Systems and methods for predicting hydrocarbon production and assessing prediction uncertainty
Methods and systems for predicting hydrocarbon production and production uncertainty are disclosed. Exemplary implementations may: obtain training data, the training data including (i) training production data, (ii) training engineering parameters, and (iii) a training set of geological parameters and corresponding training geological parameter uncertainty values; obtain an initial production model; generate a trained production model by training the initial production model; store the trained production model; obtain a target set of geological parameters and corresponding target geological parameter uncertainty values and target engineering parameters; apply the trained production model to generate a set of production values and corresponding production uncertainty values; generate a representation using visual effects to depict at least a portion of the set of production values and corresponding production uncertainty values as a function of position within the subsurface volume of interest; and display the representation.
1. A computer-implemented method for training a model using stochastic input parameters for predicting hydrocarbon production and hydrocarbon production uncertainty in a subsurface volume of interest, the method being implemented in a computer system that includes one or more physical computer processors, non-transient electronic storage, and a graphical user interface, the method comprising:
obtaining, via the graphical user interface, (i) training production data, corresponding to the subsurface volume of interest, (ii) training engineering parameters, corresponding to the subsurface volume of interest, and (iii) a training set of geological parameters and corresponding training geological parameter uncertainty values, corresponding to the subsurface volume of interest;
obtaining an initial production model;
applying the training production data, the training engineering parameters, and the training set of geological parameters and corresponding training geological parameter uncertainty values to the initial production model via the graphical user interface;
training the initial production model by using the training data;
generating a trained production model; and
storing the trained production model.
2. The computer-implemented method of claim 1 , wherein the training set of geological parameters comprises at least one of carbonate content, porosity, and silica content corresponding to the subsurface volume of interest.
3. The computer-implemented method of claim 1 , wherein the training production data comprises at least one of cumulative oil production, historical oil production, cumulative gas production, historical gas production, and cumulative oil production at a plurality of different time intervals.
4. The computer-implemented method of claim 1 , further comprising:
determining the training set of geological parameters and corresponding training geological parameter uncertainty values, corresponding to the subsurface volume of interest;
performing a coarse grid inversion to generate a coarse gridded model via data from the subsurface volume of interest;
interpolating the coarse gridded model to generate a fine gridded model and the set of geological parameters; and
determining a predicted uncertainty of the set of geological parameters based on a P-value calculation of the fine gridded model to generate the training geological parameter uncertainty values.
5. The computer-implemented method of claim 2 , wherein the training set of geological parameters further comprises an expected trend value and a statistical uncertainty of the set of geological parameters interpolated from nearby measured geologic parameter data to the collated spatial location of the production data.
6. The computer-implemented method of claim 4 , wherein the coarse gridded model comprises:
m est =[ G T G+ε 2 S ] −1 G T d=G −g d
and interpolating the coarse gridded model comprises bicubic interpolation.
7. A computer-implemented method for applying a model for predicting hydrocarbon production and hydrocarbon production uncertainty in a subsurface volume of interest, the method being implemented in a computer system that includes one or more physical computer processors, non-transient electronic storage, and a graphical user interface, the method comprising:
obtaining, via the graphical user interface, a target set of geological parameters and corresponding target geological parameter uncertainty values;
obtaining, via the graphical user interface, target engineering parameters;
obtaining a trained production model, the trained production model having been trained by training an initial production model, wherein the training data includes training production data, corresponding to the subsurface volume of interest, (ii) training engineering parameters, corresponding to the subsurface volume of interest, and (iii) a training set of geological parameters and corresponding training geological parameter uncertainty values, corresponding to the subsurface volume of interest;
applying the trained production model to the target set of geological parameters, corresponding target geological parameter uncertainty values, and target engineering parameters, via the graphical user interface, to generate a set of production values and corresponding production uncertainty values;
generating a representation using visual effects to depict at least a portion of the set of production values and corresponding production uncertainty values as a function of position within the subsurface volume of interest; and
displaying the representation.
8. The computer-implemented method of claim 7 , wherein the target set of geological parameters and corresponding target geological parameter uncertainty values comprises:
calculating a geological parameter trend and a geological parameter tread uncertainty of the target set of geological parameters by interpolating the training set of geological parameters corresponding to the subsurface volume of interest; and
wherein the training set of geological parameters is based on sampled well data.
9. The computer-implemented method of claim 8 , wherein calculating the geological parameter trend comprises:
calculating an inversion model via a generalized least-square inversion using the equation, d=Gm;
wherein m comprises trend values at gridded locations;
d comprises geological parameter data; and G comprises a matrix including the equations for an estimate of each gridded model location as a distance weighted sum of neighboring data.
10. The computer-implemented method of claim 7 , wherein the target engineering parameters comprise at least one of well length, injected proppant, drilling properties, wellbore configurations, fluid types, and well spacing at a new location in the subsurface volume of interest.
11. The computer-implemented method of claim 8 , wherein calculating the geological parameter trend uncertainty comprises:
determining a P-value, residual and total variance, F-test ratio and N eff computed in the inversion model.
12. The computer-implemented method of claim 7 , wherein applying the trained production model comprises:
analyzing each node of the trained production model and corresponding uncertainty; and
predicting the set of production values and corresponding production uncertainty values.
13. The computer-implemented method of claim 7 , wherein applying the target set of geological parameters and corresponding target geological parameter uncertainty values comprises:
applying a geological parameter trend and a statistical uncertainty of the target set of geological parameters corresponding to the new location in the subsurface volume of interest to the trained production model to generate the set of production values and the corresponding production uncertainty values.
14. A computer system for training and applying a model for predicting hydrocarbon production and hydrocarbon production uncertainty in a subsurface volume of interest, the system comprising:
one or more physical computer processors,
non-transient electronic storage,
a graphical user interface;
memory; and
one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more physical computer processors, the one or more programs including instructions that when executed by the one or more physical computer processors cause the system to:
obtain, via the graphical user interface, (i) training production data, corresponding to the subsurface volume of interest, (ii) training engineering parameters, corresponding to the subsurface volume of interest, and (iii) a training set of geological parameters and corresponding training geological parameter uncertainty values, corresponding to the subsurface volume of interest;
obtain an initial production model;
apply the training production data, the training engineering parameters, and the training set of geological parameters and corresponding training geological parameter uncertainty values to the initial production model via the graphical user interface;
train the initial production model by using the training data;
generate a trained production model;
store the trained production model;
obtain, via the graphical user interface, a target set of geological parameters and corresponding target geological parameter uncertainty values;
obtain, via the graphical user interface, target engineering parameters;
apply the target set of geological parameters, corresponding target geological parameter uncertainty values, and target engineering parameters to the trained production model, via the graphical user interface, to generate a set of production values and corresponding production uncertainty values;
generate a representation using visual effects to depict at least a portion of the set of production values and corresponding production uncertainty values as a function of position within the subsurface volume of interest; and
display the representation.
15. The system of claim 14 , further comprising:
calculating a P-value, residual and total variance, F-test ratio, and N eff based on the training geological parameters.
16. The system of claim 15 , wherein the P-value is calculated from the tail area above the F-test ratio; the F-test ratio comprises:
F
-
test
Ratio
=
(
N
e
f
f
-
1
)
(
v
a
r
i
a
n
c
e
e
xp
lained
)
v
a
r
i
a
n
c
e
residu
al
and the N eff comprises
N eff =d pre =Nd obs with N=GG −g .
17. The system of claim 14 , wherein the training geological parameter uncertainty values and the target geological parameter uncertainty values comprise:
Predicted
Uncertainty
=
(
P
v
a
l
u
e
*
v
a
r
i
a
n
c
e
total
+
(
1
-
P
v
a
l
u
e
)
*
v
a
r
i
a
n
c
e
residu
al
)
18. The system of claim 14 , wherein the initial production model and the trained production model comprise a machine learning probabilistic random forest regression.
19. The system of claim 18 , wherein training the machine learning probabilistic random forest comprises:
formulating a cost function in view of training parameter uncertainty values; and
generating a tree structure, wherein the tree structure comprises one or more nodes corresponding to one or more levels of the tree structure, and
wherein traversing to each terminal node of the one or more nodes is based on a probability determined by the cost function.
20. The system of claim 19 , wherein the cost function for determining the optimal splitting threshold for a feature k comprises:
PRSS
(
T
,
k
)
=
∑
i
=
0
N
∫
-
∞
T
(
y
i
-
y
¯
l
)
2
N
(
x
i
k
,
σ
i
k
)
dx
+
∫
T
+
∞
(
y
i
-
y
¯
u
)
2
N
(
x
i
k
,
σ
i
k
)
dx
.