Insights Into Heat Transfer For Thermal Oil Recovery And Estimation Of Key Parameters For Oil Well From Temperature Data

Thermal oil recovery is the principal enhanced oil recovery (EOR) technique currently in use to recover heavy and viscous crude oil, which is defined as a progress in which high-temperature fluids are injected the into reservoir heating heavy oil. As the oil temperature increases, its viscosity decreases dramatically, making the heavy or crude oil more mobile for EOR. Predictions of the field behavior of the process require an accurate accounting of the thermal energy injected in to reservoir. Therefore, study on heat transfer for oil well dominates the thermal recovery process. Formation and reservoir thermal properties are important parameters for evaluating thermal efficiency of thermal recovery process, and oil saturation of reservoir is also a crucial criterion for evaluating exploitation value of oil field. Moreover, gas/oil flow rate of production well can be used to evaluate oilfield production directly.In this study, based on the insight into heat transfer for thermal recovery well, a novel method was presented for predicting the formation and reservoir thermal properties, oil saturation in reservoir and oil production rate from temperature data. In addition, an experiment was built to study the heat transfer characteristic for reservoir during injection and production process, and investigate the mechanism of thermal oil recovery.Firstly, for the steam injection well, this paper made full study on the wellbore, formation and reservoir heat transfer, and built a novel heat transfer model which took into account not only the wellbore-formation heat transfer but also the wellbore-reservoir heat transfer, a steam temperature model for the entire wellbore was obtained by the built heat transfer model. Both steam pressure and quality model were also derived from investigation of the steam flow. The simulated results by the steam parameters models were quite consistent with the log data. Consequently, the built heat transfer model not only could simulate the steam parameters accurately, but also could show the difference between wellbore-formation heat transfer and wellbore-reservoir heat transfer, the steam temperature, pressure and quality would decrease more in wellbore over reservoir. In addition, this paper built a heat loss model for steam injection well with consideration of wellbore heat capacity, it was found that the effect of wellbore heat capacity on heat loss from steam to formation was quite considerable, and especially larger during short steam injection time.Based on the heat transfer model of steam injection well, this study presented an effective inversion method for estimating the reservoir thermal properties through temperature log data, the inversion method is of Monte Carlo stochastic approximation method. Sensitivity analysis was firstly conducted to investigate the sensitivity of uncertain parameters, determining the inversion sequence. The results show that the sensitivity of thermal conductivity is stronger than that of volumetric heat capacity. Furthermore, the results also show that the sensitivity of reservoir thermal properties is bigger than that of formation thermal properties. The inversion method was applied to estimate the reservoir and formation thermal properties of two steam injection wells, it was found that the relative error of thermal conductivity for the two wells were 2.9% and 6.5%, and the relative error of volumetric specific heat capacity were 6.7% and 7.0%, which demonstrated the feasibility of the proposed method for estimating the reservoir thermal properties.Above all, this study firstly presents a layered inversion method for estimating the spatial distributions of formation and reservoir thermal properties from temperature data. Consequently, the present method not only could estimate effectively the thermal properties distributions which showed variability over the whole spread in depth; but also was able to predict reservoir depth by the significant change in thermal properties distributions between formation and reservoir. It can break the situation that thermal properties are usually assumed to be depth-invariant in thermal design and numerical simulations.For reservoir oil saturation, this study also presented a novel method for obtaining the distribution of oil saturation in reservoir, which was predicted by a semi-empirical model for thermal conductivity-oil saturation with the inversion thermal properties distributions. The estimated oil saturation distribution by the presented method agreed well with the field data with relative error below 10%. Moreover, the large difference in oil saturation distribution between the reservoir and formation near reservoir can also predict the reservoir depth.For the production well, this study presents an effective inversion method for estimating the oil flow rate distribution in wellbore from temperature log data and the proposed method is based on the temperature-flow rate model of production well. Therefore, a novel heat transfer model for production well with consideration of wellbore heat capacity is built first to obtain the temperature-flow rate model. Then sensitivity analysis is applied to investigation of the correlation between the flow rate and fluid temperature for the inversion method, the result indicates that the sensitivity of flow rate is larger than that of formation thermal properties. Lastly, the inversion method is used to estimate the flow rate three production wells. The relative error mean inversion is less than 10%, and the points in middle wellbore gets lower relative error rate for flow rate estimation; the prediction error rate at most middle points is less than 3%. It indicates that the temperature log data of the middle wellbore point is more significant for flow rate inversion method, which may cut the cost of measurement by reducing the test points.For the water injection well, with the thermal properties inversion method for steam injection well, this study uses the special layered inversion method appropriate to estimate the formation thermal properties profiles water injection well from temperature log data. The proposed method is based on the water temperature model. Therefore, a novel heat transfer model for water injection well was built firstly to get the temperature model. Then sensitivity analysis was conducted to investigate the correlation between the thermal properties and water temperature, determining estimation sequence. The sensitivity analysis concludes that heat capacity showed stronger correlation with temperature than thermal conductivity, which is contrary to steam injection well sensitivity result; and the water temperature in water decreased with the depth and then increased, difference from steam temperature distribution in steam injection well, both the results indicate the difference between the water injection well and steam injection well. Consequently, the layered inversion results not only distinguished the regional difference of formation, but also illustrated spatial coverage and depth variability of thermal properties.In the last part of the paper, a novel experiment is built to investigate the heat transfer characteristic of water injection reservoir. Based on the experiment, this paper studies on the effect of injection conditions (injection temperature, injection rate) and reservoir properties (porosity and oil saturation) on the reservoir temperature distribution, and focuses on the effect of reservoir temperature on the production parameters, such as, oil flow rate, oil production, fluid production, oil-water ratio, thermal recovery efficiency, and so on, investigate the thermal oil recovery mechanism ultimately. It is found that the oil production and recovery efficiency would increase largely with improving the injection temperature, the oil production with injection temperature 75 ℃ is about twice of that of with injection temperature 20 ℃ at injection time 6000s. Inversely, the oil-water ratio and oil saturation would decrease obviously with injection temperature increasing. Above all, this paper presents a novel method predicting the transient variation of oil flow rate and saturation as well as porosity from reservoir temperature data. The prediction error for both oil saturation and flow rate are below 10% at all time points, showing high precision for the presented method.