filter the effects of first order of intraboating actions in the production field, such as gas elevation, position changes in the restriction valve, and the like, which are carried out in the production well itself. This intrapace filtering takes place regardless of whether the allocated flow data have been previously adjusted for known variations in the independent well variables (for example, in-line hours, constraint position, gas lift rate, time since restart, etc.), as discussed above.
    According to this embodiment of the invention, process 100 is then performed for the selected producer P k . The time series of differential values of normalized gradients A norm for which the producers P k are recovered, and the moving average of the normalized differential gradient A nO rm is calculated for k2 samples in the surrounding time or else including a sampling time t x ; the duration value of k2 is one of the values initialized in process 86, and is selected based on previous observation, characterization, or theory. In the decision
    51/82
    101, system 20 evaluates, so that the absolute value of the mean threshold 12. Threshold 12 is initialized in process 86, characterization, or the desired number positive value realization of the producer, in this example, how much invention, as not only but also the direction (ie, largest multiples event can be etc., potential producer. In addition iterations of the 100 process of an association carried out throughout to improve the> of the event.
    ; according to this
    Decision 101 compares each value as an algebraic value, against each of the If the AVGinon moving average. (tx) n0 positive moment greater than value + 1 at time t x AVGAnorm (tx) has value system 20 assigns the AVGA moving average nO rm threshold -12 and the threshold +12, oat x in process 102.
    Figures 9a to 9c operation of the embodiments of the invention.
    the time t x of current analysis, mobile AVGA nO rm (t x ) exceeds the similarly defined or previous observation, in order to calculate τ2 it takes either a negative starting point, in this injective theory, or it is adjusted of events. The threshold an analysis value considers the magnitude, flow, lower flow) of that, if desired, of smoothing in the joint time of K2 values, robustness of the identification and realization of the invention, the AVGA moving average nO rm (tx) thresholds + 2i , -12. has value, system 20 assigns the moving average -12, threshold +12 in process 104; if the negative is less than the threshold value -1 at x in process 106. If, „+ - has an antro value o (t x ) at time t x has vd_L ^ system 20 assigns the value 0 illustrate a simple example of 100 a 106 according to this Figure 9a illustrates an example of a time series of normalized gradient diphtherial values for a producer Pk- «and an example from Figure
    52/82
    9, a potential event that corresponds to a negative variation in the gradient (by an amount of twice the threshold τΐ, or -2) was identified at time t x - 5r and an event that corresponds to a potential positive change in the gradient (from a total of four times the threshold τΐ, or +4) was identified at time t x . None of the other times under analysis correspond to a change in the gradient above the threshold τΐ.
    Figure 9b illustrates the result of process 100, in which the moving average AVGA norm (t) over five sampling periods centered around each sampling time (ie, k2 = 5), was calculated. As shown in Figure 9b, the AVGA norm (t) value of -0.4 resulted from calculating the average of the -2 values of A norm at time t x _ 5 , with the value of -0.4 spread over the five sampling times for which the five-period centered average would include time t x -5 (no other changes in the gradient should be present within that five-period time window). Likewise, an AVGA norm (t) value of +0.8 results from calculating the average of the +4 value of A nO rm at time t x , with this value of 0.8 spread over the five sampling times in which the five-period centered average would include time t x (no other changes in the gradient should be present within that five-period time window). In the example in Figure 9b, the positive and negative thresholds + τ2, -τ2 are shown, with values of +0.5, 0.5, respectively. As is evident from the comparison of Figures 9a and 9b, the changes detected in the gradient at specific sampling moments tx-5, t x were effectively spread over the time that involves the sampling points. This spreading over time facilitates
    53/82 event detection, in a way that carries the strongest changes in the gradient more strongly.
    Figure 9c illustrates the results of decision 101 and processes 102, 104, 106 of process 72 in this embodiment of the invention. The AVGA nO rm (t) values spread from -0.4 which involve the time t x s each fall below the negative threshold -τ2 (which is -0.5, in this example), and as such process 102 is applied to each of these sampling points, which sets these values to 0. But as the AVGA norm (t) +0.8 values spread over time around the time t x exceed the positive threshold +12 (+0.5 in this example ), process 104 is performed to set the +1 value for each of these sampling times, as shown in Figure 9c. This decision threshold 101 according to this embodiment of the invention thus serves to filter out minor changes in the gradients of the measurement data, while preserving the effect of scattering in time to detect the presence of events, as will be described in more detail below.
    Referring again to Figure 7, decision 107 determines whether additional points of time over the time series of normalized gradient differential values Anorm for this producer P k remain to be processed; if so (decision 107 is yes), then the analysis time t x is advanced (process 108) and the next moving average is calculated. If not (decision 107 is no), process 72 is complete for that producer P k .
    While process 72 is described above as averaging and time smoothing of identified producer events, it is considered that similar time averaging and smoothing calculations can be applied to
    54/82 injector events identified in process 42 described above, to facilitate the association processes described below. Other measures to facilitate the analysis can also be included at this stage in the overall process. Such an additional process is a check to ensure that the events recorded and stored for a producing well do not include any events that are a consequence of closing or restarting that same well, as events of this type are clearly the result of operator intervention. In the event that producer-to-producer interactions are to be analyzed, however, events of total closure and restart in producing wells will be maintained as causal events (the response in other producers is of interest), but not as events of interest. answer. In addition, any identified events that occur in a well during closure can be filtered out at this time.
    Upon completion of process 72 (Figure 6), optional process 73 can be performed to further facilitate the identification of producer events. In process 73, system 20 operates to flicker the producer events detected in process 72 in time. As is known in the graphic processing technique, image flickering can serve to improve the fidelity of an edge of a displayed image, essentially by eliminating the effects of pixelation (that is, errors due to sampling) in the displayed image. Likewise, time flickering of the events detected in the time series that results from process 72 can reduce the possibility of further identification of the event and the causality analysis will miss the true event of
    55/82 producer due to an injector event, due to rounding error, etc. According to this embodiment of the invention, the jitter process 73 can be performed simply by creating additional time series of detected events (for example, digital representations containing data corresponding to the binary algebraic result shown in Figure 9c), with each time series additional time shifting events for a selected period of flicker (for example, at the end of a sampling period) in any direction. Each of the additional time series, together with the original result, can then be processed in the manner described below.
    Continuing with the shaking process 73 (if performed), the potential events of detected producers ρθ] _θ3 processes 70, 72 according to this embodiment of the invention are ready for causal analysis regarding potential injector events. As shown in Figure 6, candidate injector events identified in process 42 are retrieved in process 74, along with all attributes determined in process 42. As mentioned above, these attributes can include such information, for each injector or injector event, such as delay times observed by the user or system 20, between the injector event and potential producer events that resemble the injector event (for example, as identified in visualizations such as those shown in Figure 5d). The identities of these P1 to P7 producers identified as having similar corresponding events can also be retrieved, if desired. In process 76, system 20 selects a range of delay times for analysis in relation to injector events,
    56/82 within which producer events are expected to occur (if any). Process 76 can be obtained by system 20 automatically from the delay time attributes detected in process 42 and retrieved in process 74. Alternatively, a user of system 20 can enter or adjust the range of delay times to be analyzed based on in an improved visualization that focuses on isolated events and intermediate periods free of injection events, as described above; such a visualization can reveal the time intervals of inter-communicating time by plotting time lines adjacent to the injection and production data.
    The precise size and timing of events identified in time series data from producing wells is sensitive to the choice of parameters used. Effective default values for the parameters can be derived from the intrinsic values and the variability of the time series data itself. However, it has been recognized, in connection with the present invention, that parameters can be validly varied across the range of reasonable values. According to an alternative embodiment of the present invention, the process can be carried out over a number of scenarios that explore the entire matrix of reasonable value ranges, for all parameters, with the set of results throughout these post-processed scenarios to eliminate scenarios that clearly result in unviable numbers of events (that is, the noise level events in the process data will be resolved). The post-processed results can then be managed as a set of event models to locate
    57/82 isolated events as described above for injection wells, while injection data is analyzed in a similar way as described above for production data. Alternatively, a set of counting scores can be generated, as will be described below.
    After the recovery of both producer events (process 72) and injector events (process 74), system 20 below runs process 78 to identify those producer events that are within the selected range of causal delays for each of the injector events. It is envisaged that the various approaches to identifying paired injector-producer events within the causal delay time interval, and the attributes of the paired injector-producer events, can be used in connection with this invention.
    A suitable approach for use in connection with embodiments of the present invention is described in U.S. Patent No. 7,890,200, issued February 15, 2011, entitled Process-Related Systems and Methods, frequently mentioned in this instrument and incorporated by reference, in its entirety, in this patent application. According to this approach, the processed time series of injector measurements and the producer events verified against the thresholds and smoothed over time identified in process 72 are considered as process variables that have values that vary over time. The causal relationships between the process variables are identified by the process of US patent No. 7,890,200, with the help of indicating the injector events as cause events, and the corresponding producer events as the response events.
    58/82 correspondents. As described in this US patent No. 7,890,200, the confidence levels for the identified pairs of injector-producer events are calculated, along with any other statistical attributes, in order to be useful in the rest of process 44 in Figure 6.
    A generalized counting approach to identify injector-producer relationships in process 78 will now be described with reference to figure 10, starting with the selection of an injector Ij for analysis, in process 110. In this description, each of the injectors II to 15 in the field of production 6 under analysis will be interrogated sequentially, it should be understood that such data analysis can, if desired, be conducted in parallel. In process 112, an injector event is selected from the measurement data time series for the selected injector Ij, 'alternatively, if averaging, time smoothing, or another filtering process 72 is applied to injector events, the time series of injector events will correspond to the result of such processing. These injector events can be either increases in injection flow or decreases in injection flow. Once a specific injector event is selected in process 112, the time series of event indicators produced in process 72 for each producer Pl to P7 are then analyzed in process 114, along the causal delay interval selected in the process 76 to identify producer events (both positive +1 and negative negative -1) that occur within that causal delay interval that coincides with the injector event. Decision 115 is then enforced by the
    59/82 system 20 to determine whether additional injector events for the selected injector Ij continue to be analyzed; if so (decision 115 is yes), another injector event is selected in process 112, and the process
    114 is repeated. After completing the analysis of all injector events for the selected injector Ij (decision
    115 is no), system 20 then executes decision 117 to determine whether additional injectors still need to be analyzed. If so (decision 117 is yes), processes 110, 112, 114, and decision 115 are repeated for the next injector.
    After completing the identification processes for all injectors (decision 117 is no), process 116 is then executed by system 20 to count the producer events identified by process 114, for each injector-producer pair. The resulting counts can include the following values, for each injector-producer pair (Ij, P k ), such as:
    • number of causal events in injector Ij • number of response events in producer P k in response to causal events in injector Ij • number of causal events in injector Ij with no response in producer P k , and response events in producer P k to other events in different injectors. number of positive response events (increased flow) and negative response events (reduced flow) in producer P k in response to positive causal events (increased flow) in injector Ij • number of positive response events (increased flow) flow) and negative response events (reduction of
    60/82 flow) in producer P k in response to negative causal events (decreased flow) in injector Ij and so on.
    After counting process 116, system 20 performs statistical analysis process 118, to provide various statistical measures related to the responses of producer-injector pairs identified in process 114. The various statistical measures calculated in process 118 may include one or more of following options.
    • support (and percentage of support) of responses from producer P k attributed to causal events in the Ij injector • level of confidence that the association exists • chi-square parameters referring to the association • the total score or figure of merit for the strength of the association • surprise statistics for the association and so on. It is considered that experts in the art, having reference to the present specification, will be promptly able to select and apply the aforementioned statistical measures that they define as useful in assessing the strength of the identified injector-producer associations, depending on the specific field of production 6 and of experience in secondary recovery analysis according to embodiments of the present invention, and in other ways.
    Other operations can additionally be included within the identification process 78 performed by system 20, in accordance with embodiments of this invention. As mentioned above, the gradient analysis used to identify producer events in process 42 provides the benefit of first order filtering,
    61/82 intrapoço, of effects that appear as possible producer events caused by injection. These first order effects tend to as changes be removed from the analysis, and do not appear or in any other possible attribute to a significant event in the analyzed production. However, 'true response in an injection production well may occur at the same time as one due to a change in gas lift, throttle valve etc. In that injection also masking to be an intrapoço effect, change in the case position, the real response to the event of the intrapoço effect, would be filtered with true answer it is considered that, in 78 it can include the synthetic in reality, therefore, of the producer. In connection with the present invention, the insertion of an injector event of medium delay time. For example, producer process, statistics evaluated causal relationship well-behaved for injector-producer pair, but an event identified in the delay time and specific injector, increase in the elevator the insertion of a synthetic event in an estimated magnitude in process 78 can compensate for the masking of the true producing event for such a first-order effect, to compensate for the degradation in the statistical association due to the presence of a first-order intrapace effect.
    In addition, process 78 can also identify producer-producer associations, in which an event of change in the outflow of a producer Pr is determined to be strongly associated with an event of change in one or both of them in the process counts. a being in process 114 and 118 may indicate these events to producer may not be expected for an event due to some action (for example, gas) in the production well itself. A> synthetic
    62/82 flow output from a producer other than P m , and not in response to an injector event. The knowledge of such producer-producer associations can be analyzed by system 20 to further characterize the reservoir; alternatively, the system 20 and its user can disqualify or completely ignore events caused by producer-producer associations, if the objective of the general process is to assess the potential injection actions on the output of the production field 6 in isolation from the interproductive effects.
    As shown in Figure 6, in process 81, system 20 can optionally display a view of the injector-producer events identified in process 78. Figures 11a and 11b illustrate examples of such visual effects. Figures 11a and 11b each show (from bottom to top) indications of time series of events: injector II being turned on (l01_inj. ON), injector II being turned off (l01_inj.OFF), increased production at producer Pl (P01_prod. INCREASE) and reduced production at producer Pl (P01_prod.DECREASE). The presence of an event throughout each of these time series is indicated by a rectangle, with the length of the rectangle corresponding to the duration of the event. Figure 11a illustrates associations identified between increased injection events (I +) in injector II and increased production events (P +) in the PI producer by the vertical lines (for example, the E01 association) that link the events. These event indications can also optionally include a visualization of the strength of the event through color or shadow. Figure 11b illustrates the same four time series of injector II and producer Pl events, with associations
    63/82 between injector II events being turned off and reduction of production events at producer P1 indicated by vertical lines. Again, decreases in production events associated with other injector events are indicated in Figure 11b by vertical lines, which are not linked to an injector II event. These views as shown in process 81 allow the user of system 20 to visually check the identified associations; it is considered that the user can also interact with these visual effects, for example, to confirm or reject specific associations.
    Turning again to Figure 6, process 80 is now performed by system 20 to determine the measure of association strength for each injector-producer pair. The number of injector-producer pairs will, of course, correspond to the product value of the number of injectors by the number of producers (for example, production field 6 in Figure 1, five injectors II to 15 and seven producers Pl to P7 result in thirty-five injector-producer pairs).
    An example of classification process 80 according to an embodiment of the present invention is illustrated in Figure 12. In this example, the population of injector-producer pairs {!>; P k } is first classified according to its polarity behavior, which evaluates the polarity of the effects on the producer P k in response to events of both polarities in the injector Ij. The first group of injector-producer pairs 121a {Ij, · P k } includes those for which producer P k exhibits increased production flow events in response to increased injection events from the Ij injector, and also exhibits reduction events production flow in response to injection reduction events in the
    64/82 injector Ij (ie, both up-up and down-down behavior). The second group 121b includes the injector-producer pairs {Ij; P k ) for which the producer P k exhibits an increase in production flow events in response to the increase in injection events in the injector Ij, but which do not show a reduction in production flow events in response to the reduction of injection events in the injector. injector Ij (ie, up-up behavior, but not down-down). The third group 121c of injector-producer pairs {Ij, · P k } includes pairs for which producer P k exhibits reduced production flow events in response to reduced injection events in the injector Ij, but which they do not show an increase in production flow events in response to the increase in injection events in the Ij injector (ie, the down-down behavior, but not the up-up). The last group 121d includes the injector-producer pairs {Ij, · P k } that do not exhibit an increase in production flow events in response to an increase in injection events in the injector Ij nor a reduction in production flow events in response to a reduction injection events in injector Ij. The statistical ordering process 122 is applied within each group 121a through 121d. It is considered that the statistics used to perform such a classification will include the level of confidence in the existence of an association between the injector Ij and the producer P k , and support for producer events in the producer P k attributed to the injector Ij, · other statistics may , alternatively or additionally, be used as appropriate. The statistical ordering process 122 processes injector-producer pairs {Ij, · P k } within groups 121 and classifies them in the ordered list 125, according to their strength of association. As is evident from Figure 12, the list
    65/82 ordinate 125 contains the injector-producer pairs {Ij; Pk) ranked first according to their polarity response (that is, according to groups 121a to 121d, with group 121a occupying the top of the list 125 ranking, group 121b the second part of the ranking, etc.) , and with the results of the statistical ordering process 122 that classifies the individual injector-producer pairs {Ij; P k } within each of these parts of list 125. As mentioned above, other approaches and classification techniques can be used alternatively or in addition. For example, the user or operator of production field 6 may know of information that can be incorporated into other exclusion entities, for example based on geography or geology, which can be used to remove specific injector-producer associations from the ordered list 125 , regardless of statistical results.
    Following the classification process 82 (Figure 6), the detection process 44 in the overall process flow shown in Figure 3 is completed, according to this embodiment of the invention. The detection process 44 therefore performs the task of analyzing historical and current producer measurement data relevant to output flow rates in producer wells P to P7 in the production field 6 of interest, such measurement data are the direct measurements flow rates, measurements of the allocated flow rates of mixed outlets, flow rates calculated based on indirect measurements in the well (for example, temperature and pressure), or another measurement parameter, such as downhole pressure. From this analysis, process 44 detected events in these producers Pl to P7, considering the capacity
    66/82 response of these production events to injection events in wells II to 15 in the production field of 6, and organized the classification of possible injector-producer pairs according to the strength of their behavioral association. According to embodiments of the present invention, these injector-producer associations are applied iteratively to a reservoir model in process 46, in an orderly manner according to the result of process 44, to efficiently obtain a reservoir working model that can be used to assess potential continuing and secondary recovery actions.
    According to embodiments of the invention, the well-known capacitance model, or capacitance-resistivity model (CRM), is built using associations derived in process 44. To summarize, CRM typically models the cumulative production volume q (t) of a given well over time, assuming a condition of pseudo-stationary state, such as the sum of a 'primary exponential term, with the effects of injection wells in the same production field, and a term that reflects variations in downhole pressure (BHP). The typical expression of the CRM equations is given by Sayarpour et al., The Use of Capacitance-resistivity Models for Rapid Estimation of Waterflood Performance and Optimization, SPE 110081, presented in 2007 at the SPE Annual Technical Conference and Exhibition (2007), incorporated here , in your totality.
    q (r) = q (i o k TT t -t 0 P 'where t 0 is the initial moment, t is a time constant, I (t) reflects the injection flow rate over time that affects the production of the specific well, c t is the
    67/82 well compressibility, V p is the volume of pores in the well, and the p wf values are downhole pressures. When evaluating the effect of the injection flow rate measured in an injection well on the cumulative production q (t) in a production well, as reflected in the value I (t) in the CRM equation, the three gain parameters (that is, the connectivity of the injector Ij to the well), the time constant of the injection ratio of the injector Ij to the well, and the productivity constant that reflects the strength of the reservoir when related to the relationship between the injector I ó and the well, must be evaluated for each of injectors II to 15 in the production field 6. This assessment is applied to each of the producers Pl to P7, in order to model the entire production area 6. Normally, the derivation of the CRM for a given production field involves the solution of an optimization problem, which considers injection flow rates and production flow rates, to minimize the absolute error of each producer; optimization, then, provides the desired parameters (ie gain, time constant, productivity constant) for each of the injector-producer pairs in the production field, which provides a useful model in the evaluation of secondary recovery.
    Conventional CRM optimization is, however, an over-parameterized problem. As such, the effort and computational resources required to converge to a reasonable estimate of the model can be substantial. According to embodiments of the present invention, however, the derivation and evaluation of a useful CRM model of the reservoir can be done efficiently, with reasonable effort and computational resources.
    68/82
    Referring now to Figures 13, 14a and 14b, an example of the operations performed by system 20 in process 46 will be described in detail below. As shown in Figure 13, process 130 retrieves the ordered list 125 of injector-producer pairs generated in process 44, based on the associations of events observed in the measurement data and the corresponding statistical analysis of these associations. In this embodiment of the invention, a group of injector-producer pairs candidates to be applied in a first passage of the CRM derivation to production field 6, is then selected in process 132. In this first passage of process 132, this candidate group selected from injector-producer pairs includes the strongest associations from the ordered list 125, and excludes those from the weakest associations. The specific selection of process 132 can be performed in an interactive way by the user of system 20, perhaps in addition to the orientation of system 20 in its grouping of injector-producer pairs according to the strong, medium, weak, and no associations.
    Figures 14a and 14b illustrate an example from the top of the ordered list 125 for injectors II to 15 and producers Pl to P7 of production field 6 in Figure 1. In this example, Figure 14 illustrates the ordering of associations based on the increase the flow rate of the producer in response to increases in injection, and Figure 14b illustrates the ordering of associations based on decreasing the flow rate of the producer in response to reductions in injection. It is considered that the particular selection of associations for the application to CRM can be made separately (for example, a product injector pair
    69/82 selected can only reflect the increase ratio and not the reduction ratio), or both ratios can be used to select the injector-producer pair. As shown in Figures 14a and 14b, the specific injector-producer associations are grouped according to the STRONG, MEDIUM and WEAK association groups. Each association includes the identification of the injector and the producer, together with the confidence level of that association, and the indication of support for the change in the production flow attributed to that injector. In this example, the relationship between injector II and producer P1 is a particularly strong one, with the highest level of trust and support in each of the lists in Figures 14a and 14b. It is considered that the number of injector-producer pairs in each of the FORTE, AVERAGE and WEAK association groups is not fixed from field to field or varies over time. In fact, it is considered that these groups can be identified because they depend on relatively large gaps in confidence or support values that conveniently separate the various groups. Other approaches to attribute the strength of associations can be used, examples of which include strong visual matches between the subset of isolated events, the use of extrinsic information pertaining to geology, etc.
    Therefore, returning to Figure 13, the first passage of process 132 can, in this way, select the STRONG associations present in the ordered list 125 of injector-producer pair. These injector-producer pairs are then used in the optimization of CRM for the production field of process 134, performed by system 20 according to conventional CRM optimization algorithms and techniques.
    70/82
    CRM parameters for other injector-producer pairs reflect zero connectivity in process 134. After completion of CRM optimization process 134, system 20 then evaluates one or more uncertainty statistics for CRM parameters optimized in process 136 , for the parameter values obtained in this most recent passage of the optimization process 134. The assessed uncertainty statistics are considered as conventional measures of uncertainty, for example, standard error of parameter values. This first example from process 46 (Figure 3) is now complete.
    Returning to Figure 3, as this is the first instance of process 46, the result of decision 47 made by system 20 necessarily returns the result yes. Process 46 is then repeated with at least one additional injector-producer association. In the detailed flow diagram of Figure 13, in this next step, process 132 selects one or more associations from the ordered list 125 for application to the optimization process 134. For example, if the entire group of FORTE associations (Figures 14a, 14b) has been applied in the first pass of process 134, at least one association of the group 'AVERAGE (that is, the highest rated injector-producer pair in that group) will be selected in this next instance of process 46. This additional association can be an isolated association, all the AVERAGE group, or some subset of that group. The optimization process 134 is then repeated with the additional association or associations, and one or more uncertainty statistics are evaluated again for this next step of the optimization process.
    71/82
    134, which completes this instance of process 46 with the increase in the number of associations.
    For this second (and later) instances of process 46, the uncertainty statistics calculated in process 136 are compared to the values of the uncertainty statistics calculated in the most recent previous passage in process 46. Decision 47 is made by process 20 to determine whether the model fit improved to a statistically significant extent. For example, the well-known Student's t test can be applied to determine, from standard error or other uncertainty statistics calculated in the last two evaluations of the model, whether the distribution of the model parameters evaluated in this process example 136 (ie with the additional associations) is equal to the distribution of parameters of the model of the previous example, in the selected statistical significance. For example, decision 47 can assess this statistical similarity using a selected threshold level of the p-value (probability that the selected statistic from the most recent parameter distribution is at least as extreme as the statistic from the previous distribution, if the distributions are equal), with the test statistic being the standard error of the model parameters. Of course, other tests of statistical significance can be used in relation to the difference between the two sets of parameters of the model. The specific threshold level can be selected by the user a priori, or it can be selected during the global process based on previous values of the uncertainty statistics for the specific field 6 of production. If the uncertainty statistic of the CRM parameters
    72/82 evaluated reflect significantly better fit (for example, minor standard error) in the most recent passage of process 46 with one or more additional injector-producer associations (decision 47 is yes), process 46 is repeated again, including the addition of one or more injector-producer associations according to the ordered list 125. On the other hand, if the latest passage of process 46 does not improve the statistical uncertainty in the CRM parameters of the process 46 of optimization for the selected statistical significance (decision 47 is no), then the derivation of the CRM model is considered complete. The inclusion of additional injector-producer associations would not serve to improve the optimization of the CRM parameters, in any statistical significance. The values of the model parameters, from the most recent passage of process 46 (or from the previous passage of process 46, if desired), are then used in the subsequent evaluation of the CRM.
    For this reason, according to embodiments of the present invention, the difficulties of deriving the model of the injector and producer relationships of a production field from measurement data referring to flow rates are largely avoided. In particular, the difficulty in deriving a CRM model due to excessive parameterization, especially when applied to production fields that still contain a reasonable number of injection wells and production wells, is largely avoided. Only injector-producer connections that significantly affect the CRM model, to any statistically significant degree, need to be included in the optimization of the model parameters. This efficient model construction is based on real measurement data and
    73/82 automated event identification, and allows quick re-evaluation of models with measurement data obtained most recently. Furthermore, this derivation and evaluation of the secondary recovery model can easily be scaled for large production fields, with a large number of injectors and producers, without overloading the available computational resources, due to its hierarchical application of the strongest injector-producer associations of according to the statistical measures of these associations.
    L0 In view of this, referring back to Figure 3, the resulting model with its evaluated model parameters can then be used to analyze future secondary recovery actions. A proposed increase or change in the fluid injection flow in one or more injection wells 15 in the production field under analysis can be applied to the model, and the effect of the proposed change in production can be easily assessed. Examples of conventional techniques to optimize secondary recovery actions through CRM assessment and similar reservoir models are described in Liang et al., Optimization of Oil Production Based on a Capacitance Model of Production and injection Rates, SPE 107713, presented in 2007 at SPE Hydrocarbon Economics and Evaluation Symposium (2007), Sayarpour et al, The Use of Capacitance-resistivity Models for Rapid Estimation of Waterflood Performance and Optimization, SPE 110081, presented in 2007 at the SPE Annual Technical Conference and Exhibition (2007); both of which are incorporated herein by reference in their entirety. A connectivity model for the reservoir, as provided by embodiments of the present invention, can then be used to assess the efficiency of recovery actions
    74/82, by trial and error, or by an additional optimization process (for example, minimizing a cost function), or by some other technique, to maximize oil and gas production through recovery activities secondary, at minimal cost.
    The processes involved in determining a statistical model of the reservoir, according to embodiments of the present invention, may also allow for additional analysis and experimental design, in addition to the evaluation of potential secondary recovery actions. For example, the statistics underlying the ordered list of injector producer associations produced according to this change can be analyzed separately to structure optimization experiments. According to this approach, injector-producer associations that appear to be strongly linked (for example, strong support), but exhibit weak confidence that the strong association can specifically be tested, by intentionally causing injection events in that injector, to keep constant other injectors, and closely monitor the response to the producer; assessments of injector-producer interactions from such experiences can be used to further refine the effective strength of this association. According to other uses of the embodiments of this invention, candidate wells for sweeping modification, such as by means of water injection with the dispersion product BRIGHT WATER available from TIORCO, can be identified by analysis in accordance with embodiments of the present invention. The optimization of secondary recovery actions according to embodiments of the present invention can also incorporate economic cost factors, for example, the
    75/82 attributing an economic value to the injected water, and evaluating the barrels of oil produced from such injection at certain price levels, in order to arrive at an economic optimization of these secondary recovery actions. These and other uses are contemplated to be within the scope of this invention.
    Capacitance-resistivity model (CRM) assessment before event detection
    According to another embodiment of the invention, the evaluation of the reservoir model is performed before the detection of injector-producer events. Figure 15 is a flow diagram illustrating an example of such an embodiment of the invention; similar processes in this embodiment, as in the embodiment described above in relation to Figure 3 are identified in Figure 15, with the same reference numbers.
    The process of this embodiment of the invention begins, as before, with process 40, in which the measurement data referring to the flow rates of the wells of interest in the production field 6 are obtained and processed by system 20. As described above in detail in With respect to process 40, these measurement data are acquired from the appropriate data source and may include flow measurements or flow calculations for each injector II to 15 and producer Pl to P7 of the production field 6 over time, and other well measurements such as downhole pressure (BHP), unstructured or non-periodic data from fluid samples, well tests, and chemical analysis, etc. Process 40 also applies, if desired, various filters, processing, and editing this measurement data as described above, for example, to remove invalid values
    76/82 and statistical anomalies, adjustment or filtering of the data on a regular periodic basis, application of corrections of reservoir barrels, and the like.
    As described above in relation to Figure 3, system 20 then identifies injector events from the transformed measurement data, in the process of 42. The way in which system 20 performs the event identification process 42 can follow the described above in connection with Figures 3, 4a, and 4b, including the correlation and visualization approaches described above. As before, injector events of various types are envisaged to be detected in this process example 42. These events include on-off injector events corresponding to injector wells that are placed in and out of line. Injection events that occur during operation (that is, changes in the injection flow rate in an injector that is in line) can also be considered in accordance with this embodiment of the invention. In addition, as described above, isolated injection events (for example, events that occur in one injector that differ from changes in several other injectors, such as changing the injection rate in the opposite direction) can lend an extraordinary understanding of well communication. -to-well. The injector events identified in process 42 thus correspond to changes in the injection flow in one or more injectors, and can also correspond to other occurrences, such as changes in the alternative injection of water and gas into injectors, and the increase in gas production or gas / oil ratio (GOR) in producers, as described above.
    77/82
    According to this embodiment of the invention, a reservoir model is evaluated before detecting events of injector-producer pairs, to restrict the number of injector-producer pairs that require the detection of events and study of association. As such, once a set of injector events has been identified in process 42, the appropriate reservoir model is evaluated to initially identify producers who potentially have some connection and thus respond to the injector events identified in process 42. For example, a capacitance-resistivity (CRM) model is evaluated based on these identified injector events, in process 150. As is well known in the art, conventional CRM models evaluate the effect of an injection flow rate measured in a well cumulative production injector q (t) in a production well, by evaluating the three gain parameters (ie, the connectivity of the injector Ij to the well), the time constant of the relationship between the injection of the injector Ij to the well , and the productivity constant that reflects the strength of the reservoir, since it refers to the relationship between the injector Ij and the well. In process 150 according to this embodiment of the invention, the complete set of gains relative to one or more injector events identified in process 42 is evaluated; that is, the gain associated with each of the producers P1 to P7 in the production field 6, is evaluated. It is considered that the extent to which the CRM optimization convergence problem is solved in process 150 may be relatively gross, compared to what is expected by the complete assessment of the reservoir model.
    78/82
    In process 152, the CRM gains evaluated in process 150 are analyzed based on the injector events identified. More specifically, the injector-producer pairs that exhibit zero gain in the evaluation process 150 can be eliminated from further considerations in the process of Figure 15 according to this embodiment of the invention. The iterative evaluation of CRM within process 150 can be invoked to identify and confirm zero-gain pairs. In addition, system 20 (either automatically or interactively with a user's collaboration) can identify ganzero injector-producer pairs based on criteria such as the distance between the injector and the producer in the field, the presence of other geological restrictions (that is, information indicating extrinsic physical impossibility of a connection between injector and producer), and the like. As a result of the 152 process, a set of injector-producer pairs are identified, from the CRM, as having non-zero gains and, thus, would have some level of connection within the reservoir. These non-zero gain pairs are then forwarded to process 44, in which system 20 detects producer events caused by injector events within the restricted subset.
    Alternatively, process 42 can be omitted prior to CRM evaluation process 150 and analysis process 152, as the identification of injector events is not strictly necessary before CRM evaluation. In this alternative approach, the full set of gains for all available injector-producer pairs determined in process 150 are analyzed in process 152, and those with zero gain (or as explicitly
    79/82 determined or according to an alternative criterion) are removed from further analysis as described above.
    Therefore, according to this embodiment of the invention, the event detection process 44 is basically called to confirm the diagnosis or to reject the injector-producer relationships identified by evaluating a CRM in processes 150, 152, based on the level of statistical uncertainty. for each of those relationships. In addition, the event 44 detection process also allows the explicit illustration of those gains that are statistically valid, based on the examination of the producer's responses to the identified injection events. These analyzes by the event detection process 44 can be based on both primary events (injector on-off events) and also secondary events (injector running events). By limiting the set of injector-producer associations that will be examined in the event identification task performed by system 20 in process 44, event detection is much more efficient, and it is also more effective because false positive associations (events that are detected, but that have zero-gain in the CRM model) are eliminated. In addition, the CRM assessment before event detection helps to refine the extraction of effectively isolated events in the injection history due to this limitation of the association set. For example, if a number of injectors is rejected by the CRM assessment as possible influences on a given producer, the remaining smaller subset of influential injectors on that producer can be processed more effectively (for example, by analyzing the direction of change) to improve even more
    80/82 estimates of the fundamental delay time for that pair of wells, which in turn improves the identification of precise associations between wells in the production field.
    In addition, it is considered that the combination of CRM evaluation (processes 150, 152), with event detection (process 44) allows the development of an absolute test criterion for the scheduling of a production event. For example, any injector-producer pair with a non-zero gain in CRM, at a high confidence level, should exhibit at least some event matching in the event detection process. In this way, the selection of parameters and values used in the event detection process 44 to define production events can be done by evaluating parameters and values that improve the scores of the high confidence associations of these well pairs.
    For example, the injector-producer pairs indicated by process 150 as linked can be analyzed within process 44 to derive an expectation of the probable number of response events in that producer well, which can guide the selection of event marking thresholds. In this approach, major injector on-off events are well correlated with time across the production field, as all wells tend to close together, and then reopen together, in order to quickly return to full production. As such, these events often add little knowledge about connectivity. In one embodiment, the development of an event detection threshold in a given producer can use the limited set of pairs provided by the CRM assessment process 150, 152 by:
    81/82 • First, identify and remove on / off events from the flow rate time sequence of the producer;
    • For the injectors indicated by the process 150 as connected to the producer, eliminating on-off and up-down injection events in immediately previous periods of time (that is, within the delay time expected for the determined producer);
    • Repeat these two steps to mask events in the producer's time sequence;
    • Then, add the remaining elements of the connected injectors' time sequences (both the binary values of events and their magnitudes);
    • Assess the number of peaks in the time sequence of the added injection flow rate; and • Determine a useful threshold at which the added injection flow rate time sequence causes a causal event in the given producer time sequence.
    This threshold can be useful in the event detection process 44, particularly to discern the presence and importance of events in both injectors and producers.
    The results of the event detection process 44 are then used, as described above, to iteratively evaluate the reservoir CRM model (process 46 and decision 47), according to the relative statistical strengths of the associations. The analysis of future actions to be taken in the production area (process 48) is thus facilitated, as described above.
    82/82
    It is further considered that other variations and implementations alternative to the embodiments of the present invention, as would become apparent to those skilled in the art with reference to this specification, can also be applied and are within the scope of the present invention as claimed.
    Although the present invention has been described according to its preferred embodiments, it is of course anticipated that the modifications of, and alternatives to, those embodiments, the modifications and alternatives to obtain the advantages and benefits of the present invention, will be apparent to those skilled in the art. specialty, having as reference this specification and its drawings. Such modifications and alternatives are considered to be within the scope of this invention as subsequently claimed in the present patent application.
    权利要求:
    Claims (24)
    [1]
    1. Method implemented by computer for the evaluation of water injection in a subsurface hydrocarbon reservoir in which one or more producer wells and one or more injector wells were drilled, characterized by: understanding the steps of: receiving measurement data over the time corresponding to flow rates of one or more producer wells and one or more injector wells;
    identify, from the measurement data received, a plurality of associations between one of the producing wells and one of the injector wells, each of the associations identified by a measure of the strength of the association;
    classify the identified associations according to the strength of the association;
    apply one or more of the best classified associations to the capacitance-resistivity model of the reservoir;
    evaluate the capacitance-resistivity model of the reservoir in relation to the measurement data to obtain a set of model parameters and the associated statistical uncertainty;
    apply the following one or more of the associations, selected according to the order of classification of the associations, to the capacitance-
    resistivityevaluate the the reservoir; capacitance-resistivity model of reservoir, with the following one or more of interconnections applied in relation to the data in
    [2]
    2/24 measurement, to obtain a set of model parameters and the associated statistical uncertainty; and repeat the application steps of the following one or more of the interconnections and evaluate the capacitance-resistivity model of the reservoir with the following one or more of the applied interconnections, until the statistical uncertainty reflects the similarity of the model parameters since the most recent evaluation step and the model parameters of a previous evaluation step, in a selected statistical significance.
    Method according to claim 1, characterized by:
    further understand after the repeated application and evaluation steps and taking care that the statistical uncertainty reflects the similarity with the selected statistical significance:
    then evaluate an injection proposal in one or more of the injection wells, using the capacitance-resistivity model of the reservoir and the evaluated parameters of the model.
    [3]
    Method according to claim 1, characterized by:
    the statistical uncertainty corresponds to the standard error of the model parameters.
    [4]
    Method according to claim 1, characterized by:
    the measurement data for producer wells correspond to cumulative production over time.
    3/24
    Method according to claim 1, characterized by:
    the measurement data comprises downhole pressures over time.
    Method according to claim 1, characterized by:
    the classification step comprises:
    group the associations identified in a plurality of subsets according to the correspondence of the polarities of the changes in the measurement data between the injection well and the production well;
    so that a first instance of the application stage applies a first subset of interconnections that correspond to the best classified associations for the reservoir's capacitance-resistance model; and so that a second instance of the application step applies a second subset of interconnections that correspond to the next highest classification associations of the reservoir model.
    Method according to claim 6, characterized by:
    the classification stage also includes:
    among the one or more better classified of the plurality of subsets, order the identified associations according to a statistical measure of the strength of the association.
    Method according to claim 1, characterized in that
    The classification stage comprises:
    order the identified associations according to the statistical measure of the strength of the association.
    .3. Method according to claim 1, characterized by:
    further understand:
    identify, from the measurement data that correspond to the flow rates of one or more injection wells, injector events in which changes in the flow rate have occurred;
    detect, from the measurement data that correspond to flow rates in one or more producer wells, one or more producer events in which changes in the flow rate have occurred;
    identifying detected producer events that occurred within a selected interval of delay time relative to identified injector events; and to derive, from the identification of the detected producer events, associations between one of the injection wells and one of the producing wells.
    A method according to claim 9, characterized by:
    the stage of identification of detected producer events comprises, for each one or more producer wells:
    calculate the gradient in the measurement data for each of a plurality of points in time; and detecting points in time at which the gradient calculated from one point in time to another varies by more than a first threshold value.
    [5]
    5/24
    11. Sludge mill according to claim 10, characterized by:
    the step of calculating the gradient of a point in time calculating the regressive gradient of the measurement data and the corresponding measure of fit over a given number of points in time, which includes points in time prior to the point in time in question; and the detection step comprises, for each of the plurality of points in time:
    compare the point-in-time adjustment measure with the previous point-in-time adjustment measure;
    calculate, in response to the measure of the point-in-time adjustment that degrades from the measure of the point-in-time adjustment by a selected margin, the progressive gradient of the point-in-time measurement data over a given number of points in time after the point in time in question; and identifying the producer event of the point in time that responds to the progressive gradient that differs from the regressive gradient by more than the first threshold value.
    Method according to claim 11, characterized by:
    the stage of identifying a producer event also includes:
    calculate the value of the amplitude of the difference between the progressive gradient and the regressive gradient of the point in time.
    [6]
    6/24
    13. Method according to claim 12, characterized by:
    the step of identifying detected producer events, further understanding:
    calculate, after the operation of detecting points in time in which the calculated gradient changes from a point in time, the moving average of the amplitude value within a selected time window, which moves over a period of time selected in the measurement data;
    then identify the producer event in each contiguous time group in which the moving average of the amplitude value exceeds a second threshold value; and assign the algebraic unit value indicator at each point in time that corresponds to an identified producer event, the algebraic unit value indicator sign that corresponds to the polarity of the change in the gradient of the identified producer event.
    Method according to claim 9, characterized by:
    further understand:
    derive, from the identification of the producer events detected, associations between one of the injector wells and one of the producer wells.
    assign an indicator to one or more of the derived associations that indicates the strength of the association between the associated injection well and the production well.
    [7]
    7/24
    15. Method according to claim 9, characterized by the step of identifying injector events, comprising:
    display the time series of measurement data for a selected injection well on a computer system screen;
    operate the computer system in order to identify one or more potential injector events in the time series;
    accept a user input that selects one of the potential injector events;
    display on the screen, for the potential selected injector event, a portion of the measurement data time series for the selected injection well in combination with a portion of the measurement data time series for the selected production well, which portion is normalized to time and amplitude for reciprocal alignment in time; and accept, after displaying part of the time series, a user input that confirms the potential selected injector event.
    16. Method according to claim 9, characterized by:
    the step of identifying injector events, understanding:
    display the time series of measurement data for a selected injection well on a computer system screen;
    [8]
    8/24 accept a user input that indicates a potential injector event in the presented time series;
    operate the computer system in order to identify one or more potential injector events similar to the potential injector event indicated, and to identify, for a user, one or more of the potential events that are functionally isolated from intra-well effects;
    accept a user input that selects one of the potential injector events;
    display on the screen, for the potential selected injection event, a portion of the measurement data time series for the selected injection well in combination with a portion of the measurement data time series for the selected production well, which portion is normalized to time and amplitude for reciprocal alignment in time; and accept, after displaying part of the time series, a user input that confirms the potential selected injector event.
    17. Method according to claim 9, characterized by:
    further understand:
    evaluate, after the step of identifying injector events, and before the step of detecting one or more producer events, the reservoir's capacitance-resistance model in relation to the measurement data to derive gain values for each injector-producer pair; and
    [9]
    9/24 define a subset of one or more injector-producer pairs that have non-zero gain values;
    where the steps of identifying detected producer events and deriving associations are performed on the defined subset of one or more injector-producer pairs.
    18. Method according to claim 1, characterized by:
    further understand:
    correct the measurement data received based on the variations in the measurement values independent of the flow in the well.
    19. Method implemented by computer to detect events of change in flow velocity from a well in a hydrocarbon reservoir, characterized by:
    understand the steps of:
    receiving measurement data over time that corresponds to the flow rates in the well; and at each of the plurality of time points for which measurement data are present:
    calculate the regressive gradient of the measurement data and a corresponding adjustment measurement over a selected number of points in time, which include time intervals before the point in time;
    compare the point-in-time adjustment measure with the previous point-in-time adjustment measure;
    [10]
    10/24 calculate, in response to the measure of the point-in-time adjustment that degrades from the measure of the point-in-time adjustment by a selected margin, the progressive gradient of the point-in-time measurement data over a given number of points in time after the point in time in question; and identifying a flow rate change event at the point in time in response to the progressive gradient that differs from the regressive gradient by more than the first threshold value.
    20. Method according to claim 19, characterized by:
    the step of identifying the flow rate variation event, further understand:
    calculate the value of the amplitude of the difference between the progressive gradient and the regressive gradient of the point in time.
    21. The method of claim 20, characterized by:
    further understand:
    calculate, after the operation of detecting points in time in which the calculated gradient changes from a point in time, the moving average of the amplitude value within a selected time window, which moves over a period of time selected in the measurement data;
    1d en t i f i c a r, in s and g u d, the event of change in the flow rate in each contiguous group of time in
    [11]
    11/24 that the moving average of the amplitude value exceeds a second threshold value; and assigning an algebraic indicator unit value to each point in time that corresponds to an identified flow rate change event, the sign of the algebraic indicator unit value corresponds to the polarity of the change in the gradient of the identified flow rate change event.
    22. Computerized system to evaluate the injection of water in a subsurface hydrocarbon reservoir, in which one or more producer wells and one or more injector wells have been drilled, characterized by: understanding:
    one or more processing units for executing the program instructions;
    memory resource, to store measurement data over time that correspond to flow rates in one or more producer wells and in one or more injector wells; and programming memory, coupled with one or more processing units, for the storage of computer program that includes instructions that, when executed by one or more processing units, are capable of causing the computer system to execute the sequence of operations which comprises:
    receiving measurement data from the memory resource;
    identify, from the measurement data received, a plurality of associations between one of the producing wells and one of the wells
    [12]
    12/24 injectors, each of the associations identified by a measure of the strength of the association;
    classify the identified associations according to the strength of the association;
    apply one or more of the best classified associations to the reservoir's capacitance-resistance model;
    evaluate the capacitance-resistivity model of the reservoir in relation to the measurement data to obtain a set of model parameters and the associated statistical uncertainty;
    apply the following one or more of the associations, selected according to the order of classification of the associations, to the capacitance-resistivity model of the reservoir;
    evaluate the capacitance-resistivity model of the reservoir, with the following one or more of the applied interconnections, in relation to the measurement data, to obtain a set of model parameters and the associated statistical uncertainty; and repeat the operations of applying one or more of the interconnections to the subsequent one and evaluating the capacitance-resistivity model of the reservoir with the subsequent applied one or more of the interconnections, until the statistical uncertainty reflects the similarity of the model parameters from the evaluation stage latest and the parameters of the
    [13]
    13/24 model of the evaluation of a previous step, with a selected statistical significance.
    23. The system of claim 22, characterized by:
    after repeated application and evaluation operations, and capable of responding to the statistical uncertainty that reflects the similarity to the selected statistical significance, the sequence of operations further comprises:
    then evaluate an injection proposal in one or more of the injection wells, using the capacitance-resistivity model of the reservoir and the evaluated parameters of the model.
    24. System according to claim 22, characterized by:
    the classification operation comprises:
    group the associations identified in a plurality of subsets according to the correspondence of the polarities of the changes in the measurement data between the injection well and the production well;
    in which the first instance of the application operation applies a first subset of interconnections that correspond to the best classified associations for the reservoir's capacitance-resistivity model; and in which the second instance of the application operation applies a second subset of interconnections corresponding to the subsequent highest ranked associations for
    [14]
    14/24 the capacitance-resistivity model of the reservoir.
    25. System according to claim 22, characterized by:
    the sequence of operations also includes:
    identify, from the measurement data that correspond to the flow rates of one or more injection wells, injector events in which changes in the flow rate have occurred;
    detect, from the measurement data that correspond to flow rates of one or more producer wells, producer events in which changes in the flow rate occurred;
    identifying detected producer events that occurred within a selected interval of delay time relative to identified injector events; and to derive, from the identification of the detected producer events, associations between one of the injection wells and one of the producing wells.
    26. System according to claim 25, characterized by:
    the operation to identify producer events detected for each of the one or more producer wells, comprising:
    calculate the gradient in the measurement data for each of a plurality of points in time; and detecting points in time at which the gradient calculated from one point in time to another varies by more than a first threshold value.
    [15]
    15/24
    27. System according to claim 26, characterized by:
    the operation of calculating a gradient of a point in time to calculate the regressive gradient of the measurement data and the corresponding adjustment measure over a certain number of points in time, which include time intervals before the point in time; and the detection operation, for each of the plurality of points in time, comprise:
    compare the point-in-time adjustment measure with the previous point-in-time adjustment measure;
    calculate, in response to the measure of the point-in-time adjustment that degrades from the point-in-time measure by a selected margin, the progressive gradient of the point-in-time measurement data over a given number of points in time after the point in time in question; and identifying the producer event of the point in time that responds to the progressive gradient that differs from the regressive gradient by more than the first threshold value.
    28. System according to claim 27, characterized by:
    detection of producer events, further understand:
    calculate the amplitude value for the difference between the progressive gradient and the regressive gradient of the point in time;
    [16]
    16/24 calculate, after the operation of detecting points in time in which the calculated gradient changes from a point in time, the moving average of the amplitude value within a selected time window, which moves over a selected time period of the measurement data;
    then identify the producer event in each contiguous time group in which the moving average of the amplitude value exceeds a second threshold value; and assign the algebraic unit value indicator at each point in time that corresponds to an identified producer event, the algebraic unit value indicator sign that corresponds to the polarity of the change in the gradient of the identified producer event.
    29. System according to claim 25, characterized by:
    the injection event identification operation, understand:
    display the time series of measurement data for a selected injection well on a computer system screen;
    operate the computer system in order to identify one or more potential injector events in the time series;
    accept a user input that selects one of the potential injector events;
    display on the screen, for the potential selected injector event, a portion of the time series of measurement data for the injection well
    [17]
    17/24 selected in combination with a portion of the time series of measurement data for the selected production well, which portion is normalized in time and amplitude for reciprocal alignment over time; and accept, after displaying part of the time series, a user input that confirms the potential selected injector event.
    30. System according to claim 25, characterized by:
    the sequence of operations also includes:
    evaluate, after the injection event identification operation, and before the detection operation of one or more producer events, the capacitance-resistivity model of the reservoir in relation to the measurement data to derive gain values for each injector-producer pair ; and defining a subset of one or more injector-producer pairs that have non-zero gain values;
    in which the operations for the identification of detected producer events and the derivation of associations are carried out on the defined subset of one or more injector-producer pairs.
    31. Non-volatile, computer-readable medium for storing a computer program that, when run on a computer system, causes the computer system to perform a sequence of operations to evaluate the injection of water into a subsurface reservoir of hydrocarbon in which a
    [18]
    18/24 or more producing wells and one or more injector wells were drilled, characterized by:
    the sequence of operations comprise:
    access stored measurement data that correspond to flow rates in one or more producer wells and one or more injector wells over time;
    identify, from the measurement data, a plurality of associations between one of the producing wells and one of the injector wells, each of the identified associations having a measure of the strength of the association;
    classify the identified associations according to the strength of the association;
    apply one or more of the best classified associations to the capacitance-resistivity model of the reservoir;
    evaluate the capacitance-resistivity model of the reservoir in relation to the measurement data to obtain a set of model parameters and the associated statistical uncertainty;
    apply the following one or more of the associations, selected according to the order of classification of the associations, to the reservoir's capacitance-resistance model;
    evaluate the capacitance-resistivity model of the reservoir, with the following one or more of the interconnections applied, in relation to the measurement data, to obtain a set of model parameters and the associated statistical uncertainty; and
    [19]
    19/24 repeat the operations of applying one or more of the interconnections to the subsequent and evaluating the capacitance-resistivity model of the reservoir with the subsequent application of one or more of the interconnections, until the statistical uncertainty reflects the similarity of the model parameters from the stage most recent evaluation method and the parameters of the evaluation model from a previous step, with selected statistical significance.
    32. Non-volatile, computer-readable medium for storing computer programs, according to claim 31, characterized by:
    the sequence of operations, subsequent to repeated application and evaluation of operations, and in response to the statistical uncertainty that reflects the similarity for the selected statistical significance, further understand:
    then evaluate an injection proposal in one or more of the injection wells, using the capacitance-resistivity model of the reservoir and the evaluated parameters of the model.
    33. Computer readable, non-volatile medium for storing computer programs according to claim 31, characterized by:
    the classification operation comprises:
    group the associations identified in a plurality of subsets according to the correspondence of the polarities of the changes in the measurement data between the injection well and the production well;
    [20]
    20/24 in which the first instance of the application operation applies a first subset of interconnections that correspond to the best classified associations for the capacitance-resistivity model of the reservoir; and in which the second instance of the application operation applies a second subset of interconnections that correspond to the subsequent associations of higher classification for the capacitance-resistivity model of the reservoir.
    34. Non-volatile, computer-readable medium for storing computer programs, according to claim 31, characterized by:
    the sequence of operations also includes: identifying, from the measurement data that correspond to the flow rates of one or more injection wells, injector events in which changes in the flow rate have occurred;
    detect, from the measurement data that correspond to flow rates of one or more producer wells, producer events in which changes in the flow rate occurred;
    identifying detected producer events that occurred within a selected interval of delay time relative to identified injector events; and to derive, from the identification of the detected producer events, associations between one of the injection wells and one of the producing wells.
    [21]
    21/24
    35. Non-volatile, computer-readable medium for storing computer programs, according to claim 34, characterized by:
    the operation of identifying producer events detected, for each one or more producer wells, comprises:
    calculate the gradient in the measurement data for each of a plurality of points in time; and detecting points in time at which the gradient calculated from one point in time to another varies by more than a first threshold value.
    36. Non-volatile, computer-readable medium for storing computer programs, according to claim 35, characterized by:
    the operation of calculating a gradient of a point in time calculates a regressive gradient of the measurement data and a corresponding adjustment measure over a certain number of points in time that include time intervals before the point in time;
    and the detection operation, for each of the plurality of points in time, comprise:
    compare the point-in-time adjustment measure with the previous point-in-time adjustment measure;
    calculate, in response to the measure of the point-in-time adjustment that degrades from the point-in-time measure by a selected margin, the progressive gradient of the point-in-time measurement data over
    [22]
    22/24 a certain number of points in time after the point in time in question; and identifying the producer event of the point in time that responds to the progressive gradient that differs from the regressive gradient by more than the first threshold value.
    37. Non-volatile, computer-readable medium for storing computer programs, according to claim 36, characterized by:
    the producer event detection operation also includes:
    calculate the amplitude value for the difference between the progressive gradient and the regressive gradient of the point in time;
    calculate, after the operation of detecting points in time in which the calculated gradient changes from a point in time, the moving average of the amplitude value within a selected time window, which moves over a period of time selected from the measurement data;
    then identify the producer event in each contiguous time group in which the moving average of the amplitude value exceeds a second threshold value; and assign the algebraic unit value indicator at each point in time that corresponds to an identified producer event, the algebraic unit value indicator sign that corresponds to the polarity of the change in the gradient of the identified producer event.
    [23]
    23/24
    38. Non-volatile, computer-readable medium for storing computer programs, according to claim 34, characterized by:
    the injection event identification operation comprises:
    display the time series of measurement data for an injection well selected on the web of a computer system;
    operate the computer system in order to identify one or more potential injector events in the time series;
    accept a user input that selects one of the potential injector events;
    display on the screen, for the potential selected injector event, a portion of the measurement data time series for the selected injection well in combination with a portion of the measurement data time series for the selected production well, which portion is normalized to time and amplitude for reciprocal alignment in time; and accept, after displaying part of the time series, a user input that confirms the potential selected injector event.
    39. Non-volatile, computer-readable medium for storing computer programs, according to claim 34, characterized by:
    the sequence of operations, further understand:
    evaluate, after the operation of identification of injector events, and before the operation of detection of one or more producer events, the model
    [24]
    24/24 capacitance-resistivity of the reservoir in relation to the measurement data to derive gain values for each injector-producer pair; and define a subset of one or more pairs
    5 injector-producer with gain values other than zero;
    in which operations to identify detected producer events and to derive associations are performed on the defined subset of 10 or more injector-producer pairs.
    类似技术:
    公开号 | 公开日 | 专利标题
    BR112014010661A2|2019-08-27|Event flow-based statistics provision model detected
    RU2573746C2|2016-01-27|Systems and methods for well performance forecasting
    US9151868B2|2015-10-06|Reservoir architecture and connectivity analysis
    US8244509B2|2012-08-14|Method for managing production from a hydrocarbon producing reservoir in real-time
    US10203428B2|2019-02-12|Pressure transient test with sensitivity analysis
    US20110191029A1|2011-08-04|System and method for well test design, interpretation and test objectives verification
    US8229880B2|2012-07-24|Evaluation of acid fracturing treatments in an oilfield
    He et al.2016|Proxy-based work flow for a priori evaluation of data-acquisition programs
    AU2007211291A1|2007-08-09|Methods, systems, and computer-readable media for fast updating of oil and gas field production models with physical and proxy simulators
    WO2016118711A1|2016-07-28|Measuring inter-reservoir cross flow rate through unintended leaks in zonal isolation cement sheaths in offset wells
    AU2017258836A1|2017-11-30|History matching multi-porosity solutions
    EP2959322A2|2015-12-30|Improving velocity models for processing seismic data based on basin modeling
    Levitan2007|Deconvolution of multiwell test data
    Guevara et al.2019|A machine-learning methodology using domain-knowledge constraints for well-data integration and well-production prediction
    Hatampour et al.2018|Hydraulic flow units' estimation from seismic data using artificial intelligence systems, an example from a gas reservoir in the Persian Gulf
    Alfataierge et al.2019|3D hydraulic-fracture simulation integrated with 4D time-lapse multicomponent seismic and microseismic interpretations, Wattenberg Field, Colorado
    Sinha et al.2018|Statistical controls on induced seismicity
    Moreno et al.2015|Channelling detection using data-driven models
    Lee2010|Investigating statistical modeling approaches for reservoir characterization in waterfloods from rates fluctuations
    Geier et al.1996|Generalized radial flow interpretation of well tests for the SITE-94 project
    Moreno et al.2014|Injection leakage detection in mature fields using data-driven models
    Costa et al.2019|Evaluation of an uncertainty reduction methodology based on Iterative Sensitivity Analysis | applied to naturally fractured reservoirs
    US20210310345A1|2021-10-07|Methods for generating synthetic production logs for perforated intervals of a plurality of wells
    US8255816B2|2012-08-28|Modifying a magnified field model
    Romberg et al.2020|Predicting Trouble Stages with Geomechanical Measurements and Machine Learning: A Case Study on Southern Midland Basin Horizontal Completions
    同族专利:
    公开号 | 公开日
    CN103917743A|2014-07-09|
    AU2011380514B2|2016-10-13|
    EP2773845A2|2014-09-10|
    WO2013066358A2|2013-05-10|
    EA201490907A1|2014-09-30|
    US20130116998A1|2013-05-09|
    WO2013066358A3|2013-09-19|
    EP2773845B1|2016-05-25|
    EA026086B1|2017-02-28|
    US9140108B2|2015-09-22|
    AU2011380514A1|2014-05-15|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US6519531B1|2000-09-27|2003-02-11|Rod P. Batycky|System and methods for visual interpretation of well rate allocation factors|
    US6937023B2|2003-02-18|2005-08-30|Pathfinder Energy Services, Inc.|Passive ranging techniques in borehole surveying|
    CA2514516C|2003-03-26|2012-11-20|Exxonmobil Upstream Research Company|Performance prediction method for hydrocarbon recovery processes|
    GB0415144D0|2004-07-06|2004-08-11|Attar Software Ltd|Method and system for detecting events in process operating data and identifying associations between related events|
    WO2008025622A1|2006-08-30|2008-03-06|Shell Internationale Research Maatschappij B.V.|Method for controlling and/or optimizing production of oil and/or gas wells and facilities|
    US8170801B2|2007-02-26|2012-05-01|Bp Exploration Operating Company Limited|Determining fluid rate and phase information for a hydrocarbon well using predictive models|
    CN101899972A|2010-06-30|2010-12-01|中国石油大学|Establishment method of physical model capable of forecasting waterflood development of sandstone reservoirs|AU2013263330B2|2012-05-14|2016-05-26|Landmark Graphics Corporation|Method and system of predicting future hydrocarbon production|
    GB201306967D0|2013-04-17|2013-05-29|Norwegian Univ Sci & Tech Ntnu|Control of flow networks|
    US20150051838A1|2013-08-15|2015-02-19|Chevron U.S.A. Inc.|System and method of determining and optimizing waterflood performance|
    US9784885B2|2014-06-27|2017-10-10|Saudi Arabian Oil Company|Methods and systems for estimating sizes and effects of wellbore obstructions in water injection wells|
    US10012059B2|2014-08-21|2018-07-03|Exxonmobil Upstream Research Company|Gas lift optimization employing data obtained from surface mounted sensors|
    EP3674516A3|2014-08-22|2020-09-23|Chevron U.S.A. Inc.|Flooding analysis tool and method thereof|
    US10078865B2|2014-09-08|2018-09-18|Leeo, Inc.|Sensor-data sub-contracting during environmental monitoring|
    US10577894B1|2015-06-08|2020-03-03|DataInfoCom USA, Inc.|Systems and methods for analyzing resource production|
    WO2017044108A1|2015-09-10|2017-03-16|Halliburton Energy Services Inc.|Method for n-dimensional detection of spatial, temporal and gravimetric attributes and associated variation in plots, digital objects and photographs|
    CN105260613B|2015-10-26|2018-06-01|中国石油天然气股份有限公司|A kind of method for obtaining compressor set robustness|
    CN105404726B|2015-11-02|2018-02-13|中国地质大学（武汉）|A kind of capacitor model inverting inter well connectivity method and device based on Gaussian Profile|
    US10385659B2|2015-12-17|2019-08-20|Arizona Board Of Regents On Behalf Of Arizona State University|Evaluation of production performance from a hydraulically fractured well|
    US10215002B2|2016-05-05|2019-02-26|Saudi Arabian Oil Company|Determining wellbore leak crossflow rate between formations in an injection well|
    CN105971577B|2016-07-08|2018-09-04|中国石油天然气股份有限公司|Improve the method and device of fireflood injection well and producing well connectivity|
    US11263370B2|2016-08-25|2022-03-01|Enverus, Inc.|Systems and methods for allocating hydrocarbon production values|
    US10303819B2|2016-08-25|2019-05-28|Drilling Info, Inc.|Systems and methods for allocating hydrocarbon production values|
    NO20161482A1|2016-09-19|2018-03-20|Roxar Flow Measurement As|Multi-model fraction verifican|
    US11187078B2|2017-01-20|2021-11-30|Total Sa|Method for evaluating connectivity between a first well and a second well in a hydrocarbon production field and related system|
    US10584578B2|2017-05-10|2020-03-10|Arizona Board Of Regents On Behalf Of Arizona State University|Systems and methods for estimating and controlling a production of fluid from a reservoir|
    US10900344B2|2017-11-07|2021-01-26|Saudi Arabian Oil Company|Determining wellbore leak crossflow rate between formations in an injection well|
    CN108764760A|2018-06-25|2018-11-06|成都北方石油勘探开发技术有限公司|A kind of inter well connectivity analysis and method for early warning based on data mining|
    CN111119808A|2018-10-31|2020-05-08|北京国双科技有限公司|Construction method of oil field injection-production relation model and identification method of oil field injection-production relation|
    CN110500083A|2019-08-05|2019-11-26|中国石油天然气股份有限公司|A kind of oil-water well dynamic connectivity method of discrimination|
    RU2715593C1|2019-09-28|2020-03-02|Общество с ограниченной ответственностью "Тюменский институт нефти и газа" |Method of operative control of water flooding of formations|
    WO2021252070A1|2020-06-11|2021-12-16|Exxonmobil Upstream Research Company|Methods of interpreting a plurality of time-series datasets generated from operation of hydrocarbon wells|
    CN112302594A|2020-12-01|2021-02-02|中国石油大学|Method, device and equipment for determining connection structure of heterogeneous water-drive reservoir|
    RU2758326C1|2021-04-12|2021-10-28|Публичное акционерное общество «Татнефть» имени В.Д. Шашина|Method for regulating the operating mode of a well equipped with an electric center pump installation in an inter-well pumping system|
    法律状态:
    2019-09-03| B08F| Application dismissed because of non-payment of annual fees [chapter 8.6 patent gazette]|Free format text: REFERENTE AS 5A, 6A E 7A ANUIDADES. |
    2019-12-24| B08K| Patent lapsed as no evidence of payment of the annual fee has been furnished to inpi [chapter 8.11 patent gazette]|Free format text: EM VIRTUDE DO ARQUIVAMENTO PUBLICADO NA RPI 2539 DE 03-09-2019 E CONSIDERANDO AUSENCIA DE MANIFESTACAO DENTRO DOS PRAZOS LEGAIS, INFORMO QUE CABE SER MANTIDO O ARQUIVAMENTO DO PEDIDO DE PATENTE, CONFORME O DISPOSTO NO ARTIGO 12, DA RESOLUCAO 113/2013. |
    2021-10-05| B350| Update of information on the portal [chapter 15.35 patent gazette]|
    优先权:
    申请号 | 申请日 | 专利标题
    US13/288,840|US9140108B2|2011-11-03|2011-11-03|Statistical reservoir model based on detected flow events|
    PCT/US2011/061321|WO2013066358A2|2011-11-03|2011-11-18|Statistical reservoir model based on detected flow events|
    [返回顶部]