Experimental Investigation of Fracturing With Propellant in a Large Sandstone Block

Fig. 2—Lower section of the East half of the block under ambient light (left) and fluorescent light (right).

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

Propellants have been used in oil and gas wells to assist with perforating and creating near-wellbore stimulation. Propellants are electrically ignited at the perforated interval. Upon ignition, they rapidly create a large amount of gas, and the pressurization leads to breakdown of the formation. Hypotheses suggest that the pressurization leads to the creation of multiple fractures. This paper describes an experimental study with a new propellant and aims to understand the pattern of fracture creation with these propellants.


The fracture patterns created with propellant can include biwing fractures (in the direction perpendicular to the minimum principal stress), multiple/radial fractures, and explosive fractures, depending on the pressurization rate, as demonstrated in multiple small-scale experiments.

A large-scale block test had been conducted previously to observe fracture patterns with the propellant Arcite 368M. The test resulted in the creation of a dominant biwing fracture in the direction perpendicular to the minimum principal stress. This paper presents an experimental study aimed at comparing the performance of a second-generation propellant with that of Arcite 368M. The goal is to determine if the propellant provides a higher pressurization rate and if multiple fractures can be obtained with the second-generation propellant.

Laboratory Setup and Experimental Procedure

A 30×30×54-in. Colton sandstone block was prepared with a 2-in.-­diameter wellbore and loaded into a triaxial test frame. The openhole section of the borehole was etched with five small V-notches (2‑in. long), spaced 60° azimuthally and 2 in. longitudinally. A small propellant assembly was centralized in the wellbore adjacent to the openhole section. The propellant assembly, integrated with an ignition system, was sealed with a water- and pressure-tight covering. The propellant charge was 1 in. in diameter and 0.5 in. long. Multiple yard tests in which propellant charges were ignited in a confined chamber were conducted to determine the appropriate size of the propellant. The size of the propellant charge was reduced until the peak pressure was less than 10,000 psi, which had been determined to be the safe limit for the experiment to be performed in the triaxial test frame.

The wellbore was instrumented with a fast pressure transducer (recording at 10 kHz) to record the pressure transient caused by the propellant burn and subsequent fracturing events. In addition, the block was instrumented with six pressure/timing probes (also recording at 10 kHz). The purpose of the probes was to indicate arrival timing of the fracture tip, thus enabling determination of the fracture growth rate and symmetry.

Once the propellant assembly was secured in the wellbore, an upper manifold was attached. The manifold allowed communication with the wellbore—namely of pressurization, venting, and pressure monitoring data. With the full test assembly complete and secure, all personnel exited the immediate area. Final hookups were completed for remote control of the pressurization and ignition systems. System pressures were adjusted to target values, the fast data acquisition system was activated, the isolation valves on the manifold were closed, and the propellant was ignited. The propellant was ignited as soon as possible after the isolation valve was closed (less than 1 second) to minimize pressure decay from leakoff.

After ignition, the system was left to cool for 30 minutes, then the ventilation valve was opened remotely to depressurize the wellbore and ventilate any remaining combustion products. Boundary stresses were gradually reduced to nominal values, and the system was left overnight before removal of the block. Upon removal, the block was visually analyzed for fractures. Subsequently, the block was split across a horizontal plane in the middle and further along vertical planes to observe the geometry of created fractures and the region of fluid entrainment.

Observations and Results

Pressure Response. Fig. 1 shows the pressure response recorded from the wellbore along with the stresses on the block in a 120-millisecond time window. The blue curve shows the wellbore pressure rise upon propellant ignition followed by pressure decay. Main observations from the pressure response are

  • The initial peak wellbore pressure of 5,790 psi is attained in 1.4 milliseconds. This pressure could be interpreted as the fracture initiation pressure.
  • The maximum wellbore pressure reached is 6,660 psi.
  • Wellbore pressure exhibits a sawtooth pattern during the rise as well as during the decay. This pattern may indicate dual mechanisms of continued propellant burn (pressure increase) followed by initiation of new fractures (pressure reduction).
  • The pressure decays sharply until 2,600 psi, after which the decay is on a different slope.
  • The East/West (i.e., minimum) boundary stress is observed to increase, whereas the North/South (i.e., maximum) boundary stress does not change much. This is consistent with the expectation of fractures opening perpendicular to the lower East/West (i.e., minimum) stress and growing in the North/South direction (i.e., aligned with the maximum stress).
Fig. 1—Wellbore pressure and boundary stresses upon propellant ignition in a 120-millisecond window. (1) Initial peak pressure of 5,790 psi at 1.4 milliseconds; (2) maximum wellbore pressure of 6,660 psi; (3) wellbore pressure exhibiting sawtooth pattern; (4) sharp pressure decay until 2,600 psi, after which decay is on a different slope; and (5) minimum horizontal stress increases whereas maximum horizontal stress does not change.


The performance of this second-­generation propellant is compared with that of the first-generation propellant, Arcite 368M. In the experiment with Arcite 368M, a much larger propellant grain size of 1-in. diameter and 4-in. length was used. An initial peak pressure of 4,773 psi was observed with a rise time of 35 milliseconds. The much smaller burn time of the second-­generation propellant indicates faster burning characteristics compared with Arcite 368M. Also, a smaller amount of the second-generation propellant resulted in a higher peak pressure, confirming the higher calorific content of the second-generation propellant.

Visual Observations. After propellant ignition and cooldown, the upper manifold was dismantled and the propellant cartridge was removed. Fracture breakthrough is evident in the top region of the North face (aligned with the maximum stress direction) from approximately 3 in. below the midplane of the propellant location through to the top edge. Fracture breakthrough is also observed in the South face, although it is much more conspicuous. The East and West faces do not reveal any fracture breakthroughs, further consistent with the limited creation of dominant North/South fractures (aligned with the maximum stress) because of the prevailing stress field.

The rock was split into top and bottom halves by creating a fracture through the midplane of the openhole section. The dominant North/South biwing fracture is observed with no noticeable off-plane fracturing. A slight discontinuity is observed in the South wing, suggesting that different South fractures could have initiated at different axial locations along the wellbore.

The top and bottom halves of the block were further split into East and West halves by extending the existing dominant fractures to the North and South faces. Inspection of the East and West halves revealed the dominant biwing fracture pattern created with propellant ignition. Fig. 2 above shows the East half, comparing views under ambient and fluorescent light. The fluorescent light was used to investigate evidence of fluorescent dye, which was added to the wellbore fluid in addition to the visible dye. The dye clearly shows the biwing fracture surfaces, appearing primarily as multiple discrete diverging fan-shaped patterns. The dark (wet) but unstained regions appear to indicate leakoff into the matrix, either from the wellbore itself or potentially from nearby fractures that may exist close to the surfaces exposed during the block-splitting process.


Compared with previous findings, the rise time of the propellant falls in the planar fracturing region. However, the wellbore pressure response had a sawtooth pattern, indicating the possibility of creation of multiple fractures. The pressure probes in the direction of maximum horizontal stress inside the body of the block did not detect a pressure signal, indicating that the fracture propagated predominantly in the direction perpendicular to minimum horizontal stress. The sandstone block was cut open after propellant ignition to identify the fracture pattern. A dominant planar fracture in the direction perpendicular to the minimum principal stress was observed. Much shorter off-plane fractures were observed close to the wellbore in the bottom half of the block. The existence of these fractures could explain the sawtooth pattern in the wellbore pressure response. The findings and data from this experimental study can serve as important inputs to calibrate physics-based models that are aimed at modeling the process of propellant ignition and subsequent fracture growth.

The results also highlight the need to understand the performance of a propellant before deploying it in the field. The objective of deploying the propellant in the field needs to be identified clearly before the propellant is selected. The experiments showed that rise time was much shorter and peak pressure was much higher than they were with the first-generation propellant. This would signify deeper fracture penetration in the target subsurface zone compared with the first-generation propellant. However, the fracture pattern was dominantly planar. If multiple fracturing is required for the success of a treatment or a workover job in the field, the propellant chemistry would need to be altered to attain shorter rise times.

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper OTC 27563, “Experimental Investigation of Propellant Fracturing in a Large Sandstone Block,” by Sahil Malhotra, SPE, Peggy Rijken, and Alicia Sanchez, Chevron, prepared for the 2017 Offshore Technology Conference, Houston, 1–4 May. The paper has not been peer reviewed. Copyright 2017 Offshore Technology Conference. Reproduced by permission.

Experimental Investigation of Fracturing With Propellant in a Large Sandstone Block

01 June 2018

Volume: 70 | Issue: 6


Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.