On April 1994, The Government of Indonesia enacted the Government Regulation Number 19 concerning the Hazardous Waste Management which will be come into effect as from April 1, 1995. Despite all the efforts made, petroleum activities in their operations may use or explore hazardous materials which eventually will produce hazardous wastes. The material, among others are catalyst.
The Appendix of this Government Regulation had categorized the Spent Catalyst as hazardous waste, therefore the handling of this waste (material) should follow the process and procedures as indicated by the regulations.
The new oil refinery at Balongan, West Java, will utilize catalyst, thus produce spent catalyst in quite a big amount, amount, i.e. more than 30 tons per day.
The comparison of treating this bulk of wastes if categorized as hazardous waste ton non-hazardous are quite enormous, and the discussion in this paper will elaborate on the efforts made by Pertamina to appeal to the Government under the co-ordination of BAPEDAL (The Environmental Impact Management Agency), to exclude this specific Spent Catalyst from hazardous waste categories. This references are made to the definitions and criteria set up in this regulation as well as comparison studies with similar industries in other countries.
The exclusion of this Balongan refinery spent catalyst from the present status with categorized as hazardous waste to non-hazardous will eventually meet the requirements of sustainable development operation achievable in most efficient way.
The Treatment of Spent Catalyst at PERTAMINA
In processing crude oil into fuel oil such as premium, diesel oil, LPG, etc. the following processes are employed : distillation, vacuum, cracking and reforming.
The cracking process is the breaking of long CH chain into a short chain generally employing a catalyst to accelerate the said chemical reaction. In addition, the catalyst is also employed at the Hydrogen Plant unit; to help accelerate the reaction between steam and gases from the unit passing through the tube reformer, catalyst of the zinc type is also utilized
In the cracking reaction, the catalyst employed is in the form of sand which mainly contains aluminum and magnesium silica; this catalyst is getting more and more saturated, so it needs to be degenerated or replaced with the new one when it does not function any more.
Determination of the initial saturation of a hydrocarbon reservoir requires resistivity log data and saturation exponent. A number of experimental investigations has shown that reservoir wettability affect the exponent value. A remarkable divergence conclusions, however, still appear in the literature. The objective of the present study was therefore to investigate the effect of rock wettability on saturation exponent. Fractal concepts have been applied to the image of thin-sections of a core sample and used to derive the resistivity equation. The advantage of this approach is that it is independent of rock-fluids equilibrium problems that commonly influence laboratory measurements. A general equation of electrical resistivity has been developed in this study. Electrical tortuosity, clay content, and rock wettability are incorporated in the equation. The present work employed twenty thin-sections of limestone and sandstones.
It was found that the lowest exponent of 1.61 was obtained for strongly water-wet shaly sandstone and the highest value of 4.99 is for strongly oil-wet. The exponent consistently increases as the wetting condition is shifted from strongly water-wet toward oil-wet. It is close to 2.0 for clean sandstones at strongly water-wet, supporting the empirical formula of Archie.
Estimation of a hydrocarbon reserve is strongly dependent of electric log data and the value of saturation exponent used. This exponent is usually either assumed to be 2.0 regardless of the reservoir wettability or is derived from laboratory measurements of electrical properties of the cores.
Saturation exponent has been the subject to many laboratory investigations. The researchers arrived at considerably different quantitative results, particularly for wetting conditions at oil-wet side. Detailed review of this matter has been made by Anderson and Sondenaa et al.
A lack of analytical investigation has been conducted to study the effect of wettability on electrical properties of rocks containing fluids. Fractal concepts have been applied to a wide range of research area related to porous media at various scales, from microscale (pore level) to megascale (field scale). A detailed review of fractal concepts applications has been presented by Sahimi and Yortsos. Briefly, a fractal object is an object that has a geometric dimension greater than its topological (Euclidian) dimension. Fractal geometry may describe the irregurality and fragmentation of natural objects and patterns. To characterize a fractal object or set, the fractal dimension D must be determined. A simple method is by "box counting". This is simply done by drawing a mesh of squares of side dimension that covers the set. The fractal dimension is determined from the slope of the relation between count and square side dimension. which is the fractal dimension that can yield any real value. To handle the complexity of the problems, the present study was directed to applying concepts of fractal analysis to determination of the electrical properties and thus saturation exponent.
Pore structures and geometries can be quite complex in nature. The complexity stems from the irregularity of pore wall surfaces. The presence of clays may increase the complexity of pore structures. As a matter of fact fluids distribution within the pores is influenced by the wetting condition. This in turn would affect the fluids connectivity. Therefore, electric tortuosity depends not only on pores and fluids connectivity but also on the irregurality of main grain surfaces and volume and distribution of clays within the pores. As pores of a porous medium can be magnified by employing microscope, identification of fine details of the pores can be performed through the use of thin sections made. Ultimately, mathematical analysis coupled with fractal analysis is thus possible to study electrical properties of porous media.
A novel normalised plot technique is developed for reservoir characterisation and reserves estimation. This method is based on the Buckley-Leverett and Welge's fluid displacement theories. The theories suggest that under ideal conditions a normalised plot of oil recovery as a function of water or gas (displacing phase) throughput is independent of rate, drainage volume, and geometry of the system. This implies that in a homogeneous reservoir, well by well normalised recovery plots should collapse to one curve. In reality, well performance is influenced by reservoir heterogeneity, drive mechanism (bottom vs edge drive), gravity, and well locations. Therefore, the normalised curves do not often match and that leads to reservoir dynamic characterisation. The normalising factor, sweep volume, is related to the recoverable reserves (RR) that we seek to estimate. This paper presents the theoretical background of the technique and illustrates its application by presenting two field examples.
Use of decline curve analysis for reserves estimation is a common practice in the petroleum industry. Reference 3 presents a summary discussion of the decline methods. However, these techniques are not based on any physical theories. Furthermore, decline analysis assumes constant operating conditions (choke size, artificial lift, etc), which is hardly ever met in field production operations. Experience has shown that analysis by exponential decline generally yields a conservative forecast (also known as a 'buyer's forecast'). On the other hand, hyperbolic extrapolation often results in an optimistic estimate ('vendor's forecast'). The applicability of the above techniques is basically a matter of convenience. The x-cut plot analysis, a second method, is based on fractional flow theory, but the assumption that a semi-log plot of permeability versus saturation data would be a straight line is not valid in all reservoirs and fluid flow systems. A third approach used in the industry is cumulative oil versus cumulative liquid (representing water influx in a water drive reservoir) production plot. Since drainage volumes and consequently recoverable reserves generally vary from well to well, such plots cannot be compared on a one to one basis. A meaningful comparison is achieved if the above conventional cumulative plots are normalised by their respective reservoir drained volumes (hence the name normalised plot).
In 1942 Buckley and Leverett presented the basic equation for describing immiscible displacement in one dimension (also known as Frontal Advance Equation). In 1952 in another paper of fundamental importance, Welge enlarged upon Buckley and Leverett's work and derived an equation that relates the average displacing fluid saturation to that saturation at the producing end of the system. Welge also determined that pore volumes of cumulative injected fluid is equal to the inverse of the slope of the tangent on fractional flow curve. The above mathematical developments have become the basis of performance predictions of incompressible fluids in one- dimensional systems.
In reality, reservoir fluids (particularly gas) are compressible and flow dynamics hardly occur in a linear system. Kern has shown application of the Buckley and Leverett equation to reservoir geometry to predict gas flooding performance. A generalised form (variable cross-section) of Frontal Advance displacement developed by Latil shows that the Buckley-Leverett theory is applicable in a reservoir as long as the displacement takes place in a porous medium whose geometry leads to the equipotentials and isosaturations being superimposed. The pore volume traversed by iso-saturation in a variable geometry reduces to a distance travelled in a linear system.
Hydrochloric:hydrofluoric acid mixtures have been extensively used to treat sandstone reservoirs, however, their application has been restricted to restoring the natural permeability of the formation. A combination of rapid spending with clays and silicates, matrix unconsolidation in the near wellbore area and subsequent precipitation of various reaction by-products has limited the usefulness of mud acid for matrix stimulation treatments.
Several existing "retarded HF acids" mixtures have been tried with mixed results. The marginal reduction in reaction rate of these acid systems could not overcome the huge contrast in surface area between clays and quartz, which dominate HF's reaction rate. Also, various "in-situ-generated" acids were developed and pumped, with questionable to poor results due to premature or improper mixing of the solutions, both in the tubulars and in the formation.
This paper discusses the chemistry and development of a new acid that can completely replace HCl in HCl:HF mixtures. Compared with available HF acid systems, the new HF acid has lower reaction rate and limited solubility with clays, but higher reaction rate and dissolving power with quartz. As a complement, various field treatments are discussed and a complete database with production data is presented.
The framework of sandstone formations is normally made up of quartz, silicate grains, feldspars, chert and mica. These minerals are bonded together by secondary minerals that precipitated from connate water, occupying partially the original pore spaces, for instance, over growth quartz, carbonates and pore lining clays. The porous fraction and hence the permeability of the formation rock is sometimes reduced by invasion of water base filtrates, either from drilling, completion, workover or treating fluids. These fluids can damage the matrix by swelling and dispersing clays or even by inducing the precipitation of scales. Also, in high permeability formations under high differential pressure some particles can invade the formation and plug some pores.
Hydrofluoric and hydrochloric acid mixtures (generally described as mud acids) have been chosen to treat these formations since they dissolve clays from drilling mud and react with most constituents of naturally occurring sandstone. The dissolution of these materials will enlarge and interconnect pores, allowing the produced fluids to move easily. Kinetics of mud acid reaction with sandstone is heterogeneous, complex and difficult to predict. However, the overall reaction rate depends mainly on the nature of the surface reaction within the pore. Unlike quartzitic fines, clays are unstable and easily protonated thus very reactive and therefore dissolved at a faster rate by HF acids. Nevertheless, when reacted with HF acids under isolated condition, both clays and quartz have similar solubility. The solubility of these minerals with 3% HF mixtures is close to 8 times lower than the amount of limestone dissolved per gallon of 15% HCl. Thus to produce in sandstone reservoirs equivalent stimulation results to those obtained in limestone, the acid volume must be several times larger.
During matrix stimulation treatments, the main factors that contribute to matrix dissolution by injected acids are the availability of minerals and their surface area. Therefore pore lining and flocculated clays, carbonates, scales and invading particles are more vulnerable to acid attack than any framework minerals. Moreover, clays have much higher surface area than quartz, hence increasing their dissolution rate, leaving quartz minerals virtually untouched. Thus, as more acid is injected, more clays and cementitious materials are dissolved especially in the big pores and the formation becomes progressively weaker until it eventually re-compacts to lower porosity and permeability levels.
The by-products of sandstone dissolution by HF acids are amorphous compounds and complex fluosilicates, fluoaluminates and fluoride salts which by nature have extremely low solubility.
Andrew, A.S. (CSIRO Division of Petroleum Resources) | Whitford, D.J. (CSIRO Division of Petroleum Resources) | Hamilton, P.J. (CSIRO Division of Petroleum Resources) | Scarano, S. (Santos (BOL) Pty. Ltd.) | Buckley, M. (CSIRO Division of Mathematics and Statistics)
A.S. Andrew, D.J. Whitford, and P.J. Hamilton, SPE, Australian Petroleum Co-operative Research Centre, CSIRO Division of Petroleum Resources; S. Scarano, Santos (BOL) Pty. Ltd.; and M. Buckley, CSIRO Division of Mathematics and Statistics
The geochemistry of sedimentary rocks is a product of their provenance, maturity and diagenetic history. Chemostratigraphy relies on the variations in major and trace element abundances in sedimentary rocks as a correlation tool. It is particularly relevant in field appraisal where conventional correlation techniques such as seismic and biostratigraphic methods often have insufficient resolution. In the oil and gas province of the Surat Basin of Queensland, a series of Triassic multi-storey channel fills overlie Permian peat swamp and basin flood deposits. Correlation of the Triassic sands between wells is crucial to definition of the trap sequences.
The chemical data are assessed in two ways: using image analysis/statistical techniques and by normalization relative to elements which do not readily partition into natural waters.
A mathematical comparison of the elemental profiles with depth shows a good match between wells Link #1 (ca. 2116 - 2132 m) and Roswin North #1 (ca. 2143 - 2157 m). Such a correlation implies the absence of the Showground and upper Rewan Formations in Roswin North #1. In its present form, this approach cannot allow for significant amounts of missing section in either well.
Using an approach that relies on the geochemical characterization of unconformity-bounded units, it is possible to correlate between the upper unit in Roswin North #1(2143 -2147 m) with the second top unit in Link #1(2108 - 2113 m), and between Roswin North #1(2151 - 2157 m) and Link #1 (2120 - 2124 m). This correlation is consistent with the correlation based on mathematical comparison of elemental profiles.
This work makes up part of a multi-basin test of the application of chemostratigraphy to reservoir definition in Australian Basins. The Surat Basin example presented here tests the technique when applied to stacked channel sands in a marginal marine environment.
Chemostratigraphy involves the application of major- and trace-element geochemistry for the characterisation and subdivision of sedimentary sequences into geochemically distinct units, and correlation of strata in sedimentary basins. Sedimentary rocks are faithful records of the sensitive and subtle changes in provenance, environment of deposition and post-depositional changes. Such changes mean that apparently uniform successions may show primary differences in the chemistry of their constituent minerals, or in the proportions of accessory phases such as heavy minerals and clays, many of which have very distinctive trace-element contents. An ability to characterise these subtle geochemical heterogeneities enables apparently uniform thick successions to be subdivided and correlated between wells.
Chemostratigraphic correlation is particularly applicable to sequences that have very poor biostratigraphic control or to thick, rapidly deposited sequences that cannot be subdivided further by biostratigraphic data. Ambiguity and uncertainty often associated with more traditional methods of correlation such as lithostratigraphy, biostratigraphy and geophysical logging may be resolved. Geochemical correlation between wells may be established by the identification of similar geochemical characteristics. Such characterisation typically involves the abundance determination of approximately 50 elements.
Inter-well correlation may be achieved by the recognition of similar geochemical characteristics in adjacent sequences and is accomplished by:
By eliminating the need to under-ream, the combination of PDC bit technology and a newly developed tool that enables the operator to simultaneously ream and drill ahead reduced by four days and approximately US$465,000 the time and costs of an offshore Australian well.
This paper describes the development of the reaming while drilling system and its application in a highly cost-intensive well offshore Australia. A detailed operational summary, emphasizing the application/performance of the system will be presented and analyzed.
Using previous under-reaming applications as a benchmark, the operator sought a risk free and cost-effective alternative to conventional hole opening. Various designs were reviewed with the aim of developing a device that would best suit the targeted application. Every facet of the drilling process was analyzed to develop a plan of best practice. For example, the use of pseudo oil-based mud was considered critical to not only stabilize the wellbore, but also to enhance the life of the reaming/drilling system.
The system successfully drilled the targeted section, with caliper logs showing 98% of the hole being drilled to the desired size. Although partially control-drilled to control potential pack-off problems, the well recorded penetration rates similar to those expected for a 17 1/2-in. PDC bit.
From this experience, several operational recommendations arose that will permit similar intervals to be drilled even more cost effectively. The authors will discuss those recommendations.
Conventional bi-center PDC bits have long proven to be cost-effective alternatives to traditional under-reamers and their associated mechanical risks. Primarily, this technology has been employed to drill a larger hole size than the inner diameter of the casing through which it passes. Furthermore, bi-center bits have effectively reduced bit and bottomhole assembly sticking in fast-moving salt formations, and eliminate the need for a second under-reaming trip in well deepening operations, thus significantly lowering costs.
As operators move toward deeper and high-pressure, high-temperature (HPHT) wells with extra casing strings and longer intervals drilled through unstable or encroaching formations, the ability to simultaneously ream and drill becomes even more economically attractive.
Operator Western Mining Corporation Ltd. considered the emerging technology in reaming while drilling systems when planning its 3800-m TVD Bay-1 exploration well offshore Australia. The well was located off the North West Shelf on block WA 215-P some 25 kilometers southwest of Barrow Island, Fig. 1 shows the location of Bay-1, highlighting the primary offset fields used in developing the bit program.
As shown in Fig. 2, the lithology of the Bay-1 upper sections comprised interlaminated hard limestone and friable fine sandstone, overlying sticky clays, and unstable water-sensitive shales. The lower intervals encountered the Malouet and Dingo/Dupuy formations characterized by a long siltstone/shale interval and, at the base, extremely hard (60 - 70 Sec/ft) and highly abrasive sand/shale sequences.
On the basis of the proposed well plan, which utilized an additional string of 16-in liner, a drill bit program was developed with the objective of eliminating under-reaming in suitable intervals.
Relative permeability is a primary parameter for determining the reservoir behaviour of coal seams. We present numerous relative permeability curves measured in the laboratory as well as curves from numerically matching field performance. Viscous fingering provides an explanation for the high values of residual water saturation.
Relative permeability is a primary parameter in determining the response of coal-seam reservoirs. Despite the importance of relative permeability, measurements are infrequent and very few recorded examples appear in the literature. The main reason for this is that the properties of coal make the measurements difficult.
In the published literature, discrepancies exist between relative permeability curves from the laboratory and the field. High values of residual water saturation, often in excess of 80%, are observed.
In this paper we present numerous relative permeability curves measured in the laboratory as well as curves from numerically matching field performance. Some of the discrepancies in the results are postulated to be a result of the unfavourable mobility ratios that exist when gas displaces water. This leads to viscous fingering. Evidence to support this conclusion is provided in X-ray CT images of core displacements.
Viscous fingering behaviour is very dependent on heterogeneity. The extremely heterogeneous nature of coal is interpreted to be primarily responsible for the variability in relative permeability curves. We provide examples of how this is manifested in practice. Viscous fingering provides an explanation for the high values of residual water saturation.
Conclusions from this paper are important for scaling results from laboratory tests to field applications. Applications occur in coalbed methane and gas drainage from coal seams for mine safety. This work is also relevant to multiphase flow in other types of fractured rock.
Reservoir Properties of Coal
Coal is generally treated as a dual porosity rock containing micropores and a network of natural fractures. The fractures are known from mining as cleats. Some basic reservoir engineering aspects of coal have been described by Gray.
Coal is arguably the most difficult reservoir rock to work with for the reasons listed below.
1. Coal tends to be friable and very heterogeneous, so sample selection is a major issue. Procedures for sample selection have been recommended by Hyman et al.
2. The connected fracture network of coal has low porosity. Laboratory testing therefore requires accurate measurements of very small amounts of fluid.
3. Very pronounced stress-dependent permeability is observed in coal. Furthermore, as the stress load is cycled, permeability tends not to return to the original value. The reason for this has not been clearly identified, but may include fines migration.
4. Many gases strongly adsorb on coal, including methane, carbon dioxide and nitrogen.
Recorded measurements of relative permeability in coal are sparse. It has only been in the last few years that direct measurements of water relative permeability during gas injection have been reported. We reproduce below some of the previously published curves. To assist visual comparison, they have been replotted with the same aspect ratio and with water saturation on the abscissa axis.
In the early 1970s Reznik et al. reported on measurements of relative permeabilities under steady-state conditions, with water saturation both increasing and decreasing. With decreasing water saturation they were unable to measure water relative permeabilities, and instead computed values from the corresponding gas permeabilities by Corey's method. They acknowledged that the applicability of this approach to coal is tenuous. Relative permeability curves for decreasing water saturation are shown in Fig. 1a. The relative permeability to water is shown as a broken line because it was not directly measured.
The application of three-dimensional (3-D) seismic methods to reservoir characterisation has gained wide acceptance in the last ten years. Almost seventy-five percent of additional reserves are now found in mature areas by redefining reservoir boundaries and internal heterogeneities using surface and crosswell seismic methods. However the implementation of, in particular, 3-D seismic methods can be very expensive and can rival the cost of drilling a dry hole.
A solution to the problem is to physically model a reservoir in the laboratory and to collect simulated seismic reflection data over the model. Then conventional, or unconventional, seismic processing techniques can be applied to the data prior to embarking on costly field seismic data collection. The physical modelling experiments enable an improved and corrected image of the subsurface to be developed, and a more effective drilling program to be designed.
We briefly review the 3-D seismic method and indicate its value in reservoir characterisation. Examples will be given of models developed for experiments in fault plane analysis and for imaging beneath high velocity near surface layers. Introduction
The 3-D seismic method was first shown to be viable for subsurface imaging in the early 1970s (Ref. 1). This was the first successful application of physical modelling to solving 3-D seismic problems. The fact that the 3-D method has been rapidly accepted in the last 10 years has been documented (Ref. 2). The conclusion is that '3-D seismic surveys have become a cost-effective tool for mapping hydrocarbon reservoirs and can have a major impact on the volume of reserves estimated'.
The method needs the subsurface area of interest to be spatially sampled correctly with seismic traces. in marine operations this may be achieved with a ship towing many cables, such that each source and receiver combination represent a sample point. In land operations various positionings of sources and receivers may be used such that an area of the subsurface is 'covered' with seismic traces.
The examples in this paper show two very different applications of 3-D seismic to solving reservoir problems, both of which required the extensive application of physical models.
Fault Plane Analysis. When a seismic wave meets an acoustic contrast it may be reflected and/or refracted. If a wave reaches a fault with a relatively high impedance contrast across it, it will refract through the fault plane, resulting in a change in the travel path of the wave. When common midpoint gathers are produced, the wave refracting across the fault can cause changes in the trace gathers which can affect 'reflections' within the gather. When the data are stacked the result can give a different appearance to the image. Such a change could be the presence of an apparent fault at depth, which is in fact an artefact of the stacking process. When computing the reserves in a reservoir the assumption of the presence of such a fault could lead to incorrect reserve estimates. Consequently, it is very important to determine whether such an artefact does, or does not, exist.
Two dimensional (2-D) seismic data collected across a fault plane containing such an impedance contrast in the shallow section will produce a false image of a fault (Ref. 3). 3-D data collected along strike to the fault will not have raypaths passing through the fault plane. so these data will not create such an artefact. A 2-D section can be reconstituted from the 3-D data, along the plane of the original 2-D line, without the artefact indicating that the 'fault' does not exist.
This paper describes the combination of integrated petroleum engineering studies with probabilistic decision making to assess the value of information of an appraisal well in a mature field.
The implementation of a cost effective pressure maintenance in the St Joseph Field offshore Malaysia has required consideration of several options, one of which is reverse dump flooding of the oil reservoir from underlying unconnected aquifers. A reverse dump flooding scheme would require drilling of a pilot well to collect additional data on the aquifers. Justification of the pilot (appraisal) well involved a series of studies based on limited data to establish that the dump flooding process had a sufficient probability of leading to a development with better economics than a gas injection scheme.
The aquifer assessment used seismic, biostratigraphic, pressure, regional data, and trend data from logs to define a distribution of connected aquifer size and quality. Well studies defined the ranges of aquifer productivity and oil leg injectivity and resulted in a flexible well completion that could be converted to a production well. Full field reservoir simulation estimated the aquifer and well performance ranges to derive a probability distribution of likely water dump volume and incremental oil reserves. Screening economics were performed and compared with gas injection. Monte Carlo simulation of the many combinations allowed definition of the probabilities as input for a decision tree analysis. Analytical well test simulation showed that uncertainties would remain in aquifer volumes after a pilot test, and needed to be incorporated into the decision tree. Ultimately the value of information of the pilot well was quantified to show an economic justification for pursuing the dump flooding option.
St. Joseph Field (Fig. 1.) has been on production since February 1982. At end 1995, the field had produced 105 MMstb oil out of the total ultimate recovery estimated at 230 MMstb. The recovery mechanism to date has been natural depletion under gravity drainage with string-by string control of gas-oil-ratio imposed. Current production potential is 28 Mb/d. Average reservoir pressure has fallen from 1060 psia to about 600 psia with an expanding secondary gas cap, and abandonment pressure under gas lifting is approximately 350 psia.
Of the 550 MMstb oil initially in place (STOIIP) in St Joseph Field, 83% is located in the main reservoir package in the Northwest Flank, which is a simple structure dipping at 20 degrees and sealed by a major strike-slip fault (Fig. 2). The oil reservoirs are Miocene age sands (Stage IVC) with intercalated shales deposited in outer shelf to deltaic/lagoonal environments. The reservoir package is sealed against a major transverse fault. Gross reservoir hydrocarbon thickness is 1450 ft true vertical with the crest of the structure at only 1280 ft ss. Sand porosities and permeabilities are good in the oil column, generally deteriorating down flank into the aquifer. In addition, major slump features have removed large reservoir sections in the aquifer. Direct aquifer influx to the main reservoir package is weak. 3D seismic data show strong continuous reflectors underneath the reservoir package, indicating a thick Stage IVA sequence of which the upper part has penetrated by eleven wells. Log and RFT measurements indicate that Stage IVA sands are water bearing and are overpressured by 60-90 psi. The Stage IVA and IVC deposits are separated by the Upper Intermediate Unconformity (U.I.U).
Pressure Maintenance Options
Three main options for secondary recovery have been considered for St Joseph:
Water saturation (Sw) in shaly sand formations has been a subject of debate for many years and several equations have been introduced for the estimation of Sw. In the present study, several equations for the estimation of Sw are applied in the Late Carboniferous - Early Permian kaolinite-bearing Tirrawarra Sandstone, an important oil reservoir in the Cooper Basin. The results from Archie and Waxman-Smits equations display a good correlation with measured core water saturation.
Plots of water saturation from the Archie equation with water saturation from the other equations show that the Archie and Waxman-Smits equations are approximately the same (r2=98%). As can be expected, due to the low cation exchange capacity (CEC) of Tirrawarra Sandstone, the second part of the Waxman-Smits equation concerning shale conductivity is very small. As a consequence, the values of Sw from the Waxman-Smits equation are very close to those from the Archie equation. Petrographical studies indicated that kaolinite patches in the Tirrawarra Sandstone are water wet. In this case, the formation resistivity is a function of formation water in macro-porosity and irreducible water associated with kaolinite micro- porosity. In other words, when kaolinite is electrically inert, the whole rock can be considered as a clean sandstone which obeys the Archie law. Because resistivity logs cannot recognise free water within the macroporosity from bound water associated with clay minerals, calculated water saturation includes both free and irreducible water. As the irreducible water associated with kaolinite minerals is not expelled during production, some intervals of kaolinite-rich sandstones, which may never produce water, may be bypassed as non-productive zones.
A new equation is introduced for estimation of effective water saturation. The equation is based on the integration of resistivity and sonic logs with image analysis data for wells in the Moorari and Fly Lake fields, where obtaining the volume of clay is difficult and unreliable. This equation, which reduces calculated water saturation by about 10% for the Tirrawarra Sandstone, is likely to be applicable to other kaolinite-bearing sandstones.
Water saturation (Sw) is one of the most important petrophysical parameters required for reserve calculation. In clean sandstones, estimation of water saturation is rather simple due to the lack of electrically-conductive materials such as clays. The presence of clays in the shaly sandstones which display high conductivity (Ref. 1) makes estimation of water saturation complicated. In this paper, factors affecting conductivity of shaly sandstones are reviewed and some of the common equations which are used for calculation of water saturation are applied to the Tirrawarra Sandstone and a new equation to estimate water saturation in kaolinite-rich sandstones is introduced.
Electrical conductivity of a rock depends on several variables including, conductivity of rock components, conductivity of pore fluid, fluid saturation and formation factor (F). In clean sandstones, the conductivity of rock components is zero and electrical conductivity is a function of conductivity of pore fluid, fluid saturation and formation factor. In shaly sand formations, the electrical conductivity of rock components can not be considered zero, as clay minerals provide additional conductivity (Ref. 1). In these sorts of formations, conductivity of the clay minerals should also be estimated.
Formation Factor. Formation factor was first introduced by Archie (Ref. 2) and defined as the ratio of the conductivity of brine to the conductivity of fully saturated clean sandstone: