The effective approach for predicting viscosity of saturated and undersaturated reservoir oil

Predicting reservoir oil viscosity with numerical correlation equations using field-measured variables is widely used in the petroleum industry. Most published correlation equations, however, have never profoundly realized the genuine relationship between the reservoir oil viscosity and other field-measured parameters. Using the proposed systematic strategy is an effective solution for achieving a high performance correlation equation of reservoir oil viscosity. The proposed strategy begins with creating a large database of pressure-volumetemperature (PVT) reports and screening all possible erroneous data. The relationship between the oil viscosity and other field-measured parameters is intensively analyzed by using theoretical and empirical approaches to determine the influential parameters for correlating reservoir oil viscosity equations. The alternating conditional expectation (ACE) algorithm is applied for correlating saturated and undersaturated oil viscosity equations. The precision of field-measured PVT data is inspected by a data reconciliation technique in order to clarify the correctness of oil viscosity correlations. Finally, the performance of the proposed oil viscosity correlation equations is represented in terms of statistical error analysis functions. The result of this study shows that reservoir oil density turns out to be the most effective parameter for correlating both saturated and undersaturated reservoir oil viscosity equations. Expected errors in laboratory-measured oil viscosity are the main factors that degrade the efficiency of oil viscosity correlation equations. The proposed correlation equations provide a reasonable estimate of reservoir oil viscosity; and their superior performance is more reliable than that of published correlation equations at any reservoir conditions.