METHOD FOR ESTIMATING THE SERVICE LIFE OF FLEXIBLE PIPES UNDER CO2 CORROSION IN OIL PRODUCTION

Abstract

The invention consists of a methodology for calculating the service life of flexible tubes subject to the SCC-CO.sub.2 phenomenon, in which the methodology allows establishing the criticality level of each duct within the scope of the phenomenon, allowing the establishment of actions for the most critics. In addition, another important gain with the development of the methodology is related to the fact that it enables safe operation even in a degraded pipe.

Claims

1. A method for estimating the service life of flexible pipes under CO.sub.2 corrosion in oil production characterized by predicting the failure of flexible pipes based on the calculation of the crack growth rate obtained by Artificial Intelligence techniques and on the critical crack estimation.

2. The method, according to claim 1, characterized in that the failure prediction involves a first step which gathers data on operational history, flexible pipe structure and structural integrity loads; a second step gathers the operational history and data from the flexible tube structure serving as inputs for the artificial intelligence algorithm that generates a predictive model of crack growth and in parallel gathers data from the flexible pipe structure and loads of structural integrity for inputs of the critical crack size calculation model, and a third step combines both models and defines a service life predictive model by applying safety factors.

3. The method, according to claim 2, characterized in that the process of preparing the data from the database of the first step consists of: a) extracting geometric data as well as some mechanical properties of the material and the flexible pipe; b) obtaining the operational history of internal pressure, fluid temperature and CO.sub.2 content, and consequently, after performing permeation analysis, then obtaining the history of CO.sub.2 fugacity in the annulus; c) performing global analysis to obtain the effective tensile strain of the line in static condition; d) obtaining the maximum crack sizes found in tensile and pressure armatures through dissections; e) all the information mentioned above is compiled in a processing worksheet and average values are obtained for each layer; f) inputting the data processed in e) into the database.

4. The method, according to claim 2, characterized in that the second step calculates the crack growth rate using the CO.sub.2-SCC and fatigue failure modes.

5. The method, according to claim 2, or characterized in that the second step calculates the crack growth rate using an AI algorithm for the CO.sub.2-SCC failure mode, the calculation methodology consisting of presenting values of the constants of multivariable equations for tensile armature and pressure armature.

6. The method, according to claim 2, characterized in that the second step adds the fatigue failure mode to the crack growth rate if the pipe is a riser, applying to tensile armature wires using equation (1).

7. The method, according to claim 1, characterized in that the critical crack estimation estimates the stress field on the structural layer wire, whether for tensile or pressure armatures, and considering a crack geometry.

8. The method, according to claim 6, characterized by the critical crack estimate for tensile armatures at a point located on the outer surface of the wire, the calculation of the effective stress after FAT being calculated considering the steps of: a) deformation imposed on the wire during the manufacturing process; b) deformation imposed on the wire during FAT and obtaining such associated stress; c) relief of stress after FAT from the previously obtained associated stress and obtaining post-FAT stress.

9. The method, according to claim 6, characterized in that the critical crack estimate for pressure armatures is equivalent to that of the tensile armature.

10. The method, according to claim 7, characterized in that calculation of the operational stress of the critical crack in the tensile armature be composed of tensile strain and internal pressure values, wherein the tensile strain values are obtained through a global analysis of the extreme condition, and the pressure values are adopted for maximum operating values.

11. The method, according to claim 4, characterized in that the second step calculates the crack growth rate using an AI algorithm for the CO.sub.2-SCC failure mode, the calculation methodology consisting of presenting values of the constants of multivariable equations for tensile armature and pressure armature.

12. The method, according to claim 4, characterized in that the second step adds the fatigue failure mode to the crack growth rate if the pipe is a riser, applying to tensile armature wires using equation (1).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The present invention will be described in more detail below, with reference to the attached figures which, in a schematic and non-limiting manner of the scope of the invention, represent examples of embodiments. In the drawings:

[0023] FIG. 1 illustrates the traditional inner layers of unbounded flexible pipes;

[0024] FIG. 2 illustrates schematically the permeation of CO.sub.2 into layers that represent the annulus;

[0025] FIG. 3 represents schematically the crack growth by CO2-SCC as a function of time;

[0026] FIG. 4 shows an overview of the variables that affect the crack growth rate by CO2-SCC;

[0027] FIG. 5 shows an overview of the alternative variables alternative that affect the crack growth rate by CO2-SCC;

[0028] FIG. 6 illustrates an scheme for obtaining the multivariable equation for the AI algorithm crack growth rate;

[0029] FIG. 7 shows a flowchart of the general calculation for estimating the service life of flexible pipes subjected to the CO2-SCC phenomenon, once the model (multivariable equation) is defined for a given structural layer (tensile or pressure armature, for example);

[0030] FIG. 8 shows step 1 of the method in detail;

[0031] FIG. 9 illustrates in detail the achievement of the multivariable equation for predicting the crack growth rate due to CO2-SCC. Such an equation is used in step 2;

[0032] FIG. 10 illustrates the deformation process imposed on the tensile armature wires during the manufacture of flexible pipes, as adapted from U. S. Fernando, M. Davidson, K. Yan, M. J. Roy, T. Pirling, P. J. Withers e J. A. Francis, “Evolution of Residual Stress in Tensile Armour Wires of Flexible Pipes During Pipe Manufacture,” Proceedings of the International Conference on Offshore Mechanics and Arctic Engineering (OMAE), 2017;

[0033] FIG. 11A illustrates the tensile stress fields acting on the structural layer wire of a flexible pipe after seating on the tubular body;

[0034] FIG. 11B illustrates the tensile stress fields acting on the structural layer wire of a flexible pipe during FAT;

[0035] FIG. 11C illustrates the tensile stress fields acting on the structural layer wire of a flexible pipe after FAT and subjected to operational loading;

[0036] FIG. 12A shows comparisons between the actual and predicted crack sizes for tensile armature; and

[0037] FIG. 12B shows comparisons between the actual and predicted crack sizes for pressure armature.

DETAILED DESCRIPTION OF THE INVENTION

[0038] For the CO2-SCC phenomenon to occur, the interaction of four main factors is necessary, namely: a) tensile stress, which can be either static or dynamic, residual or even elastic; b) an enabling environment, i.e., the annulus containing CO.sub.2 and water; c) a susceptible material, that is, one having low tenacity and high mechanical strength, which are subject to the cracking effect; d) and finally the factor time which has an incubation period and a subsequent crack growth, leading to rupture of the layer.

[0039] The method of the present invention is intended to achieve the crack growth rate that aims to prevent the advancement of the discontinuity over time, and the critical crack size, which causes rupture of the structural layer wire (for example, tensile or pressure armature). With this information in hand, it is then possible to estimate the time required for rupture of the wire in the layer to occur, as shown schematically in FIG. 3.

[0040] The crack growth rate in structural layers depends on several variables, namely: CO.sub.2 fugacity in the annulus, temperature in the bore of the flexible pipe and external temperature, deformation associated with the manufacturing process of the flexible pipe, CO.sub.2 content of the transported fluid, internal operational pressure and factor of use of the wire of the layer during operation. These variables, in turn, may depend on the operational history, characteristics of the flexible pipe structure, manufacturing history and loads, as shown schematically in FIG. 4. Alternatively, other variables can also be considered to contribute to the crack growth rate, such as: CO.sub.2 partial pressure, operational stress, inner barrier thickness. To correlate so many variables with the crack growth rate, an artificial intelligence (AI) software was required. To feed this software, a database was created that gathers information on dissections of flexible pipes subjected to CO2-SCC and operational data, as shown in FIG. 5. Accordingly, the software searches for a multivariable equation whose answer is the crack growth rate, which, in turn, is a function of operational parameters and structure characteristics.

[0041] Then, the methodology used here obtains the crack growth rate, which is influenced by the CO2-SCC and, in the case of risers, also by fatigue, and the point of rupture (critical crack), delimiting the curve, as shown in FIG. 3.

[0042] Once the growth rate models due to CO2-SCC have been obtained, a service life calculation method can then be performed, as shown in FIG. 7. The first step of the proposed method consists of gathering all the input data necessary to perform service life calculation, namely, the operational history, structure data and loads used for structural integrity management.

[0043] In the second stage, the operational history and flexible pipe structure data serve as input for the crack growth calculation method. An equation with several variables for each structural layer of the flexible pipe is used to estimate the crack growth.

[0044] The crack growth predictive model is obtained through an artificial intelligence algorithm that uses data stored in a structured database consisting of dissection data of pipelines subjected to CO.sub.2 stress corrosion and their respective operational histories and structural characteristics. Through this AI algorithm, multivariate equations were determined to estimate the crack growth rate in the tensile armature (TC.sub.AT) and pressure armature (TC.sub.AP). Such equations are used in step 2 shown in FIG. 7.

[0045] The third step of the method consists of defining the service life by applying safety factors to the results achieved in step 2.

[0046] Generation of the database depends on information from different sources, such as dissection data, analysis results and structural characteristics of the flexible pipe. FIG. 9 presents the flowchart for obtaining relevant information for each case. The main points related to the generation of the database and obtaining the multivariate crack growth equations due to CO.sub.2-SCC are highlighted below:

[0047] A) Initially, based on the dissected flexible pipe position, the segment is identified from the installation document. From this, the structure data sheet is obtained and geometric data is extracted, as well as some mechanical properties of the material and the flexible pipe, such as yield stress, flowline tensile failure, among others.

[0048] B) Upon consulting the flow analysis report, the operational history of internal pressure, fluid temperature and CO.sub.2 content of the drained fluid is obtained. These data together with the structure information are used to perform the permeation analysis, thus obtaining the history of CO.sub.2 fugacity in the annulus.

[0049] C). Furthermore, in case of risers, a global analysis is required to obtain the effective tensile strain of the line in the static condition.

[0050] D). From dissections of the flexible pipes, it is then possible to obtain the crack sizes found in the tensile and pressure armatures (adisec.AT and adissec.AP, respectively).

[0051] E) Thus, for each case in the database, all the information mentioned above is compiled in a processing worksheet. This worksheet processes the input data and then provides the average parameters for each layer (see Table 1).

TABLE-US-00001 TABLE 1 Example of output data from a database case Operating time with Average Maxi- CO.sub.2 temper- Average mum Total fugacity Average ature Average Average Average crack Crack operating above 1 CO.sub.2 in the CO.sub.2 Internal assembly growth Appli- Size time bar fugacity borehole content pressure Average strain rate Case Structure cation Section [μm] [days] [days] [bar] [° C.] [%] [bar] fugacity rate [μm/day] Tensile armor Flowline 2300 1046 989 2.87 43.33 24.74 453.24 0.12 0.0087 2.33 Pressure armor Flowline 1387 1046 989 2.87 43.33 24.74 453.24 0.26 0.0503 1.40

[0052] F) Once each case is processed, it is entered into the database, as shown in Table 2.

TABLE-US-00002 TABLE 2 Database data % t.sub.OP t.sub.OP,fCO2,annulus>1 f.sub.CO2,annulus,average T.sub.b,average CO.sub.2,average P.sub.int,average ε.sub.m,AT Case [days] [days] [bar] [° C.] [%] [bar] [—] FU.sub.AT,average ε.sub.m,AP FU.sub.AT,average a.sub.max,AT TC.sub.AT a.sub.max,AP TC.sub.AP [—] [—] [—] [μm] [μm/day] [μm] [μm/day]

[0053] G) Finally, through consolidation of the database, analyzes in the AI software are performed to obtain multivariable equations for the tensile and pressure armatures.

[0054] As shown earlier, the crack growth rate (TC) by CO.sub.2-SCC depends on several variables. In the model considered in the methodology for using the AI algorithm, the following variables are used: 1) CO.sub.2 content of the transported fluid (% CO.sub.2); 2) Internal pressure (Pint); 3) Layer wire utilization factor during operation (FU.sub.OP); 4) Temperature in the flexible pipe bore (Tbore); 5) Deformation associated with the manufacturing process of the flexible pipe (ε.sub.m) 6) CO.sub.2 fugacity in the annulus (fCO.sub.2, annulus).

[0055] Alternatively, other variables can be considered, see FIG. 5. In this approach, the following variables are used: 1) CO.sub.2 content of the transported fluid 2) Internal pressure; 3) CO.sub.2 partial pressure in the bore (pipe core—internal diameter section) of the flexible pipe 4) Stress acting on the layer wire during operation; 5) Temperature in the flexible pipe bore; 6) Total internal barrier thickness; 7) CO.sub.2 fugacity in the annulus.

[0056] In case of risers, a fraction of crack growth due to fatigue (da/dN) is added to the growth rate for tensile armature wires using the Paris law ((N. E. Dowling, Mechanical Behavior of Materials—Engineering Methods for Deformation, Fracture and Fatigue, 4th Ed., 2013), which is given by

da/dN=CΔK.sup.m  (1)

Where C and m are material parameters of the tensile armature of a given flexible pipe manufacturer, whose value can be provided by such manufacturer or obtained through material assays. AK corresponds to the variation of the stress intensity factor.

[0057] The critical crack calculation is based on the structural integrity standard BS7910: 2015—“Guide to methods for assessing the acceptability of flaws in metallic structures”. To carry out such an evaluation, it is necessary to estimate the stress field acting on the structural layer wire (tensile or pressure armature) and to also consider a certain crack geometry. Thus, using Fracture Mechanics concepts it is possible to establish the crack size capable of causing wire rupture based on material properties and imposed loads.

[0058] To calculate the crack size, first, one must know the stress field acting on the component. In the case of flexible pipes, during manufacture of the flexible pipe, the metallic layer wires are subjected to a tortuous process of plastic deformation, where bendings in different planes and directions are applied. A schematic example of the flexible pipe manufacturing process focused on the tensile armature wires is shown in FIG. 10. Pressure armature wires go through similar deformation cycles.

[0059] This process ends up generating a complex state of residual stresses on the wires even before the flexible pipe starts operating. At the end of the process, the final laying aims to ensure that the wire is not subjected to any springback. However, from the manufacturing point of view, this control is complex and, in general, the wire laid in the pipe ends up having unknown levels of residual stress as well as a level of stored elastic energy. From the calculation methodology point of view, it is idealized that the effective stress field (σ.sub.ef, pre-FAT) acting prior to the factory acceptance test is composed of a residual portion (σ.sub.res, pre-FAT) and an elastic portion (σ.sub.e1, pre-FAT). This elastic portion is idealized as a bending stress that makes the wire to reach the desired end geometric configuration, while the residual portion is idealized as a self-balancing profile. This is shown schematically in FIG. 11A.

[0060] The pipeline manufacturing process is completed when it is subjected to FAT. During this test, the structural layers are subjected to stresses greater than those foreseen in the design condition. These stresses combined with the existing residual and elastic stresses can lead to partial flow of the wire section, as shown in FIG. 11B. Such a localized plastification then leads to a new state of effective stress after the FAT (σ.sub.ef, post-FAT), due to the relief of residual stresses. In operation, the pipelines will be subjected to external loads that will be transformed into operational stresses (σ.sub.OP), which, in turn, will be added to σ.sub.ef, post-FAT. Accordingly, a local stress state (σ.sub.local) will be formed in the wire, as shown in FIG. 11C. The operating stress is determined from conservative tensile and pressure values used in the management of the structural integrity of flexible pipelines, which are determined as follows:

[0061] A) Tensile strain: obtained from a global analysis in the extreme condition.

[0062] B) Internal pressure: maximum pressure values that may occur during operation of the flexible pipeline are adopted.

[0063] Considering a point located on the outer surface of the wire, calculation of σ.sub.ef, post-FAT,AT is performed in three stages, namely:

[0064] A.1) Deformation imposed on the wire during the method of manufacturing and pipe laying (ϵ.sub.m,AT);

[0065] A.2) Total deformation imposed on the wire during FAT (ε.sub.t,AT) and achievement of the stress (σ.sub.ef,FAT,AT) associated to such deformation;

[0066] A.3) Stress relief after FAT from σ.sub.ef,FAT,AT and obtaining of σ.sub.ef,post-FAT,AT.

[0067] The procedure for calculating the critical crack of the pressure armature is quite similar to that adopted for the tensile armature.

Examples

[0068] The following examples are presented to fully illustrate the predictability of the present invention and how to practice the same, without, however, being considered limitative of its content. They are intended to analyze the response of the methodology to the results predicted by the method with actual values. The following assessments were carried out: a) Crack growth rate: the values of maximum crack size predicted by the model were compared with values actually observed from dissected ducts; and b) Service life: in these analyses, the objective is to assess the ability to predict the occurrence or not of failures considering the cases present in the database.

[0069] Through the crack growth rate method, it was possible to estimate the predicted final size of the crack considering the cases present in the database that were numbered from 1 to 8, see FIGS. 12A and 12B. Thus, the sizes estimated through the equations were compared with the actual values obtained in the dissection of flexible pipes. FIG. 12A discloses a comparison for the tensile armature, where the predicted values for the internal armature (tensile armature 1) and external armature (tensile armature 2) are presented. It can be noted that the method can predict the crack size with excellent accuracy. The same is seen for the pressure armature by analyzing FIG. 12B.

[0070] Consistency of the methodology for service life was assessed for 8 cases present in the database. In some instances, the flexible pipeline failed while in others, the flexible lines were withdrawn before failure occurred. Thus, we observed the prediction of the methodology for the different cases studied. Thus, the ability of the methodology to predict failure and non-failure situations was assessed. This is summarized in Table 3.

[0071] As can be seen, the methodology was correct for the vast majority of cases. The only inconsistency was case 2, where the methodology predicted a failure that did not occur in operation. However, this fact, in a way, corroborates that the methodology is conservative.

TABLE-US-00003 TABLE 3 Summary of results of the analysis of service life consistency Tensile armature Pressure armature Actual Actual Case situation Model situation Model Conclusion BD 1 Failure Failure No No Consistent failure failure methodology. BD 2 No Failure No No Conservative failure failure failure and consistent methodology. BD 3 No No No No Consistent failure failure failure failure methodology. BD 4 Failure Failure No No Consistent failure failure methodology. BD 5 Failure Failure No No Consistent failure failure methodology. BD 6 No No No No Consistent failure failure failure failure methodology. BD 7 Failure Failure No No Consistent failure failure methodology. BD 8 Failure Failure No No Consistent failure failure methodology.

[0072] It should be noted that although the present invention has been described in relation to the attached flowcharts, it may be subjected to minor modifications and adaptations by the skilled person, depending on the specific instance, as long as it is within the scope of the invention as defined herein. A mitigation coefficient in service life calculation, which can be mainly applied to certain applications will be associated with the defined risk levels of the operation