After-Closure Analysis of DFITs Drives Design of Hydraulic-Fracturing Programs
This paper describes a diagnostic fracture injection test-analysis program and the suitability of the results from ACAs for use in hydraulic-fracture design.
The use of diagnostic fracture injection tests (DFITs) for prefracture investigation has become routine in the oil field, particularly for understanding reservoir properties and subsequently optimizing hydraulic-fracture design. A key component of an effective DFIT is an after-closure analysis (ACA) to assess the transmissibility of the formation and allow for effective design. This paper describes a DFIT-analysis program and the suitability of the results from ACAs for use in hydraulic-fracture design.
The Khazzan field is being developed currently and includes multiple gas-bearing formations. The primary development reservoir is the Barik sandstone, which is characterized by permeabilities on the order of 0.1 to 1 md. An additional reservoir under development is the Amin formation, which lies deeper than the Barik and is perhaps more unconventional in nature, with estimated permeabilities an order of magnitude lower than the Barik formation. Both reservoirs require hydraulic fracturing to produce at economically attractive rates and, as such, carry the same sort of challenges to reservoir understanding inherent to all unconventional plays. This was recognized in advance of the appraisal program, and an approach was taken to address these challenges in a more-holistic fashion, encompassing a full suite of data gathering, including surveillance and well testing.
One of the key tools used was DFIT along with associated ACA of the decline to determine reservoir properties. During the appraisal phase, significant rigor was aimed at ensuring high-quality data would be recorded and that an appropriate amount of time would be allocated to monitoring pressure declines to enable valid interpretations. This resulted in the ability to draw a good correlation between data gathered from the ACA operations and data collected from post-fracturing well-test data.
Methods and Process
Stimulation and Testing Sequence. The approach taken to stimulate and test the wells in Khazzan was to use a dedicated well-test unit. The overall sequence was as follows:
- Rig up well-test package
- Displace kill fluid and clean out with coiled tubing
- Perforate the target interval
- Rig up a tree-saver
- Perform DFIT and monitor pressure decline
- Perform main fracturing
- Establish post-main-fracturing-treatment pressure-decline period for fracture closure
- Rig down the tree-saver
- Clean out underdisplaced proppant with coiled tubing
- Flow the well back for cleanup and testing
- Perform a drift run with slickline to confirm hold-up depth
- Rig down equipment and handover well to operations
After the tree-saver is rigged up and the system is pressure tested, the DFIT is performed following a standardized procedure of pumping initially at a rate of 15 bbl/min until formation breakdown is observed. The rate is reduced then to 5 bbl/min until a total of approximately 25 bbl of clay-control water is pumped. The actual rates and volumes used are not fixed and can vary for specific applications (e.g., when reservoir quality is expected to be significantly lower than what is normally observed). In general, the higher rate is used to ensure a well-connected fracture at the wellbore during breakdown and then reduced to the selected DFIT pump rate that is still above the fracture-extension rate yet low enough to maximize the pump time. Some pressure is applied to the tubing/casing annulus to ensure that the completion-tubing stress-analysis safety factors are always honored.
Once the desired volume has been injected, the pumps are shut down and an isolation valve upstream of the pressure transducers is closed to allow the pressure decline to be monitored without being disturbed. Care is taken not to affect the annulus pressure during the decline period through any unintended pumping or intentional pressure bleedoff. Volume is measured by use of multiple calibrated magnetic flowmeters to ensure accuracy of the total volume pumped. The pressure-decline data are interpreted on site to determine key parameters including breakdown pressure, instantaneous shut-in pressure, net pressure, estimated closure pressure, fluid efficiency, and reservoir transmissibility.
The ACA decline was analyzed by use of the delta-time log-log plot to identify both formation linear flow and pseudoradial flow, on a plot of the decline data and its derivative on a log-log scale. Then the decline data during the pseudoradial-flow period are used to determine reservoir transmissibility. Initial reservoir pressure ideally should be known independently, rather than being determined from a Horner plot of the decline itself.
The fracture heights created during DFIT injections are not the same as those generated during the main fracture treatments, nor does the fracture height in the DFIT likely cover all the zones that encompass the log-derived effective permeability height to gas (kegH).
Petrophysical Log-Derived kegH
The petrophysical log-derived kegH is estimated by use of a rock-typing work flow and can be used reliably to estimate post-fracturing flow rate from the static reservoir log. The method defines the rock into discrete petrophysical rock types with similar hydraulic characteristics that focus on those that control the flow through the rock media. Saturation functions are derived by rock type, and rock-type-dependent gas and water relative permeability functions are used to determine effective permeability to gas. A pseudoinflow curve then is generated on the basis of the integration of the effective permeability to gas from the base to the top of the reservoir, with the maximum value termed here as the log-derived kegH.
The complete paper presents five case histories of ACA execution and interpretation, as well as integration with log-derived kegH and well-test data.
Larger Data Set
A larger subset of data can be used to illustrate the correlation between the ACA-determined transmissibility, the log-derived kegH, and the deliverability of the well after the fracturing treatments have been performed. The data presented here are from vertical Barik development wells that were treated with a single-stage hydraulic-fracturing treatment. In each case, the ACA pressure-decline data exhibited a –1.0 slope on the delta-time log-log plot, indicating that the pseudoradial-flow regime had been observed and transmissibility could be properly determined.
Fig. 1 is a plot of log-derived kegH and the ACA-determined transmissibility. The data have been normalized on a scale from 0 to 1 on the basis of the range of values determined across both parameters so that any data points falling on a unit line slope indicate a 1:1 correlation between the two. The figure illustrates that a relatively good correlation exists for a considerable number of wells and that, when the direct correlation does not match, the log-derived kegH is greater than that determined by the ACA. This trend reinforces the belief that the zonal coverage determined from the DFIT likely is affected in some cases by the limitations in fracture height achieved during the test. When the DFIT fracture height is factored in and compared with the log-derived kegH across the same interval, a better match is achieved, but it potentially will underestimate the final well deliverability for the post-fracture-stimulated well.
Conclusions and Recommendations
- Field developments that require hydraulic fracturing provide excellent opportunities to build an understanding of the key aspects of DFIT execution, data gathering, and ACA that can affect the quality and applicability of a subsequent correlation.
- The correlations that are derived can be strengthened when the technique is applied in a relatively consistent manner and integrated with complementary and independent means of determining similar information from a conventional subsurface approach.
- The execution of DFIT and subsequent ACA analysis should follow a rigorously standardized procedure in field implementation and analysis, to allow meaningful correlations to be drawn as the data set increases.
- When applied consistently and in line with relatively good reservoir quality, a database can be created of data from many wells with pressure monitoring times that are practical and not prohibitively lengthy.
- The data set generated from such a well-organized program can be used to calibrate post-stimulation reservoir performance and can complement and assist with a range of existing petrophysical work fronts.
- A robust work flow to estimate kegH from openhole petrophysical descriptions provides a valid means to calibrate the subsequent dynamic flow behavior and is a valuable tool to understand and bound the ACA interpretation.
- The application of DFIT and ACA in extremely tight wells often can be the only achievable means of estimating dynamic reservoir properties if the reservoir will not flow even after stimulation, providing a valuable piece of data to assist in appraisal decision-making.
- Created fracture height affects the correlation between ACA-determined properties and log-derived estimates of kegH or conventional well-test-determined methods, with the ACA-determined properties generally consistently lower than the latter two.
- Applications of DFIT for horizontal wells, in highly laminated reservoirs, can be quite limited and, in general, difficult to reconcile with other means of determining the same properties because of the inability to apply conventional vertical-well-surveillance methods easily.
- The use of DFIT and subsequent ACA analysis is a low-cost method of obtaining a substantial fieldwide data set during field development that can be used to assist quickly with development choices that normally may require a working production facility.
This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE-191437-18IHFT-MS, “ACA Practical Considerations: When Is It Accurate and How Should It Be Used To Improve Reservoir Stimulation,” by O.A. Ishteiwy, SPE, M. Jaboob, and G. Turk, BP; S. Dwi-Kurniadi, SPE, Schlumberger; A. Al-Shueili, SPE, A. Al-Manji, and P. Smith, BP, prepared for the 2018 SPE International Hydraulic Fracturing Technology Conference and Exhibition, Muscat, Oman, 16–18 October. The paper has not been peer reviewed.