Artificial lift is used worldwide in approximately 85% of the wells, thus its impact in overall efficiency and profitability of production operations can not be overemphasized. Selection of a particular lift method for a given application is carried out, to a large extent, by production engineers utilizing the method more readily available to them both in terms of their past experience and current access to knowledge and technology. It has been recognized for a long time, that this is a shortcoming of most engineering disciplines and results, more often than not, in sub-optimal designs. Expert systems can help engineers in quickly selecting, from updated technology, the best options available to them for the problem at hand. The work presented here describes knowledge gathering and coding techniques, general program structure, program validation runs, and field verification of SEDLA, an expert system developed to help in the selection of the best lift method for a well or group of wells.1

The computer program developed can be effectively used for the design and troubleshooting of conventional plunger lift installations.Marcano L.; Chacfn J., Intevep S. A. ABSTRACT A mechanistic model for conventional plunger lift installation design is presented. This model applies to oil wells with high gasliquid ratios and is based on the fundamental equations of momentum balance and mass conservation. The model considers that the rate of liquid fall-back produced when the plunger rises is a linear function of the average rise velocity. The obtained set of ordinary non-linear differential equations needs to be solved numerically. A Fortran computer program for both personal and mainframe based computers was implemented. This permits the calculation of fall-back losses and allows the estimation of optimum cycle time.The model, verified with field quality data, has been shown to represent observed production trends in the wells studied, in particular the shape of the fall-back versus driving pressure ratio on the plunger is reproduced correctly. Differences between liquid/gas production and model estimations fall in the 15 to 20 percent range. This result is encouraging, considering the difficulty involved in obtaining accurate field measurements. The computer program developed can be effectively used for the design and troubleshooting of conventional plunger lift installations.

A novel mechanistic model for the design of plunger assisted intermittent gas lift installations is presented. The model considers the complete production cycle: accumulation of reservoir fluids in the wellbore, annulus pressure build-up, plunger/liquid slug rise, liquid production into flowline. It is based on the fundamental mass balance and conservation of momentum equations and considers all important parameters: reservoir pressure, inflow performance, tubing and flowline characteristics, plunger characteristics, available injection pressure, surface injection choke design, gas lift valve design and performance, etc. The model solution requires numerical techniques, and it is implemented as a computer program for personal computers and mainframe workstations. One of the most innovative characteristics of the model is that it incorporates, as part of the dynamic numerical simulation, experimental plunger rise data relating instantaneous plunger velocity to instantaneous liquid slippage past the plunger. As a result, liquid "fallback" losses can be estimated and need not be assumed. The model and computer program has been verified and validated both qualitatively and with field data. Comparisons of model predictions versus field data are included. To the best of the authors knowledge, this the first published work in which a full model of this type of artificial lift installation is undertaken. Finally, it is shown how the model can be used to obtain an optimum design of a new installation, considering both gas consumption costs and crude recovery. Furthermore, examples are included to show how to use the program to effectively optimize existing installations. Introduction Plunger lift is a well known artificial lift method. It was originally developed to unload liquids from gas wells. Its use was extended to oil wells with high gas-oil ratios and those wells that stop producing due to paraffin deposition in the tubing string. More recently, plungers have been used in conjunction with intermittent gas lift. Theoretically, the main benefits of utilizing plungers with gas lift are:An increase in liquid production due to a reduction of fallback losses.An increase in lifting efficiency due to a decrease of the gas injection requirement. Although this combination of intermittent lift with plunger looks potentially very attractive, it has received very little attention in the published literature. Our literature survey showed, among other things:No publications or case studies dealing with the effects of introducing a plunger in an intermittent gas lift installation.Only one publication (White5) dealing specifically with experimental studies showing the benefit of gas lift/plunger combinations.No publications dealing specifically with the modeling of production installations using this method. It is expected that by the mid to late nineteen-nineties, 600 MBOPD will be gas lifted intermittently from the shallow waters of Lake Maracaibo. Means must be searched to produce these wells in the most efficient manner. It is within this framework that the present investigation was initiated. Its immediate objective is to develop a mathematical model of plunger lift assisted, intermittent gas lift installations (IGPL). Furthermore, the practical application of the model to optimum design of new installations and optimizations will be explored. Literature Review A recent publication (Marcano and ChacÍn12) provides a good review of previous work related to conventional plunger lift and will not be repeated here. In Figure 1, for purposes of illustration, we present an schematic drawing of a commonly used plunger lift assisted intermittent gas lift installation (IGPL). In operation, the lifting energy is provided by compressed gas from surface installations that flows into the tubing thru a gas lift valve. In conventional intermittent gas lift, the gas injected penetrates the liquid slug and assumes a particular bullet shaped profile known as a Taylor bubble. This penetration causes some of the liquid slug fluid to slip downward around the gas bubble, with the consequent loss of production (fallback). The use of a metallic plunger in an intermittent gas lift installation provides a solid interface between the injected gas and the liquid slug. There is experimental evidence4,5,6 that, under certain conditions, the solid interface provided by the plungers is responsible for a reduction of fallback losses associated with conventional intermittent gas lift. Further benefits, in terms of reduced gas injection requirements, can be realized when combining intermittent gas lift and plungers. White5 studied experimentally intermittent gas lift with and without a plunger. Some of his experimental results are presented in Figure 2. As shown, the use of a plunger can reduce liquid fallback and improve lift efficiency in a significant manner, in the low driving pressure ratio (Pc /Pt) operating range. White found that a plunger with a hole through the center (along its longitudinal axis) caused further improvement in efficiency. From this work one can conclude that the classic picture of a plunger being a moving partition between liquid slug above and gas below is neither correct nor desirable.

SAGD ESPs run at the highest motor temperatures current technology allows. However, they cool very rapidly when shutdown. High cooling rates promote motor oil volumetric contraction, eventually leading to wellbore fluid ingress and short-circuited motors. The Paper presents successful field tests designed to decrease ESP cooling rates by inducing controlled deadheads, rather than shutting down ESPs. Various extended deadhead field trials (up to 70+ days duration) validated the approach, while confirming that no deadhead related ESP damage was induced. ESP temperature changes were measured using fiber optics strings installed as part of the usual completion in 60+ wells, during a four week-long field-wide plant maintenance turn-around. While cooling rates varied somewhat from well to well, they all showed very similar behavior and were very well fitted with a log-normal distribution, R2factor > 95%. Most ESP temperatures decreased between 50°C to 120°C in a week. This data was used as a general baseline to support the deadheading field trials. An ESP was fitted internally with an RTD at the base of the motor and externally with a clamped fiber optics string. This ESP was operated normally at 55 Hz for a few months. An 8-hour shut down test established an initial base line cooling rate of 6.6°C/hour. Subsequent 6-hour deadhead tests at 30Hz and 45 Hz showed decreased cooling rates of 4.0°C/hour and 2.2°C/hour, respectively. This result clearly established the potential to deadhead at different frequencies to obtain different lower cooling rates. Finally, two extended deadhead tests (3 and 10 weeks in duration) were executed to help determine if it was possible to induce damage in SAGD ESPs by deadheading, as is usually the case in most non-thermal applications. These ESPs operated normally after the extended tests and one was dismantled upon failure, looking for any signs of deadhead damage. Results presented show that deadheading SAGD ESPs provides the opportunity to safely minimize ESP thermal cycles, which could lead to a significant improvement in ESP run life.

Maracaibo Lake wells are currently producing approximately 800 mstbd of crude oil from some 7000 wells. 90% of these well are in gas lift and approximately 760 wells are being lifted intermittently. Natural reservoir decline will increase the number of wells that will require intermittent lift at an estimated rate of 150 wells per year. Continuous research and modeling efforts have been performed over the last thirty years to understand the dominant physical characteristics and develop mechanistic computational tools for the design of the various existing intermittent lift schemes (conventional, plunger assisted, chamber lift). The present paper will review relevent published work regarding intermittent lift and describe key modeling considerations for each scheme. Furthermore, we will focus on proper selection of intermittent lift scheme as a function productivity index, reservoir pressure, wellbore equipment and completion and other inportant well parameters.