Noirot, J.C. (Shell International Exploration and Production B.V.) | van den Hoek, P.J. (Shell International Exploration and Production B.V.) | Zwarts, D. (Shell International Exploration and Production B.V.) | Bjoerndal, H.P. (Shell International Exploration and Production B.V.) | Stewart, G. (Shell International Exploration and Production B.V.) | Drenth, R. (Petroleum Development Oman) | Al-Masfry, R. (Petroleum Development Oman) | Wassing, B. (Petroleum Development Oman) | Saeby, J. (Petroleum Development Oman) | Al-Masroori, M. (Petroleum Development Oman) | Zarafi, A. (Petroleum Development Oman)
With the realization that water injection is generally taking place under fracturing conditions, tools capable of better modelling fractured injection and its impact are being developed. Models integrating rock (fracture) mechanics and traditional reservoir simulation are now applied to water injection projects with a number of applications in the Middle East. Fracture dimensions are a key input to those models. Monitoring techniques to track the evolution of induced fractures with time are also being deployed. Amongst those techniques microseismic and specific fall-off test procedures are used.
Asphaltene precipitation is a nearby wellbore phenomenon with strong adverse effects in rock permeability affecting the production operations. However, a complete characterization of the asphaltene precipitation effect, namely quantifying the degree, scale and mechanisms of permeability reduction still remains to be resolved. In this paper we report the results of a series experiments on characterizing the asphaltene precipitation effect on permeability reduction. The experiments are also unique in providing the precipitation effects on the end point relative permeability at Siw.
For quantifying the degree and scale of asphaltene precipitation a set of experiments are designed to induce permeability reduction along the flow direction. Then the plugged and unplugged pore ratios were quantified using SEM images taken at five different points from the core inlet. In all experiments, total plugging has occurred between 40 to 60-pore volume injection. It has also been found that within a distance of 0.3 total core length, approximately 80 % of the plugging has occurred indicating the ‘snowballing' nature of mechanical entrapment.
Then the two major mechanisms of asphaltene deposition namely dynamic asphaltene adsorption and mechanical entrapment were quantified by reversing the flow direction. The reversal of the flow have initially resulted in approximately 80 percent permeability recovery, indicating that 80 percent of the permeability reduction is caused by mechanical entrapment. Then a period of simultaneously acting recovery and impairment mechanisms has been observed finally followed by impairment mechanism only. After the initial recovery period, a consequent flooding of one the cores with toluene were able to establish the original permeability. This additional recovery is attributed to removal of the adsorbed asphaltenes.
A multiphase pump was recently commissioned in a satellite oil field onshore Abu Dhabi. The pump is used to boost the flow from a well, located in a remote desert location, to a crude oil gathering centre, which is located 27 km away.
This paper describes the multiphase pump installations and the unique features of these facilities. Pump specifications and key design considerations are discussed. The paper also addresses the project implementation strategy as well as early operating experience.
Huwaila field is an undeveloped oil reservoir in a remote desert location in the south of Abu Dhabi. The field was discovered in 1965.
The closest production facilities to Huwaila are in a gathering centre in the near by Bu Hasa field. The gathering centre, which is called RDS-3, is located 27 km north of Huwaila field. The terrain is hilly with undulating sand dunes. Access is only by four-wheel drive vehicles. Production from the field began using one well Hu-44.
The well was initially produced through natural flow to the surface  and through the 27 km, 8 inch flowline to RDS-3.
Since 1996, an electric submersible pump was used to produce the well. Well production during the first years of ESP was characterized by rapidly increasing water cut. Following a single well reservoir simulation study, well Hu-44 was side-tracked as an 1800 ft horizontal well.
Various options to flow the well to RDS-3 were evaluated. A project was approved to try multiphase pumping technology. The objectives of the project are :
Produce well Hu-44 via the gathering centre at RDS-3.
Evaluate the performance of multiphase pumps as a potential technology for full development of Huwaila field.
Monitor and better understand water behavior in the reservoir.
Both dynamic and positive displacement pumps were considered. A discussion on the merits of each type and the gas volume fraction GVF is given in Appendix 1 and Appendix 2. Based on the actual flow rates and GVF at Huwaila, it became clear that the process parameters are more suited to positive displacement pumps rather than to dynamic pumps.
Vendor Scope of Supply
The multiphase pump, together with the associated electrical and control systems were tendered as one package. This was done to ensure single point responsibility and to avoid interface problems between the different vendors.
In developing the technical requirements of the tender documents, several vendors and end users were approached for their experience. There are no industry standards to cover the design and manufacture of multiphase pumps. For twin screw pumps, the design is usually in compliance with API 676, which deals with positive displacement liquid pumps 
As the pump vendor was German, it was agreed in the pre-award meeting that DIN standards would be adopted for some components in-lieu of API standards. For example, the load calculation of the timing gears were in accordance with DIN 3990 and not AGMA standard 6010 as stipulated in API 676. Also, the performance test procedures were not in accordance with API 675 as will be discussed later.
Downhole demulsification of crude oil increases well production and profitability by reducing vertical flow pressure loss. This paper presents a practical, field-tested method to accomplish downhole demulsification without the mechanical complexities of the conventional downhole injection system. Chemical is injected at surface with a pneumatic dosing pump into the gas lift stream. Lift gas entering into the tubing-casing annulus carries the chemical downhole to the injection point deep into the tubing and blends it with the well fluid to prevent emulsification. Production improves after a few days of chemical injection as the downhole demulsification process stabilizes. Successive trials have proved the effectiveness of the approach in terms of improved production. In view of the simplicity and cost effectiveness of the method, it has also been applied in naturally flowing wells that produce below their potential because of emulsion. Case study and result of such trial have been discussed in the paper that shows promising success. It has broadened the application scope of the system to emulsion-infested natural flow wells.
Bahrain Oil Field
Bahrain field is an asymmetrical anticline trending in the North-South direction. The sedimentary column extends from Cambrian Saq sandstone to the Miocene reefal deposits exposed on the surface. The field was discovered in 1932 (Fig. 1) and it contains twenty-two reservoirs (Fig. 2) with the hydrocarbon content varying from tarry oil in Aruma to dry gas in Khuff reservoirs. Mauddud zone within Wasia group of formations is the major oil reservoir of the field. Gas injection into Mauddud zone continuing since 1938 has developed a large gas cap1. Besides Mauddud, other producing oil bearing formations are Nahr Umr (Ca, Cb shaly limestones and Cc, Cd sandstones), Wara (Ac sandstone), Ahmadi (Aa, Ab limestones), Rumaila (Ostracod and Magwa limestones) and Mishrif (Rubble limestone) with depth range of 2500 ft to 1250 ft. There are 700 wells in Bahrain field wherein 260 are naturally flowing, 160 on continuous gas lift, about 70 each on beam pump and intermittent gas lift, and remaining 140 are other types including injectors, gas producers, abandoned and suspended wells. Average gas lift well production varies between 40-4000 bpd of liquid with W.C. ranging from 5 to 98%. Wells flow through separate 3/4 inch line ranging in length from 100-12000 ft to the common headers at 16 well manifolds spread across the field. Lift gas is fed into individual wells at about 1050 psig through 2-inch lines from supply headers running along north-south direction.
Typical Continuous gas lift and natural flow wells (Fig. 3) have sliding sleeves and intermittent gas lift wells have mandrels in the completion string at about 2200 ft depth. Oil gravity and viscosities of various zones range from 12-37 (°API and 4-564 Cst respectively (Table-1). Wells completed in Ac, Mauddud, Ca, Cb, Cc and Cd zones, which are well within the gas cap area, are generally naturally flowing while the flank wells are on gas lift with tendency to become self-flowing. Gas lift wells completed in Cd and Ac zones, particularly in the south of the field, have severe emulsion problem when the producing W.C. is within 60% to 95% range. Recently, emulsion production have been observed in some gas lift as well as natural flowing wells which are located inside the gas cap area and completed in Mauddud, Ca, Cb and Cc zones. Other low-pressure shallow zone wells are normally produced on pump or intermittent gas lift and a few of them flow emulsified crude.
In this paper, a technique for upscaling of absolute permeability in the well vicinity on CPG (Corner-Point Geometry) gridblocks is presented. The near-well upscaling procedure is very useful for well performance prediction. It can assure the coherence of well results by using fine grid on coarse grid simulations.
The near-well upscaling procedure is developed for advanced wells in 3D. Transmissibilities are upscaled from Cartesian fine grid issued from a geostatistical to CPG coarse grid used for flow simulations. To overcome the difficulty of complex geometrical intersections between fine gridblocks and coarse gridblocks, three approaches (geometrical approach, numerical approach and topological approach) are proposed and compared for flux upscaling. The topological approach, which needs a topological structure for gridblock regrouping in the vertical direction, seems the most promising. The near-well formation damage, which can be predicted using another numerical model, can also be integrated into the near-well upscaling procedure.
This study is conducted to test and evaluate the use of current methods of reservoir characterization, namely the permeability-porosity correlation, the J-function, and the Reservoir Quality Index (RQI) concepts, for reservoir description of heterogeneous carbonate formations. These approaches were compared with a new technique developed in this paper for improved reservoir description of carbonate reservoirs. This technique is called the Characterization Number (CN) technique and it is based upon considering fluid, rock, rock-fluid properties and flow mechanics of oil reservoirs.
To compare these reservoir characterization techniques, measurements of porosity, absolute permeability, oil and water relative permeability and irreducible water saturation for 83 actual core samples extracted from eight different wells for a new oil reservoir in the U.A.E. are obtained. These experimental data are used first to develop a permeability-porosity correlation. Then, the J-function and the RQI concepts along with the newly developed CN approach are applied and evaluated for reservoir description of the UAE carbonate reservoir under investigation. The results show that the Reservoir Quality Index concept is capable of identifying the flow units while the J-function concept is quiet poor. Also, a more refined identification of flow units is obtained by using the newly-developed Characterization Number. This improved description for the Characterization Number approach may be attributed to the consideration of rock/fluid properties of flowing fluid(s) and flow dynamic conditions of its containing formation.
Introduciton and Review
Reservoir characterization techniques are quite valuable as they provide a better description of the storage and flow properties of a petroleum reservoir and thus provide the basis for developing its simulation model. Also, carbonate reservoirs, in particular, present a tougher challenge to engineers and geologists to characterize because of their tendency to be tight and heterogeneous.
Permeability and porosity of the reservoir rock have always been considered as two of the most important parameters for formation evaluation, reservoir description, and characterization. Beyond evaluating permeability and porosity, one can also use combinations of two or more rock properties to gain insight into the character of flow through porous media. The J-function and the Reservoir Quality Index (RQI) concepts are two of the ways that the oil industry has used to characterize the reservoir media. They incorporate parameters such as porosity and permeability into a single quantity that describes/characterizes the formation. The application of the J-function and/or the Reservoir Quality Index (RQI) concepts, however, may or may not determine whether a formation can be considered to have a single flow unit or multiple ones1-4.
It is well recognized that an improved and effective reservoir description is a prerequisite for efficient development of oil reservoirs. The following is a brief review of the most common techniques available for reservoir description:
1.1. Permeability-Porosity Correlation Technique
The effective porosity1 of a rock is defined as the ratio of its interconnected pore volume to its bulk volume. The permeability of the reservoir rock is defined as the ability of that rock to allow fluids to flow through its interconnected pores. The permeability-porosity relationship has always been considered as a very valuable tool for interpreting petrophysical properties of the rock, providing better reservoir description and/or enhanced reservoir characterization. A number of investigators5-8 showed that rock permeability depends mainly upon the effective porosity. For this reason, permeability is mainly affected by grain size, grain shape, grain packing, sorting, and degree of cementation.
Wyllie-Rose1,5,6, Timur1,5,7, and Morris-Biggs1,5,8 developed similar empirical correlations to calculate the permeability using porosity and irreducible water saturation for sandstone reservoirs. If one is to apply these correlations in general, he has to take into consideration the following limitations:
they are not applicable to carbonate reservoirs,
they were developed for local fields/formations
Asphaltene Onset Pressure (AOP) conditions were measured for Kuwaiti Reservoirs fluid samples at reservoir, wellhead, and well-bore temperatures along with the saturation pressures by performing a series of pressure depletion tests using the mercury-free, variable volume, fully visual JEFRI-PVT system with laser light scattering. Asphaltene Deposition Envelope (ADE) and P-T phase diagram were developed. Asphaltene onset conditions were also investigated using different precipitants with Stock Tank Oil (STO). A number of titration experiments were performed on the STO using different alkanes and CO2. Our investigation has revealed that CO2 is the most effective asphaltene precipitant followed by alkanes (C1 to C7). To investigate the molecular structure of the asphaltene molecule, advanced analytical techniques such as 1H and 13C NMR and IR spectrometers have been utilized. It was concluded that the total number of carbon atoms was 220 of which 120 were associated with aromatic and the rest were aliphatic rings. The number of aromatic and naphthenic rings were estimated to be 42 and 114 rings respectively.
Compositional models are commonly used to simulate recovery processes in which injection gas and reservoir gases have distinctly different properties. Such processes include vapourising miscible drives, condensing miscible drives and low-pressure air injection. Recovery is typically enhanced by changes in fluid properties and a reduction in interfacial tension, leading to a variation of the relative permeability endpoints. These changes are due to mass transfer, in particular the solution of intermediate hydrocarbon components in the oil. Extensions of the black oil model such as that due to Todd and Longstaff can also treat such processes, although the mechanisms considered are rather different - in particular the amount of inter-phase mixing which occurs.
In this paper we attempt to unify these two approaches. The traditional black oil model is extended to multiple gases and property data constructed which yields a measure of the surface tension after mass transfer. Once the surface tension is known, a common end-point shift treatment can be used in both compositional and extended black oil treatments.
Traditional black oil numerical methods based on saturation variables have problems in treating the effect of partial pressure on the amount of dissolved gas when the gas saturation is low. This may be overcome using mass-based variables, but these have typically been less efficient due to the addition of an extra volume balance equation. A reduced mass variable formulation is described which retains the efficiency of the saturation-based approach whilst being easy to generalize to multiple gas components.
The multiple gas black oil and full compositional methods are assessed on some generic examples of gas injection processes.
In crude oil production, the generation of heavy oil/water emulsions can result in a significant drop in the production rate due to the high viscosities of the emulsions. The strong agitation and mixing that takes place at the inlet of the well bore together with the presence of the naturally existing emulsifying agents in the crude contribute to the formation of the emulsion. A number of methods can be used to mitigate this problem such as the injection of surfactants into the formation and surface treatment. The objective of this paper is two-folds. Firstly, the results of an experimental program carried out to evaluate the efficiency of several surfactants to demulsify crude oil / water emulsions are presented. Secondly, a novel cost effective technique for injecting the surfactant into the well utilising existing gas lift facilities at the well heads is described together with a number of other techniques which have been tested for treatment of emulsions.
The experimental study was performed on crude oil/water emulsion samples obtained from two sandstone reservoirs in the Bahrain field. The measurements included effects of surfactants on the degree of demulsification, surface tension and shear viscosity. An optimum surfactant/emulsion volume ratio of around 1.5 to 2% was noticed in most cases for the specific wells investigated. The addition of surfactants considerably reduced the shear viscosity and surface tension of the emulsion. As expected the surface interfacial tensions decreased with increasing temperature.
The Salt Creek Field, located approximately 110 miles northeast of Midland, TX, is a heterogeneous, Pennsylvanian-aged carbonate build-up in the Permian Basin. Successive implementation of infill drilling and improved recovery processes, coupled with effective reservoir management, have led to oil recovery of over 50 percent and expected ultimate recovery as high as 60 percent.
Salt Creek field was discovered in 1950 and a centerline water injection program began in 1953. A 40-acre development program was initiated in 1970 and was completed in 1984. The waterflood pattern was also changed from a centerline drive to a field wide inverted nine-spot. In 1985, a reservoir continuity study led to a 20-acre infill drilling program that added over 150 wells, changing the pattern to a five-spot. A miscible carbon dioxide (CO2) flood was initiated in 1993, and then expanded to the underlying residual oil zone.
Reservoir management has been a key element in maximizing recovery at Salt Creek. Consistent acquisition and maintenance of appropriate well and reservoir data has allowed detailed zonal tracking of production and injection volumes (CO2 and water). Geologic models, in combination with production logs and production data, have been used to manage and optimize CO2 and water injection at the flow unit level. Effective communication processes have been established between the field operators and the technical staff to calibrate engineering and geologic data with field experience and to enhance real-time decision making. A 3D seismic survey covering the entire field was acquired in 2000 to better define productive field limits and reservoir geometry. Currently, a study integrating engineering and geoscience data is underway to further improve reservoir characterization. The study is expected to result in improved sweep efficiency through better conformance control and additional drillwells to capture unswept reserves.