TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractPrevious attempts at diagnosing the problem of water production in oil wells have been carried out without regard to the random and periodic nature of the problem. By treating the water-oil-ratio(WOR) as a stationary stochastic time series, a new model for diagnosing water production in oil wells based on Spectral Analysis/Fourier Transformation of the random production data is presented.Consequently, diagnostic analysis have been carried out in the time domain which sometimes gives misleading results due to the inherent "noise" associated with the analysis of fluctuating and random functions in the time domain 2,3 . A better analysis procedure could be achieved when the fluctuating time series variable is converted to frequency domain so that a frequency spectrum will be obtained which allows for easy identification of the different processes which contribute to the fluctuation observed in the time series. Analysis of this spectrum should enhance the understanding and diagnosis of the phenomenon represented by the sequence of observation Key parameters used in the diagnosis and characterization are the autocovariance function, the autocorrelation function and the spectral density function. These functions are respectively used to obtain information about the spectral bandwidth, the correlation structure and energy distribution within each mechanism of water production This method has been used to produce spectra for the diagnosis of coning using field data. Results show that coning can be modeled fairly closely as a low order autoregressive process with narrow spectrum. The position of the peak power in the spectrum and the periodicity of the autocovariance plot were used to uniquely characterize a typical coning process.

Existing technologies currently utilised in well, reservoir & facility management (WRFM) to gain detailed insights into the issues affecting the petroleum production system are discipline specific, hence focus is mainly given to analysis of each component of the production system separately. These isolated functional analyses are sometimes based on inconsistent datasets which may lead to erroneous diagnosis of production problems and could ultimately result in failed well interventions if not rigorously checked via multi-disciplinary reviews. The multi-disciplinary nature of WRFM practice requires that the discipline-specific interpretations of the well, reservoir & facility models be integrated to eliminate potential inconsistencies and improve quality of interpretations. This paper presents a new technology for WRFM workflow integration based on the principle of integration of interpretations. The benefits of integration of interpretations in WRFM data democratisation and process digitalisation are also highlighted.

As the economic importance of Natural gas continues to grow, gas well surveillance equally generates more interest to the petroleum Industry. The static and flowing Bottom-hole pressures must be known in order to predict the productivity or absolute open flow potential of gas wells. Often, these parameters are measured using down hole gauges method of gas well test. In Gas wells, down hole parameters could be estimated using mathematical expressions relating separator conditions and the configuration of the Piping system, instead of measured directly using Gas well Test procedure thus saving lots of dollars. The conventional method of Gas parameter measurement include the iterative method, the Sukkar and Cornell method, Cullender and Smith method. Others are Crawford and Fancher and the Poettman method. All these methods involve very long and cumbersome Iterative/look up procedures. This makes the job of the field men very tedious, especially when the unit system changes. Similarly, data integrity is not guaranteed and could mar the estimation of parameters because data is far from actual. These iterative/look up procedures are always simplified by field men at the expense of accuracy. Engineering problems are now becoming complex that their effective solution requires the systems approach using computers. This research work presents a Microcomputer program for the estimation of Static and Flowing Bottom hole pressure in gas wells, using the iterative method, Sukkar and Cornell method and the Cullender and Simth method. It utilizes a user-friendly window/ dialog based Visual Basic program functional with adjustable unit system. Introduction The increasing importance of gas in world economy has raised a lot of interest in gas well surveillance and parameters estimation. In oil wells, parameters like static pressure, flow bottom hole pressure and flow rate are measured using down hole gauges which is often inconvenient and expensive though more reliable. Often, in gas wells, these down hole parameters are estimated using mathematical expression relating separator conditions and the configuration of the piping system. The percentage error between the measured parameters and the estimate parameters is most times less than 5.2%, Young (1967). This justifies why the parameters used for gas well performance prediction surveillance are often estimated instead of measured directly.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractPrevious attempts at diagnosing the problem of water production in oil wells have been carried out without regard to the random and periodic nature of the problem. By treating the water-oil-ratio(WOR) as a stationary stochastic time series, a new model for diagnosing water production in oil wells based on Spectral Analysis/Fourier Transformation of the random production data is presented.Consequently, diagnostic analysis have been carried out in the time domain which sometimes gives misleading results due to the inherent "noise" associated with the analysis of fluctuating and random functions in the time domain 2,3 . A better analysis procedure could be achieved when the fluctuating time series variable is converted to frequency domain so that a frequency spectrum will be obtained which allows for easy identification of the different processes which contribute to the fluctuation observed in the time series. Analysis of this spectrum should enhance the understanding and diagnosis of the phenomenon represented by the sequence of observation Key parameters used in the diagnosis and characterization are the autocovariance function, the autocorrelation function and the spectral density function. These functions are respectively used to obtain information about the spectral bandwidth, the correlation structure and energy distribution within each mechanism of water production This method has been used to produce spectra for the diagnosis of coning using field data. Results show that coning can be modeled fairly closely as a low order autoregressive process with narrow spectrum. The position of the peak power in the spectrum and the periodicity of the autocovariance plot were used to uniquely characterize a typical coning process.