© KAPP KAPPA A Engineering - v4.02 - September 2006
© KAPP KAPPA A Engineering - v4.02 - September 2006
KAPPA
www.kappaeng.com
KAPPA is primarily a petroleum engineering software company. Our integrated software platform, Ecrin , is the industry standard for Dynamic Flow Analysis. Ecrin includes modules for Pressure Transient Analysis ( Saphir Saphir ) and Production Analysis ( Topaze Topaze ). Soon we will integrate our freestanding production log analysis module ( Emeraude Emeraude ) and release two new modules for nodal analysis and full field reservoir simulation. To seamlessly connect and process client data in the Ecrin modules KAPPA has developed Diamant Master , a server solution that centrally processes permanent gauge data and shares production data, technical objects and documents in a coherent environment for real time reservoir management. Founded in 1987, KAPPA now has over 3000 active commercial software licenses, used by over 300 companies worldwide. KAPPA is independent; 80% owned by its employees. Our main development office is in Sophia Antipolis, France and we have regional offices in Houston, Perth and Bahrain. KAPPA is present in ten other countries with local offices and agents. KAPPA offers complementary Training and Consulting Services (TCS) based near Gatwick, UK. We train hundreds of engineers every year in our chosen disciplines. KAPPA is a Microsoft Certified Partner.
Table of contents
Pressure Transient Analysis
10
KAPPA
3
Production Analysis
20
Ecrin
4
Production Logging
24
PDG Reservoir Surveillance
6
What is coming next?
30
3
Ecrin
Ecrin - Main window with a PTA (Saphir) and a PA (Topaze) session running
An integrated platform for Dynamic Flow Analysis Until recently KAPPA was developing PC applications aimed at being best-in-class. This objective still stands inviolate, but users told us they needed ergonomic tools that would integrate, navigate and communicate within a single program in order to save engineer time by avoiding process duplication, painful import/export and by cutting training time. As a result, in our fourth generation of software products, we integrated our applications into a single environment called Ecrin. In 2005, Ecrin v4.0 integrated the three modules required to process Permanent Downhole Gauge (PDG) data: data management ( Diamant ), Pressure Transient Analysis ( Saphir and Saphir NL ) and Production Analysis ( Topaze ). In 2006 KAPPA simultaneously released Ecrin v4.02 and a server application gathering, filtering and sharing PDG and production data ( Diamant Master ). Diamant Master is operated by the Diamant module in Ecrin. In complement, recent industry developments on Deconvolution were integrated in Saphir and enhanced multiwell simulation was added to Topaze. The process continues. A nonlinear version of the PA module ( Topaze NL ) is under development. The integration and enhancement of our production log analysis software ( Emeraude ) is taking place, and in the process we are developing new modules such as reservoir simulation ( Rubis ) and Nodal Analysis ( Amethyste ).
4
Why Ecrin? The inital goal of Ecrin was to concentrate the processing and analysis of PDG data into a single environment. But users needed more. PVT, models and general information are common, and it is difficult to excuse having to load the same information repeatedly for different applications; especially from the same software vendor.
Licenses and Readers The Ecrin user may turn on and off any module without license restriction. Ecrin, when started, checks the license information for each enabled module from a hardware key, FlexLM or network. If the module is licensed it will be fully executed. If it is not licensed it will execute in Reader Mode. This will allow the engineer to open the corresponding files, print, export, run a report, and even copy individual objects into licensed modules (the opposite is not true as a Reader cannot modify anything).
In other words engineers dealing with permanent gauge data and performing pressure transient and production analysis wanted this to be done in the same application.
Leave my Saphir alone... Conversely, engineers only interested in one KAPPA product may not want to have a workstation environment forced upon them. So Ecrin can also be installed as a single software product, say Saphir. Even after installation the user may switch Ecrin behavior between a standalone application and an integrated workstation.
For example, if you have a Saphir license and no Topaze license you will be allowed to open Topaze document and, using the browser, drag-and-drop the data, PVT, model, etc into a Saphir document.
Why the name Ecrin? You will have noticed that the names of our software gemstones are spelt incorrectly. In truth it is even worse as they are spelt in French (eg Saphir instead of Sapphire). The main reason was to avoid trademark problems in an industry that mostly uses English names and to avoid boring acronyms. Ecrin is the French word for jewellery box. With Ecrin you buy the gemstones and we offer you the box.
Ecrin Browser Switching between two small applications within a big one is a bit ho-hum. In reality this can be done between any Windows applications with a simple Alt-Tab. What is interesting is the ability to share and transfer objects. This is done using the Ecrin Browser. Put simply, if there is, for example, a PVT object available in Saphir, through Ecrin this can be dropped into Topaze. You need only enter the PVT data once.
Ecrin toolbar with a mixture of licensed and Reader modules
Operating Diamant Master The Diamant / Diamant Reader module in Ecrin can fully operate Diamant Master (see next section). Data and technical objects stored in the Diamant Master database may be created, edited, deleted and dragged-anddropped into any other Ecrin module. The privilege to perform operations in Diamant Master is not related to the Diamant licensing but to the privilege associated to individual users.
Another example might be the 2D-map. Sitting at the top level in Ecrin this would be available to all applications by simple drag-and-drop. This can go as far as transferring a complete Saphir document into a new Topaze document obtaining the information, pressures, rates and model on a single click.
Ecrin releases In 2005 Ecrin v4.0 integrated a data module (Diamant), a PTA module (Saphir) and a PA module (Topaze) to constitute the kernel of our workflow to process PDG data. It transpired that the workflow was right but a server application was needed to share the filtered data within a workgroup. This was the origin of the v4.02 project, which integrated Diamant Master and two technical developments that were considered urgent: deconvolution in Saphir and a pressure controlled multiwell simulation in Topaze. With Diamant Master in place, the expansion of Ecrin continues with additional features to the existing modules, the integration of PL (Emeraude) and the coming release of a reservoir simulator and a nodal analysis package (see page 30).
Ecrin browser
5
PDG Reservoir Surveillance
Diamant main window
Permanent Downhole Gauges (PDG) are a remarkable source of information of both long term production data and the capture of occasional build-ups that may be described as ‘free well tests’. Data are acquired at high frequency and over a long duration. The down side is the large number of data points gathered, which can amount to hundreds of millions per sensor, far beyond the processing capability of today’s fastest PC. There are a number of challenges: storing and accessing the raw data, filtering it and then transferring this to the relevant analysis module and finally sharing both filtered data and analyses. Diamant Master is a server solution that addresses all of these issues. Installed on a dedicated machine, it permanently mirrors raw data for fast processing, reduces the number of points with wavelet based filtration, stores and shares the filtered data and exports filtered data sets to third party databases. It also stores and shares analyses and various files amongst engineers of the same workgroup. The Diamant module in Ecrin works in two ways: • In the free Reader version, Diamant operates Diamant Master and transfers data and technical objects between Diamant Master and the Ecrin analysis modules. • As a licensed module, Diamant locally reproduces some of the features of Diamant Master for the smaller operators, working nonreal time and without the sharing capability of Diamant Master.
6
Handling PDG data
Modern PDG Data processing
What PDG data provides PDG acquire pressure data at high frequency and over a long duration. A typical data set, as shown below, will include two types of information; each spike is an unscheduled shut-in that may be treated as a ‘free’ well test for Pressure Transient Analysis. In addition the long term global producing pressure response, ignoring these peaks, can be used in association with the well production to perform Production Analysis and/or history matching. The data is there and it is already paid for. It is ‘simply’ a matter of getting at and interpreting the data.
Data preview: a quick data scan of one point in every thousand gives the user an overview and easily identifies anomalies. A selection on the data window can be done and outliers are immediately discarded. Filtration setting: within the load window an initial sample of a fixed size, typically around 100,000 points or one week of data, is extracted. In an interactive and iterative process, the software equivalent of running a pencil through a noisy path of data the engineer visually sets the wavelet filter. A post-filtration based on a maximum ∆t and ∆p is then used to reduce the number of points of the de-noised signal.
Initial sample
Filtered data
Load: the data is loaded in overlapping increments. During the load the user can visualise the overall picture, with the result of the filtration since the start and the filtration on the current increment. At any time the process can be interrupted and the filter parameters modified.
Typical PDG data response gathered over two weeks
Wavelets filtration Nice idea, one not so little problem; the available data is vast and growing. For one single gauge there are typically 3 to 300 million data points. This will bring even the fastest of today’s PCs to a grinding halt. But we need both short term high frequency data for PTA and long term low frequency data for PA. To obtain both in the same process KAPPA adapted a wavelet algorithm, acting as a high pass filter close to the pressure breaks typical to a shutin and as a low pass filter on the remainder of the data. This filter will typically divide the number of points by 100 without loosing significant information.
Global load window
Current load window
Update and partial reload: the process maintains a persistent link to the original data source. For each gauge, regularly or on user request, the process reconnects to the data source and then loads and filters incremental data using the filter as set for the particular gauge. It is also possible to change the filter setting, for new data or retroactively, or to partially re-populate a data segment over, for example, an identified build-up with a completely different filter level or even no filtration. Data analysis: filtered data can be transferred by drag and drop to an analysis module. Shut-ins are analysed and compared using the PTA module (Saphir) while producing pressures will be history matched using the PA module (Topaze).
Wavelets denoising: (1) raw data = 10,000 points; too low (2), too high (3) and selected (4) thresholds; (5) post-filtration; (6) filtered data = 70 points
7
Workflow using Diamant Master
Once the filter is set, DM will filter as a background task, regularly updating the filtered channels as soon as sufficient new points have been mirrored. The filtered data is stored in the local DM database to be subsequently sent to Ecrin analysis modules on a single drag and drop. The filtered data may also be exported to a third party database. It is possible for the Ecrin users to return to an old part of the data and request DM to reload the data in the corresponding time range with a different filter or no filter at all, hence locally repopulating the filtered data whenever needed. Diamant Master also stores all types of KAPPA technical objects and files in a hierarchic and intuitive structure, in order to be shared and re-used by Ecrin interpretation modules.
Diamant Master as seen by the end user A typical workflow using Diamant Master (DM) is shown in the figure below. DM is a permanent process installed on a dedicated machine running Windows 2000 or 2003 Server™. Engineers operate DM from Diamant in Ecrin. A Diamant license is not required to operate DM. Even as a Reader, Diamant has full control of DM; privileges and restrictions are only associated to the users identified by their Windows™ login names.
Loglog plots before and after a partial reload Diamant Master Workflow
For the end user the interface is similar to a local Diamant session, however all operations are performed and shared on the DM server. Diamant Master, at all times, remains connected to the original data sources from which it sequentially imports the raw, unfiltered data. From Ecrin, users with the correct privilege can navigate the input database and indicate which tag(s) should be imported. Data is mirrored from the raw database to a local, fast access format (BLI). At the start of deployment DM will remain in an infinite loop in order to retrieve the legacy data. Once DM has updated a given gauge it will regularly contact the new data and load on a timer set by the DM administrator. For each mirrored data set, users with the right privilege may create one or several filtered channels using the wavelet algorithm and post-filtration previously described.
WEB access Diamant Master raw and filtered data are exclusively created from the Diamant module in Ecrin, and Diamant remains the best way to handle data, technical objects and files when using KAPPA applications. However these can also be accessed from an Internet browser by connecting to the DM server IP address or its name in the domain. The engineer can view the status of the different processes and gain access to the data tables and technical objects stored by Diamant Master. It is possible to recover the filtered data in Excel™ format without using Ecrin. An ActiveX control can also be loaded to navigate the data structure in the same browser environment as Diamant.
Diamant session in Ecrin operating a Diamant Master field
Some Diamant functionalities from an ActiveX control
8
Diamant Master administration The WEB service is also used to administer Diamant Master. In the administration part of the DM local site, new database adaptors can be added, users can be created or deleted, user privileges are controlled, individual processes can be started, monitored and/or killed. The database can be cleaned, and the different timers and control operations can be modified.
Each solution would require a specific adaptor to navigate and access the data. KAPPA has implemented a unique database protocol (EDBI) that permits the connection to customized adaptors. In most cases the adaptor will be written by KAPPA. Each adaptor is delivered as a DLL that includes the data access and the user interface to navigate the database. It acts as a plug-in. At the first connection, Ecrin will automatically download the plug-in from Diamant Master and the user will navigate without further installation. The database interface also has adaptors to export the filtered data to external client databases. Mirror, Filter and Calculations: When an Ecrin user decides to mirror PDG data or to create new filtered data, the server will store the new instructions in the KAPPA database. The mirroring process (DMMP) and the filtering process (DMFP) are independent. In their operation cycle they regularly check their lists of tasks. The calculation process (DMCP) creates and permanently updates tags that are derived from other tags, for example the summation of production at well group and field levels. Virtual metering simulation of rates from pressures using a Topaze model, will be accessible in a later version of Diamant Master.
WEB based administration of Diamant Master
Diamant Master processes The diagram below shows the processes that constitute Diamant Master. These processes operate continuously and independently.
KAPPA database: Diamant Master stores objects and field information in SQL Express™ installed by default with Diamant Master. However it is possible to store DM data on a higher level of SQL server or under Oracle™. Raw and filtered gauge data are stored in fast access files (BLI), only file pointers are stored in the database.
Server process: The interface between the KAPPA storage database and any other DM module is controlled by the server process (DMS), the central process of Diamant Master. It interfaces with Ecrin clients using DCOM. It controls the link with the WEB service and stores instructions to the other processes in the KAPPA database ensuring that simultaneous requests from various users remain consistent. It protects data locked by a user against possible interference from other users.
Workflow using Diamant only For very small workgroups, the Diamant module in Ecrin has a subset of Diamant Master PDG capabilities. The database connection (EDBI), and therefore the ability to access filtered data from various sources is the same. Mirroring is allowed but incremental loads are triggered by the user. The filtration process is identical but data are stored in a local Diamant file. Direct sharing is not possible however filtered data may be exported to files. The simplified PDG data process of Diamant is shown below. It is not necessary to acquire Diamant in order to operate Diamant Master, unless local and independent data processing is required.
Diamant Master processes
Database interface (EDBI): The beauty of standards is that there are so many to choose from. So it is in the Oil Industry; there is no standard way to store PDG data. There are many providers, and each has their own data model. It is common for Operators to have several providers and hence different data models will co-exist. Most databases have low level access (ODBC, OLEDB, OPC, etc), but this is, at best, cumbersome for end users.
PDG data workflow in Diamant
9
Pressure Transient Analysis
Saphir main window
Saphir was first developed eighteen years ago by two engineers who needed a tool for their own interpretation work. Since then Saphir has grown to a dominant position with over 2000 commercial licenses across operators, service companies and consultants alike. The reason for this wide acceptance was the ease of use and the technical decision to converge modern methodology with increasing PC processing power. The Saphir methodology has always been based on the Bourdet derivative as the main diagnostic tool, complemented by matching the measured data to the model taking into account the detailed production history. The ever increasing processing power of PCs has enabled us to aggressively expand the technical capability of Saphir. This has resulted in the development of extensive and fast numerical modeling capability, extended to nonlinear problems ( Saphir NL p.18), and more recently deconvolution. Simultaneously the testing world has changed. Operators need transient data analysis but are reluctant to test in the classical sense. Today's engineers have to grab what data they can, beyond normal well test operations. This requires particular p rocessing and additional modeling capabilities, permanently updated in Saphir.
10
Main additions in v4.0 (2005) Saphir was integrated as the PTA module of Ecrin. Using the shared browser, drag-and-drop of technical objects between documents from the same or different modules became possible. A sensitivity option in the control panel matched the data with multiple ranges of parameters of the selected model in the same interpretation tab. wavelet filtration was added at load time, with some database (ODBC, OLEDB) load capabilities. The creation and editing of the rate history was also improved. New numerical models included limited entry, fractured and horizontal wells. Horizontal and vertical anisotropies and multilayer reservoirs with crossflow permitting multiple wells and partial completions were added. It was also possible to visualize geometries in 3D. Analytical models could account for horizontal anisotropy and Saphir NL could handle unconsolidated formations with reversible / irreversible between porosity / permeability and pressure relations
Interpretation using the numerical module
Key features • User-friendly software with a powerful kernel • Very short training, no retraining for occasional users • Software under constant development • Methodology based on the Bourdet derivative • Modern deconvolution (v4.02) • No limitation to the number of gauges or data points • Real time interface with acquisition systems • Powerful wavelet filtration for production data • Extensive analytical model catalog • Unique 2-D numerical module extending the modeling capabilities to situations with arbitrary outer boundary shapes, any fault trajectories, composite zones, etc • Unlimited number of analyses on different gauges, build-ups, models and/or model parameters • Fast and robust optimization routine • Gas material balance correction for closed systems • Artificial Intelligence based model adviser • Free Reader for reporting and exporting • 24-hour on-line and telephone technical support • Extensive training and consulting services • Real gas and real dead oil diffusion (Saphir NL) • Non-Darcy flow (Saphir NL) • Water+Hydrocarbon 2-phase flow (Saphir NL) • Water injectors in oil or gas reservoirs (Saphir NL) • Unconsolidated formations (Saphir NL) • Simulation with minimum pressure control (Saphir NL)
Local 3D refinement limited entry well
3D cross-section multilayer crossflow with two wells open in different layers
Main additions in v4.02 (2006) The main focus of Ecrin v4.02 was on the operation of Diamant Master and its compatibility with all Ecrin modules, including Saphir. However Saphir v4.02 had a major enhancement with the integration of recent developments in deconvolution. This method is presented on page 15 of this document. A new vertical interference external model to match the observation probe of a formation tester was also added. Build-up response
Deconvolved response
2-Porosity PSS - Picking the transition
11
Feature details
QA/QC The QA/QC plots display all the loaded gauges. They can be synchronized interactively or automatically using nonlinear regression. For any plot, the difference between all gauges and a selected reference is dynamically calculated, and serves as a basis for detailed analysis of wellbore effects such as phase segregation. QA/QC can be used to correct reservoir trends, analyze gradient surveys and correct tidal effects.
Data Load and structure Saphir can load an unlimited number of data points from an unlimited number of gauges, and individual production history from an unlimited number of wells. Data input can be made from the clipboard, ASCII (flexible format description), P.A.S. files, databases using ODBC and OLEDB, keyboard or a real time interface with acquisition systems. Data can also be loaded by drag-and-drop from any other Ecrin document. For large data sets, wavelet filtration can be applied at load time. A Saphir document is organized in a well-defined hierarchy visualized with the Data Browser. Whilst analyzing, if several gauges have been loaded, it is possible to switch the active gauge, or use several gauges simultaneously. With a single gauge selected, multiple production, injection and/or shut-in periods can be considered simultaneously and displayed together on all relevant plots. Rate editing The tested well production history may be adjusted graphically or in a spreadsheet. A comprehensive set of editing facilities includes insertion, deletion, merging, splitting and synchronization. Graphical synchronization may be set to cursor position, closest data point and intersection of two user defined straight lines. Other features include the creation of slug rates from pressures, adaptive averaging, and refining the production history from pressure breaks using wavelets. Production events (drawdown, build-up, etc) are automatically identified when the rate history is loaded then displayed with a clear naming convention.
QA/QC
Selecting data for analysis After loading and editing data, a gauge and a production or shut-in period are selected and automatic semilog and loglog plots are created. A Saphir document can contain multiple gauges and, during the analysis, the user may dynamically change the gauge or simultaneously display several gauges on all plots. Similarly, several build-ups can be compared. The Bourdet derivative is used as the main diagnostic tool. Analysis relies on matching data with a model response generated for the full rate history. Since version 4.02, it is possible to extract a deconvolved version of the selected data. The deconvolution algorithm and method are detailed on page 15. Selecting data with Deconvolution
Edit rates
Data editing The user may select all or any part of a pressure data set either from the table display or graphically (box, time range, data range, individual points, search criteria, etc). A selection processing toolkit enables deletion, postfiltration, global arithmetic operations, removal of outliers, copy to clipboard, averaging, and wavelet denoising. Edited data may be stored as a new gauge hence preserving the original.
Extraction and matching of multiple build-ups
12
Analytical and numerical models Saphir has a wide range of built-in analytical well, reservoir and boundary models (see technical references). Constant or rate dependent skin, constant or changing wellbore storage may also be added. Additional external models may also be dynamically connected, either from the KAPPA website (see technical references) or independently developed by a third party. Pick options are offered for most parameters for a first estimate by pointing to a characteristic feature of the model on the derivative plot, eg the time when a boundary effect causes the derivative to depart from IARF.
Material balance When an analytical or a numerical model is used for a closed reservoir, average reservoir pressure is calculated and displayed. For gas, the average pressure, obtained from a p/Z calculation, is used at any simulated time to calculate the reference gas properties, hence correcting the model response for material balance.
Gas material balance correction
Changing well With the changing well option, the user can assign a different well model to different phases of the production history whilst the reservoir and boundary conditions remain the same. An example of this application is in pre and post-frac tests, where the changing model will offer a unique and consistent treatment of the total response.
Model menu
Well Intake This option is used to define an intake model that Saphir can run in conjunction with the reservoir model to simulate the pressure at gauge depth, in particular at surface. As the modification is part of the model, it can be changed as required, and possible alternatives can easily be compared. Well intake correction is also integrated in the nonlinear regression.
Multilayer analytical models Saphir integrates a comprehensive multilayer option with an unlimited number of commingled layers. Each layer has its own initial pressure, and the engineer may select any standard or external model. All layers are connected to a single wellbore. Individual stabilized and/or transient rates can be loaded and associated to any combination of contributing layers. The model simulates the pressure response and the combination of layer rates that were loaded. It will allow simultaneous optimization on both pressures and layer contributions.
Multilayer interpretation
Well intake
13
Multiwell analytical models Analytical models for homogeneous and double-porosity reservoirs and various boundary effects can account for the interference of other wells. In the model menu, the user may decide to take into account, or not, the influence of these wells. Comparing the model with, or without, this interference will allow the user to decide whether nearby wells had a significant impact on the transient response. These models also account for horizontal anisotropy. Improving, comparing and sensitivity analysis After model generation, nonlinear regression is used to optimize the model parameters. Regression may be automatic, or the user may control the list of variable parameters and acceptable parameter ranges. Optimization may be performed on the loglog plot or on the whole production history. Confidence intervals may be displayed at the end of the regression process. Sensitivity analysis may be performed by running the same model for different ranges of parameters. Multiple analyses may be overlaid and compared on all plots. A loglog comparison across multiple files is also possible for buildups years apart or with nearby wells.
IPR plot
Exporting Any document data can be sent to the clipboard or exported to ASCII, Excel™ and dBase™ files. The export option provides direct access to the candidate data, including information, PVT tables, gauges, rate history, model, etc. P.A.S. files, the Alberta Energy and Utilities Board required format for electronic well test data submission, can also be exported as TRG (pressure and temperature) and AOF (IPR / AOF results) files. Reporting Saphir provides a quick way to produce a built-in report including all relevant sections of the analysis. This includes a main results summary page, history listings, and one page dedicated to each plot with the appropriate information and results. The report can be previewed, and a number of options are given to customize the fonts, change the logos, etc. When a single plot page is required, the print option can be called from the plot itself. Saphir is also an OLE™ automation server providing access to all current interpretation parameters and results from external applications such as MS-Word™. The Saphir installation comes with a template MS-Word™ document including macros to access and retrieve those values. All Saphir plots can be sent to the clipboard in WMF, BMP, JPEG, or TIFF format.
Comparing normalized build-ups years apart from different files
Test design All Saphir analytical and numerical models are available for Test Design whereby a virtual gauge is created on which a complete analysis may be simulated. Options to simulate the actual gauge response, taking into account its resolution, accuracy and potential drift can be the basis for selecting the appropriate tools or to check if the test objectives can be achieved in practice. Specialized analyses Flexible plots can be created to complement the default loglog and derivative diagnostics with options tailored to specific flow regimes. Pre-defined types are available, such as MDH, Horner, square root and tandem root. On a flexible plot the user can create straight lines, by regression or interactively, and Saphir calculates the relevant parameters. AOF / IPR AOF / IPR analyses are available for vertical (straight line, Vogel, Fetkovitch, Jones, c&n), horizontal (Joshi, Renard, Borisov, Giger, Vlis, P.S.S.), and fractured wells (P.S.S.). IPR can be used for flow after flow, isochronal, or modified isochronal tests, and includes options to display extended, stabilized, and transient IPR.
History plot preview
14
Deconvolution in Saphir v4.02
The deconvolution process will then ‘twist’ the shape of the unknown derivative response within reasonable limits until a best match is obtained. In other words, the process will ensure that the resulting derivative shape is within the bounds of what would be expected from a model.
Deconvolution in principle The old idea behind deconvolution was to create a theoretical response to an ideally constant producing rate from the real pressure response to a complex, sometimes inaccurate production history. From this response, we arrive at an idealized drawdown solution that may be subsequently and directly matched with a model over a time period much greater than any single component build-up. Therefore, with such a deconvolved response we could see much further into the reservoir, hence making earlier reserve bookings from defined limits.
Example Look at the well production history and its corresponding PDG pressure response below. The two selected buildups, of 120 and 300 hours are coherent and therefore, on the loglog plot, their derivatives are consistent. The deconvolved response is shown as white/red thick lines. The duration of the deconvolved signal is 2,500 hours. It exhibits a close system behavior that was not detected in either build-up.
Good theory but in practice pure deconvolution is unstable, especially at late time and this is just where it is most useful. However, recent publications (e.g. SPE #77688 and SPE #84290) suggested a method which, given the proper caveats, could be of significant help.
Saphir deconvolution process The unknown is the derivative response for a unit production, which we decompose as a polyline on a loglog scale. We obtain the unit pressure response by numerical integration of the derivative, and we can simulate the pressure response with the usual superposition. Another potential unknown is the initial pressure. The optimization process adjusts the derivative response in order to match the data and minimize the curvature of the derivative response, i.e. it will find the simplest derivative response that will best match the data.
Deconvolution on two consistent build-ups
Why does it work? No real magic here. In the case above the process obtains the additional late time behavior by having the simulated response honor both build-ups simultaneously.
The process also allows a certain flexibility on the (inaccurate) rate values. The optimization will also ensure that these rate changes are as small as possible. The duration of the deconvolved response is the whole time interval between the start of the production history and the last pressure point.
The additional information comes from the depletion between build-up #1 and build-up #2. If there was only one buildup, this depletion could be calculated to honor the initial pressure (if known). Before deconvolution, a good engineer would match one of the buildups with a boundary, see that the simulation is incoherent with the second build-up and add the extra boundaries until the full picture matches. Deconvolution just takes you there in only one step.
What is the catch? Be careful and remember the assumptions made. The main assumption is that superposition works and is applied to the ‘spline’ model. If the model changes in time, and in particular the skin, the process will not work. If the different build-ups are inconsistent deconvolution will not work. For multiple wells deconvolution will not work either. There is more: you create a theoretical response by matching data. Then you match a drawdown model to this response. Any mistake in the deconvolution calculation will remain for the rest of the interpretation process. So it is vital to always check the simulation match on the REAL data, at least on the history plot, and return to the original build-ups at the end of the process as a last validation.
The variables of the deconvolution process
Saphir deconvolution for the end user The engineer selects the periods, which would typically be a set of coherent build-ups, where the deconvolution will be calculated. At this point the initial pressure is input if known and rate adjustment selected or not.
Do not try to use the deconvolution any cost. It is a nice tool, but sometimes it just does not work. Caveat emptor.
15
Numerical models in Saphir
On numerical well testing Numerical well testing is not a new idea and has been tried since the advent of the simulator. In the early 1990’s the first automatic gridding designed for well testing was developed. KAPPA became involved in this at that time working under three specifications: first, the model must take minutes to build and run so that using a numerical model should not significantly change the duration of an interpretation; second, the reliability of the numerical model, and especially its automatic gridding, should be comparable to the reliability of an analytical model; and finally nonlinear regression should work.
One of the main achievements of earlier versions of Saphir was to provide the interpretation engineer with a variety of tools to get closer to the real reservoir. This was done by creating reservoir visualization, adding multiwell capabilities to some analytical models, and developing a numerical module based on an unstructured (Voronoi) grid with a modeling flexibility far beyond that of an analytical model. In Generation 4 the third dimension was added as a local grid refinement for modeling limited entry wells, horizontal wells and multi-layer formations. Horizontal and vertical anisotropies were also added.
The 2-D Map option This is the starting point for any detailed description of the reservoir, accessible from a dedicated tab in the Saphir main window. It may be used to:
These capabilities, integrated in the standard version of Saphir, correspond to the same assumptions of linear diffusion as used for analytical models. They constitute a sort of ‘super-analytical model’ extended to complex geometries, as described in these two pages.
• Simply visualize the reservoir for report purposes • Position interfering wells to be used in analytical models • Define the numerical model: reservoir shape, wells
In Saphir NL the numerical model was extended to handle nonlinearities including real gas, dead oil and two-phase flow restricted to water and one hydrocarbon phase. Saphir NL can also handle compaction and non-Darcy flow. Saphir NL is detailed in page 18 of this document. In the two present pages we will focus on the geometry.
•
1
2
Vector image built
Input bitmap
4
Thickness fields
3
Composite zones
5
Automatic gridding
7
Numerical model dialog
geometry and position, faults, composite zones in the reservoir and/or around the wells, special distribution of thickness and porosity, etc Visualize the automatic gridding and allow a small core of specialist users a high level of user control
User controlled gridding
8
2-D field display of simulated pressures
16
6
9
3-D field display of simulated pressures
Defining the numerical model After loading a bitmap (BMP) representing the reservoir (Fig. 1), the engineer first sets the scale by defining a known distance between two points. Once these dimensions are known, the tested well is positioned and the reservoir outer limits described as a closed polygon. Any segment of this polygon may be set as a sealing or constant pressure boundary.
Running the numerical model A tab in the model menu permits switching between analytical and numerical. Well parameters and permeabilities are retained, but the analytical geometry is replaced by the 2-D reservoir description (Fig. 7). 2-D and 3-D visualization / animation The simulated pressure fields may be stored for user controlled time steps, allowing a 2-D display during the simulation and later an animation (Fig. 8). The same real time display and replay is available in pseudo 3-D, the horizontal plane being the reservoir geometry and the pressure being displayed in the Z plane. The usual OpenGL display capabilities (angle, lighting, scale, gain, etc) are available (Fig. 9). When the model uses true 3-D gridding (complex wells, multilayer) a full 3-D view can be displayed representing the real 3-D geometry, where it is possible to define arbitrary cross sections, to show only certain pre-defined groups of cells (see page 11). Coloring and animation are also available in this mode.
If inner boundaries are present, any number of polyline faults may be drawn with control of individual fault transmissibility. Individual wells (vertical, horizontal and/or fractured) may also be created and positioned, and their production history entered. Later, when the model is defined, vertical and fractured wells may be individually set as fully penetrating or limited entry. Once the geometry of the problem is defined, the display of the original bitmap is turned off and the 2-D Map displays a vector description of the problem (Fig. 2).
Composite zones and thickness/porosity fields Fault polylines may also be used to delimit composite zones where separate mobilities and diffusivities can be defined. Additional composite zones may also be added around each well (Fig. 3). Porosity, thickness or permeability fields may be defined, either interactively or by importing an ASCII file. Kriging and other interpolation / extrapolation algorithms are used to define these properties at each cell (Fig. 4).
Complex well geometries and anisotropy The numerical module allows for fractured (Fig. 10), partially penetrating (Fig. 11), limited entry fracture (Fig. 12) and horizontal wells (Fig. 13). The 2-D grids around the well is replaced by a 3-D module. Saphir also permits horizontal (Fig. 14) and vertical anisotropy (Fig. 15). Multilayer The numerical model can simulate multilayer reservoirs. Any number of layers can be considered, and any number of wells penetrating those layers. The well model is restricted to vertical or hydraulic fracture. Each well can be selectively opened or closed in individual layers. Crossflow may also be modeled. The contour, faults, etc are identical in all layers.
Checking and controlling the gridding The 2-D Map displays automatic gridding, adapted to honor the reservoir contour, inner faults and wells (Fig. 5). The default is recommended but specialists may modify the basic grid geometry, size, main directions, and the local grid refinement around each well (Fig. 6).
10
Fractured well : pressure field
Limited entry well : display of one sector
14
13
Horizontal well : local grid refinement
12
11
Horizontal anisotropy
17
Fractured well with limited entry
15
Limited entry well + vertical anisotropy
Saphir NL
Water injectors A particular application of two-phase flow is the case of a water injector in an oil reservoir. Fall-offs of a water injector can be reasonably modeled using a radial composite solution. But injections are different. The interface is moving and the problem changes in time. Depending on the relative permeability tables and the level of injection rate, different behaviors will occur in time. Semi-analytical developments have been made on the subject, but with Saphir NL there is no assumption except the inherent uncertainties of relative permeabilities.
Saphir NonLinear (NL) is the highest level Pressure Transient Analysis (Saphir) module in the Ecrin suite. It addresses the specific issues of nonlinearities in Pressure Transient Analysis. Saphir NL covers the same geometries and well models as the numerical module of Saphir, however the slightly compressible fluid assumption and pseudopressures are replaced by the exact diffusion equations. Superposition in time has gone and the numerical module is run as a standard simulator seamlessly integrated into Saphir. Pressure control can be activated in the model to supersede the usual rate control when the simulated pressure goes outside predefined limits. When nonlinearities are non negligible, Saphir NL allows the engineer to solve the problem of diffusion of real gas, with or without non-Darcy, real dead oil, with or without water, and such specific conditions as formation compaction.
2-phase relative permeabilities
Using a numerical model to handle nonlinearities in well testing is not a new concept. Other groups have made attempts, with some success, to do this using local grid refinement on standard simulators. But with Saphir NL, a fast, easily handled and workable tool is at last operational.
Real gas With analytical models gas is handled using pseudotimes and pseudopressures. This works to a point, but increasingly the limitations can be seen. With Saphir NL the exact real gas diffusion equation is implemented. Because the problem is often locally linear and remains single-phase, the nonlinear solver will converge quickly and the generation time is in the order of a linear problem.
Water injector: Evolution of the water front
Non-Darcy flow For real gas, the Forchheimer equation is used to handle turbulent flow, both at well level and within the reservoir, including the case of hydraulic fractures. A specific turbulence coefficient is entered for each fracture.
Real dead oil For single-phase oil the slightly compressible fluid assumption implies that viscosity and compressibility are constant. The dead oil PVT option models it exactly.
Pressure control vs rate control When running a Saphir NL case, each well can be given a minimum and a maximum flowing pressure. Should the simulated pressure go outside the imposed limits, the engine will switch to pressure control and the rate will vary in consequence. On the history plot, the simulated rates will then be displayed.
Water+hydrocarbons The model runs with two phases, one phase being water. Both detailed PVT and the Water-Oil or Water-Gas relative permeability tables must be entered.
Formation compaction This feature allows the definition of pressure dependent permeability and porosity. This dependency may be reversible or irreversible. It is reversible when the property is a function of the current pressure values, whether increasing or decreasing. In the irreversible case the lowest historical pressure of each cell is stored and the property is related to this lowest value. Integration in the Saphir workflow A nonlinear analysis is just a particular type of Saphir analysis. Each nonlinear analysis stores, in addition to the model, its own PVT and relative permeability tables. Water injector and oil producer
18
Technical References
Built-in analytical models In the analytical model dialog, the user can select a large combination of well, reservoir and boundary models, in complement of wellbore and skin options.
Built-in numerical models User defined reservoir contour in the X-Y plane, unlimited number of segments Any contour segment sealed or at constant pressure User defined faults inside the contour with individual leakage factor True double-porosity model (duplication of grids) Composite regions with associated diffusivity, storativity and double-porosity model Horizontal anisotropy Varying thickness and porosity fields Conductive faults Multiple wells Fractured well with finite / infinite conductivity Limited entry vertical well with vertical anisotropy Fractured well with limited entry and vertical anisotropy Horizontal well with vertical anisotropy Changing storage (Saphir only) Time-dependent and rate-dependent skin Saphir and Topaze: slightly compressible liquid Non-Darcy flow for gas (Saphir NL only) 2-phase W-O and W-G (Saphir NL only) Real gas diffusion (Topaze and Saphir NL only)
Wellbore models
Well models
Skin models
Reservoir models
Boundary models
PVT correlations Z Gas Viscosity
Pb & RS Oil
Bo Co
Viscosity Rsw Bw Water Cw Viscosity
Dranchuk, Standing, Beggs & Brill, Hall-Yarborough Lee et al., Carr et al., Lee compositional Lasater, Vasquez & Beggs, Standing, Glaso non volatile, Glaso volatile, Lasater-Standing, Petrosky & Farshad Standing, Vasquez & Beggs, Glaso, Petrosky & Farshad Petrosky & Farshad, Vasquez & Beggs Beggs & Robinson, Beal Katz, Meehan & Ramey Gould, McCain, Meehan & Ramey Dodson & Standing, Osif Van-Wingen & Frick, Meehan & Ramey, Helmholtz
No storage Constant storage (Fair, Hegeman) Changing storage Finite radius Fracture - uniform flux Fracture - infinite conductivity Fracture - finite conductivity Horizontal Limited entry Constant Rate dependant Time dependant Homogeneous 2-porosity P.S.S. 2-porosity transient sphere 2-porosity transient slab 2-layer with X-flow Radial composite Linear composite Infinite Single sealing fault 1 constant pressure fault Closed circle Constant pressure circle 2 parallel faults 2 intersecting faults any angle Composite rectangle Leaky fault
External analytical models The user may dynamically connect additional external models, either delivered by KAPPA as a complement to the built-in model catalog or developed internally by the customer. 2-layers with X-flow & radial composite 2-layers with X-flow & 2-porosity 2-porosity & radial composite 2-porosity with skin at matrix blocks 3-porosity (1 fissure and 2 matrices) 3-layers with X-flow 4-layers with X-flow 4-layers with X-flow in closed system Conductive fault Horizontal well with horizontal anisotropy Horizontal well with identical fractures Horizontal well with non identical fractures Multi-lateral well Well in a reservoir pinchout Slanted well, fully penetrating Slanted well in an infinite reservoir Slanted well in a closed reservoir Radial composite 3 zones Vertical interference (v4.02)
External model definition and input
19
Production Analysis
Main Topaze screen with production history plot
Since the introduction of Topaze in 2003 Production Analysis (PA) has moved forward significantly. The old methods of assuming constant pressure or empirical decline functions have been replaced by a process employing advanced methodology such as the Blasingame plot and using true diagnostics based on the analytical and numerical modeling capabilities developed in Pressure Transient Analysis. The merging of the modeling capability of Topaze with the abundance of data from permanent gauges installed at surface or downhole has meant that users are now able to obtain answers that were previously only available from transient tests but remained hidden in long term production data. This information has the advantage that it is available at no extra cost and with no deferred prod uction. As the long term production is modeled, the evolution in time of the well productivity may also be quantified. Finally, forecasting may be based on a real model as opposed to an empirical function. In the Ecrin framework, data, technical objects, even complete analysis models may be copied from Saphir into Topaze by simple drag-and-drop providing a quick start point for production analysis and forecasting.
20
Main additions in v4.0 (2005) • Integrated into Ecrin; Drag-and-drop of technical objects within any opened documents; in particular models can be copied directly from/to Saphir • Wavelets at load time from databases using standard protocols such as ODBC or OLEDB • Load DMP2 files • Numerical model: limited entry well (vertical and fractured), horizontal well, X-Y & H-V anisotropies • Numerical models: 3-D grid display • Analytical models: extended to all Saphir models • Analytical models: horizontal anisotropy • Analytical models: multiwell solution • Sensitivity analysis by matching model with multiple range of parameters • Enhanced Arps with multiple simultaneous scales • Arps water-oil extensions : fo vs Qo, log(fw) vs Qo, 1/fw vs Qo, 1/qo vs Qo/qo • Analytical correction for water production • Enhanced output of production forecast • Load multiphase production used for intake correction • Enhanced interface, plots, annotations & report • Edit options merged with Saphir; in multiwell, they can be used to edit the production of any well • Multiple analysis report
Numerical model with multiple wells
Key features • Full analysis spectrum from decline curve to complete rate and pressure history matching • Brings the advances of the last twenty years in Pressure Transient Analysis to Production Analysis • Processing very large data sets (permanent gauges): wavelet filtration at load time • Extensive analytical model catalog • Unique numerical module extending the modeling capabilities to situations with arbitrary outer boundary shapes, any fault trajectories, composite zones, etc • Multi-well simulations • Model compatibility with Saphir • Simulate pressures from rates and rates from pressures • Fast modeling option for large data sets • Real gas numerical solution for exact material balance • Calculate and display average reservoir pressure • Time-dependent skin option • Unlimited number of analyses on different gauges, using different models and/or model parameters • Fast and robust optimization routine • Optimization on rates, cumulative production, pressures or any weighted average • Forecast of rate or pressure beyond the current history • Built-in surface to downhole correction with single and multiphase flow correlations • 24-hour on-line and telephone technical support • Extensive training and consulting services
Arps multiple scale analysis
Main additions in v4.02 (2006) Ecrin v4.02 was focused on the operation of Diamant Master and its compatibility with all the Ecrin modules, including Topaze. However Topaze v4.02 has additional enhancements including the possibility, in the numerical model, to control multiple well flowing pressures and production contributions. This allows PDG data to be used directly in the model instead of relying on production data that may only be allocated or inaccurate.
Numerical model with mutiple wells
21
Loading & editing data Topaze can load an unlimited amount of production and pressure data with independent time sampling. Data may be input as points (time, value) or as steps (duration, value), and this option can be dynamically or permanently changed later. Topaze will accept data from any type of ASCII and database files. Specific client customized data base connections developed for Diamant / Diamant Master may also be used to directly load data in Topaze. Comprehensive editing features include unique wavelet filtration that allows data reduction without losing significant trends or breaks in the pressure or rate signature. Topaze can also retrieve data by drag-anddrop from another Topaze document, a Saphir document or Diamant / Diamant Master. Arps
Well intake correction and extraction When pressures are acquired at surface or at any point distant from the sandface, the well intake option allows either the loading or generation of a well intake response to simulate sandface pressure. When the engineer extracts the data for analysis, options include the choice of pressure and rate gauge, time range, time sampling, and whether or not to correct pressures to datum.
Blasingame The Blasingame plot displays instant and average productivity index with respect to material balance time (cumulative production divided by instantaneous rate). It also calculates the derivative, in a display similar to an inverted loglog plot tending to a negative unit slope when pseudo-steady state is reached.
Model match on Blasingame plot
Loglog plot The loglog plot can be used as a diagnostic tool with exceptionally clean data. When data is more scattered some trends may however be detected. The simulated model can also be compared to the data on this plot.
Fetkovich
Fetkovich type-curves These curves are available to process data in the absence of permanent pressure measurements assuming constant producing conditions. Normalized rates and cumulative production can be superimposed on the selected type-curve. Arps plot The default scale is log(q) versus time, but other scales are available: q vs. t, log(q) vs. log(t), q vs. Q and log(q) vs. Q. The automatic match by nonlinear regression best fits the end of the data and displays the best matching decline function, which may be changed interactively. When oil and water rates are available, it is also possible to estimate the ultimate recovery from other graphs: fo vs Qo, lof(fw) vs Qo, 1/fw vs Qo and 1/qo vs Qo/qo. The simultaneous display of several scales is possible within the same plot.
Loglog plot
22
When the well productivity changes… If the simulation deviates from the data and indicates a change in the well productivity index the user may assign individual skin values to different production periods. Nonlinear regression is then applied on all skins, resulting in a relation between mechanical skin and time.
Normalized rate cumulative plot This variation of the Agarwal-Gardner plot shows dimensionless rate versus dimensionless cumulative production. A straight line at boundary-dominated flow gives a direct estimate of reserves. For gas, an iterative solution is implemented as the plotted values are in turn a function of the reserves.
Normalized rate cumulative plot
Changing skin
Model and nonlinear regression With a wide range of well, reservoir and boundary models (see technical references), Topaze offers the unique capacity to simulate pressures from the production history, or simulate rates and cumulative production from the pressure history, or both simultaneously. Nonlinear regression then allows history matching, minimizing the error in terms of pressures, rates, cumulative production or any weighted average.
Production Forecast Without data, or after history matching, a production forecast for any model may be run based on the anticipated producing pressure. Sensitivity to improvements or decay of productivity index can be simulated. Reporting, Exporting and Topaze Reader As in Saphir, Topaze has an extensive range of reporting, exporting and printing capabilities. The free and unprotected Topaze Reader allows Topaze files to be read, printed and exported without the requirement for an active Topaze license.
Model for simultaneous oil-water production When working with oil, an analytical prediction of water production can be included. The simulated water rate may be compared to the measured, and the difference may be added in the nonlinear regression. When the well drainage area changes… The same modeling and regression capacity is extended to the Topaze unstructured numerical module in order to handle complex geometries. The numerical module also allows the simulation of multiple well production, where individual wells can be pressure or rate controlled. Topaze permits 2-D and 3-D visualization of the well drainage areas and their evolution with time.
Production forecast
Common features with Saphir Topaze shares built-in analytical and numerical linear capabilities with Saphir. The Saphir external models will run in Topaze. Under Ecrin a complete Saphir document may be dragged-and-dropped into Topaze (or vice versa) to complement the analyses on the same data. For a detailed list of models refer to the Technical References p.19. For a description of the numerical linear capabilities, see Saphir p.16-17. 3-D model
23
Production Logging
Emeraude main window - Deviated well - 3 phases
Production logging is now seen as a powerful quantitative method that takes its own place in the set of data acquisition tools for the reservoir engineer, along side transient and production analysis. No longer just the tool of last resort, PL is now used as a calibration point for the reservoir model and as an important tool in the development over time of the producing intervals in the wellbore. The interpretation process has shifted into the hands of the end-user engineer due, to a great extent, to the development of client, as opposed to tool focused software; Emeraude. Production logging surveillance has given the reservoir engineer a powerful tool in the drive for the more accurate and refined reservoir characterization. Emeraude is now used by all the major service companies and all the major producers and many independent operators and service providers. Emeraude is seen as the industry standard allowing a common platform for communication and interpretation between service companies and operators. From vertical injectors to horizontal or highly deviated multiphase producers, Emeraude provides a comprehensive and intuitive set of tools, to produce results from the log data from simple through to the most sophisticated tool strings. KAPPA remains committed to the ongoing development of the industry standard PL interpretation package by remaining in close contact with tool manufacturers. Emeraude is currently a standalone package. Integration in the Ecrin environment is now taking place, for release planned in 2007.
24
Main additions in v2.42 (2005) • FSI module • Spinner response based on momentum • Surface contrib. in contributions tab (Zone Rates) • Update log option in contributions tab (Zone Rates) • Apparent downflow enabled when Vapp is positive • Apparent downflow option can be customized with possible factor and shift • Average friction true calculation • Grid customization • Drag & drop from Data Store back to a Pass • Collapse browser option • View Up/ View Down Passes option • Lateral average stores standard deviation • Improved mnemonic management
Main window with browser
Key features • Unlimited number of logging runs within a document with one or several interpretations for each • Logical data structure viewed/edited in a Data Browser • Fast and extensive plotting options with automatic and user-defined track creation • Fast learning curve • Fast path for simple cases • Methodology based on nonlinear regression offers full flexibility in the type and number of inputs • Comprehensive list of flow models from 1 to 3 phase • Specific models for apparent downflow, and flow through a standing water column • Temperature model • Global log optimization with constraints • Selective Inflow Performance (SIP) • Multiple Probe Tools support (DEFT, GHOST, CAT): image views / cross sections • FSI visualization and interpretation module • Pulsed Neutron Log (PNL) module • CGM/TIF log output • Well sketch • Formation test data QA/QC • Free reader for reporting and exporting
FSI cross-section
FSI geometry definition
Image view properties
Main additions in v4.10 (2007) The integration of Emeraude in Ecrin will bring the possibility to exchange components between Emeraude and the other modules and between Emeraude documents. It will be possible to transfer PVT, wellbore geometry, full wellbore description including the presssure drop model, as will feeding in a multilayer PTA analysis with the PL rate results in Emeraude. Beyond the benefit of the integration, Emeraude v4.10 will include a number of new options such as a complex user formula module, layout templates, and a test design option.
25
Feature details
Spinner calibration and apparent velocity The user interactively defines the spinner calibration zones and the positive and negative lines on each zone are automatically calculated. Two calibration modes are available: constant value of thresholds, or constant ratio between thresholds and intercepts. Calibration results can be viewed and edited directly with a full editing tool box. Once the spinner calibration is satisfactory, the user generates an apparent velocity channel. At each depth the apparent velocity, Vapp, results from a user controlled weighted average of the calculations for each pass.
Data load and display Emeraude can load data from LIS, LAS, and ASCII files or from the clipboard or keyboard. Versus depth logs or stationary data can be input. The load option automatically recognizes the file format. New mnemonics can be defined and become part of the software settings. Mnemonics can be filtered for the current and future loads. Log tracks are automatically created for each mnemonic after the load, and scaled automatically to display all the available measurements.
Single and zoned PVT The PVT model defined by correlations provides the properties of any phase at any temperature and pressure. It is also possible to redefine the properties for each inflow zone. In each phase dialog, there are tabs for all the relevant properties where the correlations can be viewed both graphically and inside a table, and matched to user-entered data. The PVT needs to be created only once, and can then be copied from one interpretation to another. It is possible to save this to a file and to load that same file into the PVT of other documents. Load dialog
Methodology Rate calculation is treated as a minimization problem and solved using nonlinear regression, offering full flexibility in the type and number of input measurements. Interpretations can be run from any number of sufficient inputs including: spinner apparent velocity, density, pressure gradient, capacitance, holdup of any phase, velocity of any phase, rate of any phase, and temperature. The solution rates are found by minimizing the error between measured and simulated values. Each difference, or residual can be weighted separately. Two different nonlinear regression schemes are available; the local regression solves for the cumulative rates in the wellbore at a series of user-defined depths and global regression solves simultaneously for all the zone contributions and can be used as a second stage to impose sign constraints.
Data browser
Data structure and browser editing Internally, Emeraude uses a hierarchical representation that may be visualized in the data browser. At the top of this hierarchy is the general well data that will typically include: open-hole gamma ray, T.V.D. or deviation log, zoned I.D. or caliper log and perforations. Next is the survey data, which is the basis for one or several interpretations, providing for sensitivity studies. Each survey has an associated data store where copies of channels can be made. Editing facilities include: lateral averages, depth stretch, shift, data cut and fill, merging, splicing, derivative and sampling. Spinner calibration
Interpretation models Emeraude offers a full range of flow models from single to 3-phase model which considers slippage between water and oil (Liquid-Liquid model) and slippage between the liquid mixture and the gas (Liquid-Gas). Specific models are provided to handle flow re-circulation as well as flow through standing water columns.
Flow map
View/match PVT correlation
26
Zone rates calculation The Zone Rates option calculates, views and modifies rates for any calculation zone. Results are displayed as a table or graphically, in terms of measured values against simulated values. This can be used to assess the relevance of the models when compared to the measured surface rates and to alter the model to match those rates. The user has full control of the solution model and the results. Additional results are also presented such as the flow regime, average velocities, holdups, slippage velocities, as well as cumulative rates, downhole and at Standard Conditions. Input parameters for each zone can be listed and edited. The value of the supplied reference channels and all relevant PVT properties can be viewed in tabular form. Contributions can be directly edited. When a gradio tool is used, all components of the simulated pressure gradient are output: tool friction, friction along the pipe, and acceleration gradient. A flow map is used to spot discontinuities in the simulated responses due to the change of flow regime.
Schematic and complete logs Zone Rates results are plotted as schematic logs. In addition ‘zoned simulated’ measurements are generated and plotted together with the input channel. A Complete Log output can also be obtained which, at every selected depth increment, calculates the cumulative rates by nonlinear regression. Because the complete log is everywhere faithful to the data, it can serve as a good indication of the entry points.
Multiple Probe Tools with visualization Emeraude can process multiple probe tools giving either bubble counts, holdup of a given phase or capacitance measurements. The built-in options are tailored for the Schlumberger DEFT, GHOST, and the Sondex CAT but can be customized. Image tracks can be created, and cross sections displayed at any depth. Average holdups can be calculated using an arithmetic average or a stratified average.
Zone rates option
Global Regression The default calculation scheme in Emeraude involves successively solving the cumulative rates at selected depths inside the wellbore. The contributions of the inflow zones, located in between the calculation zones, are then obtained from successive differences. Because each local regression is done regardless of the solution above or below, the overall solution may result in contributions from the same interval showing different signs which, physically, is not possible.
Multiple probe tools - Image view - Cross-section
Flow Scan Imager A specific treatment is offered for the Flow Scan Imager (FSI). The holdup information provided by the six rows of resistivity and optical probes can be viewed as an image to assess the flow distribution. Cross sections can be displayed that provide a clear view of all the tool measurements at a given depth. The cross section is updated continuously as the cursor moves along the log. On the cross section, the user can choose an interpolation method for holdups and spinners along the vertical axis (see figures on p.25). Using the user selection, the software calculates average holdups, total velocity, and possibly phase velocities and rates. Note that in this process each spinner has its own calibration values for the different fluids. Any average channel calculated from the FSI processing may then be input to the usual calculation engine.
Global Regression provides a method for solving those cases. For every producing interval, the sign of the contribution can be imposed. It is also possible to fix any particular contribution to a user-entered value, in particular a null value. Global regression can be coupled with a Genetic Algorithm to avoid local traps.
Well sketch The well completion is built by drag and drop of predefined elements, e.g. tubing, casing, packer, etc., and the depth and diameter information is entered. For the latter, Emeraude can retrieve the values from an existing ID channel. Global regression
27
For gas rates, the SIP can be based on pseudopressure instead of pressure. An unlimited number of SIP’s can be created and compared. Each zone can be assigned a different model: straight line, c&n, or Jones IPR. The SIP can use the total rate, the rate of a given phase, or the total liquid rate. The analysis can be done downhole or with surface values. Pressure datum correction can be applied and a composite IPR displayed.
Pulsed Neutron Log (PNL) interpretation Clean formation, Shaly - Single water model, Shaly - Dual water models are available. When a model has been selected, Emeraude indicates which channel inputs are needed. Those channels can be loaded from an openhole interpretation, or estimated from correlations. In the start-up phase they are given a unique constant value. They can later be redefined on user-selected zones. Zones can be defined anywhere along the log. Crossplots are created on these zones and can be used to redefine locally the input measurements. For unclean formations the plot can dynamically correct for shale. The relevant capture cross-sections are changed on a crossplot either by direct input or by dragging the plot lines. When the change is satisfactory, it is possible to modify the logs with the new value, on the selected zone only, or on the whole range. When all inputs have been defined, a water saturation log can be generated which is automatically used to display a Bulk Volume Analysis, and a Pore Volume Analysis view. Time-lapse presentation can be obtained automatically by supplying the interpretation with additional water saturations, and defining the relevant chronology.
Well sketch creation
Temperature The temperature can be used to replace a faulty spinner or a spinner ‘blinded’ by such effects as flow recirculation. The temperature model requires the definition of a formation temperature profile that may be directly input or interactively picked on the data. Fluid heat capacities and an overall heat loss coefficient are also required. The latter can be estimated on the top calculation zone when surface rates are available. It may also be included as a variable of the global regression.
Temperature Interpretation
Selective Inflow Performance (SIP) Generating a SIP analysis can be done very quickly once the reservoir zones are defined. In every selected interpretation, the SIP plot will read the rate values on the schematic logs, the pressure and the temperature values.
PNL Interpretation
Well views In many situations understanding the measurements and results may require visual representation. In a horizontal well, oscillations in the well trajectory will be responsible for significant variations in the holdup due to the influence of deviation on the slippage velocities. Displaying the holdups inside a representation of the wellbore is therefore essential. Well views can be created from the browser to display the well geometry, based on TVD or deviation, the internal diameter and zones. Holdup channels, obtained as direct tool measurements or complete log calculations, can also be added to the well view and displayed within the wellbore. SIP analysis
28
Technical References
Formation test data QAQC An option of the special panel allows the loading of reservoir pressure and reservoir permeability / mobility channels. They can be loaded together or separately; in which case the software will reconcile the depths. Each depth point can be assigned a legend, and a quality indicator. All information appears on the tracks built automatically to display the pressure and permeability. Lines can be calculated or drawn and gradients / contacts are deduced.
Pressure drop correlations Artep Aziz & Govier Beggs & Brill Choquette Dukler-Eaton Duns & Ros Hagedorn-Brown Orkiszewski Petalas & Aziz Taitel & Dukler
Export Channels present in an Emeraude document can be exported in LIS, LAS, or ASCII formats. Those files can then be transferred to third party applications, corporate databases, etc.
Liquid-Gas; mechanistic; any angle Liquid-Gas; mechanistic; vertical Liquid-Gas; empirical; any angle Liquid-Liquid; empirical; vertical Liquid-Gas; empirical; horizontal Liquid-Gas; empirical; vertical Liquid-Gas; empirical; vertical Liquid-Gas; empirical; vertical Liquid-Gas; mechanistic; any angle Liquid-Gas; mechanistic; vertical
A flowmap for Petalas & Aziz
Output The log output in Emeraude is WYSIWYG and can be sent to any MS-Windows™ printer. Single or multiple log tracks, as well as any X-Y plot can be copied to the clipboard in Bitmap or WMF format. The Log printout includes a preview option where full control is given to modify the fonts, scales and grid lines. Screen captures can be made at any point and returned to later with a single click. API logs can be produced from the print preview option and stored within the document. A built-in report can be printed and previewed that includes predefined sections. This report can be complemented by the appropriate log outputs generated as required. If further customization is required, it is possible to produce a report in MS-Word™ using the OLE™ interface of Emeraude. Through this interface, any parameter or result of the active document can be retrieved using the appropriate macros. A template MS-Word™ report is installed and can be customized as required Internet: Click & Send The Click & Send option enables the user to send directly an E-Mail with the current Emeraude file, or a subset, to any destinee, including KAPPA support. It is possible to send a compressed version of the file, with all surveys or only the active one.
Gaz PVT - Z correlation match
PVT correlations Z Gas Viscosity
Pb & RS Oil
Bo Co
Viscosity Rsw Bw Water Cw Viscosity
Log preview
29
Dranchuk, Standing, Beggs & Brill, Hall-Yarborough Lee et al., Carr et al., Lee compositional Lasater, Vasquez & Beggs, Standing, Glaso non volatile, Glaso volatile, Lasater-Standing, Petrosky & Farshad Standing, Vasquez & Beggs, Glaso, Petrosky & Farshad Petrosky & Farshad, Vasquez & Beggs Beggs & Robinson, Beal Katz, Meehan & Ramey Gould, McCain, Meehan & Ramey Dodson & Standing, Osif Van-Wingen & Frick, Meehan & Ramey, Helmholtz
What is coming next?
Amethyste
Rubis
Ecrin v4.10 (2007) and v4.20 (2008) Diamant, Saphir and Topaze have a long list of additional features defined in cooperation with our user group. In addition, Ecrin v4.10 will see the integration of Emeraude as its PL module, and Topaze NL will be released. Finally two new modules that complement the Ecrin suite will be released: a reservoir simulator ( Rubis ) and a nodal analysis package ( Amethyste ). Reservoir Simulation Rubis will be a fast, interactive 3-D / 3-phase simulator that will be positioned half way between material balance and the ‘large standard’ simulators in the market. The 3-D geometry will comprise a cartesian accumulation of layers, each layer using a 2 D, locally 3D, unstructured grid. This will be more coarse than, but compatible with, the PTA numerical models. Rubis is scheduled for Ecrin v4.10 (2007). Nodal Analysis In 2005, in collaboration with a major operator, KAPPA ported WAM, a client in-house nodal software package from DOS to Windows™. This version is available now but with a limited deployment. It will be the technical base of Amethyste, the nodal analysis module of Ecrin. Emeraude and Amethyste will share the same flow models, and it will be possible to use the results of the production log to initiate the nodal model at the level of the formation. Lift curves generated by Amethyste will be usable with a drag-and-drop in the simulator, PTA or PA analyses.
30
