Marcellus Gas Shale Project RPSEA Project 9122-04 GTI Project 211000 Technology Status Assessment February 15, 20111 Background The Marcellus Shale, stretching across Pennsylvania, Westt Virginia, Ohio, Maryland and New York, is a low permeability hydrocarbon bearing shale formation off Devoniann age. Hydrocarbon production is mainly comprisedd of gas however gas condensate does occur in some parts off the play. Main exploitation techniques utilize horizontall drilling coupled with a completion program that includess multiple hydraulic stimulation treatments. Securing waterr for hydraulic fracturing and disposal of flowback andd produced waters can be difficult to achieve. Because of thee topography and terrain, often times it is preferred to drilll multiple wells from a single well pad. This type of welll layout reduces the environmental footprint by reducing thee number of well pads and eliminating the need for additionall roads to multiple well locations. Key geologic factors influencing production from the Marcellus shale are thermal maturity of the rock, net thickness, porosity, permeability (dual permeability system), naturall fracture network system, and mineralogy as it influences the brittleness of the rock therebyy impacting the hydraulic fracturing process. The type, extent, and nature of natural fractures play significant roles in production from the Marcellus shale and as such, development of a technique for determination of the state of the fracturing is one of the major foci of this project. Similar to other shale formations, hydraulic fracturing iss an imperative for establishing economic production from Marcellus. Characterization of hydraulic fractures in the presence of natural fractures and determination of the stimulated reservoir volume is also a major objective of the project. In summary, the Marcellus Gas Shale project is a fully integrated field based project that includes geologic, reservoir engineering, and completion studies with the objective of developing an understanding of the completion procedures as influenced by shale properties. The project includes detailed geological, reservoir engineering, and hydraulic fracturing studies. A brief discussion of the state of technology in each study area follows. Geology The matrix permeability of shale formations is far too low to allow for any substantial production and high rate gas flow must be from a network of natural fractures in communication with high conductivity 1
hydraulic fractures. As such, characterization of natural fractures and their interactions with hydraulic fractures has been the subject of many studies. The state-of-the-art techniques include the use of image logs and core information and extrapolation of the information beyond the wellbore. Data from surface outcrops and downward extrapolation of the surface knowledge to reservoir depths are also common. However, all such extrapolations; be it downward from surface or outward from the wellbore to the reservoir, remain somewhat speculative. A major objective of the present project is the development of a predicting tool based on fracturing properties of the Marcellus shale lending some reliability to such extrapolations. Specifically, we intend to develop a technique for predicting the state of natural fractures and their clustering using a methodology that correlates the subcritical indexes with clustering tendencies of natural fractures. A second objective of the project is characterization of interaction between hydraulic fractures and natural fractures. In particular, the fracturing mode of sealed natural fractures when subject to hydraulic fracturing shall be studies in detail. Bauru of Economic Geology (BEG) will be in charge of these investigations. Characterization of pore spaces in shale is a complementary task in the area of geological investigations. In these efforts, the Lawrence Berkeley Laboratories shall use Focused Ion Beams (FIB) for nano-scale imaging of pores for better understanding of the pore space. Results from these studies have the promise of shedding some light on the diffusion process in shale. Completion Strategy Completion of wells in the Marcellus typically involves multistage fracturing in the horizontal section of the wellbore. Stage sequencing is typically performed with pumped down guns and bridge plugs, where the bridge plug is set after a successful frac stage in order to isolate open perforations from the rest of the wellbore, and then the guns are fired in order to perforate the casing for the next frac stage. This type of completion provides known fracture initiation points and enables a more controlled re-fracturing operation in future re-completions. Another option available, but less common, is the openhole completion utilizing openhole packers and ball activated shifting sleeves, that when shift, open up frac ports connecting the formation to the wellbore. This type of completion can greatly reduce fracturing time by not having to shut down in between frac stages. However, the internal diameter (ID) of the wellbore is incrementally reduced as a result of the seats that the balls land on and if any future intervention work is needed, aside from other constraints, the baffles will have to be drilled out in order to regain full wellbore ID. Optimum spacing in between frac stages is typically found through trial and error and confirmed by microseismic fracture mapping and production data. Vertical wells are typically completed with cemented casing and fracture stimulated with multiple frac stages, depending on number of pay intervals. Stage sequencing is achieved with bridge plugs and guns that are conveyed on wireline. Vertical wells are much less common than horizontal wells due to economics and increased environmental footprint. Hydraulic Fracturing Typical stimulation treatment in a dry gas horizontal well is comprised of 6-8 hydraulic fracture stages, with 400 thousand gallons (kgal) to 500 kgal of fluid per stage, typically being slickwater. Each stage also includes on average 400 thousand pounds (klb) to 600 klb of proppant with roughly one third to one half of total weight being 30/50 mesh white sand, and the remainder being 100 mesh silica. Typical treatment rates are between 50 to 100 barrels per minute (bpm). The 100 mesh silica is typically ramped to 3 pounds per gallon (ppg) and is pumped before the larger mesh proppant. The 30/50 mesh proppant is pumped last and is typically ramped to 4 ppg. In wells with considerable amount of gas condensate or 2
other heavier hydrocarbons, it is advantageous to pump more of the 30/50 mesh proppant or an even larger mesh proppant altogether. If larger mesh proppant is used, then higher viscosity fluids may be required to generate ample fracture width for proppant placement and enough viscosity for proppant transport if flow rate is low. Treatments that utilize both friction reduced water and highly viscous crosslinked gels, are called hybrid fracs and are gaining more acceptance in areas that produce wet gas. Recent studies have tried to determine if fracs with smaller proppant volumes (600 klb per well) or, Lite Fracs, can achieve the same gas production as large proppant volume fracs (4400 Klbs per well) (SPE- 131783). In these tests the total amount of frac fluid pumped in the Lite Fracs was about the same as the normal fracs. This was achieved by greatly reducing the proppant concentration throughout the treatment. The results indicate that much production is lost when small proppant volumes are pumped as opposed to large proppant volumes. The average estimated ultimate recovery (EUR) for the wells stimulated with Lite Fracs is 43% less than the average EUR of wells stimulated with the large volume of proppant. History matching of production has shown the effective fracture half length of the Lite Frac to be 1/10 of the large volume fracs. Unlike the Barnett shale, the Marcellus hydraulic fracturing is still in the experimental stage and it will take some time before the most efficient approach is identified. The present project aims at evaluation of two rather new approaches in fracturing of the Marcellus shale. These are Zipper-Frac and Simul-Frac approaches where two or more nearby wells are fracture stimulated alternately (Zipper-Frac) or simultaneously (Simul-Frac). These fracturing strategies are devised on the premise that temporal changes in the state of stress during hydraulic fracturing may be such that fractures from different stages attain different azimuthal directions thereby resulting in a larger stimulated reservoir volume. Fracture Imaging Microseismic fracture imagining is a common method of fracture event mapping in the Marcellus. Currently there are two methods of receiver array implementation, surface and downhole. In addition to microseismic imaging, tilt meter survey is also available however it is not very common in the Marcellus and does not provide the level of detail that microseismic imaging does. If downhole microseismic is utilized, a monitor well is required, in which case it could either be a dedicated observation well or an offset producing well that is shut-in. If a producing well is used as the observation well then lost production can have a significant financial impact. Moreover, if a dedicated observation well is drilled then it also carries a large financial burden, however, valuable reservoir information is typically gathered either through cores or logs from the pilot well. Because of environmental considerations, many wells are drilled from a single pad and extend out as to cover the greatest amount of reservoir. Surface microseismic arrays are gaining in popularity because they do not require an observation well and a single surface radial array can be used to monitor fracture treatments on multiple wells such as the ones depicted in Figure 2. Although an observation well is not needed for a surface array, a wide area of land has to be available for setting up the array, in which case it may not always be available. Additional surveying and permitting has to be performed prior to setting up the surface array and in some cases a large environmental footprint can be left if clearing of obstructions is required. Signal resolution of the surface array is reduced as compared to downhole arrays, as a result of attenuation of higher frequencies due a longer signal travel path. 3
Figure 2: Example of a typical multiple well arrangement extending from a single pad on surface and a radial surface microseismic array covering all the wells. In this project, a number of horizontal wells in a pattern similar to that shown on Figure 2 will be fracture stimulated and concurrent surface and downhole seismic imaging will be carried out. Attempts shall be made to evaluate the accuracy of event locations acquired by downhole and surface microseismic monitoring. Furthermore, a vertical seismic profile will be surveyed to create an accurate velocity structure needed for high quality processing of the surfacee data. In addition, a speculated change in velocity model due to hydraulic fracture creation will be evaluated by the use of check shots in between fracture stages. Reservoir Engineering Aside from conventional reservoir engineering techniques previously developed that combine surface 3-D seismic imaging and reservoir characteristics derived from core and log data, a new and novel approach to reservoir characterization has been developed by Dr. Shahab Mohaghegh. The new novel techniquee takes the top-down approach as opposed to a bottom-up approachh currently utilized in most of the industry leading reservoir engineering software such as Petrel. Inn a conventional reservoir simulator, large amounts of data are needed in order to accurately model the reservoir and many numerical simulation runs are required to complete the analysis. The top-downn approach uses preexisting formation and production data from at least 25 wells to build a reservoir model. Fuzzy logic then recognizes patterns and predicts the outcome of either an infill well or and offset well. Results are verified by removing a single well with known properties from the dataset and predicting its behavior based on the remaining wells in the dataset. This method aids in the development of in-fill strategies by determining the location of sweet spots for remaining reserves. Formation Evaluation Current technology enables characterization of fracture populations in shale reservoirs through the evaluation of core and wireline log data. In the case of shale reservoirs, desorption of natural gas from the organic material has significant contribution to gas productionn and log data are calibrated with laboratory desorption tests to develop what is known as shale gas formation evaluation model. In addition, estimates of total organic conten and total adsorbed gas are used for density, porosity and saturation corrections. The ongoing research in the Marcellus Shale will focus on evaluating technologies and techniques that will allow for identification of highest productivity intervalss in the reservoir and enable delineation of fracture species through integrationn of core, electric log, andd image measurements. Advanced logs such as spectral gamma and magnetic resonance imaging enable characterization of formation properties with 4
very little correlation to core data, thus reducing coring and core analysis costs. Nonetheless, core analysis is still critical for identifying natural and induced fracture species and for developing advanced methodology for calculation of shale porosity, water and hydrocarbon saturations, free and adsorbed gas saturations, matrix permeability, and other properties that can be correlated to electric log data. Summary This project relies heavily on high quality field data and aims at reliable characterization of the Marcellus shale as a whole. It will also evaluate the applicability and accuracy of current technologies focusing on hydraulic fracturing and microseismic imaging. By better understanding the composition of the Marcellus and the mechanisms that drive sustainable long term production, ultimate gas recovery from this shale can be significantly improved. The final products resulting from this project are methods and techniques for optimal drilling and completion strategies along with data that enables a deeper understanding of the play in terms of composition and producibility. 5