The use of seismic attributes has increased, especially when extracted from interpreted horizons. The various available attributes are not independent from each other but represent, in fact, different ways of presenting and studying fundamental information from seismic data (time, amplitude, frequency and attenuation). However, statistical analysis using attributes must be based on geological knowledge and not only on mathematical correlation. Petrophysical studies and seismic modeling are sources of understanding. Such knowledge is necessary to improve confidence in observed correlations with reservoir parameters and must be part of all attribute analysis.
However, the use of seismic attributes leads to several questions, for example, what do they all mean? When to use one or another? How to use them on geologic modeling? How reliable those data are? The answers to these questions are not easy, but considering about petrophysical modeling (Porosity, NTG and permeability) what is the best approach: to consider only well data, that are punctual and need to be interpolated, or try to find correlation with physical measurements (seismic data)? Not to consider seismic attributes makes one feel coming back in time, when this important tool was not available.
On a giant oilfield offshore Brazil seismic attributes (‘conventional', complex trace, polynomial decomposition, geometric and coherence) have been used to create geological models and to reduce uncertainties. The attribute choice must be performed by the geophysicist and the geologist working together, in order to check geological meaning of attribute maps, possible physical meaning of the attribute, etc. Plots of the highest correlation values should be visually inspected in order to choose the attribute with best correlation to the desired parameters.
The results show attributes have been favourable to porosity and NTG prediction, but regular (at maximum) to permeability. For permeability even if the results are not so good, the correlation are improving for the latest models (as long as new wells are used). Polynomial decomposion and complex trace attributes have shown better results.
Introduction: seismic attribute definitions and discussions
The use of seismic attribute data for prediction of detailed reservoir properties began more than 30 years ago.
In fact, a seismic attribute is any property derived from seismic reflection signal. Attributes may be compared to lithology in an attempt to devise a method of property prediction away from well control. The method of prediction can vary from a simple linear correlation to multi-attribute analysis, geostatistical methods, etc.
As an evidence of current proliferation the use of attributes, Chen and Sidney (1997) have catalogued more than 60 commom seismic attributes along with a description of their apparent significance and utility.
Although there is a rich history of seismic attributes use in reservoir prediction, the practice remains a difficult and uncertain task. The bulk of this uncertainty arises from the nature of the physics connecting a number of attributes to a corresponding reservoir property. Due to the complex and varied physical processes responsible for various attributes the unambiguous use of attributes for direct prediction will probably remain a challenge for the years to come.
In addition to the fact above described, there is the possibility of coming across statistical pitfalls while using multiple attributes for empirical reservoir property prediction. In addition, many attributes are derived using similar signal processing methods and can, in some cases, be considered largely redundant with respect to their description of the seismic signal.
This paper compares the output of several available empirical black oil model correlations against compositional model results. In this process, the limitations of these models became apparent.
Even acknowledging the imperfections of black model implementation, it is possible to improve the quality of the outputs by means of making the definitions consistent and coherent across the prediction ranges.
A new method is outlined in order to extend the validity of the models in predicting both reservoir and multiphase flow simulations.
This new method is presented here and will be extended in a separated paper.
The behavior of black oil fluid is commonly inferred from two PVT laboratory procedures: flash (or separator test) and differential liberation. Oil formation volume factor and gas solution ratios are calculated as explained by McCain. On the other hand, given a particular EOS is possible to obtain PVT fluid parameters by simulating the same laboratory procedures or making direct flash calculations at any particular condition.
The traditional calculation method outlined in 1 can be modified in a simple way to extend the validity of black oil model correlations by accounting the dew point curve. Negative gas solution ratios indicate liquid vaporization, and need not to be masked by any correction method. If we follow definitions literally, Rs diminish towards dew point and reaches a constant negative minimum at dew point and inside monophasic gas area. Oil formation volume factor can be lower than unity and in fact should be zero at dew point.
As modern calculations take into account both reservoir and multiphase wellbore and pipeline calculations, is of paramount importance to be able to accurately predict fluid properties in a wider range of pressure and temperature conditions.
The first objective of this paper is to make apparent the limitations of current PVT laboratory calculations and propose a revision.
A second objective is to present black oil model standard correlations phase diagrams together with phase diagrams calculated with EOS and acknowledge the differences and limitations of empirical correlations.
The third objective is to outline a new mathematical method to improve black oil correlations.
The following definitions extracted from Dake will be taken as references:
These parameters enable converting fluid volumes at any conditions to volumes at standard conditions.
Modern viscosity prediction methods have to satisfy the requirements for flow assurance and reliable reservoir characterization by demonstrably predicting accurate, reliable and internally consistent viscosity data. The best way to achieve this is by employing predictive methods based on the best available theory, simplified, just sufficiently to allow ready application and validated against a critical set of primary experimental data of proven accuracy.
The presented VW methodology is one such method, that is based on the kinetic theory of rigid spheres, modified to take into account the behavior of real fluids. It has no adjustable parameters, and requires no dense mixture viscosity data. In this work, the VW method was validated against a new set of natural gas viscosity data of very high accuracy. The experimental data were predicted with an rms deviation of the order of 0.5%-1%, commensurate with the experimental accuracy of the data. Overall, it is estimated that the VW method can predict the viscosity of natural gas within ±2% in the temperature region 260 K - 400 K and pressures up to 200 bars, with the accuracy deteriorating slightly at higher pressures and lower temperatures. It can be used to predict the viscosity of CO2-rich, sour and wet natural gas.
Increasing demand for natural gas has led to the need to develop a more reliable reservoir characterization and simulation. The upstream gas industry, through the gas suppliers, is also being faced with increasing demand for precision in the monitoring of gas supplies. For the exploitation and usage to be optimal, an accurate and reliable knowledge of viscosity, along with other thermophysical properties of natural gas is a prerequisite. The large number of possible natural gas mixtures, and the wide range of relevant conditions, precludes obtaining data by experimental means alone; accordingly, predictive methods are required.
The petroleum industry currently bases its prediction of viscosity on the Lohrentz-Bray-Clarke (LBC) correlation produced in the 1960's1. As is well documented in the SPE literature2-3 and elsewhere4-5, such an approach does not produce reliable and accurate viscosity predictions. For natural gas the failure of the method is mostly evident at high pressures, since the underlying assumption of the LBC method, that the residual viscosity is only a function of density, no longer holds true4,6. A number of methods have been proposed to address this structural failure with varying degrees of success2,4. Furthermore, a number of empirical correlations have been proposed for natural gas (see Ref 3 for details), as a replacement of the LBC method. They usually rely on fitting to a limited set of experimental viscosity data. The data do cover natural gas and similar hydrocarbon mixtures, but invariably a limited number of compositions is available. Also, the viscosity data used for this purpose, are not always obtained in primary instruments.
A primary instrument7 is a well-characterized experimental apparatus, with a well-defined uncertainty level, which produces data that cannot be demonstrated to be inconsistent with other data or with theory. The use of such viscometers requires knowledge of a full, fluid mechanics, working equation and requires a number of corrections to be applied to the experimental data, making the analysis expensive and time-consuming. This makes use of empirically based correlations difficult to justify unless they are fitted to the data obtained in a number of primary instruments. Even then the possibility of systematic errors in the primary data set could not be discounted. More importantly any extrapolation to mixtures and conditions outside the range they were fitted to is fraught with difficulty. Thus, currently neither the empirical correlations nor LBC type methods produce optimal, accurate and reliable viscosity values over the whole range of conditions and compositions of interest to the natural gas industry. Hence, they do not satisfy the modern industrial requirements for flow assurance and reliable reservoir characterizations, which are likely to become even more stringent in the not-so-distant future.
The analysis of alternatives of development of gas deposits and condensed is usually carried out by using the coupling between the balance of materials in predictive mode and the nodal analysis or also using numerical simulation models. In both cases it is required a suitable model of the multiphase flow behavior in vertical pipes in order to be able to predict with accuracy the flowrates the wells will produce through the different life stages of the reservoir.
The correlations for the multiphase flow, with which it is modeled the flow in vertical tubes, show a big dispersion of results in flowrates less than 50 km3std/d with gas - liquid relationships less than 10000 m3/m3. This area of low flowrates and low gas - liquid relationships is usually observed in the gas and condensed reservoirs final phase due to the decline in the production because of the fall of the static pressure and the fall of the gas - liquid relationship and the increase of the water cut.
Consequently, the dispersion in the correlations generates important differences in the production forecasts, with the associated high economic impact, because the pressures for abandonment of the wells are very different according to the multiphase correlation used.
In this work we show an exhaustive statistical analysis of the behavior of the multiphase flow correlations in vertical tubes, contrasted with measurements of pressure gradients in wells, in the majority of the cases within the ranges of flowrates and gas - liquid relationships mentioned in the previous paragraph.
From the results of this analysis, it is developed a new correlation of multiphase flow for gas - condensed - water systems that allows predicting the pressure gradients in vertical tubes within the reasonable error ranges.
We show the methodology used for the development of the correlation and the statistical analysis of the results of the application of it based on the data from the studied cases.
Successful design and implementation of a miscible gas injection project depends upon the minimum miscibility pressure (MMP) and other factors such as reservoir and fluid characterization. The experimental methods available for determining MMP are both costly and time consuming. Therefore, the use of correlations that prove to be reliable for a wide range of fluid types would likely be considered acceptable for preliminary screening studies. This work includes a comparative evaluation of MMP correlations and thermodynamic models using an equation of state by PVTsim1 software. We observed that none of the evaluated MMP correlations studied in this investigation is sufficiently reliable. EOS-based analytical methods seemed to be more conservative in predicting MMP values.
Following an acceptable estimate of MMP, several compositional simulation runs were conducted to determine the sensitivity of the oil recovery to variations in injection pressure (at pressures above, equal to and below the estimated MMP), stratification and mobility ratio parameters in miscible and immiscible gas injection projects. Simulation results indicated that injection pressure was a key parameter that affects oil recovery to a high degree. MMP determined to be the optimum injection pressure. Stratification and mobility ratio could also affect the recovery efficiency of the reservoir in a variety of ways.
Through the past decades, miscible displacement processes have been developed as a successful oil recovery method in many reservoirs. The successful design and implementation of a gas injection project depends on the favorable fluid and rock properties. The case studies using Eclipse2 compositional simulator considered the effect of key parameters, such as injection pressure, stratification and mobility ratio on the performance recovery in miscible and immiscible flooding of the reservoir. However, accurate estimation of the minimum miscibility pressure is important in conducting numerous simulation runs. MMP is the minimum miscibility pressure which defines whether the displacement mechanism in the reservoir is miscible or immiscible.
Thermodynamic models using an equation of state and appropriate MMP correlations were used in determining the MMP.
Compositional simulation runs determined the sensitivity of the oil recovery to the variations in above mentioned parameters. Significant increase in oil recovery was observed when interfacial tension dependent relative permeability curves were used. These relative permeability curves provide an additional accounting for miscibility by using a weighted average between fully miscible and immiscible relative permeability curves. The local interfacial tension determines the interpolation factor which is used in obtaining a weighted average of immiscible and miscible (straight line) relative permeabilities.
Simulation runs were performed at pressures below, equal to, and greater than estimated MMP for reservoir fluid/ injection gas system. Oil recovery was greatest when miscibility achieved.
To investigate the effect of stratification on the performance recovery of the reservoir, the base relative permeability of two layers changed. Location of the high permeable layer (up or bottom layer) in the stratified reservoir greatly influenced the efficiency of the reservoir.
Understanding the effect of interfacial tension and adverse mobility ratio on the efficiency of the gas injection project was the last case study. Injection gas and reservoir fluid compositions differed in such a way to have interfacial tension and mobility dominated mechanism. To investigate the effect of interfacial tension water was considered as a fluid with much higher surface tension values with the oil. Lower surface tension values between rich gas and reservoir fluid (interfacial tension dominated) made gas injection project a more competitive recovery method than waterflooding. In mobility dominated displacement mechanism (lean gas/reservoir fluid system) the viscous instabilities were more important than the interfacial tension effect. For this case, waterflooding with favorable mobility ratio resulted in higher oil recoveries.