There are several completion methods that can be used in horizontal applications. The two primary systems are pump down plugs ("plug and perf") with cemented casing or open-hole packers with ball seats. The simplest and most robust system is a cemented long string with pump down plugs. Benefits include providing a stable wellbore with a cemented casing string, controlling fracture initiation points via perforations, and ease of troubleshooting when/if problems arise during hydraulic fracturing. The two most significant drawbacks are increased completion time due to perforating and the risks associated with plug parts and sticking pipe during coil tubing drill-out. Open-hole packer systems generally provide less control on fracture placement but easier drill-out post-frac in comparison to the pump down method. Having no cement behind casing can cause potential issues with production casing from a long term standpoint, and there is little control for fracture initiation between stages. In a conventional or unconventional reservoir the fracturing fluid will seek the path of least resistance and stimulate that particular area. This is an issue since the goal of fracture stimulating a conventional reservoir is to stimulate the entire lateral. Trouble-shooting can also be a problem with the open-hole packer system due to the possibility of multiple issues being in play and the inability to differentiate between them. Benefits are speed of operation, since there is no perforating between stages, and reduction in debris during drill out. When choosing a system one must consider a variety of costs and benefits before making a decision. Currently the cemented casing/pump down plug method seems to provide the best cost/benefit approach for horizontal Cotton Valley completions.

Stimulation design can be a topic of much debate. To begin, let's try to stick with practical design parameters that are assumed to be true amongst the majority of industry professionals. Pumping rates directly affect fracture growth. This can be fracture height, fracture length, or a combination of the two. Higher viscosity fluid systems tend to create more fracture width and potentially height. The opposite applies for lower viscosity fluid systems. Proppant selection should include the consideration of closure stresses, and of course formation deliverability, which can be determined via basic reservoir models. These are fracture characteristics that do not take into account factors such as barriers or unconventional fracturing networks.

Now that we have discussed basic fracturing dynamics a quick historical look at previous fracturing designs in vertical wells is worth mentioning. The industry began fracture stimulation of the Cotton Valley reservoir in vertical wells in the Ark-La-Tex region several decades ago. If the mathematical fracture models utilized at that time were correct, then theoretically, the Cotton Valley would be significantly drained, making horizontal drilling a much less attractive or even uneconomic option. Apparently these vertical fracture designs yielded much more height growth and significantly less length than originally anticipated. The result is a smaller drainage radius on a per well basis. This brings us to stimulation design in horizontal wells. Since it appears likely that industry stimulation models for vertical Cotton Valley wells were extremely optimistic, one can assume modeling a horizontal completion is equally challenging. Fortunately the general assumptions for fracture dynamics stated earlier still apply. There is also reason to believe that horizontal wells create a natural environment that favors length growth versus height growth due to all of the perforations being placed within a 4-5 inch vertical area in the reservoir. Ultimately the only way to properly understand the dynamics of a fracture treatment is to utilize micro-seismic and general experience while comparing production in wells with similar reservoir mechanics.

Currently the majority of our Cotton Valley fracture treatments are hybrid (slickwater, linear and X-link) with 200' stages, 3-5 perf clusters per stage, and an adequate number of perfs per cluster depending on designed pump rate. Our rule of thumb is approximately 1.5 bpm of pump rate per perforation assuming a 0.42" hole size. Hybrid fracturing treatments theoretically allow one to capture the best of both worlds with respect to fracturing design. The Cotton Valley is a conventional reservoir with relatively low porosity and permeability, so the majority of our hybrid frac is slickwater. As our proppant concentration increases above 2ppg, linear and/or X-link systems are used with gel concentrations in the range of 20ppg - 25ppg.   Total proppant pumped ranges from 250,000#s - 500,000#s on a per stage basis, and average pumping rates range from 30 bpm - 60 bpm. Rate and total proppant pumped are a function of the fracture height needed to stimulate the reservoir. There is certainly potential for using just slickwater systems in the Cotton Valley, however, in our opinion one will generally tend to achieve flatter production curves and increase total well productivity with the use of hybrid designs (with higher sand concentrations at the wellbore) versus slickwater.

Once the stimulation phase is complete we move to the drill-out, which tends to present the most risk.   At its worst, this operation can potentially cause a loss of the entire well, loss of lateral section, or at a minimum, a very expensive fishing job. There are of course some risk mitigation tactics that can be used with coiled tubing, but Brammer has found that using a workover rig in certain plays (Eagle Ford and Cotton Valley) simplifies the drill-out, making "stuck pipe" and "friction lock" a non-issue. In normally pressured reservoirs as lateral lengths become longer, conventional clean out with a workover rig tends to become more attractive. The cost of a workover rig drill-out is less than, or equivalent to, a coil tubing cleanout, the primary drawback being the time required to rig up. We believe that this approach not only decreases costs, but greatly reduces the probability of sticking pipe during the drill-out phase of the completion.