Chapter Outline
Introduction
Characteristics of History Production Data
Principles of History Matching (HM)
History Matching Workflow
Assisted History Matching (AHM)
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 173
Introduction
What is HM? The process of adjusting/modifying the
existing reservoir model in such a way as to achieve a
better fit to the observed data.
Why HM? To make sure the reservoir model is reliable
representation of the real reservoir before it is used for
future performance prediction.
How? To solve an inverse problem to minimize the
difference between the calculated and observed data in a
sense of
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 174
Introduction
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 175
Introduction
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 176
Pure History Matching - Not Really Good!
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 177
History Matching with Flow Characteristics –
Better Result!
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 178
HM in Reservoir Simulation Workflow
Static Model Dynamic
Model
Production History
Forecasting Matching
Strategic Plan
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 179
Characteristics of History Production Data
Data source from surface and/or bottom-hole (permanent
downhole gauge - PDG)
Historical production is observed as time-series data.
Time-series data characteristics: Trend, seasonality,
cyclicity, and noise.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 180
Characteristics of History Production Data
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 181
Characteristics of History Production Data
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 182
Principles of History Matching
Select response to be matched (production data, pressure
data, etc.)
Select parameters and their range to vary to obtain match.
Identify important parameter – those which are highly
uncertain and which affect the reservoir response to be
matched.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 183
Methods of History Matching
History
Matching
Automatic History Manual History
Matching Matching
Construct a mathematical model
Define an objective function
Apply a minimization algorithm
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 184
Quantification of the variables level of
uncertainty
Variables are often considered to be indeterminate (high
uncertainty): pore volume, permeability, transmissibility,
kv/kh ratio, rel. perm. curves, aquifer properties, etc.
Variables are often considered to be determinate (low
uncertainty): porosity, gross thickness, net thickness,
rock compressibility, capillary pressure, etc.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 185
History Matching Workflow
3. Determine the historical
1. Set the objectives of 2. Determine the production data to be
the history-matching method to use in matched and the criteria to be
process. the history match. used to describe a successful
match.
4. Determine the reservoir
6. Compare the results
5. Run the simulation data that can be adjusted
of the history match
model with the best during the history match
run with the historical
available input data. and the confidence range
production data chosen
for these data.
in Step 3.
7. Change the reservoir 8. Continue with Steps 5
data selected in Step 4 through 7 until the criteria
within the range of established in Step 3 are
confidence. met.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 186
History Matching Prediction Workflow
Interpreted geology, geophysics,
petrophysics Sensitivity runs
Produce results for risk
evaluation and economics
Static reservoir description
Construction of simulation grid Prediction runs
Modification
Existing of static
wells or dynamic
continue to
until actual and model
produce and are worked match
Assign fluid model and rock physics over logically
function New well drilling can be
Tuning runs implemented
Petrel simulation model Modify properties until EOR options can be tested
model and historical
data match
Make history strategy
(actual production and pressure)
Sensitivity runs
NO
Identify uncertain parameters
Simulate production and pressure
Solution found
YES
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 187
History matching improvement tool
Production Data Integration
― Automated methods of incorporating production data into
geologic models.
Experimental Design
― Identifying key variables
Assisted History Matching
― Using the unique data in streamlines to aid history
matching
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 188
Assisted History Matching (AHM)
Assisted history matching is the process of accelerating the
matching process.
The simulation engineer specifies the parameters to vary to
obtain a match, and the response to be matched.
A computer program (like CMOST-CMG) is then given the
task of finding the values of the match parameters which
give the best history match.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 189
Why are we interested in Assisted History
Matching?
Global Flow Unit
(Fault Block)
Individual
Layer
Individual
Well
History matching
frequently comes down to
local changes
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 190
When to use Assisted History Matching
To refine the match obtained by manual history matching.
After identifying the most important unknown parameters.
Assisted history matching methods work better (and run
faster) when given a smaller number of parameters to
match.
Do not include parameters that have little or no effect
on the desired response. Some methods may fail entirely
for under-determined problems.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 191
Assisted History Matching method
Most methods involve some variation of least squares.
Seek parameters α that minimize the objective function:
n 2
F j y t y t
obs i sim i
i 1
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 192
ECLISPE keyword for HM
WCONHIST
Specific to production wells in history matching
Sets the observed rates, per phase, in surface conditions
Calculates the production rate depending on the chosen
control mode
WHISTCTL:
Allows to change only the control mode; for example, to pass
from a reservoir rate control to a surface oil rate control
WCONINJH: for injection wells
*This keywords can be created with SCHEDULE application.
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 193
ECLISPE keyword for HM
EXAMPLE
WCONHIST
- 1 2 3 4 5 6 7 8 9 10
-name stat ctrl Qos Qws Qgs VFP Qgl THPobs BHPobs
‘P1’ 'OPEN' 'RESV' 255 15 1000 0 1* 1* 150 /
/
WHISTCTL
-- 1 2
--new status BHP action
‘ORAT’ 'NO' /
WCONINJH
-- 1 2 3 4 5 6 7 8
--name phase stat Qinj BHPobs THPobs VFP Rs/Rv
‘I1’ 'WATER' 'OPEN' 1000 1* 300 2 1* /
/
11/22/2024 Dr. Mai Cao Lan - Faculty of Geology & Petroleum Engineering, HCMUT 194
