ABSTRACT

For reasons that are either incidental and/or accidental to the operation of gas transmission pipelines, transients do commonly arise. In fact, steady state operation is a rarity in practice. The equations governing the behavior of these transients constitute a system of nonlinear conservative hyperbolic equations, which are difficult to solve. For numerical expediency, attempts are often made to linearize these equations by dropping the most troublesome terms in the momentum equation. Among these, the one that is most often neglected is the kinetic energy term. A robust numerical algorithm and a computer code have been developed without neglecting any term. In the present study, eight field examples of engineering interest are simulated to provide some understanding of the behavior of gas pipeline transients under operational scenarios. The first two examples are real field cases for which measured data are available. The first pertains to the propagation of a fast transient in a 24-inch, 300-ft. long pipe; the second involves the propagation of a slow transient, with 24-hour cycle, in a 45mile long, S-inch transmission pipeline. Comparisons between the predicted results and the measured data are very good and superior to the predictions reported in the literature. The last six examples are hypothetical field cases for the same long transmission pipeline subjected to different transient initiators. Two of these examples concern line packing while the last four relate to pipeline rupture.

INTRODUCTION

Design and cost-effective operation of a gas transmission pipeline requires accounting for its response under transient load. Actual operations invariably encounter transient states. The loss of a compressor, the addition or loss of supply or sale points, pipe leak and/or rupture, and variable demand are a few of the initiators of line transients. Under isothermal conditions, the continuity and momentum equations, together with the equation of state, constitute the governing equations describing transient flow in natural gas pipelines. The assumptions usually made include isothermal flow, applicability of steady-state friction and negligible wall expansion or contraction under pressure loads. In simulating transient flow of single-phase natural gas in pipelines, most of the previous investigators neglected the inertia term in the momentum equation. This renders the resulting set of partial differential equations linear. Numerical methods previously used to solve this system of partial differential equations include the method of characteristics (MOC) and a variety of explicit and implicit finite difference schemes [1, 2, 3, 6, 93. Neglecting the inertia term in the momentum equation will definitely result in loss of accuracy of the simulation results. In order to compensate for the absence of the inertia term in the momentum equation, Yow introduced the concept of "inertia multiplier" to partially account for the effect of the inertia term in the momentum equation[4, 51. Wylie, et al. [8] simulated transients in natural gas pipelines in accordance with the concept of "inertia multiplier". Rachford and DuPont [ll] demonstrated that calculations based on the concept of "inertia multiplier" sometimes yield very

This content is only available via PDF.
You can access this article if you purchase or spend a download.