Abstract
Precise characterization of underground reservoirs requires accurate calculations of the reservoir’s petrophysical data and accurate selection of the mathematical model governing the reservoir’s dynamic. In this study, we develop a novel heuristic-clustering algorithm, namely GA–DBSCAN–KMEANS, that can be applied over pressure transient data to assess the true reservoir model out of a pool of candidates. In this algorithm, each specific reservoir model is considered a subpopulation in the GA (genetic algorithm). Then, the simultaneous optimization of all the reservoir models is sought using the proposed hybrid algorithm. During the optimization process, the population size of different models will be either decreased, increased, or unchanged based on the average quality match obtained for each model. A combined DBSCAN (density-based spatial clustering of applications with noise)–KMEANS clustering scheme is used to increase the population size for the best reservoir model in each iteration of the GA. The accuracy of the proposed algorithm was verified using several synthetic data and a real field case obtained from the open literature. The tested data were collected from different types of reservoir models, including homogeneous reservoirs, matrix-fracture dual-porosity reservoirs, and fault-limited reservoirs. For uncertainty analyses and to test the performance of the algorithm under large numbers of initializations, Monte Carlo simulations were conducted. Results of the Monte Carlo simulations unveiled high values of P10, P50, and P90 for the probability of the true reservoir model and low values of these statistics for the false reservoir models. This shows that the outcome of the proposed algorithm is not affected by the initial randomization of the solution subspaces; hence, the developed algorithm is a reliable tool in determining the most probable reservoir model from transient well testing data.
Similar content being viewed by others
Abbreviations
- λ :
-
Interporosity flow term at fractured reservoirs
- ω :
-
Matrix storativity ratio at fractured reservoirs
- μ :
-
Viscosity, cp
- c t :
-
Total compressibility, psi−1
- Ф :
-
Porosity, fraction
- h :
-
Reservoir thickness, ft
- r w :
-
Wellbore radius, ft
- q :
-
Oil flow rate, STBD
- B :
-
Oil formation volume factor, Rbbl/STB
- C :
-
Wellbore storage coefficient, bbl/psi
- k :
-
Reservoir permeability, md
- t p :
-
Produced time, h
- \( t_{{{\text{eq}}.}} \) :
-
Equivalent time at the buildup test, h
- \( \Delta t \) :
-
Well test duration, h
- N :
-
Number of chromosomes at the GA
- N r :
-
Number of realizations at the Monte Carlo simulation
- ΔQ :
-
Difference at the quantiles of the CDF plot at two subsequent Nr values
- ε :
-
Convergence criteria for the CDF plot at the Monte Carlo simulations
- r f :
-
Distance to the reservoir boundary in boundary limited reservoirs
- d f :
-
Distance to the fault
- \( K_{0} \) :
-
Modified Bessel function of the second kind with zero degree
- \( K_{1} \) :
-
Modified Bessel function of the second kind with first degree
- \( I_{0} \) :
-
Modified Bessel function of the first kind with zero degree
- \( I_{1} \) :
-
Modified Bessel function of the second kind with first degree
- \( r_{\text{D}} \) :
-
Dimensionless radius at wellbore
- \( R_{\text{eD}} \) :
-
Dimensionless radius at the reservoir boundary
- \( \bar{P}_{\text{D}} \) :
-
Dimensionless wellbore pressure in Laplace medium
- \( \bar{P}_{\text{wD}} \) :
-
Dimensionless wellbore pressure in Laplace medium after considering skin factor and wellbore storage
- \( C_{\text{D}} \) :
-
Dimensionless wellbore storage coefficient
- \( d_{\text{D}} \) :
-
Dimensionless distance from well to the fault
- S :
-
Skin factor
- x :
-
Parent in the GA
- y :
-
Offspring in the GA
- \( v_{n} \) :
-
nth variable in GA encoding
- \( {\text{gene}}\left[ m \right] \) :
-
Binary version of the mth variable in GA encoding
- GA:
-
Genetic algorithm
- DBSCAN:
-
Density-based spatial clustering with applications of noise
- IPF:
-
Interporosity flow
- PSS:
-
Pseudo-steady state
- CDF:
-
Cumulative probability density function
References
Aboosadi ZA, Rooeentan S, Adibifard M (2018) Estimation of subsurface petrophysical properties using different stochastic algorithms in nonlinear regression analysis of pressure transients. J Appl Geophys 154:93–107
Adibifard M, Tabatabaei-Nejad S, Khodapanah E (2014) Artificial neural network (ANN) to estimate reservoir parameters in naturally fractured reservoirs using well test data. J Pet Sci Eng 122:585–594
Adibifard M, Bashiri G, Roayaei E, Emad MA (2016) Using particle swarm optimization (PSO) algorithm in nonlinear regression well test analysis and its comparison with Levenberg-Marquardt algorithm. Int J Appl Metaheuristic Comput IJAMC 7:1–23
Alajmi MN, Ertekin T (2007) The development of an artificial neural network as a pressure transient analysis tool for applications in double-porosity reservoirs. In: Asia Pacific oil and gas conference and exhibition. Society of Petroleum Engineers
Al-Kaabi A, Lee W (1990) An artificial neural network approach to identify the well test interpretation model: applications. In: SPE annual technical conference and exhibition. Society of Petroleum Engineers
Allain O, Houze O (1992) A practical artificial intelligence application in well test interpretation. In: European petroleum computer conference. Society of Petroleum Engineers
Athichanagorn S, Horne RN (1995) Automatic parameter estimation from well test data using artificial neural network. In: SPE annual technical conference and exhibition. Society of Petroleum Engineers
Awotunde AA (2015) Estimation of well test parameters using global optimization techniques. J Pet Sci Eng 125:269–277
Bazargan H, Adibifard M (2017) A stochastic well-test analysis on transient pressure data using iterative ensemble Kalman filter. Neural Comput Appl 31:1–17
Bourdet D, Whittle T, Douglas A, Pirard Y (1983) A new set of type curves simplifies well test analysis. World Oil 196:95–106
Bourdet D, Ayoub J, Pirard Y (1989) Use of pressure derivative in well test interpretation. SPE Form Eval 4:293–302
Deep K, Thakur M (2007) A new crossover operator for real coded genetic algorithms. Appl Math Comput 188:895–911
Emerick AA, Silva E, Messer B, Almeida LF, Szwarcman D, Pacheco MAC, Vellasco MMBR (2009) Well placement optimization using a genetic algorithm with nonlinear constraints. In: SPE reservoir simulation symposium. Society of Petroleum Engineers
Ershaghi I, Li X, Hassibi M, Shikari Y (1993) A robust neural network model for pattern recognition of pressure transient test data. In: SPE annual technical conference and exhibition. Society of Petroleum Engineers
Ester M, Kriegel H-P, Sander J, Xu X (1996) A density-based algorithm for discovering clusters in large spatial databases with noise. In: KDD, pp 226–231
Garg A, Garg A, Tai KJCG (2014) A multi-gene genetic programming model for estimating stress-dependent soil water retention curves. Comput Geosci 18:45–56
Guyaguler B, Horne RN (2001) Uncertainty assessment of well placement optimization. In: SPE annual technical conference and exhibition. Society of Petroleum Engineers
Guyaguler B, Horne RN, Tauzin E (2001) Automated reservoir model selection in well test interpretation. In: SPE annual technical conference and exhibition. Society of Petroleum Engineers
Han Y, Park C, Kang JM (2011) Prediction of nonlinear production performance in waterflooding project using a multi-objective evolutionary algorithm. Energy Explor Exploit 29:129–142
Haupt RL, Ellen Haupt S (2004) Practical genetic algorithms. Wiley, Hoboken
Holland JH (1992) Genetic algorithms. Sci Am 267:66–72
Karkevandi-Talkhooncheh A, Hajirezaie S, Hemmati-Sarapardeh A, Husein MM, Karan K, Sharifi M (2017) Application of adaptive neuro fuzzy interface system optimized with evolutionary algorithms for modeling CO2-crude oil minimum miscibility pressure. Fuel 205:34–45
Kharrat R, Razavi S (2008) Determination of reservoir model from well test data, using an artificial neural network. Scientia Iranica 15:487–493
Lee S, Min B, Wheeler MFJCG (2018) Optimal design of hydraulic fracturing in porous media using the phase field fracture model coupled with genetic algorithm. Comput Geosci 22:833–849
Li H, Sanchez R, Qin SJ, Kavak HI, Webster IA, Tsotsis TT, Sahimi M (2011) Computer simulation of gas generation and transport in landfills. V: use of artificial neural network and the genetic algorithm for short-and long-term forecasting and planning. Chem Eng Sci 66:2646–2659
Mitchell M (1995) Genetic algorithms: an overview. Complexity 1:31–39
Park H-Y, Datta-Gupta A, King MJ (2013) Handling conflicting multiple objectives using Pareto-based evolutionary algorithm for history matching of reservoir performance. In: SPE reservoir simulation symposium. Society of Petroleum Engineers
Rahmah N, Sitanggang IS (2016) Determination of optimal epsilon (eps) value on dbscan algorithm to clustering data on peatland hotspots in sumatra. In: IOP conference series: earth and environmental science. IOP Publishing, p 012012
Rahman M, Rahman M, Rahman S (2001) An integrated model for multiobjective design optimization of hydraulic fracturing. J Pet Sci Eng 31:41–62
Romero C, Carter J (2001) Using genetic algorithms for reservoir characterisation. J Pet Sci Eng 31:113–123
Romero C, Carter J, Gringarten A, Zimmerman R (2000) A modified genetic algorithm for reservoir characterisation. In: International oil and gas conference and exhibition in China. Society of Petroleum Engineers
Sastry K, Goldberg D, Kendall G (2005) Genetic algorithms. In: Search methodologies. Springer, pp 97–125
Schulze-Riegert R, Axmann J, Haase O, Rian D, You Y-L (2001) Optimization methods for history matching of complex reservoirs. In: SPE reservoir simulation symposium. Society of Petroleum Engineers
Schulze-Riegert R, Axmann J, Haase O, Rian D, You Y-L (2002) Evolutionary algorithms applied to history matching of complex reservoirs. SPE Reserv Eval Eng 5:163–173
Srinivas M, Patnaik LM (1994) Genetic algorithms: a survey. Computer 27:17–26
Stehfest H (1970) Algorithm 368: numerical inversion of Laplace transforms [D5]. Commun ACM 13:47–49
Uyar S, Eryigit G, Sariel S (2004) An adaptive mutation scheme in genetic algorithms for fastening the convergence to the optimum. In: Proceedings of 3rd APIS: Asian Pacific international symposium on information technologies
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
There is no conflict of interest.
Additional information
Communicated by V. Loia.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A: Laplace solution of the reservoir models
Appendix A: Laplace solution of the reservoir models
In this section, we provide the mathematical formula for the pressure solution of the reservoir models used in this study. The solutions can be found in the Laplace domain in (Ershaghi et al. 1993), and we have used the numerical Stehfest Laplace inverse algorithm to generate the solutions in the time domain. For all the reservoir models, \( \bar{P}_{\text{wD}} \) is defined as follows after considering skin and wellbore storage into the wellbore pressure model:
In the above equation, s is the Laplace variable, S is the skin factor, and \( C_{\text{D}} \) is the dimensionless wellbore storage coefficient. In addition, \( \bar{P}_{\text{D}} \) is the pressure response without skin and wellbore storage and is different for each reservoir model. In the following, we have provided the different formulations used for \( \bar{P}_{\text{D}} \) for each reservoir model. In all the following equations, \( r_{\text{D}} \) is the dimensionless radius that we are investigating for the pressure. Since we are interested in wellbore pressure, \( r_{\text{D}} \) will be equal to 1.0 in all the equations.
1.1 Model 1: Infinite acting homogeneous reservoir (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree; s is the Laplace parameter.
1.2 Model 2: Infinite acting dual-porosity reservoir with PSS interporosity flow (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree; s is the Laplace parameter, and \( r_{\text{D}} \) is the dimensionless radius. \( \lambda \) and \( \omega \) are interporosity flow term and matrix storativity ratio, respectively.
1.3 Model 3: Infinite acting homogeneous reservoir with linear fault (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree; s is the Laplace parameter; \( d_{\text{D}} \) is the dimensionless distance from well to the fault; and \( r_{\text{D}} \) is the dimensionless radius.
1.4 Model 4: Infinite acting dual-porosity reservoir (PSS interporosity flow) with linear fault (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree; s is the Laplace parameter; \( d_{\text{D}} \) is the dimensionless distance from well to the fault; and \( r_{\text{D}} \) is the dimensionless radius. The function \( f\left( s \right) \) is the same as defined before for model 2.
1.5 Model 5: Infinite acting homogeneous reservoir with angular 45° fault (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree; s is the Laplace parameter. \( d_{{{\text{D}}i}} \) is the dimensionless distance defined to the ith imaginary well, and \( r_{\text{D}} \) is the dimensionless radius. \( C_{\text{D}} \) is the dimensionless wellbore storage coefficient.
1.6 Model 6: Infinite acting dual-porosity reservoir (PSS interporosity flow) with angular 45° fault (Ershaghi et al. 1993)
All the parameters are the same as model 5, except for \( f\left( s \right) \) which is the same defined before for model 2.
1.7 Model 7: Finite acting homogeneous reservoir (no flow boundary) (Ershaghi et al. 1993)
where \( K_{0} \) is the modified Bessel function of the second kind with zero degree and \( K_{1} \) is the modified Bessel function of the second kind with the first degree. Similarly, \( I_{0} \) is the modified Bessel function of the first kind with zero degree and \( I_{1} \) is the modified Bessel function of the second kind with first degree. Also, s is the Laplace parameter, \( r_{\text{D}} \) is the dimensionless radius, and \( R_{\text{eD}} \) is the dimensionless radius at the reservoir boundary.
1.8 Model 8: Finite acting dual-porosity reservoir with PSS interporosity flow (Ershaghi et al. 1993)
All parameters are the same defined in model 7, except for the \( f\left( s \right) \) function which is the same defined before for mode 2.
Rights and permissions
About this article
Cite this article
Adibifard, M., Sheidaie, A. & Sharifi, M. An intelligent heuristic-clustering algorithm to determine the most probable reservoir model from pressure–time series in underground reservoirs. Soft Comput 24, 15773–15794 (2020). https://doi.org/10.1007/s00500-020-04908-6
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-020-04908-6