It has often been reported that the peak production of a well drilled in tight formations is highly dependent on the fracture contact area. However, there is no efficient approach to estimate the fracture surface area at present. In this paper, we propose a method to calculate the fracture surface area based on the falloff data after each stage of the main hydraulic fracture treatment.
The created hydraulic fracture closes freely before its surfaces hit on the proppant pack, and this process can be recognized on the pressure falloff data and its diagnostic plots. The pressure decline rate during fracture closure is mainly caused by fluid leakoff from the fracture system into the formation matrix. For a horizontal well drilled in the same formation, we may assume the same leakoff coefficient among all stages, so the total fracture surface area can be calculated for all stages to meet the requirement of the fluid leakoff rate.
Wellbore storage effect, friction dissipation and tip extension dominate the early pressure falloff data. While the transient dominated by friction losses typically lasts about one minute, tip extension may end after about 15 minutes. Therefore, falloff data should be acquired for at least 30 minutes to observe a fracture closure trend. The fracture closure behavior can be identified on the G-function plot as an extrapolated straight line or on the Bourdet derivative in log-log plot as a late time unit slope. The behavior of the late unit slope depends on the pressure decline rate, or correspondingly, to the fluid leakoff rate. Therefore, the total fracture surface area can be estimated using hydraulic fracture design input values for formation leakoff coefficient and fracture closure stress. The calculated fracture surface area represents the combined area of primary and secondary fractures, effectively all fracture surfaces contributing to the fluid leakoff.
We applied the approach to all stages in a horizontal well that exhibit the fracture closure behavior. The approach shows promise as a straightforward way to estimate fracture surface areas that could, enable, in turn, an early estimate for the expected well performance.
Yang, Ruoyu (Southwest Petroleum University) | Guo, Jianchun (Southwest Petroleum University) | Zhang, Tao (Southwest Petroleum University) | Zhang, Xudong (Southwest Petroleum University) | Ma, Jian (Sinopec) | Li, Yang (Southwest Petroleum University)
Slick-water fracturing treatment is one of the most effective method to develop shale reservoir, which creates complex fracture system by connecting the pre-existing natural fractures. However, the proppant transport and placement behavior is quite different from that in conventional bi-wing fractures due to the low viscosity fluid system and intersections between fractures. The goal of this work is to simulate and understand the characteristic of proppant transport behavior in Complex Fractures network.
A Eulerian multiphase model is introduced to simulate the transport and settling behavior in the hydraulic fracture network, which takes turbulence effects and friction stress between the proppant particles into consideration and fully couple the fluid phase with particle phase. Simulation work was conducted to investigate the control mechanism and influencing factors for proppant transportation from main fracture into secondary and tertiary fractures.
The simulation results indicate that a small proppant dune quickly forms in the main fractures first, and almost no proppant enters the lower grade fracture until the proppant dune in the intersection reaches a specific height. With continuous injection of slurry fluid, majority of the proppant enters in the lower grade fracture which is controlled by gravity rolling from the dune in main fractures and fluid drag force, and the proppant settles quickly and gradually reach their own equilibrium height. Parametric study shows that smaller proppant density and particle size can also help proppant transport into secondary fractures and form a higher equilibrium height dune, resulting in larger effective propped area. Moreover, when the lower grade fracture is closer to the inlet entrance, the proppant is more likely to transport in, and the height of sand dunes formed in the fractures is higher.
The proppant transport process in complex fracture systems is simulated by Eulerian Multiphase Model in this paper. This study extends the understanding of the process and mechanism of proppant transport in complex fracture system and controlling factors, which helps optimize hydraulic fracturing design in shale formation.
A Coal Seam Gas Field, operated by Origin Energy as upstream operator of the APLNG project in Queensland, Australia has been on commercial production since 2005. From inception to 2012 the development concept used vertical cavitated wells or vertical fracture stimulated wells. Four pilot surface to inseam well pairs were successfully drilled and commissioned in 2012 and 2013. Based on the performance of these pilots, further development of the coal seam gas field converted from stimulated vertical wells to horizontal wells from 2015 onwards. Surface-in-seam (SIS), high-angle-sump-horizontals (HASH), and multilaterals wells drilled with coil tubing technologies were implemented in the field. The paper describes horizontal wells drilling, completion and geo-steering technology. The optimal operational strategies of horizontal wells are also described and the production performance of horizontal wells are compared to the vertical well performance. The key factors contributing to the success of horizontal drilling in the field are presented.
Sun, Wenliang (Ocean University of China, Qingdao) | Zou, Zhihui (Ocean University of China, Qingdao / Qingdao National Laboratory for Marine Science and Technology) | Zhou, Hua-Wei (Ocean University of China, Qingdao / Qingdao National Laboratory for Marine Science and Technology / University of Houston) | Rui, Yongjun (Sinopec) | Zhao, Shengtian (Sinopec) | Cui, Qinghui (Sinopec) | Zhang, Jianzhong (Ocean University of China, Qingdao / Qingdao National Laboratory for Marine Science and Technology)
Seismic tomography is an important method of subsurface velocity model building. However, it is challenging to determine the geometry of velocity layers in complex areas using only the first-arrivals. We propose a joint multiscale deformable layer tomography (MDLT) to use the first arrivals and the reflection traveltimes simultaneously in the velocity inversion. The results of synthetic tests show that the joint MDLT can accurately recover some layered velocity models, with higher stability and accuracy than the first-arrival MDLT.
Zhang, Ruyi (School of Ocean and Earth Science, Tongji University) | Liu, Haojie (Shengli Oilfield Geophysical Research Institute, SINOPEC) | Wang, Huazhong (Shengli Oilfield Geophysical Research Institute, SINOPEC) | Yang, Hongwei (School of Ocean and Earth Science, Tongji University)
Summary Based on the characteristic equation of pre-stack data frequency variation response, according to the relationship between elastic impedance and reflection coefficient, frequency-dependent viscoelastic fluid factor can be derived, so does the relationship of elastic impedance with frequency and incidence angle. Pre-stack data contain more information of amplitude and frequency. Using the frequency-dependent viscoelastic impedance equation and bayesian inversion framework, the objective function of frequency-dependent elastic impedance inversion can be established to realize the frequency-dependent impedance inversion at different angles. According to the elastic impedance equation of frequency varying viscoelastic fluid factor, the relationship between elastic impedance and frequency-dependent viscoelastic fluid factor is established, and the pre-stack seismic inversion method of frequencydependent viscoelastic fluid factor is studied. Combined with rock physics, fluid reservoir identification based on frequency-dependent viscoelastic fluid factor can be realized.
Yan, Xia (China University of Petroleum (East China)) | Huang, ZhaoQin (Heriot-Watt University) | Yao, Jun (China University of Petroleum (East China)) | Li, Yang (China University of Petroleum (East China)) | Fan, Dongyan (Sinopec) | Sun, Hai (China University of Petroleum (East China)) | Zhang, Kai (China University of Petroleum (East China))
After hydraulic fracturing, a shale reservoir usually has multiscale fractures and becomes more stress-sensitive. In this work, an adaptive hybrid model is proposed to simulate hydromechanical coupling processes in such fractured-shale reservoirs during the production period (i.e., the hydraulic-fracturing process is not considered and cannot be simulated). In our hybrid model, the single-porosity model is applied in the region outside the stimulated reservoir volume (SRV), and the matrix and natural/induced fractures in the SRV region are modeled using a double-porosity model that can accurately simulate the matrix/fracture fluid exchange during the entire transient period. Meanwhile, the fluid flow in hydraulic fractures is modeled explicitly with the embedded-discrete-fracture model (EDFM), and a stabilized extended-finite-element-method (XFEM) formulation using the polynomial-pressure-projection (PPP) technique is applied to simulate mechanical processes. The developed stabilized XFEM formulation can avoid the displacement oscillation on hydraulic-fracture interfaces. Then a modified fixed-stress sequential-implicit method is applied to solve the hybrid model, in which mixed-space discretization [i.e., finite-volume method (FVM) for flow process and stabilized XFEM for geomechanics] is used. The robustness of the proposed model is demonstrated through several numerical examples. In conclusion, several key factors for gas exploitation are investigated, such as adsorption, Klinkenberg effect, capillary pressure, and fracture deformation. In this study, all the numerical examples are 2D, and the gravity effect is neglected in these simulations. In addition, we assume there is no oil phase in the shale reservoirs, thus the gas/water two-phase model is used to simulate the flow in these reservoirs.
Production of ultra-heavy oils is economically and technically challenging due to the very high viscosity of heavy oils, sharp viscosity increase over a small temperature drop and high operating costs. Reservoir oil can't even be mobilized by steam stimulation only due to inadequate reservoir energy. Even after the oils flow to the wellbore, the viscosity of the oils may exponentially increase when transported towards the wellhead due to the geothermal temperature decrease. The liquid oil could naturally turn into solid bitumen at any point where the temperature drops. The longer the travelling distance to surface for the oil, the bigger temperature drop, the greater the oil viscosity, and the more severe production challenges.
This paper presents the challenges associated with the production of ultra heavy oil in deep reservoirs in China. Operational difficulties widely exist in mobilization of in-situ oil, flow of oil from formation to wellbore, lifting of produced fluids from wellbore to surface, and surface processing and transportation of hydrocarbons. The sandstone reservoirs, sitting at a depth from 1600 to 1800 meters and having no support of any aquifer, contain approximate 4 million metric tons of 1.02~1.05g/cm3 heavy oil reserve. The oil-bearing formations have an average porosity of 27~29%, an average permeability of 1 Darcy and an original reservoir pressure of 16~17.5MPa. The oil viscosity at reservoir conditions (80°C) ranges from 6000 to 10000 centipoises (cP). Always keeping oil at a relatively low viscosity for feasible pumping is the theme topic with the thermal oil production in this type of reservoirs.
To find fit-for-purpose solutions, challenges had been analyzed in details for each part of the entire oil producing process covering the oil flow from the reservoirs to surface. The oil viscosity change with temperatures, the impact of oil viscosity reducers on the mobility of oil compounded with steam stimulation and CO2 injection for providing the initial energy to mobilize the heated oil, optimization of horizontal wells, screening of suitable wellbore lifting technology including wellbore heating and insulation and suitable chemicals for reducing the oil-water interfacial tension, and the steam stimulation optimization had been studied carefully prior to well drilling.
So far, 26 horizontal wells were drilled with an average of 130 meters horizontal section. Production data showed daily liquid rates at 800 tons at 55% water cut for all 26 producers after one year. The average peak oil production, the average cycle oil production capacity, the average cycle cumulative oil production of a single well was 25 metric tons per day, 14 metric tons per day and 2130 metric tons respectively. The average oil-steam ratio was 1.46 with a maximum oil-steam ratio of 5.26. The technologies discussed in this paper had been proved effective to produce ultra heavy oil from 1600 to 1800 meters formations with oil viscosity at 50°C conditions ranging from 180,000 to 260,000 cP.
When the seismic wave propagates in the formation, the seismic wavelet spectrum has obvious time-frequency variation, and it changes with the offset. Therefore, utilizing prestack seismic data with abundant amplitude and frequency information, this paper proposes a new method that computes attenuation parameters Q by extracting transient seismic waves in prestack domain. Firstly, the initial wavelet model is given. Then the four order cumulative quantity matching method is used to estimate the wavelet spectrum from the amplitude spectrum of the seismic record, and the optimal Q value is obtained. Finally, using the relationship of Q versus offset (QVO), the linear regression is used to obtain the formation Q value at the zero offset, that is, the true Q value of the formation. The model test results show that the method is reliable and accurate. The application effect of the actual data shows that the attenuation parameters extracted by this method can effectively identify the oil and gas in the reservoir and make up the deficiency of the traditional method.
Presentation Date: Wednesday, September 27, 2017
Start Time: 4:45 PM
Presentation Type: ORAL
Tahe Oil field is one of the biggest carbonate oil field in Tarim basin, northwestern China. Due to the Ordovician dissolution event, paleokarst reservoir was broadly developed there millions years ago. The seismic exploration shows that strong diffraction will be generated while seismic wave travelling through the karst cave. It will turn to be “strong serials reflection” after migration process. Great success has been obtained by using this special signal (serials reflection) to predict the karst reservoir. Many other methods are employed in recent years to extract more details from the seismic. Among those, seismic inversion is a popular one.
Seismic inversion is a powerful tool for reflective seismic exploration, where the reservoir is assumed to be lateral homogeneous beneath the earth. As a consequence, the traditional inversion might not work for the karst reservoir, since it has strong horizontal heterogeneous. In this paper, we try to study the applicability of post-stack seismic inversion in karst reservoir by using the modeling data and real seismic data.
Presentation Date: Monday, September 25, 2017
Start Time: 3:55 PM
Presentation Type: ORAL
Conventional methods for velocity model building are often limited in their ability to resolve geologic features characterized by high-contrast, short-wavelength velocity variations. If left unresolved, these features lead to velocity errors that significantly degrade images of deeper targets. We propose a new approach to correct for such velocity errors. By redatuming recorded wavefields to the vicinity of the geologic features we seek to resolve, we could identify far-field wavefront phase distortions resulting from short-wavelength velocity errors. We corrected for these phase distortions by estimating static time shifts that aligned redatumed wavefields to reference wavefields computed by demigrating a guide image. Using the estimated time shifts, we could improve the image either by adjusting the redatumed wavefields or by updating the velocity model.
Presentation Date: Wednesday, September 27, 2017
Start Time: 9:45 AM
Location: Exhibit Hall C, E-P Station 3
Presentation Type: EPOSTER