Analytical and numerical study on radius of investigation in composite reservoirs Authors: Haijun Fan Published: 2024 ABSTRACT AbstractRadius of investigation (ROI) is of great significance in well testing and reserve estimation. The studies on ROI over the past decades were mainly focusing on homogeneous reservoirs. In heterogeneous formations, such as composite reservoirs and fractured reservoirs, the problem remains unsolved. Few ROI evaluation methods for composite reservoirs have been validated and commonly accepted. This paper conducts extensive analytical and numerical analysis on the pressure propagation of composite reservoirs, analyzing the limitations and misunderstandings of previous methods on the evaluation of ROI. The pressure solution of a two-zone composite reservoir is presented in the Laplace domain and is used to analyze the impact of heterogeneity parameters on pressure transient propagation and the evaluation of radius of investigation. A new ROI formula for composite reservoirs is proposed, which is in great agreement with the analytical solution. The study reveals that the inner zone of a composite reservoir cannot be treated as homogeneous when the pressure front approaches the interface of the two zones with high heterogeneity. A large mobility ratio will slow pressure propagation in composite reservoirs, while a low storativity ratio will increase the value of ROI. Reservoir simulation and real case studies are used to demonstrate the findings of the analytical study. The new study sheds insights into the pressure propagation and eases the calculation of ROI in composite reservoirs.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-024-81494-9.Keywords: Composite reservoir, Radius of investigation, Distance of influence, Mobility ratio, Reservoir heterogeneitySubject terms: Fossil fuels, Hydrology Introduction Pressure front propagation is crucial in many reservoir performance analyses. Radius of investigation (ROI), as a measure of pressure propagation, is a fundamental concept in reservoir engineering and is essentially required in pressure transient tests, well-spacing designs, and reserve estimations. ROI is defined as the distance of pressure wave traveled with the elapse of time after it was initiated from the well-bottom. It depends on many reservoir parameters, such as fluid mobility and rock compressibility. Different ROI equations for homogeneous reservoirs have been presented and applied for decades1–4. Meanwhile, the development of unconventional reservoirs necessitates the estimation of distance of investigation (DOI) for linear flow in fractured horizontal wells. Researchers have presented some new study results and equations on DOI5–11. For homogeneous reservoirs, the most commonly accepted equation in dimensionless form is Eq. (1). Where D is the coefficient, the value of which depends on the selection of different criteria. Various Dvalues have been presented over the past decades of years4. The challenge in estimating ROI of heterogeneous reservoirs arises due to different formation heterogeneities12, and no commonly accepted equation is available. Yuan et al.13developed a new RTA and DOI evaluation method for heterogeneous unconventional reservoirs using the concept of continuous succession of steady states. Fan14studied linear flow formation with an exponential permeability profile and presented ROI euqation for a typical heterogeneous reservoir. Natural fractured reservoir is one type of heterogeneous formation. Sobbi15 and Aguilera16 presented different formulas for ROI in naturally fractured reservoirs, respectively. Fan17pointed out the limitations of these two equations and provided a new ROI equation based on the solution of Warren-Root dual-porosity model18. Composite reservoirs are often encountered in situations such as geological discontinuity, well damage, polymer flooding, or thermal recovery of heavy oil. Few ROI equations for composite reservoirs have ever been presented. Wang19 studied the ROI for non-Newtonian fluids and presented an equation for ROI of composite reservoirs, as in Eq. (2). Where, k1 is the permeability of the inner zone and r1 is the radius of the inner zone. Wang stated that Eq. (2) applies to outer zone (ri>r1), and he treated the composite reservoir as homogeneous before pressure front reaches the interface of the inner and outer zones. However, as will be demonstrated in the next section, the propagation of pressure will be affected by the properties of the outer region even before the interface is reached. Equation (2) cannot reflect the impact of the storativity ratio of the composite reservoir either. The mistake when deriving Eq. (2) is the author’s inappropriate assumption of a constant formation flow rate at any location, which confused the concept of steady and unsteady state in pressure transient theory. Nie20 derived ROI formulas for multi-zone composite reservoirs. For a two-zone composite model, the following equation was proposed. Where is radius of inner zone, , , . In Eq. (3), Nie uses the same ROI formula for homogeneous reservoirs as Wang (1997) when . This assumption taken by Wang and Nie, however, has not been theoretically verified and will lead to a misunderstanding of pressure traveling and ROI calculation in composite reservoirs. For instance, in the case of an outer zone with an extremely low permeability, the interface of the two zones will behave like an impermeable barrier. The pressure propagation will be totally different from that in a homogeneous infinite-acting reservoir. In Nie’s model, the relationship of ROI with time in the II-zone is derived using Eq. (3) when . The typical curves about ROI versus time in a two-zone composite reservoir from Nie (2019) are shown in Fig. 1.Unsuitabilities in this figure include: (1) the ROI in I-zone is entirely not affected by permeability change in the II-zone; (2) the lower permeabilities in II-zone yield higher ROI values. Besides the problematic application of ROI in I-zone, the sequence of these typical curves may be wrongfully marked. The radius of investigation cannot increase with the decrease of the II-zone permeability. Typical curves of ROI in a composite reservoir with different II-zone permeabilities (From Nie20). Further, in Eq. (3), we cannot find the impact of mobility ratio and storativity ratio between the two zones. In deriving Eq. (3), Nie assumed that the flow rate at the interface of the I-zone and II-zone is equal to the constant well production rate at the sandface. Wang (1997) also assumed a constant formation flow rate, q (r, t) = qB (B is formation volume factor) when deriving Eq. (2). This assumption from the two authors is obviously invalid according to pressure transient flow theory. The formation flow rate q (r, t) varies with time and distance. It may be equal to the well bottom-hole production rate only when the time is very long at a certain distance. Understanding the flow regime is essential to the study of radius of investigation in composite reservoirs. In this paper, we first derive the analytical solution for a two-zone composite reservoir and discuss the determination of ROI for the case of constant production rate. We analyze the influences of heterogeneity parameters of a composite reservoir on pressure propagation. Then, a numerical simulation of a composite reservoir model is conducted to study the performance of pressure transient propagation. Finally, a real case study is provided to compare the calculation of radius of investigation with different methods. Methods Solution of composite reservoir under constant well production rate The composite reservoir model used here is shown in Fig. 2. The assumptions include (1) homogeneous inner and outer zones with different flow efficiency, porosity, and compressibility; (2) single phase flow under Darcy’s law; (3) ignore gravity, skin, and wellbore storage; (4) the radius of the inner zone is r1, the outer zone boundary is infinite; and (5) the well produces at a constant production rate q. Scheme of a two-zone composite reservoir in analytical model. According to the above model, the pressure solution in Laplace domain for the inner zone is (Appendix B) Where, , , , . The dimensionless variables are defined in Appendix A. The pressure gradient for inner zone is The pressure and pressure gradient solutions for the outer zone are given in Appendix B (Eq. B-24 and Eq. B-25). The following analysis of pressure propagation and flow rate distribution is based on these solutions and the corresponding numerical inverse Laplace transform. The pressure and pressure gradient profiles at different times are shown in Figs. 3 and 4. It is observed that the inner zone performance is affected by the mobility ratio (M) even before the interface of the composite reservoir is reached. This differs from the previous assumptions of Wang19and Nie20. A higher M value will increase the pressure drop before the interface. But after the interface, the pressure drop goes down, even below the value of the homogeneous model (M = 1). For pressure gradient (equivalent to flow rate), a higher M value decreases the pressure gradient before and after the interface. The impact of mobility ratio on pressure gradient is more noticeable than that on the pressure. Pressure profiles of a composite reservoir (the interface is at r1D = 500). Pressure gradient profiles of composite reservoir (the interface is at r1D = 500). At an early stage, the system behaves like an infinite homogeneous reservoir; the pressure and pressure gradient are independent of the mobility ratio. As time increases, the heterogeneity begins to affect the pressure distribution near the interface of the composite reservoir. Under an unfavorable mobility ratio, the mobility of the outer region is smaller than the mobility of the inner region (i.e., M > 1), and the interface acts as a closed outer boundary, causing a greater pressure drop, as in Fig. 3. From Fig. 4, in the outer zone near the interface, the pressure gradient is lower than that of a homogeneous reservoir, representing the requirement of flowing rate continuity. Definition of ROI in composite reservoirs under constant production rate Various criteria for the definition of ROI have been studied4, among which is the flow rate ratio or pressure ratio, which is highly flexible, but worthfully applied. Define a front-bottomhole rate ratio (FBRR), which is the ratio of flow rate at pressure front to that at the well-bottom of a well under constant production rate9,10. (For constant-production-pressure well, it is FBPR, which is the ratio of dimensionless front pressure to dimensionless bottomhole pressure.) When FBRR reaches a certain predefined value, the pressure disturbance begins to be felt or observable in the formation. Mathematically, the FBRR for a constant-production-rate well in a circular reservoir can be expressed as In which β is the predefined value of FBRR. The conventional ROI equations corresponding to different β values are outlined in Table 1. For instance, the β value of 0.37 gives the most frequently used ROI equation as Eq. (1), in which the D value equals to 2. Table 1Conventional DOI equations with different coefficients (from Fan9).AuthorDOI FormulaD valueFBRRBrownscombe (1951), Van Poolen (1964) 1.780.45Muskat (1937), Lee (1982), Streltsova (1988) 20.37Hurst (1961), Finjord (1988) 2.820.14Jones (1962) 40.02Tek et al. (1957) 4.290.01 Conventional DOI equations with different coefficients (from Fan9). In a composite reservoir, relationship of FBRR and flow rate should be redefined according to different zones. Figure 5 shows the relationship of FBRR with distance at different times. In the inner zone, as time increases, the FBRR gradually gets close to 1 (dimensionless wellbottom production rate is 1). In the I-zone far from the interface, the FBRR is almost not affected by the mobility ratio M. In this situation, the ROI equation for homogeneous reservoirs can be used. However, when the pressure front approaches or outdistances the interface, the ROI begins to be highly affected by the mobility ratio. It is obvious that the higher value of M, the longer it takes to reach a certain distance in the outer zone. The heterogeneity affects the behavior of the outer zone more than that of inter-zone, but the degree of this impact weakens with the increase of mobility ratio M. Impact of mobility ratio on flow rate ratio (FBRR) in composite reservoir (the interface is at r1D = 500). Time to reach interface in composite reservoir at different mobility ratios The previous models from Wang (1997) and Nie (2019) derive the same time to reach the interface as in a homogeneous reservoir. We can prove this is a wrong conclusion based on the analytical model. It is observed from Fig. 6 that the radius discontinuity prolongs the time for the pressure front to reach the interface under an unfavorable mobility ratio (M > 1). Change of flow rate at the interface under different mobility ratios (the interface is at r1D = 500). From Fig. 6, we can determine the time to reach the interface in a composite reservoir by drawing a horizontal line with a certain flow rate ratio (FBRR) value, such as 0.37, representing the traditional ROI equation with a D value of 2. Table 2 lists the times to reach the interface of a composite reservoir under various mobility ratios from different models. When M = 1, the time from the analytical model for composite reservoir is approximately equal to the time from ROI equation as Eq. (1) for homogeneous reservoirs. Table 2Time to reach interface in composite reservoir at different mobility ratios (r1D = 500).Time to interfaceComposite ModelHomogeneous ModelM = 1M = 2M = 3M = 4M = 1tD_r163,13076,63288,674100,26062,500 Time to reach interface in composite reservoir at different mobility ratios (r1D = 500). Define the storativity ratio of the inner and outer zones as . If the permeability and viscosity of the two zones in a composite reservoir are the same (M = 1), the ratio of diffusivity coefficient () becomes inverse of storativity ratio (). High value refers to a lower inner zone storativity, which leads to a shorter time to reach the interface, representing a faster propagation of pressure waves. Figure 7 depicts flow rate change at the interface of composite reservoir under different inverse storativity ratios. It is observed that the impact degree of storativity becomes slighter when the value increases from 1 to 4. It is also noted the impact of the storativity ratio is less sensitive than that from the mobility ratio on the flow rate at the interface of a composite reservoir. Change of flow rate at the interface under different storativity ratios (the interface is at r1D = 500). Pressure traveling in the outer zone of a composite reservoir To study the pressure traveling in the outer zone, we select a distance rD=1000 outside of the interface (r1D = 500). We also draw the FBRR change for a distance inside of the interface (rD=100) for comparison, as shown in Fig. 8. It is observed that the change of mobility ratio M has negligible impact on the inner zone away from the interface but has an obvious influence in the outer zone. The increase of M prolongs the time to reach the distance rD = 1000. By setting a critical value of FBRR, we can determine the precise time for the pressure front to reach the distance rD = 1000. Comparison of FBRR at inner and outer zones under different mobility ratios (the interface is at r1D = 500). The application of Eq. (2) is tested to calculate the arrival time to a distance outside of the interface. Assuming a constant fluid viscosity, the dimensionless form of Eq. (2) becomes Figure 9 shows the comparison of pressure traveling time by setting a distance rD=1000 (the interface of the two zones is at r1D = 500). The arrival times under different M values from Eq. (9) are higher than those from the analytical model, which implies a slower propagation velocity of the pressure front. Comparison pressure travel times at the outer-zone from different methods. (the interface is at r1D = 500, the distance in outer-zone is rD=1000). By Eq. (9), we calculated the ROI in the outer zone at different times and compared the results with the analytical solution from this work. The results are shown in Fig. 10. It is noted that Wang’s model and analytical solution give the same ROI when M = 1. However, when M is great than 1, Wang’s model generates unreliable ROI values in outer zone, especially when the time is not very long, or the distance is near to the interface. Only when the propagating time becomes very long the ROI from Eq. (9) gets close to the analytical solution. Wang adopted the ROI equation for a homogeneous reservoir before the interface is reached. However, the main problem with Wang’s method is that the calculated ROI does not have continuity from the inter-zone to the outer zone. The abrupt change of ROI at the interface of a composite reservoir from Wang’s method is inexplainable. Comparison of ROI from Wang model and analytical solution (the interface is at r1D = 500). The above analysis provides two insights into the ROI of composite reservoirs: (1) The inner zone of a composite reservoir is equivalent to a homogeneous reservoir only when the pressure front is some distance away from the interface; (2) when time becomes very long, the ROI equation will be reduced to . Therefore, I suggest the following ROI equation for composite reservoirs. In which, is the distance to interface. Ei is the exponential integral function, . M is the mobility ratio. Mathematically, when is much great than 1 and when . Before the time of pressure front reaching the interface, the value of is very large and Eq. (10) can be simplified to Eq. (1). When time is very long, Eq. (10) becomes , reflecting the impact of the mobility ratio in composite reservoirs. We compared the ROI calculation results from Eq. (10) with the analytical solution for the same composite reservoir with an interface at r1D = 500, as shown in Fig. 11. There is an excellent agreement between the ROI from Eq. (10) and the analytical solution. Meanwhile, Eq. (10) is a direct calculation of ROI in composite reservoirs, providing convenience in many reservoir engineering calculations, such as dynamic drainage area or dynamic reserve estimation. The accuracy and efficiency compared with other methods make Eq. (10) highly applicable for estimating ROI in composite reservoirs. Comparison of ROI from Fan model and analytical solution (the interface is at r1D = 500). Validation Numerical simulation Synthetic simulation models are built to illustrate the pressure performance in composite reservoirs with varying properties. The reservoir is a two-zone composite reservoir with an outer boundary of 1000 m, as shown in Fig. 12. The inner and outer zone interface is located 400 m from the well. The inner zone permeability is 20mD, and the outer zone permeability is 5mD. The simulation model uses a radial grid system, with 33 cells in r direction, 1 cell in θ and 1 cell in z direction. The well is located in the center of the reservoir, producing at a constant surface production rate of 30m3/day. Other parameters of simulation models are listed in Table 3. Reservoir permeability distribution map in simulation model (mD). Table 3Summary of parameter values in synthetic numerical simulation models.ParametersValuesParametersValuesReservoir thickness, m20Initial pressure, MPa25.84Inner zone radius, m400Outer zone radius, m1,000Inner zone permeability (mD)20Outer zone permeability (mD)5Inner zone porosity0.1Outer zone porosity0.1 Summary of parameter values in synthetic numerical simulation models. Figure 13 shows the pressure distribution of the numerical simulation at a production time of 182 days, with a higher pressure drop in the inner zone. To obtain the pressure profiles in the distance, we extracted the pressure of each radial grid block and compared the composite model with the relevant homogeneous model, as shown in Fig. 14. In the composite model, the II-zone has low permeability (M = 4), resulting in lower pressure in the I-zone. However, outside the interface, the pressure from the composite reservoir becomes gradually higher than the pressure from the homogeneous model. This is consistent with the analytical curves in Fig. 3. In homogeneous model the II-zone has the same permeability as I-zone (M = 1). The lower flow resistance in II-zone of the homogeneous model results in an easy pressure drop. The flow rate profiles of reservoir simulation are calculated from the pressure gradient distribution of reservoir simulation, as shown in Fig. 15. It is noted that the flow rate is not constant in both zones but decreases with distance in both zones. Away from the wellbottom, the flow rate from the composite model is always below the flow rate from the homogeneous model. Reservoir pressure map in simulation model (kPa, time = 182 days). Comparison of pressure profiles in simulation models (time = 182 days). Comparison of flow rate profiles in simulation models (time = 182 days). Real case study In the real case study, we use the same data of well X2 from Nie (1997). The data used to calculate ROI are listed in Table 4. The radius of the inner zone is 64.2 m, and the corresponding dimensionless value rD is equal to 1035.5. The inner zone permeability k1 is 113mD. The outer zone permeability k2 is 21.77mD. The viscosity in both zones remains consistent. The mobility ratio M is 5.19. For the comparative study, we also changed the outer zone permeability according to the reference, listed in Table 5 from Case 1 to Case 5. Table 4Data of well X2 for real case study.ParametersValuesParametersValuesInner zone radius, m64.2Outer zone radius, m-Inner zone permeability (mD)113Outer zone permeability (mD)21.77Inner zone porosity0.069Outer zone porosity0.058Inner zone viscosity, mPa·s57.42Outer zone viscosity, mPa·s57.42Inner zone total compressibility, 1/MPa0.001Outer zone total compressibility, 1/MPa0.0009Formation thickness, m15.39Wellbore radius, m0.062 Data of well X2 for real case study. Table 5Heterogeneity parameters in different cases.Casesk1, mDk2, mDMη1, cm2/sη2, cm2/sωCase 111321.775.19285733.93Case 211331.773.562851062.69Case 311341.772.712851392.05Case 411351.772.182851731.65Case 511371.771.572852391.19 Heterogeneity parameters in different cases. From the analytical solution, we derived the dimensionless arrival time to the interface, which is 553,290. In real-time, it is 74571s (20.7 h), while from Nie’s calculation, this arrival time is 10.036 h. According to the analytical method in this work, we calculated the radius of investigation for different cases, as shown in Fig. 16. It is observed that the relationship between ROI and time is affected by the change in mobility ratio even before the interface is reached. Only in I-zone far from the interface can we neglect the influence of the variation of II-zone permeability. This is different from Nie’s study, which treated the inner zone as a homogeneous infinite-acting model. Figure 16 also indicates that higher II-zone permeability (or low mobility ratio) gives a larger value of ROI. This is also consistent with analytical analysis. Radius of investigation in real case study under different outer zone permeabilities. To check the reliability of Nie’s model, we regenerated the typical curves of ROI versus time according to Eq. (3), and compared them with the analytical method in this work, as shown in Fig. 17. It is observed that Nie’s model gives a higher ROI than that from the analytical solution at the same II-zone permeability. The ROI derived from the homogeneous model (k1 = k2 = 113mD) is also shown in Fig. 17, marked as a solid black line. It is noted that when k2 = 71.77mD, the ROI from Nie’s model is even higher than that from the homogeneous model, which is unreasonable. Nie’s model generates an apparent overestimation of the radius of investigation. Meanwhile, we found that the transcendental equation in Eq. (3) is not easy to solve using the Newton iteration suggested by the author. The solution is highly sensitive to the selection of initial values. Comparison of ROI values derived from different models in real case study. Conclusions This paper studies pressure traveling behavior and the estimation of radius of investigation in composite reservoirs based on solutions in the Laplace domain. The impact of heterogeneity on pressure propagation in a composite reservoir is analyzed using analytical and numerical methods. The conclusions include: The inner zone of a composite reservoir is different from infinite acting homogeneous reservoir due to the existence of the interface and heterogeneities among different zones. ROI equations for homogeneous reservoirs cannot be simply applied to the pressure propagation in the inner zone, which will lead to discontinuity in ROI calculation. The inner zone of a composite reservoir is different from infinite acting homogeneous reservoir due to the existence of the interface and heterogeneities among different zones. ROI equations for homogeneous reservoirs cannot be simply applied to the pressure propagation in the inner zone, which will lead to discontinuity in ROI calculation. When pressure propagation approaches the interface of a composite reservoir, the pressure and flow rate profiles are highly affected by the degree of heterogeneities. In the case of an unfavorable mobility ratio (M > 1), the traveling time to reach a certain distance near or beyond the interface of a composite reservoir becomes longer. Equivalently, reduced ROI is obtained in the composite reservoir at an unfavorable mobility ratio. When pressure propagation approaches the interface of a composite reservoir, the pressure and flow rate profiles are highly affected by the degree of heterogeneities. In the case of an unfavorable mobility ratio (M > 1), the traveling time to reach a certain distance near or beyond the interface of a composite reservoir becomes longer. Equivalently, reduced ROI is obtained in the composite reservoir at an unfavorable mobility ratio. The new ROI formula in this work offers an easy-to-use direct calculation of the ROI for composite reservoirs with high accuracy and efficiency, outperforming all previous methods. The new ROI formula in this work offers an easy-to-use direct calculation of the ROI for composite reservoirs with high accuracy and efficiency, outperforming all previous methods. The results in this study apply to composite reservoirs under constant well-bottom production rate. In the case of constant well-bottom production pressure, the ROI should be defined based on FBPR (pressure-front and well-bottom pressure ratio), while other procedures will be the same as in this paper. The results in this study apply to composite reservoirs under constant well-bottom production rate. In the case of constant well-bottom production pressure, the ROI should be defined based on FBPR (pressure-front and well-bottom pressure ratio), while other procedures will be the same as in this paper. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 List of symbols Constants Formation thickness, m Modified Bessel function, first kind, zero order Modified Bessel function, first kind, first order Inner-zone permeability, L2, mD Outer-zone permeability, L2, mD Modified Bessel function, second kind, zero order Modified Bessel function, second kind, first order Mobility ratio Pressure at somewhere in the formation, MPa Initial reservoir pressure, MPa Well bottom-hole flowing pressure, MPa Well production rate, m3/day Distance to the well in radial flow, m Radius of inner-zone in a two-zone composite reservoir, m Dimensionless distance, = Distance of investigation for radial flow, m Wellbore radius, m LAPLACE transform variable replacing time Producing time, seconds Dimensionless producing time Reservoir porosity, decimal Formation fluid viscosity, mPa·s Diffusivity coefficient, cm2/s Euler constant Storativity ratio, , decimal Diffusivity ratio, decimal Author contributions H.F. proposed the new radius of investigation equation for composite reservoirs and finished the analytical and simulation validations. Data availability The main simulation parameters are listed in Table 3 and the data for analytical validation are shown in Tables 4 and 5. The supplementary raw data used in the simulation model are available from the corresponding author on request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The main simulation parameters are listed in Table 3 and the data for analytical validation are shown in Tables 4 and 5. The supplementary raw data used in the simulation model are available from the corresponding author on request.