CORE TESTS HELP PREVENT FORMATION DAMAGE IN HORIZONTAL WELLS

Tim Beatty, Barry Hebner, Randy Hiscock
PanCanadian Petroleum Ltd.
Calgary

D. Brant Bennion
Hycal Energy Research Laboratories Ltd.
Calgary

Simulating drilling mud invasion in cores helped determine and eliminate the damaging effects of various mud systems prior to the drilling of a horizontal well in Canada.

Formation damage during drilling has a propensity to cause severe productivity reductions in horizontal wells.

Formation damage is highly reservoir specific. General damage classifications according to rock and fluid type can be postulated, but specific tests are necessary to evaluate precisely the primary potential mechanisms of damage for each lithofacies.

A proper understanding of the mechanisms and severity of damage improves the selection of the optimum mud system (based on a weighted evaluation of economic, technical, operational, safety, and environmental concerns). The laboratory tests varied parameters such as mud base type (oil or water), overbalance pressure, solids content and size distribution, bridging agents, chemical adsorption, and physical incompatibility.

Horizontal wells have a much greater sensitivity to formation damage than do vertical wells in the same formation. Therefore, drilling fluids used in horizontal wells must be as nondamaging as possible, particularly in highly overbalanced operations.

In vertical well bores, localized damage induced by invading drilling fluid and solids can often be bypassed through cementing and perforating. Because of the high cost of perforating horizontal wells, cleanup must be achieved by extended flow or chemical stimulation techniques.

In long horizontal sections, reservoir drawdown may not be sufficient to remove formation damage. To this end, stimulation attempts can be costly, and in some cases, ineffective. If damage is severe, productivity may be uneconomic, and a viable drilling play could subsequently be abandoned. By conducting core displacement tests with various drilling fluids on representative reservoir samples, the least-damaging drilling fluid can be selected.

Proper testing and core preparation procedures must be followed to ensure that results are representative. This article outlines core displacement test methodology and presents the results of two core displacement studies. The findings of these studies are substantiated by actual production data. These results illustrate that although damage from a drilling fluid cannot be eliminated in horizontal drilling, it can be minimized.

FORMATION DAMAGE

Formation damage is any process that reduces the flow capacity of a zone bearing oil, water, or gas. Formation damage is a source of serious productivity reductions in many oil and gas reservoirs and causes water injectivity problems in many waterflood projects.1

Formation damage can occur when a nonequilibrium fluid or solid-bearing fluid enters a reservoir or when an equilibrium fluid is displaced at extreme velocities. Most processes used to drill, complete, or stimulate reservoirs have the potential to cause formation damage. These processes include drilling, cementing, completions and stimulation (perforating, acidizing, and fracturing), workovers (kill fluids and hot oil treatments), waterflooding or water disposal, enhanced oil recovery processes (miscible, chemical, and thermal flooding), and excessive injection or production rates.

Horizontal wells are much more susceptible to damage than vertical wells for a number of reasons, including the following:

Horizontal well bores have a substantially longer contact time with the drilling fluid than vertical wells. In a vertical well, the drilling fluid may only be in the pay zone for a matter of hours; in a horizontal well, the time may weeks.
Most horizontal wells are not cased and perforated but are completed open hole. Relatively shallow invasion damage, which would be easily perforated through in a standard conventionally cased vertical well, remains a major source of permeability reduction in many horizontal wells.
Uniform drawdowns are difficult to obtain in horizontal wells because of the length of the well in the pay zone. Except in selected zones, it is much more difficult to clean up damage from the invaded fluid or solids.
The physical mechanics of flow into horizontal wells are substantially different from those into vertical wells because vertical and horizontal permeabilities in most formations are not the same. An equivalent amount of damage causes substantially greater productivity impairment in horizontal wells than in vertical wells."

During drilling, the following mechanisms can cause severe productivity reductions in many horizontal wells:

Invasion of incompatible or damaging mud filtrates into the formation causing permeability impairment from clay swelling, clay deflocculation, or the formation of emulsions or insoluble precipitates
Fines migration induced by high fluid loss rates because of highly overbalanced drilling operations
Physical adsorption of polymer, inhibitors, or other mud additives
Adverse relative permeability effects associated with the permanent trapping of invaded water-based or oil-based mud filtrate in the near well bore region 5
The permanent entrainment of solids contained in the drilling mud.

The solids include mud additives such as barite, bentonite, artificial bridging agents, and microfines generated by the drill bit but not removed by the surface solids control equipment.

Proper simulation of drilling mud invasion into the formation allows an unbiased selection of the optimum drilling fluid to minimize formation damage in a horizontal well. Fluid rheology, solids content and size distribution, overbalance pressure, formation permeability, wettability, and other parameters can be taken into consideration for proper mud selection.

TEST APPARATUS

Fig. 1 is a schematic of the experimental apparatus used in the mud leak off tests. The core material is placed in a 3.81-cm ID flexible sleeve with a 0.635-cm wall thickness. The ductile sleeve allows a confining overburden pressure to be transferred to the core to simulate reservoir pressure.

The core, mounted within the lead sleeve, is placed inside a 7.5-cm ID steel core holder capable of simulating reservoir pressures up to 68.9 mPa. This desired overburden pressure is obtained by filling the annular space between the lead sleeve and the core holder with simulated formation water and then compressing the water with a hydraulic pump.

The ends of the core holder each contain two ports: one for fluid feed or production and the other for pressure measurement. The portion of the core holder directly adjacent to the production end of the core is equipped with a radial distribution plate to ensure evenly distributed fluid flow into and out of the core. The specially designed mud injection head of the core holder is equipped with large diameter tubing to allow easy passage of whole muds across the core face.

A large dead volume at the core face ensures that the mud undergoes continuous sheer as it flows past the sand face. The experimental system is designed to allow continuous circulation of fresh mud across the core face at any specified overbalance pressure.

The pressure differential is monitored with a Validyne pressure transducer. The transducer is mounted directly across the core and measures the pressure differential between the injection and production ends.

The pressure transducer has a range of 0-1,500 kPa and is accurate to 0.01% of the full scale value. The pressure transducer connects directly to a strip chart recorder which provides a continuous pressure profile of the test. A digital readout also appears on a multichannel Validyne terminal from which the test operator takes readings as a backup.

A Ruska displacement pump is used to inject fluids into the core. The pump can inject at rates from 1 to 8,200 cc/hr at pressures up to 68.9 mPa, with an accuracy of 0.01 cc. Back pressure on the system (for full reservoir condition tests) is controlled with a 316 stainless steel controlling back pressure regulator accurate to 0.5% of the set point value. This regulator allows for the smooth production of fluids from the system at any desired flow rate and set point pressure.

EXPERIMENTAL PROCEDURE

Although preserved core is preferred for any type of special core evaluation, in many cases this type of core is unavailable. Clean, extracted core is often used in its place. If the formation is homogeneous, 3.81-cm OD plugs are drilled from full diameter core. If the formation is heterogeneous, full diameter core is used with special cross flow heads that allow the fluid leak off to be parallel to the high permeability zones (fractures, vugs, etc.). Plugs are normally screened by air permeability measurements to select representative samples for testing.

Unoxidized, uncontaminated samples of reservoir crude oil are used for evaluating oil reservoir cores, and humidified nitrogen or methane gas is normally used to evaluate gas reservoir cores. Produced water or reconstituted equivalents are used to simulate the initial water saturation.

Wettability restoration is essential for nonpreserved samples because formation wettability can strongly influence mud performance with respect to fluid loss rates and phase trapping effects. The cleaned and extracted core samples are restored by vacuum saturation with formation brine and flooded with brine for 3-7 days to reach equilibrium. The brine is then slowly displaced with dead reservoir crude oil, after which fresh oil is slowly circulated through the core samples for 6 weeks at reservoir temperature and overburden pressure.

MUD LEAK OFF

The core samples are mounted in the equipment shown in Fig. 1. The samples are maintained at reservoir temperature, and net overburden pressure is applied to simulate the net pore pressure in the reservoir. The following procedure is then used for each sample:

Crude oil (or gas if a gas reservoir is evaluated) is displaced at a low injection rate to obtain an initial stabilized baseline oil permeability at Swi (flow direction No. 1)
In the opposite orientation (flow direction No. 2), the core is exposed to drilling mud at the specified hydrostatic overbalance pressure.
The total amount of fluid leak off is recorded and plotted against time to evaluate the effectiveness of the bridging agents present in the drilling fluid. Overbalance pressures are computed using mud density supplied by the mud manufacturers and current reservoir pressure data.
Overbalance pressure is maintained on the core until seal off or a stabilized leak off rate is achieved.
Oil flow is established in flow direction No. 1 to simulate production from the reservoir following drilling. The permeability impairment caused by mud invasion is then evaluated.
If desired, the cores then undergo petrographic evaluation (thin section, scanning electron microscope, and X-ray diffraction) to evaluate specific mechanisms of formation damage. Also, stimulation treatments (acid, oxidizing agents, solvents, etc.) can be simulated to test their effectiveness in removing damage--a useful technique for designing stimulation programs for wells already damaged.

UPPER MANNVILLE

The Countess upper Mannville "RR" pool is located in Section 35, Township 18, Range 16 W4M in Alberta. The pool has seven vertical producing oil wells and one horizontal oil well, 12D-35-18-16 W4M (Fig. 2). Discovered in 1989, the pool had an original oil in place of 1,132,500 cu m (7,123,400 bbl) with recoverable oil estimated at 566,260 cu m (50% recovery factor).

The horizontal well was drilled in July 1991 and is currently producing 77.4 cu m oil/day (487 bo/d), The well's cumulative production was 27,306 cu m of oil and 2,377 cu m of water as of Aug. 31, 1992.

The upper Mannville (Glauconitic) reservoir was deposited in a fluvial/estuarine setting dominated by tides and characterized by low sinuosity fluvial channels contained within the ancestral floodplain. The reservoir sand is typified by a sharp erosional contact with the underlying lower Mannville (Ostracod) deposits (Fig. 3). These sands contain high angle trough crossbedding with tidal mud couplets and chert-rich laminae concentrated along the foresets. The reservoir sand has been incised by a lower permeability upper Mannville channel resulting in an abrupt transition with the overlying upper Mannville sediments.

Grain size ranges from medium at the base of the unit to fine/medium at the top of the channel, giving rise to a slight overall fining upward sequence. The sand is moderately well sorted, with quartz and chert the predominant minerals (80%) and with minor percentages of feldspar (sublitharenite rock-type classification).

Clays are present in trace quantities. Kaolinite is the dominant clay type, and minor amounts of illite and chlorite also occur. Acidizing or fracturing this reservoir is unnecessary because of the high reservoir quality and the possible instigation of fines (kaolinite) migration resulting in a deterioration of permeability.

Reservoir quality is good to very good with porosities ranging from 18 to 24% and Kmax (air) from 250 to 5,200 md (Table 1). An upward-decreasing trend in vertical permeability is evidenced by the higher Kv/Kh ratios at the base of the core. This upward decrease in vertical permeability is attributed to the increased reservoir heterogeneity resulting from tidal couplets, chert-rich laminae, and calcareous cemented sands that are more prevalent near the top of the unit.

LAB TEST RESULTS

Nine mud systems (four oil based and five water based) were evaluated on nine separate restored Countess core samples. All tests were conducted at the reservoir temperature of 37 C. with 10,340 kPag of net overburden pressure. All the muds for this study were tested with the assumption of a constant overbalance pressure of 1,200 kPa for simplicity of comparison. The muds contained 3% pulverized rock flour (

A low viscosity CaCO3 325 grade invert mud had the best performance of the oilbased fluids. Because of reservoir pressure constraints, the invert systems lacked sufficient density for normal balanced drilling operations. Therefore, these muds were not actively considered for field application after the initial tests because of the high added solids content which would be required to increase the mud density.

A KCI polymer system and CaCO3 0 grade polymer system had the best performance of the water-based fluids. The KCI polymer fluid was selected as the best option for use in this horizontal well, based on economics, mud disposal requirements, and drilling considerations.

FIELD OPERATIONS

Based on offset well data, a fluid density of 1,080 kg/cu m would allow a 600 kPa overbalance pressure during the drilling of the horizontal well. Pressure data indicated that if virgin reservoir pressure were encountered, the mud density may have to be increased to 1,200 kg/cu m to maintain overbalance. To achieve this density, KCI and NaCI would have to be added. Because this horizontal well was at the start of the learning curve for Pan-Canadian Petroleum Ltd., underbalanced drilling was not considered here. The 200-mm diameter horizontal section was planned to be drilled at a flow rate of 1.2 cu m/min.

Two centrifuges and two flow line cleaners were used for solids control. The centrifuges were capable of handling a rate of 1.40 cu m/min. The flow line cleaners used 200-400 mesh screens.

Once the horizontal section was started, the well began to flow. To drill the well with 600 kPa overbalance, the mud weight was increased to 1,140 kg/cu m with KCI and NaCI. With the added salt, total solids rose from 4 to 9%. The 400-mesh screens on one flow line cleaner blinded frequently, resulting in whole mud being dumped into the sump. These screens were subsequently replaced with 200-mesh screens.

Table 3 lists the production history for the horizontal well; daily rates have averaged more than 75 cu m/day. Fig. 4 is a comparison of the horizontal well's production to the average production of the seven vertical producing wells in the field. On average, horizontal well production is approximately 400% greater than vertical well production at a comparable (20%) drawdown rate.

Although the water cut is gradually increasing in the horizontal well, it remains at less than 25% after more than 1 year of production. The advantage of low-damage horizontal drilling is evident from these production data.

CONSOLIDATED SAND

The formation name and drilling location of the well are confidential; however, the nature of the reservoir (low permeability consolidated sand) and its susceptibility to damage during horizontal drilling are relevant to this article.

The reservoir-quality sand was deposited in a marine environment with minor tidal influences. Bedding is massive with argillaceous partings in the form of tidal couplets. Reservoir sandstones occur as thin elongated sand bars. The individual bars are typically 1.5 km wide and 5.0 km long with an average thickness of 2.0 m. Economic viability of this thin pay interval requires horizontal drilling technology to enlarge the well bore drainage radius.

The reservoir is a homogeneous, very fine grained, moderate-to-well-sorted dolomitic subarkose. Detrital grains are composed predominantly of monocrystalline quartz with subordinate amounts of alkali feldspars and minor constituents of plagioclase feldspar, sedimentary rock fragments, and trace amounts of heavy minerals, mica, and phosphate. Diagenetic cements are characterized by early syntaxial quartz overgrowths and poikilotopic calcite which was subsequently replaced or partially replaced by ferroan dolomitic cements.

This reservoir is highly susceptible to formation damage during acid stimulations because of the significant amount of ferroan carbonate cements. A late-stage anhydrite cement occludes minor amounts of primary and secondary porosity created during dolomitization. Clay content is low, and the formation is not considered to be highly sensitive to freshwater-based drilling fluids.

Reservoir quality is considered average with overburden-corrected porosity averaging 13%. Matrix permeabilities range as high as 130 md and average overburden-corrected permeability (Kmax) is about 45-50 md (Table 4). Visual intergranular and moldic porosity display good interconnectedness in thin sections. Vertical permeability is adequate to drain an average 2-m thick pay zone and is not considered to be a critical factor in a horizontal well.

LAB TEST RESULTS

Table 5 summarizes the test results of eight mud formulations evaluated for this second horizontal well candidate. To simulate drilling microfines, 3% by mass of

Actual field overbalance pressures, based on mud densities, were used in this test series, except for mud NO. 8 which simulated drilling in a near-balanced condition with a low-solids invert. All tests were conducted at the reservoir temperature of 81 C. with 19,300 kpag of net confining overburden pressure.

Examination of the data led to the following interesting trends:

The reservoir is extremely susceptible to damage. Seven of the eight mud systems tested caused substantial damage, The major mechanism of the damage, based on post-test petrography and test performance, appears to be the permanent entrainment of mud solids and silicate microfines in the rock matrix.
Only the low-solids content (reduced to
A comparison of equivalent fluids using freshwater and 3% KCI also suggest some slight degree of freshwater sensitivity in the reservoir material.

Thus, the lower overbalance, low-solids content invert was selected as the optimum fluid for use in this horizontal well.

FIELD OPERATIONS

Based on offset well data, a fluid density of 920 kg/cu m would give a 300 kPa overbalance pressure on the formation. For well control, NaC1 would be added to the system rather than potentially damaging solids such as barite. Because of the volatile nature of this crude oil, underbalanced drilling was not considered. The 159-mm diameter horizontal section of the hole was planned to be drilled with a 0.7-0.9 cu m/min flow rate.

Two centrifuges and a flow line cleaner provided solids control. The centrifuges were capable of handling a flow rate of 1.4 cu m/min. The flow line cleaner used 200 mesh screens. If the solids control equipment could not maintain a density of 920 kg/cu m, the drilling operation would cease until that density were attained.

Approximately 500 m into the horizontal section, the well began to flow. A density of 1,000 kg/cu m was required to balance the formation pressure. NaC1 was added to the drilling fluid for well control purposes. Solids in the mud system increased from 4 to 9% after the salt additions.

This horizontal well is currently producing approximately 15 cu m/day of oil. Because of the low pay and permeability, vertical wells have never produced any significant oil.

ACKNOWLEDGMENT

The authors wish to thank PanCanadian Petroleum Ltd. for permission to publish this article.

REFERENCES

Bennion, D.B., Thomas, F.B., and Bennion, D.W., "Effective Laboratory Coreflood Tests to Evaluate and Minimize Formation Damage in Horizontal Wells," presented at the Third International Conference on Horizontal Well Technology, Houston, Nov. 12-14, 1991.
Babu, D.K., and Odeh, A.S. "Productivity of a Horizontal Well," Society of Petroleum Engineers Reservoir Engineering, November 1989, pp. 417-421.
Ertekin, T., Sung, W., and Schwerer, F.C., "Production Performance Analysis of Horizontal Drainage Wells for the Degasification of Coal Seams," Journal of Petroleum Technology, May 1988, pp. 625-632.
Joshi, S.D., "A Review of Horizontal Well and Drainhole Technology," SPE paper 16868 presented at the 1987 SPE Annual Technical Conference and Exhibition, Dallas, Sept. 27-30, 1987.
Bennion, D.B., Cimolai, M.P., Bietz, R.F., and Thomas, F.B., "Reductions in the Productivity of Oil & Gas Reservoirs Due to Aqueous Phase Trapping," paper presented at the CIM/44th Annual Technical Meeting of the Petroleum Society in Calgary, May 9-12.