SPE Production & Operations
Volume 23, Number 2, May 2008, 268-279

SPE-101502-PA

On Instability in Gas Lift Wells and Schemes for Stabilization by Automatic Control

View full textPDF ( 2,375 KB )

DOI  More information 10.2118/101502-PA http://dx.doi.org/10.2118/101502-PA

Citation

  • Eikrem, G.O., Aamo, O.M., and Foss, B.A. 2008. On Instability in Gas Lift Wells and Schemes for Stabilization by Automatic Control. SPE Prod & Oper23 (2): 268-279. SPE-101502-PA.

     

Discipline Categories

  • 5 Production and Operations

Summary

In this paper, we present a simple nonlinear dynamic model that is shown to capture the essential dynamics of the casing-heading instability in gas lift wells despite the complex nature of two-phase flow. Using the model, stability maps are generated showing regions of stable and unstable settings for the production valve governing the flow of produced oil and gas from the tubing. Optimal steady-state production is shown to lie well within the unstable region, corresponding to an oil-production rate that cannot be sustained without automatic control. Three simple control structures are suggested that successfully stabilize the casing-heading instability in simulations, and more importantly in laboratory experiments.

Introduction

Artificial lift is a common technique to increase tail-end production from mature fields, and injection of gas is among the most widely used methods. Gas is injected into the tubing as deep as possible and mixes with the fluid from the reservoir (see Fig. 1). Because the gas has lower density than the reservoir fluid, the density of the fluid in the tubing and, consequently, the downhole pressure decrease. As the downhole pressure decreases, the production from the reservoir increases. The lift gas is routed from the surface and into the annulus, which is the volume between the casing and the tubing, and enters the tubing through a valve, or an injection orifice. Backflow from the tubing into the annulus is not permitted by this valve. Gas lift can induce severe production flow oscillations because of casing-heading instability, a phenomenon that originates from dynamic interaction between injection gas in the casing and the multiphase fluid in the tubing. The fluctuating flow typically has an oscillation period of a few hours and is distinctly different from short-term oscillations caused by hydrodynamic slugging. The casing-heading instability introduces two production-related challenges. Average production is decreased as compared to a stable-flow regime and the highly oscillatory flow puts strain on downstream equipment. Reports from industry as well as academia suggest that automatic control (feedback control) is a powerful tool to eliminate casing-heading instability and increase production from gas lift wells (Kinderen and Dunham 1998; Jansen et al. 1999; Dalsmo et al. 2002; Boisard et al. 2002; Hu and Golan 2003; Eikrem et al. 2006; Aamo et al. 2005).

Understanding and predicting conditions under which a gas lift well will exhibit flow instability is important in every production-planning situation. This problem has been adressed by several authors by constructing stability maps, [i.e., a 2D diagram that shows the regions of stable and unstable production of a well (Eikrem et al. 2006; Poblano et al. 2005; Fairuzov et al. 2004)]. The axes may define the operating conditions in terms of gas-injection rate and production-choke opening or wellhead pressure.

In this paper, we present three different control structures for stabilizing casing-heading instability in gas lift wells. Stability is analyzed for each controller, and it is shown how feedback control stabilizes performance, at least locally, around some operating point. The performance of the controllers is demonstrated in simulations, but more importantly, stabilization is also achieved in laboratory experiments.

The paper is organized as follows: A desription of the laboratory facilities that are used in this work is given. Thereafter, the dynamics of casing-heading instability are discussed, and suitable models for analysis and design are proposed. The proposed control structures are presented along with stability analysis, closed-loop simulations, and experimental results. The paper ends with conclusions.

View full textPDF ( 2,375 KB )

History

  • Original manuscript received: 4 January 2006
  • Revised manuscript received: 14 November 2006
  • Manuscript approved: 2 July 2007
  • Version of record: 20 May 2008