A Simplified Performance-Predictive Model for In-Situ Combustion Processes
- J.F. Genrich (Sohio Petroleum Co.) | G.A. Pope (U. of Texas)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- May 1988
- Document Type
- Journal Paper
- 410 - 418
- 1988. Society of Petroleum Engineers
- 5.7.2 Recovery Factors, 4.1.2 Separation and Treating, 4.6 Natural Gas, 5.2.1 Phase Behavior and PVT Measurements, 7.1.9 Project Economic Analysis, 5.4 Enhanced Recovery, 4.1.5 Processing Equipment, 5.2 Reservoir Fluid Dynamics, 5.1.1 Exploration, Development, Structural Geology, 4.3.4 Scale, 7.1.10 Field Economic Analysis, 5.4.6 Thermal Methods, 4.1.4 Gas Processing
- 1 in the last 30 days
- 324 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
Summary. A one-dimensional (ID) model to perform rapid calculations of fluid production history for dry and wet forward in situ combustion processes is presented. The predominantly explicit, discrete timestep processes is presented. The predominantly explicit, discrete timestep formulation divides the reservoir into four zones of constant properties, separated by shock fronts. Model results are compared with a ID simulator study of a combustion-tube experiment and with field data from the Suplacu de Barcau project.
in-situ combustion has been applied for several decades as a thermal recovery method for highly viscous crudes. In the early projects, only air was injected to provide the necessary oxygen to sustain a burning front (dry combustion). Later, it was realized that recovery efficiency and project economics can often be improved by simultaneous injection of water (wet combustion) or by higher oxygen concentrations in the injected gas (enriched combustion).
The implementation of an in-situ combustion project involves substantial capital investments and operating costs. Reliable production estimates are required to assess the profitability of the process for a specific reservoir and to compare it with that of other recovery methods.
A model is developed that calculates production of oil, gas, and water from forward in-situ combustion processes for given reservoir characteristics and injection rates. The simple, ID structure yields fast computing times. The model can be used to compare production response from various process modes (dry, wet, and production response from various process modes (dry, wet, and enriched) and to determine an optimum combination of process parameters. This provides valuable assistance in deciding whether to parameters. This provides valuable assistance in deciding whether to reject a reservoir as an in-situ combustion candidate or to initiate an in-depth study with more comprehensive and sophisticated tools.
Predictive models for in-situ combustion have been formulated since early field applications of the process. Depending on their basic approach (analytical, numerical, scaled experimental, empirical, or correlatives) and the specific objectives under which they were developed, they vary not only in complexity but also in applicability regarding process mode and parameters.
This model combines analytic descriptions of fractional flow, combustion reactions, phase behavior, and heat transfer into a fast algorithm that uses a maximum number of commonly available reservoir and fluid parameters as input data.
General Process and Model Characteristics
The complex and highly interdependent mechanisms of an in-situ combustion process prohibit the development of a simple I D model that accounts for all major phenomena involved. These include a variety of chemical reactions (oxidation, decarboxylation, pyrolysis, and secondary reactions such as cracking. formation of H2S, pyrolysis, and secondary reactions such as cracking. formation of H2S, and mineral decomposition), simultaneous flow of three phases, formation and deposition of a solid coke phase, phase changes (condensation and vaporization), thermal expansion, distillation, oil swelling, and gravity override. In addition, frontal instabilities are caused by reservoir heterogeneities.
Core samples and laboratory experiments show that a reservoir subjected to in-situ combustion can be divided into the following zones (listed in downstream direction): burned zone, combustion zone, evaporation zone, condensation zone, water bank, oil bank, and zone of initial conditions. They are characterized by typical temperature and saturation profiles (Fig. 1). The model considers only four homogeneous zones (burned, combustion, steam, and cold; designated by subscripts b, c, s, and f, respectively), as shown in Fig. 2. Movement and saturations of each zone are calculated on a timestep basis from ID solutions of the fractional-flow equations by the method of characteristics. This approach has been successfully applied in simplified models for several other EO processes. It was also used by Wilson et al. to formulate a performance-predictive model for dry in-situ combustion. That performance-predictive model for dry in-situ combustion. That model considers three zones ahead of the combustion front: water bank, oil bank, and zone of initial conditions. All zones are at initial reservoir temperature, and three-phase flow is permitted only in the water bank. Water is assumed to be immobile in the other zones.
Use of the frontal-advance concept to determine average temperature, saturations, and frontal velocity for each of the four zones leads to highly implicit combinations of fractional-flow equations and energy balances. To achieve a decoupled, more explicit formulation, further simplifications have to be introduced. Model assumptions and limitations are as follows.
1. Pseudocomponents considered are water, of, initial hydrocarbon gas, oxygen, uncondensable gas other than oxygen, and the rock matrix.
2. There are two stationary (coke and rock) and three flowing phases (aqueous, oleic, and gaseous). phases (aqueous, oleic, and gaseous). 3. Except for the gaseous phase, all others are pure; i.e., they consist of only one pseudocomponent: fuel, rock matrix, water, or oil. Hence, any form of rock/fluid interaction or solubility of components in oil and water is disregarded (for example. distillation, adsorption, and rock dissolution).
4. Species other than water in the gaseous phase are handled outside of the combustion zone as a single pseudocomponent "uncondensable gas." This includes oxygen, gaseous combustion products such as CO and CO2, and inert components (nitrogen).
5. Initial hydrocarbon gas is assumed to be displaced piston-like by the uncondensable gas.
6. All other mobile hydrocarbons are regarded as a single, non-volatile pseudocomponent oil. Vaporization of lighter hydrocarbons is thus neglected.
7. There is an immobile coke phase in the combustion and burned zones.
8. No aqueous phase exists in the combustion zone; i.e., no super- wet combustion. 9. No steam is in the burned or cold zones.
10. All combustion water is transported ahead of the combustion front.
11. Capillary pressure is negligible.
12. ID horizontal linear flow exists.
13. There is no change in PV.
14. The amount of oil displaced in the steam zone is calculated directly from an empirical. residual-oil-to-steam correlation.
15. The average temperature of the burned zone is equal to the sandface temperature at the injection wells.
|File Size||694 KB||Number of Pages||9|