Liquid loading occurs when gas production declines to a rate that is insufficient to lift the associated liquids to surface. At lower rates, gas production becomes
intermittent and eventually stops entirely. However, liquid holdup in the horizontal section may impair production before loading in the production tubing becomes evident. Holdup in the horizontal section can lead to slug flow from the horizontal wellbore to the tubing and to an earlier onset of liquid loading in the tubing.
This paper presents liquid holdup data from a single onshore horizontal tight gas well, obtained through video logging. A transient multiphase flow model is then used to match the observed conditions.
The results from the transient multiphase flow model were found to be consistent with the measured data acquired from the video-logging. Sensitivity analyses
were performed with normalized trajectories representing toe-up, toe-down, undulating, and complex drilling profiles. Sensitivities to variations in the liquid-gas ratio
and the distribution of the reservoir inflow were also investigated. The results of the transient multiphase flow modelling support the conclusion that complex trajectories are more prone to production losses caused by liquid holdup.
The implications of this conclusion for trajectory optimization and tubing landing depth selection are explored. Modelling liquid holdup can lead to improvements in planning new drilling projects, mitigating the impact of liquid loading on long-term performance.
Porosity is a key reservoir parameter and high accuracy is needed to properly estimate reserves. But even though there is a long history of porosity measurements and various tools from which to derive it, this can still remain a difficult task. None of the logging tools directly measure porosity but instead respond to density, lithology and fluid. Combining different measurements can help to solve for porosity but also brings the complexity of invasion as all the tools do not have the same radial response. This problem is even more complex when dealing with gas formations as the fluid effect on the measurements is very high.
This paper looks at various methods to improve porosity computations via the integration of Nuclear Magnetic Resonance (NMR) and other porosity measurements in South China Sea gas reservoirs.
As illustrated in Fig. 1 the gas reservoirs in the South China Sea are composed of sand shale sequences with good porosity. Wells are drilled with oil base mud, invasion is shallow and gas effect is large on the neutron and density measurements. Fig. 1 shows a strong neutron-density gas effect and highlights the importance of fluid corrections to obtain reservoir porosity.
As gas effect is opposite on the neutron and density measurements, the usual workflow is to compute an apparent porosity from density assuming a water filled rock and then apply a weighted average of this apparent density porosity with an apparent neutron porosity.
Fig. 2 shows the result of such an approach and the comparison to core. One can observe that in some zones the match is good and not as good in others. One possible reason is a variation in clay content, clay contains neutron absorbers which can have a large impact on the neutron. This effect seems unlikely as these sand reservoirs are clean with minimal amounts of clay. Another possible explanation is that computing a weighted average assumes that the gas effect is similar on neutron and density, but as these tools see different volumes of rock, this assumption is incorrect when invasion is variable.
This paper illustrates the limitation of using neutron-density measurements to compute porosity and the necessity to include other measurements to reduce uncertainty.
In order to accurately compute porosity and adequately correct for invasion, one needs to integrate measurements that respond to similar volumes of rock. Fig. 3 shows the geometrical responses of density, neutron and two NMR tools. The Combinable Magnetic Resonance (CMR) tool has a single depth of investigation (DOI) while the Magnetic Resonance Scanner (MR Scanner) has three depths of investigation. One can observe that 80% of the density information comes from the first four inches away from the borehole while this same volume represents only 15% of the neutron response. This explains why shallow invasion effects cannot be properly compensated when combining neutron and density measurements, any small variations in the invasion will have different effects on the density and neutron.
NMR is another measurement which can be combined with density to compute porosity. Similar to neutron, NMR reads low in the presence of gas. This is a consequence of gas' low hydrogen index (HI) and of its hydrogen polarization deficit due to the long longitudinal relaxation time (T1) and the limited acquisition wait time.
A quantitative workflow called Density Magnetic Resonance Porosity (DMRP) combining density and NMR measurements was developed in 1998 (Freedman et al, 1998). NMR is acquired with short wait time to boost the gas effect (apparent porosity deficit) and total porosity is computed using a weighted average of density derived and apparent NMR porosity. The weights are computed from gas properties and NMR acquisition parameters.
In shale plays, as with all reservoirs, it is desirable to achieve the optimal development strategies, particularly well spacing, as early as possible, without overdrilling. This paper documents a new technology that can aid in determining optimal development strategies in shale reservoirs. We integrate a decline-curve-based reservoir model with a decision model that incorporates uncertainty in production forecasts. Our work extends previous work by not only correlating well spacing and other completion parameters with performance indicators, but also developing an integrated model that can forecast production probabilistically and determine the impact of development decisions on long-term production.
A public data set of 64 horizontal wells in the Barnett shale play in Cooke, Montague and Wise Counties, Texas, was used to construct the integrated model. This part of the Barnett shale is in the oil window and wells produce significant volumes of hydrocarbon liquids. The data set includes directional surveys, completion and stimulation data, and oil and gas production data. Completion and stimulation parameters, such as perforated interval, fluid volume, proppant mass, and well spacing, were correlated with decline curve parameters, such as initial oil rate and a proxy for the initial decline rate, the ratio of cumulative production at 6 months to 1 month (CP6to1), using linear regression. In addition, a GOR model was developed based on thermal maturity and average GOR versus time. Thousands of oil and gas production forecasts were generated from linear regression and GOR models using Monte Carlo simulation, which serve as the input to the decision model. The decision model then determines the impact of well spacing and other completion/stimulation decisions on long-term production performance.
The technology introduced in this paper can be used to help operators in unconventional reservoirs reach optimal spacing and completion strategies earlier in the lives of these reservoirs, which could accelerate production and increase reserves.
The design of the wells in Cuervito Field is generally composed of three casings, in order to access the highest yielding sands. Drilling operations are characterized by the presence of a balloning effect, causing the loss of circulation that occasionally difficult meeting the planned target for the well.
The balloning effect, breathing formation or phenomenon of loss and gain, is an elastoplastic behavior of the walls of a well that usually originates in impermeable formations. When circulating drilling fluids it creates an additional pressure that is commonly defined as the equivalent circulation density (ECD), which when imposed on the walls of the well, causes them to dilate. When circulation is stopped, the ECD decreases, the diameter of the hole shrinks along with the pressure, creating a flow return to the well. High temperatures also play a role in the balloning effect. The expansion and contraction of the holes in the Cuervito Field is highly remarkable.
In order to minimize the encountered problems and to optimize drilling operations, a series of steps were prepared in order to study Cuervito Field's 16 wells. This study contains geological and geopressure information. As a result of this study two different options were encountered: (i) Adding an additional casing or (ii) An unconventional drilling technique.
The first alternative (adding an additional casing to the well) proved to be a more complex and less elegant solution when compared to the unconventional drilling technique.
The unconventional drilling technique known as Managed Pressure Drilling (MPD), enables the control of the annular pressure profile and monitors the bottom pressure limits in order to maintain them at a constant level, while minimizing drilling fluid loss and drilling downtime.
During the drilling of the well 17 of Cuervito Field the effect balloning was presented in large scale and was implemented a pilot project to evaluate the technique of MPD. The result was successful and was applied in the following three wells in the Field, achieving the elimination of the typical problems of the area.
Cuervito Field is part of the Burgos Basin in northeastern Mexico (see Figure 1). The main producing zones of Cuervito Field are the Formations known as Yegua that is composed of Cook Mountain sands (CM); and Queen City Formations which is composed of QC2, QC3, QC3A, QC6 and QC7 sands. These sands have a permeability of less than 0.01 mD (tight gas reservoirs), regular porosity (7% - 15%) and abnormal pressures between 5000 psi - 7200 psi. Hydraulic fracturing is required for the commercial production of gas and condensate in these sands.
The top down In-Situ Combustion (ISC), involves the stable propagation of combustion front from top vertical injector to the bottom horizontal producer. Besides laboratory studies in conventional sandstones, no application of the process in fractured carbonates has been addressed yet. In this paper a successful combustion tube experiment and history match of Iranian low permeable fractured heavy oil reservoir called Kuh-E-Mond, is presented and accompanied with details of experimental and simulation model. Validated model has been modified further to investigate the feasibility of Top-down ISC in fractured reservoirs mimicking block scale combustion cells. Effects of fractures geometrical properties such as orientation, location, extension, density, spacing, disconnection and dispersion have been considered. Investigation of aforementioned geometrical properties performed for the case of presence of networked fractures (presence of both vertical and horizontal fractures). Results confirmed higher outcome in the case of optimum vertical or horizontal fractures density and spacing. Laterally located vertical fractures enhanced the process in terms of ultimate oil recovery and sweep efficiency. Longer vertical fractures and higher degree of fractures dispersion through the reservoir improved the recoveries compared to the case of extended horizontal fractures and higher degree of horizontal fractures dispersion through reservoir. Depending on the reservoir parameters (such as fracture and matrix permeability) there is an optimum length of vertical fracture in which it enhances the recovery. This means very long extension of vertical fractures could cause early oxygen breakthrough and as a result lower sweep efficiency and oil recoveries. Simulation analysis revealed that Top-down In-Situ Combustion has higher feasibility for the reservoirs with highly networked fractures such as those occurring in the Persian Gulf coast.
Unconventional oil reservoirs, such as Bakken, have gained considerable interest in recent years because they have become a great resource to produce oil and gas to meet the energy needs of North America. Performance prediction from these tight reservoirs is a challenge because of the complexity of reservoir flow, well completion, and fracture stimulation techniques. Elm Coulee field, in Bakken, is an example of such unconventional reservoirs and is located in Richland County, Montana. The field was drilled using both vertical and horizontal wells, but in recent years the use of horizontal wells has become the standard practice. The objectives of this study were: (1) evaluate the long-term (7 to 10 years) performance of horizontal wells in Montana Elm Coulee, (2) develop a better understanding of how to predict the long-term performance of younger Bakken fields in North Dakota based on the Elm Coulee experience.
Arps hyperbolic decline curve analysis was used as the main forecasting approach. In Arps analysis, EQUATION where q is the flow rate, D is the decline rate, and b is the decline exponent. It will be demonstrated that forecasts using a constant b overpredicts well performance. To match the long-term performance of Elm Coulee wells, the numerical value of b had to be decreased with time.
Analytical approaches (log-log type-curve diagnostic plots and the Fetkovich log-log normalized plot) were also used to decipher the flow regimes, and to determine the varying decline rate from long-term producing wells in Elm Coulee Field. In addition to analytical modeling, numerical modeling was also used because it is more comprehensive in utilizing a larger set of reservoir parameters such as reservoir heterogeneity variations. This is very useful in transferring what we learned from the long-term performance of Elm Coulee Montana wells to the short-term performance of wells in North Dakota by addressing both geology and reservoir property differences between these fields.
The world energy demand is continuously increasing. Therefore heavy oil and bitumen reservoirs are driving more attentions for world energy supply. There are large amount of bitumen reserves in Canada. Only very small portion of bitumen reserves (~15%) in Alberta is mineable, and rest must be recovered using in-situ techniques. Bitumen viscosity can be reduced substantially by heating or dilution with a solvent. Steam Assisted Gravity drainage (SAGD) has been employed
commercially to recover bitumen from Athabasca oil sands. This method requires large amount of water, facilities for water treatment, and natural gas to generate steam. There have been attempts to develop and optimize hybrid SAGD processes to reduce water consumption during bitumen recovery by steam. The co-injection of steam and solvent additives (e.g. ESSAGD, SAP, SAS) can improve bitumen recovery due to its viscosity reduction by dilution with solvent and heating by steam. The experimental and pilot studies with steam and n-alkane co-injection shows enhanced oil recovery, and reduction in steam consumption (Nasr et. al., 1991, 2001, 2002, 2003). The phase behaviour of the bitumen and n-alkane solvents at SAGD operating condition is very complex and there is possibility of multiphase formation or asphaltene precipitation. The recent pore scale experimental study of this process has shown evidence of asphaltene precipitation during the ES-SAGD with n-alkanes (e.g. Pentane, Hexane) (Mohammadzadeh et. al., 2010). There are few published asphaltene precipitation data for Athabasca bitumen and n-alkanes at different temperature and pressure in the literature (Sabbag et. al. 2006). These data were used to develop an EoS model for the asphaltene precipitation using the CMG-WinProp Asph/Wax multiphase flash calculation. The asphaltene precipitation during the steam and n-alkane co-injection was studied using STARS thermal reservoir simulation model. This paper explains a method to characterize Athabasca bitumen based on the experimental
SimDist data. A technique for tuning the solid solubility model parameters was addressed to develop asphaltene precipitation model for n-heptane and Athabasca bitumen. Also the asphaltene precipitation modeling with STARS, its effect on the steam chamber development and ES-SAGD performance are discussed.
Brennan, Barny (Xcite Energy Resources Ltd) | Lucas-Clements, Charles (Xcite Energy Resources Ltd) | Kew, Stephen (Schlumberger) | Shumakov, Yakov Alexandrovich (Schlumberger) | Camilleri, Lawrence A.P. (Schlumberger) | Akuanyionwu, Obinna Chiemezie (Schlumberger) | Tunoglu, Ahmet (ADTI) | Hayhurst, Steve (ADTI) | Simpson, John
Due to increased hydrocarbon demand and technological advances, production from heavy oil fields in the United Kingdom Continental Shelf (UKCS) has become possible over the past 10 years. Despite substantial reserves in the UKCS with crudes less than 20° API, most of the activity has been confined to exploration and appraisal drilling. The main reason for the restricted activity has been the high uncertainty of the reservoir and fluid properties. Operational complexities inherent to heavy oil also limit the use of conventional appraisal-well testing technology.
A method was developed to determine the most suitable technology for testing wells with heavy oil using an electrical submersible pump (ESP). The solution was applied in the Bentley field located in the UK sector of the North Sea in block 9/3b, on which final appraisal well 9/3b-6Z was flow tested in December 2010.
The technical challenges included a short weather window, maintaining fluid mobility through the surface-testing equipment, oil and gas separation for metering, obtaining accurate flow measurements, and designing the most appropriate ESP system. A combination of technologies—dual-energy gamma ray venturi multiphase flowmeter, real-time monitoring, and a novel ESP completion—provided a solution that enabled successful well test execution. A multirate test reaching a final stabilized rate of 2900 bpd, with a subsequent period of pressure buildup was accomplished in less than 2.5 days with 10 to 12° API crude. A key lesson was how to obtain the quality of data that would enable reservoir engineers to extract with confidence a
productivity index and perform pressure transient analysis for reservoir characterization. This success paves the way for development drilling to commence on the Bentley field at the end of 2011, but also demonstrates potential that can enable new heavy oil field developments.
Taylor, Robert Stewart (Halliburton Energy Services Group) | Barree, Robert David (Barree & Assocs. LLC) | Aguilera, Roberto (University of Calgary) | Hotch, Ottmar (Halliburton) | Storozhenko, Ken K. (Halliburton Energy Services Group)
If one is repeatedly conducting the exact same completion in the same lithology of the same reservoir, then initial production (IP) rates might be a good indicator of relative long-term well performance or estimated ultimate recovery (EUR). This technique is tempting to use because it is quick and simple, and allows for easy comparison. However, use of this method alone assumes the shape of production decline curves will remain consistent from one well to the next. This method can especially be fallible when different numbers of fractures are placed along a lateral with possibly variable length. In this case, the relation between IP and EUR becomes much less defined.
Basing key economic decisions on IP alone can be misleading. This paper examines situations in which IP is not only a poor indicator of ultimate well performance, but in fact shows a reverse correlation. Sizing and spacing of fracturing treatments along a horizontal wellbore as well as vertical placement of the lateral in the zone can all be key variables. In many cases, one can choose high IP or optimized economic return, but not always both. Assuming that high IP equates to greater economic return can be a critical error. It is therefore essential to those involved in well-completion design as well as financial analysts to understand the variables involved as well as their impact. A lack of understanding could lead to poor completion and/or stimulation decisions that could severely impact the return on investment (ROI).
Observations and Discussion
The relationship between IP and longer-term profitability was investigated in two ways:
1. The first was to examine actual public domain production data and observe the relationship between IP and longer-term production as a function of horizontal wellbore placement within the zone. In this case, well economics are assumed to be closely related to cumulative production for each pair of offset wells because well costs should be comparable (similar horizontal lengths and a similar number of hydraulic fracturing stages).
2. The second approach was to examine the relationship from a theoretical perspective using a fractured well reservoir simulator in combination with an economic model to determine if higher IP always translates into improved economics.
Effect of Vertical Placement of Lateral Section on Production. Figs. 1A, 1B, and 2A, 2B illustrate two examples of comparative cumulative gas production versus lateral placement relative to the top of the zone for closely offset wells in NE British Columbia, Canada. Each pair of offset wells has similar horizontal lengths and number of frac treatments performed on them. Whether total well production or production per frac are examined, the observations are the same—the well with higher IP produces less cumulative gas (Table 1). For the figures in this discussion, the production per frac treatment was used. In both of these examples, it can be seen that the well with the lateral placed closer to the top of the zone has the higher IP. However, in each case, the cumulative production after less than four months is lower on the well with the higher IP and higher lateral placement in the zone. It is believed that this is attributed to gravity drainage of the water-based fracturing fluid.
Lateral wellbores placed near the top of the zone can more quickly produce natural gas as frac fluid drains below the wellbore. However, there is a less-effective mechanism for recovery of the water that drains into the lower portion of the zone. As a result, water remaining in the fracture and formation fracture face below the lateral wellbore inhibits inflow of gas. One mitigating effect could be that water closer to the bottom of the zone might be capable of imbibing into the reservoir because of the low gas-inflow rates and high capillary pressures. Over time, this would help remove some of the water from the lower portion of the fracture and fracture face, leading to some improvement in gas inflow and total production.
Unconventional gas wells often show linear flow during their transient period, and this transient behavior can last for several years. Currently industry uses typecurve matching technique to analyze this linear flow. The typecurves in common use are based on the assumption that wells produce at constant rate. However, most of the time, the production rate is variable, and in fact is closer to constant pressure operation. The constant pressure typecurve is useful but not suitable when both rate and pressure vary. It is necessary to have an easy-to-use method for analyzing variable rate/pressure data in linear flow.
There are two objectives for this paper. First, it serves to review the formulation, typecurve, specialized plots and superposition time that are used to analyze transient linear flow. It helps readers gain a deeper understanding of the theory. Second, it serves to illustrate a practical and effective way for analyzing variable gas production data.
In this development, we studied the effect of skin on the typecurve and on the specialized plot. We converted the constant pressure solution to its constant rate equivalent by using material balance time and found it to be acceptable for all practical purposes. We converted real time to corrected pseudo-time to account for variable gas properties, and found the effect to be small in the analysis of actual production data. We investigated the effect of outliers on superposition time. In the end, we proposed an approach to analyze variable rate/pressure data during transient linear flow and confirmed the validity of our methodology using synthetic data that generated in numerical models.