Professional Documents
Culture Documents
Shadravan 2017
Shadravan 2017
Fluid design
Fig. 1—Applications of the intelligent design tool for various fluid-design experiments.
Inverse
Modeling
Fig. 3—Schematic of the proposed intelligent integrated tools for guided experiment design, characteristic prediction, and fluid-
system design.
50 40 30
70
60 45 30 25
Viscosity (cp)
Viscosity (cp)
50 40
40 20
35 20
30
30 10
20
15
10
25 0
0 400
5 20 5 10
4 20 300 4
3 200 3
2 15 15 2
10 1 5
1 100 0
0 5
Ingredient A (gal/bbl) Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(a) (b)
Fig. 4—Nonlinear estimator resulting from ANN (pink asterisks show the training data).
and covariance function) are suitable (Deutsch and Journel 1992), GPR
or Kriging is the best linear unbiased estimator. In geostatistics lit-
U ¼ Hidden layer to output layer wights ½ðq þ 1Þ m: erature, the spatial dependencies are represented by a variaogram
ð7Þ function instead of a covariance function. In signal-processing lit-
erature, this estimation procedure is known as Kalman filtering,
The activation function at neurons is chosen to be a hyperbolic which has been also applied to history-matching problems (Tarrahi
tangent function. We used the backpropagation method in sequen- et al. 2013, 2015). GPR by nature does not need any a priori
tial mode to train the network and obtain the weights W and U assumption on the form of regression functions. It is a purely data-
matrices (Bishop 2006; Theodoridis et al. 2010). To start the driven methodology that extracts the relationship between varia-
learning process, we initialize the weights as random numbers. bles from the information provided by the training data.
Estimation error for the training data is not a promising perform- It is certainly the inherent correlation and relationship of the
ance measure for ANN because it may raise the overfitting issue, parameters and measured properties that guide or train the regres-
where the constructed regression tool reproduces the training sion tool to follow a specific trend. Data-driven machine-learning
data almost perfectly but fails to properly estimate the new data methods are trained by a set of measurements (training samples).
(Hawkins 2004). The constructed regression tool should fit the Therefore, if there is no correlation or dependency between a set
training data properly and at the same time should have the gener- of parameters and the measured outputs, then the machine-learn-
alization power to predict any unknown and new cases. Therefore, ing tool cannot extract any meaningful relationship. Consequently,
to find the optimal ANN, we should split the data to the new train- the designed regression tool will be insensitive to that parameter,
ing and test data; design ANN by use of training data; and per- so it is recommended to remove the independent or less-effective
form cross validation on the test data. To obtain the best ANN parameters from the parameter set. Sensitivity analysis is a proce-
estimator, we tuned the ANN parameters (maximum error and dure that can be used to identify the most-effective parameters.
maximum iteration in the backpropagation-learning algorithm and GPR has been applied in a variety of applications, such as environ-
the number of hidden layers) through leave-one-out cross valida- mental science, geoscience, mining, and remote sensing, where it is
tion (LOOCV). Each time we leave out one aspect of the training also known as Kriging (Journel and Huijbregts 1978). Other fields of
data and design ANN with the remaining data. We repeat this pro- GPR application are hydrology, neuroscience, real-estate appraisal,
cedure N times (N is number of training data). After tuning, the and image processing. In geostatistics, a subfield of geoscience and
best-obtained average relative cross-validation error is 26% and mining, GPR or Kriging is used to populate the rock-property meas-
its corresponding training data error is 13%. Depending on urements (that is, from core at the wells), such as permeability and
LOOCV, the optimum number of hidden-layer neurons is six. porosity, over the whole subsurface resource or reservoir extent.
For simplicity, we plotted the viscosity with respect to only two It should be noted that the underlying physical process of the
input parameters and kept the third one constant. Fig. 4a shows the system (i.e., the governing equations, such as the diffusion equa-
300-rev/min viscosity with respect to fluid density and Ingredient tion in case of fluid flow in porous media) is not incorporated
A content while temperature is constant (80 F). Fig. 4b illustrates directly into the GPR formulation. However, GPR infers the inter-
the 60-rev/min viscosity with respect to temperature and Ingredient relationship of the parameters from the provided training data set.
A content while fluid density is constant (12 lbm/gal). The plotted In addition, GPR as a data-driven method requires a sample data
surface is the nonlinear regression resulting from ANN. With this set in advance. A covariance model that shapes the correlation
nonlinear-regression function, we can predict the viscosity of the between parameters is also needed to be determined beforehand,
new sample without performing extra laboratory experiments. which depends on the physical system under investigation. The
form of the covariance function is usually predetermined, such as
Gaussian Process Regression. GPR is a nonlinear-regression or an exponential, spherical, or Gaussian function (Journel and
interpolation technique that models the new estimated (interpolated) Huijbregts 1978), and its parameters can be inferred from the
values derived from the Gaussian process determined by a covari- given data set in Appendix A.
ance function (Williams and Rasmussen 2006). GPR is a statistical
machine-learning method that is also known as Kriging and is well-
established in geostatistics and computer-science literature (Deutsch Methodology
and Journel 1992; Shi and Choi 2011). To estimate the correspond- In this study, the fluid was weighted up by adding particulates and
ing value for the new input, GPR calculates the weighted average not salt. Because there are three input parameters and plotting
(linear combination) of the known values (training data) depending each of the dial readings with respect to three independent input
on the correlation (or proximity) of the new input and the training parameters is not very informative, for illustration purposes, we
data governed by the covariance function (Rasmussen and Williams plotted the dial reading with respect to only two input parameters
2006). Providing the training data and the prior assumptions (e.g., and kept the third one constant. In Figs. 5 and 6, all the output
Viscosity (cp)
12 15
Viscosity (cp)
15 10
10
10 8
10 8
6
6
5 4
5 4
2 2
0 0
5 400
4 20 300 5
3 4
2 15 200 3
10 2
1 1
0 5 100 0
Ingredient A (gal/bbl) Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(a) (b)
Viscosity (cp)
20
12 12
15 10 15
10
10 8 8
10
6
6
5 4 5
4
2 0
0
5 400
4 20 5
3 300 4
2 15 3
200 2
1 10 1
100 0
Ingredient A (gal/bbl) 0 5 Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(c) (d)
25 25
Viscosity (cp)
20 15 20 15
15 15
10 10
10 10
5 5 5 5
0 0
5 400
4 20
3 300 5
15 4
2 3
1 10 200 2
1
Ingredient A (gal/bbl) 0 5 Fluid Density (lbm/gal) 100 0
Temperature (°F) Ingredient A (gal/bbl)
(e) (f)
60 rev/min at Constant T (°F) 60 rev/min at Constant Fluid Density
35
30 35
40
35 25 40 30
Viscosity (cp)
Viscosity (cp)
30 20 30 25
25
20 15
20 20
15 10
10 10 15
5 5
0 0
400 10
5
4 20 300 5
3 4 5
15 200 3
2 2
1 10 1
100 0
Ingredient A (gal/bbl) 0 5 Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(g) (h)
Fig. 5—Nonlinear estimator resulting from GPR (pink asterisks show the training data).
Viscosity (cp)
Viscosity (cp)
25 25
30 20 30 20
20 15 20 15
10 10
10 10
5 5
0 0
5 400
4 20 300 5
3 4
2 15 200 3
10 2
1 1
100 0
Ingredient A (gal/bbl) 0 5 Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(a) (b)
Viscosity (cp)
Viscosity (cp)
40 30
40 30
30 25
25
30 20
20 20
20 15 15
10 10
10 10
0 0
5 400
4 20 300 5
3 4
2 15 3
200 2
1 10 1
100 0
Ingredient A (gal/bbl) 0 5 Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(c) (d)
70 60 70 50
60 60
Viscosity (cp)
40
Viscosity (cp)
50 50 50
40 40 30
30 40 30
20 20 20
10 30 10
0 10
0
5 400
4 20 20
3 300 5
15 4
2 3
1 10 200 2
10 1
0 5 100 0
Ingredient A (gal/bbl) Fluid Density (lbm/gal) Temperature (°F) Ingredient A (gal/bbl)
(e) (f)
Fig. 6—The predictor function to estimate viscosity resulting from GPR method (pink asterisks show the experimental data
points).
values are plotted vs. two pairs of the input parameters. For In addition, to show the estimation error or confidence interval
instance, Fig. 6e shows the 300-rev/min viscosity with respect to of estimation, in Fig. 8 the nonlinear-regression function or non-
fluid density and Ingredient A (polymer) content while tempera- linear estimator of 60-rev/min viscosity vs. Ingredient A content
ture is constant ( F). The pressure was held constant at 7,500 psi. (fluid density and temperature are constant) is plotted along with
Pressure affects the rheological properties of oil-based fluids more the associated confidence interval of 68%, corresponding to the
than water-based fluids and its effects should not be ignored, espe- estimated value 6 SD.
cially across abnormally pressured zones. The plotted surface is
the nonlinear regression resulting from GPR, and, as shown, it Conclusions
goes through the training data (pink asterisks). With this nonlin- Trial-and-error fluid design is not always a cost-effective option
ear-regression function, we can predict the viscosity of the new and lacks long-term vision to accumulate knowledge on fluid sys-
sample without performing laboratory experiments. Fig. 7 also tems. There is a need for a tool that is empowered by machine-
shows how the predicted viscosity changes vs. Ingredient A at learning methodologies. This paper investigated an intelligent
constant rev/min, temperature, and fluid density. tool that is equipped with flexible machine-learning algorithms.
45 45
40 40
35 35
Viscosity (cp)
Viscosity (cp)
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 1 2 3 4 5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Ingredient A (gal/bbl)
Ingredient A (gal/bbl)
Fig. 8—Nonlinear-regression function resulting from GPR and
Fig. 7—2D plot of viscosity vs. Ingredient A at constant temper-
the corresponding confidence interval (the highlighted enve-
ature (2008F) and fluid density (12 lbm/gal) for a fixed rev/min.
lope around the nonlinear-regression curve shows the estima-
tion error), with temperature of 2008F and fluid density of 12
lbm/gal.
Hawkins, D. M. 2004. The Problem of Overfitting. J. Chem. Inf. Comp. Seminar, Bergen, Norway, 20 April. SPE-180032-MS. https://doi.
Sci. 44 (1): 1–12. https://doi.org/10.1021/ci0342472. org/10.2118/180032-MS.
Journel, A. G. and Huijbregts, C. J. 1978. Mining Geostatistics, London: Tarrahi, M., Jafarpour, B., and Ghassemi, A. 2013. Assimilation of Micro-
Academic Press. seismic Data into Coupled Flow and Geomechanical Reservoir Models
Lee, J., Shadravan, A., and Young, S. 2012. Rheological Properties of with Ensemble Kalman Filter. Presented at the SPE Annual Technical
Invert Emulsion Drilling Fluid Under Extreme HPHT Conditions. Pre- Conference and Exhibition, New Orleans, 30 September–2 October.
sented at the IADC/SPE Drilling Conference and Exhibition, San SPE-166510-MS. https://doi.org/10.2118/166510-MS.
Diego, California, 6–8 March. SPE-151413-MS. https://doi.org/ Tarrahi, M., Jafarpour, B., and Ghassemi, A. 2015. Integration of Micro-
10.2118/151413-MS. seismic Monitoring Data Into Coupled Flow and Geomechanical Mod-
els With Ensemble Kalman Filter. Water Resour. Res. 51 (7):
Mijnarends, R., Frolov, A., Grishko, F. et al. 2015. Advanced Data-Driven
5177–5197. https://doi.org/10.1002/2014WR016264.
Performance Analysis For Mature Waterfloods. Presented at the SPE
Theodoridis, S., Pikrakis, A., Koutroumbas, K. et al. 2010. Introduction to
Annual Technical Conference and Exhibition, Houston, 28–30 Sep-
Pattern Recognition: A Matlab Approach. Burlington, Massachusetts:
tember. SPE-174872-MS. https://doi.org/10.2118/174872-MS.
Academic Press.
Osman, E. A. and Aggour, M. A. 2003. Determination of Drilling Mud
Williams, C. K. and Rasmussen, C. E. 2006. Gaussian Processes for
Density Change with Pressure and Temperature Made Simple and
Machine Learning. Cambridge, Massachusetts: The MIT Press.
Accurate by ANN. Presented at the Middle East Oil Show, Bahrain,
9–12 June. SPE-81422-MS. https://doi.org/10.2118/81422-MS
Rafieisakhaei, M., Barazandeh, B., and Tarrahi, M. 2016a. Analysis of
Supply and Demand Dynamics to Predict Oil Market Trends: A Case Appendix A—GPR
Study of 2015 Price Data. Presented at the SPE/IAEE Hydrocarbon
To perform GPR, we consider one output at a time [i.e., we con-
Economics and Evaluation Symposium, Houston, 17–18 May. SPE-
struct seven GPR models for seven dial-reading outputs (Shi and
179976-MS. https://doi.org/10.2118/179976-MS.
Choi 2011)]. Therefore, the training data for each GPR-model
Rafieisakhaei, M., Barazandeh, B., Moosavi, A. et al. 2016b. Modeling design is
Dynamics of the Carbon Market: A System Dynamics Approach on
the CO2 Emissions and its Connections to the Oil Market. Oral presen-
Training data DfXi ; yi g; i ¼ 1; 2; …; N; . . . . . . . . ðA-1Þ
tation given at the 34rd International Conference of the System Dy-
namics Society, Delft, The Netherlands, 17–21 July.
Rasmussen, C. E. and Williams, C. K. I. 2006. Gaussian Processes for
where yi is a scalar and represents one of the dial-reading outputs.
Machine Learning. Cambridge, Massachusetts: MIT Press.
Each nonlinear-regression function is a random sample (para-
metrized by the input variable) drawn from a joint Gaussian prob-
Schakel, M. D. and Mesdag, P. R. 2014. Fully Data-Driven Quantitative
ability function given the training data set D. Our goal in GPR is
Reservoir Characterization by Broadband Seismic. SEG Technical
to train a function f from data D. A Gaussian process is a set of
Program Expanded Abstracts 2014: 2502–2506. https://doi.org/
random estimation functions, f , each with probability pð f Þ that
10.1190/segam2014-0671.1.
formally can be expressed as Bayesian formulation while a set of
Shadravan, A. and Amani, M. 2012. HPHT 101: What Every Engineer or training samples D is given:
Geoscientist Should Know about High Pressure High Temperature
Wells. Presented at the SPE Kuwait International Petroleum Confer- pð f ÞpðDj f Þ
ence and Exhibition, Kuwait City, Kuwait, 10–12 December. SPE- pð f jDÞ ¼ ; . . . . . . . . . . . . . . . . . . . . . ðA-2Þ
pðDÞ
163376-MS. https://doi.org/10.2118/163376-MS.
Shadravan, A. and Tarrahi, M. 2016. Machine Learning Leads Cost Effec- where pð f jDÞ shows the posterior probability of all the regression
tive Intelligent Fluid Design: Fluid Engineering Perspective. Presented functions given the training data. pð f Þ represents the prior proba-
at the SPE Bergen One Day Seminar, Bergen, Norway, 20 April. SPE- bility of all the possible regression functions and pðDj f Þ is the
180033-MS. https://doi.org/10.2118/180033-MS. likelihood function. In Bayesian inference context, pðDÞ is only a
Shadravan, A., Narvaez, G., Alegria, A. et al. 2015a. Engineering the scaling parameter that does not affect the intended posterior prob-
Mud-Spacer-Cement Rheological Hierarchy Improves Wellbore Integ- ability density function (PDF) pð f jDÞ.
rity. Presented at the SPE E&P Health, Safety, Security and Environ- The nonlinear-regression function in GPR is a set of random
mental Conference-Americas, Denver, 16–18 March. SPE-173534- variables indexed by a continuous parameter (e.g., time, space, or
MS. https://doi.org/10.2118/173534-MS. temperature), which is also called random function f ðXÞ. The
Shadravan, A., Tarrahi, M., and Amani, M. 2015b. Intelligent Cement
assumption in GPR is that any set of regression functions has a
Design: Utilizing Machine Learning Algorithms to Assure Effective
jointly Gaussian distribution with zero mean:
Long-term Well Integrity. Presented at the Carbon Management Tech- k 1 1
f CðXÞ1 f
nology Conference, Sugar Land, Texas, 17–19 November. CMTC- pð f jXÞ ¼ N½0; CðXÞ ¼ ð2pÞ 2 jCðXÞj 2 e 2 ; . . . ðA-3Þ
440236-MS. https://doi.org/10.7122/440236-MS.
Shadravan, A., Tarrahi, M., and Amani, M. 2016. Agile Data-Driven Fluid where CðXÞ is the covariance matrix and k represents the dimen-
Design: Predicting the Properties of Drilling, Spacer and Cement sion of function f with respect to the continuous parameter. Esti-
Slurry Fluids. Oral presentation given at the 2016 AADE Fluids Tech- mating the unknown target value y0 given the set of training y
nical Conference and Exhibition, Houston, 12–13 April. AADE-16- presents a conditional-probability function:
FTCE-07.
Shi, J. Q. and Choi, T. 2011. Gaussian Process Regression Analysis for pðy0 jyÞ ¼ Nðy0 ; ry0 Þ: . . . . . . . . . . . . . . . . . . . . . . ðA-4Þ
Functional Data. Boca Raton, Florida: CRC Press.
Subrahmanya, N., Xu, P., El-Bakry, A. et al. 2014. Advanced Machine The GPR estimation value is the mean of the above condi-
Learning Methods for Production Data Pattern Recognition. Presented tional PDF (Eq. A-4). Basically GPR proposes a Gaussian PDF
at the SPE Intelligent Energy Conference and Exhibition, Ultrecht, for the new estimated value, which can be fully characterized by a
The Netherlands, 1–3 April. SPE-167839-MS. https://doi.org/10.2118/ mean and a SD value.
167839-MS. As a necessary preprocessing step to avoid scaling issues from
Tarrahi, M. and Shadravan, A. 2016a. Inverse Modeling for Fluid System different input and output ranges, we apply data normalization
Characterization through Machine Learning Algorithms. Presented at and scaling to input and output values. Target or output values are
the SPE Bergen One Day Seminar, Bergen, Norway, 20 April. SPE- assumed to be normal standard variables (resulting from a Gaus-
180034-MS. https://doi.org/10.2118/180034-MS. sian distribution with zero mean and unit variance). Moreover,
Tarrahi, M. and Shadravan, A. 2016b. Advanced Big Data Analytics the input values (input parameters) are shifted and scaled to a
Improves HSE Management. Presented at the SPE Bergen One Day [0,1] interval.
X Xmin 2 3
Xs ¼ . . . . . . . . . . . . . . . . . . . . . . . . ðA-5Þ CðX1 ; X1 Þ CðX1 ; XN Þ
Xmax Xmin 6 .. .. .. 7
Cnn ¼4 . . . 5
YY CðXN ; X1 Þ … CðXN ; XN Þ
Ysn ¼ : . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-6Þ
rY ¼ Covariance matrix of training data; . . . . . . . ðA-18Þ
To define the spatial relation of the samples in the input space 2 3
(in this study, a 3D space), a covariance function is established: CðX1 ; X0 Þ
6 .. 7
Cnu ¼6
4 .
7
5
CðXi ; Xj Þ ¼ vexpðhp Þ; . . . . . . . . . . . . . . . . . . . . . ðA-7Þ
CðXN ; X0 Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h ¼ gT L1 g; . . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-8Þ ¼ Cross covariance of training data and new input:
ðA-19Þ
g ¼ RðXi Xj Þ; . . . . . . . . . . . . . . . . . . . . . . . . . . ðA-9Þ
2 3 The estimated output is basically a linear combination of the
l1 0 0 training outputs where the coefficients are determined by the spa-
L¼40 l2 0 5 ¼ Scaling matrix; . . . . . . . . . . . ðA-10Þ tial configuration of the input (with respect to the training data)
0 0 l3 and the covariance-function shape. One of the unique advantages
of GPR compared with ANN is the ability to obtain the estimation
Xi ; Xj ¼ Experiment input values ð3 1Þ; . . . . . . . ðA-11Þ error. GPR not only provides us with the estimated output at the
2 3 new input value, but it is also able to calculate its associated error
1 0 0 or variance. With this capability, we can specify how good each
R1 ðh1 Þ ¼ 4 0 cosðh1 Þ sinðh1 Þ 5; . . . . . . . . . . . ðA-12Þ estimation is or determine the confidence of each estimation and
0 sinðh1 Þ cosðh1 Þ decide if we need to perform new laboratory experiments.
2 3
cosðh2 Þ 0 sinðh2 Þ
Estimation variance ofy0 ¼ v CTnu Cnn Cnu : . . . . . . ðA-20Þ
R2 ðh2 Þ ¼ 4 0 1 0 5; . . . . . . . . . . . . ðA-13Þ
sinðh2 Þ 0 cosðh2 Þ
2 3 This provides us with the estimation SD and consequently the
cosðh3 Þ sinðh3 Þ 0 confidence interval associated with each estimation. It should be
R3 ðh3 Þ ¼ 4 sinðh3 Þ cosðh3 Þ 0 5; . . . . . . . . . . ðA-14Þ noted that GPR is an absolute estimator; that is, estimated values
0 0 1 for the training data are equal to true outputs, and unlike ANN,
the estimation error for training data is zero. After tuning by use
and of LOOCV, the average minimum cross-validation error is 20%.
We obtained different hyper parameters (different covariance
R ¼ R3 ðh3 ÞR2 ðh2 ÞR1 ðh1 Þ; . . . . . . . . . . . . . . . . . ðA-15Þ functions) for different dial-reading outputs. For all the output
dial readings, the obtained power parameter p through cross vali-
where l1 ; l2 , and l3 are correlation lengths in three input directions; dation is 2, so the preferred covariance function is the squared-
the matrix R represents the rotation matrix in 3D, which can be exponential-covariance function.
calculated by use of the rotation angles around different axes
(D’Orangeville and Lasenby 2003); and v represents the variance
parameter. In specific cases, where the parameter p ¼ 2, the result- Arash Shadravan is a director at ReservoirFocus in Houston. He
ing covariance is the squared-exponential-covariance function (or manages various projects related to enhancing the econom-
Gaussian), and if p ¼ 1, the exponential-covariance function is ics of upstream operations in the US. Shadravan has previously
reproduced. The parameters involved in defining the covariance worked for Schlumberger, Baker Hughes, Occidental Petro-
function are called hyper parameters. By changing the hyper pa- leum, and Superior Energy Services. He has been a journal
reviewer and technical contributor as a committee member
rameters, the GPR estimator quality will change. In this study, we
in SPE conferences and workshops. Shadravan has published
tune seven hyper parameters (three correlation lengths, three rota- more than 30 peer-reviewed journal and technical confer-
tion angles, and the power) through LOOCV to obtain the optimal ence papers related to the application of machine learning in
GPR estimator. In LOOCV, we assume one of the training data production optimization, cementing, drilling fluids, and under-
points is unknown; use the other eight training data points to balanced drilling. He has been a SPE member since 2006.
design the GPR estimator; estimate the unknown data point; and Shadravan holds a master’s degree in petroleum engineering
then compare the estimated value with the true value. This proce- from Texas A&M University and a bachelor’s degree in petro-
dure is repeated for all the outputs (seven dial readings) and for leum engineering.
all the training data (nine samples). In this cross-validation proce- Mohammadali Tarrahi is a senior reservoir engineer in the Subsur-
dure, the performance measure of the GPR estimator is the aver- face Modeling and Optimization Department at Shell Global
age relative error. Solutions in Houston. He has been part of several research
The following is the relation between the semivariogram (pop- and development groups at Shell, Occidental Petroleum Corpo-
ular spatial-correlation representation in geostatistics) and the co- ration, and Microseismic. Tarrahi is the author of more than
variance function: 30 peer-reviewed-journal and technical-conference papers
regarding numerical reservoir simulation, geostatistics, reservoir
characterization, uncertainty assessment, big-data analytics,
cðXi ; Xj Þ ¼ v CðXi ; Xj Þ: . . . . . . . . . . . . . . . . . . . ðA-16Þ and statistical machine learning. He holds a PhD degree in petro-
leum engineering from Texas A&M University, a master’s degree
To estimate the corresponding output y0 of the new input pa- in petroleum engineering, and bachelor’s degrees in electrical
rameter X0 , we use the Gaussian process interpolation (also called engineering and petroleum engineering.
simple Kriging):
Mahmood Amani is an associate professor of petroleum engi-
2 3T neering at Texas A&M University at Qatar (TAMUQ), and has
y1 been on the faculty of Texas A&M University for more than 13
6 .. 7 1 XN
y0 ¼ 4 . 5 Cnn Cnu ¼ ki yi ; . . . . . . . . . . . . . . ðA-17Þ years. He moved from the main campus of Texas A&M Univer-
i¼1 sity in College Station, Texas, to Qatar to help start the Petro-
yN leum Engineering Program at TAMUQ. Amani is the founding
faculty member and the first program coordinator and 2016 SPE Middle East Regional Distinguished Achievement
chair of the Petroleum Engineering Program at TAMUQ. He Award for Petroleum Engineering Faculty as well as the 2013
considers his role in establishing the petroleum-engineering SPE Middle East Regional Service Award, and he holds two US
program in Qatar as his most-remarkable professional patents. Amani holds a bachelor’s degree in mechanical en-
achievement. Before joining academia as a faculty member gineering from Wichita State University, a master’s degree in
focused on petroleum engineering, Amani worked as a natural-gas engineering from Texas A&M University-Kingsville,
research scientist with the Texaco Exploration and Production and a PhD degree in petroleum engineering from Texas A&M
Technology Department in Houston. He was awarded the University.