Predicting the Build/Drop Tendency of Rotary Drilling Assemblies
- P.N. Jogi (Anadrill-Schlumberger) | T.M. Burgess (Anadrill-Schlumberger) | J.P. Bowling (Anadrill-Schlumberger)
- Document ID
- Society of Petroleum Engineers
- SPE Drilling Engineering
- Publication Date
- June 1988
- Document Type
- Journal Paper
- 177 - 186
- 1988. Society of Petroleum Engineers
- 1.6.6 Directional Drilling, 1.10 Drilling Equipment, 1.6.1 Drilling Operation Management, 1.4.1 BHA Design, 1.6 Drilling Operations, 1.1.6 Hole Openers & Under-reamers, 1.10.1 Drill string components and drilling tools (tubulars, jars, subs, stabilisers, reamers, etc), 4.3.4 Scale, 1.12.1 Measurement While Drilling
- 5 in the last 30 days
- 482 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 10.00|
|SPE Non-Member Price:||USD 30.00|
Today, the majority of rotary bottomhole assemblies (BHA's) for directional control are designed through practical experience and trial and error. This approach can produce satisfactory results when a great deal of local experience can be drawn on. It can prove costly, however, during drilling in a new area because of the increased number of trips and correction runs. This paper demonstrates how a BHA model can be used to predict the directional inclination tendencies of rotary assemblies, thus limiting the uncertainty associated with the traditional BHA design techniques.
The technique is demonstrated on data from 17 bit runs from three wells on the same platform in the Gulf of Mexico. Predicted tendencies from BHA descriptions alone proved to be accurate (to an error of ±0.1°/100 ft [±0.03°/10 m]) in more than half the cases. The uncertainty of other predictions appeared to depend on the hole gauge. The distance taken for a BHA to reach a stable build/drop rate after the start of a bit run depends on the length of the BHA. This factor must be taken in account in the prediction of BHA performance.
Factors that determine the behavior of a BHA have been the subject of many papers over the last 30 years.1-7 It is generally recognized that the following factors are the most important: (1) stabilizer location and gauge; (2) drill-collar stiffness; (3) borehole inclination and curvature; (4) weight on bit (WOB); (5) hole size; (6) rotary speed; (7) bit sidecutting action; (8) formation strength; and(9) formation anisotropy.
Most mathematical models are two dimensional (2D) and static.8,9 They attempt to predict the side forces at the bit and stabilizers by assuming that the BHA deforms like an elastic beam. The side force at the bit is then used to predict the build/drop tendency of the BHA. These models take Factors 1 through 4 into account. They are often analytic and tend to run in a matter of seconds on small computers.
Although 2D models have led to significant improvements in BHA design, large discrepancies frequently occur between predictions and field results. These errors are a result of Factors 5 through 9 and the three-dimensional (3D) nature of the wellbore. Attempts are sometimes made to correct the predictions with adjustable parameters as follows. Field results are first compared with predicted results. An input parameter to the model (often associated with the formation) is then varied until agreement is reached. We prefer not to use this approach unless the cause of the error is clearly identified and can be quantified or measured in real time. A parameter associated with the formation anisotropy cannot account for changes in BHA behavior if the cause of the problem is hole enlargement.
Some advancement has been made by adding the effects of 3D curvature and the dynamic effects of rotary speed.10 These models tend to be solved numerically by finite-element techniques. Because of their size and complexity, they tend to run in a matter of minutes or even hours on large computers. As such, they tend to be more suited to a research environment than to rig-site utilization where an interpretive approach and fast turnaround are required for decision-making.
The model described in this paper is a compromise between the two extremes. It is a 3D static analysis of the BHA. Because it is 3D, it can take into account complex wellbore geometry and can model the effect of downhole torque on the BHA. Geometry and torque are two of the factors that contribute to the turn of a BHA. The model is solved analytically and thus runs in a few seconds on a rig-site computer. This means that it can be used efficiently as both a design and an interpretive tool. The solution is based on the general theory of bending and twisting of elastic rods (see the Appendix).
The model provides a prediction of the side force at the bit and stabilizers, the deformed shape of the BHA, and the stable or equilibrium tendency of the BHA. The first result is used to predict the qualitative tendency (i.e., the build/drop) of the BHA for a specified wellbore geometry. It cannot be used to quantify the initial tendency reliably. Stabilizer side forces can be used to predict which sides will be subject to most wear.
The second result is used to monitor which collars are subject to large bending stresses and to determine which collars touch the wellbore. This is important in measurement-while-drilling (MWD) operations, when wellbore contact can result in excessive wear of resistivity rubbers and large stresses can result in premature failure of MWD components and drill collars.
The third result predicts the final or asymptotic behavior (build/drop raft and turn rate) of the BHA. This prediction is based on the assumption that the BHA will tend toward drilling a hole curvature such that the side force at the bit is zero. Deviations from this curvature will create side forces that will tend to push the bit back to the stable position.
All directly measurable factors (Factors 1 through 4), except rotary speed (Factor 6), are accounted for in the model. Caliper information is never available in real time; thus, in the absence of information on hole washout, the hole size, dh, is normally assumed to be equal to the bit diameter. The side-cutting action of the bit and the rock strength are omitted. We are not aware of any reliable model for these effects, despite the fact that the side-cutting action could be measured in the laboratory and rock strength can be inferred in real time from the inverse rate of penetration (ROP). For completeness, rock anisotropy has been included in the model by use of an anisotropy index described by Lubinski and Woods2 or by assuming that the anisotropy creates a moment or couple at the bit.7 In practice, we believe that because no measurements of anisotropy or bending moment are currently available with MWD tools, it is best to ignore rock anisotropy.
The theory behind the model is outlined in the Appendix. The next section describes some of the most important theoretical predictions and shows how sensitive the model is to input parameters like WOB, hole size, and stabilizer clearance. Then the model is compared with MWD directional data coming from 17 bit runs from three wells on the same platform in the gulf coast. Only the build/drop tendencies of the BHA's are analyzed at this stage. The response is also compared with hole-size information measured by caliper logs after drilling.
|File Size||6 MB||Number of Pages||10|