Design of gain scheduling control strategy for artificial gas lift in oil production through Modified Relay Feedback Test

Huda Hussein, Ahmed Al Durra, Igor Boiko

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

One of the major problems encountered in artificial gas lifting is the stability of production from wells were gas lift may show a highly oscillatory behavior, and hence production varies greatly with time. This paper presents methodology of design of a gain scheduling strategy based on a new physical model of artificial gas lift in oil wells. The model was preliminarily validated by comparing the states dynamics with other models from the literature which have been replicated. The Modified Relay-Feedback-Test (MRFT) method was used to calculate the PI controller parameters, and the model reached steady state for different operating points. Gain scheduling was then performed following the MRFT results to provide a simple PI controller tuning parameters for the automatic control and optimization of the production from the oil well, and to compensate for the casing-heading instability in the system.

Original languageBritish English
Title of host publicationVSS 2014 - Proceedings of 13th International Workshop on Variable Structure Systems
PublisherIEEE Computer Society
ISBN (Print)9781479955664
DOIs
StatePublished - 2014
Event13th International Workshop on Variable Structure Systems, VSS 2014 - Nantes, France
Duration: 29 Jun 20142 Jul 2014

Publication series

NameProceedings of IEEE Workshop on Applications of Computer Vision
ISSN (Print)2158-3978
ISSN (Electronic)2158-3986

Conference

Conference13th International Workshop on Variable Structure Systems, VSS 2014
Country/TerritoryFrance
CityNantes
Period29/06/142/07/14

Fingerprint

Dive into the research topics of 'Design of gain scheduling control strategy for artificial gas lift in oil production through Modified Relay Feedback Test'. Together they form a unique fingerprint.

Cite this