| dc.description.abstract | In reservoirs with insignificant column heights, each component will have the same chemical potential regardless of its location. Therefore, the reservoir can be assumed to have a constant composition. On the other hand, a height potential term need be added to reservoirs with significant column heights.
The compositional gradient in a petroleum reservoir can be calculated from the chemical energy, gravity and thermal driving forces if the reservoir is in a stationary state. There are two published thermal models which have contrary compositional gradient predictions, with one model predicting that thermal gradient decreases compositional variation [2], [5] and another model predicting the contrary, i.e., that the inclusion of thermal gradients increases compositional variation. This thesis reviews and discusses those models based on the literature and many discussions. The differences in these predictions are found to be due totheir use of different reference temperatures for ideal gas enthalpy (Hig), the first using 273.15K and the second using the temperature at Hig =0.
The lessons learned about the C6+ input data and the issue of volume shift in PVTSim are briefly discussed before single sample and common fluid (common EoS) compositional gradient procedures are presented. Then, the data from four reservoirs are used to compare calculated and observed compositional gradients.
In case the Gas Oil Contact (GOC) is unknown, the GOC may be calculated from a single sample compositional gradient, and then this calculated GOC is compared to the GOC from the multiple wells pressure test. It will be discussed later, using a single fluid sample compositional gradient to determine GOC in initialization is highly risky.
When the GOC is known, the observed deviations of C1, C7+, MW C7+, Bo and GOR indicate that neither the thermal nor isothermal model can well predict the compositional variation in gas and oil zones. For simplicity, the black oil model in oil reservoirs with known GOCs will mainly consider the oil zone because of the oil value. Later, the predictions of isothermal models are observed to be better than those of thermal models in the oil zone in terms of GOR, which introduces different composition in the black oil model, and Bo.
Furthermore, a reservoir study is performed to understand the aromatic effect, the geochemistry and migration, the gamma distribution fluid and enthalpy matching. The results suggest that enthalpy matching should not be performed because petroleum reservoirs may not be at a stationary state; thus, enthalpy matching may mislead to the wrong models. | nb_NO |
