New Drilling Technology

An effective approach to keeping the hole clean in high-angle wells

Applying weighted sweeps to keep the hole clean is poorly understood and often misapplied. Here is the best way to get good results

Mike Sewell, Newfield Exploration Co.; and Joe Billingsley, Baroid Drilling Fluids, Halliburton

Drilling a high-angle well is almost always accompanied by hole-cleaning issues. Traditional approaches such as high-viscosity sweeps, high rotary speeds, fast pump rates, frequent short trips and lubricants may buy time if the well reaches total depth (TD) quickly enough. However, in long, openhole intervals, hole cleaning should be effective, or the operator may face sidetracking the well.

Among operators who are drilling high-angle wells, the focus is shifting from old "cuttings removal" models to application of a sweep method that actually reaches and removes the silt bed accumulating on the low side of the hole. As early as 1986, hole-cleaning research indicated that turbulent flow produced by relatively thin drilling fluid is more effective at silt bed removal than a high-viscosity flow profile.¹ The silt bed, composed of formation sand and barite, can create the same torque, drag and overpull conditions associated with differentially stuck pipe – even in a cased hole.

Described here are: the basic application of weighted sweeps; limitations / applications of design models; and effects of viscosity, rotation speed and drillstring sliding. Basics of designing sweeps are presented, along with specific application recommendations.

"Jet Stream" Sweeps

Consistent results in silt-bed removal have been achieved with fully-circulated, low-viscosity, weighted sweeps which exceed drilling mud weight by 3 – 4 ppg and provide a 200 – 400-ft column in the annulus. The weighted sweep is built from the mud in use; and the yield point remains at its normal drilling value. Viscosity increase is minimal and results from barite additions, rather than from thickening the mud. If pumped regularly at the normal circulating rate – while rotating pipe at 80 rpm (minimum) – once the sweep reaches bottom and is allowed to return to surface, the weighted sweep can carry a significant, visible quantity of silt to the shakers and mud cleaner. Operators routinely using this method observe radically different "before and after" hole conditions.

For example, while planning a high-angle, West Cameron (Gulf of Mexico) well, Newfield Exploration Co. received a "torque prediction analysis" from the directional drilling service. For this well, where the angle would reach 55°+ with a 120° left-hand turn, torque at TD was expected to be 22,630 ft-lb. Instead, actual torque peaked at 15,000 ft-lb, Fig. 1. Weighted sweeps, rather than high-vis sweeps, were circulated throughout the drilling operation. No beads or lubricants were used. Hole cleaning issues were minimal.

Fig. 1. Weighted Sweep application on Gulf of Mexico well.

The end-of-well report included this directional driller’s description of conditions in the tortuous, final 1,300 ft of the 9-7/8-in. hole, from 9,900 to 11,200 ft (MD): "The assembly slid without any trouble, which I felt was unusual for the profile of this well."

Old Models Fall Short

Flow rate and fluid viscosity are usually evaluated first when a hole cleaning problem is suspected. While the flow profile is the basis for designing the right hole-cleaning program, many misapplications rely on inappropriate or incomplete modeling.

Many traditional models are based on an "average annular velocity," which implies a uniform flow regime. This applies to straight holes and vertical sections of directional holes, but does not apply from the kick-off point (KOP) down. Field evidence now demonstrates that flow profile deviation due to pipe eccentricity may be more important than other parameters typically considered for hole cleaning. Pipe eccentricity can markedly influence the calculated flow profile for a given point in the borehole, as shown in Fig. 2.

Fig. 2. Flow visualization for Herschel-Bulkley model (rotating) for pipe eccentricity of 0.05 (where 0.0 = Centered pipe; 1.0 = Drill pipe touching wellbore).

Power Law and Bingham Plastic models were necessary for oilfield applications when only slide rules and hand calculators were available. Computers now make it possible to use the more sophisticated Herschel-Bulkley model. Although more exact in its calculations, the latter does not yet account for drill-pipe rotation, making it the better choice for depicting the dead areas while slide drilling, Fig. 3. The Power Law model also does not account for rotation but, because it also does not address gel strengths or low-shear-rate stresses, it becomes more appropriate for analyzing rotating pipe.

Fig. 3. Herschel Bulkley vs Power Law for rotating and sliding scenarios.

Other Effects, Increased Viscosity

In conventional wells, barite "sag" began to occur when deviated wells exceeded 30°. Mud weight out was lighter than mud weight in, despite the absence of gas or water flow, which might account for density reduction. The corrective mud treatment required raising the low-shear-rate rheology (three- and six-rpm Fann viscometer readings), making the entire mud system thicker. The yield point also increased from the 6 to 10 lb/ft² range to as much as 30 lb/ft². Thicker fluids were perceived as cleaning the hole better; but exaggerated flow on the top side of the hole meant more sluggish fluid movement on the low side, and even faster silt / barite buildup.

While rotating, mud under the drill pipe is not quite stagnant, but is very sluggish, Fig. 4. The calculated flow profile in this figure, for 0.5 eccentricity, indicates a problem area under the drill pipe. Since fluid follows the path of least resistance, the thicker the fluid, the more pronounced the flow deformity. There is an optimum low-end rheological value – based on 6-rpm Fann readings – beyond which, hole cleaning becomes more difficult.

Fig. 4. How viscosity increase affects flow profile.

Weighted sweeps have an entirely different effect. The 2-D and 3-D flow profiles depicted here reveal an abnormal "jet-stream" flow in the annulus which is capable of removing fines from the low side of the hole. When drag from pipe rotation combines with gravity due to weight variance, i.e., a weighted sweep, this jet stream appears to flow through the narrow part of the annulus and scour the bottom side. Further, the increased mud weight provides more buoyancy, which helps lift fines and silts out of the hole. The success of the above mentioned case well supports this conclusion.

Rotation speed. Pipe rotation speed contributes to cuttings delivery. For a weighted sweep to be effective, the pipe should always be turning while the sweep circulates up the annulus. Increased rotation speed causes significant step increases at 100 – 120 rpm, and again at 150 – 180 rpm (Fig. 5); but there is a down side to relying on high pipe-rotation speeds.²

Fig. 5. Effects of pipe rotation speed.

Directional tools required for aggressive directional programs often limit the rotary to 60 – 80 rpm. The tools can malfunction at higher speeds and defeat efforts to achieve efficient hole cleaning. Further, increased pipe rotation speed may improve delivery of drilled cuttings to surface, but not necessarily remove or prevent fines accumulation on the low side of the hole. Pumping a weighted sweep achieves fines removal at the normal rpm range.

Sliding and the dead zone. Slide drilling can create "dead spots" where sludge and silt may build up. Since sliding is unavoidable in most high-angle wells, it is critical to understand the dynamics of silt-bed formation.

Slide drilling puts a large length of the drillstring in contact with the low side of the borehole. Even the most liberal flow profiles show small stagnant areas adjacent to the drill pipe / wellbore contact. During connections and trips, the stagnant mud pockets allow some barite or fines accumulation on the low side of the hole.

The revised Cuttings Export model – which provides optimized rheological and pipe-rotation parameters – calculates the ability of different sweeps to transport normal drilled cuttings, and rates the sweeps accordingly, Fig. 6. However, the ability to transport cuttings to surface does not, by itself, assure a clean borehole.

Fig. 6. Cuttings transport efficiency (Zero = Normal drilling).

On another Gulf well, the "drill ahead hydraulics" model predicted hole cleaning problems would occur while drilling 5,800 ft of 7-in. hole at a 40°+ angle. Weighted sweeps were pumped while sliding, but were not circulated out completely prior to surveys and short trips. When a short trip was attempted in 5,600 ft of open hole, it was necessary to backream. Multiple tight spots and packed-off areas were encountered. Some of these packoffs were attributed to a balled bottomhole assembly, causing operators to ignore the silt-bed buildup.

Another series of weighted sweeps was pumped, this time while rotating pipe, based on the model’s effectiveness indications. The sweeps were circulated out completely. The hole was cleaned, and drilling continued to TD with no further problems.

Designing a Weighted Sweep

A weighted sweep is usually 3 – 4-ppg higher than the mud weight and spans 200 – 400 ft in the annulus, contributing about 50 psi of hydrostatic pressure. The drill fluid must have good rheological and fluid-loss properties before the sweep is mixed and pumped, i.e., a loose mud with a high filtration rate will not sweep as well and may exacerbate any existing problem. The sweep must make it all the way to surface.

Normal to slightly decreased pump rates (200 gpm or less) are recommended for pumping weighted sweeps. As a rule of thumb, if the end of the sweep is less than 0.5-ppg over original mud weight (at the flowline), the sweep was probably too short or too light, or both. When building a weighted sweep with higher viscosity mud, a greater weight difference may be necessary to achieve results.

Weighted sweeps are often pumped as preventive maintenance, particularly in larger hole diameters, where annular velocity and flow profile are less than ideal for cuttings and silt-bed removal. The directional driller and drilling fluids engineer can coordinate timing of the sweep with sliding and rotating to ensure optimal results.

Shaker-screen mesh size influences how much of the removed silt bed is visible as the sweep crosses the shaker. A 210-mesh screen usually shows convincing results. On one of the first wells to experiment with weighted sweeps in an oil-based mud system, the first sweep returned 16 bbl of silt. Subsequent sweeps yielded 4 – 12 bbl per run. Drilling resumed with no further problems on the formerly high-torque, packed-off hole.

The following recommendations will assist in design and application of an effective weighted sweep. However, weighted sweeps will not always eliminate torque and drag in complicated wells. Proper well design and planning are the first defenses against a sidetrack scenario.

Recommendations

The following procedures are recommended for most-effective, weighted-sweep application:

• Calculate sweep height in the vertical part of the annulus, and added hydrostatic head in the annulus.
• Rotate at 80 rpm+ while pumping the sweep, 180 rpm+ is preferred. It is not recommended to pump a weighted sweep while sliding.
• Keep pumping until the sweep is out of the hole. If pumping stops while the sweep is in the deviated hole, it falls out of the "jet stream" and redeposits captured fines.
• Avoid increasing viscosity of the weighted sweep.
• Plan to run 3 – 4 sweeps to achieve desired results when hole-cleaning problems are indicated. One sweep will probably not be sufficient to remedy an existing problem.
• Avoid checking for flow or closing the annular with the weighted sweep in the drill pipe.
• If desired for preventive maintenance: 1) schedule sweeps every 6 – 8 hr or 500 – 1,000 ft (whichever comes first, unless the hole dictates greater frequency); barite / silt build-ups are dependent on time and footage; and 2) pump a sweep while circulating bottoms up before tripping, and another once back on bottom.

In summary, this innovative and effective practice challenges long-held beliefs about sweep viscosity and pipe rotation. Nonetheless, objections to weighted sweeps – that they will "knock the bottom out" or "take too much time" – will quickly subside as more high-angle wells are drilled with fewer sidetracks.  

Literature Cited

¹ Azar, J., A. Pilehvari and S. Shizari, "State of the art cuttings transport in horizontal wellbores," SPE Drilling and Completion, Sept. 1999, 14:3, pp. 196 – 199.

² Krepp, A., "Hole cleaning in ER wells: It’s not the mud man’s job," Drilling & Well Performance Technology Newsletter, 4th Quarter 1999, pp. 6 – 7.

The authors
	Michael Ray Sewell is a Drilling Engineer for Newfield Exploration Co. One of the founders of the company, he has over 26 years’ experience in the industry, including 12 years with Tenneco as a drilling / production / reservoir engineer. He holds a BS in PE from the University of Missouri, 1975, and is a member of SPE and SPWLA.
	Joe Billingsley is a Senior Technical Advisor for Baroid, a division of Halliburton. He holds a BS in physics from Texas A&M University and has over 22 years’ experience in the industry. Mainly, he focuses on trouble-shooting problem wells, as well as international assignments. He is a member of SPE, API and AADE.