School of Petroleum and Mining Engineering
Petroleum Engineering Department
Course title Reservoir simulation with ECLIPSE
HW # 4
Title Modelling
Date 31.03.2023
Performed by Kanatbay Darmen, 2nd year student
Almaty, 2023
INTRODUCTION
The introduction should provide a clear statement of the problem posed by the HW/Lab/Practice. The introduction needs to
present background information. A brief summary of the unique approach you used to solve the problem should be given,
possibly also including a coincide introduction to theory or concept used later to analyze and to discuss the results.
Sometimes, in modelling, we may need to decrease the number of cells in the grid (without
changing it’s geological sizes) to overview the object in a more general perspective in order to
make quicker calculations and simulations.
In this lab work, we tried to perform an upscaling process for the grid, upon our latest
lecture about the corresponding subject. During it’s implementation, such subjects as permeability
average, viscous fingering effect and polymers were reviewed.
MAIN PART
Description of the problem. What you should to obtain.
According to the requirements of this lab work, we need to perform upscaling process (3
cases) and compare the properties of each upscaled grid with the base case. Moreover, we need to
analyze the changes in results along with providing appropriate plots and FloViz models for each
case.
For the last point of this lab work, we needed to try to improve our recovery by either
adding polymers or changing completion data (by modifying base case!) and perform the
corresponding analysis.
MATERIALS AND METHODS
Describe the procedure to carry out the solving method/measurements and how the solving method/measurements
were analyzed. Show the solving procedure.
Methodology for this lab work consisted of thorough plan of desired cases, types of
averaging permeability and consequent trial/error to get the perspective of the workflow and
process.
According to the requirements, 3 cases of upscaling were desired.
Thus, the base case file has been modified into three new DATA files, with each having
different GRID dimensions and correspondingly averaged permeabilities. Despite of changed
dimensions, the geological sizes were not affected, as the equilised size was calculated according
to these new DIMENS values. Affected keywords are DIMENS, WELLDIMS, DX/DY/DZ, BOX,
TOPS, PERM, PORO, REGIONS, WELSPECS and COMPDAT.
For the first case, initial dimensions of 50/7/23 (NX,NY,NZ) were upscaled to 10/7/6.
Permeability was averaged by only using the method for flow parallel to uniform layers, or
this formula.
Permeability averaging for the first case
For the second case, base case dimensions (50/7/23) were upscaled to 25/7/12.
Permeability was averaged in two methods – with using averaging for flow parallel to uniform
layers for PERMX and PERMY, and using averaging for flow across uniform layers for
PERMZ.
- permeability average for flow across uniform layers
Permeability averaging for the second case
For the third case, base case dimensions were upscaled to 25/7/6. Permeability was
averaged by using two methods described previously.
Permeability averaging for the third case
For the last part of this lab work, base case is modified into a new DATA file such that the
recovery is improved. There were two choices of achieving this – by adding polymers or by
changing completion data. The experiment has followed the polymerization method.
Regarding the data file changes: POLYMER has been added to the list of phases; such
properties as PLYVISC, PLYADS, PLYROCK, PLYMAX and TLMIXPAR were added to the
PROPS section; WPOLYMER was added to SCHEDULE section (values for given keywords were
adopted from the practice file).
RESULTS
Present your actual results. The results should be summarized using appropriate tables and figures (graphs or
schematic). Every figure and table should have a legend that describes concisely what is contained to shown.
FIRST CASE OF UPSCALING:
Base case FloViz Model (PERMX)
Upscale #1 FloViz Model (PERMX)
SECOND CASE OF UPSCALING:
Base case FloViz Model (PERMX)
Upscale #2 FloViz Model (PERMX)
Upscale#2 FloViz Model (PERMZ)
THIRD CASE OF UPSCALING:
Base case FloViz Model (PERMX)
Upscale #3 FloViz Model (PERMX)
Upscale#2 FloViz Model (PERMZ)
ALL CASES VS BASE CASE:
FOPT (Field oil production total)
FWPT (Field water production total)
Field Oil Production Rate
WBHP for INJ
WWCT:PROD
ATTEMPT OF IMPROVING RECOVERY (LAST PART):
Field Oil Production Rate
WBHP for INJ
WBHP for PROD
WWCT:PROD
FOPT (Field oil production total)
DISCUSSIONS
The DISCUSSION interprets the results in light of the project's objectives. The DISCUSSION should also present an evaluation
of the particular approach. For example: Based on the results, how could the experimental procedure be improved? What
additional, future work may be warranted? What recommendations can be drawn?
To analyze results of upscaling cases, generated excel files (.RSM) were used to calculate
the relative error. Each case was overviewed in terms FOPR, WBHP:INJ and WWCT. In-
program formulas allowed the calculation of relative errors, which are
№ of case FOPR % FOPT % FWPT %
1 23.5 57.3 177.1
2 25.7 31.7 201.3
3 22.4 47.2 35.1
As it can be seen, relative error for FWPT gives absurd values, which may be caused by
some sharp fluctuations to higher or lower values at certain points of time. Thus, it is believed that
the third case of upscaling is better and closer to base case in terms of process calculations.
One may assume that during the upscaling process for the third case, average
permeabilities for PERMZ were the most similar to the base case PERMZ in terms of
unpermeable layers, which contributed to low relative error compared to others.
The problem here is that during all upscaling processes, layers with no permeability were
also averaged among the rest of PERM values. This might cause problems in calculations as we
might not track the fluid inside the grid correctly as it is assumed to flow throughout the certain
region despite it having unpermeable blocks. Thus, I assume using LGR would be beneficial in
this case.
In terms of improving recovery, polymerization has significantly contributed to good
results in field oil production. It is explained by the fact that adding polymer to displacing fluid
(water in our case) is increasing it’s viscosity which excludes the possibility of viscous fingering
effect. This way, more oil is swept out of the reservoir.
CONCLUSION
The CONCLUSIONS should summarize the central points made in the Discussion section, reinforcing for me the value and
implications of the work. If the results were not definitive, specific future work that may be needed can be (briefly) described.
Therefore, any conclusions should be based on observations and data already discussed.
To conclude the experiment, it needs to be stated that none of upscaling cases were close enough to
the real base case properties. Unfortunately, unpermeable zones were not excluded from the
averaging and the relative error was unacceptable. It was concluded that using LGR for non-
upscaleable regions would be more beneficial. In case of polymerization, it showed great results
for recovery.
REFERENCES:
The REFERENCES section should contain complete citations following standard form. The form of the citation depends on the
type of source being referenced, and is different for whole books, chapters in books, and articles published in a journal.
[Link]
Your HW/Lab/Practice should be send only in electronic form in doc/docx format to
classroom. The same reports will be evaluated for zero marks.