Design calculation model of oil fluid property parameter knowledge database for production logging

Abstract

The fluid property parameter calculation affects the accuracy of the interpretation the accuracy, in the interpretation of the liquid production profile. Therefore, it is particularly important to accurately calculate the physical property parameter values, in the establishment of the fluid property parameter expert knowledge base system. The main physical parameters include the following calculation methods of the oil. The oil property parameter conversion formula mainly studies the formulas such as bubble point pressure, dissolved gas-oil ratio, crude oil volume coefficient, crude oil density, crude oil viscosity, and crude oil compression coefficient. Design expert knowledge base system, it is based on the calculation methods of these physical parameters. A computational fluid property parameter model is constructed by training production log sample data. Finally, the interactive and friendly product interpretation software model was developed in 9 wells’ data. The design calculation model can increase the accuracy to achieve 95% of oil fluid property parameter. Accurately calculate fluid property parameter values.

Keywords

Production logging oil fluid property parameter calculation model

With the development of computer technology, particularly the rise of software engineering, production logging data interpretation by hand before reading, paragraph-by-paragraph comparison phase to use computer software to intelligent recognition, to increase accuracy and efficiency [1]. In Interpretation, the accurate calculation of fluid physical properties of oil is very important, so an expert knowledge base system software engineering formed a set of physical parameters of expert knowledge base oil fluids computer software modules to accurately calculate fluid physical parameters.

1. Introduction

The subject of production logging originated from Yangtze University [2]. It is a key discipline of Yangtze University. The production logging interpretation method is the advantage of Yangtze University [3]. Traditional production logging interpretation methods are mainly based on experience. It also uses the parameters of logging tools to predict the physical parameters of the production logging oil [4]. Now, it is widely used to calculate the physical parameters of crude oil by hand. The accuracy is not high of these methods. We can improve the interpretation accuracy of production logging to study a kind of calculation model for accurate parameters of fluid physical properties.

Expert knowledge base [5, 6] not only has the human brain’s capacity for logical reasoning, but also the ability to accurately calculate the projection results, along with the ability to solve practical problems, the core of which is the establishment of knowledge base and the design of inference machine [7]. Knowledge Base is a collection of store operations knowledge, which contains data, formula and algorithm of the calculation model and associated experience models. Inference machine [8], the human brain thinking process of logical reasoning [9], will object to applying relevant knowledge for effective inference to achieve the objective to solve practical problems [10]. The figure is the expert knowledge database system [11].

Figure 1.

Fluid properties expert knowledge base overall design framework.

First entering in the software modules expertise, obtaining expert knowledge through a computer program, establishing the appropriate algorithm, calculation models and formulas, then can form expert knowledge base [12]. Users enter data input ports, by computer analysis of the data and the corresponding processing, and then entered in the inference machine [13]. The logic algorithms of inference machine closely are associated with expert knowledge in logic [1, 4]. Ultimately data and the algorithm model will be input to integrated processing platform base expert knowledge base, and then the result computing by the platform will be output. Parameter conversion formula is not only an important model for expert knowledge base, but also is the base of constructing out of expert knowledge base and inference machine [15].

2. Formation crude oil physical property parameter calculation method

Calculation of formation crude oil physical parameter [16], the first bubble point pressure of oil, followed by calculation of dissolved gas-oil ratio and compressibility of crude oil formation volume factor [17], and calculation of oil viscosity and density [1, 8].

The main influence factors of Bubble point pressure are oil, gas composition and the temperature of formation.

The Eq. (1) scope of application, T:100 $\sim$ 258, and Pb $=$ 130 $\sim$ 7000 psi, and T $=$ 100 $\sim$ 258 f, and RG $=$ 0.59 $\sim$ 0.95, and API $=$ 16.5 $\sim$ 63.8 API (API crude oil density units), $R_{p}$ in ft3/bbl (as below).

The Eq. (2) is how to calculate the bubble point $P_{b}$ value.

$\displaystyle\alpha=\gamma_{g}\left(1+0.5912\times\textit{API}\times 59\times% \text{lg}\left(0.101\times\frac{145.03}{114.7}\right)\ \times 0.0001\right)$ (1) $\displaystyle P_{b}=\left(\frac{R_{p}\times e^{\frac{-1\times C_{3}\times% \textit{API}}{T+460}}}{C_{1}\times\alpha}\right)$ (2)

The Eq. (1) is as $\textit{API}\leqslant$ 30, C ${}_{1}=$ 0.0362, C ${}_{2}=$ 1.0937, $C_{3}=$ 25.724. In the case of API $>$ 30, $C_{1}=$ 0.0178, $C_{2}=$ 1.187, $C_{3}=$ 23.931.

The dissolved gas oil ratio ( $R_{s}$ ) is affected by formation environment of oil, gas properties, composition, temperature and the bubble point pressure. Under normal circumstances, the density of crude oil is low, the content of dissolved gas is higher, the ratio is small. In some cases, it is also needed to calculate the ratio of dissolved gas to oil ratio, and the ratio of gas to oil is the reciprocal of the value. The magnitude of the value can affect the density, volume factor and viscosity of the calculated crude oil. The Eq. (3) is how to calculate the $R_{s}$ value.

$\displaystyle R_{s}={\gamma}_{g}\frac{{\left[\left(\frac{P\times 145.03}{18.2}% +1.4\right)\times{10}^{0.0125\times\textit{API}-0.00091\times T}\right]}^{1.20% 48}}{5.6183}$ (3)

In the Eq. (3), the unit of $P$ is MPa, the unit of $T$ is $F$ , the unit of $R_{s}$ is m ${}^{3}$ /m ${}^{3}$ (the same below). In the case of Eq. (3), the value of $P$ uses formation pressure parameter under the condition of $P\leqslant P_{b}$ and is the bubble point pressure in other cases. Crude oil compressibility coefficient ( $C_{0}$ ), The Eq. (4) is how to calculate the $C_{0}$ value.

$\displaystyle C_{0}=\frac{-1433+5\times\left(R_{s}\times 5.6183\right)+17.2% \times T-1180\times\alpha+12.61\times\textit{API}}{P\times 145.03\times 100000}$ (4)

There are some classical formulas about oil formation volume factor ( $B_{0}$ ). Equation (5) is the calculation formula.

$\displaystyle B_{o}=0.972+1.47\times 10^{-4}\left[R_{s}\left(\gamma_{g}/\gamma% _{o}\right)^{0.5}+1.25T\right]^{1.175}$ (5)

The density of crude oil ( $P_{0}$ , g/cm ${}^{3}$ ) and the ratio of dissolved gas oil ( $R_{s}$ ) are closely associated with formation volume factor of crude oil ( $B_{0}$ ). $P_{0}$ can be determined by the definition and be calculated by the experiential formula.

$\displaystyle\rho_{o}=(\gamma_{o}+0.0002178\gamma_{g}R_{s})/B_{o}$ (6)

In general, the formula for calculating the density of crude oil is shown in Eq. (7).

$\displaystyle\rho_{o}=\frac{62.4\gamma_{o}+0.0136R_{s}\gamma_{g}}{B_{o}}$ (7)

In the case of $P<P_{b}$ , you can directly use the Eq. (7) calculate the oil density. In the case of $P\geqslant P_{b}$ , crude oil density can be calculated and based on the Eq. (7), then can be calculated by the following Eq. (8).

$\displaystyle{\rho}_{o}={\rho}_{o}\times e^{C_{0}\times(P-P_{b})\times 145.03}$ (8)

In the formula, the $C_{0}$ (CP) is oil compressibility, the $P$ (MPa) is formation pressure, the $P_{b}$ (MPa) is bubble pressure.

Beggs and other scholars put forward a more classic calculation of crude oil viscosity ( ${\mu}_{0}$ ) Eqs (7) and (8). When $P>P_{b}$ .

$\displaystyle\mu_{0}=\mu_{od}(P/P_{b})^{m}$ (9) $\displaystyle m=2.6P^{1.187}10^{-0.039P\times 10^{-3}-5}$

Figure 2.

Crude oil fluid properties inference engine design diagram.

Figure 3.

Oil viscosity inference engine design diagram.

When $P\leqslant P_{b}$ .

$\displaystyle\mu_{o}=10.715(R_{s}+100)^{-0.515}\mu_{od}^{B}$ $\displaystyle B=5.44(R_{s}+150)^{-0.338}$ $\displaystyle\mu_{od}=10^{x}-1$ (10) $\displaystyle x=10^{(3.0324-0.02023\textit{API})}T^{-1.163}$

In the formula, $\mu_{od}$ shows the value of the viscosity of crude oil under bubble point pressure.

3. The design of expert knowledge base system

3.1 The design of knowledge base inference engine

The wellhead metering data – dissolved gas oil ratio, ground oil density, ground gas density (the values for relative density values), formation salinity [19], temperature standard and standard pressure. It is as a known parameter. When calculating, first calculate the bubble point pressure, then turn computing solution gas-oil ratio, compressibility, formation volume factor, viscosity and density. Per terms of the relationship between these parameters, design expert knowledge inference junction configuration [20] shown in Fig. 2 to Fig. 5.

Shown in Fig. 2, in the calculation of fluid properties of crude oil, the inference engine obtains dissolved gas oil ratio [21], ground oil density, surface gas density (the value of the relative density value), formation water salinity, fluid pressure, fluid temperature, the standard temperature and standard pressure in the knowledge base input, combined with the relevant formulas to calculate the bubble point pressure, oil compressibility, oil volume factor, oil density [22].

Figure 4.

Expert knowledge model detailed design diagram.

Figure 5.

Calculate crude oil volume factor window.

When the bubble point pressure is less than formation pressure, crude oil viscosity can be calculated directly by the equation. When the bubble point pressure of not less than the formation pressure, it also requires a combination of dissolved oil/gas ratio parameters to calculate the values, as shown in Fig. 3.

3.2 Knowledge base detailed design

Expert knowledge base can calculate precisely the value of physical parameters of oil in the formation through physical parameters of oil on the ground. Expert Knowledge model detailed design shown in Fig. 4.

Table 1
Oil PVT comparative analysis of the results table values

Well	Bo			Oil density (g/cc)			Oil viscosity (cp)			$R_{s}$ (m ${}^{3}$ /m ${}^{3}$ )			$P_{b}$ (MPa)
name	Real data	Calculation data	Relative error (%)	Real data	Calculation data	Relative error (%)	Real data	Calculation data	Relative error (%)	Real data	Calculation data	Relative error (%)	Real data	Calculation data	Relative error (%)
w1	1.050	1.053	0.28	0.920	0.916	0.48	27.380	22.370	18.30	7.484	7.373	1.49	9.700	9.880	1.85
w2	1.040	1.050	0.92	0.921	0.920	0.01	17.900	18.050	0.85	3.400	3.410	0.29	1.560	1.556	0.29
w3	1.040	1.050	0.82	0.925	0.923	0.18	19.210	19.230	0.10	2.600	2.610	0.34	1.230	1.226	0.32
w4	1.070	1.0703	0.03	0.905	0.907	0.19	11.050	22.110	0.40	11.500	11.510	0.06	4.590	4.580	0.16
w5	1.060	1.059	0.02	0.916	0.918	0.18	22.070	22.110	0.19	14.000	14.150	1.11	5.140	5.130	0.11
w6	1.050	1.051	0.07	0.910	0.920	0.13	21.600	21.640	0.16	8.000	8.050	0.56	3.113	3.110	0.11
w7	1.060	1.063	0.30	0.910	0.920	0.27	17.830	17.900	0.41	14.000	14.330	2.38	5.230	5.220	0.11
w8	1.060	1.070	0.71	0.857	0.859	0.29	4.630	4.610	0.49	14.300	14.590	2.05	4.070	4.069	0.02
w9	1.320	1.330	0.36	0.767	0.770	0.42	0.989	0.990	0.13	100.000	99.790	0.22	28.860	28.830	0.12

First, it can design the related Rules relational tables. And then it designs the Master data rule tables by combining the Model Rules of relational tables with the Rules of relational tables and Dependent rules tables by combining the Data condition tables with the Rules of relational tables. It derives the Model table for calculating the properties by the Model rules of relational Tables. The contents of each table are shown in Fig. 4. The finally actualize the software modules of fluid property expert knowledge base by computer software algorithms. In Fig. 5, there is the windows for calculating the volume factor of crude oil, calculating the bubble point pressure, calculating crude oil compressibility, calculating crude oil volume factor, calculating crude oil density and so on. Similarly, window does not enumerate.

This software system is based on C++6.0 platform by Visual C++. It can accurately calculate the parameter values of the crude oil formation fluid properties through the software module of expert knowledge base.

4. Conclusion

It had verified the several dozen production logging of formation oil physical parameters through the software modules of fluid property expert knowledge base that I designed. There are 9 wells data, the results of which are shown in Table 1 as follows. The real data is provided by oil field in the Table 1. The calculation data is calculated by the design calculation model in the Table 1.

The fluid physical parameters of formation crude oil can increase an accuracy of more than 95% by using the expert knowledge base, which plays an extremely important role in improving the accuracy of production logging interpretation, analyzing the production capacity of each layer, calculating the yield of each oil layer.

The innovation of the paper is the research on the calculation method of physical property parameters of formation crude oil by using expert knowledge base, which can be improved the accuracy of calculation and lay the foundation for improving the accuracy of production logging interpretation.

Footnotes

Acknowledgments

This work was supported by Hubei Science and Technology Demonstration Project, Oilfield Data Intelligent Analysis and Research Center (2019ZYYD016). Supported by 2016 Knowledge Innovation Special Foundation of Hubei Province (2016CFC738).

References

Frisch

Perkins

and Quirein

, Integra-ting wellbore flow images with a conventional production log interpretation method, SPE 77782, 2002.

Song

Guo

and Shi

, Levenberg-marquardt algorithm used in reservoir damage production of logging prediction, Proceedings of the 2011 2nd International Conference on Information Technology and Scientific Management, 2011, 6, pp. 62–65.

Song

Guo

and Wang

, The BP Algorithm Used in Reservoir Damage Predictionof Speed-Sensitive by Improved, Proceedings of the 2011 2nd International Conference on Information Technology and Scientific Management, 2011, 6, pp. 58–61.

Jie

Ning

Chenghong

et al., Application of seismic data to the design of horizontal well trajectory in tight sandstone reservoirs, Oil & Gas Geology 33(3) (2012), 909–913.

Zhai

Jin

and Zheng

, The analysis and modeling of measuring data acquired by using combination production logging tool in horizontal simulation well, Chinese Journal of Geophysics 55(4) (2012), 1411–1421.

Tang

, Application of geosteering technology in the developent of Sulige gas field – case studies of the Su10 and Su53 blocks, Oil & Gas Geology 4(3) (2013), 388–393.

Piron

Boizard

Heddadj

Pérès

Hallaire

Cluzeau

, Indicators of earth worm bioturbation to improve visual assessment of soil structure, Soil & Tillage Research (2016), 78–92.

Huang

Hou

Kwok

Sun

Zhao

Lee

and Zhang

, The effects of probiotics administration on the milk production, milk components and fecal bacteria microbiota of dairy cows, Science Bulletin 62(11) (2017), 767–774.

Sun

Liang

Jin

and Shan

, Experimental and modeling study of the regular polygon angle-spiral liner in ball mills, Chinese Journal of Mechanical Engineering 30(2) (2017), 363–372.

10.

Tian

Zhu

Song

and Qian

, A method of production profile quick prediction based on typical curves: A case study of the Upper Shale large multi-layered sandstone reservoir, Rumaila Oilfield, Iraq, Petroleum Exploration and Development Online 43(6) (2016), 189–201.

11.

Patel

Arora

Mehtani

Pruthi

and Pruthi

P.A.

, Assessment of fuel properties on the basis of fatty acid profiles of oleaginous yeast for potential biodiesel production, Renewable and Sustainable Energy Reviews 77 (2017), 75–86.

12.

Barton

and Cai

, Equatorial Wave Expansion of Instantaneous Flows for Diagnosis of Equatorial Waves from Data: Formulation and Illustration, Advances in Atmospheric Sciences 34(10) (2017), 1219–1234.

13.

Liu

Wang

Deng

Zhang

and Liu

, Optimal spacing of sequential and simultaneous fracturing in horizontal well, Journal of Natural Gas Science and Engineering 29 (2016), 85–96.

14.

Liang

and Zhou

, A pilot-scale study of gas production from hydrate deposits with two-spot horizontal well system, Applied Energy 176 (2016), 48–57.

15.

Cao

Luo

Yuan

Zhang

and Tan

, Horizontal well transient rate decline analysis in low permeability gas reservoirs employing an orthogonal transformation method, Journal of Natural Gas Science and Engineering (2016), 59–67.

16.

Hou

Wei

Wang

and Zhang

, Production prediction of cyclic multi-thermal fluid stimulation in a horizontal well, Journal of Petroleum Science and Engineering (2016), 43–52.

17.

Wang

Gao

and Liu

, Multi-objective sidetracking horizontal well trajectory optimization in cluster wells based on DS algorithm, Journal of Petroleum Science and Engineering (2016), 85–99.

18.

Wei

Jia

and Wang

, A new method in predicting productivity of multi-stage fractured horizontal well in tight gas reservoirs, Journal of Natural Gas Geoscience (2016), 158–169.

19.

Cheng

Gao

Jin

Chen

and Jiang

, A study to assess the stress interaction of propped hydraulic fracture on the geometry of sequential fractures in a horizontal well, Journal of Natural Gas Science and Engineering (2016) 235–247.

20.

Cui

Ren

Zhang

Ezekiel

Enechukwu

Wang

and Zhang

, Geothermal exploitation from hot dry rocks via recycling heat transmission fluid in a horizontal well, Energy 128 (2017), 156–162.

21.

Cao

Luo

Zhang

and Tan

, Characteristics of transient production rate performance of horizontal well in fractured tight gas reservoirs with stress-sensitivity effect, Journal of Petroleum Science and Engineering (2017), 178–186.

22.

Cao

Liu

and Hu

, Multiple parameters? estimation in horizontal well logging using a conductance-probe array, Flow Measurement and Instrumentation 40 (2014), 152–160.

Design calculation model of oil fluid property parameter knowledge database for production logging

Abstract

Keywords

1. Introduction

3.1 The design of knowledge base inference engine

Table 1 Oil PVT comparative analysis of the results table values

Footnotes

Acknowledgments

References

Table 1
Oil PVT comparative analysis of the results table values