Nagoo, A. S. (Nagoo & Associates) | Kulkarni, P. M. (Equinor) | Arnold, C. (Escondido Resources) | Dunham, M. (Bravo Natural Resources) | Sosa, J. (Jones Energy) | Oyewole, P. O. (Proline Energy Resources)
In this seminal work, we reveal for the first time an extensively field-tested, demonstrably accurate and simple analytical equation for the calculation of the critical gas velocity limit (or onset of liquid flow reversal) in horizontal wells as an explicit and direct function of diameter, inclination and fluid properties. For the independently verifiable and first-of-its-kind multi-play field validation study, we carefully assimilate a very large database of actual horizontal gassy oil and gas liquid loading wells from several unconventional U.S. shale plays with different bubble point and dew point fluid systems and varying gas-to-liquid ratios and varying water cuts. The shale plays in our validation database include the Eagle Ford, Woodford, Cleveland Sands, Haynesville, Cotton Valley, Fayetteville, Marcellus and Barnett formations within their associated Western Gulf, South Texas, Arkoma, Western Anadarko, East Texas, Appalachian and Permian basins. Then, after summarizing our comprehensive field testing results, practical production optimization applications of the new analytical equation and advanced use cases of interest are further highlighted in various liquid loading prediction and prevention scenarios.
As opposed to prior critical gas velocity calculation methods (droplet reversal-based, film reversal-based, flow structure stability/energy), video observations both in the lab and the field clearly show continuously-evolving, co-existing and competing flow structures even with simple fluids without mass exchanges. Therefore, this work avoids skewed assumptions on demarcating the prevailing or dominant flow structure. Instead, the new analytical equation developed is based on an analysis of the major forces in the flow field, namely the axial buoyancy vector, the convective inertial and the interfacial tension forces, in combination with an assumption of the onset of liquid flow reversal based on flow field bridging (Taylor instability). Since the new analytical equation was formulated using these minimalist assumptions, this unique characteristic results in the highest predictability obtainable for the critical gas velocity calculation because there is the least amount of uncertainties (fudge factors). The consistent accuracy of the equation against our extensive horizontal well liquids loading database verifies this fact. Moreover, the simplicity of form of the equation makes it easy to use in that every practicing engineer in practice can perform fast hand or spreadsheet calculations. In effect, this equates to having a model as simple as the Turner model but now with additional direct functions of diameter and inclination. Also, the results clearly invalidate the need for artificial variables (such as interfacial friction factor) that cannot be directly measured in any experiment. In terms of usage, the new model is used in liquid loading prevention scenarios such as end-of-tubing (EOT) landing optimization and tubing-casing selection. Evidently, this work proves that no complex, computer-only procedure is necessary for accurate critical gas velocity calculation. This finding has significant speed and improved answer-reliability implications in strong favor of the presented simple equation for use in artificial lift, production optimization and digital oilfield software in industry, in addition to being ideally suited for ‘physics-guided data analytics’ applications in real-time production operations environments.
Temizel, Cenk (Halliburton) | Energy, Aera (Halliburton) | Betancourt, Dayanara (Halliburton) | Aktas, Sinem (Turkish Petroleum) | Susuz, Onur (Turkish Petroleum) | Zhu, Ying (University of Southern California) | Suhag, Anuj (University of Southern California) | Ranjith, Rahul (University of Southern California) | Wang, Aaron (Halliburton)
Hydraulic fracturing is a very robust tool in terms of not only increasing production in tight reservoirs but also accelerating production from reservoirs with higher permeability and porosity. The success of a hydraulic fracturing treatment is highly correlated with control of the created fracture geometry. Control of fracture growth and geometry is challenging in formations where the boundary lithologies are not highly stressed compared to the pay zone, allowing out-of-zone migration of fractures. As several factors influence the growth and geometry of fractures, including the reservoir, wellbore, and fluid/proppant parameters, it requires a good understanding of reservoir parameters, including stress distribution along with appropriate use of corresponding wellbore components and fluid/proppant for successful and efficient results. Our objective is to outline the main differences between different fracture models along with the key parameters and their significance in fracture performance. Fracture treatment designs involve selecting fracturing fluids, additives, proppant materials, injection rate, pump schedule, and fracture dimensions. Although hydraulic fracturing has become more important due to development of unconventional resources in tight formations, the use of fracture models before implementation of the treatment is limited, leading to undesired fractures either with limited growth or propagation out of zone. Literature lacks a study that combines the evaluation of different fracture models in an optimization process where economics is taken into account with an objective function maximizing the net present value (NPV), providing detailed information on the financial side of this technical phenomenon, too.
While marine organic-rich mudstones (aka black shales) have been effectively described in recent years, developing depositional models has lagged. Without depositional models, predictability of facies and properties remains a major problem.
Sequence stratigraphy provides an answer. Sea level change controls sedimentation and circulation. Failure of masking sedimentation determines where marine black shales are expressed, and explains why they preferentially occur in carbonates and in the Paleozoic. In basinal organic-rich mudstones, which lack subaerial exposure surfaces, sequences can be identified by recognition of systematic variation in the rate of deposition. Episodicity and nondeposition are important considerations; and several different environments may be expressed within a black shale. But methods beyond simple observation of “unconformities” are necessary. Several parameters directly reflect rate of deposition, and together can be a powerful indicator of the depositional framework. Each comes from a different aspect of reduced sedimentation. The abundance of phosphatic fossil debris is a function of dilution by sedimentation. Illite crystallininty is a function of the length of time it is exposed to bottom water, regardless of oxidizing conditions. The relative abundances of organic matter type is a function of the length of time exposed to oxidizing conditions and the reciprocal rate of burial in reducing conditions. Other lines of evidence may also contribute to the model. While individually they may be ambiguous, the ability to correlate different signals from different processes reinforces the interpretation. The depositional model is testable against sequence models and against sequences recognized on adjacent shelves, constraining the intensity and frequency patterns of the sequences identified in the basin.
A sequence-based depositional model can help to identify lateral and vertical changes in rock properties within the basinal shales, particularly as they apply to distribution of organic matter type and content (which determine “sweet spots”), porosity, cements and bedding properties. Both actualistic and probabilistic models may be developed and may be helpful with risk analysis. While detailed analysis of every well is impractical, the application of models derived from key sections can greatly enhance predictability.
Quan, Tracy M. (Boone Pickens School of Geology, Oklahoma State University) | Puckette, James (Boone Pickens School of Geology, Oklahoma State University) | Rivera, Keith (Boone Pickens School of Geology, Oklahoma State University) | Otto, Brice (Boone Pickens School of Geology, Oklahoma State University) | Adigwe, Ekenemolise (Boone Pickens School of Geology, Oklahoma State University)
Assessment of unconventional hydrocarbon resources relies in part on characterizing the origin of the reservoir.
This includes determination of the original depositional conditions of the shale unit and any subsequent modifications to the organic matter present. Our research has found that sedimentary nitrogen isotope (d15N) measurements can be reliable proxies for determining water column redox state during deposition, as well as for compartmentalization and post-depositional fluid migration during catagenesis.
This paper summarizes our investigations into the use of d15N measurements as proxies to evaluate depositional redox conditions and fluid migration pathways, and suggests applications for these measurements in characterizing unconventional resource plays. Case studies were performed using d15N measurements from cores obtained from the Devonian-aged Woodford Shale and Caney Shale in Oklahoma, and Ohio Shale in Kentucky, and this data was combined with other geochemical and lithological measurements, including trace metals, thermal maturity, and gamma ray logs. Our data shows that bulk sedimentary d15N values primarily reflect water column redox conditions during deposition, as supported by correlation with redox-sensitive trace metals and other redox proxies.
Comparison of d15Nbulk values on a basin-wide scale indicates that interpretation of d15N as a redox proxy is consistent despite differences in the thermal maturity across the basin. Separation of d15Nbulk into inorganic and organic nitrogen isotope fractions appears to provide information about fluid migration pathways during catagenesis, as well as data regarding compartmentalization within an unconventional resource unit. As a result, measurement and interpretation of d15N values as part of a multi-proxy geochemical analysis can provide important details regarding the depositional and catagenic history of an unconventional resource play that may be essential to assessing the reservoir.
Cozyris, K.M. (Baker Hughes) | Churcher, P.L. (Lighthouse Oil & Gas Canada Management Inc.) | Piland, J.R (Lighthouse Oil & Gas LP) | Maharidge, R.L. (SPE) | Adamson, M.D. (SPE) | Lew, R. (Baker Hughes)
The effective optimization of fracture stimulation treatments in horizontal wells requires the integration of a wide range of engineering and geological data to be successful. This process begins with a thorough understanding of not only the reservoir rock properties, but also the properties of the confining nonreservoir rock. This geological information is used in the design of non-damaging completions fluids, to predict the fracture height growth and fracture half length, and to interpret the results from diagnostics, post job stimulation data and production performance. This paper documents the work conducted to design and implement multistage horizontal fracture stimulations in the Cleveland and Tonkawa sandstones located in Dewey County, Oklahoma. Methods used in this process included: 1) review of historical treatment and production data available in the area to help identify current best practices, 2) SEM, thin section and X-ray diffraction petrography to define the mineralogy, 3) core floods to determine formation damage mechanisms, 4) laboratory proppant pack floods to screen for the effectiveness of chemical additives (such as surfactants), 5) core floods using reservoir rock to determine regain permeability and flow back performance, 6) determination of mechanical rock properties, such as Young's modulus, Poisson's ratio, Brinell hardness and triaxial stress, from multiwave sonic and laboratory testing for use in fine-tuning the fracture stimulation model parameters, 7) design, acquisition and analysis of initial injection and falloff tests, 8) fracture stimulation modeling to predict the fracture geometry created by the fracture design, and 9) analysis of hourly flow back and tracer data to determine the effectiveness of the treatments in accessing the maximum amount of reservoir rock. The objectives of this work were to engineer an optimized treatment design (that would result in significant gains in initial well productivity and long term ultimate hydrocarbon recovery) and also to develop and refine new potential best practices.
Most of the shale reservoirs in US land are naturally fractured. The fracture intensity and types vary from one shale to another. Even within the same shale in the same field, the heterogeneity of fracture intensity can be often expected to be high in a horizontal well. The current popular geometrical completion design can potentially ignore the existence of natural fractures. Hence, maximizing stimulation efficiency without understanding existing natural fractures can be a challenge. In this paper, study was made of the majority of the published case studies related to natural fractures in the US shales in the last 20 years. The evidence of natural fractures from either outcrops or subsurface data, i.e. core, borehole images, or other data is summarized for each studied shale. The latest studies show that the hydraulic fracture propagation can be strongly influenced by existing natural fractures regardless of whether they are open or closed. The roles of existing fractures in the shales include: 1) providing enhanced reservoir permeability for improved productivity if they are open and effectively connected by hydraulic fractures; 2) promoting much better fracturing network complexity regardless of whether they are open or closed prior to the stimulation; 3) giving possible negative impact sometimes, i.e. high water cut, if they are connected with wet zones below or above the reservoirs. It can be concluded that engineered completion designs that employ accurate knowledge of natural fracture data, in-situ stresses, and other reservoir and completion quality indicators as inputs can provide opportunities for enhancing stimulation efficiency and fracturing network complexity. This in turn can lead to better connectivity to a larger reservoir volume and access to more drainage area in the shales.
The US shale gas story actually featured natural fractures. William Hart, a local gunsmith, drilled the first commercial natural shale gas well in US in Fredonia, Chautauqua County, NY in 1821, in shallow, low-pressure rock with fractures . The well was first dug to a depth of 27ft in a shale which outcropped in the area, then later drilled to a depth of 70ft using 1.5 inch diameter borehole. The produced gas was piped to an innkeeper on a stagecoach route. Then the well was produced without any stimulation for 37 years until 1858 when it supplied enough natural gas for a grist mill and for lighting in four shops. It was a combination of the idea from Mr. Hart to drill the well and the presence of the natural fractures in the gas shale that made the 1st commercial shale gas discovery possible in shale gas history.
EQT Production has been an active driller in the Lower Huron Shale in the Appalachian basin for nearly 100 years and has close to 5000 producing vertical wells in Kentucky alone. In 2006, EQT began drilling horizontal wells and has drilled over 400 Lower Huron Shale and over 100 Cleveland Shale wells in Kentucky through mid-2010. EQT has experimented with various completion designs and fracture methods in order to maximize production from these wells. EQT traditionally has performed both foamed nitrogen and nitrogen gas hydraulic fracture treatments, with the selection between the two dependent on reservoir pressure data taken from vertical offset wells drilled and completed in past decades. EQT also performed 33 hybrid (ultra-high quality foam) treatments on Lower Huron wells and compared their production against the production from their offsets. EQT studied the effect that frac stage length has on well production. The method in which the open-hole packers were set was also analyzed and found to be influential in well performance. From these studies, EQT has identified ways to improve well performance for a minimal cost by altering the completion design and has implemented these design changes in their completion programs.
Wilhide, Scott (EQT Production) | Smith, Jeremy (EQT Production) | Doebereiner, Daniel (EQT Production) | Raymond, Bradford (Weatherford) | Weisbeck, Denis H. (Weatherford) | Ziemke, Brian (Halliburton)
This paper presents the first use of a Rotary Steerable System (RSS) using air as the drilling fluid. The case history is in an ongoing unconventional gas development in the Appalachian Basin of the northeast United States. The RSS has been integral to increasing lateral lengths and corresponding increases in production while reducing development costs. Since 2006, EQT Production has developed Devonian reservoirs in Kentucky, Virginia and West Virginia using underbalanced, horizontal drilling techniques. The low (200-500 psi) bottom hole reservoir pressure does not allow drilling with fluid. Typical horizontal wells have had 3,000 ft laterals drilled with air, positive displacement motors, and electromagnetic telemetry (EM) MWD systems; that remains the predominate drilling technique. The air compatible RSS was employed to increase lateral length and production.
Well 568478, in Letcher County, Kentucky became the first well to have a horizontal section drilled underbalanced using RSS with dry air. To date, eight wells have been drilled with lateral lengths from 3,800 -6,000 ft. Preliminary results show that the reserves developed are proportional to the lateral length drilled. The rates of penetration (ROP) with the RSS in the additional footage were similar to or greater than those in shorter laterals drilled with a motor.
The RSS, with an integral EM MWD, has proven to be a technically feasible option in air drilling environments. The comparable ROP and resulting lower lateral costs per foot achieved with the RSS allows drilling horizontal wells with air beyond their previous limits, enabling greater production footage from fewer wellbores. It further allows drilling portions of the reservoir previously unreachable due to surface constraints such as topography.
Many of the learnings to implement the RSS technology also applied to the use of the positive displacement motors (PDMs) previously used. These learnings were applied to the motors and significant improvements were made in drilling horizontals with these tools. Motor capability was pushed beyond limits previously established in the areas being drilled.
As a result of the success with RSS and PDMs on extended laterals, EQT Production is extending the lateral length from 3,000 ft to 5,500 -7,000 ft on a plurality of the future wells drilled in the Devonian section.
Schepers, Karine Chrystel (Advanced Resources International) | Nuttall, Brandon C. (Advanced Resources International) | Oudinot, Anne Yvonne (Advanced Resources International) | Gonzalez, Reinaldo Jose
Shale gas and other unconventional gas plays have become an important factor in the United States energy market and are often referred to as statistical plays due to their high heterogeneity. They present real engineering challenges for characterization and exploitation, and their productivity depends upon an inter-related set of reservoir, completion and production characteristics.
The Devonian Ohio shale of eastern Kentucky is the State's most prolific gas producer. The gas shale underlies approximately two-thirds of the state, cropping out around the Bluegrass Region of central Kentucky and having a sub crop beneath the Mississippi Embayment in western Kentucky.
This paper describes the reservoir modeling and history matching of a Devonian Gas Shale Play, eastern Kentucky, its potential for CO2 enhanced gas recovery and storage.
A geologic model of the shale has been compiled from mineralogical, petrographic, core, production, and wireline data. The COMET3 multi-phase, dual porosity simulator is being used to investigate CO2 injection into the shale for enhanced gas recovery. To accomplish this, a subset of wells surrounding the potential injection site has been selected for further study. These eight wells cover approximately 5,300 acres of productive shale. The reservoir was subdivided into the Upper Ohio and Lower Huron members. To capture geological heterogeneity, gas production rates for these wells served as a proxy to characterize fracture permeability using geostatistical methods. Well production was history matched applying an automated process. Finally, several CO2 injection scenarios spanning huff-n-puff to continuous injection were reviewed to evaluate the enhanced gas recovery potential and assess the CO2 storage capacity of these shale reservoirs.
Increased emissions of carbon dioxide (CO2) are being linked to global climate change and are generating considerable public concern. This concern is driving initiatives to develop carbon management technologies, including the geologic sequestration of CO2. One option for sequestration may be the Appalachian Basin's Devonian black shale, a continuous, low-permeability, fissile, fractured, organic-rich rock that is both the source and trap for natural gas (primarily methane). In gas shales, natural gas occurs as free gas in the fracture system and is adsorbed on clay and kerogen surfaces, very similar to the way methane is stored within coal beds.
In general, successful applications of horizontal wells have been limited to high-permeability reservoirs and unconventional formations such as coal, chalk, and shale. Conversely, few tight-gas-sandstone reservoirs that require stimulation have realized sustained success with horizontal completions. One example of such success is the Cleveland Sand of north Texas and the Oklahoma Panhandle. Very recently, some success with horizontals has been observed in the Bossier and Cotton Valley Sands of East Texas and north Louisiana. Horizontal wells are commonly two to four times more expensive to drill and complete than offset vertical wells, yet they are theoretically capable of up to three to five times the production. Higher gas prices have lead to potentially better economics for horizontal wells (Mulder et al. 1992). However, research shows that in practice, many of these wells typically produce only 10 to 30% more than offset vertical wells. With costs more than double those of vertical wells, the economics is obviously unfavorable.
This paper discusses ways to identify and manage risks when planning, drilling, and completing horizontal wells in tight-sandstone formations to improve success.
Evidence has shown that shortcuts and blanket approaches do not work usually in these completion environments. A multitude of lithological and depletion possibilities exist as risks that need to be identified and managed through appropriate application of integrated drilling and completion technologies. Each risk may require different drilling and completion considerations in order to succeed. There is simply no recipe for repeat success.
A detailed method is presented to identify, understand, and manage risk associated with horizontal wells drilled in tight-gas-sandstone reservoirs. The method will address all of the complex subjects that need to be considered for the successful placement and completion of a horizontal well, including reservoir description (both static and dynamic), well design, drilling, stimulation, and production. It will also illustrate consequences of what may happen if these issues are not considered properly. Through this method, horizontal-well feasibility and economic results can be determined. If a horizontal well has been determined to be viable economically, this method can consistently provide a solution as to what the best completion type (vertical or horizontal) is to recover reserves and enhance recovery efficiency in tight-gas-sandstone reservoirs.