PIE Release Notes
(Refresh this page to ensure display of the latest version)
Release date: 29-Nov-2021
1) Enable the mouse-wheel in PIE to scroll a table of data or move between input fields. When displaying a table of data, holding down the "shift" key while using the mouse-wheel will extend a selection of table lines.
2) With 4K high DPI displays now available, the size of the PIE "desktop" window is now adjusted to account for large "zoom factors" set in the Windows settings for this type of display. This adjustment will also accommodate a non-standard display aspect-ratio. This update ensures the default size of a PIE plot window does not overlap other windows displaying analysis results.
3) For a multi-well simulation, the plot legend will now show the name of the "observer" well instead of a generic "pressure" or "rate" label.
Release date: 4-Oct-2021
Small update for a change to the built-in network license manager. Not for general release - this change will be part of all subsequent updates.
Release date: 28-Sept-2021
1) PIE has been converted to a 64-bit application, and the 32-bit version has been retired. This change only affects the internal operation of the program. The user-interface and the program capabilities remain the same. The move to 64-bit is to enable future developments using new software development tools and methods.
2) The well productivity-index plots (Plots -> "Well PI vs. Time", or "Well PI vs. Cum.Prod") have been improved. Shut-in periods are now skipped. The y-axis plot scales are adjusted to ignore "spikes" in the PI. Note that the spikes are still plotted, but the default y-axis scales are set to just display what is most important. The plots scales can still be changed (via "zoom" or "Graphics -> Set Scales") to view more or less of the plotted data.
3) Selecting a type-curve model is now done from a list of all available models, and the list can be "filtered" to quickly locate a model with specific well/reservoir attributes. This change to the user-interface ensures a particular model can be located within the extensive type-curve model catalogue. Models can also be selected by entering a quick-select model ID (e.g. "001001" is the homogeneous model). This change allows users familiar with the program to remember a model ID rather than a combination of selection options. Note that the model ID is also shown in the list of saved results under the "Results Manager".
4) A new "GUIDES" button has been added to the graphics tool-bar when displaying a derivative-plot while doing a type-curve simulation. This button allows a characteristic line (half-slope, 1/4 slope, stabilisation, or spherical) to be added to the derivative-plot along with the simulated response. This is useful when trying to understand how a type-curve model can yield a characteristic line (e.g. multi-layer horizontal-wells exhibiting a 1/4 slope), or for plots put into a report where a particular flow-regime needs to be highlighted.
5) The "Analysis" options to do diagnostics, select a type-curve model, or re-load previously saved results can now be selected while still in the ASCIIprep data-processing mode. This eliminates the need to first display an analysis-plot prior to selecting those options.
Release date: 5-Dec-2020
1) The Windows "manifest" for PIE now shows the program as being "DPI Aware" i.e. PIE knows how to handle high resolution displays where a "scale factor" is being applied. This Windows "scale factor" is set under "Display Settings" and the option to make text bigger by a percentage amount. On a 4K "UHD" screen (typically 3840 x 2160 pixels), this scale factor can be quite large (250%) which will make the old version of PIE appear "fuzzy" because Windows "zooms" tells PIE it is running on a lower resolution display, then "zooms" the output by the scale factor. With this update, PIE tells Windows to skip zooming and let PIE handle the display at its full resolution. The end result is a much sharper PIE display.
2) Even so, on a high resolution display the text font used by PIE is too "thin" and small making it difficult to read. So I have increased the font size and weight so it appears a bit larger and much darker than before. That gives a better contrast that is easier to see.
3) On multiple monitors, the "Help" function has been updated to ensure it creates a help display on the same monitor that PIE is running on. Also, the help window size is now scaled to the PIE main window to avoid Windows from resizing it too small on a high resolution display.
Release date: 1-Sept-2020
1) the "date" and other time graphics functions now work correctly in the gauge-comparison module. Previously, a gauge reference-time that differed from the global reference-time lead to reporting incorrect date/time values on a comparison plot.
2) the "date" and other time graphics functions are disabled in the Main -> Reports -> "Plot Manager". This fixes a potential bug where the plot being displayed is unrelated to the current global reference-time, which would cause these time functions to display meaningless (and misleading) values.
3) An important safeguard has been added to protect a PIE-file from being corrupted during a "garbage collection" cycle. In previous versions of PIE, when the main PIE window was closed (using the "X" box in the upper-right corner), PIE would continue to work in the background doing a "garbage collection" cycle to clean-up and compact the PIE-file. This background operation could potentially be interrupted in the middle of a critical rebuilding step which results in corruption of the PIE-file. With this PIE update, closing the main PIE window will cause the program to stop immediately and not attempt a PIE-file clean-up. In addition, when PIE is stopped normally via the "File" menu, there is now a new user-interface to clearly show the progress of the PIE-file clean-up and when the critical update step is being performed. A new option "Compact PIE-File" has also been added to the "File" menu to allow the user to force a PIE-file clean-up to potentially reduce the PIE-file size (which is useful after deleting plots or other data during an analysis).
4) The right-click context menu in ASCIIprep now includes the "Add Flow-Periods to the Test" option. This saves a number of clicks to access that frequently used option from the ASCIIprep pull-down menu.
5) For graphics functions, the middle mouse button is now equivalent to hitting the "Return" key on the keyboard. This will speed-up the use of options to add individual pressure points to the analysis-data, and other operations where the "Return" key is used to indicate that a sequence of mouse selections is to be accepted and applied.
Custom version - not for general release.
Released date: 20-Apr-2020
1) A minor update to fix a missing "HELP" button on one of menus in the Monte-Carlo option
2) An important update to fix a floating-point bug that can happen with newer INTEL CPU's. The "AVX2" extensions in these newer systems have changed the way "denormallised" floating-point numbers are handled, which can cause random floating-point exceptions that crash PIE when using the Optimiser. This update now includes a more detailed floating-point initialization that ensures the CPU configuration is set to handle denormalised numbers correctly.
Custom version - not for general release.
Release date: 20-Aug-2019
1) Fixed bugs that displayed right-click context menus or dialog-boxes in the wrong location when using multiple monitors (e.g. PIE is on one monitor, but a right-click menu appears on a different monitor).
2) Fixed a bug where the PIE "help" window would appear at the wrong location when using multiple monitors.
3) Fixed a bug where deleting a range gauge-data rate measurements under ASCIIprep would display the PIE "null" value (-9999) as rate values instead of being removed from the display.
4) The general purpose "TIME-1", "TIME-2", etc. time markers (Graphics -> "Time Markers" option) will now appear on analysis-plots and the "Data Plot". This fix allows an event seen on an analysis-plot to be correlated to a time the overall well rate and pressure history.
5) Selecting the option Analysis -> "Diagnostic" will ensure that the analysis starts over again, with no previous results accidentally carrying over into a new analysis.
6) The results obtained from diagnostic and model-parameter estimates on the derivative and superposition plots now include the "productivity-index at P*" and "delta-p skin" where those estimates are appropriate.
7) Fixed a bug in the calculation of "delta-p skin" so a gas-test analysis uses the pseudo-pressure transform to compute this value.
8) Fixed a bug in the type-curve simulation option for a gas-test that allowed the simulation to proceed with an input initial-pressure that was outside the the pressure range of the PVT table.
9) Added a "Monte-Carlo" uncertainty analysis to deconvolution as per paper SPE164870.
Release date: 16-Dec-2018
Most changes are to the internal operation of the program. The important UI changes are:
1) Because large negative skin-factors seem to be popular these days, the model validation logic now includes checks on the "equivalent wellbore radius" to ensure that does not exceed reservoir boundary distances or an interface between zones. This change should prevent crashes or odd looking model behaviour.
2) The deviated-well model has been improved. However, the integral average used by this model was found to "smear-out" the transition between linear-flow and the final radial-flow over the reservoir thickness. This effect is very subtle (which is why it was not noted until now), but is something to be aware of. A bit of research is under way to update this model in a future release.
3) The gas PVT menus no longer issue spurious prompts about updating analysis-constants or saving changes.
4) There are now separate analysis-constant input displays for oil, gas, and multi-phase (i.e. the option "Analysis-Constants" under the "Analysis" menu). The main change is for gas tests - there are a number of options that allow the viscosity, compressibility, and volume-factor values to be set by entering two pressure values so PIE can do a PVT table lookup to set the relevant constants. Also, the values displayed for a gas-test depends on the "test-type" i.e. when using a pseudo-pressure you no longer are prompted to enter a gas volume-factor or viscosity at the average pressure.
5) If the gas PVT tables are changed (via the "Edit PVT Data" option on the EDIT menu), then the analysis-constants will automatically be updated to correspond to the new PVT tables. In previous versions of PIE it was possible to have the analysis-constants used to compute results out-of-sync with the pseudo-pressure used to display the data.
6) The pressure values used to look-up gas PVT properties for the analysis-constants can be set to a special "USER_SET" value in order to bypass any automatic updates and allow analysis-constants to be set to specific values. This ensures you always have control of the analysis-constants in case there are special circumstances where these values must be set independent of the gas PVT tables.
7) When saving/restoring analysis results for a gas-test, the settings used in the analysis-constants are also saved. That also means the pressure values in the analysis-constants are honoured i.e. the pressure values are used to update the constants to the current PVT tables except when the pressure values are given the "USER_SET" value.
Small bug-fix update
Release date: 14-Feb-2016
1) A new Installer which simplifies the
number of steps. Note that the PIE entry in the "All Programs" menu
under the "Start" button is now a single entry labelled "PIE
Well-Test Analysis" instead of a sub-folder.
2) A new compiler and development environment is now being used to build PIE. At some point in the future, a 64-bit version of PIE will be available.
3) Selecting the "Cancel" button in the analysis-data manager will un-do all deletions of the current analysis-data and analysis-data saved in the PIE-file. The deletions will only take place when the user selects the "OK" button or selects an option from the PIE menus.
4) Starting a test design will now display the "Design Manager" which allows the user to set the analysis-data ID and descriptive comment that will be used for the design. The design will be saved in the current PIE-file (which can be changed by using the option File -> "New PIE-File..."). The "Design Manager" also allows previously saved designs to be deleted.
5) A user-defined table in the gas PVT set-up can now use a linear-interpolation to prepare the look-up table instead of a cubic-spline. This is useful for PVT properties with a discontinuous slope or other non-smooth feature.
6) Fixed a bug where the multi-layer multi-perforation models failed to generate a solution when the number of layers is equal to one.
7) fixed a bug in the solution for the "inclined well", "multi-layer with cross-flow", "radial" model (and associated MLDEVW models) where a correction for FWBR did not converge correctly for some cases. This also improves the solution and speeds the calculations. Added checks on the input parameters to trap cases where the setup causes the wellbore diameter to intersect a layer interface.
8) The old Fortran input/output has been replaced with system-level file operations to speed-up manipulation of large PIE-files and large sets of gauge-data. This change improves the performance of PIE when opening a PIE-file that is located on a network file-server. This change also fixes bugs where PIE would appear to "freeze" when using the option File -> "Save Test As..." option, saving a large set of gauge-data, or when PIE cleans a PIE-file on closing that file.
9) fixed a bug in MSTATE that could reset boundary fog-factors to a NULL value.
10) The license setup can be verified from the HELP menu via "About PIE & License" option, and PIE will prompt you for a correct serial-number if the current value is invalid. If the wrong serial-number is entered accidentally during installation, it can be fixed using this setup option. The setup also allows a short-cut to be created, register the ".pie" file-name extension, and see the current license serial-number.
Release date: 8-Aug-2015
1) the number of layers has been increased
to eleven (11) in the general multi-layer no-cross-flow model and the
"lambda" multi-layer model with cross-flow. This change to the
multi-layer models applies to the "Layer-Rate Simulation" option and
to the use of these models in a normal analysis. Note that the non-linear
regression "Optimiser" has been updated to handle the larger number
of parameters associated with an 11 layer system, but users should be careful
when performing a regression with a large number of "unknowns".
2) the "Layer-Rate Simulation" option has been improved to display the "pass data" on the layer-rate plot. This allows the simulation to be matched to measured PLT pass-data. Also note that this update includes a major change to the way the multi-layer model set-up is stored in the PIE-file such that the new data has both forward and backward compatibility.
3) the tide analysis option has been fixed to display the correct results when using the "gravity-potential" function (i.e. an "earth tide" measured in an on-shore well), and the on-line help for tide processing has been improved. If no extra parameters (storage coefficient, permeability, Biot's constant, etc.) are input, then the pore compressibility reported is based just on the attenuation factor. If *all* of the extra parameters are entered, then the results take into account the rock properties and the small flows into and out of the well-bore.
4) the ASCIIprep option to select pressure data for analysis now allows a "fixed sampling frequency" option that picks pressure values on a "points per hour" basis. This makes it easier to get a reasonable distribution of points over a draw-down period with many rate-changes.
5) a bug in the linear-composite model has been fixed. The solution would revert to a homogeneous model unexpectedly when the mobility and storivity ratios were the reciprocal of each other. It is unlikely that anyone has encountered this bug (a mobility equal to the reciprocal of the storivity doesn't make much physical sense). However, some effort was put into fixing this bug because it causes an abrupt "break" in behaviour as reciprocal condition is approached.
6) in the process of fixing the linear-composite model, the solution tolerances for that model were improved. This slows the calculation of this model, but a type-curve simulation will now show a smooth response for all cases.
7) Along with the increase in the number of layers for the "lambda" cross-flow multi-layer model, the option Main -> "Multi-Layer Lambdas" has also been updated to handle the increased number of layers. This option is a utility that allows lambda values to be converted to "Kv'Kh" values under various assumptions in order to help visualise the description implied by a multi-layer model. This utility also can be used to compute lambda values given a set of "Kv/Kh" values for each layer.
Release date: 22-Nov-2014 (the 2014.10b version)
1) A new "pressure verses distance" option is located under the "MAIN" menu which will display an animation of the pressure profile along a defined cross-section through the reservoir. Before selecting the "pressure vs. distance" option, a set of analysis-data needs to be prepared that includes a valid rate-history. The animation set-up will go through all the steps needed to create an "animation table" which contains all the "slices" of "pressure vs. distance" for all the time values in the type-curve simulations that span the rate-history. Once this animation table has been prepared, graphics functions "PLAY" and "PAUSE" can be used to start/stop the animation. While the animation is playing, the repeated clicking on the graphics functions "FASTER", and "SLOWER" control the animation speed. Alternatively, the "Animation Controls" function allows the animation speed to be set to a numeric value. The HELP documentation can be used to get more details for each plot and graphics function.
Note that the "pressure vs. distance" option works with analysis-data that includes offset wells. This multi-well feature will display pressure profiles between wells.
Note that the complete set-up for the "pressure vs. distance" option is saved with the analysis-data in the PIE-file.
2) The ASCIIprep data-processing "wizard" now includes a set of "pre-filters" that can be used to reduce huge sets of raw gauge-data to a more manageable size. If a rate-history has been set-up, different filters can be applied to flow and shut-in periods. The filters are selected by checking the "Pre-Filter the data" option shown on the first screen of the ASCIIprep wizard.
3) Any of the interference-test type-curve models (i.e. models that allow the pressure response to be calculated at an arbitrary point in the reservoir) can now be used with analysis-data that includes offset wells. Also, all interference-test models will now revert to the well-bore solution when the origin of the co-ordinate system is specified. Because the interference-test models use the storivity (the product of porosity, total-compressibility, and thickness) as an input parameter, this new feature allows a multi-well system to regress the storivity to find a best-match to the data.
4) A new "Press. Resid as Skin/Turb" plot is now available, which is used to display the difference between a type-curve simulation and the measured pressure data as either a "delta-skin" or "delta-turbulence". These two "views" of the simulation residual allow the user to decide if the differences relate to a constant "skin" effect, or a rate-dependent effect. For details about this new analysis plot, please read the on-line documentation that is displayed when the "Help -> Help(General)" option is selected from this plot.
5) A new option "Manage PFA and Overlays" under the "ASCIIprep" and "Analysis" menus has been added. This option displays the flow-periods in a table to simplify selection of a particular flow-period for the period-for-analysis (PFA) when working with a long production history containing many flow-periods. In addition, flow-periods can be marked as "important" to the analysis, and the table filtered to just show those important flow-periods. Finally, flow-periods can be included or excluded from the "auto-overlay" display on the derivative and superposition plots. If the derivative or superposition plot are currently displayed, then each include/exclude change in the table results in an update of the analysis plots
6) A "Layer-Rate Simulation" module (under the MAIN menu) which is used to display the rates from individual layers in a multi-layer no-crossflow system. The type-curve model used for this simulation is the "general multi-layer" model which allows each layer to be a complex completion or reservoir type. This module uses a "wizard" to set-up the simulation, together with other information about the reservoir layers. The simulation results are displayed on a "Layer-Rate Plot" which shows the rates from each layer together with a comparison of the sum of those layer rates against the total surface rates (i.e. the rate-history).
7) A "Synthetic PLT" option in the "Layer-Rate Simulation" module that is used to set the time of a PLT "survey" in the well-history. Once the survey time is defined, a "PLT Survey" plot of rate vs. depth can be displayed. In addition, measured PLT data can be entered as one or more sets of "PLT pass" data, with each pass assigned a time in the well-history. When the survey time corresponds to a PLT pass, then the PLT Survey plot will display both the simulated and measured rate vs. depth data. Use the "PLT Survey" plot and the PLT pass data to match the transient rate response from a layered system to a set of measured PLT data.
8) As of version "2014.10a", the PIE installer has been modified such that all user-settings (line colours, plot options, last PIE-file used, etc.) are retained when the program is re-installed. When you first install PIE version 2014.10, these settings will NOT be saved, but subsequent installations of version 2014.10 or higher will retain the settings.
9) A bug has been fixed where a type-curve simulation of an "impulse-test" rate-history (very short flow followed by a long shut-in) would generate a noisy response towards the end of the shut-in period. An impulse-test is a difficult numerical problem to solve so, even with this bug-fix, it is still possible to have some noise on the derivative-plot of the simulation response. However the bug-fix does ensure a smooth overall trend of the simulation response.
Release date: 1-Nov-2013
1) Up to three separate cases can now be specified for type-curve simulation. This change also required changes to the way parameters entered in the "Analysis Results" window with mouse-clicks, and additional options to display either the three simulation cases or the parameter estimates.
2) All of the basic data and results contained in a PIE-file can now be exported to an EXCEL spreadsheet in the "XML" format. This simplifies the transfer of data and results from PIE to another application. See the "Create XML Spreadsheet" option under the "FILE" pull-down menu.
3) The display of gauge-data has been improved when switching between PIE-files to avoid leftover plotted points from the previously displayed data.
4) The screen drawing speed can now be adjusted via the File -> "Output Preferences" option. There is a new field called "screen detail" where a value equal to "1.0" will be full detail, and higher values reduce the level of detail drawn on the screen (less detail equals faster drawing speed). This change addresses extremely cumbersome analysis-data preparation when working with very large sets of ASCIIprep gauge-data i.e. editing options can become unresponsive due to the slow window redrawing that occurs whenever PIE needs to refresh the display. This problem was especially true for gauge-data approaching the program limit of 40 million rate/pressure points.
5) The way the markers are drawn on the plots has been improved by using the "polyline" functions in Windows. This change yields marker drawing speeds that are a factor of 2 or 3 times faster. This is an important graphics performance improvement when dealing with ASCIIprep gauge-data plots displaying millions of points.
6) PIE automatically maintains an up-to-date plot legend by scanning all the plot data whenever key drawing operations are done. However, with a very large set of gauge-data, this scan can be very slow. As a consequence, editing the analysis-data when working with a large set of gauge-data can become almost impossible (e.g. changing a rate value will trigger the slow legend scan). A check-box item has been added to the "Legend" menu (Format -> "Plot Legend...") which allows the automatic legend scan to be turned off when working with large amounts of plotted data.
7) Some plots would display a legend with duplicate entries. The File -> "Output Preferences" option includes a check-box to have PIE remove these duplicate legend entries.
8) Some models allowed a zero well-bore storage coefficient to be entered, but this caused a floating-point exception in the Optimiser non-linear regression. All models now require a non-zero well-bore storage coefficient.
Release date: 5-Dec-2011
1) A new Deconvolution Algorithm has been added to the program based on the work by Levitan et al. This algorithm has a number of important features that make it particularly useful for analysing a long well history where there has been significant pressure depletion in the reservoir. For gas reservoirs, the algorithm includes all the techniques described in SPE 134261 "Deconvolution of Pressure and Rate Data from Gas Reservoirs with Significant Pressure Depletion". These techniques ensure the response-function obtained from the algorithm will display the correct late-time behaviour by removing artefacts associated with changing fluid properties. This is a robust deconvolution algorithm that allows the user to obtain a reliable estimate the reserves being accessed by a particular well, and obtain a clear understanding of the reservoir characteristics.
The key features in the deconvolution algorithm are controlled by a set-up "wizard" that is located under the "Deconvolution..." option on the "Analysis" menu. In particular, note the options to define a pore-volume other additional options to control the deconvolution calculations. The "Help" button on this screen will display an extensive description of how these options work.
2) The PIE Program Installer has been updated to conform with Windows-7 security requirements. The application files are placed in the current users "local application data" folder, and all settings are stored in locations where the user has full access rights. These changes eliminate the "User Account Control" prompts for administrator credentials when installing the program.
3) The PIE application has been validated for use on 64-bit Windows systems. The program itself remains a 32-bit application, but some small changes have been made to ensure normal operation on a 64-bit system.
Release date: 1-Sept-2010
1) Improved Offset Well Options have been implemented to improve the display of data for a group of wells. The EDIT menu option to "Edit Offset Wells" now includes a time/rate/pressure table to input data for each offset well. The data for each offset well can now be individually displayed on derivative, superposition, and data plots.
2) Better Offset-Well Control options have also been implemented to individually select the offset-wells that are part of the simulation and deconvolution. The "Offset Wells" option on the Analysis menu now displays the wells that can be turned on (a check mark) or off (no check mark). In addition, a "Location and Correction" table is available to set the co-ordinate location for each well and a pressure correction to be applied to the display of the well pressure-data. The pressure correction ensures that any static corrections between wells are accounted for on the various analysis plots.
3) A new Extended Simulation Option is now available to control the type-curve simulation behaviour when the period-for-analysis is a subset of the entire well history. In previous versions, two simulations were carried out to span the period-for-analysis and then all of the well-history after the period-for-analysis. For complex models, this extra simulation could take some time to complete. The 2010.09 version of PIE now has an "Extended Simulation" option on the Analysis menu which will show a check-mark if the extended simulation is to be done, or just the period-for-analysis is the check-mark is not displayed. Selecting the option will toggle the check-mark on or off.
4) The Simulation Time algorithm has been improved. This algorithm sets-up the time-grid for a simulation taking into account the number of flow-periods and the rate-changes in the offset-well rate histories. The new algorithm ensures that a more detailed and sensible grid is used in those flow-periods being displayed on the analysis plots (as selected by the period-for-analysis and the "auto-overlay" option).
5) There is a new "Splash Screen" which displays a message indicating the 2010.09(bp) version of PIE is a proprietary product. The installation process will also display an end-users license agreement warning that this version of PIE contains BP proprietary technology.
Release date: 9-Feb-2009
A major update to increase the time resolution (1 second in 68 years) to avoid rounding errors when displaying gauge-data over a very long time span.
Release date: 31-Oct-2007
1) a "Flow-Period Properties" Function is now available on any data plot. This function can be accessed by a "right click" on a particular flow-period and selecting the "Flow-Period Properties" option from the menu. Alternatively, this function can be accessed by the keyboard combination "Control-W" and then clicking on a particular flow-period. The "Flow Period Properties" function will display a menu showing the duration, rate, start-time, and a flag indicating if the flow-period is valid for analysis. The menu also displays a series of check-boxes allowing particular settings to be assigned to a flow period. These settings are part of the analysis-data for a test, and are saved along with the test data to a PIE-file. Each of these new settings are described in the next two topics of this release note.
2) the Period-for-Analysis (PFA) can now be changed using the "Flow Period Properties" settings. If the PFA setting is checked prior to starting an analysis, then it is remembered and applied when the first analysis plot is displayed. If this setting is checked during an analysis, then the PFA is changed to the new flow-period and the data-plot is displayed showing the new PFA and ready for the selection of other analysis plots from the "Plots" pull-down menu. Note that this method for setting the PFA will always use a rate-history that starts from the first flow-period in the analysis-data. If a different start for the rate-history is wanted, then old "Period-for-Analysis" option on the "Main" menu is still available which allows a specific start and end for the PFA to be selected from a data plot.
3) an Auto-Overlay Manager has been added to manage which flow-periods in the analysis-data are "overlaid" for comparison on the derivative and superposition analysis plots. The "Flow-Period Properties" option includes separate "auto overlay" check-box settings for the derivative and superposition plot. Each flow-period that is valid for analysis and has these settings checked will automatically be overlaid onto the derivative and/or superposition plot.
The overlaid curves can be temporarily turned "on" or "off" using the option "Auto-Overlay on/off" under the "Graphics" pull-down menu. Note that a check-mark is shown to indicate when the auto-overlay manager is "on". Plots that auto-overlay several sets of data can still be saved, copied, and overlaid like any other PIE plot using the plot manager.
All of the flow-periods available for auto-overlay can be viewed using the option "Auto-Overlay Selection" under the "Graphics" pull-down menu. This option will display a table showing all of the flow-periods in the test (or just the shut-in periods) along with a flag to show if the auto-overlay manager will "include" or "exclude" a flow-period when constructing the derivative or superposition plot. A range of lines on this table can be selected in order to "include" or "exclude" a group of flow-periods from the auto-overlay manager. This "Auto-Overlay Selection" option is particularly useful for manipulating the auto-overlay settings for a long production history containing many individual shut-in's.
4) the Plot Legend has been improved to display the date-time associated with each set of data on the plot. This change was made to work in conjunction with the auto-overlay manager in order to clearly identify which sets of data are displayed on a particular plot. Note that the plot legend is displayed by a right-click on the plot, then selecting the "Plot Legend..." option.
5) the Graphics Text Labels can now be given individual colour and font characteristics. This change clears up a peculiarity in the PIE display caused by the legacy of using GKS graphics index-based drawing attributes.
Release date: 9-Aug-2006
1) Areal Anisotropy is now included with the homogeneous, double-porosity, linear-composite, horizontal-well, and inclined-well models. The corresponding interference-test models also include areal anisotropy to allow for multi-well simulation. This anisotropy is shown as the ratio "Areal ky/kx" i.e. the permeability in the y-axis direction divided by the permeability in the x-axis direction.
2) the Type-curve simulation will now extend past the end of the current period-for-analysis on the data-plot. This simplifies matching later portions of the test data while analysing an earlier shut-in. The data plot is scaled to show the current period-for-analysis. The extended simulation can be shown by increasing the plot scales using the "Set Plot Scales" option under the "Graphics" menu.
3) the Graphics Display has been re-written to allow 2147483648 different colours. Previously, this was limited to four different attributes which would result in duplicate colours for complicated overlaid plots. The new scheme is guaranteed not to repeat colours when creating a overlaid plot. The Graphics Output has also been re-written to use a higher resolution that is more suitable for a printer.
4) The ASCIIprep Input "Wizard" has been modified to allow a new gauge to be created while keeping the current gauge rate data. This is handy for keeping the same rate-data when preparing data from multiple gauges. Also, the "reference date/time" value is used when reading data in a date-time format. This allows you to define a common decimal-time scale across different sets of data.
5) The Copy Option for graphics now comes in two different styles. The normal "Copy" will use whatever output preferences have been set. The new "Copy Plot Only" option will copy just the plot itself with no analysis results. This is handy when preparing reports because it avoids repeated selection of the "Output Preferences" option to switch between styles.
6) The Analysis Results Window has been modified so values can be deleted (i.e. set to a "null" value). This is useful when doing specialised analysis (e.g. tide processing) and different calculation options need to be invoked based on the number of valid parameters entered into the analysis.
7) The Period-for-Analysis option has been modified to correct some awkward behaviour. You can now "drag" the mouse over a range of flow-periods and the program will select a sensible period-for-analysis. Tolerances have been included so a near-miss on a flow-period will not result in an odd selection. Normally, you would drag the mouse over the shut-in period that you want to analyse. The old style of "double clicking" on a shut-in period can still be used.
Release date: 30-Jun-2005
1) The Data Processing in PIE now includes a "rate set-up" facility within the data processing wizard. This option will take raw rate measurements and work out a sensible rate-history for the test based on the accuracy of the data, and a desired target number of flow-periods for the rate-history. This rate set-up facility can also be accessed directly from the "ASCIIprep" menu option "Scan to set-up Rate-History" for use at any point in the data preparation process.
This rate set-up facility does not guarantee shut-in periods are exactly aligned with the pressure data, so it is still important for the user to zoom-in on the start of each PBU that will be used the analysis and ensure the shut-in starts at the correct time. This is done by clicking on the vertical-line in the rate-history and dragging the rate-change to the correct point in the pressure data.
2) the ASCIIprep menu options have been improved. The option to "Add a Flow-period to the Test" will now accept a click-and-drag operation to mark-out a period within the data to become a flow-period. Moving a rate-change in the rate-history can be done by clicking once on the vertical-line marking the rate-change, and then using a "shift-click" in the pressure data to move this time to a point found by the "maximum gradient" selection method.
3) "Un-Do" and "Re-Do" options can now be accessed with the "Control-Z" and "Control-Y" keyboard combinations.
4) A Plot Legend is available for all plots displayed by the program. To add a legend to a plot, right-click anywhere in the plot window, select the "Plot Legend..." option, and then select the desired position of the legend on the plot. To re-position the legend, just click on the legend itself to re-display the positioning options. Future versions will include more options to edit the legend entries and set the position of the legend within the plot frame.
5) User Input is now "localised" to account for different style of number formats (e.g. thousands separator, different decimal points, etc.) based on the computer regional settings selected by the user. The raw-data processing has always had this "localisation", but this new feature extends this numeric processing to all input.
Release Date: 24-May-2004
1) The Data Processing in PIE has been re-arranged and simplified. An extra step in the input wizard will ask to identify new data being read or "pasted" from the clipboard, and allow an "automatic pressure data selection" option to be invoked. Only three steps are required for data processing; read the raw data, set-up a rate-history, and then save the data to a PIE-file.
The Analysis-Data Manager under the EDIT menu is now the central point where multiple sets of data are managed in the program. Re-loading a set of data will automatically re-load and display the associated set of gauge-data.
2) the ASCIIprep menu has been revised so all the main options are more easily available with sub-menus for the extra functions. This revised menu is used in place of the graphics "buttons" to provide a more legible description of what each function does. The ASCIIprep menu is now the main menu for manipulating the rate and pressure data for a test.
3) "Un-Do" and "Re-Do" options have been added to the EDIT menu for all rate and pressure editing operations. The program will track an unlimited number of changes until the analysis is started or the data is saved to a PIE-file.
4) The set-up for an Interference-Test has been simplified so the observation well with the pressure measurements can be given a rate-history made up of a single zero rate. The production or injection at the surrounding wells in the interference-test are set-up using the "Offset Wells" option under the EDIT menu.
6) The Data Processing in PIE has been re-arranged and simplified. An extra step in the input wizard will ask to identify new data being read or "pasted" from the clipboard, and allow an "automatic pressure data selection" option to be invoked. Only three steps are required for data processing; read the raw data, set-up a rate-history, and then save the data to a PIE-file.
The Analysis-Data Manager under the EDIT menu is now the central point where multiple sets of data are managed in the program. Re-loading a set of data will automatically re-load and display the associated set of gauge-data.
7) the ASCIIprep menu has been revised so all the main options are more easily available with sub-menus for the extra functions. This revised menu is used in place of the graphics "buttons" to provide a more legible description of what each function does. The ASCIIprep menu is now the main menu for manipulating the rate and pressure data for a test.
8) "Un-Do" and "Re-Do" options have been added to the EDIT menu for all rate and pressure editing operations. The program will track an unlimited number of changes until the analysis is started or the data is saved to a PIE-file.
9) The set-up for an Interference-Test has been simplified so the observation well with the pressure measurements can be given a rate-history made up of a single zero rate. The production or injection at the surrounding wells in the interference-test are set-up using the "Offset Wells" option under the EDIT menu.
1) A new "Deconvolution" option is a major addition to the program. Deconvolution is a very useful regression calculation that reveals whether or not the analysis-data has a unique underlying type-curve response that explains the observed pressure behaviour. All deconvolution options are located on the "Deconvolution" sub-menu under the "Analysis" pull-down menu. Each option starts a "wizard" interface to provide a structured sequence of steps to set-up the deconvolution algorithm.
1) A new 2-PHASE WATER-INJECTION MODEL has been added to the program. The model correctly models the process of water-injection taking into account the movement of the "flood front" and the initial behaviour when starting injection into a reservoir at the initial water saturation. The model can be used with an arbitrary injection history. The model also includes temperature-dependent viscosity to represent the change in fluid mobility as cold injection fluid warms up to the reservoir temperature. The 2-phase water-injection model is selected by choosing an "Injection" well-type in a "homogeneous" reservoir with a "radial" reservoir geometry.
2) A new MULTI-WELL OPTION has been added to the program that greatly simplifies the set-up of complicated problems. A new option has been added to the EDIT pull-down menu where the name, location, and rate-history for up to nine offset wells can be added. Then the analysis proceeds as normal with the pressure interference from the offset wells taken into account when doing a type-curve simulation or non-linear regression.
The HELP system has been modified to include
movies in the "How To.." topics under the HELP pull-down menu. These
movies illustrate the sequence of options needed to carry out a particular part
of an analysis. A web-based version of these "How To..." topics can
also be seen at:
4) The TIDE EFFECTS option under the "Analysis" pull-down menu has been improved to use a new tide potential-function which computes the gravitational force caused by the relative orbital motion of the earth, moon, and sun. This allows for better removal of tide effects from the analysis data.
5) The TIDE EFFECTS option has also been split into two separate functions; one function to set the parameters for the tide potential-function, and a second function to shift the time at which a peak or trough affects the data.
6) A new TIDE PROCESSING module has been added under the "MAIN" pull-down menu. This module contains a comprehensive set of options for extracting the tide response from a set of bottom-hole pressure data. Options are also available for correlating the tide response with surface ocean-tide measurements or with the tide potential-function. When the bottom-hole tide response is correlated with surface data, the reservoir total-compressibility can be computed based on the attenuation between the surface and bottom-hole tide response. Monitoring the total-compressibility over time can be related to changes in fluid saturations.
1) The operation of the "ASCIIprep" menu for gauge-data preparation has been changed significantly. The interface for reading data from a text-file or from the "copy and paste" clipboard has been changed to follow the same steps as used in the EXCEL "Text to Columns" wizard. Gauge-data is now sorted by the program allowing raw data to be added in any order.
2) An "Estimate a Rate-History" function has been added to the "ASCIIprep" menu to automatically set-up a rate-history from the gauge-data. This function is now the main method for starting preparation of the analysis-data from the gauge-data.
3) Options for Permanent Gauge-Data processing have been developed to smooth out and "de-noise" large volumes of bottom-hole pressure data. These options are based on a "wavelet" transform of the data which allows specific types of noise to be reliably removed from the pressure-data.
4) The GRAPHICS interface has been improved. A uniform set of lines styles and colours are used for both the screen display and print-out (WYSIWYG). The user now has full control of the graphics colours, styles, and fonts. Various plot elements can be altered by using a mouse "right-click" or "double-click".
5) A "PLOT-STYLE" has been added to hold all of the information required to define the appearance of a plot. Each plot has its own independent plot-style. A global "default-style" can be defined to control the appearance of new plots, or to return a plot to the default-style. This default-style is automatically saved so it can be loaded back whenever PIE is started.
6) A "RADIAL-COMPOSITE INTERFERENCE-TEST" model has been implemented to compute the pressure response at an arbitrary point in a 2-zone radial-composite reservoir. This model include rectangular no-flow reservoir boundaries that can be used to model a single fault, parallel faults, or a closed drainage-area.
1) A "General Multi-Layer, no cross-flow model" has been added to the program. This model represents a multi-layered system where no there is no cross-flow between layers in the reservoir, but there is cross-flow through the well-bore. A different well, reservoir, and boundary model can be assigned to each layer of the system. The available layer models range from a simple vertical-well homogeneous reservoir to a complex linear-composite system or an inclined well. This model is selected using the "Type-curve" option under the "ANALYSIS" pull-down menu and then choosing a "General" well type, in a "Multi-Layer no Cross-flow" reservoir with a "Radial" geometry.
1) A new "Analysis Data" manager that will allow different sets of rate and pressure data to be stored in a single PIE data-file. This allows the user to work with a single PIE-file to set-up and store different views of the analysis-data (e.g. a detailed sub-set of a full well history, different rate-history configurations, pressure data based on different gauges, etc.) . The new Analysis-Data manager is located under the "EDIT" pull-down menu and allows a single PIE-file to act as a central repository for all of the data and results associated with a particular well.
2) A new linear-composite type-curve model has been added to the program. This model represents a reservoir with up to five "strips" of rock each having different properties. The model also covers interference-tests and multiple producing segments within the basic linear-composite reservoir geometry.
3) An Impulse-test plot using the method described by Peres, Onur, and Reynolds has been added under the "PLOTS" pull-down menu. This plot will allow parameter estimates to be made and will display the simulated impulse pressure response for a selected type-curve model.
4) The "pseudo-time" transform for the analysis of gas-wells is now available under the "select test-type" option on the EDIT pull-down menu.
5) PIE can now read and write the basic static, rate, and pressure data to and from the more recent text based PAN data-files (the older binary PAN-files cannot be read or written). These files are accessed when a "PAN file" file-type is selected from the drop-down list on the standard file-name dialog. Note that no results are read from or saved to the PAN-file; this will be part of a later update.
1) Several new type-curve models have been added to the program since the 09/97 version. The inclined-well model was completely re-written to use a finite well-bore radius solution and include reservoir boundaries. The same inclined-well model with boundaries is also available as an interference-test model. The hydraulic-fracture models are now available in a double-porosity as well as a homogeneous reservoir.
2) A new "Reservoir-Limits" plot has been added to the "PLOTS" pull-down menu. This plot is used to determine the pore-volume connected to a well. The plot used in PIE has been modified from the traditional reservoir-limits "pressure verses time" display in order to handle arbitrary data with varying rates (see chapter 12.8 of the user manual).
3) PIE is "Year-2000" compliant. The program can handle any date format including ambiguous 2-digit dates for years on or after 2000. Preferences for decoding different date formats are saved in the registry so they can be automatically restored when the program starts.
1) A new multi-layer, multi-lateral horizontal-well model has been added. This model allows up to four horizontal well segments to be located at arbitrary depths and orientations in a multi-layer reservoir.
2) A new multi-layer inclined-well model. This model allows an inclined producing interval of an arbitrary length to be positioned in a multi-layer reservoir.
3) A new multi-layer multi-perforation partial-penetration model has been added. This model allows up to four vertical perforated intervals to be arbitrarily positioned in a multi-layer reservoir.
4) New Sand-Face-Rate-Convolution (SFRC) plots to analyse continuous bottom-hole rate and pressure data.
5) Options to remove tidal-effects from the data or add tidal-effects to a test design.
6) Two new interference-test models have been added. The first is a multi-layer horizontal-well interference test model where the pressure response is calculated between two horizontal wells with an arbitrary orientation. The second is a multi-layer partial-penetration interference-test model where the pressure response is calculated between an arbitrarily located observation interval and one to three producing intervals.
Two new type-curve models have been added:
A "General-Composite" model that allows continuous or discontinuous changes reservoir permeability, porosity, thickness, viscosity, or compressibility. This model can be also be used for interference tests. The property variation can be described as a series of step-changes, as a series of linear changes, or as a combination of both step and linear changes. The technical details of this model have been published in SPE 30554.
The Fetkovich & Arps decline-curve model can now be used as a pressure-function to match build-up tests.
One of the fundamental theoretical problems in well-testing is the analysis of gas-well tests where there is significant reservoir depletion. The pseudo-pressure transform corrects the effect of non-linear behaviour of gas viscosity on the calculation of permeability. However, this transform does not correct the material-balance error caused by the non-linear changes in gas compressibility for a well in a closed reservoir. Hence, using the usual transient solutions for a gas-well test where there is significant reservoir pressure depletion results in a material-balance error for the calculation of average-pressure and pseudo-steady-state productivity.
A new solution method has been implemented in PIE that applies a correction factor to the transient solution that ensures the overall material-balance of the system is honoured. This correction is based on new theoretical methods to 'de-superpose' different boundary-conditions on the normal convolution calculations used for transient test analysis. The end result is a PIE simulation that correctly computes the linear "P/Z vs. cumulative production" characteristic of gas-reservoir depletion as well as the correct transient response.
To select this new solution method, set the test-type from the "Premier Menu" to the new option "Pseudo-Pressure with Material-Balance Correction".
New plots are available to display the well productivity-index (PI) as a function of time or cumulative production. These plots are prepared using the true average-pressure as determined by exact material-balance equations. As a convenience, the average-pressure of the system is also displayed with the well productivity-index.
These plots are very useful in identifying trends in productivity and when a well has reached pseudo-steady-state flow. The average-pressure display is a handy way to estimate depletion (particularly for gas-wells) using the exact material-balance equations.
New "History-Match" and "Production-Prediction" options are available to compute rates and pressures from arbitrary production constraints. The production constraints can specify a fixed rate, a fixed flowing pressure, or a maximum rate for a given flowing pressure. These constraints can be applied in any combination over the duration of the production data.
These new options allow PIE to solve general production problems using a new "analytic simulator". For example, a quick production forecast can be made using the maximum-rate constraint to limit production to a 'plateau' rate until a limiting bottom-hole flowing pressure is reached after which the production decline is calculated. The analytic simulator allows well behaviour to be analysed using whatever data is most appropriate for the analysis.
The "History-Match" option allows the Optimizer to adjust the model parameters to obtain a best match of the analytic-simulator output to measured production data. For example, a simulation could be set-up to compute the rates during a draw-down from specified flowing pressure constraints and the pressures during a build-up from a zero fixed rate constraint. The Optimizer would adjust the model parameters for this simulation set-up to find a best-fit to measured rate data during the draw-down and to the pressure data during the build-up.
The "History-Match" and "Production-Prediction" options are accessed from the 'MAIN' pull-down menu. These options work just like a regular type-curve simulation or test-design. The only new item to look for is the "Simulation Constraints" option under the 'HISTORY-MATCH' or 'PROD.PREDICTION' pull-down menus. This is where the simulation constraints (i.e. fixed-pressure, fixed-rate, or maximum-rate) are set-up.
The 'analytic simulator' used for the new "History-Match" and "Production Prediction" options has been extended to include the effect of off-set 'pulser' wells. This allows PIE to do a small-scale 'full field' analytic simulation to account for the effects of the 'pulsers' on the production behaviour at an observation well. The rate and/or pressure behaviour at the observation well is computed given an arbitrary set of production constraints; just like the new "History-Match" or "Production-Prediction" options.
The "Multi-Well Simulation" option is located under the 'MAIN' pull-down menu and is used to match the analytic-simulator to a set of measured production data. New options to set-up the 'pulser' wells and to set-up the paths between the pulsers are located under the 'MULTI-WELL' pull-down menu. There are a number of additional plots located under the 'PLOTS' pull-down menu to display results at the observer and/or pulsers.
The "Multi-Well Design" option is located under the 'MAIN' pull-down menu and is used to prepare a test design for a multi-well system. The operation is just like the "Multi-Well Simulation" except that the simulation can be converted to a synthetic set of analysis data and the non-linear regression options are turned off.
The following new type-curve models have been implemented:
- a multi-layer
horizontal-well model with cross-flow
between reservoir layers.
- a composite-channel
model in a channel or closed-
rectangle reservoir geometry.
interference-test model with reservoir
cross-flow. This model allows the pressure in any one
layer to be observed at a distance from the active well.
hydraulic-fracture model with no reservoir
cross-flow. This model allows individual hydraulic-
fractures to be specified in different layers with all
layers coupled together at the well-bore.
- multi-layer horizontal-well
model with no reservoir
cross-flow. This model allows individual horizontal
wells to be specified in different layers with all
layers coupled together at the well-bore.
- all of the multi-layer
models with no reservoir cross-
flow now use closed-rectangle boundaries in each layer.
This geometry together with the boundary 'fog-factors'
allows complex reservoir limits to be specified in each
The use of skin-factors in gas-well tests has been made much more sensible by defining all input skin-factors to be mechanical-skin (i.e. skin with no turbulence effects). This means a change in the period-for-analysis does not affect a simulation because mechanical-skin is the same regardless of the rate-history. This combination of parameters is also more 'intuitive' because increasing either the mechanical-skin or turbulence gives a larger pressure drop in a simulation.
Skin-factors calculated and displayed by the program (e.g. Global-Skin for a horizontal-well) are now defined to be total skin-factors i.e. they include the effect of turbulence. This allows the overall effect of the input mechanical-skin and turbulence factor to be seen as an overall skin-factor.
The objective function for regression of data spanning several flow-periods has been improved. The value of initial-pressure can now be fixed to a particular value or constrained to a range of values in order to force the optimizer to look at solutions that have a specified initial pressure-drop to the first data point.
Processing raw gauge-data from an ASCII-file was kept as a separate module in previous versions. This new ASCII-Reader is integrated into the program in a more flexible way and allows raw data from several gauges to be read and stored in a PIE- file. The ability to store gauge-data means you only need to read the raw-data once and can easily return to the raw-data to re-select data for analysis.
When storing gauge-data in a PIE-file, the program compresses the data by a factor of 2 to 4 depending on how the data is represented in the ASCII-file. In a test case, a set of raw data in a 1.8 Megabyte ASCII-file reduced to a 560 Kilobyte PIE-file that included the data selected for analysis.
The interface to the new ASCII-Reader is entirely graphical and uses a new 'ASCIIprep' pull-down menu on the menu-bar. You can perform all of the main processing operations from this pull-down menu. The graphics-functions at the bottom of the plot contain the 'tools' needed to carry out specialized operations on the data, however, you should not need to use them for most cases.
A new facility has been added to allow you to reduce the volume of gauge-data. This is handy for cutting down the amount of raw-data in those portions of the test where a lot of detail is not required.
With the facility to save the data from several gauges in a PIE-file comes a new facility to compare and synchronize gauges. This new module is accessed from the 'Main' pull-down menu by the option 'Compare Gauges'.
After selecting a base and comparison gauge, any specified range of data can be compared by plotting the difference between the two sets of gauge-data. This difference is plotted together with the average and standard deviation of the difference data. There are a number of graphics functions that allow the comparison-gauge to be shifted relative to the base- gauge in order to minimize the 'scatter' in the difference data and synchronize the two gauges. The applied time shift can then be used to update the comparison-gauge.
Note that for very large sets of raw-data on low-performance computers, you will need to reduce the gauge-data to something reasonable (e.g. 20,000 points) since the difference calculations are very intensive. To this end, you might want to save one full set of gauge-data and one set of reduced data for comparison.
PIE will no longer support plain MS-DOS computers or 'dumb' terminals (e.g. Tektronix, HP, VT240, etc.). These systems are just too limited to handle the new features in PIE. There is no plan to support these systems in future, although the Tektronix terminal may be available on special request.
The supported environments are:
- Microsoft Windows
version 3.1 or higher
- MOTIF on any UNIX system
- Plain X-Windows on any UNIX system
This should cover most computer systems in use today.
The multi-layer model with reservoir cross-flow has been completely re-written to be more robust and flexible. The new solution includes reservoir boundaries in any configuration and accepts a inter-layer cross-flow coefficient (lambda) of zero. This means that you can model multi-layer systems with groups of layers in communication and with reservoir boundaries.
This feature completely changes the interpretation section of PIE. The main objective of this change was to gather the most commonly used options together into a simple graphical interface. This change cuts out a large number of highly specialized options and concentrates on the basics.
When the "Derivative Analysis" option is selected from the "Premier Menu", the derivative plot is displayed. The "DIAGNO" key contains the diagnostic functions eg. UNITSL, STABIL, etc. The "MODEL" key brings up the menu to select a type-curve model. Once a model is set, two extra functions become available; "ESTIMA" for the special function used to estimate model parameters (eg. Lambda and Omega for double- porosity) and "TCSIM" for the functions to control a type- curve simulation and the Optimizer.
To view the test on different plots, the "PLOTS" key contains functions to access a superposition or data plot. For 'window' systems, these plots are contained in their own window and analysis operations can be switched between plots simply by 'clicking' on the plot. Analysis and type-curve simulations done on one plot are updated to be shown on the other plots.
A number of derivative analysis functions have been enhanced to remove the need for specialized analysis plots. For example, the UNITSL function will compute a storage coefficient and use this coefficient to set the end of well- bore storage marker, ENDWBS, thereby removing the need to use the specialized delta-p vs. delta-t wellbore storage plot.
For those of you who prefer the old PIE menu system, this can be accessed from the PREMIER MENU using option 1 to select a 'General Analysis'. Note that specialized plots like Odeh- Jones and specialized analysis like Interference tests can only be accessed from the 'General analysis' user-level.
For MacIntosh and MS-Windows users, the program operation has been improved to take better advantage of 'ease of use' facilities in the window interface on these systems. For example, selecting a file is done using a standard file- selection dialog instead of typing in a file name. All printing is now done through the standard system facilities.
On Microsoft Windows systems, individual windows can be 'Minimized' or 'Maximized'. This is handy to arrange the display of windows to your liking. Also note that graphics windows can be re-sized as desired with the plot filling the new window. As a small feature, the windows can be moved around by 'dragging' the title-bar of the window, but this operation does NOT change the active window. This feature allows inactive windows to be re-sized and positioned over a large active window so different views of the test can be arranged as desired.
Copy and Paste clipboard editing facilities have been added so the plots and text from PIE can easily be transported to other Mac or Windows program. The Paste operation is only allowed for text eg. a table of pressure data can be pasted into a PIE. The addition of these clipboard facilities removes the need to print and gather PIE output separately from the program used to generate a report.
With further releases, the menus in PIE will take on a more sophisticated look ie. nice little boxes, buttons, and colours. For this release, the main priority was ensure the major facilities were operational.
A pressure gauge model has been introduced to the type-curve simulation menus. The gauge model defines a transfer function that adds the effect of gauge resolution, noise, drift, and 'lag' to a type-curve simulation. Look for the graphics function 'GAUGE' or the menu option 'GAUGE MODEL' when doing a type-curve simulation.
This is a very important addition to PIE as it provides a means to design a test within realistic hardware limitations and to check that characteristics on the analysis plots are truly significant reservoir behaviour. It is strongly recommended that this option be used with AT LEAST the gauge resolution and noise parameters set. This will ensure that the analysis is done on data that is within the gauge measuring capabilities.
A default list of gauge types can be selected when using the gauge model (eg. a Standard Quartz gauge). These data come from a few published sources and are probably reasonable values, however, this is by no means guaranteed. The parameters for each gauge are stored in the PIE configuration file, hence, changes can easily be made. If any users have information about the correct parameters for a gauge type, then contact SMC to update their version with this data and have that gauge included in future releases.
Previous versions allowed you to compute the response for a multi-well test (ie. more than one pulser) but did not provide adequate facilities for using the Optimizer to find a match. This has now been fixed by subtracting the response of all the active wells from the Observer data except for the pulser selected for optimization. This results in a display of the net pressure response for a selected pulser. As an added feature, a flowing observation well can be matched to itself ie. all of the pulsers are subtracted from the observer data leaving just the observer pressure transient behaviour.
In addition, a new menu has been added that allows wells to be selectively switched 'on' or 'off'. This is handy in order to quickly check the response from individual pulsers. Also, an option for a 'Homogeneous' reservoir is available which forces all pulsers to have the same KH and storivity regardless of changes to any one pulser.
Complicated, but these options are a big improvement.
A partial-penetration infinite-conductivity hydraulic-fracture with a fracture skin model has been added. This is selected as a sub-type of a hydraulic-fracture well in a homogeneous reservoir with a radial geometry. This model is particularly good for 'penny shaped' fracs in very thick reservoir sections (eg. acid fracs in chalk reservoirs).
The partial-penetration model has been improved by using the 'fast Greens functions' implemented in the last version for the Horizontal well model. This helps to make the partial- penetration model compute faster.
The addition of the Results Manager with the last release allowed the user to store as many different cases as desired. However, one over-sight was to not store the model static data (eg. perforation thickness, horizontal drain-length, etc.) with each case. Hence, it was possible to save a case, change the static data, re-load the saved case, and obtain different results. This problem is now cured and each case that is saved includes the current model static data. When the results for a case are viewed with the 'V' command, the model static data is also displayed.
Note that results saved with earlier versions will not have the model static data saved. Hence, the default model static data entered on the appropriate menu will be used. If you re- save the results with this version, the static data is saved with the results.
As PIE has matured, a good number of reporting options have been added. These are still in the appropriate places in the program, but have also been gathered into a new Report Menu. This is where you can load a PIE data-file, display/print the data, review/print stored analysis plots, review/print stored analysis results.
In addition, a general X-Y data plot option has been added to the Report Menu. This allows you to set up a plot of any sort of data you desire (eg. a static gradient survey). A very useful 'trick' is to set-up the general plot to have the same type and context as a PIE analysis plot. This allows the analysis plot to be restored as a general plot where you can add new data or manipulate the plotted data using the XYDATA or DEL-XY graphics functions.
Four new test-types have been added to the program; multi- phase Perrine analysis, multi-phase oil pseudo-pressure analysis, multi-phase gas pseudo-pressure analysis, and multi- phase water pseudo-pressure analysis. The multi-phase pseudo- pressure functions are defined either as a gas pseudo- pressure, a Fetkovich pressure-squared function, or by a user- defined table.
When doing a multi-phase analysis, the data processing has been changed to allow you to enter oil, gas, and water rates either by direct input or by processing a particular phase- rate in the ASCII file-reader. The 3-phase rate data can be viewed or edited using a number of different terms; GOR, WOR, BS&W, OGR, etc. The options used to display the data show all three phases for the rate data.
The Perrine test-type means the analysis is done with a total sand-face flow rate to obtain a total mobility. This total mobility is then 'split' into individual phase permeabilities depending on the flowing fraction of that phase.
The pseudo-pressure test-types allow the pressure data to be transformed to account for the non-linear terms in the equations for multi-phase flow. In general, some information about the reservoir flow is needed (eg. solution-gas drive) to create these pseudo-pressure functions. This problem has NOT been addressed in this release and the user will need to enter a pseudo-pressure function for the particular case being analyzed. For an oil pseudo-pressure analysis, the Fetkovich pressure-squared pseudo is available (see SPE paper 14237).
The turbulence coefficient has been left in as part of the parameters for multi-phase analysis. This means that oil wells can now be analyzed with a 'rate dependent skin'. For now, turbulence must be a positive value ie. a higher skin with a higher rate. However, if the demand is there, this can be changed to allow a negative turbulence factor.
The user can select a particular phase-rate to be set when processing raw data from an ASCII file. This allows the rate data for main flowing phase to be set-up from raw data files instead of tedious entry by hand.
To avoid the pesky prompts about saving/loading results, a new Results Manager has been added. At any time, you can select a 'Results Management' option from an interpretation menu in order to view, print, delete, save, or load results in memory and/or previously saved results. As many results as desired can be saved for the SAME model. If you forget to save results, the result manager will automatically save results for the current model. Each set of results, including those currently in memory, are identified by a user-defined comment string.
The number of basic analysis functions that can be performed for any type-curve models has been increased. At any time, the user can calculate (using the appropriate plot) the following:
- storage coefficient
- the end of wellbore storage
- permeability and skin
- fracture length and conductivity
- permeability anisotropy (Kv/Kh)
- distance to a boundary
- width of a linear flow channel (parallel boundaries)
- intersection angle of two boundaries
To accompany this unification, graphics function keys that are NOT appropriate for use are NOT shown. Hence, the user should never be faced with selecting graphics functions that will not do a calculation because the wrong model has been selected.
All models that use a skin factor, except for the 3 and 4 layer models, now include turbulence effects. This includes the horizontal and inclined well models. 1) New PIE User Interface
Out is the teletype operation and in is a fully portable window interface developed specifically for PIE running under MS-DOS. The following functions are supported:
- Menus are displayed in a 'box' and the input cursor is located at the point where the data is entered. The different levels in the program are represented as a 'stack' of windows on the screen. This ensures that the user can see where they are in the program.
- all commands are meaningful on all screens eg. the 'T' command will scroll a table to the top line or move the cursor to the first field on a display with no scrolling table.
- when a mouse is available, the user can click on ANY word on the screen to use that word as input (a word is defines as characters delimited by a blank, comma, or semi-colon). If the user clicks on a data-input field, the cursor will move directly to that field.
- when a mouse is available, clicking on another window will cause the program to do the necessary options to move to that menu eg. a click on the PREMIER menu while at the ASCII data-file reader menu will return the program directly back to the PREMIER menu.
- for systems without a mouse, a set of 'Exception' commands are available
Support for systems without a mouse will improve as time allows more complex 'drivers' to be written for various systems (eg. VAX).
The currently published horizontal-well solutions using Green's functions are for a line-source well. This is not a problem where the reservoir permeability and Kv/Kh ratio are reasonable. However, when long horizontal wells are used to exploit low permeability fields with a low Kv/Kh ratio and negative skin, this model will fail as the difference between a line-source and finite-wellbore radius becomes large for small values of Td during vertical radial-flow.
To remedy this problem, a new Green's function was developed for a finite wellbore radius source in a slab reservoir. This allows a finite-wellbore radius horizontal well model to be solved by Green's function methods.
Note that this model is only available for a horizontal well in an infinite reservoir. If reservoir boundaries are included (even with a fog-factor of zero) then the line-source solution is used instead.
In addition to a finite-wellbore solution for the horizontal well, a much faster algorithm for computing the horizontal- well Green's function solution has been found. This makes the horizontal-well model easier to use as type-curves or the Optimizer work much faster. In addition, the horizontal well in a rectangular reservoir now computes as fast as the infinite acting solution. This is particularly useful with the Well Deliverability analysis or Prodution Prediction options for analyzing/predicting long-term horizontal-well production for a specified drainage area.
A new model has been added to represent an infinite- conductivity vertical fracture with skin. This new model has a very fast solution, hence, it can be used for the initial analysis of hydraulic fractures with skin instead of the very complex semi-analytic solution for the finite-conductivity fracture with skin model. Use this model as a possible solution for tests that show a derivative-plot storage 'hump' followed by a 1/2 slope.
Two new options have been added to the Premier Menu; option 10 for Deliverability Analysis and option11 for Production Prediction. These modules use dimensionless rate superposition calculations to compute simulated rates given a pressure history.
Deliverability Analysis allows you to take measured time, rate and pressure data and match the rate data with any of the models available in PIE. To support conventional decline curve analysis, the Fetkovitch and Arps decline Type-Curves are available. The standard decline curve analysis plots can be used to display the results of the simulation.
Production Prediction is similar to the Deliverability Analysis except that you only need to specify the pressure history from which rates will be calculated. This allows you to take well and reservoir data and make a prediction of the production rates.
Four new models have been added to the program; a three-layer reservoir with cross-flow, a four-layer reservoir with cross- flow, a linear composite reservoir, and the Fetkovitch Decline Curves.
The three and four layer reservoir models are particularly useful for the analysis of partial penetration with a gas-cap or aquifer. You can specify the individual layer properties so layers can represent zones containing different fluids. A new example of this kind of analysis has been added to the "Worked Examples Reference" which uses the example data file, FOURLY.EG.
The multi-layer models in PIE have been extended to include no reservoir cross-flow systems with 2, 3, or 4 layers. These models can have a closed circular boundary in each layer with a different radius specified for each layer. These circular boundaries each use a 'fog-factor' ie. they can vary from constant pressure to no-flow. In addition each layer can have a different initial pressure. You choose one these models by selecting a vertical well in a 'Layered, No-Crossflow' reservoir from the model selection menu.
The main reference for these models is SPE 18125, "Transient Pressure Behaviour of Commingled Reservoirs" by F.J. Kuchuk. The initial condition used for these models represents a well prior to perforation, ie. a uniform initial pressure in each layer with no production from any layer. The individual layer initial pressures are represented by differential pressures relative to layer 1.
These models are very complex (up to 19 unknowns for a 4 layer reservoir) hence, they should only be used where specific scenarios are to be modelled. Don't expect these models to provide a unique analysis of an unknown reservoir.
A new module has been added which allows you to design or simulate an interference test. Up to 9 wells can be used in the interference test. Any of the wells in the test can be selected at any time to be the observation well. An observation well can be flowing or shut-in.
Two types of Odeh plot are provided with the Diagnostic plots and Conventional Analysis plots. These plots allow all pressure data, both flows and shut-ins, to be plotted on one display. Any number of straight-lines can be drawn on the plot to determine permeability and skin factor. For gas-well tests, the turbulence coefficient and mechanical skin factor are calculated. These plots provide an alternative AOF analysis for gas wells.
This new model allows you 'import' an externally defined type- curve model into PIE via an ASCII data file. The ASCII data file contains keywords defining:
- the name of the
- the name of a parameter (eg. Cde(2S))
- the default log-log axis scales
- options for time-match calculations
- the array of dimensionless time values
- one or more arrays of dimensionless pressure values for different parameter values.
This user defined model is fully incorporated into PIE with no restrictions on its use ie. multi-rate type-curves and type-curve simulations can be done.
An example ASCII data file for a homogeneous well in a closed circular reservoir is supplied with this release (file CIRCLE.CRV on the EXTRA disk).
A type-curve fit Optimizer has been added to the Interpretation module. This is a non-linear least-squares regression package that will match a given type-curve model to the data.
While this new feature can be a labour saving utility, it can also be abused ie. used to fit a model without thinking about why the model is appropriate. Refer to the documentation in the user manual for more information.
The Optimizer works with all of the current models availible in PIE; even the 'User Defined' model. It can also be used to match a two-well interference test.#
The internal format of the time data values has been changed from decimal hours to integer seconds. This switch allows the program to compute delta-t values exactly to the nearest second over a large range of absolute time. This change fixes the problem of rounding of small delta-t values over large flow periods of several months.
In general, you will notice no difference in the operation of the program. It is possible that some data files with very small delta-t values will be loaded to give two time values the same. If this occurs, use the data modification option to look at the pressure data and change/delete the point with the same time value as an adjacent point. Save the data file after making this change.
When you save a data file with this version, the time data are stored as integer values. This is incompatible with previous versions. This new version will recognize an old data file and convert the time data accordingly. However, old versions of the program will not read the new data files correctly.
ENSURE ALL USERS OF PIE UPGRADE TO THIS RELEASE OF THE PROGRAM. WHEN A FILE CONFLICT OCCURS (NEW FILE USED WITH OLD VERSION) UPGRADE THE INSTALLATION TO THIS VERSION.
A new ASCII file reading module has been added to the program which replaces the old tabular system of reading a large ASCII data file. The main features of the new system are:
- reads the entire ASCII file into PIE (max limit 1,999,800 points; larger limit on request...)
- Manual or Automatic data selection algorithms
- selects data and creates a rates history
- can read ASCII rate data and calculate cumulative average rates from this data when creating the rate history
- all operations done with a graphical interface
- bad ASCII data can be deleted using the 'ZAP' graphics function.
The big advantage to a graphical interface is the entire test can be seen and the individual test flow periods can easily be selected. In general, this new ASCII file reading module greatly simplifies loading data into PIE from large service company files.
A new simplified gas PVT menu has been created which allows quick setup of PVT properties. Moreover, once the PVT data has been set for a particular data file, it remains set for all future uses of the program with that data file.
There is only one menu to work with when setting up the gas PVT data. Default property values based on correlations are calculated once the pressure range of the PVT tables, gas composition, specific gravity, and temperature have been set. You can then modify these defaults by using different options for the correlations or by entering a table for a particular property. The PVT data can be viewed in a table or on a plot.
You can now perform more analysis functions on the derivative analysis plot in the Interpretation menu. The new functions are:
SPHERE - spherical flow analysis for Kv/Kh
LAMBDA - derivative minimum defines the double-porosity parameter, Lambda.
CHANEL - define a 1/2 slope for the width of a channel
FAULT - define the deviation from the stabilization due to a reservoir hetrogenaity
ANGLE - define the final stabilization for intersecting boundaries to set the angle between boundaries.
Also note the SKIN key has been removed. Each key that sets a stabilization uses both the time and pressure of the point picked on the plot to define both permeability and skin factor. Therefore, selection of a stabilization requires a point that lies within the data on the plot and not just a pressure level.
You can now define two straight-lines on a square-root plot for the analysis of parallel, no-flow reservoir boundaries. An additional function key, 3-SIDE, is used to define the second line which sets the distances for a parallel boundary closed at one end. The analysis calculation on this plot calculate the channel width and the well location within the channel.
The AUTOSL key ALWAYS defines a line representing a linear flow in a channel open at both ends. The 3-SIDE key ALWAYS defines a line representing linear flow in a channel closed at one end. Therefore, use the AUTOSL and/or 3-SIDE keys to set the appropriate analysis line or lines.
A number of graphics function keys and a new option in the data modification menu allow you to store graphics plots in a PIE data file. Plots are stored by simply selecting the PLTSAV key under the MISC key. The complete plot (including analysis lines and text) is stored in the current PIE data file.
Stored plots can be recovered so an analysis can be resumed at the point the plot was saved. Also, stored plots can be overlaid onto the current plot to allow comparisons of multiple sets of data on a single display. Note that this overlay feature can be used to overlay a plot stored in a different PIE data file than the current one.
After loading a PIE data file in the main data manipulation menu, use option 7 to access a new option under the Other data modifications menu. The new option is for Review of stored plots and this allows you to display any stored plot in the current PIE data file. After selecting a plot for display, you can overlay other plots, modify the plot title, re-position the plot text, or make a hardcopy.
This is a new model that represents two reservoir layers with no communication between the layers in the reservoir. Each layer is within a closed circular boundary with a boundary type ranging from no-flow (fog-factor=1) to constant pressure (fog-factor=-1). Intermediate values of the boundary type fog- factor represent a transmissibility change which is modeled by the radial composite equations. The radius of the closed circle can be different for each layer.
A typical application of this model is a well with two layers and one layer pinches out. This will set up differential depletion between the layers resulting in wellbore crossflow during a pressure build-up.
This new model is a variation of the partial penetration model. All of the parameters are the same except for the specification of the pressure measurement location at a point above or below the perforated interval. Use this model to match a vertical pulse test to determine the Kv/Kh ratio.
This model was re-written to provide an accurate solution for extreme mobility ratios. Also the meaning of the static data was changed to clarify the definintion of inner and outer zone properties.
The re-write of the radial composite solution will result in some differences from the previous solution for mobility ratios greater than 5 or less than 0.2. The differences should not be sufficient to require re-analysis of a test.
The new meaning of the various static data for this model are the reverse of previous versions. The inner zone static data are the well static data entered from the data manipulation menu. The outer zone static data are set when you select the radial composite model. Note that this is the reverse from previous versions, hence, old analysis that use this model should ensure the well static data and type-curve static data are correct.
This model requires specification of the individual layer properties. In the old version, this was done by setting the data for one layer and the program back-calculated the other layer properties from the well static data entered in the data manipulation module. This was on the basis that the well static data was the average properties for combined two layers.
This version now allows you to set the data for each layer when you select a two-layer model. Once you have entered the data, the current well static data in memory is updated to be consistent with your data. You can make this well static data permenant by re-storing the test data in a PIE data file. You can recover from an error by re-loading the PIE data file since the data file is not altered by any of these steps.
Some changes were made in the use of the perforation skin factor for this model. The type-curve calculations assumed the perforation skin entered by the user did not include the anisotropy skin, but the specialized analysis for horizontal well perforation skin did include the anistropy skin. This conflict has been resolved in this release however, it is only important for very small Kv/Kh values (eg. 0.0001).
This type curve has been added to the availible models. It can be used for both Homogeneous and Double Porosity reservoirs. In conjunction with this model, specialized analysis for the derivative plot, Horner plot, Superposition plot, and Semi-log plot have been added. The specialized analysis allows the mobility ratio, perforation skin factor, and injection zone radius to be calculated.
This type-curve has been added to allow analysis of inclined wells where early-time effects of a transition between radial flow perpendicular to the wellbore to pseudo-radial flow are present. The solution is found using Green's functions. Both Homogeneous and Double-porosity reservoirs can be modeled.
This type-curve is the limiting behaviour of the Inclined Well type-curve, but takes advantage of a simpler solution for a 90 degree inclination. In addition, specialized analysis on the derivative, Horner, Superposition, and Semi-log plots has been added to allow estimates of Kv/Kh ratio and perforation skin factor. Both homogeneous and double-porosity reservoirs can be modeled.