Investors and operators like to draw attention to the lengthy history of oil production from the Permian. Benefits include well-known reservoirs and lower cost brownfield options for services and infrastructure. Today's boom mirrors prior periods of intense Permian activity, but using prior comparisons of tight oil reservoir behavior can result in dangerous conclusions.
The Permian houses thousands of vertical wells that have been online for decades. The actual terminal decline rates of those wells can be modeled with empirical data, and our analysis shows that they sit between 5% and 10% annually. However, pure field data for horizontal tight oil wells does not go back as far, so terminal decline values from vertical wells or general ‘shale’ declines are often applied to Wolfcamp type curves as a proxy. This is a risky practice.
Building evidence suggests that the most active Wolfcamp sub-plays may eventually still have annual decline rates greater than 10% five or more years into their lifespan. Not recognizing this and modeling with the lower proxy value from older, analogue tight oil plays could result in overstating production potential and overvaluing projects. Wolfcamp players without exposure to other basins may need to resort to M&A to fill production gaps. We are already seeing signs of this in 2018 deals with Concho and Diamondback using M&A to acquire larger undrilled acreage footprints.
In many cases, Permian tight oil wells realize more than 40% of the estimated ultimate recovery (EUR) within only 36 months of being online. Nearly 50% net present value (NPV) is realized by year five. Rightfully so, we have observed the analytical focus being placed on each well's first few years. This emphasis has also been driven by the limited number of tight oil wells with more than five years’ worth of production history. Less than 20% in the Permian tight oil wells have been producing more than 60 months.
As the Wolfcamp play matures, the later-life performance of wells starts to matter more. Even with record rig counts, the number of new wells drilled each year becomes a smaller proportion of the total wells contributing to supply.
Pineda, Wilson (BP) | Wadsworth, Jennifer (BP) | Halverson, Dann (BP) | Mathers, Genevive (BP) | Cedillo, Gerardo (BP) | Maeso, Carlos (Schlumberger) | Maggs, David (Schlumberger) | Watcharophat, Hathairat (Schlumberger) | Xu, Weixin (Wayne) (Schlumberger)
Deepwater depositional environments in the Gulf of Mexico can be very complex. Accurate determination of depositional facies is important in these capital-intensive fields. The most common reservoir facies are laterally extensive sheet sandstones with thin mudrock layers, channel complexes (isolated or amalgamated) and channel-levee complexes (often with poor reservoir communication). Reservoirs are often complicated by steep dips close to salt domes and the presence of potential fluid conduits due to faults or fractures. Borehole images aid in determining the character of the sediments, as well as improve net sand calculations, and illuminate the geology in the near wellbore region both in structure and depositional environment, and to provide valuable geomechanics information for the determination of the stress vector.
A well was recently drilled through one of these deep water sediment sequences in the Gulf of Mexico with an oil-based mud (OBM) system. An extensive acquisition program included a series of logging while drilling (LWD) and wireline images. In addition to the current LWD lower resolution borehole imaging tools, a new LWD dual physics OBM imager was deployed for the first time in this field. Five different types of physics were acquired, including lower-resolution images from nuclear measurements (gamma ray, density and photoelectric) and the high-resolution images from dualphysics OBM imager (DPOI) which is based on resistivity and ultrasonic measurements. Wireline high-resolution OBM resistivity images were also acquired. This paper shows a comparison of images collected with the new DPOI versus traditional LWD images and high-resolution wireline resistivity images.
Comparisons of the types of features observed from the various imaging tools were made, showing how the differences in physics, resolution and time of logging affects the images, as well as the impact these factors can have on subsequent interpretations. Four main categories of features are included in comparisons between the tools: sand-rich sections, consistently dipping mudrocks, chaotic zones and fractures/faults. The different images allow fuller interpretation of the gross sequence. In general, the higher the resolution, the more detailed and confident the interpretation is, particularly where the hole conditions are good. In degraded borehole sections, the LWD acquisition was beneficial for obtaining images as early as possible, when damage was at a minimum. The impact of the differences in the physics depends on the properties and contrasts being imaged. This is observed with fractures - both conductive and resistive examples can be seen on both LWD and wireline images. The ultrasonic images are complementary with both low and high amplitude fractures seen, providing more confidence in the fracture interpretation.
Merza Media, Adeyosfi (Schlumberger) | Muhajir, Muhajir (Pertamina Hulu Energi Tuban East Java) | M. Wahdanadi, Haidar (Joint Operating Body Pertamina Petrochina East Java) | Agus Heru, Purwanto (Joint Operating Body Pertamina Petrochina East Java) | Anugrah, Pradana (Schlumberger) | Dedi, Juandi (Schlumberger)
Most of sedimentary basins in Indonesia contain productive carbonate reservoirs. Geologically, the reservoirs are mostly part of a reef complex and carbonate platform, with basinal areas situated mainly in the back arc of the archipelago. Many of the productive carbonate reservoirs have dual porosity systems with widely varying proportions of primary and secondary porosity. Carbonates of the Tuban formation in Platinum field represent two carbonate buildups identified with similar effective porosity but different productivity. This paper describes a method for characterizing secondary porosity distribution at the wellbore and field scales to address the productivity difference between the northern and southern carbonate buildups in this field.
To resolve the challenges in characterizing secondary porosity in a carbonate formation, an integrated workflow was developed that consists of combination of quantitative and textural analysis based on borehole images at the single-wellbore scale and the seismic inversion result to control lateral distribution at the field scale. Analysis based on borehole image log provides high-resolution porosity characterization based on its size, interconnectivity, and type. The result of the single-wellbore analysis will be distributed at the field scale with control of a seismic attribute such as acoustic impedance (AI). Acoustic impedance is built with stochastic seismic inversion to provide a higher-resolution result compared to the deterministic seismic inversion method.
The result of the analysis based on borehole images at the single-wellbore scale shows most of the northern carbonate buildup wells demonstrate high development of porosity from interconnected vugs, leading to a relatively high permeability interval. In contrast, the southern carbonate buildup wells demonstrated low secondary porosity development. Low secondary porosity development is related to cemented zones and the predominance of claystone facies in a well. Later, the result of the single-wellbore scale analysis was distributed at the field scale with seismic attribute control such as AI. The Platinum field shows a negative correlation between AI and porosity with a value of -0.769; hence, the acoustic impedance from stochastic seismic inversion can be used to control the porosity distribution. The secondary porosity model shows a distinct difference between the northern and the southern carbonate buildups. The northern carbonate buildup has higher average secondary porosity compared to the southern carbonate buildup. The result was confirmed with production data; the northern carbonate buildup has higher productivity compared to the southern carbonate buildup.
This integrated workflow provides a comprehensive and high-resolution analysis of secondary porosity distribution at the single-wellbore scale and the field scale. Thus, this workflow can reduce uncertainty during reservoir characterization, well placement, and production planning.
Amer, AimenAi (Schlumberger) | Sajer, Abdulazziz (Kuwait Oil Company) | Al-Adwani, Talal (Kuwait Oil Company) | Salem, Hanan (Kuwait Oil Company) | Abu-Taleb, Reyad (Kuwait Oil Company) | Abu-Guneej, Ali (Kuwait Oil Company) | Yateem, Ali (Kuwait Oil Company) | Chilumuri, Vishnu (Kuwait Oil Company) | Goyal, Palkesh (Schlumberger) | Devkar, Sambhaji (Schlumberger)
Producing unconventional reservoirs characterized by low porosities and permeabilities during early stages of exploration and field appraisal can be challenging, especially in high temperature and high pressure (HPHT) downhole conditions. In such reservoirs, the natural fracture network can play a significant role in flowing hydrocarbons, increasing the importance of encountering such network by the boreholes.
Consequently, the challenge would be to plan wells through these corridors, which is not always easy. To add to the challenge, well design restrictions dictate, the drilling of only vertical and in minor cases deviated wells. This can reduce the possibility of drilling through sub-vertical fracture sets significantly, and once seismic resolution is considered, it may seem that all odds are agents encountering a fracture network.
This article addresses a case where a vertical well is drilled, in the above-mentioned reservoir setting, and missed the natural fracture system. The correct mitigation can make a difference between plugging and abandoning the well or putting it on production.
The technique utilized is based on a borehole acoustic reflection survey (BARS) acquired over a vertical well to give a detailed insight on the fracture network 120 ft away from the borehole. Integrating this technique with core and high-resolution borehole image logs rendered an excellent match, increasing the confidence level in the acoustically predicted fracture corridors.
Based on these findings new perforation intervals and hydraulic stimulation are proposed to optimize well performance. Such application can reverse the well decommissioning process, opening new opportunities for the rejuvenation of older wells.
Some of the first high-pressure/high-temperature (HP/HT) development wells from Elgin and Franklin have been exposed to sustained casing pressures in their "A" annulus, threatening the integrity of the wells. The sustained pressure in the annulus was attributed to ingress through the production casing of fluids from the overburden chalk formations of the Late Cretaceous. The mechanism triggering the ingress into the "A" annulus was uncertain until access to the production casing was achieved. A recent campaign to abandon development wells of Elgin and Franklin that had sustained "A"-annulus pressure brings new evidence on the mechanism causing the ingress. Temperature surveys have been acquired in the production tubing to identify the fluid-entry points in the production casing. Multifinger calipers have been run in the production casing, revealing several shear-deformation features. These deformations are localized along various interfaces, and are attributed to the stress reorganization associated with the strong reservoir depletion. A detailed analysis of the surveys shows that fluid ingress is occurring at distorted casing connections, if located close to weak interfaces along which shear slip occurs. The shear deformation is suspected to cause a loss of the sealing capacity of the connection, leading to gas ingress into the "A" annulus. This conclusion emphasizes the need to consider any potential for localized shear deformations in designing casing for HP/HT wells.
The integration of seismic with production data becomes more and more demanding in reservoir development. Traditionally, the integration either uses a seismic attribute to correlate with the production data or uses the reservoir model to reconcile with the production history, particularly in the area of time-lapse seismic. However, how to integrate 3D seismic with production data remains a challenge. In this paper, an RGB fusion of seismic attributes with production data is proposed. Instead of using three seismic attributes for the three base colors, R(red), G(green) and B(blue) for fusion, the new approach replaces one component, for example G, with the production data or a dynamic attribute. The production data are always at the well locations. In order to integrate with seismic attributes, the production data-based attribute, such as water-cut, is converted to an estimated spatial distribution over time. For history matching review purposes, a direct extraction of the dynamic properties from the simulation model can be applied. The estimated dynamic attribute, which is derived from either the production data or the simulation model, is integrated into the RGB fusion. Via visualization over time, it would be a good tool to compare the spatial variation of the production attribute with that of the seismic attributes. The proposed approach is applied to a mature field. It shows good consistency between the reservoir dynamics with the seismic attributes. This is helpful in understanding the dynamic behavior and the potentials by combining seismic attributes and production data.
Presentation Date: Monday, October 15, 2018
Start Time: 1:50:00 PM
Location: 209A (Anaheim Convention Center)
Presentation Type: Oral
Summary Recent field experiments have demonstrated that distributed acoustic sensing (DAS) can be used to record strain in fiber optic cable at mHz frequencies. However, the effect of fiber optic cable construction on strain transfer has not been evaluated. The laboratory experiments presented here were designed to mimic fiber optic cable cemented into a borehole that intersects bedrock fractures. Hydraulic stress on the fracture is expected to stretch the cemented borehole which can be sensed by DAS. In the laboratory, we cemented five different fiber optic cable constructions into a pipe and then strained the pipe periodically using stepper motors.
Panja, Palash (Department of Chemical Engineering, and Energy & Geoscience Institute, University of Utah) | Velasco, Raul (Energy & Geoscience Institute, University of Utah) | Deo, Milind (Department of Chemical Engineering, University of Utah)
In this work, we estimate the Stimulated Original Oil In Place (SOOIP) of hydraulically fractured horizontal wells in prominent shale plays. This is done by compiling production data from hundreds of wells belonging to the Bakken, Niobrara, Wolfcamp, Eagle Ford, Bone Springs, and Woodford totaling over 2,500 wells. Additionally, we present probabilistic distributions of SOOIP with mean, standard deviation, P10, P50, and P90 estimates for each play.
To circumvent the challenge of data availability for each well, we use the findings of a previous study where all reservoir unknowns are grouped into two major parameters. One of these parameters, alpha, is a function of the stimulated reservoir volume, compressibility, and pressure drawdown, where the last two are unknowns. While alpha is determined with high confidence for each well, we account for the uncertainty of compressibility and drawdown values across wells by assuming a normal distribution for these parameters. Lastly, by incorporating 1 million Monte Carlo samplings and a Mersenne Twister random number generator we estimate SOOIP distributions for each play with varying degrees of confidence.
The final results show that the Niobrara and Bakken have the highest mean SOOIP values per well while the values for the Woodford and Bone Springs are the lowest among all six plays considered. Volumetric calculations using data from the literature qualitatively corroborate these findings. New insight on the stimulated volumes per well for prominent shale plays can be derived from these results as they correlate to horizontal well length, formation thickness, and hydraulic fracture half-lengths in each play.
Scaling precipitation in the form of calcium or iron sulfate (CaSO4 – FeSO4) appears to be a common problem in most of the wells completed in the Yeso formation in the Permian Basin. Accumulation of these forms of scale in the perforations and the downhole equipment can create severe production losses and therefore they require constant, some times costly, treatments. Additionally, their presence in the wellbore hinders the efficiency of the acidizing procedures, performed by the operators to further increase the productivity of their wells.
The objective of this paper is to present an innovative workflow, which significantly reduces the scaling effects and allows the acidizing of these carbonate reservoirs to be more efficient. Results are avialble for more than 20 wells, from Occidental Petroleum’s assets in the North New Mexico region. Details regarding the evolution of the procedure, the types and amounts of the chemicals are extensively discussed in this paper.
The procedure starts with the candidate well selection, which preferably are located in the highest OOIP areas to maximize the efficiency. The second step involves the water sampling and analysis to determine the types of scaling and consequently the required chemical treatment. The innovation in this step has to do with the time that converters are left in the wellbore and the extra chemicals that have been added compared to the previous procedures. The last step involves the deployment of the acid downhole, where three different methodologies have been tested and evaluated based on their efficiency and well returns.
The proposed approach has been successfully applied to more than 20 wells and the results are encouraging showing an average incremental oil production of ~600% while the execution cost remains very low.
Monthly production data can be easily obtained from public and commercial databases, and are essential information to assess the performance of operators across different fields. Additionally, analyzing the production of the wells via maps is helpful in the visualization of general aspects of the field geology and potential identification of sweet spots. In this context, decline curve models are a feasible choice to process the available production data, because this is the only data required by these models, they have a reduced number of parameters which can be promptly history matched and analyzed, and they provide production forecasts and estimated ultimate recoveries (EUR's). As summarized in the work of Arps (1945), decline curve analysis has been applied to predict oil production since the beginning of the last century. Arps (1945) presented differential equations for the rate-time relationship, which resulted in the exponential, harmonic and hyperbolic models. Although initially these models were derived from empiricism, subsequent works attempted to explain those equations from a fluid flow perspective (Fetkovich, 1980; Camacho-Velázquez, 1987; Camacho-Velázquez and Raghavan, 1989). However, in unconventional reservoirs, the Arps equations are not a reasonable extrapolation for the extended transient flow period, because it does not account for a transition to boundary dominated flow and results in an infinite EUR (Lee and Sidle, 2010).