Well-Placement Timing, Conductivity Loss Affect Production in Multiple-Fracture Wells

Topics: Hydraulic fracturing Shale oil
Fig. 1—Side view of the reservoir model showing the relative position of the three wells in the base-case model. The horizontal sections of Wells 1 and 3 are placed in Layer 2, and that of Well 2 is placed in Layer 4 (z-scale is exaggerated).

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

Horizontal wells in liquids-rich shale plays are now being drilled such that lateral and vertical distances between adjacent wells are significantly reduced. In multistacked reservoirs, fracture height and orientation from geomechanical effects coupled with natural fractures create additional complications; therefore, predicting well performance using numerical simulation becomes challenging. This paper describes numerical-simulation results from a three-well pad in a stacked liquids-rich reservoir (containing gas condensates) to understand the interaction between wells and production behavior.

Numerical Simulation

The reservoir simulator used for this study was designed to handle ­unstructured-grid-based simulation cases. Most of the numerical reservoir simulators that are used for modeling horizontal wells with multiple hydraulic fractures are based on structured grid cells in which the hydraulic fractures are modeled as symmetric biwing fractures perpendicular to the wellbore. In most cases, they use local grid refinement (LGR) to incorporate the hydraulic fractures into the model, which generally works well if a single well is involved and the grids are not tied to the Earth models. However, if the wells and the hydraulic fractures are not orthogonal to the Earth-model grids or if the reservoir contains nonorthogonal secondary fractures (natural or induced), modeling them with LGR becomes a challenge. When multiple wells that are not parallel are drilled from a single pad, properly representing the wells with hydraulic fractures in a structured-grid-based reservoir-simulation model becomes even more challenging. By contrast, the unstructured-grid-based reservoir models are not restricted by any of these limitations—they can have any geometry, size, or orientation for the wells and can include primary hydraulic fractures, secondary fractures, and open natural fractures. Instead of using grid cells that are parallel in shape, the unstructured grid cells can have any arbitrary shape. Incorporating realistic induced hydraulic fractures in a reservoir model is easier if unstructured grids are used.

Reservoir-Simulation Model

The reservoir model consisted of five horizontal layers with varying properties in each layer. The horizontal sections of Wells 1 and 3 go through the middle of Layer 2, and that of Well 2 goes through the middle of Layer 4. Each well is completed exactly the same way.

Fig. 1 above is a side view of the base-case reservoir model with the layout of the three wells. The position of the horizontal section is shown by a yellow dot at the end of each green line, representing the vertical section. The yellow horizontal line goes through the middle of the reservoir (which is also the middle of Layer 3). The hydraulic fractures in the base-case model are confined within certain layers—the fractures for Wells 1 and 3 are confined within Layers 1 and 4, and the fractures for Well 2 are confined within Layers 2 and 5. The base-case model had a total of 419,500 unstructured grid cells. In the base-case model, the horizontal distance between two parallel laterals was 1,000 ft and the vertical distance between two adjacent parallel laterals was 200 ft; hence, the actual distance between two adjacent parallel wells was approximately 1,020 ft. The gap between the fractures from two adjacent wells is 200 ft in the base-case model


Effect of Staggering of the Wells. For this study, a new case was created with zero vertical offset between wells (i.e., with the horizontal sections of each well at the same vertical position in the middle layer of the reservoir). To preserve the overall fracture height at 300 ft, two additional layers were created by dividing the uppermost and lowermost layers in half.

The base case shows less pressure change at the location of Well 2 because its fractures do not reach the top layer directly, as a result of staggering (the top of hydraulic fractures of Well 2 are 100 ft below the top surface in the model). The pressure distribution around the fractures for the case in which the wells are in the same horizontal plane and the total fracture height equals that of the fracture height for the base case is more uniform compared with the base case. Finally, the pressure distribution for the remaining case is also more uniform, and, because all the fractures reach the top layer, the pressure change is more intense.

The cumulative-oil-production increase is 8.3%, and the cumulative-gas-production increase is 12.1%. The case in which the wells are placed in the same plane at the center of the reservoir but the fractures penetrate all of the reservoir layers shows the highest increase in production of oil and gas. Although it is not a direct comparison with the two other cases because of the extra fracture height, this case represents a ­cumulative-oil-production increase of 39% and a cumulative-gas-production increase of 22%.

Effect of Delaying Completion of Well 2 by 1 Year. The second scenario investigated the effect of delaying the completion of Well 2 (the well in the middle) by 1 year. Along with the base case, three other cases were created with different fracture lengths for Well 2 in view of the fact that it is possible to create fractures of different lengths than the original case because of stress-regime changes. Additionally, a case was created with just two wells (Wells 1 and 3). The individual rate plots for oil and gas indicate a spike in both oil and gas production when the third well is started. Overall, delaying completion, if not accompanied by creating fractures of equal length, as with the two original wells, results in lower cumulative oil and gas production than the base case, which had three wells completed at the same time.

Effect of Including Natural Fractures in the Model. A natural-fracture generator that can work with calibrated or stochastic fracture models was used to populate the reservoir model with natural fractures. The natural-fracture network is generated with a total of 356 natural fractures.

The presence of natural fractures in the model (i.e., connected ones) resulted in higher cumulative production for both oil and gas (3.8% increase for oil and 1.1% increase for gas). Connecting to the existing natural-fracture network and keeping it open during the production period helped improve overall recovery.

Effect of Matrix Permeability and Condensate/Gas Ratio (CGR). Two new cases were created to evaluate the effects of changing matrix permeability and CGR. For the first case, the matrix permeability of each of the five layers is increased from the base case by a factor of 10 and the simulation is run for 5 years. For the second case, a lighter reservoir fluid with a CGR of 50 STB/MMscf (one-fifth of the base-case CGR) was used with the same equation of state but with different pseudocomponents than the base case. A significant increase in cumulative oil production is seen when matrix permeability is increased by a factor of 10 (66.2% increase for oil and 39.6% increase for gas). For the case in which the CGR is reduced by a factor of five, the cumulative oil production is reduced by 14.2% and the cumulative gas production increases by 10.5%.

Effect of Conductivity Loss in the Fractures and in the Matrix. Two new cases were created to evaluate the effects of conductivity loss in the fractures and in the matrix, which occurs during the life of a hydraulically fractured well in any unconventional reservoir. In most simulators, this effect is simulated by using separate (or the same) transmissibility-multiplier functions for the matrix and the fractures. The first of the two new cases looking at the effect of conductivity loss is created from the base case, and the second is created from the case with matrix permeability that is 10 times that of the base case. In both of these new cases, the same two transmissibility-multiplier tables were used. For the base case, a loss of conductivity resulted in only a 0.2% loss in cumulative oil production but a 67.0% loss in cumulative gas production. For the high-permeability-matrix case, permeability damage resulted in an 11.5% loss in cumulative oil production and a 65.5% reduction in cumulative gas production. For both values of matrix permeability, the loss of cumulative gas production is more severe than the loss of cumulative oil production.


This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 180448, “Effects of Well-Placement Timing and Conductivity Loss on Hydrocarbon Production in Multiple-Hydraulic-Fracture Horizontal Wells in a Liquids-Rich Shale Play,” by Shameem Siddiqui, SPE, Doug Walser, SPE, and Ron Dusterhoft, SPE, Halliburton, prepared for the 2016 SPE Western Regional Meeting, Anchorage, 23–26 May. The paper has not been peer reviewed.

Well-Placement Timing, Conductivity Loss Affect Production in Multiple-Fracture Wells

01 March 2018

Volume: 70 | Issue: 3