TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractIt is well known that the deliverability of gas condensate wells can be impaired by the formation of a condensate bank once the bottomhole pressure drops below the dew-point. There have been many excellent laboratory studies on gascondensate relative permeability that describe this phenomenon, but integrated laboratory-simulation-field studies that compare systematic predictions to field performance are few. This work presents a careful and comprehensive evaluation of deliverability of several wells in a low permeability gas condensate reservoir. This sandstone reservoir has permeability ~ 10 md, porosity ~ 10%, and condensate yield ~ 50 bbls/MMscf.Our approach consists of the following steps -(i) careful selection of core samples that cover the range of expected responses; (ii) design of fluid systems that mimic reservoir fluids, but at lower temperatures; (iii) appropriate relative permeability measurements k rg = f(k rg /k ro , N c ) for a range of flow conditions; (iv) fitting this data to several different relative permeability models for use in reservoir simulators;(v) use of analytical spreadsheet tools to calculate deliverability; (vi) detailed single well compositional models with realistic geology and boundary conditions extracted from full field models, honoring complex producing rules and differential depletion; (vii) comparison of prediction to four wells, three vertical and one inclined.We have been able to reasonably predict the performance of these wells. The productivity reduction of the wells were found to be in the range of 80%, the majority of which occurred in the initial phases of production. Our ability to reasonably predict the performance has given us confidence that our approach, including measuring only the relevant portion of the relative permeability curves and using synthetic fluids, may be sufficient.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractSignificant productivity loss occurs in gas-condensate wells when the bottom hole flowing pressure drops below the dewpoint pressure. The decline in productivity is due to nearwell accumulation of condensate in the reservoir rock, which is significant even for wells producing very lean gas with liquid dropout values less than 1%. Many different methods such as hydraulic fracturing, dry gas injection and solvent injection have been proposed and implemented to stimulate such wells. However, all of these methods offer short-lived stimulation and are sometimes not profitable. New experimental core flooding data using chemical treatments show that the steady-state gas and condensate relative permeability in both outcrop and reservoir sandstones can be increased by a factor of 2 to 3 over a wide range of temperature (145 to 275 °F). Spectroscopic data show that the sandstone surface remains modified by the chemical even after flooding the core with large volumes of gas. A relative permeability model that includes effects such as the decrease in the residual condensate saturation after treatment and the effect of capillary number is presented. Fine-grid compositional simulations of a single-well treatment were done using the calibrated relative permeability model to investigate the performance of chemical treatments under field conditions as a function of variables such as treatment radius. These simulations show that chemical treatments have the potential to greatly increase production at low cost relative to the increased revenue since only the near-well region blocked by the condensate needs to be treated.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractSignificant productivity loss occurs in gas-condensate wells when the bottom hole flowing pressure drops below the dewpoint pressure. The decline in productivity is due to nearwell accumulation of condensate in the reservoir rock, which is significant even for wells producing very lean gas with liquid dropout values less than 1%. Many different methods such as hydraulic fracturing, dry gas injection and solvent injection have been proposed and implemented to stimulate such wells. However, all of these methods offer short-lived stimulation and are sometimes not profitable. New experimental core flooding data using chemical treatments show that the steady-state gas and condensate relative permeability in both outcrop and reservoir sandstones can be increased by a factor of 2 to 3 over a wide range of temperature (145 to 275 °F). Spectroscopic data show that the sandstone surface remains modified by the chemical even after flooding the core with large volumes of gas. A relative permeability model that includes effects such as the decrease in the residual condensate saturation after treatment and the effect of capillary number is presented. Fine-grid compositional simulations of a single-well treatment were done using the calibrated relative permeability model to investigate the performance of chemical treatments under field conditions as a function of variables such as treatment radius. These simulations show that chemical treatments have the potential to greatly increase production at low cost relative to the increased revenue since only the near-well region blocked by the condensate needs to be treated.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractIt is well known that the deliverability of gas condensate wells can be impaired by the formation of a condensate bank once the bottomhole pressure drops below the dew-point. There have been many excellent laboratory studies on gascondensate relative permeability that describe this phenomenon, but there are few integrated laboratorysimulation-field studies that compare systematic predictions to field performance.We present extensive experimental relative permeability data sets on some sandstone reservoirs. These data span the k rg /k ro and capillary number parameter space. We discuss the experimental procedures, and the design of fluid systems that mimic reservoir fluids, but at lower temperatures. Next we demonstrate various steps involved in our approach by modeling a gas condensate well with field production history. Here we first measured relative permeability data on core samples from the reservoir and fit them to capillary number dependent relative permeability models. Then, we performed detailed single well compositional modeling with realistic geology and boundary conditions. Finally, we compared the predictions to actual production data, and found that the match was quite good. The productivity reduction was found to be in the range of 80%, the majority of which occurred in the initial phases of production. Our ability to reasonably predict the well performance has given us confidence that our approach, including measuring only the relevant portion of the relative permeability curves and using synthetic fluids, may be sufficient.

Predicting production from gas-condensate wells requires an accurate relative permeability model when a condensate bank forms. At high flow rates typical of many gas-condensate wells, the relative permeability is rate dependent. Such rate dependence can be modeled using a capillary number to calculate the decrease in residual saturations and the corresponding increase in relative permeability as viscous forces become dominant over the interfacial forces. New steady-state relative permeability data have been measured over a wide range of capillary numbers including very high values corresponding to the near-well region. These measurements have been made on several reservoir rocks as well as outcrop rocks and over a range of temperature, pressure, connate water saturation and hydrocarbon composition typical of gas-condensate reservoirs. PVT data of gas-condensate fluids can be used to predict the ratio of the gas to the condensate relative permeability and this simplifies the measurements and modeling since only data corresponding to the pressures near wells are needed. A relative permeability model developed at the University of Texas was tested using both new data and data from the literature. With only one parameter set, essentially all of the data for all rocks and conditions was fit within experimental uncertainty. Introduction In gas condensate reservoirs, when the bottomhole pressure in flowing wells falls below the dew point pressure of the fluid, a liquid hydrocarbon phase called condensate is formed and trapped by capillary forces. The liquid condensate continues to accumulate, occupying portions of the rock pores that otherwise would be available for gas flow, and thus impeding gas flow, until a critical liquid saturation is reached that is similar to the value for residual oil saturation that would form in the same rock under the same flow conditions. Once the critical liquid saturation is exceeded, both the condensate and gas flow towards the wellbore, but condensate continues to accumulate until a steady-state saturation is reached that is somewhat higher than the critical condensate saturation. Condensate banking can reduce the well productivity significantly, in several instances by a factor of 2 to 4. Afidick et al.,1 Barnum et al.,2Engineer3 and Ayyalasomayajula et al.4 have reported field data that show significant productivity loss due to condensate accumulation. Several investigators4–13 measured the effect of capillary number on gas-condensate relative permeabilities. However, most of the laboratory data are at low capillary numbers. Much less data are available at high capillary numbers corresponding to the condensate banks near the production wells. In this paper, we present gas-condensate relative permeability data over a wide range of capillary numbers measured on both sandstone and limestone rocks. Capillary Number Dependent Relative Permeability Model Pope et.al.14 presented a relative permeability model for gas and condensate relative permeabilities as a function of capillary number. The relative permeability krl of each phase lis calculated by interpolating between the measured value at low capillary number and a straight line corresponding to a very high capillary number: (1) where (2) The residual saturation of each phase l is modeled as a function of trapping number as shown below: (3)