- Email: [email protected]
Unusual behavior of produced gas oil ratio in low permeability fractured reservoirs
Journal of Petroleum Science and Engineering 144 (2016) 76–83
Contents lists available at ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Unusual behavior of produced gas oil ratio in low permeability fractured reservoirs Palash Panja a,b,n, Milind Deo b a b
Energy & Geoscience Institute, University of Utah, United States Department of Chemical Engineering, University of Utah, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 19 June 2015 Received in revised form 14 October 2015 Accepted 14 March 2016 Available online 15 March 2016
Leading unconventional plays in the US such as Eagle Ford, Bakken and Niobrara have average gas oil ratio (GOR) ranging from 500 SCF/STB to 4000 SCF/STB. The behavior of produced GOR is difficult to characterize for unconventional reservoirs. Initial reservoir pressure (Pi), well operating pressure and fluid properties directly related to Pressure Volume Temperature (PVT) such as bubble point pressure (Pb ), initial GOR (Rsi), GOR at Pb (Rsb), GOR at flowing bottom hole pressure (Rsw) are the key factors affecting produced GOR from low permeability (10–5000 nD) reservoirs. Gas production may be controlled and kept in the desired production window by maintaining the flowing bottom hole pressure (Pwf or BHP). A
Keywords: Gas oil ratio Shales Hydraulic fracture Bottom hole Pressure Bubble point pressure Oil recovery
(
single characteristic factor affecting the produced GOR is found to be 1 −
Rsw Rsb
)
⎛ P ⎞ ⎜1 − wf ⎟ ⎜ Pb ⎟⎠ ⎝ . ⎛ P ⎞ ⎜1 − wf ⎟ ⎜ ⎟ Pi ⎠ ⎝
(1 − ) considers the fluid PVT effect with operating condition, second part ⎛⎝1 − Rsw Rsb
⎜
P wf ⎞ Pb
⎟
⎠
The first part
accounts for the
P wf ⎞ ⎛ proximity of operating pressure with bubble point pressure and third part ⎜1 − P ⎟ is the drawdown ⎝ i ⎠ effect. Production behavior in terms of produced GOR can be predicted using this single factor. Produced GOR increases with time when this factor exceeds a certain value, while, little or no deviations from the initial GOR are observed for lower values of the factor. It should be noted that the predictive factor does
not depend on reservoir matrix permeability. Initially, a factor
(Pb − P wf ) (P i − P wf )
comprising only pressure terms
was developed which failed to capture the behavior of produced GOR. Initial reservoir pressure and flowing bottom hole pressure (Pwf) are varied to study a wide range of reservoirs and production conditions. The oil rates, recovery factor and produced GOR are the key production parameters for this study. Suitability of this factor is validated by comparing simulation data with field data. The optimum well operating pressure can also be determined using this factor to maximize recovery. Deviation of GOR from its initial value is higher for low permeability reservoirs. Higher gas and oil are recovered from reservoirs with higher initial gas oil ratios (Rsi). & 2016 Elsevier B.V. All rights reserved.
1. Introduction Obtaining higher estimated ultimate recovery (EUR) of oil is the primary concern of an operating company. Dissolved gas helps to improve the mobility of oil by making the oil lighter. Oils classified as Black oils contains certain amount of dissolved gas, which evolves when the reservoir pressure drops below bubble point pressure in the course of production. Gas flows easily through n Corresponding author at: Energy & Geoscience Institute, University of Utah, United States. E-mail address: [email protected] (P. Panja).
http://dx.doi.org/10.1016/j.petrol.2016.03.005 0920-4105/& 2016 Elsevier B.V. All rights reserved.
porous medium than oil because of its higher mobility and starts dominating two-phase flow by suppressing the flow of oil. On the other hand, free gas sustains the pressure in the reservoir. Produced GOR gradually increases with time as the reservoir pressure declines in the reservoir. The reasons behind the increase of GOR and decline of oil rate with time are first explained by Millikan (1926) by compiling field data and he also discussed the drawdown effect on produced GOR. Well placement and proper design of flow column were the two sets of factors to determine GOR (Albertson and Schaeffer, 1931). Different possible ways to control producing GOR to optimize the ultimate recovery without compensating with rate of production were
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Nomenclature GCI fp fpn GOR kfx kfy kfz km Pb
Gas-Oil-Ratio Characteristic Index (dimensionless) Difference in Pressure Ratio Factor (dimensionless) Normalized Difference in Pressure Ratio Factor (dimensionless) Gas Oil ratio (SCF/STB) Fracture Absolute Permeability in X-direction (mD) Fracture Absolute Permeability in Y-direction (mD) Fracture Absolute Permeability in Z-direction (mD) Reservoir Absolute Permeability (nD) Bubble Point pressure (psi)
demonstrated (Marsh and Robinson, 1929; Sullivan, 1937) incorporating field data. Multiphase black oil flow equation (Muskat, 1945) was solved numerically (Arps and Roberts, 1955) for initial GOR up to 2000 SCF/STB to obtain oil recovery and the method was validated by comparing the results with actual field data. It was concluded that a definite relationship could be established between oil recovery and reservoir fluid properties like initial GOR, API gravity and type of reservoir rock. Compositional simulations are used (Brinkman and Weinaug, 1956) to predict the GOR and formation volume factors for dissolved gas drive reservoir at saturated conditions. Jones-Parra and Reytor (1959) developed a material balance method incorporating gravity segregation to show the effect of GOR on production from fractured limestone reservoirs. Their results proved that oil rates declined less at higher initial GOR. Levine and Prats (1961) concluded that the produced GOR is independent of reservoir permeability by numerically solving partial differential equations describing solution gas drive reservoirs. Another study by Prats and Levine (1963) showed that the produced GOR for vertically fractured reservoir is higher than unfractured reservoir. A new method to forecast GOR dependent on oil rate and consistent with reservoir mechanisms was developed (Lawal et al., 2006). Laboratory experiments (Busahmin and Maini, 2010) showed that the performance of foamy heavy oil system is affected negatively with increase in initial GOR which were counter-intuitive. Ultra-low permeable reservoirs like shales behave differently from conventional reservoir with high reservoir permeability. Downhole well sensors (Al-Khelaiwi et al., 2014) were used to calculate produced gas oil ratio to avoid faulty gas meter reading at separator on surface. Impact of nano-pores (Khoshghadam et al., 2015) on gas oil ratio was investigated by compositional simulation of liquid rich shale oil reservoir. In this study, we characterize the qualitative behavior of produced GOR from ultra-low permeability fractured reservoirs such as shales. For this purpose, various factors combining pressure terms and/or fluid properties terms are examined. Two possible factors are attempted initially, to capture the behavior of produced GOR; one factor is able to predict the behavior of produced GOR correctly. We also investigated the effect of dissolved GOR on production performance for different reservoir permeabilities and flowing bottom hole pressures.
2. Characterization of GOR A variety of criteria such as constant pressures (initial, bubble point or bottom hole pressures), constant pressure differences and constant ratio of pressure differences are tried initially in this study. Discussing the results of this initial screening study is out of scope of this article. It is observed that none of the them is the valid criteria for characterizing the performance (oil rate, GOR and
Pi PVT Pwf Rpc Rsb Rsi Rsw X Y Z
77
Initial Reservoir pressure (psi) Pressure Volume Temperature (dimensionless) Flowing Bottom Hole pressure (psi) Cumulative Gas Oil ratio (SCF/STB) Gas Oil ratio at Bubble Point Pressure (SCF/STB) Initial Gas Oil ratio (SCF/STB) Gas Oil ratio at Flowing Bottom Hole Pressure (SCF/ STB) Reservoir Dimension in X-direction (ft) Reservoir Dimension in Y-direction (ft) Reservoir Thickness (ft)
oil recovery) but combining them in particular manner can exhibit a conclusive relationship which can predict qualitative behavior of production performance. The ratio of pressure differences (fp) is first examined as shown by Eq. (1).
fp =
(Pb − Pwf ) (Pi − Pwf )
(1)
The oil rate and oil recovery can be explained using ratio of pressure difference factor (fp) (shown in Eq. (1)). However, without incorporation of GOR into the factor, it is ineffective to correctly predict the behavior of the produced GOR. This fact will be established in the later part of the article as we analyze the results. The depletion of long transient state reservoir can be represented by the extent of the difference between the bubble point pressure and flowing bottom hole pressure since the average reservoir pressure for this kind of reservoir is not representative of the behavior. The difference in the GOR at Pwf and GOR at Pb is one characteristics parameter to predict the behavior of produced GOR from ultra-low permeable fractured reservoir. All these factors are accommodated with Eq. (1) into a single characteristic factor, GOR characteristic index (GCI) as given by Eq. (2).
( (
Pwf
1− P ⎛ R ⎞ b GCI=⎜ 1 − sw ⎟ ⎝ Rsb ⎠ 1−Pwf P i
) )
(2)
The factor, GCI, can be used only when flowing bottom hole pressure is less than the bubble point pressure. When flowing bottom hole pressure is above the bubble point pressure, the well produces initial GOR because reservoir is undersaturated. The value of GCI is presented in Table 1 for the different cases in this study. The effect of this factor is clearly observed in the qualitative behavior of produced GOR which is discussed later.
3. Experimental design The common industry belief is that the drawdown i.e. the difference between the reservoir pressure and flowing bottom hole pressure is the key factor to dictate the production behavior. In this study, the significance of location of bubble point pressure is also considered. To get rid of the established idea of drawdown, the experiments are designed in such a way that drawdown i.e., the difference between initial reservoir pressure and flowing bottom hole pressure are same for all runs. The same drawdown for all experiments ensures that drawdown is not affecting the changes in production performance. The location of bubble point pressure with respect to flowing bottom hole pressure and initial reservoir pressure is a significant factor in the production performance of oil and gas. To create different scenarios of experiment,
78
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Table 1 Design of experiments by varying Rsb, Pi and Pwf. Sr. no.
Rsb (SCF/STB)
Pi (psia)
Pb (psia)
Pwf (psia)
(Pi Pb) (psia)
(Pb Pwf) (psia)
(Pi Pwf) (psia)
fp
GCI
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
500 1000 2000 500 1000 2000 500 2000 2000 500 1000 790 2500 500 1000 1000 500 2000
5765 7015 9515 5300 5300 5300 6017 9767 9863 6113 6650 6000 8500 5200 5500 5150 4865 4871
1265 2515 5015 1265 2515 5015 1265 5015 5015 1265 2515 1990 6265 1265 2515 2515 1265 5015
965 2215 4715 500 500 500 1217 4967 5063 1313 1850 1200 3700 400 700 350 65 71
4500 4500 4500 4035 2785 285 4752 4752 4848 4848 4135 4010 2235 3935 2985 2635 3600 144
300 300 300 765 2015 4515 48 48 48 48 665 790 2565 865 1815 2165 1200 4944
4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800
0.0625 0.0625 0.0625 0.1594 0.4198 0.9406 0.0100 0.0100 0.01 0.01 0.1385 0.1646 0.5344 0.1802 0.3781 0.4510 0.2500 1.0300
0.0683 0.0209 0.0071 0.4085 0.7129 0.8976 0.0018 0.0002 0.0 0.0 0.0974 0.1984 0.2975 0.5124 0.6002 0.7997 0.9227 0.9891
Fig. 1. The relative values of initial reservoir pressure, bubble point pressure and flowing bottom hole pressure on Rs-PVT diagram (a) case 1, operation above bubble point pressure, initially undersaturated reservoir (b) case 2, operation below bubble point pressure, initially undersaturated reservoir (c) case 3, operation below bubble point pressure, initially saturated reservoir.
initial reservoir pressure, bubble point pressure, flowing bottom hole pressure and GOR (Rsb) at bubble point are varied. Initially we divided the simulations into various categories as discussed in the
next section to inquire the factor(s) which would characterize the qualitative behavior of production performance mainly GOR. Different categories are shown in Table 1 and discussed here.
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
3.1. Constant pressure differences The difference between initial reservoir pressure and bubble point pressure (Pi Pb) and the difference between bubble point pressure and flowing bottom hole pressure (Pb Pwf) are kept constant for serial numbers 1–3 as shown in Table 1. This is known as constant pressure difference (Δp) case.
79
rate which is not economic. The most practical location of the initial reservoir pressure, bubble point pressure and flowing bottom hole pressure for economic production is case 2. Wells are normally operated well below the bubble point pressure to obtain the economic production rate.
4. Reservoir simulation 3.2. Constant end pressures For serial numbers 4–6 in Table 1, initial reservoir pressure and flowing bottom hole pressure are constant. This can be called constant boundary pressures (or constant P) cases. In constant pressure cases (serial no. 4–6 in Table 1), the difference between initial reservoir pressure and bubble point pressure (Pi Pb) and the difference between bubble point pressure and flowing bottom hole pressure (Pb Pwf) vary depending on the location of bubble point pressures. 3.3. Special cases Serial number 7–18 are special cases to study the effect of characteristic factor, GCI which is defined in Eq. (2). In the special cases, Pi, Pb and Pwf are taken such a way that GCI covers a wide range i.e., from near zero to one as shown in Table 1. This ensures that all possibilities are considered in this study. In some cases such as serial number 9 and 10, well is operated above bubble point pressure i.e., flowing bottom hole pressure is greater than Pb as shown in Fig. 1a. The GCI becomes zero for these cases as shown in Table 1. In these cases, reservoir remains in undersaturated, and produced GOR is always equal to the initial GOR (Rsi). Well can be operated at different pressure depending on initial reservoir pressure and bubble point pressure. Similarly, the values of fp and GCI vary greatly on the relative locations of Pi, Pb and Pwf. Pressures (Pi, Pb and Pwf) in all the 18 cases given in Table 1 can be located in P versus Rs diagram. This relative locations lead to three different possibilities for the ratio of pressure difference as described in Fig. 1a–c. To understand all cases given in Table 1 in the context of Fig. 1, serial numbers are displayed in the figure along with the corresponding values of fp and GCI. Operating well near bubble point pressure suppresses gas production but it also yields very low oil Table 2 Reservoir and completion parameters used in all of the simulations. Reservoir top (ft): Matrix permeability, km (nD): Reservoir dimensions X(ft), Y(ft), Z(ft): Fracture permeability (mD): Fracture width (ft): Fracture height (ft) Fracture orientation Initial oil saturation (%): Reservoir porosity (%):
12,800 10, 100, 1000, 5000, 1000, 750, 200 kfx ¼ kfy ¼ 150; kfz ¼300 0.05 Reservoir height Parallel to YZ plane 84 5
Based on field data available for Eagle Ford from Railroad Commissions of Texas, different initial GORs (300, 500, 1000, 1500, 2000 and 2500 SCF/STB) are selected to represent various reservoir fluids. Ultra-low permeable reservoirs (10, 100, 1000 and 5000 nD) with one horizontal well and one vertical fracture located in the middle of the reservoir are simulated. The fracture height and fracture width are equal to the reservoir height and width respectively. The reservoir is extended to 2000 ft in the xdirection, 750 ft in the y-direction, and 200 ft in the z-direction; with a top depth of 12,800 ft The properties of the reservoir are summarized in Table 2. The matrix permeability, initial reservoir pressure, flowing bottom hole pressure and fluid properties are varied while the fracture permeability, fracture width, fracture orientation, matrix porosity and initial hydrocarbon saturation remain constant. Minimum grid block size (Panja et al., 2013) is used to obtain converged results to get rid of any grid effects.
5. Validation of the factor, GCI Validation is an integral part of model development. Any model should be validated before it is accepted for future application. Real field data is from Eagle Ford is used to check the applicability of GCI. The production data are collected from Railroad Commissions of Texas. The reservoir and fluid parameters such as initial reservoir pressure, bubble point pressure, initial GOR (Rsi) and operating conditions are not available in public data base. These parameters are generated by superimposing simulated gas rate and oil rate with field data. Various cases are simulated with different Pi, Pb, Pwf, Rsi and reservoir permeability as shown in Table 3. Prior knowledge of the field (Panja et. al., 2015) is applied during choosing these parameters. Two different flowing bottom hole pressures are investigated for each field case. The pairs, (1,2), (3,4), (5,6) and (7,8) have same properties except the flowing bottom hole pressure. Comparisons of field data from Eagle Ford with simulation results for eight cases ( Table 3) are shown in Fig. 2. It should be noted that the rates are for non-interfering single fracture and net to gross ratio (NTG) is considered as 0.4 for Eagle Ford. If there are five clusters per stage in a 16-stage well, with about 80 possible fractures, the rate for 50 nd to 450 nD reservoirs is expected to be between 2.5 mstb/month and about 5 mstb/ month in the first year. This depends on the initial dissolved GOR,
Table 3 Plausible reservoir conditions and operational parameter for various lease in Eagle Ford. Lease unit
Simulation case
Pi (psi)
Rsi (SCF/STB)
Pb (psi)
Pwf (psi)
Permeability (nD)
GCI
Hawn holt
1 2 3 4 5 6 7 8
3000 3000 5300 5300 5300 5300 8000 8000
300 300 800 800 950 950 1650 1650
765 765 2015 2015 2390 2390 4140 4140
100 200 650 950 750 1000 1000 1250
450 450 200 200 400 400 50 50
0.80 0.60 0.53 0.34 0.55 0.42 0.66 0.58
JP head bower C Meyer Davenport
80
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Fig. 2. Superimposition of field data with simulation results for (a) Oil rate (b) Gas rate.
6. Result and discussion The matrix permeabilities chosen for this study are 10 nD, 100 nD, 1000 nD and 5000 nD. Initial reservoir pressure, bubble point pressure, flowing bottom hole pressure and GOR (Rsb) at bubble point are already selected to create a wide range of values of GCI as given in Table 1. Total 72 simulations ( 4 permeabilities for each18 cases given in Table 1) are run to investigate the effect of dissolved GOR curve and operation parameters on production performance in terms of oil rate, oil recovery and produced GOR. Commercial simulator IMEX, a Computer Modeling Group Black Oil simulator, is used to conduct the study.
7. Controlling produced GOR
Fig. 3. Comparison of GOR behavior between field data and simulated results.
drawdown, bubble point pressure and a number of other factors. Analysis of the Eagle Ford oil rate indicates that the reservoirs fall within the 50–450 nD permeability range. It is evident from Fig. 2a and b that the oil rates and gas rates are not very much affected by flowing bottom hole pressures ( in the range used here) for the reservoirs of 50–450 nD permeability. Superimposing GOR with the field data is also necessary to generate the reservoir conditions. The comparison simulated results and Eagle Ford field data of produced GOR is shown in Fig. 3. The GOR values for Eagle Ford are much more variable from 300 to 1650 SCF/STB. GCI is calculated for each case as shown in Table 3. Ratio of cumulative GOR and initial ratio produced from JP Head Bower C Unit and Meyer Unit are less compared to same from Hawn Holt and Davenport unit as shown in Fig. 3. These performances are also correctly predicted by GCI. The values of GCI for JP Head Bower C Unit and Meyer Unit (0.34–0.55) are less than the values of GCI for Hawn Holt and Davenport unit (0.58–0.80). Higher GCI indicates that increase in GOR with respect to initial GOR is higher. Although the initial GOR of Hawn Holt is 300 SCF/ STB only, the rise of GOR (ratio of GOR and initial GOR) with time is higher than the other field. This performance also confirms that initial GOR is not the only factor to decide behavior of the produced GOR. GCI can be used reliably to analyze qualitative behavior of produced GOR.
It is desired to operate wells to control the GOR, hence, prescribing flowing bottom hole pressure is one quest of this study. GCI can also be utilized to propose the flowing bottom hole pressure to operate wells in a manner to optimize the GOR. The flowing bottom hole pressure can be controlled to maintain desired produced GOR by setting the value of GCI. Assuming that the dissolved GOR varies linearly with pressure for saturated reservoir, following formula for the flowing bottom hole pressure is derived as shown in Eq. (3).
Fig. 4. Variability of Pwf/Pb with the factor GCI.
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
⎤ ⎡ ⎤2 ⎛P ⎞ ⎛P ⎞ Pwf ⎡ =⎢ 1 − 0. 5 ⎜ b ⎟ GCI⎥− ⎢ 1 − 0. 5 ⎜ b ⎟ GCI⎥ − (1−GCI) ⎝ Pi ⎠ ⎝ Pi ⎠ Pb ⎣ ⎦ ⎣ ⎦
81
(3)
Knowing the initial reservoir pressure and bubble point pressure of reservoir fluid, the value of flowing bottom hole pressure is prescribed using Eq. (3) to maintain desired produced GOR. The variation of Pwf/Pb with the factor, GCI is shown in Fig. 4. The relationship between Pwf/Pb and GCI are established for various ratios of Pb/Pi. It is clear that Pwf/Pb is inversely proportional to GCI. Higher value of GCI predicts lower value of Pwf/Pb i.e., lower value of flowing bottom hole pressure compared to bubble point pressure which leads to produce more gas. The curve approaches to linear when Pb/Pi moves towards unity. Little or no deviation of produced GOR compared to initial GOR is observed if well is operated at the pressure higher than 0.5 Pb to 0.7 Pb depending on the Pb/Pi fraction. For example, if any well in a reservoir with initial pressure of 5500 psi and bubble point pressure of 3500 psi must be operated at pressure higher than about 0.65 Pb i.e., 2275 psi ( as predicted by Eq. (3) and Fig. 4) to keep the produced GOR near to initial GOR, but rate of oil production may not be economic for this high flowing bottom hole pressure (2275 psi). If factor value of 0.6 is a chosen, the flowing bottom hole pressure is near 0.3 Pb i.e., 1050 psi which will enhance the rate of oil production in expense of higher produced GOR. Optimization of oil rate depends on the operators considering many factors like ability to handle gas, long term production strategy etc. 7.1. Effect of initial gas oil ratio (Rsi) The effect of initial GOR on produced GOR, oil rate and oil recovery are discussed for three different initial GORs (500, 1000 and 2000 SCF/STB) while ratio of pressure difference is kept constant at 0.0625 although GCI varies. Slope of dissolved GOR curve(dRs /dp) of 0.4 and four reservoir permeabilities (10, 100, 1000 and 5000 nD) are considered. The effect of initial GOR on cumulative GOR is shown in Fig. 5a and b for the ratio of pressure difference is 0.0625. It is intuitive that higher GOR is produced from reservoirs with higher initial GOR. The deviation of GOR from initial GOR is not noticeable with different matrix permeabilities for very low value of GCI (0.007–0.07). The differences in the oil rate for different initial GORs are not significant for lower matrix permeability (10 and 100 nD). Higher oil rate is achieved with higher initial GOR and with higher reservoir permeability. Higher initial gas dissolved
Fig. 6. Effect of initial gas oil ratio on oil recovery with constat fp of 0.0625.
in oil phase always supports the production by sustaining the reservoir pressure for long time. The mobility of oil is also improved with dissolved gas by making oil phase lighter. Simultaneously, more gas is produced on the surface as dissolved and free gas. Higher initial GOR finally enhances the oil rate in cost of higher gas production. The oil recovery is also improved from higher initial GOR as evident in Fig. 6 for the ratio of pressure difference of 0.0625. The initial GOR effect on oil recovery is very much noticeable for higher permeability reservoirs (5000 and 1000 nD). Highest oil recovery is obtained from reservoir with initial GOR of 2000 SCF/ STB and reservoir permeability of 5000 nD. As described earlier, the higher initial dissolved gas provides higher energy in the reservoir, thus helps to produce more oil. Lowest amount of oil is recovered from the reservoir with initial GOR of 500 SCF/STB. The effect of initial GOR on oil recovery is not very prominent for 10 nD reservoir. 7.2. Effect of GCI The effect of GCI is already established in earlier sections with
Fig. 5. Effect of initial gas oil ratio with constat fp of 0.0625 on (a) cumulative gas oil ratio (b) Oil rate.
82
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Fig. 7. Cumulative gas oil ratio for (a) GCI of 0.054, 0.068 and (b) Undersaturated reservoir, GCI¼0 (Dash line: dRs/dp¼ 0.8 (SCF/STB)/psi, Solid line: dRs/dp¼ 0.4 (SCF/STB)/psi).
validation with field data. The factor GCI is of particular interest for quantitative analysis of produced GOR. Behavior of produced GOR is dependent not only on the operating conditions but also on the fluid properties and initial reservoir conditions. The qualitative nature of produced GOR cannot be explained solely by the ratio of pressure difference, fp. The factor, fp is useful to explain the oil rate and oil recovery only for different initial GOR. Produced GOR is normalized with initial GOR to compare among the fluids having different initial GOR (500, 1000 and 2000 SCF/STB) in the same scale. As shown in Fig. 4, the deviations of produced GOR depend on value of GCI. The simulated results are analyzed here for various GCI. 7.3. Little or no deviation of GOR Produced GORs have neutral deviation when produced GORs vary within 72% of initial GOR. The neutral deviations are depicted in Fig. 7a and b. The neutral or little deviations are found when the factor (GCI) is not greater than 0.07. A special case for neutral deviation can be identified when the flowing bottom hole pressure is above the
bubble point pressure i.e. reservoir is in undersaturated conditions. Neutral deviation occurs when the flowing bottom hole pressure is close to bubble point pressure. Sufficient amount of gas does not evolve out of oil phase and the oil phase contains initial amount of dissolved gas only. 7.4. Positive deviation of GOR Positive deviation of GOR is commonly observed in the field. Positive deviations of produced GOR are shown in Fig. 8a and b. Positive deviation is noticed when large portion of reservoir is below bubble point pressure and flowing bottom hole pressure is sufficiently below bubble point pressure. In these reservoir conditions, gas and oil both phases flow simultaneously. The values of GCI are greater than 0.2 for positive deviation. Effect of GCI on cumulative RATO in terms of deviations is summarized in Table 4.
8. Conclusion The complex behaviors of produced GOR, oil rate and oil
Fig. 8. Positive deviation of cumulative gas oil ratio for GCI of (a) 0.24, 0.41, 0.41, 0.92 and (b) 0.89, 0.99 (Dash line: dRs/dp¼ 0.8 (SCF/STB)/psi, Solid line: dRs/dp ¼0.4 (SCF/ STB)/psi).
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Table 4 Effect of GCI on cumulative GOR. Range of GCI Comments cumulative GOR Below 0.07 Above 0.2
No or little deviation Positive deviation
recovery from ultra-low permeable fractured reservoir are illustrated with various PVT properties by varying important parameters like initial dissolved GOR slope of dissolved GOR, operating conditions and reservoir conditions. A single universal predictive factor – GCI, which helps understand the GOR behavior of lowpermeability reservoirs is defined. The following conclusions are made; Effect of Rsi
Higher initial dissolved GOR yields higher produced GORs. Effect on oil rate is not very significant. Higher oil recovery is obtained from higher initial GOR and higher reservoir permeability. The value of using GCI – the single universal factor
Little deviations from initial dissolved GOR are observed in production GOR when factor is less than 0.07.
Positive deviations of produced GOR are observed when factor is greater than 0.2.
Higher deviations (positive and negative) are noticed from reservoir with lower permeability reservoir. GCI is effective measure of the qualitative behavior of produced GOR accommodating reservoir conditions, fluid properties and operating condition. Well can be operated to produce desired GOR by using the value of GCI prescribed for particular GOR window. The value of GCI is specific to the other parameters like relative permeability, critical gas saturation, porosity of reservoir etc. but its characteristic application on produced GOR would be identical.
Acknowledgments The authors gratefully acknowledge the academic license to CMG products from Computer Modeling Group, Calgary, Canada. Financial support to Mr. Palash Panja through the Conoco Phillips Fellowship is also acknowledged.
83
References Al-Khelaiwi, F.T., Muailu, H.A., Alfakher, A., 2014. Gas Oil Ratio (GOR) Estimation Using Downhole and Surface Well Sensors. Albertson, M., Schaeffer Jr., W.A., 1931. Control of gas-oil ratios in the yates field, Pecos County, Texas. J. Soc. Pet. Eng. Arps, J.J., Roberts, T.G., 1955. The Effect of the Relative Permeability Ratio, the Oil Gravity, and the Solution Gas-Oil Ratio on the Primary Recovery From a Depletion Type Reservoir. Brinkman, F.H., Weinaug, C.F., 1956. Calculated Performance of a Dissolved Gas Drive Reservoir by a Phase Behavior Method. Busahmin, B.S., Maini, B.B., 2010. Effect of solution-gas-oil-ratio on performance of solution gas drive in foamy heavy oil systems. Canadian Unconventional Resources and International Petroleum Conference, Society of Petroleum Engineers, Calgary, Alberta, Canada. Jones-Parra, J., Reytor, R.S., 1959. Effect of Gas-Oil Ratio on the Behavior of Fractured Limestone Reservoirs. Khoshghadam, M., Khanal, A., Lee, W.J., 2015. Numerical Study of Impact of NanoPores on Gas-Oil Ratio and Production Mechanisms in Liquid-Rich Shale Oil Reservoirs, Society of Petroleum Engineers. Lawal, K.A., Uwaga, A.O., Osoro, O.F., 2006. A Systematic Methodology for Extrapolating Gas-Oil Ratio During Declining Oil Production, Society of Petroleum Engineers. Levine, J. S., Prats, M., 1961. The Calculated Performance of Solution-Gas-Drive Reservoirs. Marsh, H.N., Robinson, B.H., 1929. Means of Controlling Gas-oil Ratio. Millikan, C.V., 1926. Gas-Oil Ratio as Related to the Decline of Oil Production, with Notes on the Effect of Controlled Pressure. Muskat, M., 1945. The production histories of oil producing gas-drive reservoirs. J. Appl. Phys. 16, 147–159. Panja, Palash, C, T., Deo, Milind, 2015. Factors controlling production in hydraulically fractured low permeability oil reservoirs. Int. J. Oil Gas Coal Technol. Panja, P., Conner, T., Deo, M., 2013. Grid sensitivity studies in hydraulically fractured low permeability reservoirs. J. Pet. Sci. Eng. 112, 78–87. Prats, M., Levine, J.S., 1963. Effect of vertical fractures on reservoir behavior- results on oil and gas flow. J. Soc. Pet. Eng., SPE-593-PA Sullivan, R.J., 1937. Gas-Oil Ratio Control in Flowing Wells. Eighteenth Annual Meeting, American Petroleum Institute, Cicago.
Contents lists available at ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Unusual behavior of produced gas oil ratio in low permeability fractured reservoirs Palash Panja a,b,n, Milind Deo b a b
Energy & Geoscience Institute, University of Utah, United States Department of Chemical Engineering, University of Utah, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 19 June 2015 Received in revised form 14 October 2015 Accepted 14 March 2016 Available online 15 March 2016
Leading unconventional plays in the US such as Eagle Ford, Bakken and Niobrara have average gas oil ratio (GOR) ranging from 500 SCF/STB to 4000 SCF/STB. The behavior of produced GOR is difficult to characterize for unconventional reservoirs. Initial reservoir pressure (Pi), well operating pressure and fluid properties directly related to Pressure Volume Temperature (PVT) such as bubble point pressure (Pb ), initial GOR (Rsi), GOR at Pb (Rsb), GOR at flowing bottom hole pressure (Rsw) are the key factors affecting produced GOR from low permeability (10–5000 nD) reservoirs. Gas production may be controlled and kept in the desired production window by maintaining the flowing bottom hole pressure (Pwf or BHP). A
Keywords: Gas oil ratio Shales Hydraulic fracture Bottom hole Pressure Bubble point pressure Oil recovery
(
single characteristic factor affecting the produced GOR is found to be 1 −
Rsw Rsb
)
⎛ P ⎞ ⎜1 − wf ⎟ ⎜ Pb ⎟⎠ ⎝ . ⎛ P ⎞ ⎜1 − wf ⎟ ⎜ ⎟ Pi ⎠ ⎝
(1 − ) considers the fluid PVT effect with operating condition, second part ⎛⎝1 − Rsw Rsb
⎜
P wf ⎞ Pb
⎟
⎠
The first part
accounts for the
P wf ⎞ ⎛ proximity of operating pressure with bubble point pressure and third part ⎜1 − P ⎟ is the drawdown ⎝ i ⎠ effect. Production behavior in terms of produced GOR can be predicted using this single factor. Produced GOR increases with time when this factor exceeds a certain value, while, little or no deviations from the initial GOR are observed for lower values of the factor. It should be noted that the predictive factor does
not depend on reservoir matrix permeability. Initially, a factor
(Pb − P wf ) (P i − P wf )
comprising only pressure terms
was developed which failed to capture the behavior of produced GOR. Initial reservoir pressure and flowing bottom hole pressure (Pwf) are varied to study a wide range of reservoirs and production conditions. The oil rates, recovery factor and produced GOR are the key production parameters for this study. Suitability of this factor is validated by comparing simulation data with field data. The optimum well operating pressure can also be determined using this factor to maximize recovery. Deviation of GOR from its initial value is higher for low permeability reservoirs. Higher gas and oil are recovered from reservoirs with higher initial gas oil ratios (Rsi). & 2016 Elsevier B.V. All rights reserved.
1. Introduction Obtaining higher estimated ultimate recovery (EUR) of oil is the primary concern of an operating company. Dissolved gas helps to improve the mobility of oil by making the oil lighter. Oils classified as Black oils contains certain amount of dissolved gas, which evolves when the reservoir pressure drops below bubble point pressure in the course of production. Gas flows easily through n Corresponding author at: Energy & Geoscience Institute, University of Utah, United States. E-mail address: [email protected] (P. Panja).
http://dx.doi.org/10.1016/j.petrol.2016.03.005 0920-4105/& 2016 Elsevier B.V. All rights reserved.
porous medium than oil because of its higher mobility and starts dominating two-phase flow by suppressing the flow of oil. On the other hand, free gas sustains the pressure in the reservoir. Produced GOR gradually increases with time as the reservoir pressure declines in the reservoir. The reasons behind the increase of GOR and decline of oil rate with time are first explained by Millikan (1926) by compiling field data and he also discussed the drawdown effect on produced GOR. Well placement and proper design of flow column were the two sets of factors to determine GOR (Albertson and Schaeffer, 1931). Different possible ways to control producing GOR to optimize the ultimate recovery without compensating with rate of production were
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Nomenclature GCI fp fpn GOR kfx kfy kfz km Pb
Gas-Oil-Ratio Characteristic Index (dimensionless) Difference in Pressure Ratio Factor (dimensionless) Normalized Difference in Pressure Ratio Factor (dimensionless) Gas Oil ratio (SCF/STB) Fracture Absolute Permeability in X-direction (mD) Fracture Absolute Permeability in Y-direction (mD) Fracture Absolute Permeability in Z-direction (mD) Reservoir Absolute Permeability (nD) Bubble Point pressure (psi)
demonstrated (Marsh and Robinson, 1929; Sullivan, 1937) incorporating field data. Multiphase black oil flow equation (Muskat, 1945) was solved numerically (Arps and Roberts, 1955) for initial GOR up to 2000 SCF/STB to obtain oil recovery and the method was validated by comparing the results with actual field data. It was concluded that a definite relationship could be established between oil recovery and reservoir fluid properties like initial GOR, API gravity and type of reservoir rock. Compositional simulations are used (Brinkman and Weinaug, 1956) to predict the GOR and formation volume factors for dissolved gas drive reservoir at saturated conditions. Jones-Parra and Reytor (1959) developed a material balance method incorporating gravity segregation to show the effect of GOR on production from fractured limestone reservoirs. Their results proved that oil rates declined less at higher initial GOR. Levine and Prats (1961) concluded that the produced GOR is independent of reservoir permeability by numerically solving partial differential equations describing solution gas drive reservoirs. Another study by Prats and Levine (1963) showed that the produced GOR for vertically fractured reservoir is higher than unfractured reservoir. A new method to forecast GOR dependent on oil rate and consistent with reservoir mechanisms was developed (Lawal et al., 2006). Laboratory experiments (Busahmin and Maini, 2010) showed that the performance of foamy heavy oil system is affected negatively with increase in initial GOR which were counter-intuitive. Ultra-low permeable reservoirs like shales behave differently from conventional reservoir with high reservoir permeability. Downhole well sensors (Al-Khelaiwi et al., 2014) were used to calculate produced gas oil ratio to avoid faulty gas meter reading at separator on surface. Impact of nano-pores (Khoshghadam et al., 2015) on gas oil ratio was investigated by compositional simulation of liquid rich shale oil reservoir. In this study, we characterize the qualitative behavior of produced GOR from ultra-low permeability fractured reservoirs such as shales. For this purpose, various factors combining pressure terms and/or fluid properties terms are examined. Two possible factors are attempted initially, to capture the behavior of produced GOR; one factor is able to predict the behavior of produced GOR correctly. We also investigated the effect of dissolved GOR on production performance for different reservoir permeabilities and flowing bottom hole pressures.
2. Characterization of GOR A variety of criteria such as constant pressures (initial, bubble point or bottom hole pressures), constant pressure differences and constant ratio of pressure differences are tried initially in this study. Discussing the results of this initial screening study is out of scope of this article. It is observed that none of the them is the valid criteria for characterizing the performance (oil rate, GOR and
Pi PVT Pwf Rpc Rsb Rsi Rsw X Y Z
77
Initial Reservoir pressure (psi) Pressure Volume Temperature (dimensionless) Flowing Bottom Hole pressure (psi) Cumulative Gas Oil ratio (SCF/STB) Gas Oil ratio at Bubble Point Pressure (SCF/STB) Initial Gas Oil ratio (SCF/STB) Gas Oil ratio at Flowing Bottom Hole Pressure (SCF/ STB) Reservoir Dimension in X-direction (ft) Reservoir Dimension in Y-direction (ft) Reservoir Thickness (ft)
oil recovery) but combining them in particular manner can exhibit a conclusive relationship which can predict qualitative behavior of production performance. The ratio of pressure differences (fp) is first examined as shown by Eq. (1).
fp =
(Pb − Pwf ) (Pi − Pwf )
(1)
The oil rate and oil recovery can be explained using ratio of pressure difference factor (fp) (shown in Eq. (1)). However, without incorporation of GOR into the factor, it is ineffective to correctly predict the behavior of the produced GOR. This fact will be established in the later part of the article as we analyze the results. The depletion of long transient state reservoir can be represented by the extent of the difference between the bubble point pressure and flowing bottom hole pressure since the average reservoir pressure for this kind of reservoir is not representative of the behavior. The difference in the GOR at Pwf and GOR at Pb is one characteristics parameter to predict the behavior of produced GOR from ultra-low permeable fractured reservoir. All these factors are accommodated with Eq. (1) into a single characteristic factor, GOR characteristic index (GCI) as given by Eq. (2).
( (
Pwf
1− P ⎛ R ⎞ b GCI=⎜ 1 − sw ⎟ ⎝ Rsb ⎠ 1−Pwf P i
) )
(2)
The factor, GCI, can be used only when flowing bottom hole pressure is less than the bubble point pressure. When flowing bottom hole pressure is above the bubble point pressure, the well produces initial GOR because reservoir is undersaturated. The value of GCI is presented in Table 1 for the different cases in this study. The effect of this factor is clearly observed in the qualitative behavior of produced GOR which is discussed later.
3. Experimental design The common industry belief is that the drawdown i.e. the difference between the reservoir pressure and flowing bottom hole pressure is the key factor to dictate the production behavior. In this study, the significance of location of bubble point pressure is also considered. To get rid of the established idea of drawdown, the experiments are designed in such a way that drawdown i.e., the difference between initial reservoir pressure and flowing bottom hole pressure are same for all runs. The same drawdown for all experiments ensures that drawdown is not affecting the changes in production performance. The location of bubble point pressure with respect to flowing bottom hole pressure and initial reservoir pressure is a significant factor in the production performance of oil and gas. To create different scenarios of experiment,
78
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Table 1 Design of experiments by varying Rsb, Pi and Pwf. Sr. no.
Rsb (SCF/STB)
Pi (psia)
Pb (psia)
Pwf (psia)
(Pi Pb) (psia)
(Pb Pwf) (psia)
(Pi Pwf) (psia)
fp
GCI
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
500 1000 2000 500 1000 2000 500 2000 2000 500 1000 790 2500 500 1000 1000 500 2000
5765 7015 9515 5300 5300 5300 6017 9767 9863 6113 6650 6000 8500 5200 5500 5150 4865 4871
1265 2515 5015 1265 2515 5015 1265 5015 5015 1265 2515 1990 6265 1265 2515 2515 1265 5015
965 2215 4715 500 500 500 1217 4967 5063 1313 1850 1200 3700 400 700 350 65 71
4500 4500 4500 4035 2785 285 4752 4752 4848 4848 4135 4010 2235 3935 2985 2635 3600 144
300 300 300 765 2015 4515 48 48 48 48 665 790 2565 865 1815 2165 1200 4944
4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800 4800
0.0625 0.0625 0.0625 0.1594 0.4198 0.9406 0.0100 0.0100 0.01 0.01 0.1385 0.1646 0.5344 0.1802 0.3781 0.4510 0.2500 1.0300
0.0683 0.0209 0.0071 0.4085 0.7129 0.8976 0.0018 0.0002 0.0 0.0 0.0974 0.1984 0.2975 0.5124 0.6002 0.7997 0.9227 0.9891
Fig. 1. The relative values of initial reservoir pressure, bubble point pressure and flowing bottom hole pressure on Rs-PVT diagram (a) case 1, operation above bubble point pressure, initially undersaturated reservoir (b) case 2, operation below bubble point pressure, initially undersaturated reservoir (c) case 3, operation below bubble point pressure, initially saturated reservoir.
initial reservoir pressure, bubble point pressure, flowing bottom hole pressure and GOR (Rsb) at bubble point are varied. Initially we divided the simulations into various categories as discussed in the
next section to inquire the factor(s) which would characterize the qualitative behavior of production performance mainly GOR. Different categories are shown in Table 1 and discussed here.
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
3.1. Constant pressure differences The difference between initial reservoir pressure and bubble point pressure (Pi Pb) and the difference between bubble point pressure and flowing bottom hole pressure (Pb Pwf) are kept constant for serial numbers 1–3 as shown in Table 1. This is known as constant pressure difference (Δp) case.
79
rate which is not economic. The most practical location of the initial reservoir pressure, bubble point pressure and flowing bottom hole pressure for economic production is case 2. Wells are normally operated well below the bubble point pressure to obtain the economic production rate.
4. Reservoir simulation 3.2. Constant end pressures For serial numbers 4–6 in Table 1, initial reservoir pressure and flowing bottom hole pressure are constant. This can be called constant boundary pressures (or constant P) cases. In constant pressure cases (serial no. 4–6 in Table 1), the difference between initial reservoir pressure and bubble point pressure (Pi Pb) and the difference between bubble point pressure and flowing bottom hole pressure (Pb Pwf) vary depending on the location of bubble point pressures. 3.3. Special cases Serial number 7–18 are special cases to study the effect of characteristic factor, GCI which is defined in Eq. (2). In the special cases, Pi, Pb and Pwf are taken such a way that GCI covers a wide range i.e., from near zero to one as shown in Table 1. This ensures that all possibilities are considered in this study. In some cases such as serial number 9 and 10, well is operated above bubble point pressure i.e., flowing bottom hole pressure is greater than Pb as shown in Fig. 1a. The GCI becomes zero for these cases as shown in Table 1. In these cases, reservoir remains in undersaturated, and produced GOR is always equal to the initial GOR (Rsi). Well can be operated at different pressure depending on initial reservoir pressure and bubble point pressure. Similarly, the values of fp and GCI vary greatly on the relative locations of Pi, Pb and Pwf. Pressures (Pi, Pb and Pwf) in all the 18 cases given in Table 1 can be located in P versus Rs diagram. This relative locations lead to three different possibilities for the ratio of pressure difference as described in Fig. 1a–c. To understand all cases given in Table 1 in the context of Fig. 1, serial numbers are displayed in the figure along with the corresponding values of fp and GCI. Operating well near bubble point pressure suppresses gas production but it also yields very low oil Table 2 Reservoir and completion parameters used in all of the simulations. Reservoir top (ft): Matrix permeability, km (nD): Reservoir dimensions X(ft), Y(ft), Z(ft): Fracture permeability (mD): Fracture width (ft): Fracture height (ft) Fracture orientation Initial oil saturation (%): Reservoir porosity (%):
12,800 10, 100, 1000, 5000, 1000, 750, 200 kfx ¼ kfy ¼ 150; kfz ¼300 0.05 Reservoir height Parallel to YZ plane 84 5
Based on field data available for Eagle Ford from Railroad Commissions of Texas, different initial GORs (300, 500, 1000, 1500, 2000 and 2500 SCF/STB) are selected to represent various reservoir fluids. Ultra-low permeable reservoirs (10, 100, 1000 and 5000 nD) with one horizontal well and one vertical fracture located in the middle of the reservoir are simulated. The fracture height and fracture width are equal to the reservoir height and width respectively. The reservoir is extended to 2000 ft in the xdirection, 750 ft in the y-direction, and 200 ft in the z-direction; with a top depth of 12,800 ft The properties of the reservoir are summarized in Table 2. The matrix permeability, initial reservoir pressure, flowing bottom hole pressure and fluid properties are varied while the fracture permeability, fracture width, fracture orientation, matrix porosity and initial hydrocarbon saturation remain constant. Minimum grid block size (Panja et al., 2013) is used to obtain converged results to get rid of any grid effects.
5. Validation of the factor, GCI Validation is an integral part of model development. Any model should be validated before it is accepted for future application. Real field data is from Eagle Ford is used to check the applicability of GCI. The production data are collected from Railroad Commissions of Texas. The reservoir and fluid parameters such as initial reservoir pressure, bubble point pressure, initial GOR (Rsi) and operating conditions are not available in public data base. These parameters are generated by superimposing simulated gas rate and oil rate with field data. Various cases are simulated with different Pi, Pb, Pwf, Rsi and reservoir permeability as shown in Table 3. Prior knowledge of the field (Panja et. al., 2015) is applied during choosing these parameters. Two different flowing bottom hole pressures are investigated for each field case. The pairs, (1,2), (3,4), (5,6) and (7,8) have same properties except the flowing bottom hole pressure. Comparisons of field data from Eagle Ford with simulation results for eight cases ( Table 3) are shown in Fig. 2. It should be noted that the rates are for non-interfering single fracture and net to gross ratio (NTG) is considered as 0.4 for Eagle Ford. If there are five clusters per stage in a 16-stage well, with about 80 possible fractures, the rate for 50 nd to 450 nD reservoirs is expected to be between 2.5 mstb/month and about 5 mstb/ month in the first year. This depends on the initial dissolved GOR,
Table 3 Plausible reservoir conditions and operational parameter for various lease in Eagle Ford. Lease unit
Simulation case
Pi (psi)
Rsi (SCF/STB)
Pb (psi)
Pwf (psi)
Permeability (nD)
GCI
Hawn holt
1 2 3 4 5 6 7 8
3000 3000 5300 5300 5300 5300 8000 8000
300 300 800 800 950 950 1650 1650
765 765 2015 2015 2390 2390 4140 4140
100 200 650 950 750 1000 1000 1250
450 450 200 200 400 400 50 50
0.80 0.60 0.53 0.34 0.55 0.42 0.66 0.58
JP head bower C Meyer Davenport
80
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Fig. 2. Superimposition of field data with simulation results for (a) Oil rate (b) Gas rate.
6. Result and discussion The matrix permeabilities chosen for this study are 10 nD, 100 nD, 1000 nD and 5000 nD. Initial reservoir pressure, bubble point pressure, flowing bottom hole pressure and GOR (Rsb) at bubble point are already selected to create a wide range of values of GCI as given in Table 1. Total 72 simulations ( 4 permeabilities for each18 cases given in Table 1) are run to investigate the effect of dissolved GOR curve and operation parameters on production performance in terms of oil rate, oil recovery and produced GOR. Commercial simulator IMEX, a Computer Modeling Group Black Oil simulator, is used to conduct the study.
7. Controlling produced GOR
Fig. 3. Comparison of GOR behavior between field data and simulated results.
drawdown, bubble point pressure and a number of other factors. Analysis of the Eagle Ford oil rate indicates that the reservoirs fall within the 50–450 nD permeability range. It is evident from Fig. 2a and b that the oil rates and gas rates are not very much affected by flowing bottom hole pressures ( in the range used here) for the reservoirs of 50–450 nD permeability. Superimposing GOR with the field data is also necessary to generate the reservoir conditions. The comparison simulated results and Eagle Ford field data of produced GOR is shown in Fig. 3. The GOR values for Eagle Ford are much more variable from 300 to 1650 SCF/STB. GCI is calculated for each case as shown in Table 3. Ratio of cumulative GOR and initial ratio produced from JP Head Bower C Unit and Meyer Unit are less compared to same from Hawn Holt and Davenport unit as shown in Fig. 3. These performances are also correctly predicted by GCI. The values of GCI for JP Head Bower C Unit and Meyer Unit (0.34–0.55) are less than the values of GCI for Hawn Holt and Davenport unit (0.58–0.80). Higher GCI indicates that increase in GOR with respect to initial GOR is higher. Although the initial GOR of Hawn Holt is 300 SCF/ STB only, the rise of GOR (ratio of GOR and initial GOR) with time is higher than the other field. This performance also confirms that initial GOR is not the only factor to decide behavior of the produced GOR. GCI can be used reliably to analyze qualitative behavior of produced GOR.
It is desired to operate wells to control the GOR, hence, prescribing flowing bottom hole pressure is one quest of this study. GCI can also be utilized to propose the flowing bottom hole pressure to operate wells in a manner to optimize the GOR. The flowing bottom hole pressure can be controlled to maintain desired produced GOR by setting the value of GCI. Assuming that the dissolved GOR varies linearly with pressure for saturated reservoir, following formula for the flowing bottom hole pressure is derived as shown in Eq. (3).
Fig. 4. Variability of Pwf/Pb with the factor GCI.
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
⎤ ⎡ ⎤2 ⎛P ⎞ ⎛P ⎞ Pwf ⎡ =⎢ 1 − 0. 5 ⎜ b ⎟ GCI⎥− ⎢ 1 − 0. 5 ⎜ b ⎟ GCI⎥ − (1−GCI) ⎝ Pi ⎠ ⎝ Pi ⎠ Pb ⎣ ⎦ ⎣ ⎦
81
(3)
Knowing the initial reservoir pressure and bubble point pressure of reservoir fluid, the value of flowing bottom hole pressure is prescribed using Eq. (3) to maintain desired produced GOR. The variation of Pwf/Pb with the factor, GCI is shown in Fig. 4. The relationship between Pwf/Pb and GCI are established for various ratios of Pb/Pi. It is clear that Pwf/Pb is inversely proportional to GCI. Higher value of GCI predicts lower value of Pwf/Pb i.e., lower value of flowing bottom hole pressure compared to bubble point pressure which leads to produce more gas. The curve approaches to linear when Pb/Pi moves towards unity. Little or no deviation of produced GOR compared to initial GOR is observed if well is operated at the pressure higher than 0.5 Pb to 0.7 Pb depending on the Pb/Pi fraction. For example, if any well in a reservoir with initial pressure of 5500 psi and bubble point pressure of 3500 psi must be operated at pressure higher than about 0.65 Pb i.e., 2275 psi ( as predicted by Eq. (3) and Fig. 4) to keep the produced GOR near to initial GOR, but rate of oil production may not be economic for this high flowing bottom hole pressure (2275 psi). If factor value of 0.6 is a chosen, the flowing bottom hole pressure is near 0.3 Pb i.e., 1050 psi which will enhance the rate of oil production in expense of higher produced GOR. Optimization of oil rate depends on the operators considering many factors like ability to handle gas, long term production strategy etc. 7.1. Effect of initial gas oil ratio (Rsi) The effect of initial GOR on produced GOR, oil rate and oil recovery are discussed for three different initial GORs (500, 1000 and 2000 SCF/STB) while ratio of pressure difference is kept constant at 0.0625 although GCI varies. Slope of dissolved GOR curve(dRs /dp) of 0.4 and four reservoir permeabilities (10, 100, 1000 and 5000 nD) are considered. The effect of initial GOR on cumulative GOR is shown in Fig. 5a and b for the ratio of pressure difference is 0.0625. It is intuitive that higher GOR is produced from reservoirs with higher initial GOR. The deviation of GOR from initial GOR is not noticeable with different matrix permeabilities for very low value of GCI (0.007–0.07). The differences in the oil rate for different initial GORs are not significant for lower matrix permeability (10 and 100 nD). Higher oil rate is achieved with higher initial GOR and with higher reservoir permeability. Higher initial gas dissolved
Fig. 6. Effect of initial gas oil ratio on oil recovery with constat fp of 0.0625.
in oil phase always supports the production by sustaining the reservoir pressure for long time. The mobility of oil is also improved with dissolved gas by making oil phase lighter. Simultaneously, more gas is produced on the surface as dissolved and free gas. Higher initial GOR finally enhances the oil rate in cost of higher gas production. The oil recovery is also improved from higher initial GOR as evident in Fig. 6 for the ratio of pressure difference of 0.0625. The initial GOR effect on oil recovery is very much noticeable for higher permeability reservoirs (5000 and 1000 nD). Highest oil recovery is obtained from reservoir with initial GOR of 2000 SCF/ STB and reservoir permeability of 5000 nD. As described earlier, the higher initial dissolved gas provides higher energy in the reservoir, thus helps to produce more oil. Lowest amount of oil is recovered from the reservoir with initial GOR of 500 SCF/STB. The effect of initial GOR on oil recovery is not very prominent for 10 nD reservoir. 7.2. Effect of GCI The effect of GCI is already established in earlier sections with
Fig. 5. Effect of initial gas oil ratio with constat fp of 0.0625 on (a) cumulative gas oil ratio (b) Oil rate.
82
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Fig. 7. Cumulative gas oil ratio for (a) GCI of 0.054, 0.068 and (b) Undersaturated reservoir, GCI¼0 (Dash line: dRs/dp¼ 0.8 (SCF/STB)/psi, Solid line: dRs/dp¼ 0.4 (SCF/STB)/psi).
validation with field data. The factor GCI is of particular interest for quantitative analysis of produced GOR. Behavior of produced GOR is dependent not only on the operating conditions but also on the fluid properties and initial reservoir conditions. The qualitative nature of produced GOR cannot be explained solely by the ratio of pressure difference, fp. The factor, fp is useful to explain the oil rate and oil recovery only for different initial GOR. Produced GOR is normalized with initial GOR to compare among the fluids having different initial GOR (500, 1000 and 2000 SCF/STB) in the same scale. As shown in Fig. 4, the deviations of produced GOR depend on value of GCI. The simulated results are analyzed here for various GCI. 7.3. Little or no deviation of GOR Produced GORs have neutral deviation when produced GORs vary within 72% of initial GOR. The neutral deviations are depicted in Fig. 7a and b. The neutral or little deviations are found when the factor (GCI) is not greater than 0.07. A special case for neutral deviation can be identified when the flowing bottom hole pressure is above the
bubble point pressure i.e. reservoir is in undersaturated conditions. Neutral deviation occurs when the flowing bottom hole pressure is close to bubble point pressure. Sufficient amount of gas does not evolve out of oil phase and the oil phase contains initial amount of dissolved gas only. 7.4. Positive deviation of GOR Positive deviation of GOR is commonly observed in the field. Positive deviations of produced GOR are shown in Fig. 8a and b. Positive deviation is noticed when large portion of reservoir is below bubble point pressure and flowing bottom hole pressure is sufficiently below bubble point pressure. In these reservoir conditions, gas and oil both phases flow simultaneously. The values of GCI are greater than 0.2 for positive deviation. Effect of GCI on cumulative RATO in terms of deviations is summarized in Table 4.
8. Conclusion The complex behaviors of produced GOR, oil rate and oil
Fig. 8. Positive deviation of cumulative gas oil ratio for GCI of (a) 0.24, 0.41, 0.41, 0.92 and (b) 0.89, 0.99 (Dash line: dRs/dp¼ 0.8 (SCF/STB)/psi, Solid line: dRs/dp ¼0.4 (SCF/ STB)/psi).
P. Panja, M. Deo / Journal of Petroleum Science and Engineering 144 (2016) 76–83
Table 4 Effect of GCI on cumulative GOR. Range of GCI Comments cumulative GOR Below 0.07 Above 0.2
No or little deviation Positive deviation
recovery from ultra-low permeable fractured reservoir are illustrated with various PVT properties by varying important parameters like initial dissolved GOR slope of dissolved GOR, operating conditions and reservoir conditions. A single universal predictive factor – GCI, which helps understand the GOR behavior of lowpermeability reservoirs is defined. The following conclusions are made; Effect of Rsi
Higher initial dissolved GOR yields higher produced GORs. Effect on oil rate is not very significant. Higher oil recovery is obtained from higher initial GOR and higher reservoir permeability. The value of using GCI – the single universal factor
Little deviations from initial dissolved GOR are observed in production GOR when factor is less than 0.07.
Positive deviations of produced GOR are observed when factor is greater than 0.2.
Higher deviations (positive and negative) are noticed from reservoir with lower permeability reservoir. GCI is effective measure of the qualitative behavior of produced GOR accommodating reservoir conditions, fluid properties and operating condition. Well can be operated to produce desired GOR by using the value of GCI prescribed for particular GOR window. The value of GCI is specific to the other parameters like relative permeability, critical gas saturation, porosity of reservoir etc. but its characteristic application on produced GOR would be identical.
Acknowledgments The authors gratefully acknowledge the academic license to CMG products from Computer Modeling Group, Calgary, Canada. Financial support to Mr. Palash Panja through the Conoco Phillips Fellowship is also acknowledged.
83
References Al-Khelaiwi, F.T., Muailu, H.A., Alfakher, A., 2014. Gas Oil Ratio (GOR) Estimation Using Downhole and Surface Well Sensors. Albertson, M., Schaeffer Jr., W.A., 1931. Control of gas-oil ratios in the yates field, Pecos County, Texas. J. Soc. Pet. Eng. Arps, J.J., Roberts, T.G., 1955. The Effect of the Relative Permeability Ratio, the Oil Gravity, and the Solution Gas-Oil Ratio on the Primary Recovery From a Depletion Type Reservoir. Brinkman, F.H., Weinaug, C.F., 1956. Calculated Performance of a Dissolved Gas Drive Reservoir by a Phase Behavior Method. Busahmin, B.S., Maini, B.B., 2010. Effect of solution-gas-oil-ratio on performance of solution gas drive in foamy heavy oil systems. Canadian Unconventional Resources and International Petroleum Conference, Society of Petroleum Engineers, Calgary, Alberta, Canada. Jones-Parra, J., Reytor, R.S., 1959. Effect of Gas-Oil Ratio on the Behavior of Fractured Limestone Reservoirs. Khoshghadam, M., Khanal, A., Lee, W.J., 2015. Numerical Study of Impact of NanoPores on Gas-Oil Ratio and Production Mechanisms in Liquid-Rich Shale Oil Reservoirs, Society of Petroleum Engineers. Lawal, K.A., Uwaga, A.O., Osoro, O.F., 2006. A Systematic Methodology for Extrapolating Gas-Oil Ratio During Declining Oil Production, Society of Petroleum Engineers. Levine, J. S., Prats, M., 1961. The Calculated Performance of Solution-Gas-Drive Reservoirs. Marsh, H.N., Robinson, B.H., 1929. Means of Controlling Gas-oil Ratio. Millikan, C.V., 1926. Gas-Oil Ratio as Related to the Decline of Oil Production, with Notes on the Effect of Controlled Pressure. Muskat, M., 1945. The production histories of oil producing gas-drive reservoirs. J. Appl. Phys. 16, 147–159. Panja, Palash, C, T., Deo, Milind, 2015. Factors controlling production in hydraulically fractured low permeability oil reservoirs. Int. J. Oil Gas Coal Technol. Panja, P., Conner, T., Deo, M., 2013. Grid sensitivity studies in hydraulically fractured low permeability reservoirs. J. Pet. Sci. Eng. 112, 78–87. Prats, M., Levine, J.S., 1963. Effect of vertical fractures on reservoir behavior- results on oil and gas flow. J. Soc. Pet. Eng., SPE-593-PA Sullivan, R.J., 1937. Gas-Oil Ratio Control in Flowing Wells. Eighteenth Annual Meeting, American Petroleum Institute, Cicago.