Highly engineered cemented liner completions outproduce openhole offsets by a factor of 1.8 to 1, across a range of reservoir quality.

Operators have looked to the improved productivity that horizontal wells promise in low-permeability formations to offset the added costs of drilling and completion. However, they cannot always expect the increased formation exposure provided by a horizontal wellbore to replace the benefits of a stimulation treatment. Only a few of the horizontal drilling programs undertaken on the assumption that a long, openhole lateral can effectively eliminate the need to case, cement and stimulate a well have achieved their goal.
A case history of a multiyear horizontal completion project in West Texas revealed the false economy of not cementing casing in horizontal wells drilled to low-permeability formations. This project, which compared the production performance of cemented and noncemented wells drilled across three reservoir quality categories, revealed the increase in initial well cost was more than justified by dramatic improvement in cash flow, and that the better the reservoir quality, the more the improvement.
Background
The study area is in southwestern Midland County, Texas, in the Permian Basin. The producing interval is the Devonian Thirtyone formation, at a depth of about 12,000 ft (3,660 m). The pay zone lithology is a dense cryptocrystalline limestone with matrix microporosity of 3% to 5% resulting from digenetic alteration. Gross pay intervals range from 200 ft to 300 ft (61 m to 92 m), with net pay averaging from 30 ft to 60 ft (9 m to 18 m). Very little natural fracturing exists in the Thirtyone formation, and permeability is on the average of 0.045 md, making it a prime candidate for hydraulic fracturing.
The study compared 20 wells with conventional openhole-liner completions (OHLC wells) and nine wells with cemented liners (CLC wells). Researchers grouped wells of both types for comparison according to reservoir quality: infield wells had the highest permeability and porosity, flank wells were intermediate, and edge wells had the lowest permeability and porosity (Figure 1).
Many of the early horizontal wells in the area were re-entry recompletions of poorly performing vertical wells. Only those wells with larger-diameter vertical cased sections - 7-in. casing to about 11,000 ft (3,355 m) - that could accommodate larger horizontal laterals were included in the analysis. The horizontal laterals for these wells varied from 2,800 ft to 3,600 ft (854 m to 1,098 m) and typically were a 6.125-in. hole with a 4.5-in., 11.6-lb liner.
Operators initially oriented the horizontal wellbores of the wells drilled and evaluated in this study perpendicular to the expected natural fracture orientation. However, as drilling expanded, no significant natural fracturing was observed, and operators considered other factors in selecting a desired orientation. Because the reservoir appeared to maintain the greatest continuity and lateral extent along depositional strike, a wellbore aligned parallel to depositional strike was considered to improve the likelihood of fracture stimulation effectiveness by increasing the opportunity for hydraulic fractures to be transverse to the borehole. Additionally, this allows the bit to more easily remain within the targeted pay interval and contact more reservoir rock. Operators drilled the CLC wells parallel to strike, while the OHLC wells included parallel-to-strike and perpendicular-to-strike orientations.
Openhole completion procedure
Operators stimulated initial openhole completions in the main field area using a common manifold arrangement of 28% HCl with friction reducer and alternated stages of titanate-crosslinked fracturing fluid. Equal volumes of acid and frac fluid were pumped with gross acid volume calculated on 60 gal/ft of horizontal section. A typical job totaled 216,000 gal of acid. Tubing was run to the end of the lateral, and fluids were pumped down the tubing and tubing-casing annulus simultaneously. Annular injection rates averaged 60 bbl/min to 80 bbl/min at 3,500 psi to 4,500 psi. Tubing injection rates were 15 bbl/min to 20 bbl/min at 7,000 psi to 8,000 psi. Post-treatment logs indicated stimulation was limited to the heel and toe areas. Satisfactory initial production rates (3 MMscf/d to 4 MMscf/d) declined to about 1 MMscf/d during 3 to 6 months. Hole enlargement in the heel area typically eliminated access to the horizontal section once the tubing was pulled into the vertical section.
Operators in the flank and edge well areas attempted to improve results and reduce costs using an uncemented tubing string with predrilled ported sections for the stimulation. After fracturing, they pulled the tubing and produced the wells openhole. Other efforts involved reducing the acid concentration and varying the volumes pumped, and using branched, openhole, multiple laterals of various orientations.
Completion design
To effectively acid fracture long horizontal intervals, some mechanism must divert the stimulation fluid to specific points along the exposed formation. While cementing is a consistently reliable method, it can negatively impact completion effectiveness if it is slow to dissolve and inhibits fracture initiation. To avoid this, operators used a nonstandard acid-soluble cement (ASC) in the wells with cemented liners. This cement dissolves at a much faster rate, creating an area of communication in the annulus immediately adjacent to the perforations. The addition of carbonate weighting material achieves increased solubility. Foaming agents also can help increase the yield (reduce cost) and reduce the density of this weighted cement formulation. The carbonate weighting material can be formulated with a variable mesh size distribution to help bridge off fractures and reduce cement losses into high-permeability secondary porosity features.
Operators selected perforating and acid fracturing intervals based on mud log and porosity log information. Single-trip, multiple-zone, tubing-conveyed perforating (TCP) was used. About 12 zones were shot simultaneously per TCP trip into the average 3,600-ft (1,098-m) lateral section. Halliburton's Point Source Cluster Perforating (PSCP) method enabled even distribution of the treatment fluids between zones. Fewer perforations are made close to the lateral's heel section, and perfs are added to successively deeper clusters of perfs approaching the toe, where the differential pressure is lowest.
An in-situ crosslinked acid (ICA) system stimulated the CLC wells. The ICA system involves the pumping of a thin-gelled (1.5% to 2% vol polymer) 17% to 20% HCl acid base designed to leak into the formation matrix along the fracture face, creating wormholes. When it reaches a pH of 2.5 to 3.5, the fluid rapidly viscosifies, halting leak-off and preserving live acid for reaction further down the fracture. The fluid uncrosslinks and reverts to original viscosity at a pH of about 3.5 to 4. The result is longer etched lengths and deeper live acid penetration.
Operators stimulated the CLC wells with a gross acid volume of about 45 gal/ft to 50 gal/ft of lateral length. The initial acid stage was ungelled 20% HCl followed by a water-based, cross-linked, gelled fracturing fluid. The third stage was the ICA system, followed by 20% HCl to enhance near-wellbore fracture conductivity and enlarge pathways created by the previous stages. The fifth stage was an overflush.
Acid fracture treatments were pumped at 110 bbl/min to 120 bbl/min. Early treatments tagged with multiple radioactive isotopes indicated the desired fluid distributions were being obtained, and in subsequent wells this practice was discontinued.
Reservoir quality
For each of the reservoir quality categories, conventional OHLC wells were compared with CLC method wells based on the cash flow generated by an average well performance profile. Because both gas and condensate were produced, the economic comparisons were based on barrels of oil equivalent (Table 1).
Analysts normalized the production profile of each well within a group to a common starting point and averaged it to create a single production profile for each completion type and reservoir quality category. The well production history lengths varied, ranging from about 10 months to 30 months. Most of the wells were drilled during 1997 to 2001. The analysts then generated a simple discounted cash flow stream. Due to the variation in lease net revenue interests, operating costs and gas contracts, no attempt was made to compare the wells on a rate-of-return basis.
Despite a US $400,000 increase in initial capital costs related to the more sophisticated CLC employing the ASC, PSCP and ICA completion protocols, the CLC wells outperformed the OHLC wells on a cash flow basis (Figure 2). The relative improvement also increased as the quality of the reservoir (and the formation productivity) increased. The wells completed in the highest quality area realized an incremental benefit of nearly $8 million, discounted, after only 2 years of production, based on $25 oil. In the poorest reservoir quality area, where neither OHLC nor CLC wells reached a positive cash flow after 2 years, the incremental benefit was still more than $500,000. While the average CLC well in the poorest quality area would reach positive discounted cash flow after 27 months, the OHLC well never would within the 48-month period modeled.
While the OHLC approach costs less initially, the method severely limits stimulation effectiveness. The limitation increases with increasing reservoir quality. Effective stimulation should be a primary completion goal, even when it requires more intensive engineering control during completion.
Editor's note: This article is based on SPE paper 76725, presented at the Society of Petroleum Engineers' Western Regional/American Association of Petroleum Geologists Pacific Section Joint Meeting in Anchorage, Alaska, May 20-22, 2002.

Comments