Reliability-Based Design And Application Of Drilling Tubulars

Abstract

This paper describes a reliability based design of drilling casing and tubing in the load and resistance factor design format. The approach is based on the fundamental principles of limit state design. The paper identifies the limit states of pipe performance in well applications, discusses stochastic modeling of the load and resistance variables, and describes calibration of the design check equations. In the calibration analysis, uncertainties in the various design variables, e.g., kick load intensity and steel mechanical properties, are determined from analysis of the field and laboratory data and represented by appropriate statistical distributions. The reliability based design procedures are complemented by the pipe specifications and quality assurance procedures, also described in the paper. Application of the load and resistance factor design of casing is illustrated by an example problem. BACKGROUND Oil country tubular goods (OCTG) are subjected to a variety of loads during their service lives. These loads originate from various operations, e.g., running, cementing or producing, and accidental conditions such as the kick, or lost returns. Variability of the drilling tubular's strength and loads is well recognized [1, 2]1. The strength uncertainty, for example, arises due to the inherent variability of material properties, workmanship, and handling of tubulars during installation. The load uncertainty is associated with a designer's inability to estimate loads precisely. The objective of design is to estimate the "minimum" strength and the "maximum" load over the life of a tubular and make sure that they are separated by an adequate margin. Traditional design utilizes experience based safety factors, along with a characteristic set of loads and strengths, to assure the safety of tubular design. These procedures work well when a large database of experience supports the design. However, outside of the historical experience, or when new materials and novel applications are considered, for example, deep sour wells, the basis of judgment essential to establish a safety factor is lacking. As a result, the safety of these designs cannot be assured to a known extent. A number of other issues question the validity of the traditional approach to produce optimum casing and tubing designs. Exploration and production well designs consider a distinctly different knowledge base for their load estimation. The consequences of failure of as-intended performance are also different for various applications of a given casing or tubing string. Obviously, the safety margin requirements should also vary. Also, the traditional safety factor approach considers the stress at a selected point to be the appropriate design criterion. However, it can be shown that in well design, as indeed in many other structural applications, capacity of a structural member is better characterized by its gross or total failure behavior, rather than the stress at a selected point. This is due to the fact that load bearing capacity and stress are often not linearly related.

Numerical and theoretical analysis of burst pressures for casings with eccentric wear

“…The ABAQUS software (6.10) is capable of simulating these nonlinear processes. Therefore, ABAQUS [9] was used to perform finite element analysis to investigate the ultimate burst strength of a finite, threedimensional walled casing that contained geometrical imperfections. It was assumed that the deformed casing was imperfect under internal pressure and that the cross section remained eccentric.…”

“…The plastic zone continues to spread until the outer fibres of the casing yield. Experiments by Gulati[9] show that, under excessive internal pressure, the ultimate burst capacity of a casing is approximately 2%-3% larger than the yield strength of the outer fibre of the casing.…”

Analysis of Ignorance Factors in Design Criteria Subject to Model Uncertainty

“…The sensitivity of the collapse failure mode to imperfections, especially when the onset of plasticity precedes instability, makes it difficult to "predict" collapse loads. In a major development of reliability-based criteria for casing and tubing pipes (Gulati et al, 1994), the ultimate capacity of moderately thick and thick tubes loaded by externa! pressure is, therefore, calibrated based on the results of well executed experimenta.…”

“…The effect of iit depends critically on the variability of the "loading term" 6p in (18). Severalload domains were considered in the study (Gulati et al, 1994) as part of the overall zonation scheme of the design set. As an example, we only consider salt loading (direct contact with a fiowing salt formation), which imposes the most severe collapse load on casing.…”