Estimation of Initial Oil in Place for Buzurgan Oil Field by Using Volumetric Method and Reservoir Simulation Method

Abstract


Introduction
In oil and gas exploration and production, it is essential to have a reliable estimation of Initial Oil in Place (IOIP).Reliable reserve estimation depends on the petrophysical data used in the volumetric method for estimating IOIP as well as in the simulation method (Saleh et al., 2019).Well logs are wellknown in the petroleum sector for their importance in oil and gas exploration and reservoir appraisal.The quantity of oil and gas in the subsurface is calculated using well logs.Porosity, hydrocarbon saturation, thickness, area, and permeability are the most important petrophysical criteria to consider when evaluating a reservoir.Furthermore, reservoir geometry, formation temperature and pressure, and lithology can all have a part in a reservoir's appraisal, completion, and production (Egbai et al., 2012).In current study, neural network technique is used to determine the reservoir petrophysical properties depending on data of core and well log from 13 wells at Buzurgan oilfield.Consequently, calculating IOIP requires a petroleum engineer to verify the value of rock porosity that may containing the oil, saturations of water and oil, reservoir thickness and the ratio of net to gross of the reservoir (Saleh et al., 2019).Reserve management, exploitation, exploration, and production all rely heavily on reservoir estimates.The purpose of reserve estimation is to primarily study a reservoir in order to estimate, appraise, and analyze the IOIP as well as the reservoir's historical and present performance (Omoniyi and Adeolu, 2014).There are many methods to calculate volume of oil in place such as volumetric method, material balance method, decline curve analysis method and simulation method (Ali et al., 2018).The method chosen is determined by the timing of development and production, the quantity of data available, and the features of the reservoir.
Volumetric Method is estimated IOIP depending on reservoir rock properties of the selected reservoir, including its porosity, net to gross and initial water saturation as well as reservoir thickness.On the other hand, the Material Balance Method, is a dynamic method that predicts IOIP depending on production and pressure data.A Decline Curve Analysis Method (DCAM) such as Draw Down Test (e.g.Limit Test) can be able for estimating IOIP (Saleh et al., 2019).However, there is another a modern and quick method that can be used to calculate (IOIP) which is the simulation method (Ali et al., 2018).The volumetric technique is concerned with assessing reserves based on volumetric rock data, such as area, thickness, and pore volume.In addition to other key information such as water saturation, and fluid content inside the pore volume.This yields a volumetric unit or ton estimate of the quantity of hydrocarbons in situ.Because not all of the oil in the reservoir can be retrieved, the eventual recovery can be predicted using an appropriate recovery factor (Robert et al., 2016).
The objective of current study, is to determine the IOIP for Mishrif Reservior-Buzurgan oilfield by volumetric method and simulation method using Petrol 2017 software during increase the number of drilled wells to more than 100 wells as wells as the newest data from these wells.Also, the comparison between the results of current study and last previous study is be done.

Volumetric Method
Estimating IOIP by volumetric method can be working directly afterward first exploration processes, before start of production.So, this method depends on determination of reservoir volume that based on dividing the reservoir geophysical maps into finer bulk grids or parting the contour map area into pieces (Ali et al., 2018).The reliability of IOIP forecasting is highly dependent on the knowledge of regional geology and the quality of the seismic analysis, both of which will improve as more wells are drilled and more precise descriptions and geologic and petrophysical maps of reservoirs become available (Saleh et al., 2019).The dimensions of the reservoir, pore volume inside the rock matrix, and the fluid content within the void space are all needed for the volumetric technique.This will provide an accurate estimation of the hydrocarbons in place, from which may calculate the ultimate recovery by using the recovery factor (Robert et al., 2016).Thus, equation 1 typically used to calculate IOIP by volumetric method.
The petrophysical properties for reservoir understudy is calculated using neural network technique from 13 cored and 29 logged wells.The interactive petrophysics (IP) software was used to apply the neural network technique.The needed data was gamma ray (GR), resistivity log (deep and shallow), bulk density (RHOB), sonic log (DT), neutron log (NPHI), core porosity and Core permeability.In addition, neural map software is used to digitize the structural maps of MB21 unit of Mishri reservoir for north and south dome.
Typically, the calculation of IOIP which affected by many parameters such as reservoir porosity, water saturation, net to gross, oil formation volume factor and, depths of oil-water-contact.Consequently, the understanding of the impact of the petrophysical properties will help for accurate estimation of IOIP, and for saving time and money while field exploration and development period.In previous study by (CNOOC Iraq Ltd.) Company, the estimated IOIP was approximately equal to (741.7*10 6 sm 3 ) depending on data getting from less than 30 wells (CNOOC, 2013).

Reservoir Simulation Modelling
Generally, precise value of oil initial in place can be estimated under reservoir conditions by simulation method.Thus, reservoir engineers can take an appropriate understanding of reservoir behavior with precise prediction for future estimation and making investment decisions (Ali et al., 2018).In this study, a black oil reservoir simulator has been used with the target of modelling and simulating the reservoir (West Virginia University, 2014).Petrel 2017 software is a useful software which utilized to calculate volumes of reservoir fluid oil recovery and making oil prediction.In addition to petrophysical properties, the needed data was counter map, well head and well tops for building surfaces and layering of all units.In addition, PVT analysis reports as well as capillary pressure and relative permeability versus water saturation reports were organized for simulation method.As the surfaces of all Mishrif reservoir unit were created, the simple gridding step was done depending on the units surfaces and specific increment, so the grid dimensions is set as X=50 m and Y=50 m, the layering processes take account of dividing Mishrif's units in sublayers where MA, MB11, MB12, MB21, MC1 and MC2 units is divided into MA-1, MA-2, MA-3, MB11-1, MB11-2, MB12-1, MB12-2, MB21-1, MB21-2, MB21-3, MB21-4, MB21-5, MB21-6, MB21-7, MB21-8, MC1-1, MC1-2, MC1-3, MC1-4, MC2-1, MC2-2, MC2-3, and MC2-4 respectively, based on a variety of petrophysical parameters.Formation evaluation results which include effective rock porosity (φ), water saturation (Sw) and net to gross ratio (N/G) were up-scaled by arithmetic averaging method.In addition, property modelling step is achieved using Sequential Gaussian Simulation which is an adequate averaging method allowing for finer scale heterogeneity and better control as well as prediction of porosity, water saturation and net to gross in regions far away from the well control points., then the oil water contact was initiated for all units, and finally, the volume calculation is obtained depending upon all previous process as well as the value of oil formation volume factor.

Area of Study
The Buzurgan oil field is located South-Eastern part of Iraq close to Iran border, 40 Km Northeast from Amara as shown in Fig. 1.The Mishrif Formation is an important reservoir in southern Iraq.In fact, the structure of Buzurgan field has two domes which are north and south domes.The dimensions of north dome is 16 km * 6 km while the dimensions of south dome is 23 km * 8 Km (Hussein, 2012).The Mishrif Formation have been bounded by two formations, which are Khassib Formation at top while the Rumaila Formation at bottom.Henec, the existed wells are 85 producer, and 16 injectors.

Materials and Methods
The systematic plan of this research concerned many steps in orders to achieve accurate estimating of IOIP.

Petrophysical Properties
The steps to precise determination of petrophysical properties included determination of shale volume, porosity, and water saturation.Environmental correction should be done for selected logs data.Shale volume determination was a key step to detect the shale volume that effects on the calculation of the porosity and water saturation.Corrected porosity (φCorr) was calculated using neural network technique depending upon the log porosity (φL) as a results of interpretation obtained from log data such as GR, shale volume, resistivity log, sonic, neutron and density log porosity as well as core porosity (φC).Water Saturation (Sw) was obtained using Archie equation.Permeability (K) calculation is an important property because of its effect on the production rate of hydrocarbon therefore the neural network technique was used to estimate the permeability depending on log data, φCor and core permeability (KC).Determination of cut-off petrophysical parameters are the crucial control for processing the reserve of oil in the reservoir, therefore the calculation of cut-off values for permeability, porosity, water saturation and shale content were obtained depending on plotting many crossplot between K vs. φCor, φCor vs. Sw and φCor vs. Vshale.Thus, the cutoff values for permeability, porosity, water saturation and shale content are 0.1 md, 8%, 65% and 23%.The Net to Gross (N/G) is calculated depending on an acceptable cutoff values for permeability, porosity, water saturation and shale content.

Geological Modeling
To estimate the IOIP, a high resolution three-dimensional geological model for Mishrif Formation has been built by using Petrel-2017 software.This program is mainly used for 3D geo-modeling as an input for successive simulation models (Schlumberger, 2017).In the present study, building of 3D geological model is defined in the following steps:

Data preparation
The first step for building geological model is preparing the related subsurface data which gained from the drilled wells into the Mishrif reservoir during the development stage such as unit's tops, well's target depth, contour maps, well type, inclination, and azimuth.In addition, the result of petrophysical calculation from cores and logs and the fluids contact are needed.

Construction of structural model
The construction of geological model is needed to install the accurate and uniform boundary limits.Thus, the polygon lines are utilized within the geologic model.Fourteen 2D geological surfaces are created, indicating the top of all Mishrif Formation units and its barriers (Mishrif, MA, b1, MB11, b2, MB12, b3, MB21, b4, MC1, b5, MC2, b6, and Rumaila), as illustrated in Table 1.
Three-dimensional grid structure of Mishrif Formation involves of 212 grid cells through x-axis, 138 grid cells through the y-axis, and 29 sublayers through the vertical direction.The total number of grid cells is 848,424.The increment for areal gridding size was established to be 50m×50m, as illustrated in Fig. 5.The horizons are generated for all available layers in Mishrif reservoir.Consequently, 14 horizons and 13 zones have been built for all geological surfaces that have been generated as shown in Fig. 6.The segmentation of horizons of Mishrif reservoir units by layering every single zone and subdivided it into many small layers.These small layers will increase the number of fine grids in 3D grid system to represent the high vertical heterogeneity clearly, as shown in Fig. 7.

3D Modeling of petrophysical properties
The arithmetic averaging approach was used to scale up the porosity, net to gross, and water saturation scaling up for each well's well log data.The geometric averaging approach, on the other hand, is employed for permeability scaling up.This geometric averaging is very sensitive to modest reservoir permeability values and provides accurate findings using a one-dimensional averaging approach, then apply appropriate statistical method for estimating and averaging the log data for every un-sampled grid cells.
Consequently, the porosity model was generated depending on the results of porosity that are determined from well logs, confirmed and matched with porosity from core measurements.In this study, the statistical technique used is Sequential Gaussian Simulation (SGS) algorithm.This method allows for finer scale heterogeneity control as well as forecasting the porosity in regions far away from the drilled wells, as shown in Fig. 8 for MB21 unit.Fig. 9 presents the histogram of porosity for MB21 unit of Mishrif reservoir where the x-axis represent the porosity values and the y-axis represent the number of porosity values repetition.Consequently, the porosity values for MB21 unit shows a good matching between imported well logs calculated, and up-scaled porosities.However, it has been remarked that the porosity in most of Mishrif reservoir units increase in the crest of north and south dome and decreases in direction to the flanks.The porosity values for all units are shown in Table 2.In regard of water saturation model, ninety-four wells in Mishrif Formation is upscaled, as shown in Fig. 10 for MB21 unit.The minimum water saturation can be observed in the crest of South dome and northeastern parts of North dome whereas the maximum water saturation is located in the flanks in most of Mishrif's units.Table 3 presents the mean water saturation values for every units.In addition, net pay is an important parameter in building the geological model, therefore it contributes for estimating the IOIP volume and so non-reservoir rock will not be characterized during the calculations (Worthington, 2009).Fig. 11 shows net to gross (N/G) property modelling for MB21 unit.The result gives a good indication for increasing of N/G in the crust of South and North domes due to good properties (porosity and water saturation) and decreasing towards Middle part of reservoir.Table 4 presents N/G mean values for all six Mishrif units.

OWC model
Fluid contacts is considered one of the important parameters for forecasting the hydrocarbons initial in place as well as in redevelopment planning methods (Bora et al., 2012).the OWC model is created depending on the water saturation data versus depth which are already interpreted by IP software.Water saturation data plotted versus depth for every unit to detect the accurate OWC.The average OWC for all units at both north and south domes were generated depending on 29 wells as shown in Table 5. OWC surfaces model are created and extended throughout the Buzurgan oilfield, as shown in Fig. 12.

Results of Estimation of Initial Oil in Place (IOIP)
Estimation of IOIP is considered the main target of current study.Thus, IOIP is calculated by using two method; Volumetric Method and Simulation Method.

Static Model
The IOIP is essentially established using volumetric method.The calculation depends upon petrophysical property models, oil water contact model, and initial oil formation volume factor.The volumetric method was utilized to determine the initial hydrocarbon in place for every unit into Mishrif reservoir.In the present study, the result shows that IOIP equal to 731*10 6 surface cubic meters (sm 3 ) for whole Mishrif reservoir.Table 6 present the IOIP value for every Mishrif 's unit individually.

Reservoir Simulation Method
In the current work, reservoir dynamic model was used to determine the IOIP for dynamic simulation model.The reservoir simulation method is a modern method in petroleum industry to calculate the IOIP and making prediction and history matching as well, the software is more acceptable since it generate any other information in case its missing.
Up to December 2020, more than 100 wells have been drilled at Mishrif reservoir/Buzurgan oilfield.A reservoir simulation model was constructed depended upon all oil production and water injection vertical and horizontal wells which drilled from 1976 to 2020.In addition, accurate data required are, capillary pressure versus water saturation, and units OWC.Typically, the adjustment of Pc values was a crucial factor to modify the IOIP result.Thus, the accurate Pc values presented in Table 7.One of the most important variables in estimating hydrocarbon reserves is having a precise understanding of the capillary pressure distribution on drainage processes as capillary pressure data is largely utilized to determine initial fluid contacts and transition zones.Thus, the adjustment of Pc curve a a result of higher field water production rate and field water cut values from simulated model before adjusting than observed data, therefore multiply the Pc values by factor 0.84 reduced the field water production rate and field water cut values to match the observed data.
The history matching between simulated model and observed data is obtained with good results.Thus, the static wells pressure matching result are achieved for Bu-3, Bu-4, and Bu-13, as presented in Fig. 13.In reservoir engineering, capillary pressure data is largely utilized to determine initial fluid contacts and transition zones.

Discussions
Once the simulated model has been established, the result of IOIP was obtained with acceptable result which is equal to 746*10 6 Sm 3 .As south dome is bigger than north dome in bulk volume as well as MB21 unit own a good petrophysical properties, IOIP in south dome is founded to be higher than in north dome of Mishrif Formation.Consequently, the result of IOIP is close to that calculated by volumetric method.Layer MB21 is considered the main unit of IOIP which have 71.8% of whole IOIP of Mishrif reservoir while MA, MB11, MB12, MC1 and MC2 units represent 1.23%, 7.8%, 0.27%, 12.9% and 6%, respectively.However, the accurate history matching is achieved depending on collective data of the oil and water production, water injection, fluid properties and well log data.Finally, the history matching was a crucial issue that make simulated method is more accurate than volumetric method.Thus, to perform the future reservoir prediction strategies as well as water, polymer or surfactant flooding to improve sweep efficiency and increase oil recovery depending on the outcomes of reservoir geological and the accurate history matching of simulated model of current study.

Conclusions
two different methods were used to calculate the IOIP for the Mishrif reservoir; volumetric and simulated methods.The achieved results from the volumetric method were equal to 731*10 6 sm 3 and relatively close to results obtained from simulation method which was equal to 746*10 6 sm 3 with a small percentage error equal to 2%.In the current study, the modification in capillary pressure values was very important to get an accurate estimate of IOIP while comparing with result of the volumetric method.However, the results showed an increase in estimated IOIP values reach to 0.6% during used more data obtained from 100 wells.MB 21 has good reservoir characteristics, according to the distribution of petrophysical properties as porosity and water saturation.MB21 unit is founded to be own highest IOIP equal to 525*10 6 sm 3 whereas MB12 unit is founded to be own lowest IOIP equal to 2*10 6 sm 3 .Finally, the simulation method can provides more accurate results while calculation volume of oil in place as a result of adjusting of capillary pressure versus saturation values which reduced by 16% to achieve the best matching between the simulated model and a real observed data for pressure, production and injection fluid which reflect true situation of reservoir behaver as well as maintain the IOIP at real value.

Table 1 .
Units, barriers and sublayers for the Mishrif reservoir

Table 2 .
Porosity values for all Mishrif units

Table 3 .
Water saturation values for all Mishrif units

Table 4 .
N/G values for all Mishrif units

Table 5 .
Average OWC for Mishrif Reservoir at North and South domes.

Table 6 .
IOIP for Mishrif reservoir units by Volumetric Method.

Table 7 .
Capillary Pressure (Pc) before and after Adjustment values.