api rp 11v5 final draft

Page Proof 3: June 12, 2008

Recommended Practices for Operation, Maintenance, Surveillance, and Troubleshooting of Gas-lift Installations

FT API RECOMMENDED PRACTICE 11V5

THIRD EDITION, JUNE 2008

DRA

DRAFT



Recommended Practices forOperation, Maintenance,Surveillance, and Troubleshootingof Gas-lift Installations

Upstream Segment

FT API RECOMMENDED PRACTICE 11V5THIRD EDITION, JUNE 2008

DRA

DRAFT

Special Notes

API publications necessarily address problems of a general nature. With respect to particular circumstances, local,state, and federal laws and regulations should be reviewed.

Neither API nor any of API's employees, subcontractors, consultants, committees, or other assignees make anywarranty or representation, either express or implied, with respect to the accuracy, completeness, or usefulness ofthe information contained herein, or assume any liability or responsibility for any use, or the results of such use, of anyinformation or process disclosed in this publication. Neither API nor any of API's employees, subcontractors,consultants, or other assignees represent that use of this publication would not infringe upon privately owned rights.

Classified areas may vary depending on the location, conditions, equipment, and substances involved in any givensituation. Users of this recommended practice should consult with the appropriate authorities having jurisdiction.

Users of this recommended practice should not rely exclusively on the information contained in this document. Soundbusiness, scientific, engineering, and safety judgment should be used in employing the information contained herein.

API is not undertaking to meet the duties of employers, manufacturers, or suppliers to warn and properly train andequip their employees, and others exposed, concerning health and safety risks and precautions, nor undertaking theirobligations to comply with authorities having jurisdiction.

Information concerning safety and health risks and proper precautions with respect to particular materials andconditions should be obtained from the employer, the manufacturer or supplier of that material, or the material safetydata sheet.

API publications may be used by anyone desiring to do so. Every effort has been made by the Institute to assure theaccuracy and reliability of the data contained in them; however, the Institute makes no representation, warranty, orguarantee in connection with this publication and hereby expressly disclaims any liability or responsibility for loss ordamage resulting from its use or for the violation of any authorities having jurisdiction with which this publication mayconflict.

API publications are published to facilitate the broad availability of proven, sound engineering and operatingpractices. These publications are not intended to obviate the need for applying sound engineering judgmentregarding when and where these publications should be utilized. The formulation and publication of API publicationsis not intended in any way to inhibit anyone from using any other practices.

Any manufacturer marking equipment or materials in conformance with the marking requirements of an API standardis solely responsible for complying with all the applicable requirements of that standard. API does not represent,warrant, or guarantee that such products do in fact conform to the applicable API standard.

All rights reserved. No part of this work may be reproduced, stored in a retrieval system, or transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior written permission from the publisher. Contact the Publisher, API

Publishing Services, 1220 L Street, N.W., Washington, D.C. 20005.

Copyright © 2008 American Petroleum Institute


DRAFT

Foreword

This document is under the jurisdiction of the API Committee on Standardization of Production Equipment(Committee 11).

This document presents recommended practices for the operation, maintenance, surveillance, and troubleshooting ofgas-lift systems. Other API specifications, API recommended practices, and Gas Processors Suppliers Association(GPSA) documents are referenced and should be used for assistance in design and operation.

Nothing contained in any API publication is to be construed as granting any right, by implication or otherwise, for themanufacture, sale, or use of any method, apparatus, or product covered by letters patent. Neither should anythingcontained in the publication be construed as insuring anyone against liability for infringement of letters patent.

This document was produced under API standardization procedures that ensure appropriate notification andparticipation in the developmental process and is designated as an API standard. Questions concerning theinterpretation of the content of this publication or comments and questions concerning the procedures under whichthis publication was developed should be directed in writing to the Director of Standards, American PetroleumInstitute, 1220 L Street, N.W., Washington, D.C. 20005. Requests for permission to reproduce or translate all or anypart of the material published herein should also be addressed to the director.

Generally, API standards are reviewed and revised, reaffirmed, or withdrawn at least every five years. A one-timeextension of up to two years may be added to this review cycle. Status of the publication can be ascertained from theAPI Standards Department, telephone (202) 682-8000. A catalog of API publications and materials is publishedannually and updated quarterly by API, 1220 L Street, N.W., Washington, D.C. 20005.

Suggested revisions are invited and should be submitted to the Standards Department, API, 1220 L Street, NW,Washington, D.C. 20005, [email protected].

iii


DRAFT


Contents

Page


DRAFT

1 Gas-lift Operating System Components and Potential Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Gas-lift System Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Gas-lift System Operating Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3 Surface Facility Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.4 Metering and Control Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.5 Gas-lift Valve Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.6 Well Equipment Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51.7 Gathering System Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.8 Well Testing Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.9 Production Handling Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.10 Information Handling Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.11 Surveillance and Control Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2 Gas-lift Operating Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.1 Under-lifted and Over-lifted Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.2 Ineffective Gas Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142.3 Unstable Gas-lift Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 Types and Causes of Unstable Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.5 Other Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3 Surface Gas-lift Compression, Dehydration, and Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.1 Compression Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.2 Gas Dehydration Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.3 Gas-lift Distribution System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4 Gas Injection Metering and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274.1 Gas Metering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274.2 Injection Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

5 Gas-lift Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325.1 Unloading Valves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325.2 Operating Valve(s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

6 Well Equipment—Tubulars, Completion, and Wellhead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356.1 Casing Annulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366.2 Tubing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376.3 Completion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386.4 Wellhead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406.5 Wellhead Monitoring and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

7 Gathering System—Flowline and Manifold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437.1 Flowline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437.2 Manifold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

v


DRAFT

Page

8 Well Production Rate Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478.1 Well Test Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478.2 Well Test Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528.3 Well Test Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

9 Production Handling System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569.1 Oil Handling System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569.2 Water Handling System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579.3 Gas Handling System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

10 Guidelines for Collecting and Using Operating Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5710.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5710.1 Well Test Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5710.2 Downtime Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6110.3 Pressure and Temperature Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6110.4 Injection Pressure and Rate Measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6610.5 Wellhead Production Pressure, Temperature, and Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

11 Manual and Automated Well Operation and Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6811.0 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6811.1 Manual Operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6811.2 Automated Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

12 Procedures for Initial Unloading and Kick Off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7112.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7112.1 General Unloading Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7112.2 Unloading Continuous Gas-lift Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7212.3 Restarting (Kick Off) Continuous Gas-lift Wells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7412.4 Unloading Intermittent Gas-lift Wells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7512.5 Restarting (Kick Off) Intermittent Gas-lift Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

13 Procedures for Adjusting (Fine Tuning) Gas-lift Injection Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7613.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7613.1 Continuous Gas-lift Wells with Steady Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7613.2 Continuous Gas-lift Wells with Variable Injection Pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7713.3 Intermittent Wells with Time Cycle Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7713.4 Intermittent Wells with Choke Control7913.5 Do Not Use Flowline Chokes79

14 Gas-lift Troubleshooting Tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8014.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8014.1 Two-pen Pressure Charts, or Equivalent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8014.2 Acoustical Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10114.3 Tagging Fluid Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10314.4 Flowing Pressure Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

15 Recommended Practices for Dealing with Wells That Produce Sand. . . . . . . . . . . . . . . . . . . . . . . . . . . . 10915.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10915.1 Recommended Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

16 Typical Locations of Gas-lift Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11016.0 Purpose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11016.1 Gas-lift Injection or Inlet Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11016.2 Gas-lift Production or Outlet Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11216.3 Downhole Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Page


DRAFT

17 Possible Causes and Cures of Common Malfunctions of Gas-lift Systems . . . . . . . . . . . . . . . . . . . . . . 117

18 Gas-lift Troubleshooting Checklist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Figures1 Gas-lift System Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Problem of Injecting Gas Through an Upper Gas-lift Valve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Problem of Under Injection—Injecting Too Little Gas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Problem of Over Injection—Injecting Too Much Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Determining the Optimum Gas Injection Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Problem of High Injection Pressure Causing an Upper Valve to Open. . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Problem of Under Injection—Causing Operating Valve to Flood. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 Hydrate Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Water Content vs. Temperature and Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2510 Finger Distribution System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2611 Looped Distribution System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2612 Unloading a Continuous Gas-lift Well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7413 Adjustable Choke for Continuous Gas-lift Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7714 Adjustable Choke with a Pressure Regulator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7815 Installation of Pressure Measurement Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8116 Continuous Gas-lift—Good Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8217 Continuous Gas-lift—Wellhead Backpressure Too High . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8218 Intermittent Continuous Gas-lift—Good Operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8319 Comparison of Intermittent and Continuous Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8320 No Gas Injection Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8421 Continuous Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8422 Continuous Injection—Frozen Gas Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8523 Well is Flowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8524 Continuous Injection—Kick Off After Idle Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8625 Flowing Well—Loads Up Periodically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8626 Continuous Gas-lift—Normal Backpressure Higher Than Test Backpressure . . . . . . . . . . . . . . . . . . . . . 8727 Continuous Injection—Well Shut In. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8728 Continuous Injection—Well is Heading Periodically. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8829 Continuous Injection—Well is Unloading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8930 Intermittent Injection—Varying Injection Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9031 Intermittent Injection—Varying the Injection Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9132 Intermittent Injection—Injection Pressure Too High . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9233 Intermittent Injection—Well Loading Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9334 Intermittent Injection—Well Choked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9435 Intermittent Injection—Well Has Leaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9536 Intermittent Injection—Well Has Tubing Leak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9637 Intermittent Injection—Well Has Large Tubing Leak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9738 Intermittent Injection—Gas Injection Pressure Too Low . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9839 Intermittent Injection—Plugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9940 Intermittent Injection—Injection Rate Too Low . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10041 Typical Acoustical Recording. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10142 Typical Acoustical Recording. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10243 Flowing Pressure Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10644 Results of Flowing Pressure and Temperature Surveys Conducted During Intermittent

Operation of High Capacity Well to Locate Operating Valve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10845 The Gas-lift System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

DRAFT

Introduction

These recommended practices are offered to assist gas-lift system operators, analysts, technicians, engineers, andothers in understanding how to effectively plan, operate, maintain, troubleshoot, and provide surveillance of gas-liftsystems and gas-lift wells.

This document may be used in a gas-lift training course or as reference material. It can be obtained in booklet form asan API publication, or on CD ROM or cassette in Adobe PDF format.

These recommended practices discuss continuous gas-lift with injection in the casing/tubing annulus and productionup the tubing. Annular flow gas-lift (injection down the tubing and production up the annulus), dual gas-lift (two tubingstrings in the same casing), and intermittent gas-lift are mentioned; however, most of the discussion focuses on“conventional” continuous gas-lift. Many of the recommended practices in this document may be pertinent to the otherforms of gas-lift, but they should be considered and used with caution. Other recommended practices will addressdual gas-lift (API 11V9) and intermittent gas-lift (API 11V10).

This document includes:

— Gas-lift Operating System Components and Potential Problems.

Sections 1 through 11 describe the several components of an operating gas-lift system and discuss a number ofproblems that may be encountered and must be addressed to operate a gas-lift system effectively and efficiently.These sections are new to this edition of the document. A comprehensive checklist of system components isprovided and associated problems are discussed. The list can be used when troubleshooting or de-bottleneckinga gas-lift system.

These sections are recommended for use as:

— part of a training course dealing with gas-lift system operation;

— a review before beginning a major gas-lift system study;

— a review before designing and/or modelling a gas-lift system;

— a review before trying to troubleshoot difficult gas-lift system problems.

— Recommended Practices for Gas-lift Operation, Maintenance, Surveillance, and Troubleshooting.

Sections 12 through 17 are revisions/upgrades of information that has been in existence since the first edition ofthis document. These sections contain recommended practices for common gas-lift operations:

— initial unloading of the completion or workover fluid from the annulus of the gas-lift well;

— re-starting or kick off after a period of downtime;

— adjusting or fine-tuning the gas injection rate for optimum operation.

These sections discuss commonly used gas-lift troubleshooting tools. They conclude with sections that reviewthe potential locations of gas-lift problems, a table of possible causes and cures of some common gas-lift systemproblems, and a troubleshooting checklist.

These sections are recommended for use as:

— part of a training course dealing with gas-lift system operation;

— part of a training course dealing with gas-lift system maintenance;

— a review before trying to troubleshoot a difficult gas-lift operating problem.



DRAFT

Recommended Practices for Operation, Maintenance, Surveillance, and Troubleshooting of Gas-lift Installations

1 Gas-lift Operating System Components and Potential Problems 1.0 Purpose

Each gas-lift system consists of many components that must be understood, operated, and maintained. Thesecomponents must function together for the system to be efficient and economic in artificially lifting wells.

This document discusses the system components and presents a summary of recommended practices forrecognizing and dealing with many of the problems that may cause upsets and/or inefficient operations in continuousgas-lift. The checklist below may be used as a quick reference guide for spotting and addressing common problemsthat affect continuous gas-lift. This document does not address intermittent gas-lift; it is being addressed in API11V10.

1.1 Gas-lift System Components

The primary components of the gas-lift production system are:

— surface gas-lift compression, dehydration, and distribution system;

— gas injection metering and control equipment;

— gas-lift valves;

— well equipment—tubulars, completion, and wellhead;

— gathering system—flowline and manifold;

— well production rate testing facility;

— production handling system.

Each of these components is shown on Figure 1 and is discussed in the sections that follow, along with specificrecommended practices to reduce inefficiencies and unstable operations.

1.2 Gas-lift System Operating Problems

These are gas-lift system-wide operating problems and recommended practices to address them. They are cross-referenced to the section where they are discussed.

a) Wells are being under-lifted (lifting too shallow, too little gas), see 2.1:

— provide monitoring and control equipment and procedures;

— compare actual vs. design/optimum gas-lift performance;

— assure gas-lift valves and other equipment are working properly;

— allocate the available supply of gas to the most profitable wells in an optimal way when the supply is limited.

1

2 API RECOMMENDED PRACTICE 11V5


DRAFT

b) Wells are being over-lifted (too much gas), see 2.1:

— provide monitoring and control equipment and procedures;

— compare actual vs. design/optimum gas-lift performance;

— reduce compression or sell excess gas; do not try to inject all of it if overall gas supply is excessive.

c) Gas-lift distribution is ineffective (lack of good distribution control), see 2.2:

— evaluate overall system constraints, including gas-lift compression, distribution, wells, and handling systems;

— distribute or allocate available gas to each well by considering both the gas distribution system and the wells.

Figure 1—Gas-lift System Components

RECOMMENDED PRACTICES FOR OPERATION, MAINTENANCE, SURVEILLANCE, AND TROUBLESHOOTING OF GAS-LIFT INSTALLATIONS 3


DRAFT

d) Gas-lift wells are unstable (injection and/or production heading), see 2.3 and Section 4:

— provide monitoring equipment to detect unstable operation;

— determine the cause(s) of the instability;

— eliminate or reduce injection heading; it is very inefficient.

e) Wells have other problems (equipment, instrumentation, etc.), see 2.5:

— consider all components of the gas-lift system;

— monitor and calibrate all system components;

— establish and practice quality assurance on equipment selection and installation.

1.3 Surface Facility Problems

Typical problems with the surface gas-lift compression, dehydration, and distribution system, and recommendedpractices to deal with them are as follows.

a) Compression problems, see 3.1:

— maintain the compression facility to provide a consistent supply of lift gas at a stable pressure;

— monitor the compressor output and adjust the injection rates to optimally balance the demand for gas intothe wells with the supply of gas from the compressor plant or other source(s);

— perform routine compressor preventive maintenance to maximize system availability.

b) Dehydration problems, see 3.2:

— dehydrate gas-lift gas to less than 7 lb (3.175 kg) of water per million cubic ft (28,317 cubic m) of gas to avoidhydrate formation. The recommended water content for cold climate operations is 3 lb (1.36 kg) of water permillion cubic ft (28,317 cubic m);

— design the system to avoid the need for large pressure drops. API recommends that, for the sizing of pipesreceiving gas-lifted production, a “surge factor” of 40 % to 50 % be added to the estimated steady-state flowrate, compared to 20 % for naturally flowing wells;

— purge liquid from gas-lift distribution lines periodically;

— install insulation to reduce ambient temperature effects on freezing problems.

c) Distribution system problems, see 3.3:

— use a “finger” style gas-lift distribution system (see Figure 8);

— make the volume of the system as large as economically feasible;

— avoid combining continuous and intermittent lift in the same gas-lift distribution system unless automaticcontrol is used;

— measure the total gas into each group of wells to assist with allocation and troubleshooting.



DRAFT

1.4 Metering and Control Problems

Typical problems with the gas-lift injection metering and control equipment, and recommended practices to deal withthem are as follows.

a) Gas injection metering problems, see 4.1:

— use properly installed, ranged, well maintained, and accurately calibrated meters;

— measure injection pressure at the wellhead, downstream of any obstacles;

— pay special attention to gas measurement accuracy during well tests and pressure surveys.

b) Gas injection control problems, see 4.2:

— operate close to the design conditions of the gas-lift installation;

— redesign and re-install the gas-lift valves if the well's conditions have changed enough to disturb effectiveoperation;

— use a properly ranged gas flow controller to provide consistent, stable flow.

1.5 Gas-lift Valve Problems

Typical problems with the unloading and operating gas-lift valves, and recommended practices to deal with them areas follows.

a) Unloading valve problems, see 5.1:

— check the well performance of each well to detect heading or valve operating problems;

— use small-ported unloading valves to facilitate deeper valve transfer and prevent over injection at uppervalves;

— or, use downstream chokes in unloading valves to minimize throttling, avoid over injection, and minimizedamage to the valve port and seat during the unloading process;

— follow the unloading procedures in Section 12 to work down to the desired operating valve and avoid valvedamage during unloading operations.

b) Operating valve problems, see 5.2:

— use an orifice rather than a gas-lift valve in high productivity index (PI) wells [more than 0.5 B/D/psi (.012 m3/D/kPa)] to prevent throttling and permit a wider range of gas injection rates;

— use a gas-lift valve in low PI wells to prevent over injection and low casing gas pressure, which causeshydrates at chokes or excessive gas withdrawal from the distribution pipeline;

— use a choke downstream of the orifice to protect the orifice and avoid over injection;

— redesign and re-install the operating valve/orifice if well conditions change enough to disturb effectiveoperation.



DRAFT

1.6 Well Equipment Problems

Typical problems with gas-lift well equipment such as tubulars, completion equipment, and wellheads, andrecommended practices to deal with them are as follows.

a) Casing annulus problems, see 6.1:

— use a casing scraper during workovers to clean debris from the casing wall;

— circulate fluid to insure that the annulus is clean;

— hydrostatically test to gas-lift injection pressure to assure casing integrity;

— avoid attempting to lift two well completions in one casing annulus. (That is, if possible, avoid dual gas-lift. Ifdual gas-lift is required, refer to API 11V9.)

b) Tubing problems, see 6.2:

— circulate fluid to clean the tubing of corrosion products, scale depositions, paraffin, and asphaltine. If tubinghas excessive deposits, pull and replace it with a clean string;

— remove any unnecessary obstacles, such as unnecessary safety valves;

— use a mechanical set permanent or retrievable packer that holds in both directions;

— use an on-off tool with a profile nipple that permits an X-plug with equalizing prong to be set in the profile.The plug can be set when the tubing and valves must be removed; this eliminates kill fluid damage to thereservoir formation.

c) Completion problems, see 6.3:

— run flowing bottomhole pressure (FBHP) and static bottomhole pressure (SBHP) surveys at least annually toassess and track well performance and reservoir productivity;

— stimulate a well if its productivity becomes impaired and a pressure build-up test indicates that skin may bethe problem;

— minimize pressure surges and heading in a well that has sand producing reservoir rock or a sand controlscreen or gravel pack.

d) Wellhead problems, see 6.4:

— minimize flow restrictions such as bends, choke bodies, etc.;

— provide safe and easy access for wireline work.

e) Wellhead monitoring and control problems, see 6.5:

— measure the wellhead production pressure on a consistent basis;

— consider continuous measurement of the well's production rate using multiphase metering or production flowrate estimating technology.



DRAFT

1.7 Gathering System Problems

Typical problems with gathering system equipment, including the flowlines and manifold, and recommended practicesto deal with them are as follows.

a) Flowline problems, see 7.1:

— keep the flowline clean and avoid unnecessary restrictions;

— use appropriate treatments if needed to remove scale deposits, paraffin, etc.;

— avoid using one flowline for more than one well; this is to minimize excessive backpressure and avoidcomplications in monitoring the production of each well.

b) Manifold problems, see 7.2:

— minimize any unnecessary restrictions or pressure losses;

— keep all manifold valves fully open or fully closed;

— check the manifold for leaking valves using sonic or infrared detectors.

1.8 Well Testing Problems

Typical problems with well production rate testing equipment, and recommended practices to deal with them are asfollows.

a) Well test scheduling problems, see 8.1:

— test each well often enough to detect changes in performance;

— test each well long enough to obtain accurate results;

— co-ordinate well testing with other activities such as pressure surveys;

— use automatic or semi-automatic well testing to improve testing accuracy and reduce testing labor.

b) Test separation problems, see 8.2:

— purge time is required to thoroughly flush the previous well's production from the system;

— maintain the test separator backpressure consistent with the production system pressure;

— check and calibrate the well test meters.

c) Test measurement problems, see 8.3:

— make well testing a high priority so it receives the attention it deserves.



DRAFT

1.9 Production Handling Problems

Typical problems with production handling equipment, and recommended practices to deal with them are as follows.

a) Oil handling problems, see 9.1:

— measure total oil produced from each set of wells for comparison with estimates and for allocation andtroubleshooting purposes.

b) Water handling problems, see 9.2:

— measure or estimate total water production from each set of wells for comparisons and for allocation andtroubleshooting purposes.

c) Gas handling problems, see 9.3:

— measure gas production from each set of wells for comparisons and for allocation and troubleshootingpurposes.

1.10 Information Handling Problems

Typical problems with information handling, and recommended practices to deal with them are as follows.

a) Well test information problems, see 10.1:

— detect and evaluate “good,” vs. “questionable,” vs. “bad” well tests;

— re-test wells if the data is questionable;

— fix the problem that creates a bad test and then re-test the well;

— use the good well tests to evaluate actual performance and allocate gas-lift production.

b) Downtime information problems, see 10.2:

— detect and account for all downtime;

— use this data to allocate production to the wells, calculate production deferment due to downtime, andprioritize remedial work;

— keep unplanned downtime to a minimum;

— use well downtime plus facility downtime to develop an operating factor for groups of lifted and naturallyflowing wells. The operating factor (fraction of online time) multiplied by daily oil production gives the averagecalendar-day contribution from each well group.

c) Pressure and temperature survey information problems, see 10.3:

— obtain pressure surveys annually, or when conditions change;

— follow the procedures in Section 10 to obtain accurate results;

— obtain a pressure build-up if the inflow performance has changed.



DRAFT

d) Gas injection pressure and rate measurement information problems, see 10.4:

— monitor and continuously record the injection pressure and rate;

— gather this information during well tests, pressure surveys, and unloading;

— use this and related information to perform gas-lift surveillance and build accurate gas-lift models.

e) Wellhead (production) pressure and temperature information problems, see 10.5:

— monitor and continuously record the production pressure;

— gather this information during well tests, pressure surveys, and unloading.

1.11 Surveillance and Control Problems

Typical problems with the gas-lift surveillance and control system, and recommended practices to deal with them areas follows.

a) Manual operating problems, see 11.1:

— provide competent, dedicated people;

— provide on-going training in all aspects of gas-lift operation;

— provide quality measurement and control equipment;

— perform periodic system reviews to identify bottlenecks and opportunities;

— become familiar with the following checklist for use if problems develop in gas-lift wells.

1) is the master valve open?

2) is the wing valve open?

3) is the operating gas-lift valve open?

4) is the downhole safety valve open?

5) is the well's flow rate slugging?

6) is there a choke in the flowline?

7) is the gas-lift injection rate correct?

8) is the gas-lift injection pressure reasonable?

9) is the gas-lift injection pressure heading or surging?

10) is the flowing wellhead pressure reasonable?

11) is the flowing wellhead temperature reasonable?

12) is the separator pressure reasonable?

13) is hydrate forming at chokes or low spots in the piping?

14) is well test data (oil/water/gas) reasonable and reliable?



DRAFT

b) Production automation, see 11.2:

— use production automation to improve operational effectiveness.

2 Gas-lift Operating Problems2.0 Purpose

A gas-lift system is intended to operate with a stable supply of lift gas and the gas should be delivered to the wells ata relatively constant, known pressure and rate. All other components in the system; the gas-lift valves, tubulars,wellheads, flowlines, and separation and treating facilities, must be sized properly for stable operation.

If one or more of the system components are not working as intended, one or more problems can arise:

— wells may be “under-lifted,” resulting in lost or deferred production;

— wells may be “over-lifted,” resulting in excess consumption of gas and/or lost production;

— lift gas may be ineffectively distributed, resulting in sub-optimum system performance, even if all other parts ofthe system are designed adequately;

— wells may be unstable; this can lead to a large number of problems.

2.1 Under-lifted and Over-lifted Wells

A gas-lift well is “under-lifted” when the depth of gas injection is shallower than necessary for optimum lift and/or therate of gas injection is too low. Both of these conditions lead to ineffective lift gas utilization and lower than desiredproduction.

A gas-lift well is “over-lifted” when the rate of gas injection is too high and/or the gas injection pressure is too high.These conditions lead to ineffective lift gas utilization. Less than optimum gas-lift exists when a relatively largeincrease in injection gas results in a small gain or even a loss in production. The objective is to use the total availableinjection gas in an optimum way to maximize the total oil production rate.

When the compressor capacity exceeds optimum injection gas required for the group of wells, the criteria of flowstability can be applied. Stability requires a velocity of about 6 ft/s (1.83 m/s) in the gas-lifted mixture in the tubingstring above the injection point.

2.1.1 Recommended Practices

The following practices are recommended for under-lifted and over-lifted wells.

— Monitor the performance of each gas-lift well using flowing and static pressure surveys, in conjunction withaccurate well test and injection information, to determine (or estimate) the current producing gradients, depth(s)of gas-lift injection, flowing bottomhole pressure(s), and inflow performance. Use this data to create a calibratedcomputer model for use in surveillance and troubleshooting. Apply the same procedures to natural flow wells thatwill eventually be artificially lifted.

— Use a calibrated computer model, with current operating conditions, to detect a well that is being under-lifted, andto predict the optimum gas-lift injection rate, pressure, depth, and expected increase in production from solvingthe problem.

— Maintain the desired gas-lift injection control with a gas measurement system and an automated control valve tomaintain the desired injection rate and pressure.



DRAFT

— Find over-lifted and under-lifted wells using flowing gradient surveys and production tests that include injectiongas measurement. Gas saved in one group of wells may be profitably used in other wells.

— Over-inject in some wells (if necessary on a temporarily basis) to obtain stability by achieving a mixture velocity ofapproximately 6 ft/s (1.83 m/s) above the point of injection. If this condition persists however, the well should bere-designed to overcome the instability problem. Most gas-lift computer models will calculate the mixture velocity.

— Evaluate the effects on surface facilities caused by over-injection in wells, if this practice is used to maintainstability.

— Assure that the gas-lift valves are properly constructed, assembled, and set before they are installed. Use gas-liftvalves manufactured to API 11V1 specifications (or ISO 17078-2 when it is published). Use API 11V7 for repairand reassembly (or ISO 17078-2 when it is published).

— Check and probe test individual valves to assure conformance with design specifications for critical applications,or where there are on-going problems. See API 11V2, Gas-lift Valve Testing and Modeling. This document will bereplaced by ISO 17078-2 when it is published.

— Identify wells that are being under-lifted and check for an inadequate supply of gas due to pipeline restrictions orlimited compressor capacity. Evaluate the wells and establish a production priority and/or optimum gas-liftallocation. (See API 11V8.)

— Determine the frequency of compressor shutdowns. The wells may not be able to work down and maintain stable,deep injection if there are frequent interruptions. It may be necessary to redesign the unloading valves with smallerports or chokes and adjust the gas injection control. Compressor downtime should not be greater than 5 %.

2.1.2 Well Under-lifted—Injection Too Shallow

For a well to be lifted properly, the gas must be injected at the design (optimum) injection depth. The desired point of liftis usually within three joints [about 100 ft (30.48 m)] above the production packer, but the actual point of lift is based on:

— depth of the well;

— reservoir pressure;

— well's productivity (inflow performance);

— available gas-lift system pressure.

If the injection is occurring through an upper unloading valve or valves, or through a hole or leak, the well's FBHP cannot be drawn down to the desired value, and the production rate will be less than intended or desired (see Figure 2).This condition can be caused by a number of problems including:

— lower than desired injection gas pressure;

— improper gas-lift mandrel spacing;

— valves that are not assembled or set properly; valves that are installed in the wrong mandrels; use of the wrongtypes of valves, or ports that are too large;

— leak in an upper unloading valve, upper mandrel, tubing, or connection;

— well not properly unloaded; it never reached the intended operating valve;

— other problems that do not allow the well to work down to its design injection depth are restrictions at the surface,inadequate injection control, and frequent shutdowns.



DRAFT

2.1.3 Well Under-lifted—Injection Rate Too Low

Continuous gas-lift design specifies the injection rate needed to achieve a desired flowing pressure gradient from thedesign point of gas entry to the surface. If the gas injection rate is too low, the gradient will be too heavy and the wellwill not be drawn down to the desired flowing bottomhole pressure, even if injection is at the desired operating valveor orifice. This may result in the well producing less than the potential rate (see Figure 3).

NOTE By injecting through an upper gas-lift valve, even with the same amount of gas, the operating bottomhole pressure isincreased and the production rate is reduced.

Figure 2—Problem of Injecting Gas Through an Upper Gas-lift Valve

NOTE By not injecting enough gas, even through the same valve, the operating bottomhole pressure is increased and theproduction rate is reduced.

Figure 3—Problem of Under Injection—Injecting Too Little Gas



DRAFT

This condition can be caused by:

— rate of gas available to the well is less than design;

— gas composition/density is significantly higher or lower than design specific gravity;

— injection gas pressure available to the well is lower than design;

— flow restrictions at the surface;

— control choke or valve at the surface is improperly set and the injection rate is less than design;

— hydrates forming in the choke or gas line serving the well;

— the gas-lift valve port size, or choke size, is too small to accommodate the desired gas injection rate;

— increased water cut requiring additional injection gas to maintain the desired pressure gradient.

2.1.4 Well Over-lifted—Injection Rate Too High

A too high gas-lift injection rate should be avoided for practical reasons, as follows.

— May allow injection gas to enter an upper valve.

Excess gas may raise casing pressure above the set point for upper unloading valves, causing one to open withloss of efficiency and production rate.

— May lose production.

Excess gas can also increase friction in the surface flowline and tubing. This can cause a higher FBHP resultingin less drawdown and a lower production rate. The gas rate can become so great that the effect of friction lossexceeds the reduction in fluid density and the pressure gradient actually begins to increase, the FBHP rises, andthe liquid rate decreases. When gas is injected into a continuous gas-lift well, it mixes with the liquid in theproduction stream (usually the tubing). Gas is less dense than liquid, thus it lightens the pressure gradient(reduces the mixture density), allowing the well to produce with lower production pressure drops. As gas injectionis increased, the production pressure continues to drop, up to a point. This point is called the minimum pressuregradient (see Figure 4). The pointed labelled “Ideal Operation” may be considered to be on the minimum practicalpressure gradient. The injection gas rate for a continuous gas-lift well should normally be substantially less thanthe rate required to reach the minimum gradient.

— May increase costs.

The cost for compressing, dehydrating, distributing, injecting, recovering, and treating gas rises as the rate of gasinjection increases; and usually the production rate also increases. At some point, usually well before theminimum gradient is reached, the incremental cost of injection becomes equal to and then exceeds theincremental value of the increased production. This is called the economic limit or “point of diminishing returns”(see Figure 5). Even before this point, the incremental profit may fall below an acceptable limit. From aneconomic point of view, a well should not be operated beyond the point of acceptable incremental profit (see API11V8, Section 7).

2.1.5 How to Determine Optimum Gas-lift Injection Rate

The optimum gas-lift injection rate is related to the economic limit. This is usually substantially less than the minimumgradient which can be modelled with a validated computer program, as discussed in API 11V8, Section 4 andSection 8. Determining the economic limit requires reliable cost and revenue data.



DRAFT NOTE By injecting too much gas, the gradient is made heavier, which results in higher operating bottomhole pressure and areduced production rate.

Figure 4—Problem of Over Injection—Injecting Too Much Gas

NOTE The optimum production rate occurs where the incremental cost of injection is equal to the incremental value ofproduction.

Beyond this point, very little value (Δp) is gained for a significant increase in cost (Δi).

Figure 5—Determining the Optimum Gas Injection Rate



DRAFT

2.1.6 Well Over-lifted—Injection Pressure Too High

The surface gas-lift injection control valve is a flow rate control device, either automatic or manual; the rate iscontrolled by adjusting the opening.

If the gas rate is measured and regulated with a meter, flow controller, and automated control valve, a uniform ratecan be achieved. The position (opening) of the valve can continually change to maintain a desired injection rate.

A control problem can occur in gas-lift wells. The gas is primarily controlled on the surface. However, it is also(secondarily) controlled by the downhole operating valve or orifice. If more gas is injected at the surface than can behandled by the downhole valve or orifice, the pressure upstream of the gas-lift valves, in the casing annulus, must riseas the volume of gas in the annulus increases. As this pressure rises, the opening “forces” on the unloading valvesincrease. When one of these unloading valves (usually the deepest in the well) opens, gas from the annulus will enterthe tubing at this point (see Figure 6). This unloading valve becomes the “operating” valve, or the well has multipointinjection. This is inefficient and can lead to an unstable or heading condition that can cause several other problems.Unstable operation is discussed in detail in 2.3.

Over-lifting a gas-lift well should be avoided. It can be inefficient and cause increased operating costs; and it cancause severe operating problems.

2.2 Ineffective Gas Distribution

Ideally, the rate of gas available in a field should be equal to the sum of the optimum injection rates into the gas-lift wellsin the field. In real life this situation never occurs. Either more gas is available than optimally needed, resulting in somedegree of over injection, or less gas is available than optimally desired, resulting in some degree of under injection.


The following practices are recommended for effective gas distribution.

— Evaluate performance of the system, in addition to each well. Periodic evaluation can detect patterns of problemsassociated with ineffective gas distribution. For example, all wells in one part of a field may be under injected if

NOTE If the injection pressure is too high for any reason, it can cause an upper valve to open, leading to inefficient and verylikely unstable lift operation.

Figure 6—Problem of High Injection Pressure Causing an Upper Valve to Open



DRAFT

they are far from the source of the gas or hydrate is occurring in the pipeline at a low spot. Or some wells may beunstable if they are downstream of other unstable wells.

— Use a map of the distribution system to indicate which wells are normal, and which are under-lifted, over-lifted,and/or unstable.

— Use system-wide solutions to address distribution bottlenecks, gas shortage or oversupply problems, orineffective gas dehydration indicated by hydrate formation.

— Develop a plan for the optimum distribution of the gas that will be available with every scenario of compressoroutage. See API 11V8 for a detailed discussion of this process. With an automatic control system, the allocationcan be more easily optimized.

Gas should normally be distributed to the wells with the objective to maximize oil production; but optimization can alsobe expressed in terms of gas compression costs vs. the value of oil produced.

If there is over supply, the excess gas should be sold rather than being injected ineffectively, or injected into wells thathave a stability problem. In some cases, gas can be sold at an intermediate pressure that is less than the final gas-liftpressure, thus avoiding some of the compression costs.

If there is an under supply, the lower priority wells (high water cut, low productivity, etc.) should be given relatively lessgas than the high priority wells (low water cut, high productivity, difficult to re-start). If there is a temporary compressoroutage, shut in some of the lowest priority wells so that the better wells can continue to be produced effectively. Inmanually operated gas-lift systems, valves with small ports, or chokes, can be installed to limit the gas injection rate toassure that less gas is given to low priority wells.

The topic of gas-lift optimization—determining the optimum rate of gas injection into each well, and optimallydistributing the gas in cases of short supply—is discussed in API 11V8, Section 7.

2.3 Unstable Gas-lift Operation

Unstable operation or heading (not to be confused with the slug flow pattern in multiphase flow) has long beenrecognized as a serious problem confronting gas-lift wells designed for continuous operation.


The following practices are recommended for unstable gas-lift operation:

— use pressure transducers to gather and display injection or casing pressure, and producing or tubing pressure;

— use a gas meter and flow computer to obtain gas injection rate (and upstream pipeline pressure and differentialpressure);

— note patterns of instability and use these in gas-lift system analysis, as discussed in 2.2;

— see 2.4 to evaluate the type(s) or cause(s) of instability.

2.3.2 Unstable or Heading Gas-lift Operations Can Cause a Number of Problems

The following are possible problems arising from unstable or heading gas-lift operations.

— Excessive use of lift gas.

Unstable operation or heading is associated with gas being over injected during parts of the cycle and underinjected during other parts. Gas is being wasted during the over injection periods and used very ineffectively



DRAFT

during the under injection periods. The overall result is a very poor total ratio of injection gas to total liquidproduction. Unstable gas-lift heading is never as effective as steady flow in terms of lift gas utilization.Surprisingly, while the use of lift gas may be very inefficient when heading is occurring, one of the cures may beto increase the rate of gas injection, at least on a temporary basis, in conjunction with other steps to improve theoverall gas-lift design. This is discussed further in 2.4 under the topic of injection heading.

— Loss of production.

When a gas-lift well is heading, the pressure in the flow tube (normally the tubing) is fluctuating, and as a result,the FBHP is fluctuating. Production rate, based on the pressure drop from the reservoir to the well bore, is afunction of the time-weighted average FBHP, which may be significantly higher than the minimum bottomholepressure. Thus, a heading gas-lift well produces less than a stable well that is operating with a constant,minimum FBHP.

— Damage to the well bore.

Unstable operation can lead to fluctuating inflow velocities and excessive stresses on reservoir formation rockand sand control systems. Sand production can, in turn, cause damage to the well bore and gas-lift equipment,and to surface equipment. Excessive sand production can plug a producing well and lead to total well failure.

— Upsets to fluid handling facilities.

Heading can become severe, resulting in large pressure and rate surges that can cause upsets in gathering, fluidseparation or testing, and treating equipment. These upsets can, in turn, result in liquid carryovers, gas flaring,meter damage, rupture of a burst disc, or blowing open thief hatches on tanks.

— Upsets to gas handling facilities.

System upsets caused by severe unstable operation can affect gas compression, dehydration, and distributionsystems, causing increased compression and handling costs, gas flaring, and lead to nuisance shutdowns.

— Upsets to other wells on the system.

Heading can be induced in an otherwise stable well. Heading in one well may cause gas pressure fluctuationsthat interfere with and have a negative effect on other wells on the same distribution system.

— Difficulties in gas and liquid measurement and control.

The majority of gas measurements are made with differential pressure devices (e.g. orifice meters) and systemsthat record the static pressure, differential pressure across the device, and sometimes temperature. Thesedevices work best when the variables are stable. When a well is heading, the static pressure is fluctuating andthe differential pressure may be varying rapidly and widely, which greatly diminishes gas measurement accuracy;however an improvement is gained when a flow computer with rapid sampling is used.

Unstable well operation leads to surging production rates which make accurate tubing pressure and well testmeasurements difficult to obtain.

— Difficulties in analyzing gas-lift performance.

A good method for evaluating and analyzing gas-lift performance is with a flowing pressure survey, where theproduction pressures are measured under flowing conditions. When a well is heading, these surveys are moredifficult to obtain and analyze. The results may be meaningless unless pressure data is collected throughout theheading cycle at each depth (see Section 10).



DRAFT

— Difficulties in controlling and optimizing gas-lift performance.

All theories and methods for continuous gas-lift system design, operation, troubleshooting, and optimization arebased on stable operation. The theories all consider the steady state operating pressure, rate, and depth; notsome range of pressures, rates, and depths. When one or more wells in a system are unstable, the gas-liftmodels are less accurate, and the gas-lift operators cannot depend on these models for guidance.

2.4 Types and Causes of Unstable Operation

Two basic types of unstable operation or heading in continuous gas-lift wells are production (tubing) heading andinjection (casing) heading. The two types of heading and their causes and cures should be recognized andunderstood so they can be differentiated and addressed properly.

The goal in continuous gas-lift is to inject gas at a constant pressure and rate. With intermittent gas-lift, the objective isto produce slugs of liquid from the well by intermittently opening and closing the operating gas-lift valve andintermittently injecting volumes of gas into the production stream. This latter operation is not considered unstable,even though the well is “slugging.”


The following practices are recommended for differentiating and addressing types and causes of unstable operation:

— analyze a well’s heading or unstable operating characteristics to determine if the well is experiencing production(tubing) heading, injection (casing) heading, or both tubing and casing heading;

— indicate wells with heading problems on a distribution system map to determine if the problem is random or iscommon to several wells in a particular part of the system. If the problem is common to several wells, it may becaused by an upstream restriction, instability in the injection system, or a problem with the backpressureregulator that controls the pressure in the injection system;

— work through the heading analysis procedures to determine the most likely cause(s) of heading;

— take the recommended steps to eliminate or minimize the heading problem. Heading is always less efficient thanstable operation and severe heading can cause upsets or harm surface facilities.

2.4.2 Production (Tubing) Heading

A stable well produces with approximately constant production (tubing-head) pressure and flow rate. An unstable wellexperiences production (tubing) heading indicated by a varying tubing-head pressure and a varying flow rate vs. time.

The tubing-head pressure and flow rate may vary slightly, or the well may produce in slugs with periods of noproduction in between heads. The heads may be moderate, or they may be severe with the tubing-head pressurevarying by several hundred psi (kPa) and the instantaneous flow rate varying by several hundreds or thousands ofbarrels per day (m3/day).

Tubing heading can occur without casing heading, and can occur in natural flowing wells. The usual cause is theproduction rate is too low for the tubing size in the well; the low velocity initiates excessive liquid holdup. Productionheading may also be caused by inflow from more than one zone; or by restrictions in the surface facilities; or it may becaused or seriously aggravated by casing heading.

When a well has tubing heading, the pressure will be fluctuating at all depths in the tubing, including at the gas-liftvalves. Since the gas flow rate through a valve or orifice is a function of the downstream pressure, the flow rate willchange as the tubing pressure changes (up to critical flow). As the gas flow rate from the casing annulus to the tubingchanges, the casing pressure changes, which is defined as casing heading.



DRAFT

Methods to prevent or cure tubing heading include:

— determine if tubing heading is occurring by itself, or in conjunction with casing heading. Use pressure transducersto evaluate the fluctuations;

— cure tubing heading by choosing a tubing size that increases mixture velocity. An option may be to insert a coiledtubing string inside of the original production tubing string;

— cure tubing heading that is induced by casing heading by reducing the valve or orifice port size to create criticalflow. The port reduction breaks the link between the casing heading and the induced tubing heading. If casingpressure fluctuation is eliminated, the problem of tubing heading may be cured or reduced;

— do not choke the tubing at the surface to cure tubing heading, other than to protect surface facilities from severeslugging. Choking will increase the pressure in the tubing, increase the amount of gas required to produce thewell, and may reduce the production rate from the well;

— Inject more gas to alleviate tubing heading if adequate gas is available and this results in more economic overallproduction. A strategy based on operating experience is to increase the injection gas rate to raise velocity abovethe injection point to approximately 6 ft/s (1.83 m/s).

2.4.3 Injection (Casing) Heading

A continuous gas-lift well should have a constant injection (casing) pressure and gas injection rate. Casing heading isdefined as a varying injection (casing-head) pressure, which may be associated with a varying gas injection rate.However, even if the injection rate is held constant with a flow rate controller, this may or may not eliminate casingpressure heading. Tubing heading is often caused or aggravated by casing heading.

Casing heading occurs when the gas-lift well’s components (surface injection control device, unloading gas-lift valves,and/or operating gas-lift valve or orifice) are not correctly sized or adjusted for steady-state operation. The mostcommon cause is an operating valve or orifice that is sized too large. This allows the gas injection rate into the tubingto fluctuate. This causes the tubing flow rate and pressure to vary and further aggravates casing pressure (and thecasing gas inflow and outflow rate) heading.

Stable operation is attained when the actual flow rate into the casing equals the flow rate out of the casing annulus. Aslight imbalance will gradually change the volume of gas in the casing annulus; but a large imbalance leads to casingheading as the flow rate through the operating valve or orifice is alternately larger and smaller than the rate of gasinjected at the surface.

This problem may be compounded when the operator wants to increase the production from the well. When thesurface control valve is opened further, the injection rate and the casing pressure increase. Production may increaseand tubing heading may be decreased. However, the casing pressure cannot be increased too much or one or moreunloading valves may reopen.

Casing heading can also be intensified by under-injection which leads to self-intermitting at the operating valve (seeFigure 7), by production pressure operated (PPO) valves that are more sensitive to changes in tubing pressure, bysurface gas distribution system problems caused by heading in adjacent wells, by upsets in the gas compression, orby backpressure control regulator malfunction.

Casing heading can usually be prevented or cured by the following steps.

— Use flowing and static surveys and production tests to find the depth of gas injection, obtain the inflowperformance or PI, and create a valid compute model.



DRAFT — Use the model to evaluate the improvement from injecting more gas or injecting deeper. Do not be surprised if

this analysis indicates that more injection gas will be required to stabilize the well's operation.

— Redesign the gas-lift valve pressure settings, port sizes, and/or mandrel locations to permit gas injection from theoptimum depth, with no interference from upper unloading valves. Use an “orifice” valve at the operating point forhigh PI wells and a gas-lift valve for low PI wells. Do not oversize the orifice, as this can cause heading in boththe tubing and the casing.

— Eliminate outside effects such as upsets from other wells, fluctuations, bottlenecks, or other problems in thedistribution system that may be causing or aggravating the heading problem.

— Install chokes downstream of the ports in unloading gas-lift valves to prevent over injection. This can preventover injection and a resulting reduction in casing pressure, which can then cause the valve to close. It can alsoprevent throttling (reduction in gas passage as the valve begins to close). Furthermore, it can help protect thevalve's seat, stem, and ball from erosion during the unloading process. Moreover, it can allow the field to maintainonly one port size in their gas-lift valve inventory since various required rates can be achieved by choosing theappropriate choke size.

— Consider installing a smaller tubing string to increase the relative velocity in low-rate wells. Be aware that sizessmaller than 2 3/8-in. (6.06 cm) can cause wireline problems.

— Use an automatic control valve to control the injection rate. The controller is sensitive to changes in both supplypressure and rate, and thus prevents heading from occurring or becoming aggravated when outside interruptionsoccur.

2.5 Other Problems

There are additional problems.

NOTE If the injection rate is too low, the production gradient will increase and may flood the operating valve, causing theinjection pressure to rise and the next valve to open.

Figure 7—Problem of Under Injection—Causing Operating Valve to Flood



DRAFT


The following practices are recommended for problems with the gas-lift system:

— treat the entire gas-lift system as a “system” consisting of many parts that all must work together, rather than asindividual parts which can be considered individually;

— establish and maintain an automatic monitoring system to routinely check and/or calibrate all components of thesystem;

— establish quality assurance procedures for the assembly, repair, setting, testing, and installation of gas-lift valvesand assure that these procedures are followed.

2.5.2 Typical Problems

Problems encountered in gas-lift systems are as follows.

— Equipment malfunction.

Implement preventive maintenance on a regular basis to minimize unscheduled down time. Compressor andpiping components are mechanical and are subject to malfunction due to wear, corrosion, erosion, or plugging.

— Instrument malfunction. Maintain and calibrate the pressure, flow, and temperature measurement devices.Sometimes the failure is obvious but often the device appears to be working, but the information is faulty. Periodiccalibration is required.

— Equipment selection, assembly, or application.

Quality assurance is best provided by the following steps.

— Provide written standards to the gas-lift valve shops for valve assembly and cleaning, repair, and re-assembly, pressure charging the bellows, testing, etc. Assure that available API specifications andrecommended practices are followed. See API 11V7, Repair, Testing, and Setting Gas-lift Valves. Thisdocument will be replaced by ISO 17078-2 when it is published.

— Inspect the valves to insure that proper chokes are installed, if chokes are being used.

— Test the valves in the gas-lift shop or on location before they are installed in the well to insure they areproperly set to meet design specifications. This testing can also assure that the valves will open as designed,will close as designed, can provide the desired rate of gas passage, and do not leak. See API 11V2, Gas-liftValve Testing and Modeling. This document will be replaced by ISO 17078-2 when it is published.

— Monitor the installation procedure to insure the valves are run and set in the correct order and at the correctdepths. (For example, sometimes mandrels are to be skipped. Be sure that the wireline operator is aware ofthis.)

— Hydrates.

Hydrates are prevented by using dehydration to reduce water vapor content in the gas; this prevents watercondensation at cold temperatures. Pressure drops occur throughout a gas-lift system, and temperature dropsalways accompany pressure drops. If the right temperature and pressure conditions are present to condensewater, hydrates can form resulting in blockage of gas flow. This is addressed in 3.2.



DRAFT

— Gas-lift models.

Operation of a comprehensive gas-lift system depends on being able to model many aspects of the overallprocess, from the reservoir-to-well inflow, the vertical flow in the well, and flow through the surface lines andfacilities. Choice and proper use of the best models is necessary to fully understanding the gas-lift operation. Thissubject is dealt with in API 11V8, Recommended Practice for Gas-lift System Design and PerformancePrediction.

3 Surface Gas-lift Compression, Dehydration, and Distribution3.0 Purpose

Gas-lift compression and dehydration facilities, and pipeline distribution networks, are often already installed;changes can be justified only if production can be increased or downtime associated with gas-lift failures can beeliminated.

However, the compression, dehydration, and distribution systems should be evaluated for effectiveness, andimprovements should be justified if they will result in sufficient reductions in operating expense or increases inproduction.

3.1 Compression Facility

The compression facility should consistently deliver gas to the gas-lift distribution system at an adequate rate to allowa constant system pressure. This rate and pressure should be monitored to verify this condition and detect anyproblems. A master gas-lift distribution meter (or meters) should be used, in conjunction with a gas flow computer, tocontinuously monitor both the system gas input rate and pressure.


The following practices are recommended for a compression facility:

— design or configure the gas compression facility to deliver gas at a stable rate and constant pressure. Implementa compressor monitoring system to measure the flowrate and pressure of gas entering the distribution system. Ifone compressor goes down, some wells must be immediately reduced or shut down to prevent system upsets;

— set the pressure regulator on the compressor output to the gas sales system at the pressure required for the gas-lift distribution system or a minimum of 20 psi (137.9 kPa) above the sales line pressure. This will preventfluctuations or upsets in the sales system from upsetting the distribution system while allowing the excessreservoir gas to flow to the sales line;

— send excess gas to the sales line, or re-inject it into an oil or gas reservoir, rather than increase the distributionsystem pressure and over-lift some or all of the gas-lift wells;

— design the compression facility to deliver sales gas before it enters the gas-lift stage(s) if the sales line pressureis considerably lower than the gas-lift pressure;

— design or configure the compression system to have 10 % to 20 % excess capacity to prevent short-termcompressor outages from adversely affecting the gas-lift distribution system;

— implement a compressor monitoring system to measure the flowrate and pressure of gas entering the distributionsystem. The rate and pressure entering the system is important information; especially if the compressor facilityis running fully loaded. If one compressor goes down, the rate into the system will be reduced and the injectionrate into some wells must be immediately reduced or shut down to prevent system upsets;



DRAFT

— provide a make-up or recycle gas supply to maintain a stable compressor suction pressure. If the separator gasfrom the field is fluctuating, possibly due to heading wells, this make-up supply will permit the compressor tocontinue stable operation;

— implement a control system to prevent the compressor facility from delivering excess gas into the distributionsystem. Maintain a constant pressure in the gas-lift delivery system. Send any excess gas to the sales line or tosome alternative use, or re-cycle it rather than allowing it to be delivered into the distribution system to result inincreasing the distribution system pressure and over-lifting some of the wells. This requires a pressure set pointfor the desired pressure on the distribution system. Measure the volume of any gas sent to the sales line;

— perform routine compressor maintenance and implement a preventive maintenance program to minimize anyunscheduled compressor outages;

— coordinate any planned compressor maintenance with well operations so a balance can be maintained betweenthe gas rate being delivered to the distribution system and the sum of the injection rates into the wells beingserved by the system.

3.1.2 Compression Facility

The gas-lift compression facility should be maintained using accepted troubleshooting and preventive maintenancetechniques with an objective of at least 95 % on-line time. Frequent upsets will cause corresponding upsets in thegas-lift system that may result in lost production and require significant operator effort. Several days may be requiredfor a gas-lift system to return to stable operation following a compressor upset.

When a compressor goes down or must be shut down for maintenance, the gas distribution must be adjusted asquickly as possible to keep from adversely affecting the wells in the system. The gas going to the poorer wells shouldbe cut back so as not to adversely affect the better wells. In the extreme, it may be necessary to shut in the poorerwells during the outage.

Compressor outage in a large gas-lift system requires a contingency plan that can be implemented quickly. If asystem is operated manually, the best plan is to shut down the least profitable wells as quickly as possible. If a systemis operated with computer controlled distribution and injection, optimal well-by-well adjustment may be possible.

Gas compressors are often used both for sales and gas-lift distribution. The pressure should be regulated to meet themaximum requirement; either gas-lift or sales line pressure. The pressure regulator should be kept in good workingcondition, and should be set at least 20 psi (137.9 kPa) above the required sales line pressure so that upsets in thesales system will not affect the pressure of the distribution system.

A gas-lift system pressure chosen years ago, when conditions in the field were different, may benefit from a systempressure increase for more effective gas-lift. Wells that may benefit include ones:

— that have high productivities;

— are producing with high water cuts;

— are producing from high in the hole because the system pressure is too low to permit working down to the bottomgas-lift valve.

In other cases, the pressure should be decreased if the current system pressure is too high for the current well andreservoir conditions. This would normally only be true if the wells have low productivities and/or have depletingreservoir pressures.

Changing system pressure may be as easy as changing the set point on the sales pressure regulator. However,system pressure increase may require a change of the compressor cylinders or stages, and/or the distribution system



DRAFT

piping, meters, control valves, and gas-lift valves. To decrease the system pressure, the regulator must be reset andthe gas-lift valves may have to be reset.

Raising gas-lift system pressure requires a review of the pressure ratings of the piping, flanges, and valves in thesystem. For example, many systems are rated at ANSI/ASME 400, 600, or 900 corresponding to pressures of 960 psi(6,618 kPa), 1440 psi (9,928 kPa), or 2160 psi (4,893 kPa), respectively. These are common pressure ratings forpiping, flanges, and valves.

The compression horsepower required increases with the pressure ratio (compression ratio). The capital cost andoperating expense is a function of the amount of gas being compressed. If less injection gas is required at a higherdistribution system pressure, then overall compression costs are reduced.

Another way to improve gas-lift efficiency is to minimize the compressor suction pressure. When suction pressure isreduced, this may permit the operating wellhead pressure to be reduced, which improves the efficiency of the gas-liftwells. A lower limit to suction pressure can be calculated (an example is given in API 11V8, Section 6).

3.2 Gas Dehydration Facility

Most gas-lift systems use dehydrated gas; however, in some fields wet un-dehydrated gas is used. This may lead tocorrosion or formation of hydrates and plugging of piping if temperatures fall below the hydrate formation temperaturefor the system pressure. Hydrates are solid crystals that form when gas molecules are trapped within a cage of watermolecules. If water is present, hydrates can form at 80 °F (26.67 °C) in high-pressure systems.

A large temperature drop accompanies a large pressure drop, due to the Joule-Thomson effect; this can causehydrates in the gas. Hydrate formation at the chokes or in the distribution lines causes increased pressure drops,which makes the freezing problem worse. The blockage can rapidly diminish the flow of gas and cause the well to beshut down. Figure 8 (courtesy GPSA, Engineering Data Book) shows potential hydrate conditions.

Liquid in the gas-lift distribution system can also cause problems with measurement or control. Liquid tends to collectin any low spot in the distribution system; when a sufficient quantity has collected, it moves as a slug. Liquid sluggingsignificantly affects and/or damages gas meters, controllers, and gas-lift valves and/or orifices. Inaccuratemeasurement, ineffective gas control, permanent damage to equipment, and/or severe production heading in thetubing can result from slugging.


The following practices are recommended for a gas dehydration facility:

— dehydrate all gas that enters a gas-lift distribution system. The objective is to lower the dew point below thelowest expected temperature to prevent hydrate formation. In many fields, this can be achieved by dehydrating to7 lb (3.18 kg) of water per million standard cubic ft (28,317 m3) of gas. Even drier gas may be needed in very coldweather, or when CO2 is present, where the guide is 3 lb (1.36 kg) per million standard cubic ft (28,317 m3).Figure 9 (courtesy GPSA, Engineering Data Book) shows water content versus dew point temperature with gaspressure as the parametric lines;

— design gas-lift systems to avoid excessive pressure drops in the surface piping distribution system. Minimize thepressure drop across the surface control choke or control valve by using a gas-lift valve (rather than an orifice) asthe operating valve on low PI wells so that casing pressure will be higher;

— design gas-lift systems with liquid traps at low spots where liquid can accumulate, especially upstream of gasflowrate meters or control devices. Install blow down valves in the traps so any accumulated liquid can be



DRAFT periodically removed from the system. Large piping distribution systems, particularly offshore, should have piglaunchers and receivers to remove liquid and debris by periodic pigging;

— if the presence of liquid cannot be avoided, periodically purge the distribution system to remove accumulatedliquid. Purge orifice meter pressure and differential pressure taps periodically;

— use seal pots (liquid traps) on orifice meter pressure and differential pressure taps and purge periodically. Installthe meter recorders or transmitters at higher elevation above the orifice tube and insure the connectinginstrument piping has no traps;

— use methanol to avoid temporary freezing problems. Also insulate control valves or chokes where large pressuredrops may occur. Methanol is a hydrate point depressant that lowers the dew point. The best approach is to avoidthe problem and cost by using effective dehydration;

— use a heater (line, gas, or catalytic) as an alternative to avoid hydrate formation if effective dehydration is not anoption.

3.3 Gas-lift Distribution System

The best type of distribution system is one with a large volume. This ensures that relatively small changes in inlet oroutlet rates will not cause corresponding rapid changes in pressure throughout the system.

Figure 8—Hydrate Conditions



DRAFT

Figure 9—Water Content vs. Temperature and Pressure



DRAFT


The following practices are recommended for an unstable gas-lift distribution system:

— use a “finger” distribution system, Figure 10, with each well served by an individual injection line from a manifold;

— measure and control the injection rate to each well at the manifold. Measure the injection gas pressure at thewellhead just before the gas enters the casing annulus;

— if a manifold is used, the injection rate to each well can be measured and controlled at the manifold, to save cost.However, unless the individual injection lines are short, avoid the temptation to measure the injection pressure atthe manifold. This should be measured at the wellhead just before the gas enters the casing annulus;

— where a “finger” system can not be used, consider a “looped” system, Figure 11, to expand the volume of thedistribution system to minimize upsets caused by minor fluctuations in inlet or outlet rates;

— use larger diameter distribution piping if combining continuous and intermittent gas-lift wells on the same system.If possible, schedule the intermittent cycles to minimize upsets, which could occur due to concurrent,uncoordinated injection cycles;

— Avoid combining both continuous and intermittent gas-lift wells on the same system. If this is necessary, set theintermittent controllers to open at different times, or automatically schedule the intermittent cycles from a centralcontrol location to minimize the upsets which could occur due to concurrent, uncoordinated injection cycles.

Figure 10—Finger Distribution System

Figure 11—Looped Distribution System



DRAFT

3.3.2 Distribution System

Pressure pulsations can reduce production if the piping size of the distribution system is small. The system volumecapacity may be increased by connecting the annulus of a temporarily abandoned well to the system. Verify theintegrity of the casing and wellhead, and mount a pressure transducer on the casing head. There should be norestrictions (chokes, orifices, control valves, etc.) between the distribution system and the storage annulus. Thisadded distribution system volume will help smooth out pressure fluctuations in the system, and it will serve as aconvenient gas storage location if the wells must be shut in for a hurricane or other reason. Having some high-pressure gas trapped in the distribution system can help return the wells to production, since after a system isshutdown, there may be insufficient gas production to bring the system back onto operation.

The recommended “finger” configuration with a large distribution pipeline minimizes interference between wells sincethey are isolated from each other through the manifold. With this system, the pressure, temperature, and rate of gasentering the manifold from the distribution system should be measured. This can be useful in troubleshootingproblems in the overall distribution system and in determining the amount of gas available to the group of wells beingserved by the manifold. The injection pressure and production pressure should be measured at the wellhead.

Fields using a “series” or “looped” system can suffer interference between wells, which may be partially mitigated byincreasing the volume of the system with larger pipe diameter. This can be aided by using the annuli of non-producingwells or by using parallel looped lines. These systems should have the gas flow rate, casing pressure, and tubingpressure data measured at the well site.

In any continuous gas-lift distribution system, and especially in “series” or “looped” systems, it is important to avoidpressure fluctuations and rate surges, as these can upset the wells and may become aggravated as instability in onewell can lead to instability in adjacent wells. In these systems, it is important to address heading problems (see 2.5)and to avoid use of intermitters or intermittent gas-lift. If intermittent gas-lift must be used to produce some of thelower rate wells, the intermittent cycles should be scheduled and controlled centrally, according to a schedule, tominimize severe interference and fluctuations. Or, if possible, the intermittent wells should be isolated from thecontinuous gas-lift wells, on a separate portion of the distribution system.

Just as with “finger” distribution systems, it is important that the pressure and rate be measured and recorded at keypoints or branches in the system. This can be beneficial in troubleshooting problems in the system and with the wellsserved by a particular branch of the system. However, accurate gas measurements may be difficult in systems withfluctuating pressures.

4 Gas Injection Metering and Control4.0 Purpose

Gas-lift requires points where the gas being injected into each well is measured and controlled. These provide theoperator with control over system operation and data for finding under-performing wells. Gas-lift measurement andcontrol must be performed as correctly, accurately, and consistently as possible.

4.1 Gas Metering

Gas injection measurement is important in gas-lift optimization and troubleshooting. Continuous injection pressureand rate measurement of a single phase fluid is more accurate than the separator gas measurement during a welltest, which often is slugging and with gas that is saturated with liquid. Other variables that can be measured at thewellhead are injection (casing-head) pressure, injection gas temperature, producing (tubing-head) pressure,producing temperature, and in some cases production rate.



DRAFT


The following practices are recommended for gas metering:

— measure the gas injection pressure and rate for each well;

— use a properly installed, properly ranged, well-maintained, accurately calibrated flow rate measurement device;

— use flow computers to obtain improved accuracy for well analysis. Use multi-range differential pressuretransmitters that extend the maximum and minimum rate capability of the orifice meter;

— use a computer automation system to gather information;

— pay special attention to obtaining accurate gas injection pressure and rate information during each well test.

4.1.2 Metering Location

The meter location depends on the type of distribution system. A “finger” system normally has the meter at themanifold. If a “series” or “looped” system is used, the meter is at each well. The rates should be measured upstreamof each well's control valve so the “static” pressure measurement is the upstream distribution system pressure; andtemperature can also be recorded to improve measurement accuracy. The data serves both rate measurement andanalysis of the distribution system pressure fluctuations.

4.1.3 Metering Devices

The orifice meter is the most commonly used device in the oil field for measuring gas injection. This meter tube andorifice plate device is satisfactory when coupled with a flow computer or calibrated chart recorder:

An orifice meter, flow computer, or chart recorder should meet the following criteria.

— applicable API, GPA, or ASME standards should be followed;

— the orifice meter should be properly located with respect to chokes, control valves, pipe, and bends;

— the orifice meter instrument tubing to the pressure transducers or chart recorder should avoid low spots that cancollect condensed moisture. Seal pots can be used to trap condensation;

— chart recorders should be equipped with a three-way or five-way manifold for easy and safe blow down andpurging. The tubes should be routinely purged;

— the orifice meter tube and orifice plate should be correctly sized to measure the expected gas flow rate. Thedifferential pressure, or variation from maximum to minimum, should fall in the midrange of the computer orrecorder chart, not near the bottom or the top of the scale. Use multi-range transmitters for flow computers;

— the orifice plate should not be painted, dirty, scaled, scarred, nicked, worn, bent, or damaged. When they are notin use, orifice plates should be stored to prevent damage or corrosion;

— if chart recorders are used, they should be calibrated on a routine basis, and the recorder’s pens should beroutinely checked to maintain proper inking. The recorder should be equipped with a 24-hour clock drive for welltesting and charts changed before and after well testing. 8-day clocks are adequate for routine surveillance butthe measurements are difficult to read;

— use the injection gas rate from the flow computer or chart in conjunction with a well test (discussed further in10.1) for gas-lift surveillance. Also, formation gas/oil ratio is determined by subtracting the gas injected during the



DRAFT

well test from the total gas recovered (produced plus injected) during the test. Thus, both readings must beaccurate to determine the reservoir gas production rate;

— operators can use recorders with linear charts or L-10 (square root) charts. L-10 charts have the advantage thata 100-in. (256-cm) water column (maximum range) differential pressure can be directly obtained from the chartreading. However the static pressure must be calculated with equations to convert from square root to linear.Some charts have both scales for convenience.

4.1.4 Metering Alternatives

Microprocessor based gas measurement devices can also serve for gas flow rate control. The flow computer can givemore accurate gas measurements than chart recorders, especially for heading wells. They require less time formaintenance, calibration, and reading, and they can communicate with a remote computer for display, use, andstorage of the pressure, rate, and volume information at a central location.

Other metering alternatives are available, but orifice meters and flow computers are proven methods. Other methodsinclude flow nozzles, Venturi meters, vortex shedding meters, and turbine meters.

4.2 Injection Control

The term “gas injection control” implies controlling the gas rate into the well's casing annulus for stability reasons andcontrolling the injection rate into the production stream for gas-lift optimization. There is a significant difference that isimportant to understand.


The following practices are recommended for injection control:

— operate the well close to the original design conditions of injection pressure and rate until reservoir pressure andwater cut changes require gas-lift revisions;

— consider pulling and redesigning the unloading gas-lift valves, or the operating gas-lift valve, when reservoir andwater conditions change, rather than attempting to compensate with a limited control capability, since operating agas-lift well far from its design settings may lead to unstable, inefficient operation. If tubing must be pulled tochange valves, potential reservoir formation rock damage and reduced production should be evaluated;

— use a gas flowrate controller so that the desired injection rate can be consistently maintained with minimumimpact of either upstream or downstream pressure fluctuations or upsets;

— use of a controller that can incorporate pressure overrides or limits to restrict the range of casing-head pressuresthat the controller will permit;

— a gas-lift control device (variable orifice, choke, control valve, etc.) is not a measurement device; the pressuredrop across such as device gives only an estimate of the rate. There are some exceptions to this. Somecompanies produce calibrated control devices that can also provide flow rate measurements. Normally these areconsidered when there is inadequate space to install separate flow measurement and control devices;

— finally, a gas-lift controller is not a shut-off valve. Do not depend on it to be gas tight.

4.2.2 Gas-lift Control

A continuous gas-lift well is designed based on a defined production and gas injection rate and water cut. The well’sactual conditions will vary with time from the design assumptions, which will require injection rate variation but still



DRAFT

remain within the design limitations. However, when gas injection rates no longer match port sizes or pressuresettings, instability or non-optimum operation ensues.

A flowing pressure survey and validated well model should be used to evaluate whether a redesign of gas-lift valvesand/or the operating gas-lift orifice will better meet the new actual and/or desired operating conditions. Theeconomics of the changes and the potential for formation damage can be simulated with the computer model. Newcontrol equipment can be implemented if necessary.

The original design constraints or conditions are important; this information should be made readily available to thosewho must operate the gas-lift system over the long term.

4.2.3 Gas-lift Control Challenges

Manual variable chokes are adjusted until the desired differential pressure reading is obtained on the flow computeror chart recorder. The control setting obtained in this manner is only valid for one set of conditions of upstream anddownstream pressure. It is difficult to maintain consistent control, since the injection rate associated with a certaindifferential pressure pen reading on one day will not be the same as the rate associated with the same reading onanother day under different conditions.

The rate of gas passage through a variable choke or orifice depends on the upstream pressure, the downstreampressure, and the size of the restriction. (This is true as long as there is not critical flow through the restriction. Criticalflow occurs when the downstream pressure is less than approximately 60 % of the upstream pressure. When thisoccurs, the flow rate depends only on the upstream pressure and the size of the restriction.) If any of these variableschange, which is likely on a continuous basis in a dynamic gas-lift system, due to pressure variations in the deliverysystem, injection rate adjustment to another well, wells heading, or production pressure variations, this will cause therate of gas injection into the well to change.

Injection control is complicated since surface choke size and gas rate into the casing are further affected by therestriction through the operating gas-lift valve or orifice and its pressures, both upstream and downstream. Thiscontrol of the injection at two locations can lead to instability, or heading, or inability to pass the adequate rate of gas.

The surface control presents a dilemma: use a high kick-off casing pressure and small pressure drops to “work down”and inject gas into the production stream as deep as possible, or use a high kick-off pressure and larger pressuredrops for control, but with resulting shallower injection depth. The initial unloading or kick-off casing pressure and theultimate operating casing pressure may differ for IPO gas-lift valves by several 10s or 100s of psi (100s or 1,000sof kPa). This is discussed further in 5.1.

The higher the pressure drop at the surface, the greater the degree of “control” of the injection, and the better thedegree of isolation between the well and upsets in the upstream distribution system. Unfortunately, this can causehydrate formation, even in “warm” climates.

A small surface pressure drop relative to the pressure drop across the downhole operating gas-lift valve or orifice(have the surface restriction or choke larger than the downhole restriction) is the best design. The ideal case is tocontrol the rate of gas passage into the producing stream at the downhole valve or orifice, but this design method isvery difficult to accomplish. If the surface pressure drop is large relative to the downhole pressure drop (surfacerestriction is small relative to the downhole restriction), serious heading can result.



DRAFT

4.2.4 Control Option—Gas Controller

Automated gas flow rate controllers provide better control options than fixed-bean chokes, manual control valves, orvariable orifices:

— use a gas flowrate controller that can maintain a constant flow rate in the face of varying upstream anddownstream pressures by dynamically adjusting the size of the opening based on feedback from the gasmeasurement device;

— use a pressure controller for a constant downstream pressure, but only for wells with PPO valves or a singleorifice point of injection into the tubing;

— a third alternative is a combination device that would provide flow control with a pressure override or pressurelimiting capability to avoid re-opening upper gas-lift valves.

A flow controller provides an improvement over manual control methods; however, this provides no guarantee ofdesired performance. If the flowrate into the casing is greater than the flowrate capacity of the operating gas-lift valveor orifice at the downhole pressures, instability and heading or multi-point operation, both of which are undesirable,will still occur. Similarly, a gas-lift valve or orifice with greater capacity than the inlet gas rate will lead to anotherinstability and heading. However, if the valve or orifice is resized, the flow controller is a significant improvement overmanual flowrate control.

A gas flowrate controller can be a relatively expensive device. However, as mentioned above, microprocessor-basedunits can combine the gas measurement and gas flowrate control into one unit, so overall costs may be minimized.

Use of a gas flowrate controller has another advantage. The operator can test the well at various injection rates anddetermine the optimum rate for each well. This can be accomplished by using multi-rate well testing, where the well isproduced at various injection rates and the corresponding production rates are measured and analyzed.

4.2.5 Control Strategy

Use of a manual control device requires a surface pressure drop (during operation, not during unloading) of at least50 psi (345 kPa) but not more than 150 psi (1,034 kPa) in a 1,000 psig (6,895 kPa) system, and from 100 to 200 psi(690 kPa to 1,380 kPa) in a 1,440 psi or 2,200 psi (9,928 kPA or 13,790 kPa) system. These pressure drops willusually be achieved if adequate “unloading” casing pressure drops are included in the gas-lift valve design to permitthe system to effectively “work down” to the operating gas-lift valve or orifice. The design of gas-lift valves forunloading is discussed in 5.1.

If the pressure drops at the surface are much less than these guidelines, any small change or upset in either theupstream or downstream pressure can cause a significant change in the gas injection rate. If the surface pressuredrop is much more than these guidelines, the surface restriction is too small, relative to the downhole restriction in theoperating gas-lift valve or orifice, and serious casing heading can result. The downhole valve or orifice should bechanged to a smaller size.

Surface control and heading wells have some or all of the following characteristics: low bottomhole pressure, poorinflow performance, low production rate, or gas-lift valve ports (or chokes) that are too large. The pressure available inthe system may be more than is needed for the low productivity, heading wells. To restrict the pressure and rate ofgas injection into these wells, the surface restriction may have to be very small and the surface pressure drop mayhave to be very large (sometimes several hundred psi) relative to the pressure drop across the operating gas-lift valveor orifice. This situation is almost guaranteed to cause or aggravate a serious heading problem, and may causehydrate problems if any water vapor is present in the gas.

The potential use of a pressure controller was mentioned above. Such a device would attempt to maintain a constantdownstream or casing-head pressure. To do this, it would need to vary the injection rate as either the upstream



DRAFT

pressure or the downstream pressure varied. Since the gas flow is actually controlled at two locations, the secondbeing the downhole operating valve or orifice, the variable rate would affect the downhole operation and the flow intothe production stream, leading to instability in the system. Also, the casing-head pressure of a gas-lift well mustchange as the well unloads. Therefore, there would need to be a separate unloading scheme, since strict pressurecontrol would not permit unloading, with the possible exception of where PPO gas-lift valves are used. In general, thisapproach is not recommended.

The possible use of a flowrate controller with a pressure override or pressure limiting capability was discussed above.As indicated, the unloading process is basically a pressure-controlled process, whereas the actual control of a gas-liftwell depends on control of the gas injection rate. There may be times in the operation of a well when it must either beunloaded or kicked off, or the operating point may need to be changed. In these instances, it may be necessary tolimit, control, or influence both the injection rate and the injection pressure. No precise guidelines can be given in thisarea, since each situation is different, but the ability to control both flowrate and pressure may be necessary for themost effective means of overall gas-lift control.

5 Gas-lift Valves5.0 Purpose

Gas-lift valves are used in most gas-lift wells, except for very shallow wells or high pressure systems where gas canbe injected deep in the well or around the end of the tubing with no valves. Valves are primarily used to “unload” thewell (remove completion or other fluid from the gas injection conduit—usually the casing/tubing annulus) down to thedepth where gas is to be injected into the production conduit (usually the tubing). A gas-lift valve or orifice controls therate of gas injection into the production stream.

The process of gas-lift design consists of spacing the gas-lift mandrels and calculating the set pressures of the gas-liftvalves. The objectives are to space the mandrels and select the types of valves, port sizes, set pressures, and chokesizes to achieve the desired well operation.

5.1 Unloading Valves

After allowing the well to “unload,” the unloading valves should stay closed so the gas will be injected below them andthrough the operating gas-lift valve or orifice.


The following practices are recommended for unloading valves:

— check the performance of each gas-lift well frequently with pressure surveys and use the data to create acalibrated computer model. If a pressure survey can not be easily run, such as on some offshore and sub-seawells, it may be possible to do this with information from downhole pressure and temperature gauges, which areoften installed in these types of wells;

— adjust input gas control and eliminate wellhead restrictions if a well is heading, or is found to be operating throughone or more upper, unloading valve(s). If adjustments do not resolve the heading, consider pulling andredesigning the unloading valves with smaller ports and/or change the operating valve or orifice to better matchthe well's current operating condition (see also under-lifted wells in 2.1);

— review Section 12 for additional recommendations pertaining to unloading valves and initial unloading practices.

5.1.2 Unloading Valve Problems

Unloading mandrel depths and valve pressure settings are based on the original well conditions. A problem couldoccur when changes occur in the well, or in the surface gas-lift distribution system including the availability of lift gas



DRAFT

to the well. These changes subject the unloading valves to different pressures and/or temperatures, and may causeone or more of them to re-open, some or all of the time, causing one or more of the instability or inefficiency problemsenumerated earlier.

The majority of the unloading valves in wells designed for continuous gas-lift operation are injection pressure (casingpressure) operated (IPO) valves. For these valves, ± 90 % of the opening force (depending on the specific valve portsize and configuration) comes from the casing or injection pressure. During the unloading process, these valves canbe closed by a reduction in the casing or injection pressure as gas is injected deeper in the well down to the operatinggas-lift valve or orifice. However, when conditions change, the unloading valves may be reopened with increasedcasing pressure.

Production pressure (tubing pressure) operated (PPO) valves are frequently used in dual gas-lift wells and may beused in situations where the surface gas-lift injection pressure and rate can be very closely controlled, or interestingly,in cases where there is virtually no control of the surface injection rate or pressure and the desire is to let the “well”control its own operation; that is, to let the gas-lift operation be controlled by the production pressure demands.Typically a stronger well with a higher (heavier) production pressure gradient (high production pressure) will tend to liftfrom higher in the hole and a weaker well with a lower (lighter) pressure gradient (lower production pressure) will tendto lift from deeper in the well.

In a well with IPO unloading valves, a much smaller increase in the casing pressure than in the tubing pressure isneeded to cause the valve to re-open. For instance, if a valve has a production effect factor of 0.1 (and thus aninjection effect factor of 0.9), a 100 psi (689 kPa) increase in the injection pressure will increase the opening “force” by90 psi (621 kPa), whereas a similar increase in production pressure will increase the opening “force” by only 10 psi(69 kPa). Just the opposite is true for production pressure valves. (The words “opening force” are placed in quotesbecause this is really not a force, but rather an effective pressure that acts against the bellows pressure to try to openthe gas-lift valve.)

This leads to the main reason for using IPO valves in preference to production pressure valves. It is easier to controlpressure in the casing than in the tubing, and therefore it is easier to control the unloading process and keep the wellfrom becoming unstable due to operational pressure fluctuations. If the casing pressure has changed from “design”conditions, or if it fluctuates for any reason as discussed before, serious problems can arise since relatively smallchanges in casing pressure can cause one or more upper unloading valves to re-open, part or all of the time. Also, ifthe injection pressure changes, this will change the injection rate through the valve. This can further negatively impactthe performance of the well.

When an upper unloading valve does reopen inappropriately, there is usually a large initial pressure drop across thevalve and a large instantaneous gas injection rate (from casing to tubing). This causes a large “head” in the tubingpressure as the large “bubble” of gas expands and rushes to the surface. It also causes a large “depression” in thecasing pressure since the gas leaves the casing much faster than it can enter the casing at the surface through thesurface control mechanism. This can be prevented by using a choke in the unloading valves to permit them to passonly the amount of gas needed during the unloading process.

There are several reasons why a well may have upper, unloading valves re-open at inappropriate times, or pass gascontributing to instability or non-optimum operation.

— The nominal casing pressure may now be higher than the design casing pressure. This can happen in oldersystems that are serving fewer gas-lift wells than before.

— The casing pressure fluctuates and is occasionally higher than the design value. This can happen with casingheading and may be the most common cause of interference with upper, unloading gas-lift valves.

— The tubing pressure may now be higher than the design tubing pressure. This can happen in several ways. Thewell may have been stimulated and its productivity improved. The well may now be producing much more waterthan before. The well may have tubing heading and the tubing pressure may be temporarily higher due to a



DRAFT

“head” caused by a valve opening down the hole. The well may have been temporarily shut down and is in needof being restarted.

— The set pressure may now be lower due to a leak in the bellows. This happens very rarely. Usually, the bellows iseither intact or has failed completely, in which case the valve is almost always fully closed.

— The valve port or stem may be “cut” (eroded) such that the valve is no longer gas tight, even if closed.

— There may be a leaking check valve, gas-lift valve seal, packer seal, or tubing joint, and fluid (oil and/or water)may be intermittently entering the casing annulus and then being forced back into the tubing through a miniunloading process.

— The gas-lift unloading valve design may not be correct for the current conditions of the well, or the valves may nothave been actually set and installed as they were designed.

— The valve spacing may not be correct for the current conditions and it may not be possible for the well to unloaddown to the operating gas-lift valve.

There can also be a reason that upper unloading valves may not re-open when they are needed to re-unload (or kickoff) a gas-lift well that has been temporarily closed in. If the well has been producing at a high rate, the temperature ofthe upper unloading valves may have increased due to the temperature of the produced fluid. This increasedtemperature will increase the bellows pressure and keep the valves closed. In such cases, it may be necessary towait long enough for the upper unloading valves to cool down so they can be re-opened and the unloading or kick-offprocess can be initiated.

5.2 Operating Valve(s)

The operating valve is intended to be open so that gas can be injected into the production stream (usually the tubing),unlike the unloading valves which are designed to stay closed after the well is unloaded and on production. Manyoperators prefer to use an orifice (gas-lift valve with no bellows or stem) for high PI wells. This flow device is always“open” for gas injection into the tubing, but does have a check valve to prevent fluid movement from the tubing backinto the casing/tubing annulus.


The following practices are recommended for operating valve(s):

— locate the operating valve or orifice depth based on gas injection pressure and tubing pressure derived frominflow performance associated with reservoir pressure and production rate;

— invest in a number of relatively closely spaced mandrels in the expected vicinity of the operating valve depth, sothe actual depth of lift can be fine-tuned when the well is placed on gas-lift, or when operating conditions change.This is especially important in wells with high productivity indices;

— use an orifice as the operating valve on PI wells where gas injection should not be restricted, rather than a gas-liftvalve; this can prevent throttling and other problems that can be caused by valve action. Alternatively, a valve canbe used if the valve's set pressure is designed to keep the casing pressure high on low PI wells, thus preventinghydrate problems at the surface choke or excessive gas injection;

— use a choke in the orifice to limit the amount of gas injection to the desired rate. (This may be preferable to usinga small orifice, since this can limit flexibility.) Do not use such a small choke that it may become plugged with rustor scale. Also, do not use a smaller choke than required for the anticipated maximum injection rate;



DRAFT

— change the operating valve or orifice to the correct depth or the correct size when well conditions change. Withwireline techniques, the change is relatively easy, but with tubing retrievable valves, an economic evaluation isrequired. An improperly spaced, sized, or set valve or orifice can seriously affect gas-lift operational efficiencyand profitability.

5.2.2 Operating Valve Problems

Possible operating valve problems are as follows:

— an operating valve that is too shallow gives the same result as injecting through an upper, unloading valve. Thewell cannot be drawn down sufficiently, and poor production results. The valve may be too shallow if the reservoirpressure has declined since the well was designed, or if there is some reason why a valve cannot be placed in adeeper mandrel;

— if an operating valve is too deep, the well may not be able to unload down to it, or may not be able to stay on itcontinuously. The well is forced to operate through an upper unloading valve, with the attendant losses inproduction and inefficiencies.

An operating valve or orifice is difficult to space correctly because future operating conditions are never knownprecisely. One method is to space several side pocket mandrels relatively close together [within 300 ft to 500 ft(91 m to 153 m) vertical depth of each other] in the vicinity of the expected operating depth, especially if the wellmay have a high productivity. The “best” operating depth can be established by using the wireline to install theoperating valve or orifice at each mandrel, one at a time, until the “best” production is obtained. Another methodis to place additional mandrels with dummies below the initial operating orifice to plan for reservoir pressuredecline and the requirement of a deeper operating point later in the well’s life;

— an operating valve or orifice that has too small a gas passage capability, either due to a small port, or a smallchoke, or both, will not be able to pass the desired rate of gas. An upper valve will eventually re-open, causingheading and inefficiency. Also, small orifices or ports can become plugged with solids such as rust, scale, orsand. It is recommended to not use a valve port size smaller than 1/4 in. (0.64 cm) and a choke size smaller than8/64 in. (0.32 cm);

— an operating valve or orifice that has a large port may pass too much gas, thus creating instability or casingpressure heading. Over injection may be required to maintain stable operation. To address this condition, pull andchange the size of the valve or orifice.

6 Well Equipment—Tubulars, Completion, and Wellhead6.0 Purpose

All gas-lift wells must have a gas injection conduit (normally the casing-tubing annulus), a production conduit(normally the tubing), a completion zone between the producing formation and the well bore, and a wellhead toconnect the injection and production conduits to the surface injection and production lines. In addition, most wellshave some wellhead instrumentation. Optimized gas-lift operation requires these items to be properly sized for flowcapacity and equipment clearance.

Follow API RP 5C1, Recommended Practice for Care and Use of Casing and Tubing, Eighteenth Edition, May 1999.For design and ordering, use API Spec 5CT, Specification for Casing and Tubing, Seventh Edition, October 1, 2001 orISO 11960:2001, Petroleum and natural gas industries—Steel Pipes for Use as Casing or Tubing for Wells. Wellheadrequirements are in API Spec 6A/ISO 10423:2003, Specification for Wellhead and Christmas Tree Equipment.



DRAFT

6.1 Casing Annulus

Most gas-lift wells have the gas injected down the casing-tubing annulus and production flows up the tubing. If theinverse is true, most of the annulus comments in this section pertain to the tubing and the comments in 6.2 pertain tothe casing.


The following practices are recommended for the casing annulus:

— leave the casing annulus clean after the original completion or a workover by using a casing scraper andcirculating the well with clean (low solids) salt water or oil;

— verify the pressure integrity of the casing annulus by hydrostatic testing to the gas-lift gas kick-off pressure;

— design the completion tubulars for an adequate annular flow passage for the gas and sufficient clearance of side-pocket mandrels. Do not attempt to operate with an annular area that is too small. Suitable installation sizes wouldbe 2 3/8-in. (6.03-cm) OD tubing in 4 1/2-in. (11.43-cm) OD casing, 2 7/8-in. (7.30-cm) OD tubing inside 5 1/2-in.(13.97-cm) OD casing, 3 1/2-in. (8.89-cm) OD tubing inside 7-in. (17.78-cm) OD casing, and 4 1/2-in. (11.43-cm) ODtubing inside 9 5/8-in. (24.45-cm) casing;

— avoid simultaneous gas-lift of both sides of a dual well, or be prepared for the difficulty of downhole control of gasinto each string. In theory, two production zones with two tubing strings in a single wellbore appear to be costeffective; however, in practice, successful gas-lift of both sides of a dual well is difficult to accomplish. (Arecommended practice on dual wells, API 11V9, is currently under development.)

6.1.2 Casing Annulus

The casing annulus must be clean to be an effective conduit for injection gas. If it is not, trash or solids can be injectedthrough the gas-lift valve(s), along with the gas, and either plug or erode the valve(s).

The casing annulus must have good pressure integrity to prevent gas leaks into shallow zones, and to prevent leaksfrom high pressure zones into the casing that can result in the “production” of fluid through the gas-lift valve(s). Thiscan cause plugging or erosion of the gas-lift valve(s), or leaks and erosion in the tubing.

The casing must have adequate clearance for the eccentric offset of side-pocket or conventional mandrels and theannulus must be sized to permit gas-lift injection at the desired rate(s) and pressure(s). Most installations haveadequate size, but if the annular area is too small, caused by tubing collars or side-pocket mandrels that are too largerelative to the casing ID, high rate injection can create friction pressure loss in the annulus. Casing pressure at thedepth of the operating valve or orifice may not be sufficient to inject the desired amount of gas into the tubing, whichcould reduce production or cause lift from an unloading valve high in the well.

Casing can be so large that an excessive gas volume is injected at the unloading or operating valves. If this occurs,the size of the gas-lift port can be reduced, or a choke can be installed in the operating gas-lift valve or orifice to limitthe rate of gas injection to the design rate.

In some wells, casing can provide a common annulus for a dual or triple completion. In this case, one casing annulusis being used to serve more than one producing well, each with its own special requirements. Typically a dual gas-liftcompletion has no downhole control for each string. Small ports in the valves, flowing surveys in each string, and welltesting plus a gas measurement and control system will improve success for effective lift of both strings.



DRAFT

6.2 Tubing

Most gas-lift wells have production flowing up the tubing and gas is injected down the casing-tubing annulus. If thisinverse is true, most of the comments made here pertain to the casing and the comments made in 6.1 pertain to thetubing.


Verify the pressure integrity of the tubing by hydro-testing as the tubing is run into the well during completion orfollowing a workover. Dummies can be set in wireline mandrels or plugs set in the tubing to isolate valves and checksfrom excessive hydrostatic pressure.

— If the tubing is under or over-sized for the amount of fluid being gas-lifted, evaluate changing the tubing size toavoid inefficient and uneconomic operation. Consider potential reservoir formation damage during thisevaluation.

— If tubing corrosion is a problem, have the problem analyzed by a qualified corrosion engineer.

— If paraffin or scale deposition is a problem, establish a routine program to inhibit its formation or remove theparaffin or scale build-up. Severe deposits require the tubing to be replaced with a clean string so the damagedstring can be returned to a cleaning facility.

— Use safety valves only as required by regulations or for prudent operations and safety near roads, populatedareas, and marine environments.

6.2.2 Tubing String

Tubing must have good pressure integrity in the pipe body, connections, gas-lift mandrels, or other devices in thestring. Gas entering through a leak has the same effect as operating through an upper valve. Also, fluid can washthrough a leak causing erosion and string failure as well as potentially eroding a hole in the casing.

Tubing size is one of the most important selections in gas-lift design and optimization. A size too small preventsproducing the desired rate due to excessive friction pressure losses; a size too large reduces the rising fluid velocity,resulting in excessive fallback and higher FBHP. This may result in tubing heading, reduced production, or require ahigh rate of gas injection to maintain stability.

Tubing sized for the initial production rate, which was much higher than is currently available from the well, is acommon problem. The production performance of the well can often be improved by reducing the size of the tubing, ifthe workover and tubing cost can be justified. Note that tubing sizes smaller than 2.375 in. (6.03 cm) O.D. are seldomjustified, and frequently result in wireline difficulties.

Tubing size can be reduced either with a smaller diameter tubing string, or by running an insert string and then liftingup either the smaller coiled tubing string or the annulus between the original tubing and the insert string. This optionmust be carefully evaluated for operating costs, valve pulling costs, workover costs, insert string plugging, anddeposition of solids in the annulus.

Corrosion can be a severe problem; it can increase tubing roughness which can increase pressure losses due tofriction, and lead to tubing leaks or accelerated erosion at the gas entry point inside wireline mandrel pockets orconventional mandrels. One solution is to inject gas-dispersible inhibition chemical with the injection gas. Anotherremedy is batch treating down the tubing; a third option is plastic-coated or corrosion resistant alloy (CRA) tubing inthe well. CRA is expensive and used infrequently; if corrosion problems persist, consult a corrosion engineer ortechnician.



DRAFT

Paraffin deposition in the tubing is a common problem when the reservoir oil composition has a high wax content.Usually deposited at shallower, cooler depths, the paraffin restricts the flow area and reduces production. A cost-effective common remedy is to physically remove the paraffin by wireline scrapping and cutting; another option ismelting it with hot water or oil. Flowlines also have paraffin deposition problems and should be flushed with hotwater or oil.

Scale deposits may occur deeper in the well, often near the depth of gas-lift injection. Scale inhibition chemicals toprevent scale deposition can be injected in batch treatments or gas-dispersible chemical can be injected with gas-lift gas.

Safety valves are required in offshore flowing wells and may be required in gas-lift wells to prevent a blow out if thewellhead becomes damaged. They are used due to regulations or for prudent operation, even onshore near roads orpopulated areas, but they should be eliminated if the well is non-flowing, low-risk, and if the device is not mandated byregulation.

Installation of an annular safety valve is mandated in several regions of the world. These are installed to minimizeunwanted discharges at the surface (high pressure injection gas inventory held in the casing annulus during operationand/or leakage of reservoir fluids into the annulus) during a blow out situation.

6.3 Completion

“Completion” is the term used for the interface between the reservoir and the wellbore through which all of the fluidmust enter the well. Completion zone configuration options include open-hole, perforated pipe, wire-wrapped screeninside perforated pipe, gravel pack, or an artificially (chemically) consolidated interval.


The following practices are recommended for the completion.

— Use the wellbore schematics that document the completion used in the gas-lift well.

— Use pressure build up tests and skin analysis to evaluate excessive pressure losses across the completioninterval caused by clay swelling, plugging by formation material, asphaltines, scales, or waxes, reduced relativepermeability due to free gas saturation, or emulsion blockage.

— Compare the well’s measured productivity and inflow performance relationship (IPR) with theoretical calculations,with the historical trend on the well, and with those of similar wells in the field.

— Run SBHP and flowing gradient plus bottomhole pressure (FBHP) surveys, and associated well tests, annuallyor whenever a significant change in a well's productivity is observed or expected.

— Minimize pressure surges and heading if the well has a mechanical sand control system or the reservoirformation rock produces sand.

6.3.2 Completion Pressure Losses

Pressure losses in the production system should be minimized so the fluid can be lifted to the wellhead and throughthe surface piping to the production facility. Excessive pressure losses cause production rate to decline and gasusage to increase.

Excessive pressure losses during flow across the completion interval are particularly harmful since the well’sproductivity is reduced. Pressure loss occurs near the wellbore as the radial flow stream traverses the increasinglysmaller cross-sectional area, but excessive loss is termed “skin” and is caused by the following conditions.



DRAFT

6.3.2.1 Reservoir Conditions

The following reservoir conditions cause completion pressure losses:

— “free” gas saturation resulting from flowing bottomhole or reservoir pressures that are below the bubble point.Gas vapor impedes liquid flow and reduces relative permeability to oil near the wellbore;

— the advent of “free” gas production whereby the gas uses some of the native “permeability” and reduces therelative permeability to oil;

— water production, whereby the water uses some of the native “permeability” and reduces the relative permeabilityto oil;

— blockage of pore throats reduces permeability. This can occur when sand or clay fine particles are transportedfrom the reservoir, or rearrangement of sand grains is caused by high pressure drops and flow rates near thewellbore;

— swelling of antigenic (smectite and others) clay in contact with fresh water;

— deposit of heavy asphaltine hydrocarbons or chemical compounds.

6.3.2.2 Completion Design Problems

The following design problems cause completion pressure losses:

— incorrectly perforated reservoir zone, perforation debris, or perforation density (shots per ft) that is too low;

— a poorly installed sand control system, one with an excessive pressure drop and fluid velocity that can causeerosion, or an improperly sized screen plugged with fine material, heavy hydrocarbons, or chemical deposits;

— a plugged sand control device, with fine material, formation sand, heavy hydrocarbon deposits, or chemicaldeposits;

— inadequately evaluated fluids that cause reservoir rock plugging, or formation damage, such as incompatibledrilling, completion, treating, inhibitor, surfactant, or stimulation fluids;

— formation damage due to incompatible treating fluids, inhibitors, or surfactants used during sand controloperations, simulation treatments, etc.

6.3.2.3 Mechanical Sand Control Systems

Some important factors to remember about mechanical sand control systems are as follows:

— they can be damaged by excessively high pressure drops and associated high velocities and drag forces thatcan cause erosion;

— improperly implemented start up procedures that cause pressure surges (rapid changes in pressure drop) thatcan cause rearrangement of particles and plugging.

6.3.3 Steps to Minimize Pressure Losses

Surveillance testing should be performed routinely to find and evaluate sources and causes of pressure losses.



DRAFT

— Use frequent SBHP and FBHP surveys coupled with production tests to calculate the well’s inflow performance.Compare the inflow performance with theoretical calculations based on Darcy's Law, with historical inflowperformance data for the well, and with similar wells in the same reservoir. If reduced from before, or lower thansimilar wells, use a pressure build-up analysis to evaluate the skin and potential treatment to improve theproductivity and minimize the pressure losses across the completion interval.

— Low permeability could be the cause of low inflow performance, based on the pressure build-up analysis.Fracture stimulation may be required to improve well performance in low permeability reservoir formation rock.

— Skin and associated formation blockage, as determined by the pressure build-up analysis, may be reduced witha matrix acid stimulation. Consult a stimulation engineer and test the crude oil with proposed treating fluids toinsure that an emulsion is not created in the reservoir rock.

— Free gas reduces the relative permeability to oil and can hinder production. Gas is liberated in a depletion driveor gas cap reservoir, or when pressure drops below the bubble point. Both continuous and intermittent gas-liftcan be viable options to produce the well. Another option may be to cycle the well. The well can be produced,and then shut-in for a period of hours or days until the pressure builds. When the well is placed back onproduction it will produce until the gas saturation returns. This process of cyclic operation may actually increaseproduction while requiring less overall injection gas. The proper cyclic period would be different for each well andwould need to be determined by trial and error.

— If the poor productivity or IPR is due to induced or man-made factors, it is usually possible to correct the problem.A workover or stimulation job may be required and these require consultation with the appropriate engineeringstaff.

6.4 Wellhead

The wellhead is designed for safety and well control, but not necessarily for effective gas-lift operation.


The following practices are recommended for design of the wellhead.

— Design and maintain the wellhead to minimize flow restrictions and pressure drops.

— Do not use chokes on gas-lift wells; remove choke nipples and old choke bodies if they restrict flow.

— Provide a crown valve (swab valve) for easy access to the well for wireline work. It may be advisable to have apermanently installed lubricator on the well or readily available. Provide easy access to the annulus to permit gasinjection and circulation of the well.

6.4.2 Factors for Good Wellhead Design

— Minimize flow restrictions and pressure drops.

A pressure increase of 1 psi (6.89 kPa) at the wellhead may be magnified to 2 psi to 5 psi (13.79 kPa to34.47 kPa) pressure increase at the perforations. If the PI of the well is 0.5 bbls/day/psi (0.012 m3/day/kPa), theproduction rate is reduced by 0.5 barrels (0.08 m3) per day for each 1.0 psi (6.89 kPa) increase in the operatingbottomhole pressure. Eliminating pressure restrictions can increase production, especially from high PI wells.

Revise the natural flow wellhead by removing restrictions such as chokes and choke bodies (if they represent asignificant restriction). Master and wing valves should be full opening and never partially closed. Excess piping,90 ° elbows, and swedges should be minimized or removed. The objective is a smooth flow path from the tubingto the flowline with minimal pressure losses during normal operation.



DRAFT

— Provide for easy well entry.

Frequent wireline operations are required to obtain flowing pressure surveys and maintain gas-lift valves. Tofacilitate easy access to the well with minimal disruptions, the wellhead should be equipped with a crown valvefor easy installation of a lubricator, which will allow wireline access to the well without shutting it in.

— Provide for effective wellhead measurement.

Wellhead measurements are discussed in 6.5. The wellhead itself must be designed to permit thesemeasurements safely and effectively.

6.5 Wellhead Monitoring and Control

Pressure monitoring of the casing-head and tubing-head pressures is required for effective gas-lift operation.


The following practices are recommended for wellhead monitoring and control.

— Measure the wellhead injection (casing) and production (tubing) pressures and record on a two-pen chart or in anautomated data gathering system, especially during well tests and pressure surveys. The data can be used toevaluate stability and is needed for optimization.

— Be aware of other well-site measurement needs and provide for them when they are required.

6.5.2 Wellhead Measurements

Three wellhead measurements are required and others may be valuable.

— Casing-head pressure.

The casing annulus pressure should be measured at the wellhead, downstream of all input control chokes orvalves. Pressure versus time should be recorded using a two-pen chart or an electronic measurement systemusing a sensor and transmitter. This data is required for a gas-lift model, is used to estimate the depth of theoperating valve, and is used to evaluate casing pressure stability.

— Tubing-head pressure.

The tubing-head pressure should be measured at the top of the wellhead, upstream of any choke bodies or otherrestrictions. Record this pressure versus time on the same two-pen chart or electronic system as the casingpressure. This data is also required for the computer model calculations of tubing pressure and flowingbottomhole pressure, as well as to evaluate tubing pressure stability.

— Gas-lift injection rate.

This measurement is accurate when gas flow is steady, the orifice tube and plate are properly sized, and thechart recorder is calibrated. Or it can be determined when an electronic flow rate computer is used. Injection gasmeasurement is required for control, optimization, and modeling to estimate operating valve depth and FBHP:

— the injection rate must be determined, and integrated over time, to know how much gas is being injected;

— the injection rate is a variable in any model of the producing performance to determine the operating gas-liftvalve and FBHP.



DRAFT

— Other measurement.

— Multiphase meters.

These meters are sometimes installed at wellhead manifolds on offshore platforms to improve test frequencyor to provide testing capability. Pressure and temperature differential are being used to estimate productionrate in locations where well testing capability does not exist.

— Other production rate meters.

Three-phase metering systems are available to measure oil, water, and gas. In some cases, it may bepractical and economic to install such a meter on each gas-lift well.

Or, it may be cost effective to install one meter to serve multiple wells, with the wells “switched” to use the 3-phase meter on a periodic basis.

It is also possible to estimate production based on differential pressure across a restriction for wellsurveillance purposes. This process is being used effectively by some operating companies.

— Temperature.

Steady-state temperatures can be helpful for estimating wellbore temperatures used in valve designcalculations. Wellhead temperatures obtained after shut-in periods can be used to estimate the transienttemperatures at each unloading valve. Hot steady-state temperatures are obtained during the flowingsurvey, and geothermal temperatures can be obtained when the well is shut-in for an extended period. Thetransient temperature is between the hot steady-state and geothermal temperatures. Also, steady statewellhead temperature is an indicator of production rate and water cut, and can be used by the operator toinfer changes.

The steady-state producing temperatures that are measured with a wireline gauge are not equal to thetransient temperatures that are “seen” by the shallow unloading valves during the initial unloading process.These “unloading” temperatures are different than the steady state producing temperatures and thegeothermal temperature at the depth of the unloading valves. They may be between the steady statetemperature and the geothermal temperature, or they may be lower than the geothermal temperature whilecold annular liquids are being displaced downhole in cold climates

Another reason to measure the temperature is it may be correlated to the production rate. This correlation isnot well understood, but in some cases gas-lift operators are able to infer changes in production rate fromchanges in producing temperature.

While there may be little need to continuously measure the producing temperature, it is important to measureit from time to time. The reason is the gas-lift valves, which have nitrogen-charged bellows, are sensitive tothe producing temperature. Most temperature models are simple and a measurement of the producingtemperature, at different producing conditions, is necessary to calibrate the temperature model. It is alsoimportant to measure the producing bottomhole temperature. This is discussed in 10.3.

— Surface casing pressure.

Surface casing pressure should be monitored periodically for safety and environmental protection to assurethat a casing leak has not occurred.



DRAFT

In some cases, it is important, or required by law or good practice, to monitor the surface casing pressure.This has no direct operational use, but is done for safety and environmental protection to assure that acasing leak has not occurred.

— Safety systems.

Surface safety systems, surface controlled gas-lift valves, batteries, or solar panels are other wellhead itemsthat may require measurements.

— Surface-controlled gas-lift valves.

Some companies use gas-lift valves controlled from the surface, using either an electrical or a hydrauliccontrol signal. Typically, it is possible to control the position (opening) of one or more downhole valves fromthe surface. Conventional valves have one port size with the capability to open, throttle, and close. Thesesurface-controlled valves allow the port size to be adjusted remotely from surface over a range of specific,desired positions from fully closed to fully open. In some cases, it is also possible to measure the actualopening of the valve(s) in real-time and the downhole pressure at the valve(s). This technology will likelyhave wide application, especially on high-rate deep-water or sub-sea gas-lift installations.

— Various other items.

There are other items that may need to be measured or monitored at a gas-lift well site, such as: the batteryvoltage of a solar-power pack or an aid-to-navigation system, or the air pressure on a pneumatic controlsystem.

7 Gathering System—Flowline and Manifold7.0 Purpose

Flowline piping is a necessary component for gas-lift wells. An offshore platform may have less than 100 ft (30.48 m)from the wellhead to the manifold. Onshore, inland marine, or shallow water fields may have 1,000s of ft (100s ofmeters) to the production station. Except for single-well fields or leases, most gas-lift wells produce into a manifoldwhere several wells come together to be switched into either the well test system or the bulk production separator.

Additional gas-lift gas in low pressure surface flowlines and manifolds can significantly increase friction loss, thereforemonitoring and pressure measurement to find and eliminate excessive loss is an important surveillance task.

7.1 Flowline

The flowline transports the produced oil, water, and gas from the well-site to a manifold where the production can berouted to a well-test or production separation system. One-well leases may not have a manifold.


The following practices are recommended for flowline transports.

— Size flowlines to be too large, rather than too small, when initially sizing for gas-lift. Exceptions to over sizing arelong subsea tiebacks or other long production lines where severe slugging may adversely impact the receivingfacilities:

— flowlines that are too small may be subject to erosion or plugging, especially if the well is slugging orproducing solids or paraffin;



DRAFT

— flowlines that are too large can allow solids to settle out; this can cause corrosion under the deposits andmay cause line plugging.

— Monitor the pressure losses through the flowline and take corrective steps to clean the line if measured upstreampressures are significantly greater than pressures calculated with a computer model.

— Verify flowline pressure integrity with hydrostatic tests. Offshore flowlines should be rated to shut-in wellheadpressure to avoid possible leaks or failures.

— Avoid producing more than one well through a flowline at the same time.

— Use a computer optimization model that includes flowline pressure losses to avoid erroneous results when thegas injection rate is varied.

— Include hilly terrain and elevation changes, such as risers, in flowline modelling since multiphase flow is affectedby gas that collects in the high spots and liquid that settles in the low spots.

— Any extra pressure losses in the flowline that result in a higher wellhead backpressure, which in turn results in ahigher operating bottomhole pressure and less production inflow from the reservoir.

7.1.2 Flowline Considerations

Flowlines should transport gas, oil, and water from the well to the production station with minimum pressure drop andwithout deposits of sand, paraffin, or scale accumulating in the pipe.

— Flowline size.

The pressure drop through a flowline is directly related to its effective cross-sectional area and the flow rate ofgas and liquids. Small flowlines, or those plugged with solid deposits, create an excessive pressure drop thatraises wellhead and wellbore pressure, which diminishes effective gas-lift operation. A small flowline may sufferfrom erosion, especially if the well is slugging or producing solids.

If a flowline is too small for a well’s production rate, a parallel line, larger line, or nearby inactive flowline can beused to match well and flowline capacity.

If a flowline is too large for the well, slugging can occur and cause testing problems, or solids can settle in lowspots causing plugging or corrosion under the deposits.

— Flowline length.

Flowline length is dependent on the number and location of production stations. The pressure drop through aflowline is directly related to its length, and the flow rates of gas and liquid. There is usually no way to affect thelength of the line. If a flowline must be long (some may be several 1,000s of ft (100s of m) or miles (kilometers)long), it is important that an adequate flowline diameter be used.

Flowline elevation changes, such as hills, or well jacket to sea bed to production platform, should be included inmodelling since any change in elevation may increase pipeline slugging and pressure losses.

— Flowline cleanliness.

Remove sand, paraffin, or scale by periodic pigging (large diameter lines) or flushing with hot water or treatingfluid.



DRAFT

— Pressure monitoring.

Monitor the pressure drop from the wellhead to the manifold, especially for long flowlines, or if there are paraffinor sand problems. Take corrective steps if the pressure loss becomes too large. Normally, it is sufficient tomonitor the producing wellhead pressure. If it is too large, relative to the separator pressure, the difference mustbe the pressure losses in the flowline and manifold. However, in some cases, it might be worthwhile to monitorthe pressure drop across the flowline itself, by monitoring the wellhead pressure and the pressure at the inlet tothe manifold.

— Pressure integrity.

Pressure test flowlines on gas-lift wells to the gas-lift injection pressure. This possibility exists if some failure wereto occur and full gas-lift system pressure is applied to the line. Many times, flowlines are quite old, since gas-liftmay be applied late in a well’s life. Therefore, it is important that the pressure integrity of the flowlines be checkedfrom time to time, for safety and environmental protection reasons.

The pressure design basis for flowlines should be ANSI/ASME B31.8 for the maximum working pressure forvarious pipeline materials and diameters. Also, the requirements of API 14C for wellhead safety systems shouldbe considered. This specifies safety systems for various flowline services.

— Multiple wells in one line.

”Temporarily” producing two or more wells through a common flowline may be expedient, but should be avoidedsince the wells cannot be tested independently and the pressure losses may reduce production. This should beavoided for the following reasons:

— the additional flow through the line will necessarily increase the pressure losses in the line;

— the only way to conduct a well test on one of the wells will be to temporarily shut down the other well(s).

— External corrosion.

External corrosion problems should be minimized with coated and wrapped pipe, or in severe cases withadditional protection such as cathodic protection or a sacrificial anode system.

— Flowline model.

Flowlines should be in the computer model that includes the reservoir, wellbore, and separator. As gas-lift gas isadded to reservoir fluid, the wellhead backpressure can increase significantly due to friction from the increasedvelocity. A model without the flowline component can give a large error during optimization calculations.

Flowline pressure loss calculations compared to measured results can indicate restrictions in the flowline, suchas paraffin, scale, or sand deposits that would require some corrective or preventive action such as pigging orflushing.

— Any attempt to optimize gas-lift production with a computer optimization model must take into account theeffects of the flowline. As the liquid and/or gas flow rate increases, the pressure losses in the flowline alsoincrease. This results in a higher wellhead backpressure for the higher flow rates. The optimization programmust be able to take this increased backpressure into consideration.

— A model can be used to determine the restriction to flow that exists in the flowline, and the need to take somecorrective or preventive action such as pigging, flushing, etc. If the actual pressure losses are higher thanpredicted, this may indicate a reduced effective flow area, which may be the result of partial plugging of the line.



DRAFT

A flowline model should include bends, restrictions, risers, changes in terrain, and other major factors, such aschanges in elevation, within the limitations of the computer program.

7.2 Manifold

The manifold (or header) permits the production from several wells to be routed to either a well test or a productionseparation facility. Several production separators may be in the facility—for high pressure flowing wells, intermediatepressure flowing wells, and low pressure flowing and gas-lift wells. Gas-lift wells should always be routed into lowpressure test and production vessels.


The following practices are recommended for the manifold:

— minimize any pressure losses in or through the manifold;

— keep all manifold valves fully open or fully closed;

— check periodically for manifold valve leaks using sonic or infrared detectors. Multi-ported valves should bechecked frequently;

— use automatic computer-actuated valves to permit automatic well testing.

7.2.2 Manifold Considerations

The following are manifold considerations.

— Manifold size.

The piping and valve sizes and configurations (bend, tees, 45s) in the manifold should minimize pressure losses.

— Flow restrictions.

Manifold valves should be full opening and should always be kept either fully open or fully closed. Chokesinstalled on the flowline upstream of the manifold should be removed from gas-lift wells. Choke bodies should beremoved unless a choke may be required to prevent severe slugging that would destabilize separator orcompressor operation.

— Valve leakage.

Manifolds contain many valves, and often the valves are multi-ported which creates a high potential for valveleakage. Leakage invalidates well test information needed for optimization and reservoir management. Gas-liftwells that still have valves connected to high or intermediate pressure headers should be disconnected or haveblinds installed to prevent leakage, since the pressure differential across the manifold valves can be substantial.

Periodic infrared or ultrasonic testing with portable devices is the easiest option, but another method is switchingall wells to the production separator and monitoring the well test separator for indications of flow. Automated two-way or three-way valves should be rigorously monitored since they are more susceptible to leakage.

— Ease of switching from production to test.

Operations staff should be consulted for advice on methods to switch the wells between the productionseparation facility and the test facility. This applies to manual, automatic, or semi-automatic well testing.



DRAFT

8 Well Production Rate Testing8.0 Purpose

Well production rate testing is required to obtain accurate information for operation, maintenance, troubleshooting,and optimization of gas-lift installations. It should be conducted monthly or more frequently if possible.

Flowing gradient pressure surveys coupled with production rate testing, injection gas measurement, and tubing/casing pressures, should be conducted annually to provide the data needed for well performance analysis.

8.1 Well Test Scheduling

When and how to test the production rates of each gas-lift well are important decisions, since in most fields the welltest is the only time when there is a real measure of the well's production rate(s). Well test scheduling depends onregulatory requirements, the number of test devices (separators or multiphase meters) available, well stability, thetime required per test, and changing reservoir conditions.


The following practices are recommended for well test scheduling:

— test each well monthly or more frequently to establish its operating performance and gas-lift efficiency. Stablewells with relatively steady reservoir conditions may be tested less frequently;

— test at normal production conditions until the production rate and water fraction are stable long enough to obtainaccurate results, but not so long as to waste testing time. Typically, the minimum acceptable test duration is fourhours;

— avoid testing when the well or production system has been down for remedial treatments or maintenance; waituntil stability has returned. Coordinate well tests with other well activities, so wells are only tested when they arein a normal, testable condition, and so well test information is collected in conjunction with other wellmeasurements, such as flowing pressure surveys;

— use automatic well testing to improve data quality and analysis, and to easily switch wells into and out of test.

8.1.2 Well Test Objectives

The five fundamental objectives for testing gas-lift wells may require different well test procedures to obtain thedesired information are as follows.

— Objective #1: Determine the well's production and injection gas rates.

Test the well as it normally produces, either in steady flow or in fluctuating, unstable flow. Measure the rates of oil,water, total recovered gas and the associated injection gas, injection casing pressure, and production tubingpressure. To determine these values, the well test meters must be accurate and working properly, the well testsystem must be thoroughly purged before the well test begins, and the well must be tested long enough andunder the right conditions (see below) to obtain accurate measurements. However, since the well's water cut,gas/oil ratio, and inflow performance may not change rapidly, it may be sufficient to perform such a test once amonth or even less frequently.



DRAFT

The well’s production and injection data are used for several reasons.

— Determine the economic contribution of each well based on production revenue and operating cost.

— Optimize and/or allocate injection gas to the well.

— Allocate total field oil, water, and gas production to each well based on its well test and downtime, comparedto the sum of the well tests and total field production.

— The pressure of the test system may be different than the pressure of the bulk production system and thesum of well tests can vary from the field production. If possible, the test separator and production separatorshould operate at the same pressure.

The objective here is to evaluate the well’s current production potential, not the effectiveness of the gas-liftsystem. That comes next. For this test, the well should be produced as it normally does. If it is normally unstable,it should be produced in this same way, so the well's normal production is determined.

— Objective #2: Check the gas-lift performance.

Check the current performance of the well and gas-lift system with frequent tests of short duration so thatcorrective action can be initiated if a problem is detected. Here, the absolute measurement of oil vs. water is notimportant, as long as the total liquid production rate can be determined along with the rate of gas injection,recovered gas, and injection and production pressures. To check these values, it may not be necessary to purgethe well test system and it may only be necessary to conduct a short test if the well is stable, or over only a fewheading cycles if it is unstable.

It may be pertinent to conduct the two different types of well tests on a routine basis. To check the performance ofthe well and gas-lift system, it may be pertinent to conduct short tests or checks as frequently as possible. Then,to measure actual production rates and associated variables, it may be pertinent to conduct longer tests on a lessfrequent basis. In this way, the performance of each well can be checked more frequently, and the actualproduction rates can be measured more accurately. The normal approach, which is to test each well just longenough to obtain accurate readings, means that performance checks may not be performed as often as desired,and production measurements may not be determined as accurately as desired.

— Objective #3: Check the well and its gas-lift performance with a pressure survey.

Check the well’s production and injection rates (the same as the first objective) in conjunction with a flowingpressure survey. The objective is to associate gas-lift performance with gas-lift valve (or valves) injection depth.Is the well optimized with deep injection, or inefficient due to shallow injection? If the well is unstable, is thiscaused by a too large port or orifice, by multi-pointing, or by a leak in a valve or mandrel? The well should be ontest while the flowing pressure survey is being run (or within one day of the survey), and the downhole flowingpressure must be measured at the valve depths.

The flowing gradient survey and production test data are used to validate a computer model for the well,including a multiphase outflow correlation.

— Objective #4: Determine the well's inflow productivity.

Determine the well’s inflow productivity from the reservoir to the well, calculated from the production rate plus theflowing and SBHP surveys. The objective and process are different from the previous test. For this test, the wellmust be producing in a stable manner. This test is used to gather data for several reasons:

— evaluate the well's inflow and outflow performance;



DRAFT

— determine possible skin restriction to inflow if a pressure build-up survey is conducted between the flowingand static pressure surveys;

— evaluate the reservoir processes, especially secondary recovery (e.g. water flooding or gas injection) withhistorical data comparison of pressures and productivity;

— calibration of the vertical pressure profile model used for the well.

The goal is to measure the inflow productivity, and not the actual performance of the well or its gas-lift system;thus the well must be producing with a stable flowing bottom hole pressure. If the well is not normally producingin a stable manner, it must be stabilized using appropriate measures, such as temporary over injection before thetest is conducted.

— Objective #5: Determine gas-lift response.

Determine the well’s gas-lift response by conducting a “multi-rate” test, where the well is produced at severaldifferent gas-lift injection rates and the associated production rate for each injection rate is measured. The plot ofproduction rate versus injection rate is known as the gas-lift performance curve and is used in the gas-liftoptimization process after deep injection has been achieved using objective #3 methods.

The remaining parts of this section deal with the “normal” approach to well testing; that of accurately determiningthe well's current production rates.

8.1.3 Well Test Frequency

The necessary well test frequency is related to changing well, gas-lift system, and reservoir conditions. Steady, slowlychanging reservoir pressures or water fractions permit less frequent tests, and a stable, high pressure gas-lift systemwith adequate gas capacity will reduce the number of tests needed for optimization. Test frequency should beconsistent with the need to:

— test each well for the period of time needed to obtain repeatable production and injection rate data;

— coordinate testing frequency and test time with other well and facility operations;

— perform routine maintenance on the well test facility;

— have operating personnel available to perform the test operations;

— avoid severe weather that may endanger personnel and invalidate tests.

Several external factors may impact well test frequency, as follows.

— Number of wells served by the well test facility.

The number of wells connected to a test facility and the number of facilities in a field influence test frequency.These reflect the conflict between the desire to conduct frequent well tests and the desire to minimize the cost ofwell test facilities. For monthly well testing with a 24-hour test duration, the number of wells served by anindividual well test facility should be limited to 20, so there is time for retesting when needed. Use of automaticwell testing is recommended to handle the tests in a reasonable time period. (This is discussed further below.).

— Mandated well test schedules.

Mandated tests required by government regulations, company policies for obtaining production data for reservoirmanagement, or accounting requirements may determine frequency.



DRAFT

Stable wells in a mature field may not need the same test frequency as a well that is unstable or has a history ofoperating problems. Mandated schedules should be “adjusted” if rules or policies and operational conditionspermit.

— Special tests.

Special tests are conducted after well work has been performed, or a re-test is needed to determine if a“questionable” test was due to an operational problem with the well test facility or was indicative of a wellproblem.

8.1.4 Well Test Duration

Wells should be tested at normal production conditions until the production rate and water fraction are stable for longenough to determine their values. The following points affect test duration.

— Well production rate.

Higher rate and lower rate wells do not need the same test time.

— Higher rate wells tend to be more stable and purge facilities more quickly. Production rate and water fractionconsistency may be reached in less time. Historical data should be monitored closely since performancedegradation could be economically important.

— Lower rate wells may be unstable with slugs of liquid and gas; they may require time to re-stabilize afterswitching from production to test facilities; and they may need time to purge the previous well’s fluids fromthe test system. This may required a total time of 24 hours or more. However, performance changes mayonly have a small economic impact.

It may be possible to test a higher rate well for only a few hours, whereas it could require days to accurately testa lower rate well.

Offsetting this must be the value of the well being tested. It may be important that a high rate well be testedaccurately, since even a small change in its performance may mean the difference of several barrels (m3) of oil.Conversely, even a significant change in a lower rate well may not mean much from an economic point of view.

In addition to considering the overall production rate from the well, the oil cut must be considered. If a well producesa high rate, but if 95 % + of this is water, it may require several hours to obtain an accurate measurement of the oilproduction rate. However, it may not be necessary to accurately determine the oil cut (or water cut) very often. If thetotal fluid rate remains stable, and if it can be assumed that the oil cut remains essentially the same over the shortterm (weeks or months), it may be possible to conduct frequent short-duration tests to check on overallperformance, and periodic longer-duration tests to accurately determine the well's oil cut.

— Well stability.

Well rate and pressure stability aid test accuracy and reduce time, especially if continuous data collection is used.In some cases, it is possible to sample a well’s production rate instantaneously to obtain a reasonable well test.Since most well test facilities do not permit instantaneous measurements or the gathering of samples, butactually measure the volume produced over a period of time, it is necessary to test even the most stable wells fora few hours.

Unstable heading or intermitting wells require longer test times. If a well is producing with consistent cycles, theduration of the well test must be an even multiple of the cycle duration.



DRAFT

— Well test switching method.

Computer controlled data gathering and automatic switching permits the test duration to fit the conditions of eachwell. Monitoring the well's production rate during the test permits ending the test when the rate has stabilized.

Manual well test switching links test duration to the availability of the people to perform the switching operationrather than to well characteristics.

8.1.5 Well Test Coordination

Wells tests should be coordinated with related well, field, or facility activities. These related activities can occur in twocategories as follows.

— Related well operations.

— Flowing pressure (and/or temperature, and/or flow meter) surveys that are used for gas-lift well analysis.

— Maintenance work or batch chemical treating that would interrupt well production.

— Changes to the operation of a well which change its production rate require time until the well has stabilizedafter the change.

— Changes intended to increase rate or optimize gas-lift gas should be followed by a series of tests to evaluatethe change.

— It is necessary to coordinate data gathering activities. A well test on a gas-lift well is not complete unless theinjection rate, injection pressure, and production pressure are also measured during the test. This mayrequire coordination to be sure these items are measured and collected during the test.

— Unplanned well activities/problems.

Activities and/or problems can occur that will invalidate the test. If tests are conducted manually, this may beunavoidable. If an automatic well test (AWT) system is being used in conjunction with a system to monitor thewells, it may be possible to observe and respond to these types of events.

— A well is shut in or dies.

— A control parameter, e.g. injection rate, is changed.

— A change at the manifold, e.g. another well switched into test, or excessive valve leakage.

— A change of lift gas rate in an automatically distributed system due to a compressor start up or shutdown, ora production station shutdown. In such a system, maintain a constant “test” injection rate into the well(s) thatare being tested. The change in supply or demand should be absorbed by the other wells in the system.

8.1.6 Automatic Well Testing

Automatic well testing (automatic switching, data gathering, and processing) permits well switching to occur at anytime. Manual or semi-automatic testing (manual switching combined with AWT data gathering) is constrained bypersonnel and weather since the wells can only be switched into and out of test when people are available.

— AWT systems allow test durations to be optimized to be long enough to give accurate results, but not so long asto waste test time. With shorter time periods, wells can be tested more frequently.



DRAFT

— Test results are more accurate since meter data is gathered continuously and results are calculatedautomatically.

— Test results are more complete if wellhead data for injection rate, injection pressure, production pressure, andtemperature are gathered at the same time.

— Test data can be aborted if problems exist before or occur during the test.

— Special tests can be scheduled.

— Test results can be immediately reported and plotted to give operating and engineering personnel betterinformation on well performance.

— Test results can be more easily integrated with models of the gas-lift well and/or system for evaluating wellperformance and optimizing gas-lift distribution.

8.2 Well Test Equipment

Well test equipment, either separators or multiphase meters, should be operated and maintained properly to obtainusable gas-lift well performance measurements.


The following practices are recommended for well test equipment:

— include purge time to flush the previous well's fluid out of the piping and the test facility;

— maintain the separator backpressure in the test system the same as the production system;

— use chemicals and heat in the test system the same as in the production system;

— take samples from the oil outlet during the test to quantify the oil-water emulsion content or to confirm the waterfraction from a two-phase vessel;

— size well test meters properly, or use dual outlets including both meters and outlet regulator valves when ratessignificantly vary. Apply routine maintenance and calibration;

— use three-phase rather than two-phase test separators, if possible;

— use multiphase meters for platforms or remote locations not suitable for test vessels.

8.2.2 Equipment Details

Well test facility components; the separator, treater, heater, chemical injection pump, or multiphase meter, providerates of the separate components—oil, water, and gas. Production separators may only separate liquid (oil + water)from gas, or perhaps free water, oil emulsion, and gas. However, well tests require accurate measurement of all threephases whether three-phase or two-phase test separators, or multiphase meters are used. Equipment details includesizes, instrumentation controls, and chemical treatments.

Important equipment considerations are as follows.



DRAFT

— Size of separator.

Vessel diameter should be sized to handle the highest rate of gas-lift plus reservoir gas and the vessel lengthshould give the volume needed for liquid residence time to permit gas-liquid separation and oil-water emulsionbreaking in three-phase vessels. Other well test considerations are as follows.

— Purge the prior well’s fluids from the piping and the separation facility to improve the accuracy of data for thewell being tested. The purge time is a function of the production rate of the current well on test, not theprevious well.

— Maintain and calibrate the level controllers to provide gas-oil and oil-water interface stability. In mostseparation facilities, these devices control when the various fluids will be “dumped” past their respectivemeters.

— Install the meters to operate at separator pressure and place the control “dump” valve downstream of themeter.

— If level controllers are used, it may only be necessary to purge the system long enough to obtain two or more“dumps” to assure that fluid from the new well is being measured.

— Separator backpressure.

The test separator pressure should be the same as the production separation pressure so the well will produceagainst the same pressure and not have to re-stabilize. If a higher pressure drop through the production manifoldand separator inlet piping create more wellhead backpressure, test vessel pressure can be adjusted. Ifproduction must flow from the test separator outlet into the inlet of the production separator, test pressure may besignificantly higher than the production separator pressure.

— Use of chemical and heat.

Add chemicals or heat to obtain three-phase separation or to maintain consistent practices in both productionand test.

— Calibration and size of meters.

Many different types of meters are used to measure oil, water, and gas production. These meters must beroutinely checked, maintained, and calibrated to produce consistent, accurate test results. Size the meters for theexpected flow rates and, if necessary, use two different sizes, one for high rate and one for low rate wells withcorresponding outlet control valves sized for each rate.

— Size of multiphase meters.

Multiphase meters should be capable of processing the large gas fraction to be expected from low pressure gas-lift operations where gas is the dominant fluid. High and low rate wells must be tested within the manufacturer’sdefined accuracy.

8.3 Well Test Measurements

Well test measurements are obtained from the combination of the meters or sensors that measure the data to becollected for evaluation of gas-lift well performance.


The following practices are recommended for well test measurements.



DRAFT

— Collect data accurately and thoroughly review for correctness, completeness, and application to gas-liftperformance evaluation since test measurements require significant effort and expense. Eliminate data that is notpertinent or useful so that staff can concentrate on items that aid optimization.

— Use three-phase separators, net oil computers, or mass meters to measure the oil and water production. If two-phase test separators are used, sample the liquid to determine the percentage of oil and water in the producedliquid; however, this process is usually less accurate.

— Use multiphase meters to obtain gas, oil, and water test measurements where long test lines or lack of testvessels prevent accurate tests.

— Correct oil and gas test measurements from the separator pressure and temperature to standard conditions; testresults should always be reported at standard conditions.

8.3.2 Test Data

Methods to gather the needed information vary based on the operators’ equipment in production service and theinstallation of automated data collection.

— Batch test record.

The volumes of oil, water, produced gas, and injected gas are measured for the test period, often by obtaining thedifference between the “final” and “initial” meter readings. These volumes are converted to 24-hour rates with asimple time ratio. Either the average injection and production pressures, or “representative” pressures, aregathered during the test period.

This method gives the average production (and injection) rates during the well test period. If the well is stable, thismethod is sufficient, but if the well is not stable, these average rates may not represent the well's actualperformance.

— Continuous test record.

Continuous (orthochronistic) test data collection provides rate and pressure variation throughout the test; this canbe useful in evaluating the time for the well to become stable and in understanding and diagnosing unstableperformance. This method is applicable to automatic or semi-automatic well testing.

If a well test facility is operated manually, continuously obtaining rate versus time is impractical. Even if a facility isoperated automatically, the batch “dump” volumes of oil or water through a meter do not represent the productionrate until they are averaged on a longer-term basis.

There are intermediate alternatives that may be useful. If a well is being tested for 12 hours, the test data can begathered in four three-hour sets, six two-hour sets, or twelve one-hour sets. In this way, there can be four, six, ortwelve “mini” well tests which may give an idea of the variability of the well's performance, while overcomingsome of the limitations of true instantaneous metering. If this method is used, the durations of the “mini” testsshould be related to the duration of the well's cycles if it is heading or intermitting.

— Wellhead data.

Wellhead measured data should include injection rate, injection pressure, production pressure, and productiontemperature, gathered at the same time as the rate data. These values should be gathered throughout theduration of the test so that appropriate averages and variations can be obtained.



DRAFT

— Data measurement accuracy.

— Oil should be recorded to the nearest tenth of a barrel per day, if the total oil production rate is less than50 barrels (7.95 m3) per day. Otherwise, it should be recorded to the nearest barrel per day; with ameasurement accuracy of ± 5 %.

— Water should be recorded to the nearest barrel per day; with a measurement accuracy of ± 5 %.

— Recovered (injected + produced) gas should be recorded to the nearest five standard cubic ft per day(0.14 m3/day), if the total rate is less than 500 MCF (14,158 m3) per day. Otherwise, it should be recorded tothe nearest ten MCF (283 m3) per day; with a measurement accuracy of ± 5 %.

— Injected gas should be reported and measured to the same accuracy as recovered gas.

— Produced reservoir gas is obtained by subtracting injected gas from recovered gas. However, since themetered volumes of recovered gas and injected gas are normally large compared to the volume of producedgas, a meter error in either can cause a large error in the calculated produced gas. If the recorded producedgas rate is not realistic or is a negative number, an alternative approach is to estimate the produced gasbased on the oil volume and a gas/oil ratio determined from a PVT analysis.

— Injection and production pressures should be recorded to the nearest psi (6.89 kPa), although it is probablynot possible to be accurate to more than the nearest 10 psi (68.95 kPa).

— The total production, based on well tests and downtime, should match with actual measured field productionwith an accuracy of ± 5 %. Greater variation should require an investigation into the causes.

— Well test rates of oil, water, and gas should be converted to standard barrels (m3) and standard cubic ft (m3)from separator pressure and temperature based on the PVT data for the reservoir crude oil. Processsimulation may be required to add the gas-lift to the reservoir fluid composition and obtain the newcomposition with its properties.

— Other well test information.

A number of items should become part of the well test record if they are known, or if they are measured with the test.Several of these items, such as the gravities, are determined infrequently, with special tests:

— date;

— time of well test completion;

— test duration;

— number of cycles during the test, if well is heading;

— range of injection pressures, if well is heading or cycling;

— range of production pressures, if well is heading or cycling;

— pressure base used for calculating standard gas volumes;

— oil API gravity.

— water specific gravity;



DRAFT

— recovered gas specific gravity;

— recovered gas temperature;

— injected gas specific gravity;

— injected gas temperature;

— wellhead production temperature;

— wellhead choke size, or choke body size, if any is used;

— type of well control (continuous—choke, continuous—automatic, intermittent—choke, intermittent—time-cycle controller);

— indication of the degree of sand production, if any;

— indication of the degree of paraffin production, if any;

— indication of amount of H2S or CO2 production, if any;

— estimated actual depth of gas-lift injection, if this is available;

— estimated operating bottomhole pressure, if this is available;

— comments pertaining to recent well operations, unique conditions, etc.

9 Production Handling System9.0 Purpose

The production handling system gathers, separates, treats, and transports the oil, water, and gas produced by gas-liftwells. It must be properly sized and operated to permit effective gas-lift operation.

9.1 Oil Handling System

Handling, treating, and measuring produced oil is top priority since this is the primary source of income. When oil isnot treated and sold on the location where it is produced, well testing for oil and gas production rates and waterfraction becomes more important since these tests are the basis for allocating the total oil, gas, and water measuredat the central treating site. Accurate well test measurement improves allocation of the total oil production back toindividual wells. The allocation problem may be compounded if three-phase separation or multiphase meters are notused. Gas-liquid separation with the combined oil/water mixture transported to the treating facility, or free waterseparation with the remaining oil/water emulsion transported for treating, can reduce test accuracy and the accuracyof subsequent allocation.

Gas-lift well performance evaluation depends on well tests. However, gas-lift optimization and reservoir managementalso require knowledge of the total production from the well and the average production rate over time in order toevaluate changes in performance and to evaluate the economics of any proposed well or gas-lift system work.


The following practice is recommended for the oil handling system.



DRAFT

— Obtain accurate well test measurements for each well for allocation of the daily oil production (sales) back toeach well. Compare the total production to the estimated production based on well tests and downtime andinvestigate any cases where the two differ by more than 5 % to 10 %.

9.2 Water Handling System

Water production measurement and treatment are as important in gas-lift optimization and reservoir management asis oil production. However, in most oil fields, produced oil is measured accurately, while produced water is notmeasured accurately and in some cases not measured at all. Since accurate evaluation of the well, gas-lift system,and reservoir performance depend on the total fluid production—gas, oil, and water—produced water must bemeasured during well tests with the same effort as produced oil.

Water production metering is important but does not require the custody transfer measurement accuracy used for oil.Reasonable care to provide total liquid rates and volumes is sufficient so that production facilities can be properlysized and maintained and reservoir voidage can be determined.


The following practice is recommended for the water handling system.

— Obtain water measurements for each well so that total water production from all wells in the facility can beallocated back to each well. Compare the total production to the estimated production based on well tests anddowntime.

9.3 Gas Handling System

Recovered gas (total of reservoir plus gas-lift gas) should be measured with methods similar to those for produced oiland water. Total gas for all wells in the facility should be allocated back to the individual gas-lift wells based on the welltest.


The following practice is recommended for the gas handling system.

— Obtain measurements of total gas recovered from each well and injected gas for each well so that the total gasproduction for the facility can be allocated back to each well. Compare the total production to the total injectiongas and the estimated production based on well tests and downtime.

10 Guidelines for Collecting and Using Operating Information10.0 Purpose

Gas-lift well and system operating information is a valuable resource that must be collected and used in optimizingthe overall gas-lift operation. Information sources are well tests for total recovered gas, oil, and water; downtown usedto allocate total facility production and to evaluate equipment reliability; flowing gradient surveys used to find the pointof lift and to calculate inflow performance; injection rate and pressure in conjunction with tests; wellhead pressuresand temperatures.

10.1 Well Test Information

Section 8 presents recommended practices for conducting well tests. This section discusses recommended practicesfor using this information in gas-lift well operation, analysis, troubleshooting, design, and optimization.



DRAFT


The following practices are recommended for the well test information:

— use accurate and timely well test information for operations, engineering, and accounting purposes;

— select from the five types of well tests the one that matches the desired test objectives. Each well test must becarefully planned and conducted to obtain the correct information. If the wrong type of test is conducted at thewrong time, insufficient information may be obtained for the intended purpose.

10.1.2 Test Types and Applications

10.1.2.1 Production Rate and Injection Gas Rate Evaluation Test

This test measures the production and injection rates and determines the characteristics of the well; is it steady ordoes it have pulsing, unstable flow. The objective is to measure the well's production rates of oil, water, and total gas,the injection gas rate, and the injection and production pressures. For a Production Rate Evaluation Test, the wellmust be tested long enough, and under the right conditions, to obtain accurate, reliable results. The well should beproducing the way it normally produces, whether in steady or unsteady flow. If it is normally unstable, it should beproduced in its normal unstable manner. The objective is to learn how the well actually produces, not how it shouldproduce. If production variables such as water cut, GOR, and reservoir pressure are not changing frequently, this testmay only be required once per month, or even less frequently. These tests may be run as infrequently as quarterly, iflocal rules and regulations permit.

10.1.2.2 Gas-lift Performance Test

This test checks the current performance of the well and gas-lift system. Typically, a production performance test,which is sometimes called a “short cycle test” or a “gross test” is conducted for approximately an hour to detectproblems in the well or in the supply of gas. The purpose is not to accurately determine oil, water, gas, water cut,GOR, etc. It is to check the operation of the well. If it is lifting OK, no further action is needed at this time. If it is notlifting OK, a Production Rate Evaluation Test can be scheduled to determine the cause(s) of any problems. Thesetests should be run as frequently as possible, especially with an automated test system.

10.1.2.3 Gas-lift Performance Test with a Pressure Survey

This test combines a rate evaluation test with a flowing pressure gradient survey. The objective is to evaluate theeffectiveness of the gas-lift design for the well by determining where gas is being injected, if multi-point injection isoccurring, or if deeper injection depth is possible with the existing kick off unloading pressure.

10.1.2.4 Formation Productivity Evaluation Test

A test to determine formation productivity is run in conjunction with a flowing and SBHP survey. The objective is toevaluate the well’s inflow performance. For this test, the production pressure and rates must be stable. If a gas-lift wellis normally unstable, it must be stabilized, by extraordinary means if necessary, to obtain stable operation. Theobjective of this test is to evaluate the formation productivity, not the normal operation of the well. Such a test shouldbe conducted at least annually.

10.1.2.5 Gas-lift Response Test

This is a multi-rate test that measures liquid production rate versus injection gas rate. The optimum injection gas ratecan be selected after a series of changes, with each change followed by a stabilization time, and a production test. Thisoptimization process should be conducted after a gradient survey confirms that deep lift has been attained. In somefields the production rates of wells can be measured (or estimated) on a continuous basis using a multi-phase meter ora differential pressure technique. Thus a well can be “on test” all of the time. See more discussion of this in 10.5.



DRAFT

— Continuous test measurement is possible with multiphase meters if one has been installed on each well. Themanufacturer should be consulted on measurement accuracy as related to the very high gas fraction caused bythe gas-lift gas at normal low pressure test conditions.

— A “good” well test is one that accurately reflects a well’s production performance. All “good” tests should bereported and stored to be used for allocation of total field production and for historical trending. Allocation alsorequires reporting and storage of production (or down) time.

— A “bad” test is any set of data that does not reflect well performance, due to metering problems, well test problems,data gathering errors, or facility disruptions. A “bad” test should not be reported as a well test for the well.

— A “good” or “bad,” or perhaps “questionable” test may be found by comparing its data to a recent trend. If the testrate is very different from the last “good” test, the well or testing system may have a problem. The well testsystem should be checked and repaired, if necessary, and the well retested as soon as practical. If the earlier testis confirmed, both tests should be submitted. If the earlier test is not confirmed, it should be discarded.

— Well test information should be stored in tables within spreadsheets or databases and plotted on graphs ofresults versus time. The data should be available in electronic format for use in or analysis by other programs andshould be supplemented with other information such as associated production or downtime, pertinent welloperations, notes or indications of changes in well operating practices, facility modifications, water or gasinjection changes, etc.

10.1.3 Test Data and Calculations

Well test information should consist of the data given below, as a minimum, in the units preferred by the operatingcompany. Production rate per day is listed, indicating that short time tests are extrapolated to a daily rate. Problemwells should have the data reported at least hourly, rather than a simple difference of the meter reading at the end andbeginning of the test. The report should use the stabilized rate data and discard that portion when the well waschanging rate.

TGLR is the total gas to liquid ratio. FGLR is the formation (reservoir) gas to liquid ratio, where reservoir gas is thedifference between total gas and injected gas. If the total gas measurement is unreliable, reservoir gas can beestimated from the PVT data. IGLR is the injection gas to liquid ratio. Another calculation is oil divided by injection gas(bbl oil/million scf) (m3/m3) or gross liquid divided by injection gas (bbl liquid/million scf) (m3/m3).

The calculated ratios based on test data can be used to rank wells for purposes of shutting in production when gassupply is temporarily reduced.

Well CSGWell-head

Well-head

Gross Liquid Water Oil Water Total INJ TFLR FGLR IGLR

Press Press Temp Rate Rate Rate Gas Gas

psig psig °F bbl/d % bbl/d bbl/dmillionscf/d

millionscf/d scf/bbl scf/bbl scf/bbl

A04 919 116 113 1600 25 1200 400 3.13 2.84 1956 181 1775

B05 1201 160 180 4620 70 1386 3234 2.25 1.91 487 74 413

Well CSGWell-head

Well-head

Gross Liquid Water Oil Water Total INJ TFLR FGLR IGLR

Press Press Temp Rate Rate Rate Gas Gas

kPa kPa °C m3/d % m3/d m3/d m3/d m3/d m3/m3 m3/m3 m3/m3

A04 6336 800 45 254 25 191 64 88632 80420 369 34 335

B05 8281 1103 83 735 70 220 514 63713 54085 86 14 78



DRAFT

10.1.4 Well Test Usage by Operating Personnel

Operating personnel or well analysts use well test information to perform the following:

— review well test quality and remove tests that are bad due to metering or other “extraneous” problems. A well testhistory, either in tabular or preferably in graphical form, is used to compare to the current test. Once the test resultis validated, it can be transferred to engineering and other operating personnel;

— schedule and maximize use of facilities for both quick performance and longer evaluation tests or tests inconjunction with flowing surveys. Automated, real-time testing speeds the validation process for a well in test bypermitting a check of conditions which may affect the well test rates. Bad tests can be aborted, improving use ofthe facility;

— check on problems or faults in the well fluid gathering, testing, or gas-lift system;

— optimize the current operating performance of the well relative to gas-lift injection rate;

— evaluate well or system re-design to improve performance by changing injection strategies, re-designing gas-liftvalve port size or set pressure, deciding mandrel positions for gas-lift valves, dummies, or an orifice, orincreasing gas injection pressure or rate. Accurate well test information provides operations, in conjunction withengineering, a basis for making changes.

10.1.5 Well Test Usage by Engineering Personnel

Engineers, engineering technicians, or well analysts use well test information as follows:

— analyze, troubleshoot, and evaluate well production and artificial lift system performance to improve oilproduction and reduce cost;

— determine the well's IPR in conjunction with static and FBHP data. The well's IPR is an important factor in gas-liftoptimization and subsequent allocation of injection gas to the wells served by a distribution system. IPR is alsoused to evaluate the well's productive capacity;

— evaluate formation damage (skin) in conjunction with a pressure build up or other transient test. The well’s IPR isrequired to accurately optimize (or allocate) the total injection gas to the wells served by a distribution system;

— allocate field production back to individual wells. Allocated production for each well is based on its well testrate(s) and downtime that the well had during the month. The allocation factor is based on the ratio of the sum ofestimated production based on well tests and down time to the actual measured production from the field.Allocated production of oil, water, and gas is used for monthly data and for cumulative results. Each well’sproduction is summed to give the total production from any desired combination of wells in a reservoir, secondaryrecovery project, field, etc.;

— evaluate and prioritize workover prospects, recompilations prospects, well abandonment plans, and the designor evaluation of reservoir recovery processes;

— estimate reserves produced from the reservoir and adjust the remaining reserve volumes;

— size and/or evaluate the need to adjust the sizes and operating characteristics of gathering facilities, well testfacilities, production separation and handling facilities, production treating facilities, sales facilities, compressionand dehydration facilities, etc.



DRAFT

10.1.6 Well Tests Usage by Others

Accountants and regulatory staff use well tests and associated information to report produced volumes to pertinentregulatory agencies, and for paying royalties and taxes.

10.2 Downtime Information

Downtime (non productive time) is defined as any time when the well is not producing due to scheduled shutdown,wellbore equipment failure, facility shutdown, weather related shutdown, or other problems. If a well is cycling orintermitting, the “off” time that occurs between production heads or cycles is considered part of the normal productionprocess, and not as downtime.


The following practices are recommended for handling downtime:

— detect and account for downtime, including the cause(s) when this is known. Investigating and addressing thecauses can help reduce downtime;

— use downtime information to prioritize well servicing and maintenance work;

— report this information to operations, engineering, and accounting personnel.

10.2.2 Use of Downtime Information

The following are uses of downtime information.

— By operating personnel.

Operating personnel are responsible for using downtime, well tests, and associated pertinent information so theycan plan and prioritize well maintenance and servicing work. The information should also go into a data base foraccess by engineering and other personnel.

— By engineering personnel.

Engineering personnel review downtime and investigate its causes, so they can assist in diagnosing persistentwell or system problems and plan corrective actions.

— By others.

Accountants need downtime and well test data to allocate produced oil, water, and gas to the individual wells.

10.3 Pressure and Temperature Surveys

Flowing and static pressure and temperature surveys are among the most important types of information concerninggas-lift wells. Often, these surveys are not obtained frequently enough, or are obtained incorrectly, resulting in poor ormisleading information.



DRAFT


The following practices are recommended for pressure and temperature surveys:

— obtain a flowing pressure and temperature survey, and an associated static pressure survey, in every gas-lift wellon a periodic basis; do this annually where possible;

— use the guidelines below when obtaining a flowing/static pressure and temperature survey. Maintain a detailedlog of time versus depth which can be matched to the time versus pressure/temperature data on the wirelinegauges. Analysis of problems or abnormal occurrences can be evaluated upon completion of the survey;

— obtain a pressure/temperature survey when a significant change has occurred in a well, or a significant problemhas developed, and pressure/temperature data is required to analyze and diagnose the problem;

— use pressure/temperature surveys and well test data to monitor inflow performance. If inflow declines, use apressure build-up test to evaluate potential formation damage (skin);

— analyze the surveys and advise operators of the well performance and indicated depth of lift. Developrecommendations for valve changes or redesigns.

10.3.2 Philosophy of Downhole Pressure/Temperature Surveys

Pressure/temperature surveys that are accurate and complete may be the most important piece of information abouta producing gas-lift well. A poor or inaccurate downhole pressure/temperature survey is worse than no survey at all.

There are four types of pressure/temperature surveys in gas-lift—a flowing pressure/temperature survey, a FBHPsurvey, a static pressure survey, and a pressure build-up. The first three are important in gas-lift surveillance. Thefourth gives information about the reservoir and wellbore inflow performance.

Pressure/temperature surveys are expensive with costs for the wireline unit, pressure and temperature gauges,wireline operator, pressure analyst, and deferred production, if the well must be temporarily shut-in for the survey.Risk is inherent in the operation since the gauges or the wireline could become entangled or lost and require a fishingoperation.

These costs and risks may result in a reduction in the frequency and total run time of surveys. Management pressuremay be applied to return the well to production, reducing pressure build up time. This may result in erroneous data inthe surveys and be much more costly, in the long run, than spending the extra time necessary to obtain good surveys.The difference between a good gradient survey and a poor one may be a few minutes or hours; it may be severaldays in the case of a build up test. A good survey may permit significant improvements in the understanding and theperformance of the well. A poor survey may tend to confuse the understanding of the well's performance and lead toa decision to do nothing, or worse, do the wrong thing.

Downhole pressure/temperature surveys are run by downhole pressure analysts or by service companies. Wirelinepersonnel must have clear instructions on stop depths and times for pressure/temperature surveys and a copy of theguidelines for recommended wireline procedures. Engineering or technical personnel should clearly communicate therecommended survey objectives and procedures to the wireline personnel.

Pressure surveys can be run with mechanical gauges or electronic gauges. The mechanical devices are accurateenough and cost less than electronic devices. Most surveys are run with the electronic devices since they are moreaccurate, are rapidly downloaded from the memory unit to the computer, and provide both pressure and temperaturedata (mechanical devices require one gauge for pressure and one gauge for temperature).



DRAFT

The recommendation is to spend the time and the money necessary to obtain high quality downhole pressure/temperature surveys even if it means that the surveys can't be obtained as often. The recommended procedures forobtaining good downhole pressure surveys are given below.

10.3.3 Guidelines for Running Downhole Pressure Surveys

The following guidelines and procedures are recommended for obtaining accurate downhole flowing and staticpressure/temperature surveys for both gas-lift and naturally flowing wells.

— Preparing the survey instruments.

Run tandem pressure gauges and one temperature gauge if using mechanical devices. If one gauge fails, thesecond will still have the downhole pressure information. Run one electronic gauge to obtain both pressure andtemperature, with the option of a tandem pressure gauge.

Run sinker bars as needed to overcome pressure surges when entering the wellhead from the lubricator.

Use pressure gauges with the maximum pressure range and time clock duration that are suitable for the range ofthe data, if using mechanical gauges. This improves the resolution of the pressure recordings and the accuracyof the readings. Electronic gauges measure exact pressures and temperatures, but the time increment for datagathering should be at 30-second intervals.

Run a continuous temperature gauge with a pressure gauge to record the temperature profile if using mechanicalgauges. Electronic gauges have temperature sensors included.

Recalibrate each mechanical pressure gauge before each downhole pressure survey, or on a set schedule of atleast once each three survey runs. Calibrate the electronic gauge with its software prior to each use.

— Planning the survey.

Run a flowing pressure survey prior to a static survey when obtaining both pressures. The flowing survey, inconjunction with the static survey, is used to determine the IPR of the well. Obtain the flowing survey data whilerunning the gauges into the hole if the well will be shut in for a pressure build-up or to obtain a SBHP.

In gas-lift wells, run a temperature survey in conjunction with a flowing pressure survey. This can aid indetermining the operating gas-lift valve(s). The temperature gauge should be run in tandem with the pressuregauge so that both devices record the same producing conditions.

Rig up on the well and insert the pressure/temperature gauges for the flowing survey without shutting in the well.If the well must be shut in to rig up and insert the gauges, restart the well and produce it in a normal (hopefullystable) fashion before the flowing survey is run. The flowing tubing pressure should be equal to or cycling in thesame manner as during normal production. For a gas-lift well, compare the surface two or three pen chart (casingpressure, differential pressure, and tubing pressure) to assure that the well is producing normally before theflowing pressure survey is begun.

— Planning the survey stop depths and times.

Prepare a detailed prognosis for the stop depths and the time at each stop. Provide operations and wirelinepersonal with a schematic of the well showing the depths (measured and true-vertical) and internal diameters ofall subsurface well equipment.

If there is a reasonable idea of which gas-lift valve(s) the well is lifting from, and if no leaks are suspected, it maynot be necessary to make a flowing stop below each of the upper gas-lift valves. Stop at least every 1,000 ft(304.80 m); include stops at mandrels, and mid points between mandrels. Stop 30 ft (9.14 m) above and below



DRAFT

each valve in the vicinity of the operating valve(s). If a well is producing in a stable manner with a steady casingpressure, differential pressure, and tubing pressure the stops above and below each valve should be for at least10 minutes to assure that a stable reading is obtained, but need not be longer.

If the well is heading, stop at each valve for about 1.1 to 1.2 times the heading period or for sufficient time toobserve the minimum and maximum pressure in slugging, surging wells. For example, if the wellheads once perhour, each stop should be for 1.1 to 1.2 hours, or for the time to read minimum and maximum pressure on thewellhead. If it requires two runs to obtain a survey, because of the time involved, the bombs can be pulled duringthe shut-in time to obtain the static reservoir pressure, unless a pressure build-up survey is being run.

Run a flowing survey in a severely surging well or intermittent lift well by shutting the well and lowering gauges tothe perforations. Open the well and obtain two hours of flowing time to allow the well to re-stabilize, then start outof the well, stopping below the mandrels and at mid points between mandrels.

Heading may be associated with low fluid production rates, low reservoir pressures, and/or low FBHPs due to afluid entry restriction. Usually the well will be lifting from the bottom gas-lift valve or orifice, or from a valve orvalves near bottom. In heading wells, it may not be necessary to obtain flowing pressure stops near the uppervalves in severely heading wells to reduce risk of wireline tangling or gauge damage.

However, a well may be heading for other reasons. Unless it is known that the well is lifting from one or more ofthe bottom valves and/or orifice, it will be necessary to stop above and below each valve.

There may be reluctance to run a flowing pressure survey if a well is heading. If the tubing pressure chart showsshort pressure “kicks,” a gas-lift valve is probably opening intermittently. Stopping just above this valve could causeproblems. In this case, the pressure recorders should be run to bottom with the well shut in. The SBHP should beobtained first, and then the well should be kicked off, allowed to “stabilize” or return to its “normal” production mode,and a FBHP obtained. Flowing pressures should be obtained just above and below the bottom gas-lift valve ororifice and any other valves that may be open all or part of the time. Then the well can be shut-in again while thepressure recorders are pulled. To further prevent problems, extra weight bars should be added, and a “no-blow”device used to prevent the pressure recorders from being blown up the hole.

— Obtaining bottomhole pressure data.

An important reason for running a flowing pressure survey is to obtain an accurate FBHP that can be used, inconjunction with an associated well test and the SBHP, to determine the well's IPR.

The well must be producing stably and the pressure measurement tool must be left in place long enough toobtain an accurate FBHP. In addition, obtain enough flowing pressure readings below the point of injection toextrapolate the flowing pressure gradient to the mid-point of the perforations.

If the SBHP must be run first, ensure that the well is restarted and fully stabilized or returned to its “normal”producing state before moving the gauges off bottom. Note: The process of shutting in the well before the flowingsurvey is not recommended except in situations where it is the only viable option. It may be very difficult, and timeconsuming, to return the well to its “normal” production mode.

It is important, for accurate determination of the well's IPR, that a steady, “reservoir” FBHP be measured. If thewell is heading, the fluctuations in measured bottomhole pressure will obscure the true FBHP and IPR. Ifpossible, a heading well should be stabilized before the FBHP is measured, even if it is necessary to useextraordinary means. One way to do this, for many wells, is to temporarily over-inject too much gas. Manyheading wells can be temporarily stabilized in this fashion. If this method is used to obtain a stable FBHP, a stablewell test must be obtained at the same time.

Obtain a SBHP survey by shutting off the gas injection first and allow the injection pressure to reduce to theclosing pressure of the operating valve (or orifice) before closing the wing valve. This will prevent placing full



DRAFT

injection pressure on the tubing, possibly pushing some fluid back into the formation, and invalidating the staticpressure survey.

Leave a well shut in for 24 hours for a SBHP survey. The shut in time for a build up test depends on thepermeability of the well, ranging from 24 to 48 hours for a high PI well, and from 72 to 200 hours for a low PI well.The shut-in time should not be too short. This can result in an incomplete build-up and erroneous skin andpermeability calculations. Consult with reservoir engineering personnel for recommended shut-in times.

If a well has been shut-in for too short a period of time, and if pressure build-up data was being obtained, it maybe possible for an experienced bottomhole pressure analyst to extrapolate the shut-in pressure data to obtain anestimate of the actual SBHP.

Compare 24-hour static pressures with each other, and compare extrapolated reservoir pressures from build uptests with each other.

— Obtaining the static pressure gradient.

Obtain static pressure stops coming out of the hole. The number of stops and the depths at which they are takenwill determine whether the gas/oil and oil/water contacts can be located. Recognition of the oil/water contact isimportant; the correct gradient is needed when extrapolating to obtain the static bottom hole pressure at datum,or at the mid-point of the perforations. Stops should be made for 10 minutes and at least 500 ft (152.4 m) apart inthe bottom of the well, or closer stops should be made if necessary to locate a contact, and 1000 ft (304.80 m)apart in the upper part of the well.

Use pressure extrapolation to determine the “static” fluid level which will exist when the well is off production for aperiod of time. The static level is related to the gas-lift “kick-off” requirements when a well has been stopped andmust be restarted (kicked off) on gas-lift.

— Using the pressure data.

Read the flowing and static pressure stops versus times and translate to the associated depths at the specifictime. Correct the measured depth to true vertical depth for proper interpretation of the pressure data. (Thiscorrection may not be necessary in all cases, if the computer programs used to analyze the pressure data canmake the necessary correction automatically.)

Use the flowing and static pressure survey data to analyze the gas-lift well performance.

Prepare a report including the following information:

— downhole pressure/temperature survey readings;

— surface tubing and casing pressure recordings;

— well test results;

— flowing and SBHP and temperature plots;

— IPR plot from a calculation based on the PI;

— a set of equilibrium curves based on the new survey;

— a check of the current gas-lift design and any recommendations for changes;

— pressure build-up analysis performed and any resulting recommendations.



DRAFT

10.3.4 Recommended Procedures for Running Pressure Surveys

The following step-by-step procedures are recommended to run a flowing and static pressure survey:

— rig up the lubricator;

— pull the wireline-retrievable subsurface safety valve or storm choke, if the well has one. Many offshore gas-liftwells have a safety valve if they can flow. Wells that cannot flow may not be required to have a safety valve. Ifone is not required for safety purposes, leave it out of the well as it creates an additional pressure drop. If a safetyvalve is required lock it open so it won’t close on the wireline;

— review the well schematic with internal diameter data. Run a sinker bar or gauge ring to the end of the tubing,then slowly lower the sinker bar through the perforated interval and tag the deepest obstruction. Do not use animpression block as this may interrupt flow in the tubing;

— insert the pressure/temperature gauges into the lubricator, set on the wellhead, open the crown or master valve,let pressure equalize for 15 minutes or until the well is stable or normal;

— record the pressure in the lubricator with a dead weight tester or electronic calibration meter before running in thehole. Use the dead weight tester electronic device to measure the casing pressure;

— after the survey is over, record the pressure in the lubricator for 15 minutes.

Once the well is flowing or gas-lifting in a steady or normal manner, place the well on test. Conduct the well test for theentire period that the flowing survey is in progress. Also ensure that the surface recording system or chart(s) of casingpressure, tubing pressure, and gas injection rate (for gas-lift wells) are recording properly. An automated datagathering system will improve the quality of the production test data. If the well has a flowline monitoring device, besure that the flowing record of the well is saved with the pressure survey.

10.4 Injection Pressure and Rate Measurements

Gas injection rate and pressure must be known to analyze and control a gas-lift well.


The following practices are recommended for injection pressure and rate measurements:

— monitor and record the gas injection pressure and rate during production tests or continuously with an automatedsystem for gas-lift optimization. The format can vary from mechanical recorders to flow computers to real timedata base systems;

— analyze this information to detect potential operating problems, both in valve design and gas compressors ordehydrators;

— collect injection gas information to evaluate well tests, pressure/temperature surveys, and the initial unloadingprocess.

10.4.2 Importance of Injection Rate and Pressure Information

Pressure and gas injection rate, combined with well test and flowing survey data, are necessary for gas-liftsurveillance, problem recognition, performance analysis, and optimization. This information is needed to determinewhich valve(s) are injecting, if there is over or under injection, and the liquid or oil rate per unit of injection gas. Thisdata gives formation gas, determined by subtracting the injected gas from the total recovered gas.



DRAFT

Injection temperature may be important if the temperature can vary significantly [more than 20 °F (11 °C)] from day today, from day to night, or seasonally, or if hydrate formation (gas freezing) is a problem. Normally, occasionaltemperature sampling is sufficient.

Gas-lift well problems can arise any time and resolution is simplified if the injection pressure and rate are alwaysmonitored. Options include a two-pen chart recorder attached to an orifice meter for differential and line pressure,preferably with 24- or 48-hour chart drives since seven-day charts provide too little resolution for effectivetroubleshooting. An electronic gas flow computer or automated data gathering system is a more effective method.

Continuous monitoring of gas injection pressure and rate is needed for optimization, but this data is also requiredduring: a) well tests, b) pressure surveys, and c) initial unloading. The recording mechanism—chart recorder, gas flowcomputer, automation system, etc.—should be checked periodically to be sure it is working properly.

The information should be collected and analyzed as part of the test or survey. If gas is being metered and recordedwith an orifice meter and chart recorder, the chart must be integrated to determine the volume and rate of gasinjected.

Gas injection pressure and rate responses are symptomatic of certain problems and operators should be trained torecognize them from the shapes of the injection pressure and rate charts or curves.

10.5 Wellhead Production Pressure, Temperature, and Rate

Wellhead producing pressure, and in some cases the producing temperature, are required to analyze a gas-lift well.The production rate can be measured on a continuous basis with a multiphase meter or comparable device, ifoperating conditions require this information.


The following practices are recommended for wellhead production pressure, temperature and rate:

— monitor and record the wellhead production pressure continuously during tests and include a sensor for this datain an automated system. The format can be a mechanical two-pen recorder, an electronic recorder, or a real timedata base;

— analyze this information, in conjunction with the injection pressure and rate, to detect potential operatingproblems, such as heading or severe slugging, or waxing or other plugging of flowlines;

— collect production pressure information to analyze well tests, pressure surveys, and the unloading process;

— consider measuring or estimating the production rate on a continuous basis using a multiphase meter or acomparable device.

10.5.2 Importance of Production Pressure and Temperature

Production pressure should be measured in conjunction with injection pressure and rate for gas-lift well surveillanceand troubleshooting. Diagnosis of routine problems, well test analysis, pressure survey interpretation, and evaluationof the unloading process are aided with production (wellhead) pressure data.

Production wellhead temperature can vary significantly due to rate changes, gas-liquid slugging, or water cutincreases. Continuous monitoring can aid detection of these changes, but normally, occasional temperature samplingis sufficient.

Some operators measure production temperature and use it to indicate changes in flow rate, since the surfacewellhead temperature is directly related to the liquid production rate.



DRAFT

Production pressure should be measured with the same frequency and accuracy as injection pressure - whether witha mechanical recorder, an electronic device, or an automation system. As with injection pressure, productionpressure can exhibit characteristic responses in the face of certain types of operating problems. Understand thetypical problems in each field and ensure that each operator is trained to recognize them.

10.5.3 Importance of Production Rate

Incentives to measure or estimate the production rate on a continuous basis for gas-lift wells are as follows:

— know if the well is producing, and the producing characteristics, such as steady flow or heading;

— have a direct basis for allocating “actual” sales volumes back to each well;

— use the continuous rate to validate the well test results;

— when confidence in gained in this measurement (or estimate), it may radically change the well test process. Itmay be possible to use well testing to confirm and calibrate the continuous production rate measurement; thenuse it as a primary well surveillance mechanism;

— continuous production rate measurement permits real-time optimization if injection gas measurement is alsocontinuous. The gas-lift injection rate can be adjusted, in real time, to optimize the actual production of the well.

11 Manual and Automated Well Operation and Control11.0 Purpose

Gas-lift systems must be properly monitored and controlled to be profitable. This discusses the limitations of manualoperation and the advantages of automated monitoring and control.

11.1 Manual Operations

Manually operated gas-lift wells are started or stopped and gas-lift injection rates are set, adjusted, and measuredmanually by operating personnel. Manual operations are labor intensive and may be a restriction to optimization inlarge fields.


The following practices are recommended for manual operations:

— develop a process flow diagram of the gas-lift system, which is a complex integration of multiple, oftencompeting, facilities, wells, and processes. Each component of a system must be identified, both in terms ofdesign rates and physical attributes of pressure/temperature;

— provide talented, trained, dedicated people to design, operate, maintain, and optimize the complex gas-liftsystem;

— provide key people with on-going training in gas-lift monitoring, measurement, surveillance, troubleshooting,design or re-design, and optimization;

— provide quality equipment and measurements of well tests, injection pressures and rates, production pressures,and flowing and static pressure surveys;



DRAFT

— perform in-depth, periodic gas-lift system reviews looking for bottlenecks, patterns of poor performance (e.g.hydrates, heading, etc.), wells that must be re-designed to meet their current operating conditions, andopportunities to optimize gas-lift injection rates.

11.1.2 Limitations of Manual Operations

Manual gas-lift limitations are as follows:

— time to detect, diagnose, and correct problems depends on the company’s priorities and methods for datagathering and analysis;

— injection rate set points are difficult to maintain and depend on personnel to make adjustments in the face ofchanging conditions in the distribution system or wells;

— injection, production, and well test data may not be gathered simultaneously and may not represent consistentevents. Effective testing would require all data be gathered at the same time;

— real time data and associated gas-lift optimization (allocation) is difficult to accomplish;

— personnel have many competing priorities; thus to find, train, and make available gas-lift staff is a challenge.

Fortunately, there are alternatives to manual operation that can be considered, at least in (financially) important gas-lift fields. These are discussed in 11.2.

New fields should be automated, but some existing fields are and will continue to be operated manually for economic,cultural, social, geographical, contractual, or other reasons. In these cases, guidelines that can make gas-liftoperations as effective as possible should be applied.

11.2 Automated Operations

Automation of gas-lift wells and systems can take many forms, from a simple monitoring system, to a system for bothreal-time monitoring and control, to an integrated, real-time, optimization system. The same operations described inmanual operations must be conducted, but sensors, communication links, and computers augment the role ofpersonnel, which are often limited by other priorities, weather, or worker resource shortages. Personnel are stillrequired, but their task is to implement and maintain the automated system and use it to optimize the gas-liftoperations.


The following practices are recommended for automated operations:

— identify information that can be automated for improved data gathering and real time reporting with associatedcomputer analysis. Provide this information to well analysts, engineers, or technicians that can provide support toimprove production;

— develop lists of operating points that can be controlled automatically, thus eliminating weather or personnelavailability as a factor in starting and stopping wells, switching them into and out of test, or adjusting gas injectionrates;

— evaluate operating personnel for those that can help implement and maintain the automation system,communications links, and computers. Train them for these tasks and train the remaining operators to use thisequipment;

— evaluate the production gains from faster data gathering, analysis, and control changes that optimize gas-lift;



DRAFT

— conduct the same recommended practices listed for manual operations; only use rapid computer analysis tospeed problem solving.

11.2.2 Types of Waste that Gas-lift Automation Can Prevent

Automated operations can enhance the profitability of gas-lift operations, relative to manual operations, by helping toimprove their overall effectiveness and efficiency, through detecting and eliminating waste and by keeping waste fromre-occurring.

— Waste of gas through over injection in some wells.

Over injection is a typical problem. “If some gas is good, more gas must be better.” This logic is sometimesfollowed (often inadvertently) until some wells are losing efficiency and production. Automation can keep a well atthe economic optimum amount of gas with control of the injection rate. Reliable costs for gas compression andtreating, water treating and injection, and oil treating must be gathered to do the economic optimum calculation.

— Waste of production opportunity through under injection in some wells.

Under injection leads to less than optimum production. Limited compression and supply of gas requiresdistribution to the wells based on the economic optimum discussed above, with automated control of injectiongas to each well.

— Waste of production opportunity or waste of gas through ineffective gas-lift well operations such as heading, and/or not lifting from the deepest possible gas-lift valve.

Heading can be quickly identified with automated data gathering using tubing and casing pressures. Surgingpressures cause the reservoir pressure drawdown and average production rate to be less than it could be understable operation, and sand movement may be aggravated. Automated control permits adjustments to minimize oreliminate heading. Injection at the deepest gas-lift valve will also minimize heading, reduce gas usage, and givethe best or “optimum” production. A flowing survey is required to validate the deep point of lift; then automationcan help maintain it.

— Waste of production opportunity through undetected and/or unnecessary well and/or equipment downtime.

Downtime can be quickly detected with casing and tubing pressures, and gas measurement. Unscheduleddowntime may occur because of a compressor shutdown, a blocked gas-lift injection line, a blocked flowline, or asafety valve malfunction. Automation can be used to detect abnormal downtime and alert operations in a timelymanner.

— Waste of operating time and cost through ineffective data gathering, problem recognition, and problem diagnosis.

Problems are more quickly detected with automated gas-lift systems. Distribution system, well, measurement, orcontrol problems can be diagnosed with the pressure, temperature, and rate data that is gathered, aided bysoftware tools that use the data. Operators and analysts can spend their time solving known problems ratherthan looking for problems and then trying to gather the information necessary to diagnose their causes.

— Waste of capital investment through under-sizing or over-sizing of equipment.

Equipment or pipeline under sizing can lead to production losses or inaccurate well tests. A too small distributiongas line causes excessive pressure losses with reduced wellhead injection pressures, which limits the possibledepth of injection in the wells. Lower and less than optimum production results. Automated data gathering can



DRAFT

identify capacity restrictions and permit planning for expansions. Separator capacities can be more effectivelyutilized with an AWT system, which can test more wells thus saving the cost of additional test equipment.

— Waste of both gas and production through undetected system bottlenecks.

Automated data gathering can augment nodal analysis programs used to evaluate a system to detectbottlenecks. A too small or too complex gathering system causes high producing wellhead pressures, whichreduces production and causes non-optimum operation. These nodal analysis programs are only as good as theinformation they use. The automation system can provide the accurate information needed to check for and helpeliminate bottlenecks.

— Waste of both gas and production when gas is not controlled to continuous and intermittent gas-lift wells on thesame system.

Automated gas rate control is needed when low reservoir pressure, low PI, or relatively low production ratesrequire intermittent gas-lift or other artificial lift methods such as chamber lift, plunger lift, gas-lift with insert coiltubing, etc. Automated injection gas control can permit intermingling these various forms of lift with continuousgas-lift and can minimize interference between wells. This is not easy, but it has been successfully undertaken bysome operators. If it can allow more effective production of some wells, it may be worth considering.

— Waste of both gas and production through not matching the “best” types of gas-lift valves and other equipment tothe specific problems at hand.

Currently, the choice of the “best” type of gas-lift equipment, especially gas-lift valves, is at least partially cultural.“We have always used this type of valve.” This may lead to inefficiency if equipment is not applied as well as itcould be. Automation can help apply the “best” equipment for each particular situation.

12 Procedures for Initial Unloading and Kick Off12.0 Purpose

The casing annulus and tubing are full of a completion or workover (or “kill”) fluid following completion or re-completion. This fluid must be removed, down to the depth of the desired operating gas-lift valve or orifice, by aprocess called unloading, before a well can be placed on either continuous or intermittent gas-lift.

Whenever a gas-lift well is stopped or shut-in for a period of time, the fluid level in the production string will rise to thelevel that can be supported by the shut-in bottomhole pressure. When a gas-lift well is placed back on production, thewell must be re-started, by a process called kick off, to resume gas-lift injection from the desired operating gas-liftvalve or orifice.

12.1 General Unloading Recommendations

The completion or “kill” fluid which fills the tubing-casing annulus after the well is initially completed or worked overmust be removed from the well by displacing it from the annulus to the tubing, through the gas-lift valves, and thenproducing it up the tubing and out of the well. This process of removing the completion fluid is necessary to permit gasto be injected down the annulus and into the tubing at the depth of the operating gas-lift valve or orifice. This processis called unloading. Note that a standing valve is normally not used in a continuous gas-lift well. This unloadingprocess may require several unloading valves in lower pressure systems; it may be accomplished with one valve ororifice in a high pressure system.

Gas-lift in low to moderate pressure systems uses the upper gas-lift valves as “unloading valves” so the well can belifted (unloaded or worked down) to the desired depth of injection. The number of unloading valves is matched to theinjection pressure available in the field. In some cases, high pressure gas for unloading can be temporarily supplied



DRAFT

by a nitrogen source. With high pressure systems, the number of unloading valves is minimized. In a very highpressure system, the well can lift from a single valve or orifice.

Unloading is a critical process in the life of a gas-lift well. For effective gas-lift, the well must be unloaded down to thedepth of the desired operating gas-lift valve or orifice. If the well is not unloaded correctly, this depth may not bereached. Even more seriously, improper unloading may result in damage to the unloading valves thereby preventingoperation from the deeper desired point of injection, which will cause the well to lift at a less than the optimumproduction rate.


The following practices are recommended to improve the unloading process and to reduce the potential of damage tothe unloading gas-lift valves:

— use clean completion and workover fluids. Circulate the well bore clean, and leave filtered fluid in the tubing-casing annulus. Unfiltered fluids are often a source of solids that can either cut or plug the gas-lift valves;

— circulate the wellbore clean of any drilling mud before installing the gas-lift valves. If mud was left in the annulusof a well that is completed with side-pocket mandrels and dummies, a circulating plug should be set below thebottom mandrel (to keep the mud off of the formation), and circulate the annulus clean before running the gas-liftvalves. Circulating mud through an empty gas-lift mandrel pocket may damage the mandrel. The use of acirculating gas-lift valve or orifice valve (fully open) will prevent damage to the mandrels;

— do not install gas-lift valves in gas-lift mandrels during completion or re-completion activities if acidizing,fracturing, or other treating is planned. They should be installed by wire-line operations after the completion orworkover rig has completed its work. If the production tubing is run in clean completion fluid, and no further welltreatments are planned, the valves may be installed with the mandrels;

— do not circulate fluid from the casing annulus to the tubing on a well with gas-lift valves in place, since flow acrossthe valves may occur, with the likelihood of gas-lift valve and/or check valve damage. However, if the completionfluid has been replaced with a clean fluid, this may be permitted. In any case, liquid flow rates through valvesshould be minimized to minimize risk of erosion;

— blow or wash scale, trash, welding slag, and other debris from the gas injection line before hook-up, especially onnew lines. This precaution prevents the introduction of debris into the meter, control device, or the annulus whereit could cut or plug the gas-lift valves. A screen can be installed upstream of the control choke or valve to filter outdebris;

— bleed a small amount of liquid or gas back through the casing valve if there is pressure on the casing annulus; itis good practice to flow out any debris that may have accumulated in this “dead area” during the workover ordrilling operations. Bleed these fluids into a tank or disposal system;

— check separator capacity, stock tank liquid level, and all valves between the wellhead and the tank battery toensure they are operating properly and are fully open;

— check the safety relief valve on the gas gathering system.

12.2 Unloading Continuous Gas-lift Wells

Use care in unloading a continuous gas-lift well since more gas-lift valves are damaged at this time than at any othertime during the life of the well. Prevent excessive liquid rate displacement across gas-lift valves to reduce potentialvalve failure due to sand and liquid cutting.

The injection gas rate (and, therefore, the pressure) should be increased gradually to maintain a low fluid velocitythrough the open gas-lift valves. If full gas-lift injection rate or pressure is exerted on top of the fluid column in the well,



DRAFT

a pressure differential (equal to the gas-lift line pressure minus the wellhead pressure) will occur across each valve inthe installation. Damage to the valve stem and/or seat can result from the high fluid velocity through the valves ifunloading gas rate is not controlled. The top valve is most susceptible to damage because the liquid column from thesurface mostly passes through this valve.

Restricting the input gas during unloading will also prevent an overload of the surface facilities. Some installations aredesigned with the upper gas-lift valves having chokes and/or smaller ports than the operating valve to reduce the fluidflow rate and subsequent gas flow as the upper valves are uncovered. If a flowline choke is required to prevent severeslugging and overloading of surface facilities, the choke should be installed as far away from the well as possible.


The following procedures should help avoid placing excessive pressure differentials across the unloading gas-liftvalves and are recommended for initial unloading of continuous gas-lift wells:

— install a two-pen pressure recorder, or comparable electronic device, to record the injection gas pressures andproduction pressures at the surface. Measure the injection gas pressure downstream of the injection choke orcontrol valve, and measure the production pressure upstream of any flowline choke or choke body;

— bleed the wellhead pressure down slowly to the flowline pressure (separator pressure) if the wellhead productionpressure is greater than the flowline pressure;

— remove or open the flow line choke depending on the well's expected reaction to gas-lift. An adjustable choke orpositive choke should be left on the wellhead connection to the flow line only if the well is expected to flownaturally after it is “unloaded” and started with gas-lift, or if overloading of surface facilities is a possibility.Remove the choke if the well does not flow. On some offshore gas-lift wells, a wellhead choke is frequently leftpartially closed during the initial unloading process—to prevent big slugs from hitting the surface productionfacilities and shutting the system down;

— control the lift gas injection rate slowly into the well so that it takes about 8 to 10 minutes for a 50 psig (345 kPa)increase in gas pressure. Continue this rate of injection until the gas-lift injection pressure reaches about 400 psig(2,758 kPa). Increase the injection gas rate into the well, when the pressure reaches about 400 psig (2,758 kPa),so that it takes about 8 to 10 minutes for each additional 100 psig (689 kPa) increase in the gas pressure;

— increase the rate in increments of about 30 % of the final target rate. Inject at 30 % of the final rate for the first twovalves, 60 % of the next two valves, then 90 %, and finally 100 %;

— use the lowest gas injection rate until gas is injected into the tubing through the top unloading valve. An injectiongas pressure drop (or reduced rate of pressure increase) and the return of aerated fluid from the productiontubing (or casing if the casing annulus is the production conduit) provides this indication;

— increase the injection gas rate slowly, after uncovering the first and second gas-lift valves, using the incrementsoutlined above. In some instances, the gas injection rate to initially unload a well may be more than the estimateddesign rate;

— monitor the unloading process continuously by watching the injection pressure. It should be possible to see theunloading process as it proceeds to each deeper unloading valve. The surface casing pressure should drop aseach new valve is uncovered. This casing pressure drop is necessary to close the upper unloading valves and topermit the unloading process to “work down” the gas-lift design to the depth of the operating gas-lift valve ororifice (see Figure 12);

— adjust the surface gas-lift injection control device to “fine tune” the well's performance after the well is unloaded tothe depth of the operating gas-lift valve or orifice (see 12.0). Note that wells without packers will require unloadingfollowing each shut-in period.



DRAFT 12.3 Restarting (Kick Off) Continuous Gas-lift Wells

A gas-lift well that has been off production or shut-in for a period of time will have a static fluid level in the tubing thatis supported by the shut-in or SBHP. This fluid level in the tubing may produce a pressure that is greater than normalpressure in the tubing at the depth of the operating gas-lift valve or orifice, and at the depths of at least some of theshallower unloading gas-lift valves up the hole. If this pressure is high enough it may prevent gas injection into thenormal operating gas-lift valve or orifice. The well must undergo an “mini” unloading process to once again “workdown” to the operating depth. This “mini” unloading process is known as the kick-off process. After any period of downtime, a gas-lift well must be “kicked off” to be returned to normal production.

Normally, the surface controller or choke is not re-adjusted and the last injection rate is used to restart and kick off thewell. If a well has been off production for more than one day, trash may have accumulated on the top of the fluidcolumn caused by leakage from tubing to the casing. When the well is restarted (kicked off), this liquid, plus anyassociated trash, must be injected back through the gas-lift valve(s) or orifice.


The following practices are recommended when restarting wells:

— open the wellhead wing valve slowly to bleed any wellhead (production) pressure to the flowline;

— open up and increase the gas-lift injection gas slowly to avoid any pressure surges. Gradually rebuild the casingpressure and observe whether production from the tubing is initiated, based on the sound of fluid flow and atemperature increase;

— if the well has been shut in for more than one day, or if there is a reason to believe that fluid and/or trash mayhave accumulated in the annulus, follow the guidelines for unloading a newly completed well (see 12.2);

Figure 12—Unloading a Continuous Gas-lift Well



DRAFT

— if the shut down has been temporary, the casing pressure build up during the “mini” unloading process can besafely achieved over a shorter period of time—between one-half to one hour;”

— observe the final, stable operating injection pressure to verify that the well has returned to the desired operatinggas-lift valve or orifice. If it has not and the operating pressure is near the initial unloading pressure, then the wellmay be lifting from an upper valve and it may be necessary to restart the “kick-off” process at a reduced injectiongas rate;

— do not “rock” a well by applying gas injection pressure to the tubing in an attempt to depress the fluid level backinto the formation so the well can be restarted more easily. This can cause damage to the formation and/or thesand control system;

— allow time for the upper unloading valves to cool if a well has recently been producing at a high rate. The upperunloading valves may be hot due to the temperature of the fluid. Time to allow them to cool is required so theycan be reopened to allow the restart or “kick-off' process to resume.

12.4 Unloading Intermittent Gas-lift Wells

Intermittent gas-lift wells can be designed for time cycle control, choke control, or automatic control. All control optionscan be used to unload the well, and all procedures must ensure that no excessive fluid rates occur across the gas-liftvalves during the initial U-tubing or unloading operation. Time cycle control is the most common form of intermittent lift.

The type of gas-lift valve and the ratio of the casing annulus capacity to tubing capacity must be properly matched forchoke controlled unloading. Despite the type of operation, continuous or intermittent, the unloading principle is thesame—to ensure that no excessive pressure differentials occur across the unloading gas-lift valves during the initialU-tubing or unloading operation.

The port size of the operating valve must be larger than the input choke size for proper intermittent unloading andoperation. Large input choke size and small valve port size will give continuous gas-lift and potentially causemultipoint operation.


The following practices should help to avoid placing excessive pressure differentials across the unloading gas-liftvalves and are recommended for initial unloading of intermittent gas-lift wells. This process should be closelyobserved by the operator.

— Install a two-pen pressure recorder or other suitable measurement system to record the injection gas pressuresand production pressures at the surface. Measure the injection gas pressure downstream of the intermitter,choke, or control valve; measure the production pressure upstream of any flowline choke or choke body.

— Bleed the wellhead pressure down to flowline pressure (separator pressure) slowly if the wellhead productionpressure is greater than the flowline pressure.

— Remove or fully open the flowline choke. Remove the choke housing as the internal diameter of the body is smallcompared to tubing diameter and can cause the liquid slug to stall at the wellhead. Streamline the wellhead toflowline piping by removing all unnecessary elbows, which may also stall the liquid slug.

— Set the time cycle controller or input choke to inject gas at a rate that will cause a 50 psig (345 kPa) increase incasing pressure over an 8- to 10-minute time period. Continue at this rate until the casing pressure is about400 psig (2,758 kPa).



DRAFT

— Adjust the controller or input choke, once 400 psig (2,758 kPa) is reached on the casing, such that a 100 psig(689 kPa) pressure increase is achieved in an 8- to 10-minute time period. Continue this rate until the top valve isuncovered, as indicated by an injection gas pressure drop and fluid return on the production tubing.

— Do not exceed two or three cycles per hour for the first 12 to 24 hours. The injection time should be adjusted toapproximately 3 minutes, and should stop when the liquid slug reaches the wellhead. Initially, this may be morethan enough gas volume, but will be about right as the well unloads to deeper valves.

— Adjust the injection rate to be 1/3 of the design operating gas injection rate for the first two valves for wells onchoke control. Then increase the injection rate to 2/3 of design for the next two valves. After 12 to 18 hours at thereduced injection rate, adjust the injection gas volume to the full amount expected to be used for lifting the well atthe “target” production rate.

— Follow these guidelines for unloading only. The well should be adjusted to “fine tune” its performance asdiscussed in Section 13 of this document.

12.5 Restarting (Kick Off) Intermittent Gas-lift Wells

Intermittent gas-lift is applied to wells with a low SBHP or low PI, so the fluid level cannot quickly rise in the tubing.However, if a well has been off production for a period of time, the fluid column can be at a pressure greater thancasing gas pressure. One potential problem is the possible settling of solids (e.g. sand) on the standing valve. If thisoccurs, it may be necessary to use a bailer to clean out the sand to avoid the risk of damaging the standing valve. Theprocedures recommended for restarting continuous lift wells should also be applied to intermittent lift wells.

13 Procedures for Adjusting (Fine Tuning) Gas-lift Injection Rates13.0 Purpose

When a gas-lift well is first unloaded or kicked off, it may be unstable (surging pressures or heading) and far fromoptimum. This section provides recommended practices for adjusting or fine-tuning gas-lift injection rates to achievestable, optimum operation, in both continuous and intermittent gas-lift wells. Further recommended practices forintermittent gas-lift wells are provided in API 11V10.

13.1 Continuous Gas-lift Wells with Steady Pressure

Where the pressure of the gas-lift distribution system is regulated or controlled and is relatively stable, a positive oradjustable choke, or a control valve, is normally used to control the rate of gas-lift injection into the well. An adjustablechoke, control valve, or automated control valve is recommended since adjustments in choke size or control valveopening do not require an interruption in the gas flow. The choke may be manually adjusted or automatically controlled.

Excessive choking to control the injection rate may cause freezing problems. This can usually be reduced oreliminated by using a dehydration process in the gas system, installing a gas heater upstream of the choke, injectingmethanol upstream of the choke, or building a heat exchanger around the choke. The latter method will permit the hotproduced fluids to flow around the gas line and warm the incoming gas.


13.1.2 The following practices are recommended for continuous gas-lift wells with steady pressure:

— start with the gas injection choke or control valve sized slightly larger than required to inject the design gas rate;

— reduce the injection choke or control valve size/opening in small increments until the production rate declines;

— readjust the choke to the size that yields the optimum fluid production rate;



DRAFT — obtain a production rate well test and a static and FBHP survey once the well is stabilized. See 10.3 for

recommended practices to obtain this survey;

— determine the optimum gas-lift injection rate with this information. See API 11V8, Chapter II, Section 6.0 forrecommended practices to optimize gas-lift injection rates.

13.2 Continuous Gas-lift Wells with Variable Injection Pressures

The adjustment procedure here is similar to 13.1. Adjust the choke or control valve to obtain the optimum fluidproduction rate with the least amount of injection gas. The difference here is that a pressure-reducing regulator maybe installed “downstream” of the gas-lift source (compressor or other source) to regulate the supply gas to a constantpressure so that the injection gas rate into the gas-lift wells can be more accurately controlled.

The limitation of this method is that the “source” gas-lift system pressure upstream of the regulator must be higherthan the gas injection pressure required by the wells. The gas-lift wells must be designed to operate at a surfaceinjection pressure lower than the normal minimum pressure produced by the gas-lift supply system.

Flow regulators (e.g. gas flow rate controllers) maintain a reasonably constant injection rate despite minor fluctuationsin upstream or downstream pressure. As with pressure reducing regulators, the injection gas pressure must begreater than the operating injection gas pressure required by the downhole gas-lift valve design.

The best approach is to assure that the pressure in the gas-lift distribution system remains essentially constant at alltimes. This can be achieved with an automated gas-lift control system.

13.3 Intermittent Wells with Time Cycle Control

After an intermittent gas-lift well is unloaded, the time cycle or time clock controller should be adjusted to achieve theminimum injection gas required for the desired production rate. Then the injection gas cycle frequency and theduration of gas injection should be checked periodically to assure continued efficient operation.

Figure 13—Adjustable Choke for Continuous Gas-lift Control

Choke

Supply Gas



DRAFT If the liquid production rate changes, surface control of the injection gas must also change to maintain a minimumIGLR. If this ratio is too high because the spread of the operating gas-lift valve is too large, a change in the injectioncycle frequency should be evaluated before redesigning the installation.

Decreasing the injection gas cycle frequency increases the time for fluid to accumulate above the operating gas-liftvalve. The increased liquid slug height at the instant the valve opens results in increased tubing pressure at theoperating valve depth, thus lowering the required injection (casing) pressure to open the operating valve. Theinjection gas volume per cycle is reduced because of decreased valve spread and more liquid is recovered per cycle.These two things work together to yield a lower IGLR. However, there must be enough gas injected on each cycle tolift the liquid slug fully to the surface.


The following practices are recommended for determining the proper cycle frequency and duration of gas injectionimmediately after the installation is unloaded, and any time during the life of the well:

— adjust the controller for a duration of gas injection that will ensure more injection gas volume per cycle than isnormally required [approximately 500 cu ft/bbl (94.35 m3/m3) per 1,000 ft (304.8 m) of lift]. Adjusting the controllerto stay open until the slug reaches the surface will result in more gas being injected into the casing than isactually needed, and could allow an excessive reduction in the pressure of the gas-lift distribution system;

— reduce the number of injection gas cycles per day slowly until the well will no longer produce the desired rate;

— reset the controller for the number of injection gas cycles per day immediately before the last setting. Thisestablishes the proper injection gas cycle frequency;

— reduce the duration of gas injection per cycle until the production rate decreases; then increase the duration ofgas injection by 5 to 10 seconds to account for fluctuations in injection gas line pressure;

Figure 14—Adjustable Choke with a Pressure Regulator

Choke

Regulator

Gas



DRAFT

— adjust an injection gas time-cycle controller as outlined above, provided the gas-lift supply system pressureremains relatively constant. If the pressure varies significantly, the controller should be adjusted to inject thedesired gas volume per cycle with the minimum normal line pressure. When the line pressure is above theminimum pressure, some excess amount of injection gas will be used on each cycle.

13.4 Intermittent Wells with Choke Control

Choke control is another option for operating intermittent gas-lift wells. However, not all intermittent installations canbe unloaded or operated effectively with this method. The type of operating gas-lift valve and the ratio of casingannulus capacity to tubing capacity must be suited for this type of operation. The choke size should be smaller thanthe port size of the operating gas-lift valve to permit the injection pressure in the casing to decrease to the valveclosing pressure after the valve has opened.

The initial surface choke size should be sufficient to pass the amount of gas-lift gas needed for the design productionrate. The injection rate depends on the surface pressure and the casing pressure when the gas-lift injection valve isopened.

The final selection of the surface choke or opening through a metering valve must be determined by trial and erroruntil the desired operation is attained. Since an IPO gas-lift valve suited for choke control is opened by both injectiongas pressure and production pressure, increasing the injection gas pressure will decrease the production pressurerequired to open the valve. After an operating valve closes and the slug of liquid reaches the surface, the injection gasand production pressure begin to increase. The rate at which the gas pressure increases is dependent on the chokesize in the injection gas line, whereas the increase in production pressure at valve depth is a function of welldeliverability and tubing size.

If the injection line choke is too large, the valve will open at a higher gas pressure than that required for adequate gasstorage in the casing. The production pressure (slug height) will not reach a value that will result in the lower gaspressure needed for minimum injection gas usage. By decreasing the choke size, the well has a longer time to deliverfluid into the tubing. This, in turn, increases the production pressure (slug height) at valve depth and reduces the gaspressure required to open the valve.

Choke control of the injection gas is all that is needed for most PPO valve installations. The gas pressure is allowed tovary with the choke size rather than attempting to maintain a fixed gas pressure for production control.


As indicated above, not all intermittent installations can be operated by choke control. The use of a time cyclecontroller may be necessary if choke control is unsuccessful. A step-by-step procedure for choke control is as follows:

— start with the gas input choke size slightly larger than required to transmit the design injection gas volume;

— reduce the injection choke size in small increments until the production rate declines;

— place the well on production using the choke setting that gives the optimum fluid volume with the least number ofcycles. This should also be the least injection gas volume.

NOTE When using choke control in an intermittent gas-lift well, an adjustable choke to control the input gas volume is generallypreferable, since it is not necessary to interrupt the flow of gas to make input changes, as it is with a positive choke.

13.5 Do Not Use Flowline Chokes

Do not control the production rate of gas-lift wells or attempt to stabilize gas-lift wells by using flowline chokes.Flowline chokes increase the wellhead backpressure that the gas-lift operation must overcome. This can reduce the



DRAFT

well's production rate and lead to excessive gas usage, since the higher wellhead pressure stream can cause theupper gas-lift unloading valves in the well to reopen by increased production pressure effect.

This valve interference generally results in one or more of the following negative effects:

— increased gas usage;

— reduced fluid production rate;

— fluctuating injection and/or production pressures.

In most cases, the well's production rate can be better regulated by making adjustments in the gas-lift injection rate.

14 Gas-lift Troubleshooting Tools14.0 Purpose

There are recommended practices and tools available to assist in trouble-shooting or evaluating gas-lift performance.These can be used individually or collectively to indicate downhole conditions. Some techniques have limitations thatmust be understood, so the results do not misrepresent the actual situation.

The flowing pressure survey is the best method for determining and evaluating gas-lift performance, if it is runproperly. There are operating costs and mechanical risks associated with running surveys. To maximize the return oninvestment, candidates for this work must be prioritized.

14.1 Two-pen Pressure Charts, or Equivalent

The most significant pressures on a gas-lift well are the injection and production pressures. From these, which maybe recorded on a two-pen chart or measured continuously with pressure measurement instrumentation, the downholepressures in the casing and tubing can be calculated and compared to the operating characteristics of the gas-liftvalves. From this, it is possible to estimate the point(s) of injection into the production stream. This is reasonablyaccurate if the well is producing stably (not heading or surging) and the computer-based pressure profile model hasbeen calibrated with a flowing pressure survey.

Accuracy of the recorded injection and production pressures is important. Check the accuracy of the two-pen recorderor pressure transducers using a calibrated pressure gauge or dead-weight tester. Two-pen recorder charts and/orsurface pressure recordings can be used to optimize surface gas injection control, locate surface problems, and helpidentify downhole problems.

14.1.1 Where to Install a Two-pen Pressure Recorder or Pressure Measurement Devices

The injection pressure recording or measurement instrument should be connected at the well, not at the compressoror gas distribution header. It is important to know the pressures in the well, not the pressure at some distanceupstream of the well. The instrument should be installed downstream of the gas-lift input choke, time cycle controller,or control valve so that the true surface injection gas pressure is recorded.

If it is not possible to measure the injection pressure at the wellhead, it is better to measure it at an injection manifoldthan to not have it at all. There should be no restrictions between the pressure measurement location and thewellhead. And, if possible, a model of the pressure loss in the injection line should be used to estimate the actualwellhead pressure from the pressure measured at the manifold.

The production pressure recording or measurement instrument should be connected at the well, not at the battery,separator, or production header. It should be installed upstream of the choke body or any other restrictions. This istrue even when no choke bean is used since many choke bodies are not full opening.



DRAFT If the production pressure cannot be measured at the wellhead, a measurement at the production manifold is better thanno measurement at all. The measurement should be taken upstream of any choke, control valve, or other restriction.

14.1.2 Chart Interpretation

The following examples contain typical two-pen pressure recorder charts and injection/production pressure plots forboth continuous and intermittent gas-lift wells, with comments on interpretation.

Some problems, such as “blowing around,” create a unique pattern on a two-pen chart which can be useful inidentifying and remedying problems; however, one should concentrate on the underlying causes of fluctuations in theinjection and production pressures rather than attempting to compare chart patterns from other wells. Few two-penplots are exactly alike.

14.1.3 Examples of Pressure Charts from Continuous Gas-lift Wells

The following examples in 14.1.3 and 14.1.4 are merely examples for illustration purposes only. (Each companyshould develop its own approach.) They are not to be considered exclusive or exhaustive in nature. API makes nowarranties, express of implied for reliance on or any omissions from the information contained in this document.

Users of examples should not rely exclusively on the information contained in this document. Sound business,scientific, engineering, and safety judgment should be used in employing the information contained herein. Whereapplicable, authorities having jurisdiction should be consulted.

Figure 15—Installation of Pressure Measurement Devices

2-pen Recorder

Input Chokeor

Controller

ChokeBody



DRAFT Figure 16—Continuous Gas-lift—Good Operation

Operation Continuous flow, casing choke control, tubing flow.

Type of Well High productivity, high bottomhole pressure.

Type of Gas-lift Valves Injection pressure operated.

Trouble None.

Recommendation Leave well alone.

Remarks Good continuous flow operation. Well has a high working fluid level. The low flowing wellheadpressure should be noted. Well producing 2,000 bbl (320 m3) fluid per day, with 95 % water, fromwater drive reservoir, through 2 7/8 in. (7.30 cm) tubing.

Figure 17—Continuous Gas-lift—Wellhead Backpressure Too High

Operation Continuous flow, casing pressure control with regulator, tubing flow.



Trouble Inadequate production.

Recommendation Reduce backpressure.

Remarks Excessive backpressure may be due to one or more of the following:

1) choke in flow line;

2) restriction in flow line (paraffin, sand, etc.);

3) flow line too small or too long;

4) separator pressure too high;

5) too many sharp bends in flow line;

6) highly emulsified fluid;

7) excessive input gas.



DRAFT Figure 18—Intermittent Continuous Gas-lift—Good Operation

Operation Intermittent continuous flow, time cycle control, tubing flow.


Type of Gas-lift Valves Injection pressure operated, intermittent (choked).

Trouble None.


Remarks Well is injecting 30 seconds every 5 minutes. Well was first placed on tubing control. The fluid washeavily emulsified and the well was not making its production. After placing it on time cycle controlthe emulsification stopped and the well made its production.

Figure 19—Comparison of Intermittent and Continuous Injection

Operation Intermittent injection vs. continuous injection, tubing flow.

Type of Well Borderline production rate.


Trouble Inadequate production.

Recommendation An intermittent and continuous gas injection comparison.

Remarks Compare intermittent to continuous injection of gas to determine most efficient production rate.



DRAFT Figure 20—No Gas Injection Control

Operation Tubing control without choke in gas line, tubing flow.



Trouble Supply gas was closed to well. Well was off production 5 1/2 hours.

Recommendation Always use some method to control injection gas into well.

Remarks Work was being done on the gas system. Well should never be operated without a choke or sometype of control on injection gas.

Figure 21—Continuous Injection




Trouble None.


Remarks The well had been shut in overnight, and the gas had been turned on shortly before the chart waschanged. The casing pressure was at 460 psig (3,172 kPa) at the beginning at 10:1 5 a.m. Therewas a gradual pressure rise to 468 psig (3,227 kPa) due to fluid temperature increase affectingvalve. At 2:45 p.m. the casing pressure increased to 480 psig (3,309 kPa) and a “kick” can benoted on the tubing pressure. This was due to an upper valve becoming the operating valve. At10:00 a.m. the next morning the casing pressure had increased to 490 psig (3,378 kPa) due totemperature effect.



DRAFT Figure 22—Continuous Injection—Frozen Gas Input




Trouble Choke on gas line froze.

Recommendation A gas heater might be installed ahead of the choke, or a jacket might be welded around the choketo permit the hot flowline fluids to pass over it, or the well might be placed on intermittent injection.

Figure 23—Well is Flowing

Operation Continuous flow, tubing flow.



Trouble None, well is flowing.


Remarks Well is flowing; no gas is being injected.



DRAFT Figure 24—Continuous Injection—Kick Off After Idle Period




Trouble Well was closed in to repair flow line.

Recommendation None.

Remarks When the master valve was opened the tubing pressure was 250 psig (1,724 kPa). Flowimmediately started but the pressure declined to 210 psig (1,448 kPa) at the peak of U-tube. Asthe gas cleared through the gas-lift valve, the tubing pressure increased to a maximum of 345 psig(2,379 kPa), then fell off and finally stabilized at 285 psig (1,965 kPa).

Figure 25—Flowing Well—Loads Up Periodically

Operation Continuous flow, tubing control, tubing flow.



Trouble Well is flowing, but loads up with water periodically.

Recommendation Operating satisfactorily.

Remarks The tubing control element is set to inject gas into the well when the pressure decreases to160 psig (1,103 kPa). It can be noted by the rise in casing pressure opposite the drop in tubingpressure.



DRAFT Figure 26—Continuous Gas-lift—Normal Backpressure Higher Than Test Backpressure




Trouble Well is being tested in test separator. Test separator pressure is lower than normal backpressurein production separator.

Recommendation Remove high normal backpressure or test against same high backpressure for accurate flow test.

Remarks It is not possible to obtain an accurate production test on the well under the above conditions.

Figure 27—Continuous Injection—Well Shut In




Trouble Well is closed in.

Recommendation Check to see why it is closed in.

Remarks On checking, it was noted that the well had produced its monthly allowable and had been closedin. This can hurt some oil wells. It may be better to cut the daily production and produce the wellconstantly.



DRAFT Figure 28—Continuous Injection—Well is Heading Periodically




Trouble Not serious, well is periodically “heading.”

Recommendation Check to see if system gas pressure fluctuates

Remarks Reasonable good operation. Well has a tendency to “head” which could be caused by erraticvalve operation or a fluctuating system pressure.



DRAFT

Figure 29—Continuous Injection—Well is Unloading

Operation Unloading of a continuous gas-lift well

Type of Well Continuous gas-lift well


Trouble Supply gas was closed to well. Well was off production 5 1/2 hours.

Recommendation Validate that the well unloaded to the design operating depth.

Remarks A choke was used on the gas line to control the rate of gas injection. When the gas was firstturned on, an immediate surge of fluid returned from the tubing as the well was completely full ofsalt water. When the liquid volume displaced in the annulus stabilized to the gas volume rate ofthe injection gas, the tubing pressure remained at 50 psig until the top valve was uncovered andgas entered the tubing. A surge in tubing pressure is noted as each valve is uncovered. The wellfinally stabilized on the 4th valve.



DRAFT

14.1.4 Examples of Pressure Charts from Intermittent Gas-lift Wells

These two-pen recorder charts illustrate most of the common problems that may occur in intermittent gas-liftoperation. These may be used to spot problems before they become severe. The charts are hand drawn so examplesof malfunctions could be exaggerated for clarity.

Figure 30—Intermittent Injection—Varying Injection Frequencies

Remarks a) Cycle frequency too long. Tubing kicks are low and thick. A “thick” tubing kick may beassociated with a long cycle frequency. Or, it may be an indication of a long flowline, a flowlinewith too small an ID, a restriction in the flowline, or even a surface choke.

b) In this case, increased cycle frequency yields tall thin tubing kicks and more production.

c) Cycle frequency too fat. Tubing pressure does not have time to reduce to normal.



DRAFT

Figure 31—Intermittent Injection—Varying the Injection Pattern

Remarks a) Erratic gas system pressure. The pressure has declined after timer was adjusted so that nowtwo injections are required per cycle.

b) Timer is then opened for longer injection. When gas system pressure increases, too much gasis used.

c) To help stabilize gas system pressure, use choke and timer.

d) Injection frequency too fast; gas-lift valve is not loaded so does not open until second injection.Too much gas is evident in tubing kick. Reduce injection frequency for better operation.



DRAFT

Figure 32—Intermittent Injection—Injection Pressure Too High

Remarks a) Injection rate too high. May cause more than one gas-lift valve to open. This condition is seenon the casing pressure by a change in the pressure decline rate after a gas-lift valve closes.The multiple “points” on the tubing pressure also indicate this condition.

b) Too much gas. Tubing kicks are too high and too thick. Casing pressure decline is slow.



DRAFT

Figure 33—Intermittent Injection—Well Loading Up

Remarks a) Well loading up. Evidence of excessive fluid load when gas-lift valve opens early. Problem isshown by shorter and wider tubing kicks until the lower valve becomes submerged andoperation continues on an upper valve. A decline in produced fluid results.

b) Well unloading. This illustrates how the fluid load decreases from a maximum when a gas-liftvalve first operates to a minimum when the valves operate the last time just before transferringto the next lower valve.



DRAFT

Figure 34—Intermittent Injection—Well Choked

Remarks a) Choked well (flowline choke). Restriction of choke causes slug velocity to be slow andpressure reduction period to be long. Also, tubing pressure is too high.

b) Flow line restriction. About the same effect as choke. Tubing pressure changes are gradualbecause restriction is distant from wellhead.



DRAFT

Figure 35—Intermittent Injection—Well Has Leaks

Remarks a) Leak in surface intermitter. Good operation is maintained.

b) Small leak in tubing string. Between each cycle, the casing pressure declines slowly after thegas-lift valve closes. Tubing kicks are very good.



DRAFT

Figure 36—Intermittent Injection—Well Has Tubing Leak

Remarks a) Leak high in tubing. Leak is small since tubing kicks are normal. First sign of leak is seen whencasing pressure continues to decrease after gas-lift valve closes. When gas to casing is shutoff casing declines to a value near the tubing pressure.

b) Leak low in tubing. Operating pressure about the same as above. Difference shows when gasto casing is shut off. Then casing pressure declines to a value well above the tubing pressure.(Fluid seal over the valve.)



DRAFT

Figure 37—Intermittent Injection—Well Has Large Tubing Leak

Remarks Large leak in tubing string. At first, it shows as a small leak, then leak is such that the casingpressure sometimes fails to open the gas-lift valve. When the leak exceeds the cycle gasrequirement, the casing pressure declines well below the normal range and a saw tooth pattern istraced. The tubing pressure reaches a steady, elevated pressure.



DRAFT

Figure 38—Intermittent Injection—Gas Injection Pressure Too Low

Remarks Gas line pressure becomes too low. Casing pressure fails to get high enough. Tubing kickschange from good slugs, to small slugs, to a misty spray.



DRAFT

Figure 39—Intermittent Injection—Plugging

Remarks a) Plugged valve. Very slow decline of casing pressures is an indicator of this problem. The tubingpressure kicks are rounded and misty because of excessive fall back. As condition gets worse,the casing pressure stays above valve closing pressure and tubing pressures stabilize. Then,only gas is obtained with no fluid.

b) Plugged tubing, very similar to situation A, but tubing pressure reflects injection cycles. Verylittle fluid is produced.



DRAFT

Figure 40—Intermittent Injection—Injection Rate Too Low

Remarks a) Not enough gas. Fall back is excessive so fluid recovery is small. Tubing pressure hasrounded, sluggish kicks. Casing pressure operating spread is too small.

b) Not enough fluid. Casing pressure operating spread is normal, but tubing pressure is roundedand sluggish.



DRAFT

14.2 Acoustical Surveys

The well “sounder” or sonic fluid level device is an acoustical instrument which works on the echo or reflectionprincipal, similar to seismic data acquisition. The sound pulse is initiated at the wellhead by either an explosion or animplosion with the echoes or sound reflections being picked up by a microphone and recorded on a strip chart or bycomputer.

The firing head and microphone of the sonic device are generally connected to a valve in communication with thetubing-casing annulus. The sound waves travel down the annulus. Each tubing collar, gas-lift mandrel, hole in thetubing or casing, and the static fluid level in the annulus reflects the sound to varying degrees. These sound echoesare received by the microphone, amplified, and recorded on a moving strip chart or by computer.

The liquid level in the well reflects most of the sound and is recorded as a very large deflection compared to otherreflections caused by collars or mandrels, etc. The following illustrations show typical sonic recordings.

Figure 41—Typical Acoustical Recording

Upper CollarsAccented

Deep CollarsAccented

(for Greater Accuracyin Deep Wells)

Liquid LevelAccented



DRAFT

14.2.1 Applications of Acoustical Surveys

Well sounding devices can determine a variety of diagnostic information:

a) find casing or tubing fluid level;

b) estimate the operating gas-lift valve; actually the deepest point to which the well has been unloaded;

c) estimate the SBHP if there is no packer;

d) locate the approximate depth of leaks in the tubing (fill casing, unload to leak, “sound” the casing annulus fluidlevel);

e) locate mandrel depths;

f) determine if tubing-casing communication exists below the tubing’s static fluid level.

Figure 42—Typical Acoustical Recording

Start

Fluid Level

Rebound

ReboundSonic

Reflection

Finish

Rebound

#6 Valve102 Jts.

#7 Valve105 Jts.

50 Jts.

60 Jts.

#2 Valve

96 Jts.

90 Jts.

82 Jts.

80 Jts.

70 Jts.

68 Jts.

#3 Valve

#4 Valve

#5 Valve

45 Jts.#1 Valve

30 Jts.

40 Jts.

20 Jts.

1 Jt.



DRAFT

14.2.2 Limitations of Acoustical Surveys

The fluid level in the casing does not always indicate the operating valve depth. The casing fluid level only indicatesthe deepest point to which a well has been unloaded, if there is no leakage of fluid from the tubing back in to thecasing. This level does not necessarily indicate the current point of operation.

The information provided by well sounding devices can be a help in gas-lift troubleshooting, especially whencombined with, and used to verify, the other methods.

14.3 Tagging Fluid Level

A wireline unit may be used to “tag” the tubing fluid level to determine the point of operation in an intermittent gas-liftwell, or “tag bottom” to find possible obstructions, sand bridges, or fill above the perforations.

This can locate the source of many problems. If the tubing fluid level is below the bottom valve, or an obstruction islocated in the tubing, the cause may be identified. Unfortunately, this method is not accurate enough to determine thespecific operating valve, and can sometimes be misleading.

14.3.1 Application of Tagging Bottom with a Wireline Unit

A wireline unit can give information concerning the operation, or cause of improper operation of a gas-lift well asfollows:

a) plugged or obstructed tubing;

b) paraffin, scale, or other deposits;

c) well-bore material, e.g. sand, with has filled into or above the perforated interval;

d) abnormally low fluid level, at or below the gas-lift valve, can indicate SBHP decline or formation damage;

e) abnormally high fluid level, above the intended operating valve, can indicate tubing leaks, or gas-lift valvemalfunctions.

14.3.2 Limitations of this Method

Tagging the fluid level in an intermittent gas-lift well with wireline tools can give an incorrect estimation of the operatingvalve. Fluid feed-in will sometimes raise the fluid level before the wireline tools can be lowered in the hole. In addition,fluid fallback will always occur after the lift gas has been shut off. Both will cause the observed fluid level to be abovethe operating valve.

Ensure that the input gas valve is closed prior to closing the wing valve, or the gas-lift system pressure will drive thefluid level back down the hole and below the point of operation, thus giving erroneous fluid level data. The wing valveonly needs to be closed if the flowline check valve is leaking.

14.4 Flowing Pressure Surveys

See Section 10 of this document for detailed recommended practices for running downhole pressure surveys. Thissection provides additional guidelines.

In this survey a pressure-recording instrument is run in the well under flowing condition while the well is being tested.On surging or heading wells, a no-flow (or no-blow) device should be run with the tools to prevent the tools from being“blown up the hole.” The no-flow device is equipped with “dogs” or slips that are activated by sudden movement up



DRAFT

the hole. However, they are not 100 % reliable, so care must be taken when running a flowing pressure survey in aheading well.

The pressure-recording instrument is stopped above and below each gas-lift valve for a period of time; and recordsthe pressures at each valve. Many operators consider the stops below the gas-lift valves most important. Stops canalso be made above the valves if the well is not heading or surging too strongly. Making additional stops betweenvalves can be helpful in plotting and interpreting the survey. From this information, the point of gas injection can bedetermined along with the FBHP. This survey is the most accurate way to determine a gas-lift well's performance.Flowing pressure surveys can accurately determine or provide information as follows:

a) depth of gas injection;

b) FBHP;

c) PI or IPR of the well. (This can be calculated when the production test data and SBHP data are gathered inconjunction with the flowing pressure data.);

d) location of tubing leaks within the range of the pressure stops;

e) a base line reference of well performance to aid in identifying future problems;

f) provide information for the redesign of valve spacing.

The flowing pressure survey in the well should be run with the well producing as close to normal as possible, so it canbe used to identify normal operating conditions and gas-lift problems. However, the FBHP survey should be run withthe well producing as stable as possible, so the well's productivity can be accurately determined. This may requiretwo separate surveys, or a change in the well's operation between the part of the survey used to evaluate the gas-liftsystem performance and the part used to evaluate the well's inflow productivity.

14.4.1 Objective of a Flowing Pressure Survey

Flowing pressure surveys help define what is occurring in the well under a given set of conditions, so predictions canbe made about well performance under different sets of conditions (a new valve spacing, a higher operating pressure,a different gas injection rate, or a deeper point of injection).

To accomplish this, the well test conditions must accurately duplicate normal producing conditions. The survey mustbe initiated under normal flowing conditions. Do not shut in the well prior to the survey. If necessary, a “crown valve” or“swab valve” must be added to the tree so the wireline unit can rig up the lubricator without shutting in the well.

a) Enter the well under flowing conditions.

Entering the well under flowing conditions reduces the possibility of inaccurate results. The well must be producingunder conditions that are representative of normal operations during the survey. Conducting a survey on a gas-liftwell that is unloading will yield inaccurate data. (However, purposefully running pressure and temperatureinstruments during unloading can give transient pressure and temperature data that can be used to help designgas-lift valve settings.)

b) Test the well during the survey.

To get productivity data, a well test must be conducted during measurement of the drawdown conditions that existduring the survey. The test should be as long as possible. A 24-hour test can be obtained if the test is started theday before the survey. The required test data includes measurements of oil and water per day, total return gasvolume, injection gas volume, and both fluid and gas gravities. The well must be producing stably during this part ofthe test.



DRAFT

c) Record tubing and casing pressures.

A calibrated two-pen recorder chart, or comparable device, should be used to document tubing and casingpressures during the survey. This information is needed to determine gas-lift valve performance.

d) Locate leaks.

If a tubing leak is suspected, at least one stop should be made between each mandrel. It may be necessary tomake more stops in the area of the suspected leak. This will allow more accurate location of the leak.

e) Run a small-diameter pressure-recording instrument.

Running a small-diameter pressure-recording instrument is desirable to minimize the restriction to flow. This servestwo purposes.

1) It will minimize a reduction in flow rate due to a restriction.

2) It will minimize the chances of the tools being “blown up the hole.” Most pressure recording instruments areavailable in 1 1/4-in. (3.18-cm) diameter, with some being available in 1 3/4-in. (4.45-cm) diameter.

f) Take safety precautions.

Since the well is entered under flowing conditions, take precautions to reduce the chances of the tools being“blown up the hole.”

In addition to small-diameter pressure recording instruments, weighted stems and “no-flow latches” should be usedon wells that may surge or head. Upward movement of the tool string activates a set of slips in the no-flow latch toprevent being “blown up the hole.” Be sure the diameter of the stem and/or tools above the no-flow latch is notgreater than the diameter of the stem and tools below the no-flow latch. If the largest restriction to flow is above theno-flow latch, the force trying to blow the tools up the hole will be above the no-flow latch and thereby prevent itfrom activating.

If the pressure-recording instrument goes in the hole freely for the first 100 ft (30.48 m), no further troubleshould be encountered because maximum fluid velocities occur near the surface of the well. 15 ft (4.57 m) of1 1/4-in. (3.18-cm) stem should be sufficient in most cases. A typical assembly might consist of a rope socket,5 ft (1.52 m) of 1 1/4-in. (3.18-cm) stem, “no-flow” latch(es), 10 ft (3.05 m) of 1 1/4-in. (3.18-cm) stem, andpressure recording instruments [if 1-in. (2.54-cm) pressure recording instruments are being used, substitute1-in. stem (2.54 cm)].

g) Using tandem instruments.

Running tandem temperature/pressure recording instruments can save time, effort, and money by providing aback-up if one instrument fails or is reporting erroneous pressures. Most electronic instruments record bothtemperature and pressure. This allows back-up capability for both temperature and pressure with only twoinstruments in the tool string.

h) Making temperature recording.

Recording temperatures in conjunction with flowing pressure surveys can sometimes show the cooling effect at thepoint(s) of gas entry into the production stream. Temperature surveys may identify problems such as valvesworking improperly due to being designed with erroneous temperature data or locating leaks due to the coolingeffect associated with the expansion of gas.



DRAFT

Many electronic instruments have built-in temperature recording capability. However, if they do not, it will benecessary to run a separate temperature-recording instrument in tandem with a pressure instrument.

In addition to the stop in the lubricator, it is a good practice to make a stop below the mud line for offshore wells andseveral hundred ft below the permafrost level for wells in Arctic areas so that a true flowing temperature gradientcan be determined.

i) Instrument pressure recording range and clock rotation.

For mechanical pressure recording instruments, the maximum anticipated pressure in the well should not exceedapproximately 75 % of the maximum pressure range of the instrument. This practice will eliminate inaccuracy dueto recording low pressures on an instrument designed to record high pressures.

j) Fast clock rotation.

A fast clock rotation makes it easier to detect and interpret pressure changes. It is desirable to run as fast aspossible depending on the depth of the well, the number of stops, and the desired time on “bottom.” Fortemperature elements that operate on vapor pressure, the recorded temperature must be in the upper half of therange of the instrument for good sensitivity.

14.4.2 Plotting Survey Results

Graphically plotting the results of a flowing pressure survey aids in interpretation. Figure 43 shows a flowing survey ona continuous gas-lift well. The well is operating from the 5th valve making 1,000 barrels (159.0 m3/day) per day asindicated by the sharp break in the flowing pressure gradient.

Figure 43—Flowing Pressure Survey

0

1000

2000

3000

4000

5000

6000

7000

8000

90000 400 800 1200 1600 2000 2400 2800

Pressure (psig)

Dep

th (f

t)

6

5

4

3

2

1

Casing pressure flowing0.12 psi/ft (avg.)

0.44 psi/ft (avg.)

Working fluid level

Tubing = 21/2"

Fluid = 1000 bbls/day

Input gas-fluid ratio = 400/1



DRAFT

Figure 44—Results of Flowing Pressure and Temperature Surveys Conducted During Intermittent Operation of High Capacity Well to Locate Operating Valve

2000

2400

2800

3200

3600

4000

4400

4800

5200

5600

6000

120 130 140 150 160 170 180 190

Temperature (F)

Dep

th (f

t)

Minimum flowing tubingtemperature traverseValve at 3063'

Valve at 3921'

Valve at 4534'

Valve at 5025'

Valve at 5333'

Maximum flowing tubingtemperature traverse

B—Plot of Minimum and Maximum FlowingTubing Temperatures Versus Depth

2000

2400

2800

3200

3600

4000

4400

4800

5200

5600

6000

200 300 400 500 600 700 800 900

Pressure (psig)

Dep

th (f

t)

Minimum flowing tubing pressuretraverse

Valve at3063'

Valve at 3921'

Valve at 4534'

Valve at 5025'

Valve at 5333'

Minimum casingpressure

Maximum casing pressure

Maximum flowingtubing pressuretraverse

A—Plot of Minimum and Maximum FlowingTubing Pressures Versus Depth



DRAFT

The fluid level in the casing was “sounded” with an acoustical instrument and found just above the 6th valve. Widegas-lift mandrel spacing prevented operating from the bottom valve. Re-spacing the mandrels increased production to1,600 barrels/day (254.4 m3/day). The surface and subsurface pressure measurements combined with casing fluidlevel data gave an accurate picture of well performance.

Figure 44 shows flowing pressure and temperature surveys conducted on an intermittent gas-lift well. Thetemperature data compliments the pressure survey. The operating valve(s) are hard to determine from the pressuresurvey alone; however, the temperature survey clearly indicates operation from the valves at 4,534 ft (1,382 m) and5,025 ft (1,532 m).

As the tubing pressure goes from the minimum to maximum value at 5,025 ft (1,532 m), it slightly exceeds the casingpressure. This could indicate initial operation from the valve at 5,025 ft (1,532 m) and valve interference from thevalve at 4,534 ft (1,382 m), or that the well is lifting from 4,534 ft (1,382 m) with the lower valve opening slightly afterthe 4,534 ft (1,382 m) valve opens. In either case, most of the gas appears to be entering at 4,534 ft (1,382 m) asindicated by the temperature survey. Temperature fluctuations at the operating valve are not nearly as easy torecognize on high-volume wells.

1) Continuous gas-lift wells.

a) Verify the well has been flowing to the well test separator until it is stabilized.

b) Put the well on test at a stabilized rate before running the bottomhole pressure survey. The timing of the gas andfluid test, two-pen pressure measurements, separator gas meter measurements, and input gas measurementsshould correspond with the pressure traverse. Mark all charts to show the beginning and the end of the pressuresurvey.

Some operators like to make stops both below and above each gas-lift valve. If a temperature survey isconducted to locate the depth of gas injection, the instrument must be stopped at the depth of the mandrel orimmediately above it to record the cooling effect of gas entering the production stream. High flow rates relativeto the tubing size may prevent or minimize the chances of detecting a temperature change.

c) Pressure/temperature recording instruments should be equipped with one and preferably two “no-blow” latches,if the well may be surging or heading.

d) Install the lubricator and pressure recording instruments. Record the surface pressure in the lubricator for 15minutes. Calibrate the pressure in the lubricator with a dead-weight tester. Run the survey instruments, makingstops 15 ft to 30 ft (4.57 m to 9.14 m) below each gas-lift valve for 15 minutes. Shorter or longer stops may bedictated by well conditions. For example, if a well is surging or heading, each stop should be for 1.1 to 1.25times the heading period. Do not shut the well in while rigging up or recording flowing pressures in the tubing.

e) Leave the recording instruments on bottom for at least 30 minutes, or longer if a static pressure is to be read.

f) Casing pressure should be determined with a dead-weight tester or recently-calibrated two-pen pressurerecorder or calibrated pressure gauge.

g) If a SBHP is desired, shut off the lift gas and close the wing valve. Leave instruments on bottom until thepressure stabilizes.

If a tubing leak is suspected, make one or more stops between valves near the depth of the suspected leak sothe location of the leak can be determined.



DRAFT

2) Intermittent gas-lift wells.

a) Flow the well to the test separator for 24 hours to obtain a stabilized production rate.

b) Put the well on test before running the bottomhole pressure survey. The well test should be for a minimum of sixhours. Production test information, the two-pen pressure measurement, the separator gas measurement, andthe input gas measurement should be tuned with the pressure traverse.

c) Pressure/temperature recording instruments should be equipped with one and preferably two “no-blow” latches.

d) Install the lubricator and pressure recording instruments. Record the surface pressure in the lubricator during atleast one complete cycle. Calibrate the pressure in the lubricator with a dead-weight tester. Run theinstrument(s), making stops 15 ft to 30 ft (4.6 m to 9.1 m) below each gas-lift valve. Do not stop above the valvein intermittent gas-lift wells. Be sure to record a maximum and minimum pressure at each gas-lift valve byremaining at each valve for at least one complete cycle. Do not shut in the well while rigging up or recordingflowing pressures in tubing.

e) Leave instruments on bottom for at least two (2) complete intermitting cycles.

f) High and low tubing and casing pressures should be checked with a dead weight tester, recently calibrated two-pen recorder, or calibrated pressure gauge.

g) If a SBHP is desired, shut off the gas and allow the well to equalize to the separator pressure. Leaveinstruments on bottom for a minimum of 12 hours.

14.4.3 Using Surveys to Calibrate Computer Models

As described elsewhere in these recommended practices, pressure surveys can be used to calibrate computermodels of pressure profiles in wells. These calibrated models can then be used in troubleshooting to, for instance,predict the depth of gas injection from the casing to the tubing.

If it is not possible to run a pressure survey due to the condition or location of the well, e.g. in some sub-sea wells, itmay be possible to use information from installed downhole pressure and temperature gauges to calibrate the model.These gauges are installed in many sub-sea wells.

Another way to assist with calibrating computer models may be with downhole distributed temperature surveys.These measure the temperature from the top to the bottom of the well while the well is producing. They can showlocation(s) where gas is entering the tubing.

15 Recommended Practices for Dealing with Wells That Produce Sand15.0 Purpose

15.1 Recommended Practices

The following practices are recommended for dealing with wells that produce sand:

— consider gas-lift as one of the better methods for artificially lifting wells that produce sand or other solids; sand isactually produced from the well to the surface without passing through the gas-lift valves;

— do not intermittently operate wells that produce sand; pressure surging may aggravate sand problems.Continuous gas-lift with constant rates is best for wells that produce sand. The fluid flow velocity must be



DRAFT

sufficient to keep the sand suspended and bring it to the surface. Sometimes this requires higher than normalgas/liquid ratios;

— produce the well at lower rates to prevent excessive sand entry from the formation into the wellbore, if productionof sand from the well is sensitive to the well’s production rate; that is, if more sand is produced at higher liquidproduction rates. Choking or controlling the injection gas to obtain lower stabilized production rates mayaccomplish this;

— do not install chokes in the flowline as this may worsen the sand problem. Improperly sized chokes in flowlinescan cause surging and reduce velocities thus allowing the sand to fall out in the tubing, causing plugging orbridging in the tubing. If chokes are used in the flowlines, they should be used in conjunction with sand probes.Through trial and error, the well should be produced with the largest choke possible that will not result insignificant sand production;

— do not use standing valves in continuous gas-lift wells. If they are in place, remove them to eliminate possibleplugging with the sand;

— frequently check and clean the flowline, manifold, test separator, and bulk production separator if a well producessand;

— avoid any sharp piping bends, restrictions, or other locations that could be subject to sand erosion. There havebeen some severe accidents caused by sand erosion of flowlines, manifolds, and other parts of the pipingsystem.

16 Typical Locations of Gas-lift Problems16.0 Purpose

Here are typical sources and/or locations of gas-lift problems, and appropriate steps to address them. Problems ingas-lift wells are usually associated with one or more of three areas: gas injection or inlet problems, well production oroutlet problems, and downhole problems.

Inlet problems include: input choke or control valve sized too large or too small, fluctuating line pressure, or pluggedchoke. Outlet problems include: high backpressure due to presence of a flowline choke, a closed or partially closedwing or master valve, or a plugged flowline. Downhole problems include: a cutout valve, leaks, restrictions in thetubing string, or sand covered perforations.

Often these problems can be observed and diagnosed on the surface. If problems are not found on the surface,check if the problems are downhole and if they are wellbore or gas-lift equipment problems. Carefully troubleshoot agas-lift well before scheduling a workover rig!

16.1 Gas-lift Injection or Inlet Problems

16.1.1 Injection Gas Choke Too Large

Check if the casing pressure is at or above the design operating pressure. If the pressure is too high, it can causereopening of upper unloading gas-lift valves and excessive gas usage and/or unstable (heading) operation.

16.1.2 Injection Gas Choke Too Small

Check for reduced fluid production because of insufficient gas injection. This condition can prevent the well from fullyunloading, or it can cause heading. If gas is injected through an upper unloading valve, due to incomplete unloading,the well will produce at a lower than desired rate.



DRAFT 16.1.3 Low Injection Pressure

Verify that the injection gas line is supplying full pressure to the wellsite and the injection pressure is above the designoperating pressure required by the gas-lift valves. This condition can occur if the control valve is sized too small, isplugged, or is “frozen.” Freezing can also occur at low spots in the injection line between the manifold and the well.

Eliminate choke freezing by dehydrating the gas, continuous injection of methanol in the lift gas, heating the gasupstream of the injection choke, or use of a heat exchanger.

Check the gas injection rate to separate freezing problems from low-injection pressure problems due to a hole in thetubing or a cutout valve.

16.1.4 High Injection Pressure

Verify that the gas-lift input control valve or choke is not too large. Check for excessive gas usage due to reopeningupper gas-lift unloading valves.

If a high injection pressure is accompanied by a low injection rate, the operating gas-lift valve or orifice may have a toosmall choke or may be partially plugged. Alternatively, there may be a high tubing pressure that is reducing thedifferential pressure between the casing and the tubing. A high injection pressure can also be caused by highbackpressure at the wellhead preventing the well from unloading to deeper valves. To address a high tubing orwellhead backpressure, remove any flowline choke or restriction.

16.1.5 Verify Wellhead Pressure Measurements

Verify the accuracy of pressure measurements; they can cause false indications of high or low injection pressures.Check the wellhead casing and tubing pressures with a calibrated gauge or dead weight tester.

Figure 45—The Gas-lift System

Down Hole

Inlet Outlet

GasHeader

Controlleror

Casing Choke

Meter Valve

Wing &MasterValve

Remove Choke Sep.

Comp.

ProductionHeader

Tubing

Casing

Packer

Perforations

Gas LiftValve



DRAFT

16.1.6 Low Gas-lift Injection Rate

Verify the gas-lift injection line is supplying full injection pressure at the well site, and the injection pressure is abovethe operating pressure required by the gas-lift design. Also, ensure the gas-lift line valve is fully open and the injectionvalve is not too small, frozen, or plugged.

Check the production rate of the well if the surface conditions appear normal. A higher than anticipated productionrate, and the resulting higher temperatures at the depths of the gas-lift valves, will cause the valve dome (set)pressures to increase. This could restrict the gas input by throttling the operating valve(s) partially closed. This cannothappen if the operating point is an orifice.

16.1.7 Excessive Gas-lift Injection Rate

Verify that the injection control valve is not too large or that the injection pressure is not too high. If the injectionpressure is above the design pressure, this may cause one or more of the upper unloading gas-lift valves to be open.High wellhead pressure can cause multi-point injection from upper unloading gas-lift valves and result in excessivegas usage.

Verify there is no tubing leak, or a cutout valve or mandrel. These can permit a high injection rate, but will usually alsoresult in an injection pressure that is lower than the design operating pressure of the valves.

16.1.8 lntermitter Problems

Set the intermitter cycle time to obtain the maximum fluid production rate with the minimum number of gas injectioncycles per day. Injection duration should be adjusted to minimize the “tail gas” produced after the liquid slug reachesthe surface. Verify that the intermitter has not stopped - whether it is a manual wind or battery-operated model.

Eliminate problems inherent with mechanical clocks by using electronic intermitters. Intermittent gas-lift wells that areproducing more than 250 bbls/day (39.75 m3/day) should be evaluated for continuous gas-lift application.

16.2 Gas-lift Production or Outlet Problems

16.2.1 Valve Restrictions

Verify all valves in the Christmas tree, at the wellhead location, and at the production header or manifold, are fullyopen, and that an undersized valve or reduced port-sized valve is not in the line [e.g. a 1-in. (2.54 cm) valve in a 2-in.(5.08) flowline]. Other restrictions may result from a smashed or crimped flowline. Check places where the linecrosses a road.

Keep wellhead backpressure to a minimum since this pressure must be overcome by the gas-lift system. Anyincrease in wellhead pressure results in a corresponding increase in the FBHP. This reduces the differential pressurefrom the reservoir into the wellbore, thereby reducing the well's inflow or production rate. A high backpressure mayresult in use of excess gas in an attempt to overcome the high pressure.

Verify that no choke is in the flowline. Even with no choke bean in a choke body, the I. D. of the choke body is usuallyless than the full I. D. of the flowline and can cause a pressure drop. Remove the choke body if possible.

Minimize the number of 90 ° turns; they cause a higher backpressure and should be removed or streamlined wherefeasible.

Minimize high backpressures that result from paraffin, salt, scale, or sand build-up in the flowline. Hot oiling or piggingthe line will generally remove paraffin. It may not be possible to remove scale, depending on the type of scale. If thisexists, check with a chemical engineer or the flow assurance group.



DRAFT

Reduce the pressure in undersized or long flowlines by “looping” the flowline with an inactive line. The may apply tocases where the flowline I. D. is smaller than the tubing I. D.

16.2.2 High Backpressure

Correct a partially stuck check valve in the flowline; it causes excessive backpressure.

Remove as many restrictions from the system as possible. High wellhead pressure can cause the well to lift fromupper unloading valves and prevent unloading to the operating valve or orifice, resulting in excessive gas usage andreduced production.

16.2.3 Separator Operating Pressure

Keep the production separator pressure as low as possible for gas-lift wells. Often a well may be flowing into a high orintermediate pressure system when it dies and is placed on gas-lift. Verify that it is switched to the lowest pressuresystem available.

16.3 Downhole Problems

16.3.1 Hole in Tubing

Detect and correct holes in the tubing. Indications of a hole include an abnormally low injection pressure and anexcessively high injection rate, usually accompanied by a reduced production rate. Heading is another symptomalthough the intensity of heading may range from very little to severe. Similar symptoms can be caused by a leakingor cut gas-lift valve or mandrel, and a leaking tubing hanger or packer leak.

Detect holes in the tubing above the fluid level by bleeding down the casing pressure. When the pressure is bleddown, the tubing gas head will begin to bleed off through the hole to the casing.

Deeper holes below the well's static fluid level present another problem as reservoir inflow may come into play. As thecasing pressure is bled off, tubing fluid moves through the hole and enters the casing. The tubing pressure thenbleeds off at a slower rate than the casing pressure since the reservoir will be feeding into the tubing and replacingtubing fluid that was lost to the casing.

Weaker reservoirs will not be able to replace this tubing fluid as fast as the casing can be bled off. This may result inan overall drop in the tubing pressure and indicate that a hole exists. The drop in tubing pressure will usually not besustained for long as the reservoir will eventually catch up and pressure the tubing back up to its previous value. Astrong reservoir will be able to replace the tubing fluid as fast as it moves into the casing during bleeding. This willresult in a constant tubing pressure during the casing bleed-down even though a hole is present.

For this reason, an acoustical sounding device should be used to monitor the depth of the casing fluid level whilebleeding the casing pressure. Any rise in the casing fluid level proves that there is tubing-casing communication(hole) somewhere below the tubing’s static fluid level.

Confirm a hole in the tubing by the following procedure:

— equalize the production and casing pressures by closing the wing valve with the gas-lift injection gas still turned on;

— use an acoustical sounding device to find the initial casing fluid level after the pressures are equalized;

— shut off the gas-lift injection valve and rapidly bleed-off the injection pressure on the casing;



DRAFT

— monitor the production pressure. If it bleeds down as the injection pressure drops, then a hole or leak in thetubing is indicated. If the production pressure holds, find the casing fluid level at the new casing pressure usingthe acoustical sounding device;

— check for a rise in the casing fluid level. This is indication of tubing-casing communication or a packer leak;

— verify that there is no hole if the production pressure and casing fluid level hold. If no hole is present, both thecheck valves and gas-lift valves will be closed as the injection pressure bleeds to zero.

16.3.2 Estimate the Operating Gas-lift Valve by Surface Pressure Closing Method

This method can be used to estimate which is the operating gas-lift valve. IPO valves will continue to transmit gasuntil the injection pressure drops to the closing pressure of the valve. The operating gas-lift valve can sometimes bedetermined by shutting off the injection gas and observing the pressure to which the injection pressure decreases andstabilizes. This pressure is the surface closing pressure of the operating gas-lift valve.

By comparing this pressure with the design surface closing pressure of each valve, the operating valve can often bedetermined. However, note that the temperature and production pressure will also affect the “closing” pressure of thevalve. The observed surface “closing” pressure is not the same as the design closing pressure unless the productionpressure and temperature at the depth of the valve are equal to the “design” production pressure and temperaturevalues.

This method gives an approximate indication of the operating valve; it is not as accurate as a flowing pressure survey.As the injection gas is turned off and the casing pressure begins to decrease, the tubing pressure will also changedue to the decreased gas injection into the tubing. This will affect the closing pressure of the valve.

16.3.3 Check for a Well “Blowing Dry Gas”

Check for gas injection through an upper valve, with no production. The term “blowing dry gas” means the well isinjecting gas but has no production. When using IPO valves, verify the injection pressure is not above of the designoperating pressure. This may cause one or more upper unloading valves to open, and gas may be injected above thefluid level in the tubing. In addition, upper valves can be open if the temperature is well below the design value.

Verify that no hole exists in the tubing by the method described above. In addition, “sound” the casing fluid level witha sonic fluid level detection device. If the upper unloading valves are not held open by excess injection pressure, if nohole exists, and if the casing fluid level shows that the well has been fully unloaded to the operating valve or orifice,the well is probably lifting from the bottom valve or orifice.

Verify operation from the bottom valve or orifice by comparing the surface closing pressure with the design pressureof the bottom valve or orifice. The bottom valve is usually designed (“flagged”) so its surface operating pressure andclosing pressure are significantly less than the other valves in the string.

If the well is equipped with PPO valves and has an IPO valve on the bottom, blowing dry gas can indicate operationon the bottom valve, after the possibility of a hole in the tubing has been eliminated, because PPO valves will notremain open if there is no fluid pressure.

If the well is blowing around and is operating from the bottom valve, this indicates a lack of inflow from the formation.It is advisable to tag bottom to see if the perforations have been covered by sand or debris. If the well is equipped witha standing valve, verify the standing valve is not plugged. Standing valves are not used in continuous gas-lift wells.

16.3.4 Troubleshoot a Well Not Taking Gas

Eliminate the possibility of a frozen or plugged input choke, a closed injection gas valve, or closed valves on the outletside. If PPO valves are used without an IPO valve or orifice on bottom, this condition may indicate that all the fluid has



DRAFT

been lifted from the tubing and not enough production pressure remains to open the valves. Check for inflowproblems.

If IPO valves are used, verify the well started producing above the design fluid rate. A higher production rate mayhave caused the temperature to increase sufficiently to close (”lock-out') the valves. If high temperature is theproblem, the well will produce periodically then quit. If this is not the problem, verify the valve set pressures are nottoo high for the available injection pressure.

If the well has a surface or sub-surface safety valve, verify that it is open.

16.3.5 Troubleshoot a Well That is Flowing in Slugs

Evaluate several factors that can be cause this situation. With IPO valves, one cause may be port sizes are too large.This might occur if, for example, a well was initially designed for intermittent lift and was placed on continuous gas-liftdue to higher than anticipated fluid production rates. Large production pressure effects will exist because of the hightubing effect factors of the valves; the well will lift until the fluid gradient is reduced below the point where theproduction pressure would keep the operating valve open. Injecting through an oversized orifice at the operatingdepth can also cause slugging.

Temperature effects may also cause slugging. If the well started producing at a higher than anticipated fluid rate, thetemperature could increase; this could cause the valve set pressures to increase and cause them to close or “lockout.” When the temperature cools sufficiently, the valves will open again; the well will produce in cycles or heads.

Check production inflow from the formation. With IPO valves that have a high production pressure effect, or with PPOvalves, slugging can occur with limited inflow from the formation. The valves will not open until the tubing (production)pressure has risen far enough, thus creating a condition where the well will self intermit whenever adequate inflowhas occurred.

Evaluate tubing size. Heading can be caused if the tubing size is too large for the production rate being lifted.

Evaluate multi-pointing. Slugging can result from any condition that causes a well to work from two or more valves,such as an excessively long valve spacing in wells with a low injection rate, varying injection pressure, insufficient gasinjection rate, valve interference caused by flowline chokes, or incorrectly designed gas-lift valve set pressures andtemperatures.

16.3.6 Evaluate a Well That is Stymied and Will Not Unload

This condition occurs when the fluid column (tubing) pressure is heavier (higher) than the available gas-lift injectionpressure.

Apply injection gas pressure to the top of the fluid column (usually by using an equalizing line). This may temporarilydrive some of the fluid column back into the formation, thereby reducing the height of the fluid column, and allow thewell to unload with the available gas-lift pressure. Clearly, this will not work if there is a standing valve in the tubing.The check valves in the gas-lift valves will prevent the fluid from being displaced from the tubing back into the casing.

Do not use this process, which is referred to as “rocking” the well, if there is any chance of sand production. This candamage the sand control system in the well. It may also cause plugging of the formation interface to the wellbore byinjecting fine material back into the formation.

For PPO valves, “rocking” the well in this fashion may open an upper valve and permit the unloading operation tocontinue.



DRAFT

Sometimes a well can be “swabbed” to allow unloading to a deeper valve, but “swabbing” should be avoided if at allpossible due to the chances of sticking a swab cup, inadvertently unseating and pulling a retrievable gas-lift valve withthe swab, getting “blown up the hole,” or “sucking” sand into the wellbore.

Use these methods with extreme caution. Wells can be seriously damaged by pushing fluid back into the reservoirand/or by “rocking.” Unless there is a history of successfully using these techniques in a field, they should only beused as a last resort.

Consider these other possible approaches to unload a “stymied” well:

— find a way to temporarily reduce the wellhead backpressure;

— find a way to temporarily increase the gas-lift injection pressure. For example, ft may be possible to use a nearbygas well or a nitrogen truck to initially unload the well. Alternatively, it may be possible to temporarily raise thedischarge pressure from the compressor by temporarily raising the set pressure on the sales gas pressureregulator. (However, this may cause many unintended problems in a gas-lift system that serves many wells.);

— finally, it may be less expensive to pull the tubing and re-space the gas-lift mandrels than to risk causing damageto the formation or the sand control system in the well.

16.3.7 Correct any Valve That is Hung Open

Identify this if the gas-lift pressure will bleed below the surface closing pressure of any valve in the hole; but testsshow that no hole is present.

Try shutting the wing valve on the flowline, allow the casing pressure to build up as high as possible, and open thewing valve rapidly. This will create a high differential pressure across the seat of the operating valve and may removeany trash holding the valve open. Repeat the process several times if required.

Valves can sometimes be held open by salt deposition. If this is suspected, pump several barrels (m3) of fresh waterinto the casing to dissolve the salt. Where fresh water sensitive formations exist, verify this technique will not allow thewater to contact the formation.

If the above actions do not help, the gas-lift valve may be cut or eroded, or the bellows may be de-pressured. In thiscase, pull and replace the valve.

16.3.8 Correct Too Wide Mandrel Spacing

In some instances, a well will not fully unload due to excessively wide mandrel spacing. This occurs when a wellproduces at a higher rate or has a higher reservoir pressure than anticipated during the valve design.

This situation is essentially the same as the “stymied” well discussed above. If necessary, after considering all of therisks, the following procedures may be tried:

— try “rocking” the well as discussed when the well is “stymied,” as this will sometimes allow working down to lowervalves. But be careful!;

— if a high-pressure gas well (or a nitrogen truck) is nearby, use the high pressure for unloading. If IPO valves areused, this will require sufficient lift gas volume to over-ride the gas passage capability of the upper valve(s) toallow the injection pressure to build up;

— produce the well to an atmospheric tank or pit to minimize wellhead backpressure;



DRAFT

— try re-spacing the gas-lift mandrels, or as a last resort installing a pack-off gas-lift valve or orifice between existingmandrels, if the problem is severe and cannot be solved by one of the other methods;

— installing a pack-off valve, or any other action that places a hole in the tubing, is not recommended as a long-termsolution. If the well is stymied due to too wide mandrel spacing, and if there is enough extra production potentialto justify a workover, the recommended action is to work over the well to re-space the gas-lift mandrels so thewell can work down to the desired operating depth.

17 Possible Causes and Cures of Common Malfunctions of Gas-lift Systems

Malfunction Possible Cause(s) Possible Cure(s)

CommunicationBetween Casingand Tubing

Gas-lift valve cut or stuck open

— Replace leaking or unseated valve

— Check unloading procedures to avoid erosion

— Consider using different (harder) materials for seats

Gas-lift valve or dummy unseated

— Replace unseated valve or dummy

Gas-lift mandrel leak — Pull tubing, replace leaking mandrel

Packer leak — Reset packer

Tubing or tubing head leak — Pull tubing, repair leak, and re-run

Circulating sleeve leak — Close circulating sleeve

— Use pack-off for short-term correction

InjectionPressure Above/BelowNormal or Heading

Lifting from a higher gas-lift valve

— Analyze well and adjust to work down

Gas-lift valve plugged — Pull and replace valve

Higher temperature keeping valves open

— Pull and redesign valves for higher closed temperature

— Check by running a temperature survey

Gas-lift valve set pressures increase

— Pull valves, check for tail plug or gasket leaks

Well lifting from upper gas-lift valve

— Pull valve(s) and replace them, or if necessary, redesignor re-space them

Well lifting from more than one valve, injection pressure is heading

— Redesign valves and re-tune injection to preventmulti-valve operation and heading

Well lifting from orifice, but heading

— Port size too large, pull and replace

— Try lowering casing pressure

Small fluid per intermittent cycle

— Reduce intermittent cycle frequency

Injection meter reading incorrect

— Check meter, orifice, differential pressure taps,connecting lines



DRAFT

Malfunction Possible Cause(s) Possible Cure(s)

High BackPressure atWell Head

Plugged flowline — Check for partially closed valves, plugged check valves, paraffin,or sand

High separator pressure — Check backpressure regulator

Using too much gas — Adjust injection control to optimize gas

Choke in flowline — Remove any chokes or restrictions

Flowline too small — Loop line or replace with larger line

Sudden Dropin Production(Valve OperationAppears Normal)

Plugged formation — Clean out well, consider stimulation

Plugged tubing — Tag bottom. If plugged, clean out or pull tubing

Too much or too little gas — Adjust injection control to optimize gas

Standing valve stuck open/closed

— Pull standing valve and clean it

Subsurface safety valve closed

— Correct cause of premature closing, reset safety valve, pull valveis no longer needed

Fluid SlugVelocity LessThan 1,000 ft(304.8 m)in IntermittentGas-lift Well

Fluid load heavy — Increase injection cycle frequency

Low injection pressure — Increase pressure or space valves closer

Operating valve partially plugged

— Flush with fresh water or solvent

Tubing partially plugged — Cut paraffin or flush well with solvent

Poor injection valve response

— Increase injection gas rate to achieve rapid injection pressurebuild-up

Operating valve port too small

— Pull valve and replace with larger port



DRAFT

18 Gas-lift Troubleshooting Checklist

Gas-lift Troubleshooting ChecklistPage 1

Field: Well: Date:

Gas Injection Inlet or Source Problems

Possible Gas Injection Source/Supply Problems

[ ] Injection control choke or valve size may be too large — Upper valve(s) may be reopening — Too much gas may be injected

[ ] Injection control choke or valve size may be too small — May not be able to unload the well — Too little gas may be injected

[ ] Injection control choke or valve may be plugged — Choke or control valve may be frozen

[ ] Bad pressure gauges may be causing excessive or insufficient gas injection pressure

[ ] Gas-lift supply may be turned off or shut down

[ ] Gas-lift line pressure may be too low

[ ] Gas-lift line pressure may be fluctuating

[ ] Gas-lift intermitter being used on continuous gas-lift well

[ ] Gas-lift intermitter may be stopped — Cycle or injection time may be wrong

[ ] Gas-lift intermitter may be malfunctioning

[ ] Other problems/comments:

Corrective Action:



DRAFT


Gas-lift Well Production Outlet Problems

Possible Gas-lift Well Production Problems

[ ] Master valve or wing valve may be closed

[ ] There may be high wellhead backpressure due to: — A flowline choke — Flowline choke body may be too restrictive — An excessive number of 90º turns — A long flowline — A plugged or partially plugged flowline, with sand or paraffin — Emulsion — Hilly terrain — A number of canal crossings — A small flowline internal diameter — A valve closed at the header or manifold — A valve with a restricted internal diameter — A plugged or jammed flowline check valve — A valve leaking at the header allowing backpressure from other wells — Separator operating pressure may be too high — Separator orifice plate may be sized too small


Corrective Action:



DRAFT


Downhole Problems

Possible Downhole Gas-lift Problems

[ ] There appears to be no fluid feed-in. Fluid is standing at or below bottom gas-lift valve

[ ] Perforations appear to be covered by sand or other material

[ ] There may be a restriction(s) in the tubing string

[ ] The bottom valve may not be set deep enough

[ ] There may be a valve worn, cut out, or unseated

[ ] There may be a dummy cut out or unseated

[ ] There may be a leaking pack-off gas-lift valve

[ ] There may be a gas-lift mandrel leak

[ ] The gas-lift valve pressure(s) may be set too low. Cannot close one or more valves


Corrective Action:

[ ] The well is lifting from the bottom valve but: — Well may be under producing due to excessive wellhead backpressure — Well may be under producing due to low casing pressure or insufficient lift gas — Well is unstable due to too large an orifice or choke

[ ] There may be a tubing leak

[ ] Mandrel spacing may be too wide to allow the well to be unloaded

[ ] The gas-lift valve pressure(s) may be set too high. Cannot open one or more valves

[ ] Excessive injection or production pressure appears to be re-opening valve(s) up the hole

[ ] The well has dual gas-lift and: — One side of the dual may be robbing most of the gas — Too high or too low temperature may be affecting the valves in one string


DRAFT


DRAFT

Bibliography

[1] API Specification 11V1, Specification for Gas Lift Equipment

[2] API Recommended Practice 11V2, Gas-lift Valve Performance Testing

[3] API Recommended Practice 11V7, Recommended Practice for Repair, Testing, and Setting Gas Lift Valves

[4] API Recommended Practice 11V8, Recommended Practice for Gas Lift System Design and PerformancePrediction

[5] API Recommended Practice 11V10, Recommended Practices for Design and Operation of Intermittent andChamber Gas-lift Wells and Systems

[6] ISO 17078-2 1, Petroleum and natural gas industries—Drilling and production equipment—Part 2: Flowcontrol devices for side-pocket mandrels

1 International Organization for Standardization, 1, ch. de la Voie-Creuse, Case postale 56, CH-1211, Geneva 20,Switzerland, www.iso.org.

123


DRAFT

DRAFT

Invoice To (❏ Check here if same as “Ship To”)

Name:

Title:

Company:

Department:

Address:

City: State/Province:

Zip/Postal Code: Country:

Telephone:

Fax:

Email:

❏ Payment Enclosed ❏ P.O. No. (Enclose Copy)

❏ Charge My IHS Account No.

❏ VISA ❏ MasterCard ❏ American Express❏ Diners Club ❏ Discover

Credit Card No.:

Print Name (As It Appears on Card):

Expiration Date:

Signature:

Quantity Title Total

Subtotal

Applicable Sales Tax (see below)

Rush Shipping Fee (see below)

Shipping and Handling (see below)

Total (in U.S. Dollars)

★ To be placed on Standing Order for future editions of this publication, place a check mark in the SO column and sign here:

Pricing and availability subject to change without notice.

Date:

SO★ Unit Price

❏ API Member (Check if Yes)

Ship To (UPS will not deliver to a P.O. Box)

Name:

Title:

Company:

Department:

Address:

City: State/Province:

Zip/Postal Code: Country:

Telephone:

Fax:

Email:

Mail Orders – Payment by check or money order in U.S. dollars is required except for established accounts. State and local taxes, $10 processing fee, and 5% shipping must be added.Send mail orders to: API Publications, IHS, 15 Inverness Way East, c/o Retail Sales, Englewood, CO 80112-5776, USA.

Purchase Orders – Purchase orders are accepted from established accounts. Invoice will include actual freight cost, a $10 processing fee, plus state and local taxes.Telephone Orders – If ordering by telephone, a $10 processing fee and actual freight costs will be added to the order.Sales Tax – All U.S. purchases must include applicable state and local sales tax. Customers claiming tax-exempt status must provide IHS with a copy of their exemption certificate.Shipping (U.S. Orders) – Orders shipped within the U.S. are sent via traceable means. Most orders are shipped the same day. Subscription updates are sent by First-Class Mail. Other options, including next-day service, air service, and fax transmission are available at additional cost. Call 1-800-854-7179 for more information.Shipping (International Orders) – Standard international shipping is by air express courier service. Subscription updates are sent by World Mail. Normal delivery is 3-4 days fromshipping date.Rush Shipping Fee – Next Day Delivery orders charge is $20 in addition to the carrier charges. Next Day Delivery orders must be placed by 2:00 p.m. MST to ensure overnight delivery.Returns – All returns must be pre-approved by calling the IHS Customer Service Department at 1-800-624-3974 for information and assistance. There may be a 15% restocking fee.Special order items, electronic documents, and age-dated materials are non-returnable.

Effective January 1, 2008.API Members receive a 30% discount where applicable.The member discount does not apply to purchases made for the purpose of resale or for incorporation into commercial products, training courses, workshops, or othercommercial enterprises.

Available through IHS:Phone Orders: 1-800-854-7179 (Toll-free in the U.S. and Canada)

303-397-7956 (Local and International)Fax Orders: 303-397-2740Online Orders: global.ihs.com

2008 PublicationsOrder Form


DRAFT

API provides additional resources and programs to the oil and natural gas industry which arebased on API Standards. For more information, contact:

API MONOGRAM® LICENSINGPROGRAMPhone: 202-962-4791Fax: 202-682-8070Email: [email protected]

API QUALITY REGISTRAR(APIQR®)> ISO 9001 Registration> ISO/TS 29001 Registration> ISO 14001 Registration> API Spec Q1® RegistrationPhone: 202-962-4791Fax: 202-682-8070Email: [email protected]

API PERFORATOR DESIGNREGISTRATION PROGRAMPhone: 202-682-8490Fax: 202-682-8070Email: [email protected]

API TRAINING PROVIDERCERTIFICATION PROGRAM(API TPCPTM)Phone: 202-682-8490Fax: 202-682-8070Email: [email protected]

API INDIVIDUAL CERTIFICATIONPROGRAMS (ICP®)Phone: 202-682-8064Fax: 202-682-8348Email: [email protected]

API ENGINE OIL LICENSING ANDCERTIFICATION SYSTEM (EOLCS)Phone: 202-682-8516Fax: 202-962-4739Email: [email protected]

API PETROTEAM (TRAINING,EDUCATION AND MEETINGS)Phone: 202-682-8195Fax: 202-682-8222Email: [email protected]

API UNIVERSITYTM

Phone: 202-682-8195Fax: 202-682-8222Email: [email protected]

Check out the API Publications, Programs,and Services Catalog online at www.api.org.

Copyright 2008 – API, all rights reserved. API, API monogram, APIQR, API Spec Q1,API TPCP, ICP, API University and the API logo are either trademarks or registeredtrademarks of API in the United States and/or other countries.

THERE’S MOREWHERE THIS CAME FROM.


DRAFT


DRAFT

Product No. G11V53


api rp 11v5 final draft

Documents