rubis guided session #4 - kappa eng · pdf filethe rubis guided session #2 illustrates how a...

Ecrin v4.20 - Doc v4.20.01 - KAPPA 1988-2011 Rubis Guided Session #4 RubGS04 - 1/14
Rubis Guided Session #4
A01 Introduction
The Rubis Guided Session #2 illustrates how a model built in Topaze can be dragged over to
Rubis, extended and copied back to Topaze. The present guided session will illustrate further
the interaction between Rubis and the other modules of the Ecrin workstation: this time, a
sector model will be created, copied over and analysed in Saphir, and the interpretation results
will be used to improve the history match in Rubis.
A basic knowledge of both the well test analysis concepts (deciphering a loglog plot, for
instance) and the basic Saphir options is required. In addition, it is assumed that an active
Saphir NL licence is available in addition to the Rubis licence.
Key functionalities presented: creation of a sector in Rubis, drag & drop to Saphir,
use of the Rubis sector in Saphir, multi-run document, using the restart facilities to
run a simulation from a given time with the corresponding conditions, results
comparison between runs.
B01 Creating the Rubis Sector
B01.1 Motivation for a Rubis sector analysis
Load the file RubGS04.krp:
Fig. B01.1 RubGS04.krp file content

This file contains a single run named Initial history match, in which a simple 2 phase black-
oil/water model has been built and simulated. The reservoir contains three wells: one producer
named Prod 1, and two water injectors Inj 1 and Inj 2:
Fig. B01.2 Reservoir geometry
Long-term monitoring bottomhole pressure gauges are available at all wells, as well as daily
measurements of the oil rate at the producer and water rates at the injectors. In the
simulation the producer target is the pressure gauge, whereas both injectors are controlled by
the water injection rates.
A closer look at the Prod 1 well simulation results shows that a very good overall match has
been achieved:
Fig. B01.3 History match at well Prod 1: the well is controlled by the surface oil rates
(bottom) whereas the bottomhole pressure (top, fast modelling option is used) are fitted.

A single production shut-in is present at the end of the well history, at the end of the year
2005:
Fig. B01.4 Looking at the first year of production of Prod 1
in the Well Data page: 8 buildups are clearly visible
In the following, we will import the buildup measurements in Saphir and perform a
rudimentary well test interpretation in order to better understand the flow behavior around the
well Prod 1, and possibly use its outcome to improve the history match. The Saphir
interpretation file will be automatically created by the Rubis sector to PTA option.
B01.2 Creating the Rubis sector file
Move to the 2D tab, a composite limit defines a region around the well Prod 1:
Fig. B01.5 2D view of the reservoir with its composite limit

Create a sector ( ) in the region containing Prod 1:
Fig. B01.6 2D view of the reservoir after creation of the sector
As we will see later, the Rubis sector model dragged to Saphir fully uses pre-existing restarts
created by the original Rubis run when those exist, in order to better respect any reservoir
history prior to the selected sector date. However, Rubis cleans out restarts when those have
been saved in a Rubis file carrying a different version flag (e.g., a restart saved in Rubis 4.20.x
is wiped out upon reload in version 4.20.y is y > x). So you need to check whether the
RubGS04 file contains stored restarts: move to the restart settings page of the Simulation
run settings dialog by clicking on the Run settings button in the control panel. If the Existing
restarts table contains restarts, you can move to the next step. Otherwise, you need to re-
simulate the Rubis case, after having checked that the Store restarts flag is enabled.
Exit and press to initialize, then on to run the simulation.
Back in the 2D tab, double-click on the sector icon to enter the sector edition dialog:
Fig. B01.7 Sector edition dialog

This dialog defines the basic settings required for the creation of the Saphir interpretation file
(which well should be taken as reference well, which productions are to be transferred, etc...).
In our case:
Prod 1 will be the reference well of the
Saphir file (it is the only well in the sector
anyway):
We do not need to copy the complete well
history as we want to investigate the last
build-up. Change the copy start date from
January 1st 1999 to October 1st 2005.
Click on Synchronize to adjust the start
date to an existing restart:
Accept all other default settings and click on Create Saphir Document >> to switch to Saphir:
Fig. B01.8 Saphir document built from the Rubis sector
Save the Saphir file.

B02 Saphir Analysis of the Rubis Sector
B02.1 Generating the Rubis sector model in Saphir
The oil production and pressure data measured at the Prod 1 well from September 26th, 2005
have been transferred to the newly created Saphir file.
We do not need to start the Rubis sector simulation from the very beginning of 1999 because
the initial pressure and saturation maps corresponding to the reservoir status of September
26th, 2005 have been carried over: the simulation in Saphir will use this status as an initial
state.
Extract the final build-up:
Fig. B02.1 After extraction of the build-up
The Rubis sector analysis contains the full Rubis two-phase black-oil model, except that the
geometry is limited to the extracted sector. The PVT and Properties icons visible in this
analysis exhibit the same behaviour as in Rubis and lead to the same dialogs.

For instance, click on Properties in
Saphir to enter the Rubis sector model
Properties dialog, and in this dialog
edit the permeability dataset of the
Bottom property set ( ):
This is to be compared with the same
dialog visible in the original Rubis file -
note that the data point coordinates
have changed: the 2DMap origin has
been moved during the transfer of the
Rubis sector in order to have it coincide
with the tested well in the final Saphir
document.
Similarly, it is possible to edit the perforations and the wellbore model of all wells in the Wells
option. Note that the Geometry and Data icons are not accessible anymore, as both
geometry and data have been carried over from Rubis upon the sector creation.
The Controls icon is not accessible either in
the Wells dialog, but well controls may still
be edited in the Wells controls page of the
Run Settings dialog ( ):

In the Time stepping control page, keep the
default End of simulation to 2108 hr, and
modify the output fields time period to 100
hr:
Accept all defaults in the Numerical settings page, and click on OK to validate. Simulate the
Rubis sector model to obtain the following match:
Fig. B02.2 After simulation of the Rubis sector
Note that unlike the full Rubis history match, the Rubis sector has been simulated on a fine
grid in order to better reproduce the early time flow features around the well, as can be seen
in the 3D plot:

Fig. B02.3 Pressure field at 2000 hours in the sector
note the fine simulation grid around the well
B02.2 Comparison between Rubis sector, analytical and numerical models
Create a new analysis named Analytical by copy of the current sector analysis in the copy
only the 2D geometry will be carried over because all Saphir linear models are limited to single
phase (without gravity):
Fig. B02.4 Creating a new analysis
Extracts the same build-up as before and do an auto-model (Shift + ) followed by an
auto-improve (Shift + ) to match a simple analytical model (storage and skin vertical well
in an infinite acting homogeneous reservoir) on the data.

Fig. B02.5 Match with an analytical model
A good match is obtained with the parameters values indicated above, with a zero skin slightly
smaller than the original model coming from Rubis.
On the other hand, it can be seen on the history plot that the infinite acting reservoir model is
not able to reproduce the production period pressure trend visible on the data, as Rubis sector
starts from the non uniform restart pressure map.
Create now a new analysis by copy of the Analytical one, and name it Numerical linear. Open

rubis guided session #4 - kappa eng · pdf filethe rubis guided session #2 illustrates how a...

Documents