Delaying the start of new wells is understood to reduce the net present value (NPV) of a section, but variations in the arrangement of infill wells have not been examined thoroughly. In this paper, the effect of timing and pattern of well placement on NPV is studied. Three scenarios were evaluated: an infill scheme in which future wells are drilled between existing wells, a linear scheme in which future wells are immediate offsets of existing wells, and a hybrid scheme that is a combination of the infill and linear schemes.
Introduction
All previously published work using reservoir modeling to ascertain optimal well density assumes idealized conditions: zero well-to-well interference through fracture communication and simultaneous production initiation for all wells. Previous work postulated how infill-well drilling can improve cumulative recovery from shale-gas reservoirs because the distance between wells and the type of completion technique affect the amount of additional possible recovery. NPV considerations are important when operators consider either infill drilling or enhanced recovery methods to improve cumulative recovery from reservoirs that have gone through primary recovery. Reservoir simulation can help determine areas where depletion has not occurred and thus are optimal for potential infill-well development.
The base case was chosen to be five wells per section (1,056 ft between wellbores) because of previous work. This paper determined how the economic results vary for three different field-development schemes for this optimal-well-spacing scenario.
Basis of Modeling
The numerical reservoir simulations used in this study capture flow from the stimulated reservoir volume (SRV) and external reservoir volume (XRV) within the drainage volume, as shown in Fig. 1 (above).
Average reservoir properties for the models are given in Table 1. The properties were averaged across 160 wells in a shale play in North America. The completed well length was the same for all cases, 3,840 ft, and the lateral was placed in the middle of the pay. The base case had a matrix permeability of 50 nd, a fracture spacing of 80 ft (which corresponds to 48 fractures on the 3,840‑ft lateral), and a fracture half-length of 500 ft. The gas-production rate for each well in the section was 3,000 Mscf/D.
Base Case
The base case, described in earlier work, is shown in Fig. 2. The optimal number of wells per section is determined using normalized incremental cumulative recovery, which is defined as the number of wells beyond which any additional well provides a recovery factor less than 50% of that of the first well. Thus, the optimal number of wells per section for the base case is five wells, as confirmed by the peak in NPV. In this paper, all sections are assumed to require five wells for optimal development, on the basis of Fig. 2.
Effect of Field Development
Drilling schedules were studied to understand the effect of time on the value of the asset and on the optimal number of wells. Wells are not drilled simultaneously, and they are rarely fractured simultaneously (though this is increasing in popularity). Also they are often not flowed back simultaneously in a real reservoir. As a result, three different timing schemes (i.e., infill, linear, and hybrid) with perfectly adjacent hydraulic fractures were considered. Several criteria can be considered to determine when to bring the wells on line (e.g., flow regime, fixed points in time, percentage of cumulative production). The technique investigated in this paper was based on the quartile gas cumulative midpoint from the numerical model (at 12.5, 37.5, 62.5, and 87.5% of total cumulative production). Quartile gas cumulative midpoint was chosen to mitigate the operating conditions that vary by operator. In other words, if Operator A severely chokes a well and Operator B opens the well up, linear time may not be an accurate determinant of subsurface phenomena.
The results are designed to help operators make the critical decision: Should I drill this next well or not? In order to evaluate that decision properly, the economics at the moment of the decision is the main deciding factor. In other words, Well A and Well B have been producing for approximately 1 year, and a decision must be made whether an infill well should be placed between them now that the drainage areas are known. The decision to drill the next well must be made on the basis of the return on the capital expended to drill the well; therefore, NPV is the economic metric used in this paper. The cost of each well is incurred on the day the well begins producing, meaning that, once the first well in a section is drilled, the cost for the second well is incurred the moment the second well comes on line. All NPV values are based on a 20-year production life, so, despite the delayed onset of any individual well, the NPV associated with that well is still evaluated over a 20‑year period. The effect on the NPV of each drilling scheme for each type of completion has been analyzed to help operators maximize their investment.
Infill Timing Scheme
The infill-drilling case results in new wells being drilled between producing wells. In other words, Well 1, Well 3, and Well 5 (i.e., the first, third, and fifth wells) are drilled and hydraulically fractured simultaneously and then the two infill wells—Well 2 and Well 4—are drilled and fractured at a later date. From a modeling standpoint, this means the two infill wells—Well 2 and Well 4—will come on line once 12.5% of the estimated ultimate recovery (EUR) for each of the initial three wells is produced.
Recovery factor was not sensitive to timing. NPV provided far more detail regarding the viability of each successive well. Because the trends for each value of reservoir permeability were similar, one value of reservoir permeability for each type of drilling scheme was analyzed. For the 5-nd reservoir, infill wells resulted in significant NPV reduction somewhere between 37.5 and 62.5% of total cumulative production of the initial wells.
A similar trend exists for both the 50- and 500-nd cases. The magnitude of NPV loss is lower for the higher permeabilities because the NPV value of the infill wells reduces up to 70% for both the 50- and 500-nd cases compared with 98% for the 5-nd case.
Linear Timing Scheme
A linear timing scheme is also presented in which the first well is drilled, fractured, and put on production. The second well, and each successive well, is completed and put on production at an offset from the preceding well at a later date, as dictated by the infill timing. The same timing method was applied to the linear case, such that the second well comes on line at 12.5% of the EUR of the first well, the third well comes on line at 12.5% of the EUR of the second, and so on for each percentage of cumulative production.
The rate of NPV decay for the linear cases increases as the percentage of total cumulative production increases. For linear offset drilling, the risk of reducing NPV increases clearly as time progresses. In fact, a decrease of approximately 50% occurs when the initial well has produced approximately 62.5% of the total anticipated EUR. A 67% reduction can be seen when 87.5% of the total cumulative production has been reached.
The rate of NPV decay is larger for the 5-nd case and smaller for the 500-nd case. A maximum of 50% reduction exists for the 500-nd case, even at 87.5% total cumulative production. However, the 5-nd case exhibits a reduction of up to 95% for the 87.5%-of-total-cumulative-production case.
Hybrid Timing Scheme
A hybrid scheme was also examined in which the wells are placed with room for one infill between them and then the infill wells are drilled. In other words, Well 1 is drilled in the far left of the section block. Then, Well 3 is drilled in the middle of the section (the time at which Well 3 is drilled remains at 12.5% of the total cumulative production of Well 1, then 37.5%, and so on). Well 5 is the third well drilled, and it is placed at the far right of the section. Then, Well 2 becomes an infill well between Wells 1 and 3 and Well 4 is the last well drilled, placed between Wells 3 and 5. The hybrid scheme is merely a combination of the linear scheme and the infill scheme.
The NPV variation for the 500-nd cases for the hybrid scheme shows a decline in NPV for all wells, even Wells 3 and 5 (the second and third wells drilled, respectively), which appear to be in virgin reservoir. The reason for the NPV decrease between Well 1 and Well 3 was the addition of Well 2 at a later time. Well 2 robbed from Well 3 but not from Well 1 because Well 1 had sufficient time to drain its area fully, whereas Well 3 had not depleted its acreage fully.
The 5- and 50-nd cases exhibited the same trend.
The linear and hybrid schemes illustrated approximately the same loss in NPV. The trends for the infill-drilling scheme at the various permeabilities show that the NPV for the section drops at all reservoir-permeability values for greater than 62.5% of cumulative production. This drop can be attributed to the loss of NPV for the two interior wells.
This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 165722, “An Examination of Infill- and Offset-Drilling Patterns in Field Development for Shale Wells,” by Vivek Sahai, SPE, Greg Jackson, SPE, Farshad Lalehrokh, SPE, and Rakesh Rai, SPE, Weatherford, prepared for the 2013 SPE Eastern Regional Meeting, Pittsburgh, Pennsylvania, USA, 20–22 August.