Abstract The interactions of heavy oil and injected steam in the mature steamflood at the Kern River Field have been extensively studied to gain insight into the effect of steam on compositional changes of oil during the recovery process and to provide input for compositional thermal simulation. Steam distillation behavior of this 13 API California oil between 300 and 500 F under a variety of process conditions, along with extensive analysis of distilled hydrocarbons were incorporated to give a more in depth description of what is happening to the oil and what changes are occurring in the distillates and produced oil. This information was further integrated with analysis of the field distillate, 'casing blow,' to describe what is happening in the field. The results show that steam distillation is temperature dependent and is more important than originally thought. The data developed in this study is a basis for improvement of numerical thermal models with potential for better designed steamfloods and reservoir management. The results may also impact certain logging techniques used in steamfloods and possible heavy oil upgrading techniques. Kern River oil is more than 10% distillable at 300 F and 15% distillable at 400 F in dynamic laboratory steam distillation tests at steam throughputs of 4 times the initial volume of oil. Distillate physical properties of density, viscosity, molecular weight, and hydrocarbon composition of the distillates changed significantly. Distillate density, viscosity, and molecular weight properties increased in value with increasing steam throughput, and at higher temperatures. This information is important in the tuning of equations-of-state, including hydrocarbon-water interaction parameters for compositional thermal simulation. Analysis of the field distillate, 'casing blow,' showed properties similar to laboratory distillates at low steam throughputs. The observation of a light field distillate production in a mature steamflood compared to laboratory measurements implies that the casing system temperature is a major controlling factor in 'casing blow' composition and quantity. Background The phase equilibrium behavior of reservoir fluids is an important phenomenon in petroleum production, particularly in enhanced oil recovery processes. However, phase behavior for heavy oils (< 15 API) under steamflood has generally been felt to be either unimportant or to be a minimal effect to be neglected. A major question exists about whether or not the phases and fluids in a steamflood are in equilibrium. Proper modeling of a reservoir production process would be expected to include knowledge of the phases and their equilibrium compositions. In heavy oil, devoid of significant C1 - C6 composition, it has been sufficient in previous practice for steamflood modeling purposes to treat the oil as a dead oil or as a non-volatile phase. A history match numerical study of steamflood performance in the Kern River Field treated the oil as non-volatile, and was conducted without the inclusion of hydrocarbon compositional effects. Through the classic works of Willman et al., Volek and Pryor, and Closmann and Seba steam distillation has been shown to be an important component mechanism in the overall steamflooding process. The practical limit of how much of a reservoir fluid can be distilled, is obtained in dynamic steam distillation experiments developed by Wu and Brown, extended by Hseuh, Hong, and Duerksen, and refined by Wu et al. This body of work demonstrates that steam distillation is an operative mechanism in laboratory models, but it has been difficult to translate this to a quantitative contribution to the field steamflooding recovery process. Laboratory steam distillation experiments have generally been conducted as dynamic tests, that may or may not be near equilibrium. Experiments near equilibrium with extensive analysis of the phases will yield values for the vapor-liquid equilibrium (VLE) ratios (K-values), another way of assessing the importance of compositional changes in steamflooding. A major recent steam distillation study by Northrup and Venkatesan has been completed on the South Belridge oil. P. 359^