Surface Drilling Data Can Help Optimize Fracture Treatment in Real Time

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

Like what you're reading?

Sign up for the new JPT Unconventional Insights newsletter to get the latest news to your inbox.

The objective of optimizing a fracture design is to spend the least amount of money and get the most productivity out of the reservoir by stimulating and contacting as much reservoir rock as possible. This paper presents a unique work flow that addresses in real time the challenges of perforation and fracture-treatment design while accounting for the lithologic and stress variability along the wellbore and its surroundings.

Surface Drilling Data in Fracturing Design and Analysis

Existing fracturing-design tools often make simplistic assumptions because of a lack of input data. These designs still use layer-cake models and sometimes rely on the data of a nearby well rather than data measured at the considered well. As cost-cutting efforts accelerate in unconventional wells, expecting a log at every well will not be feasible. Because the outcome from the fracturing design heavily depends on specific geomechanical properties, stresses, and surrounding natural fractures, changing the current industry practice of using a nearby well and assuming all subsurface properties to be the same for all stages is imperative. The software used by the authors is able to derive a 3D distribution of the rock properties required as input in the fracturing design. This fracturing-design input available at any well is made possible by using surface drilling data to compute geomechanical logs, pore pressure, stresses, porosity, and natural fractures.

Transforming Drilling Data Into Log Properties. The surface drilling data commonly available on any rig includes weight on bit, rate of penetration, rotational speed, and torque. Transforming this data into valuable input for fracturing design starts by computing the corrected mechanical specific energy (CMSE), which removes frictional pressure losses along the drillstring to obtain accurate estimations at the bit. Confined compressive strength (CCS) is then estimated. When the CMSE and the CCS are estimated correctly, the drilling efficiency is computed. Deviations from expected pore pressures are recognized by comparing a hydrostatic trendline with the drilling-efficiency data.

The methodology to derive pore pressure from drilling data is based on the concept that the energy spent at the bit to remove a unit volume of the rock is a function of the differential pressure to which the rock is subjected while drilling. The differential pressure provides useful information for deriving the pore pressure. Feeding the rock-strength and pore-pressure information to the real-time geomechanical model helps identify the differential stress variability along the wellbore. Other rock mechanical properties also can be estimated using various lithologies derived from rock-strength information.

Propagating the Derived Log Properties Into 3D Geocellular Models. The reservoir properties computed from the surface drilling data are distributed in a 3D stratigraphic geocellular model using geostatistics or neural-network-modeling tools incorporated with the drilling data and completion design (Fig. 1). The new fracturing design uses an actual 3D distribution of the rock properties along the 3D fracture planes and does not make the common assumption of constant properties. Because the rock properties needed for the fracturing design are now available in a 3D stratigraphic geocellular model, a box grid is used to extract the necessary rock properties around the horizontal well that will be considered for the fracturing design. The design will benefit from the availability of variable rock properties in 3D, which will have a major effect on the resulting fracture geometry. Table 1 of the complete paper outlines the ways in which the new work flow surpasses many of the shortcomings of existing fracturing-design software.

Fig. 1—Geostatistics of neural-network-modeling tools transform rock properties derived from surface drilling data into 3D models that honor the stratigraphic framework.

Building Geoengineered Completions

As the need for achieving better cluster efficiency is recognized, experimentation begins with various tools for geoengineering stages. Once again, surface drilling data provide a cost-effective solution in which multiple outputs can be used to geoengineer the completion.

Among the outputs provided by the surface drilling CMSE work flow, the CMSE and the stress brittleness can be used as a profile log to grade the rock and design a geoengineered completion. Once the stage spacing has been finalized, the next step is to design and optimize the number of perforations and cluster design in any stage. The number of shots per stage can be designed on the basis of a desired perforation friction, injection rate, and injection-fluid density. Limited-entry design is a technique in which engineers limit the number of entry holes in order to create sufficient backpressure, which makes it easier to pump bigger jobs for designed injection rates. The user can optimize a limited-entry perforation design through different approaches, termed aggressive, moderate, or conservative. Once the number of perforations has been finalized, the cluster design is computed on the basis of the shots-per-foot tool.

Once the stage and perforation designs have been completed, the stress brittleness could be considered for designing cluster placement. The variability of the reference log along the wellbore then is used to design the cluster spacing in a way that ensures perforations are placed in rock with similar fracture gradients to improve cluster efficiency.

Current industry software requires a manual input of the completion design. This shortcoming is addressed in the current work flow by adding the capability to optimize either geometric or a geo­engineered completion design automatically on the basis of a reference log.

Accounting for Natural Fractures and Heterogeneous Stress Fields

The new fracturing design relies on a full-continuum geomechanical solution that uses a continuous natural-fracture model derived by using any type of fracture indicator, including those derived from surface drilling data. Unlike other approaches based on discrete fracture networks, the one used with this fracturing design requires a validated natural-fracture model. As a result, the geomechanical simulation accounts for the geologic realities of the reservoir and, consequently, can predict the microseismicity, confirming the validity of the natural-fracture model and the physics used to model the interaction between the hydraulic and natural fractures. Given this validation, the resulting strain can be used to compute the envelope that will provide the geomechanical half-lengths that capture the asymmetric behavior caused by the lateral stress gradients.

Proppant Transport and Optimal Treatment

One of the key results of the design is the proposed treatment schedule to achieve the target fracture geometry. The engineer also can analyze the evolution of the key parameters (asymmetric half-length and heights, net pressure, and width of each cluster) during fracturing and the resulting fracture geometry at each cluster, which includes the asymmetric fracturing heights reflecting the use of the 3D distribution of key rock properties and the 3D distribution of fracture conductivities. When the treatment is already executed, the pressure data could be matched and, if microseismic data are available, could also be used to understand the difference between propped lengths and microseismic events better.

In addition to multiple new geological and geomechanical constraints added to the fracturing design, the data input, treatment setup, computation of the fracturing geometry at each cluster, and the proposed treatments to achieve a certain fracturing job are performed in a few hours, which contrasts dramatically with current software that, even with fewer geologic and geomechancal constraints, may take days to set up and run. The implementation of these fast computational approaches introduces another way of using real-time fracture modeling.

Real-Time Geologically and Geomechanically Constrained Fracture Modeling

Real-time fracture modeling starts with a predesign job in which all surface drilling data are transformed into geological and geomechanical 3D models to be used as input in the fracturing design. The proposed treatment is executed, and, as the pressure data are collected in real time, the error between the simulated response and the actual response is computed. When there is a major difference, the real-time fracturing-design tool is used now as a fracturing-analysis tool; the measured data are matched by the model to provide the updated geometry. The only fracturing parameters adjusted during the history match are

  • Leakoff factor, a multiplier applied to the 3D distribution of the natural fractures or effective permeability used as input in the fracturing design
  • Perforation friction
  • Wellbore friction

These parameters are almost impossible to measure before the fracturing job and will be the only matching parameters; all others remain fixed because they were the result of an extensive reservoir-modeling effort augmented with surface drilling data.

The pumping is monitored in real time and its measured pressure is compared with the predicted pressure. If the error between the measured and simulated pressures becomes large, the three matching parameters are adjusted to ensure that the model pressure matches the measured one. When this match is achieved, the engineer can alter the fracturing design if a suboptimal fracture develops because of unexpected subsurface conditions. Once the real-time fracturing design confirms the new geometry resulting from the proposed changes, and it satisfies the engineer’s needs, then the change in rheology is executed during the fracturing and the new model will match the new subsurface response.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 189810, “Real-Time Completion Optimization of Fracture Treatment Using Commonly Available Surface Drilling and Fracking Data,” by Mohit Paryani, Djamel Sia, Bhavina Mistry, Drew Fairchild, and Ahmed Ouenes, FracGeo, prepared for the 2018 SPE Canada Unconventional Resources Conference, Calgary, 13–14 March. The paper has not been peer reviewed.

Surface Drilling Data Can Help Optimize Fracture Treatment in Real Time

01 June 2019

Volume: 71 | Issue: 6