Hepguler, Barua, and Bard recently described the integration of a commercial reservoir simulator and a commercial production network simulator using the Parallel Virtual Machine interface. This paper outlines a somewhat different procedure, based on the same interface, that has been used to couple the same commercial black oil reservoir simulator ECLIPSE to a different surface model, namely the gas deliverability forecasting model FORGAS. For a given case, this new procedure yielded a significant reduction in computation time (a factor of about 30), when compared to the Hepguler et al model, without any loss in accuracy or function. A distinct advantage of this new coupled system is that it is available commercially, unlike the previously described integrated models which were proprietary developments, not available to other companies. P. 285

Over-simplification of the hydraulic component of the field production gathering network can often introduce serious errors in the reservoir simulation forecast. One principal problem involves determining the number of wells required to achieve a desired production target rate. Neither the engineer nor the automated drilling queue logic in the reservoir simulator can easily determine the effects of increased surface backpressure on existing wells caused by the connection of new wells to the surface gathering network. The integrated simulation of the reservoirs and pipeline gathering network connecting wells to the 200 MMscfd (5635 E03 m3/d) Sexsmith, Alberta sour gas processing plant demonstrates an approach taken to provide answers to problems typical of gas production operations. The paper reviews alternate approaches pursued prior to committing to the integration of a reservoir simulator and a pipeline network model. A review of the history matching procedures conducted to "tune" the linked models to match historical production rates and pressure losses in the network are presented. Information which is "passed" between models is presented as well as suggested software improvements to the integration of the reservoir and pipeline network models. P. 291

This report was prepared as an account of work sponsored by an agency of the United States Government.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractAs the demand for natural gas has increased, so has the need for accurately forecasting reservoir deliverability. Some common methods for forecasting include (a) production decline analysis; (b) a tank type reservoir model linked to a surface network model; and, (c) a reservoir simulator linked to a surface network model. The selection of the most appropriate method should consider the economic risk associated with the forecast and the availability of pertinent data necessary for the method being applied. In reality, other considerations such as time constraints, availability of certain in-house models, or the familiarity the engineer may have with one or more forecasting methods may be the governing factor in deciding which method might be applied. The paper compares the production forecasts generated for the three forecasting methods referenced evaluating such parameters as reservoir permeability and geometry and the effect of changing backpressure and well spacing. An assessment of the impact of these parameters and the relative magnitude difference of the forecast are presented to help provide guidelines for selecting a method, and appreciating the accuracy of methods in certain circumstances.This study found that production decline analysis generates the most conservative forecast of the three methods unless information is available to characterize the transient period in advance of the depletion decline. Production decline analysis, by virtue of its inherent assumptions, is unable to accurately predict the changes in volumes that result from changing surface operating conditions. A tank reservoir model linked to a surface network model is a good tool for field optimization, as long as the reservoirs are homogeneous and have good permeability. However, reservoir geometry can severely influence the interpreted results from such models, leading to erroneous conclusions. A reservoir simulator linked with a surface network model is the preferred choice for forecasting and optimizing the system for heterogeneous and/or low permeability reservoirs as long as sufficient reservoir data are available.

This is the first of two papers describing the application of simulator-optimization methods to a natural gas storage field development planning problem. The results presented here illustrate the large gains in cost-effectiveness that can be made by employing the reservoir simulator as the foundation for a wide-ranging search for solutions to management problems. The current paper illustrates the application of these techniques given a deterministic view of the reservoir. A companion paper will illustrate adaptations needed to accommodate uncertainties regarding reservoir properties. IntroductionAlthough reservoir simulation is a well-established component of reservoir management throughout much of the petroleum industry, little use has been made of reservoir simulation coupled with systematic optimization techniques, i.e. simulation-optimization.The main advantage of applying optimization tools, per se, to decision-making problems is that they are less restricted by human imagination than conventional case-by-case comparisons. As the number of competing engineering, economic, and environmental planning objectives and constraints increases, it becomes difficult for human planners to track complex interactions and select a manageable set of promising development strategies for examination. Using optimization techniques, the search can range over all possible combinations of variables, locating strategies whose effectiveness is not always obvious to planners.The advantage of coupling the reservoir simulator to these optimization tools is that the search for strategies can be based on the simultaneous evaluation of reservoir performance measures and other economic/environmental/policy considerations. It is no longer necessary to treat technical decisions driven by simulator forecasts of reservoir response and these other components of the decision-making process as separate steps.The single biggest obstacle to the application of optimization techniques using a reservoir simulator as the forecasting tool is the computational time required to complete a single simulation. Even the examination of 10 variations on a well-field design becomes cumbersome when a single run requires hours to complete. Extending the use of these simulators into optimization regimes involving hundreds or thousands of runs poses a computational problem bigger than most organizations are willing or able to tackle.The ANN-GA solution to this problem is to train artificial neural networks (ANNs) to predict selected information that the simulator would normally predict. A heuristic technique such as the Genetic Algorithm (GA) then searches for increasingly better strategies (for example, the most productive in-fill drilling pattern), using the trained networks to evaluate the effectiveness of each strategy in place of the

The objective of this paper is to couple wellbore and surface production facilities models with reservoir simulation for a shale reservoir that contains dry gas, condensate and oil in separate containers. The goal of this integration is to improve liquid recoveries by dry gas injection and gas recycling. Methods published up to now to investigate possible means of improving recovery from shales have concentrated on laboratory work and the reservoir itself, but have ignored the surface and wellbore production facilities. The coupling of these facilities in the simulation work is critical, particularly in cases involving condensate and oil reservoirs, gas injection and recycling operations. This is so because a change in pressure in the reservoir is reflected almost immediately in a change in pressure in the wellbore and in the surface installations. The development presented in this paper considers multi-stage hydraulically fractured horizontal wells. Dry gas is injected into zones that contain condensate and oil. Gas stripped from the condensate production is re-injected in the condensate zone in a recycling operation. The study leads to the conclusion that liquid recoveries can be maximized by utilizing continuous and huff and puff gas injection schemes. In general, huff and puff injection provides better results in terms of production and economics. Molecular diffusion is found to play a crucial role in continuous gas injection operations. Conversely, the effect of this phenomenon is negligible in huff and puff gas injection. This research demonstrates that proper design of wellbore and surface installations, including for example downhole pumps and compressors, is important as they play a critical role in the performance of production and injection operations, and in maximizing recovery of liquids from shale reservoirs. The novelty of the methodology developed in this paper is the coupling of models that handle surface facilities, wellbores, numerical simulation including oil, condensate and dry gas reservoirs, gas injection and gas-condensate recycling operations. Essentially the shale containers, wellbore and surface facilities are ‘talking’ to each other continuously. To the best of our knowledge this integration for shales has not been published previously in the literature.

This is the first of two papers describing the application of simulator-optimization methods to a natural gas storage field development planning problem. The results presented here illustrate the large gains in cost-effectiveness that can be made by employing the reservoir simulator as the foundation for a wide-ranging search for solutions to management problems. The current paper illustrates the application of these techniques given a deterministic view of the reservoir. A companion paper will illustrate adaptations needed to accommodate uncertainties regarding reservoir properties. IntroductionAlthough reservoir simulation is a well-established component of reservoir management throughout much of the petroleum industry, little use has been made of reservoir simulation coupled with systematic optimization techniques, i.e. simulation-optimization.The main advantage of applying optimization tools, per se, to decision-making problems is that they are less restricted by human imagination than conventional case-by-case comparisons. As the number of competing engineering, economic, and environmental planning objectives and constraints increases, it becomes difficult for human planners to track complex interactions and select a manageable set of promising development strategies for examination. Using optimization techniques, the search can range over all possible combinations of variables, locating strategies whose effectiveness is not always obvious to planners.The advantage of coupling the reservoir simulator to these optimization tools is that the search for strategies can be based on the simultaneous evaluation of reservoir performance measures and other economic/environmental/policy considerations. It is no longer necessary to treat technical decisions driven by simulator forecasts of reservoir response and these other components of the decision-making process as separate steps.The single biggest obstacle to the application of optimization techniques using a reservoir simulator as the forecasting tool is the computational time required to complete a single simulation. Even the examination of 10 variations on a well-field design becomes cumbersome when a single run requires hours to complete. Extending the use of these simulators into optimization regimes involving hundreds or thousands of runs poses a computational problem bigger than most organizations are willing or able to tackle.The ANN-GA solution to this problem is to train artificial neural networks (ANNs) to predict selected information that the simulator would normally predict. A heuristic technique such as the Genetic Algorithm (GA) then searches for increasingly better strategies (for example, the most productive in-fill drilling pattern), using the trained networks to evaluate the effectiveness of each strategy in place of the

No abstract