Oil Reservoir Simulations

Format	JSON, HDF5
License	CDLA-Permissive
Domain	Sequence Modelling, Time Series Analysis
Number of Records	30000 simulations
Data Split	Split by data type, including SPE9-TRIANGLE, SPE9-MAX and other supporting data files.
Size	4.7 GB
Author	Jiri Navratil, Georgios Kollias, Andres Codas
Data Origin	IBM Research
Dataset Version	Version 1 – April 10, 2020
Dataset Coverage	The ORSD data collection contains tens of thousands of sequences (simulations) generated by a physics-based simulator of an oil-reservoir field.
Business Use Case	Build, train and develop machine-learning models so people can accurately predict output sequence, given an input sequence. Build three-dimension visualization of an oil-reservoir field.

Feature	Values
WELSPECS	P00 PG 18 10 9110.00 OIL 60
COMPDAT	P00 18 10 1 8 OPEN 1* 1* 1.0 1* 0 1* Z
WCONPROD	P00 OPEN ORAT 1070.9358266204822 1* 1* 1* 1* 1000.00 1* 0
WECON	P00 100 1* 0.90 550 1* WELL
TSTEP	90
WELSPECS	I00 IG 24 2 9110.00 WATER 60
COMPDAT	I00 24 2 1 15 OPEN 1* 1* 1.0 1* 0 1* Z
WCONINJE	I00 WATER OPEN RATE 1064.5992284885779 1* 6000.0000
TSTEP	90
WELSPECS	I01 IG 19 11 9110.00 WATER 60
COMPDAT	I01 19 11 1 15 OPEN 1* 1* 1.0 1* 0 1* Z
WCONINJE	I01 WATER OPEN RATE 713.0405723954925 1* 6000.0000
TSTEP	90

Feature	Description
SPE9-MAX	This dataset contains simulations generated by random drilling sequences that distribute uniformly over the entire surface grid of the SPE9 RM (24x25). The aim of the SPE9-MAX is to evaluate a ML model as an interpolator, i.e., a predictor of previously unseen sequences drawn from the same distributions as well as the same geological region. The dataset contains 30,000 such simulations partitioned into training, development, and test sets.
SPE9-TRIANGLE	This dataset contains simulations generated by random drilling sequences that distribute uniformly over a constrained triangular portions of the reservoir surface. The aim of the SPE9-TRIANGLE is to evaluate a ML model as an extrapolator, i.e., a predictor of sequences drawn from a different geological region. This dataset also contains 30,000 such simulations partitioned into training, development, and test sets.
Sample -> WELSPECS	The WELSPECS keyword defines the general well specification data for all well types, and must be used for all wells before any other well specification keywords are used in the input file. The keyword declares the name of the well, the initial group the well belongs to, the wellhead.
Sample -> COMPDAT	COMPDAT defines how a well is connected to the reservoir by defining or modifying existing well connections.
Sample -> WCONINJE	The WCONINJE keyword defines injection targets and constraints for wells that have previously been defined by the WELSPECS keyword in the SCHEDULE section. Note that wells can be allocated to a group when they are specified by the WELSPECS keyword. Wells defined to be under group control will have their injection rates controlled by the group to which they belong, in addition to any well constraints defined for the wells using this keyword.
Sample -> WCONPROD	The WCONPROD keyword defines production targets and constraints for wells that have previously been defined by the WELSPECS keyword.
Sample -> WECON	WECON defines the economic criteria for production wells that have previously been defined by the WELSPECS and WCONPROD keywords.
Sample - TSTEP	TSTEP keyword advances the simulation to a given report time after which additional keywords may be entered to instruct OPM Flow to perform additional functions via the SCHEDULE section keywords, or further TSTEP keywords may be entered to advance the simulator to the next report time.