Describing the volumetric behavior of a gas reservoir requires the knowledge of its composition; however, this information is often limited. This paper presents a workflow for predicting gas specific gravity in absence of comprehensive compositional information. The approach used is to express the real gas law in terms of specific gravity and solve for it by means of the Newton Raphson algorithm. The proposed method requires as input a live fluid density measurement and the non-hydrocarbon content. The Z factor is stated as an implicit function of pseudo-reduced properties that in turn can be explicitly expressed in terms of specific gravity; this is accomplished by means of suitable correlations. The pseudo critical properties determination accounts for the presence of non-hydrocarbon fractions thus this method is not limited by the amount of impurities concentration.

We applied the proposed workflow to a field case corresponding to a sour gas condensate reservoir from which fluid samples are available. Experimental data includes gas chromatography analysis, multistage separator test, and 790 data points comprising Z factor and fluid density measurements from CCE experiments. Specific gravity computation uses fluid density derived from static gradient measurements and an approximate knowledge of the impurities content. A number of 25 combinations of Z factor and pseudo-critical properties correlations were tested for which the convergence rate of the algorithm was 99.9%. Overall error for the specific gravity prediction ranged between −1.5% and 0.5%. CGR and saturation pressure behavior with depth were adequately reproduced.

The proposed methodology allows us for reliable determination of key fluid properties in absence of compositional information by using already and routinely available field data; no expensive additional laboratory work is required. Derived properties can be used for engineering studies purposes and for production testing validation.

You can access this article if you purchase or spend a download.