96150641-ANSI-HI-9-6-5

40
9 Sylvan Way Parsippany, New Jersey 07054-3802 www.pumps.org ANSI/HI 9.6.5-2000 ANSI/HI 9.6.5-2000 American National Standard for Centrifugal and Vertical Pumps for Condition Monitoring

Transcript of 96150641-ANSI-HI-9-6-5

9 Sylvan WayParsippany, New Jersey07054-3802 www.pumps.org

AN

SI/H

I9.

6.5-

2000

ANSI/HI 9.6.5-2000

American National Standard for

Centrifugal and Vertical Pumpsfor Condition Monitoring

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

This page intentionally blank.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

ANSI/HI 9.6.5-2000

American National Standard for

Centrifugal and Vertical Pumpsfor Condition Monitoring

Secretariat

Hydraulic Institute

www.pumps.org

Approved February 23, 2000

American National Standards Institute, Inc.

Recycledpaper

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

Approval of an American National Standard requires verification by ANSI that therequirements for due process, consensus and other criteria for approval have been metby the standards developer.

Consensus is established when, in the judgement of the ANSI Board of StandardsReview, substantial agreement has been reached by directly and materially affectedinterests. Substantial agreement means much more than a simple majority, but not nec-essarily unanimity. Consensus requires that all views and objections be considered,and that a concerted effort be made toward their resolution.

The use of American National Standards is completely voluntary; their existence doesnot in any respect preclude anyone, whether he has approved the standards or not,from manufacturing, marketing, purchasing, or using products, processes, or proce-dures not conforming to the standards.

The American National Standards Institute does not develop standards and will in nocircumstances give an interpretation of any American National Standard. Moreover, noperson shall have the right or authority to issue an interpretation of an AmericanNational Standard in the name of the American National Standards Institute. Requestsfor interpretations should be addressed to the secretariat or sponsor whose nameappears on the title page of this standard.

CAUTION NOTICE: This American National Standard may be revised or withdrawn atany time. The procedures of the American National Standards Institute require thataction be taken periodically to reaffirm, revise, or withdraw this standard. Purchasers ofAmerican National Standards may receive current information on all standards by call-ing or writing the American National Standards Institute.

Published By

Hydraulic Institute9 Sylvan Way, Parsippany, NJ 07054-3802

www.pumps.org

Copyright © 2000 Hydraulic InstituteAll rights reserved.

No part of this publication may be reproduced in any form,in an electronic retrieval system or otherwise, without priorwritten permission of the publisher.

Printed in the United States of America

ISBN 1-880952-46-7

AmericanNationalStandard

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

iii

ContentsPage

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

9.6.5 Centrifugal and vertical pumps for condition monitoring

9.6.5.0 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

9.6.5.0.1 Use of this document. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

9.6.5.0.2 Monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

9.6.5.0.3 Control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

9.6.5.1 Power monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.1.2 Means of power monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.1.3 Power monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.1.4 Power control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.2 Temperature monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

9.6.5.2.2 Means of monitoring temperature . . . . . . . . . . . . . . . . . . . . . . . . . 4

9.6.5.2.3 Specific applications of temperature monitoring . . . . . . . . . . . . . . 4

9.6.5.2.4 Sealless pump liquid temperature . . . . . . . . . . . . . . . . . . . . . . . . . 4

9.6.5.2.5 Temperature monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . 5

9.6.5.2.6 Temperature control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

9.6.5.3 Corrosion monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

9.6.5.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

9.6.5.3.2 Means of corrosion monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

9.6.5.3.3 Corrosion monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . 6

9.6.5.3.4 Corrosion parameter control limits. . . . . . . . . . . . . . . . . . . . . . . . . 6

9.6.5.4 Leakage detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

9.6.5.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

9.6.5.4.2 Means of monitoring leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

9.6.5.4.3 Leakage monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

9.6.5.4.4 Leakage control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

9.6.5.4.5 Leakage control limits for gas seal consumption. . . . . . . . . . . . . . 7

9.6.5.5 Pressure monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

9.6.5.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

9.6.5.6 Vibration monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

9.6.5.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

9.6.5.6.2 Means of vibration monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

9.6.5.6.3 Vibration monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

iv

9.6.5.6.4 Vibration control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

9.6.5.7 Lubricant analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

9.6.5.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

9.6.5.7.2 Measuring metal particles from wear . . . . . . . . . . . . . . . . . . . . . . 9

9.6.5.7.3 Measuring contamination of lubricant . . . . . . . . . . . . . . . . . . . . . 10

9.6.5.7.4 Measuring lubricant degradation . . . . . . . . . . . . . . . . . . . . . . . . . 10

9.6.5.7.5 Lubricant sampling techniques . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.7.6 Lubricant monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.7.7 Lubricant control limits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.8 Shaft position monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.8.2 How to monitor shaft position . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.8.3 Shaft monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.8.4 Shaft control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.9 Rate of flow monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.9.2 Measuring rate of flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.9.3 Rate of flow monitoring frequency. . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.9.4 Rate of flow control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

9.6.5.10 Maintenance inspection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

9.6.5.10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

9.6.5.10.2 Maintenance inspection practice . . . . . . . . . . . . . . . . . . . . . . . . . 12

9.6.5.10.3 Maintenance inspection frequency . . . . . . . . . . . . . . . . . . . . . . . 13

9.6.5.11 Speed (rpm) monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

9.6.5.11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

9.6.5.11.2 Speed measurement methods . . . . . . . . . . . . . . . . . . . . . . . . . . 14

9.6.5.11.3 Speed monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

9.6.5.11.4 Speed control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

9.6.5.12 Bearing wear monitoring of plain bearings in sealless pumps . . 14

9.6.5.12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

9.6.5.12.2 Means of detecting bearing wear . . . . . . . . . . . . . . . . . . . . . . . . 14

9.6.5.12.3 Bearing wear monitoring frequency. . . . . . . . . . . . . . . . . . . . . . . 15

9.6.5.12.4 Bearing wear control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

9.6.5.13 Pre-installation hydrotest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

9.6.5.13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

9.6.5.13.2 Means of conducting a pre-installation hydrotest . . . . . . . . . . . . 15

9.6.5.13.3 Hydrotest monitoring frequency . . . . . . . . . . . . . . . . . . . . . . . . . 16

9.6.5.13.4 Hydrotest control limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

9.6.5.14 Design review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

9.6.5.14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

v

9.6.5.14.2 Design review practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

9.6.5.14.3 Design review frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Appendix A Condition Monitoring, Failure Modes . . . . . . . . . . . . . . . . . . . . . . 18

Appendix B Condition Monitoring Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Appendix C Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Tables

9.6.5.1 — Severity levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

9.6.5.2 — Frequency of monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

9.6.5.3 — Application guidelines for leakage monitoring systemsmechanical seals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

This page intentionally blank.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

vii

Foreword (Not part of Standard)

Purpose and aims of the Hydraulic Institute

The purpose and aims of the Institute are to promote the continued growth andwell-being of pump manufacturers and further the interests of the public in suchmatters as are involved in manufacturing, engineering, distribution, safety, trans-portation and other problems of the industry, and to this end, among other things:

a) To develop and publish standards for pumps;

b) To collect and disseminate information of value to its members and to thepublic;

c) To appear for its members before governmental departments and agenciesand other bodies in regard to matters affecting the industry;

d) To increase the amount and to improve the quality of pump service to the public;

e) To support educational and research activities;

f) To promote the business interests of its members but not to engage in busi-ness of the kind ordinarily carried on for profit or to perform particular servicesfor its members or individual persons as distinguished from activities toimprove the business conditions and lawful interests of all of its members.

Purpose of Standards

1) Hydraulic Institute Standards are adopted in the public interest and aredesigned to help eliminate misunderstandings between the manufacturer,the purchaser and/or the user and to assist the purchaser in selecting andobtaining the proper product for a particular need.

2) Use of Hydraulic Institute Standards is completely voluntary. Existence ofHydraulic Institute Standards does not in any respect preclude a memberfrom manufacturing or selling products not conforming to the Standards.

Definition of a Standard of the Hydraulic Institute

Quoting from Article XV, Standards, of the By-Laws of the Institute, Section B:

“An Institute Standard defines the product, material, process or procedure withreference to one or more of the following: nomenclature, composition, construc-tion, dimensions, tolerances, safety, operating characteristics, performance, qual-ity, rating, testing and service for which designed.”

Comments from users

Comments from users of this Standard will be appreciated, to help the HydraulicInstitute prepare even more useful future editions. Questions arising from the con-tent of this Standard may be directed to the Hydraulic Institute. It will direct allsuch questions to the appropriate technical committee for provision of a suitableanswer.

If a dispute arises regarding contents of an Institute publication or an answer pro-vided by the Institute to a question such as indicated above, the point in questionshall be referred to the Executive Committee of the Hydraulic Institute, which thenshall act as a Board of Appeals.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

viii

Revisions

The Standards of the Hydraulic Institute are subject to constant review, and revi-sions are undertaken whenever it is found necessary because of new develop-ments and progress in the art. If no revisions are made for five years, thestandards are reaffirmed using the ANSI canvass procedure.

Units of Measurement

Metric units of measurement are used; corresponding US units appear in brackets.Charts, graphs and sample calculations are also shown in both metric and US units.

Since values given in metric units are not exact equivalents to values given in USunits, it is important that the selected units of measure to be applied be stated in ref-erence to this standard. If no such statement is provided, metric units shall govern.

Consensus for this standard was achieved by use of the CanvassMethod

The following organizations, recognized as having an interest in the standardiza-tion of centrifugal pumps, were contacted prior to the approval of this revision ofthe standard. Inclusion in this list does not necessarily imply that the organizationconcurred with the submittal of the proposed standard to ANSI.

A.R. Wilfley & Sons, Inc.Afton Pumps, Inc.ANSIMAG IncorporatedBechtel CorporationBlack & Veatch LLPBrown & CaldwellCarver Pump CompanyCascade Pump CoChas. S. Lewis & Company, Inc.Chempump Division, Crane Pumps &

SystemsCheng Fluid Systems, Inc.Cuma S.A.Dean Pump Division, Metpro Corp.DeWante & StowellDow ChemicalEssco PumpsExeter Energy Limited PartnershipFairbanks Morse Pump Corp.Ferris State Univ. Construction and

Facilities Dept.Flowserve CorporationFluid Sealing AssociationFranklin ElectricGrundfos Pumps CorporationIllinois Department of TransportationITT Fluid Handling (B & G)ITT Fluid TechnologyITT Flygt CorporationIwaki Walchem CorporationJ.P. Messina Pump and Hydr. Cons.John Crane, Inc.

Krebs Consulting ServiceKSB, Inc.Lawrence Pumps, Inc.M.W. Kellogg CompanyMalcolm Pirnie, Inc.Marine Machinery AssociationMarshall Eng. Prod. Co. (MEPCO)Moving Water Industries (MWI)Ortev Enterprises Inc.Pacer PumpsPacheco EngineeringPatterson Pump CompanyPinellas County, Gen. Serv. Dept.Price Pump CompanyRaytheon Engineers & ConstructorsRed JacketReddy-Buffaloes Pump, Inc.Scott Process Equipment Corp.Settler Supply CompanySkidmoreSouth Florida Water Mgmt. Dist.Sta-Rite Industries, Inc.Sterling Fluid Systems (Canada) Inc.Stone & Webster Eng. Corp.Summers Engineering, Inc.Systecon, Inc.Taco, Inc.The Process Group, LLCUniversity of MontanaVal-Matic Valve & Manufacturing Corp.Yeomans Chicago CorporationZoeller Engineered Products

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

ix

Although this standard was processed and approved for submittal to ANSI by theCanvass Method, a working committee met many times to facilitate the develop-ment of this standard. At the time it was developed, the committee had the follow-ing members:

Chairman – R. Barry Erickson, ITT Goulds

Other Members

Ed Allis, ITT AC PumpGreg Case, Price PumpBill Beckman, Floway PumpsPat Moyer, ITT, Fluid Handling (B & G)Ray Perriman, Sundstrand Fluid

HandlingDennis Rusnak, Flowserve Corp.Arnold Sdano, Fairbanks Morse

Alternates

Jerry Lorenc, ITT GouldsAllan Budris, ITT AC Pump

Jim Roberts, ITT, Fluid Handling (B & G)

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

This page intentionally blank.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

1

9.6.5 Centrifugal and vertical pumps for condition monitoring

9.6.5.0 Scope

This guideline is for centrifugal and vertical pumps,including both sealed and sealless pump designs asstated in each section. It is intended to be used as atool in implementing process safety management, aswell as general availability improvement programs.

9.6.5.0.1 Use of this document

It is the user’s responsibility to identify the need forimplementing pump condition monitoring practices.The user is also responsible for identifying thoseparameters they wish to monitor. This document doesnot require any monitoring be done, but will provideinformation relevant to making such decisions, andprovides suggestions for carrying out the monitoringprocess.

This guideline discusses the indicators that can bemonitored on centrifugal and vertical pumps to identifypump failure modes. Common means of measuringthose indicators have been defined. Control limits havebeen recommended, where appropriate, for those indi-cators whose limits are not defined in other HydraulicInstitute Standards.

There are a number of potential failure modes for cen-trifugal and vertical pumps. For each failure modethere can be several possible causes. To anticipate theoccurrence of each cause, one or more of the follow-ing 14 indicators may be monitored. The failuremodes, causes, and indicators are listed in AppendixA. The inverse, namely indicators, causes and failuremodes, are listed in Appendix B.

In addition to those indicators listed below, changes inpump sound can sometimes be used to indicate somechanges in pump performance. However, interpreta-tion of change in sound is usually subjective in nature.

Fourteen indicators of the various failure modes havebeen identified:

1) Power

2) Temperature

3) Corrosion

4) Leak detection

5) Pressure

6) Vibration

7) Lubrication analysis

8) Shaft position

9) Rate of flow

10) Maintenance inspection

11) Speed (rpm)

12) Bearing wear detection

13) Preinstallation hydrotest

14) Design review

When monitoring a pump, it is more important toestablish a baseline to which all future measurementscan be compared than to simply determine if the pumpis still operating within established commissioning lev-els. This applies to both new and reconditioned equip-ment. Trending is more important than the absolutelevel of the indicator. Soon after the pump is put intoservice, a baseline should be established. The indica-tors that one chooses to monitor at an established fre-quency can then be compared to the baseline. Thechange and rate of change of the trended indicator willgive the user indications of the pump’s current state,and how much longer it will continue to operate.

9.6.5.0.2 Monitoring frequency

The frequency at which an indicator should be moni-tored is determined by the severity of the conse-quences of failure (severity level), and the probabilityof failure.

The severity level indicates the consequences of fail-ure. There are three factors considered: safety, envi-ronmental and economic.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

2

• Safety consequences include those effects onimmediate employees as well as others in thecommunity.

• Environmental consequences may involve localregulations and national or company standards.

• Economic consequences: Included are the costsof lost production as well as the costs incurred incorrecting the failures.

Table 9.6.5.1 suggests three levels for each factor: low,medium, and high. Users should assign a severitylevel for each factor based on their circumstances. Theseverity level for an application is determined by thehighest individual level.

In determining the probability of a pumping system fail-ure, the following factors should be considered.

• Historical data

• Indicator levels at start up

• Recent trends

• Corrosive/erosive character of liquid

• Equipment redundancy

• Operating proximity to BEP rate of flow

It should be recognized that the probability of failure maybe revised at any time, particularly if trends change.

The user should categorize the probability as low,medium, or high.

By entering the chart in Table 9.6.5.2, a monitoring fre-quency can be determined.

9.6.5.0.3 Control limits

Separate control limits are recommended for alarmand shutdown conditions.

Alarm limit is defined as the indicator value at whichone wants to be notified of changes (either increase ordecrease) in the measured indicator level. The alarmlimit for each indicator should be set as a change fromthe baseline value and not from established commis-sioning acceptance levels. The pump does not need tobe shut down, but more detailed and/or increased fre-quency of monitoring should be instituted.

The shutdown limit is defined as the indicator value atwhich the unit needs to be shut down and securedimmediately. Continued pump operation at indicatorlevels in excess of the shutdown limit will shorten themean time between failures of the unit and greatlyincrease the likelihood of a catastrophic pump failure.

The alarm and shutdown limits suggested in this docu-ment are guidelines. The best knowledge base foreach pump installation is the user’s experience andknowledge. Experience may warrant higher or loweralarm/shutdown limits, depending on the design of theequipment, system requirements, past indicator his-tory, and failure record.

In subsequent sections of this document, each indi-cator is discussed. Each section includes an intro-duction discussing the reasons for monitoring,methods of monitoring, frequency of monitoring, andcontrol limits.

Appendix B contains a cross-reference between indi-cators, causes and failure modes.

Table 9.6.5.1 — Severity levels

Economic Consequences Low Medium High

Safety Consequences Low Medium High

Environmental Consequences Low Medium High

Table 9.6.5.2 — Frequency of monitoringa

a Using plant experience may dictate more or less frequent monitoring than shown in this table.

Probability of Failure:

Low Medium High

Severity Level:

Low Annually Annually Semiannually

Medium Annually Semiannually Monthly

High Semiannually Monthly Monthly

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

3

9.6.5.1 Power monitoring

9.6.5.1.1 Introduction

Monitoring the power consumed by a pump can giveadvanced indication of the following failure modes: roll-ing element bearing failure, coupling failure, shaftbreakage, and hydraulic deterioration.

9.6.5.1.2 Means of power monitoring

There are several ways of monitoring the power usedby a pump. Some of the instrumentation/systems areoutlined below.

Torque shaft: The most direct method is to install atorque shaft with integral speed pickup between thedriver and the pump. This system will directly sensethe speed and torque required by the pump. Sometorque shaft readouts will even calculate the actualpower absorbed.

Power meter: This measurement is useful if thepump is driven by an electrical motor, either directlycoupled to the pump or through a gearbox, belt orhydraulic coupling. Electrical transducers are typi-cally installed in the electrical motor starter to mea-sure voltage, current, and phase angle. Multiplyingthem results in the power supplied to the motor.This approach will not only monitor the powerincrease or decrease in the pump as parts deterio-rate or drag, but also indicates if anything is hap-pening to the general health of the electrical motorand/or the gearbox, belts or hydraulic coupling.

Electrical current: Similar to a power meter, but onlythe motor current is monitored. The line voltage andpower factor are assumed to remain constant,allowing one to calculate the power supplied to themotor. While this method monitors the condition ofthe pump and motor, it is susceptible to errorcaused by variations in the electrical supply grid.

Strain gauges: Strain gauges applied to either thepump shaft near the coupling, or the drive outputshaft, with proper telemetry or slip ring equipment,will give an indication of the torque required to drivethe pump. This approach is similar to using a torqueshaft except that a longer baseplate is not requiredto accommodate the length of the torque shaft.Some accuracy is sacrificed. If the pump speed isconstant or is known, the power required by thepump can be calculated.

Measurement practice: Monitoring the pump powerusage alone will only indicate whether or not thepower is changing. Power changes can also resultfrom changing the hydraulic operating conditions,and mechanical or hydraulic deterioration.

Along with power measurements, the operating condi-tion of the pump needs to be monitored, or as a mini-mum, the pump must be operated at the sameconditions when data is recorded.

To adequately monitor the operating condition ofthe pump; rate of flow, total head, NPSHA andpumpage specific gravity need to be measured.When variable speed devices are used, speedshould also be measured.

Once it has been determined that changes in thepower consumed by a pump are not the result ofchanges in its operating conditions, then the changescan be attributed to a pump failure mode. Furtherstudy of the changes in power and other failure causeswill be required to determine the exact mode of failure.

9.6.5.1.3 Power monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.1.4 Power control limits

Before selecting alarm and shutdown limits, the accu-racy, repeatability, and stability of the power measure-ments must be evaluated. Similarly, the accuracy,stability and repeatability with which the pump’s oper-ating condition can be set also must be evaluated.Consideration should also be given to the power level,with higher power pumps requiring tighter tolerances.

The following control limits are recommended:

Alarm — 5% to 10% change from baseline

Shutdown — 10% to 80% change from baseline

The above levels should not fall below the normal min-imum power required by the pump or above the normalmaximum power levels.

9.6.5.2 Temperature monitoring

9.6.5.2.1 Introduction

Temperature is a relatively simple and inexpensiveparameter to monitor. It can be used to monitor the fol-lowing failure modes: bearings, seal faces, corrosion,

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

4

NPSH variation, cooling-loop blockage, decoupling ofmagnetic couplings and motor-winding insulationbreakdown.

9.6.5.2.2 Means of monitoring temperature

Thermocouple probes or strip resistance temperaturedetectors (RTDs) are primarily used. When monitoringtemperatures at locations where there can be rapidchanges (e.g., vaporization of cooling flow or bearingrubs), special care must be taken. Temperature at thesource of heat input may be up to several hundreddegrees higher than measured temperature of liquid,gas or metal a short distance away. There is also atime delay between temperature rise at the hot spot asheat flows toward the measurement spot. With properattention to locating the sensor near the heat input,and minimizing the thermal inertia of the sensor andheat transfer path, temperature can be used as anindicator of machine condition.

9.6.5.2.3 Specific applications of temperature monitoring

9.6.5.2.3.1 Motor winding temperature

Motor insulation deteriorates faster above its ratedtemperature. Temperature is dependent on load, fre-quency of starts, and cooling effectiveness. Windingtemperature is more often measured by thermocou-ples located at the center of the end turns. Strip RTDsand thermistors are also used as alternate sensormethods. Motors are often fitted with thermostats thatshut off power when a preset temperature is reached.

9.6.5.2.3.2 Temperature sensitive liquids

Temperature sensors are used to measure tempera-ture of liquids to ensure that the material is maintainedin the liquid state. For high freezing-point liquids, caremust be taken to locate the sensor in the least heatedarea and allow time for temperature stabilization. Forpolymerizing liquids, the sensor must be mounted inpotential hot spots (areas of high liquid shear or nearexternal heat inputs).

9.6.5.2.3.3 Rolling element bearing temperatures

Rolling element bearing temperatures can be mea-sured by one of three methods:

• Thermocouples immersed in lube oil active flowareas to measure lubricant temperature.

• Thermocouples touching the outside raceway ofthe bearing.

• Sensors attached to or touching the outside surfaceof the bearing housing. This is the least reliablemethod of measurement.

9.6.5.2.3.4 Liquid film bearing and seal faces temperatures

Temperatures of contacting surfaces of sleeve bear-ings, thrust bearings and mechanical seal faces canchange rapidly when the liquid film is not supportingthe load correctly. For hydrodynamic radial bearings,even thermocouples installed in drilled holes as closeto the radial bearing surface as possible are often tooslow in showing excessive temperature rise to be ableto shut down before bearing failure. Use of proximityprobes to detect radial shaft position and motion ispreferred. For thrust bearings and mechanical sealfaces, thermocouples located very near the contactsurfaces can frequently detect bearing distress beforefailure.

9.6.5.2.3.5 Pumped liquid temperature rise

Measurement of liquid temperature at suction and dis-charge can indicate temperature rise accurately whenrate of flow is sufficient to avoid recirculation and suc-tion heating (above 10% or so of the best efficiencypoint [BEP] flow). At lower rates of flow, mixing fromdischarge to suction can result in relatively low indi-cated temperature rise. In these cases, the suctiontemperature may have increased so that NPSHA hasbeen reduced and flashing may be occurring. Locatingthe suction thermocouple 20 diameters or moreupstream will also avoid suction recirculation heatingerror. Significant increase in temperature rise acrossthe pump indicates a drop in flow that is likely to causeflashing.

9.6.5.2.4 Sealless pump liquid temperature

Avoidance of liquid flashing in the bearing area of seal-less pumps is critical since the bearings are liquidlubricated. Reduced cooling flow can result in anincrease in cooling-liquid temperature and flashing.Slow increases in temperature can be detected bywell-placed thermocouples. Rapid increases may notbe sensed quickly enough to avoid bearing damage.

Liquid is also often used as a coolant in the magneticgap between the inner and the containment shell. Ifflashing occurs, heat removal capacity is greatlyreduced and excessive heating may take place.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

5

9.6.5.2.4.1 Sealless pump temperature damage

Magnets in magnetic couplings lose strength at ele-vated temperatures. If excessive heating occurs,the magnets will weaken until they are no longerstrong enough to transmit the torque. Process liq-uid temperature changes as well as coolant-loopblockages may result in elevated coupling tempera-tures. Temperature sensors may be installed on theoutside of the containment shell or in the contain-ment shell cooling loop to monitor the magnetic cou-pling environment.

The following are some of the situations that may beencountered:

a) Dry run: with no liquid to dissipate heat, tempera-ture will rise rapidly.

b) Running against a closed discharge valve, temper-ature will rise slowly.

c) Decoupling of magnetic drive. In this case, theslippage of the inner magnet ring relative to theouter will generate eddy currents in the inner rotorand resulting heating can damage the magnetsboth thermally and mechanically. Coupling tem-perature will rise quickly.

d) Internal flow holes blocked by solids/polymerizedpumpage. Coolant temperature will rise at a ratedepending on the amount of blockage.

e) Solids between the inner magnet ring and the con-tainment shell. Containment shell temperature willrise slowly.

f) Internal and external rubbing on containment shell.Containment shell temperature will rise rapidly.

g) Excessive temperature can damage the insulationof canned motor wiring.

9.6.5.2.5 Temperature monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.2.6 Temperature control limits

Alarm — ±10% from baseline

Shutdown — ±20% from baseline

These limits should be considered as initial guidelines.Specific process parameters may dictate more or lessrestrictive limits.

9.6.5.3 Corrosion monitoring

9.6.5.3.1 Introduction

Pumps may be monitored for corrosive attack to pre-vent pressure boundary failure and is therefore animportant aspect of maintaining a pump’s reliability.Because of the difficulty in monitoring, potential forcatastrophic failures and the insidious effects of corro-sion, it is imperative that proper materials are selected.Difficult or unusual pumpages may require the re-selection or changes in materials based on the resultsof monitoring corrosion rates. Choosing the right mate-rials for a service should be done by consulting withknowledgeable corrosion engineers, and, when donecorrectly, some of the burden for corrosion monitoringcan be eliminated, or at least reduced.

9.6.5.3.2 Means of corrosion monitoring

9.6.5.3.2.1 Corrosion by visual/dimensional inspection

Visual inspection is the easiest and most economicalmethod of monitoring corrosion and most forms of cor-rosion can be detected by this method. However,stress corrosion cracking usually occurs without anyvisible signs thus resulting in a sudden and sometimescatastrophic failure. Visual inspection of pump inter-nals can reveal the degree of general corrosion occur-ring as well as signs of localized corrosion such aspitting and crevice corrosion. Particular attentionshould be paid to complete inspection of fasteners ascorrosion often takes place in areas hidden from view.Pressure boundary leakage may expose non-wettedfasteners to corrosive pumpage.

Visual inspection can be supplemented with dimen-sional checks of key components, which can then beused to calculate the amount of general corrosion thatthe pump is experiencing.

Since most pump manufacturers provide a corrosionallowance in the design of their equipment, the amountof dimensional change over a given time incrementcan be projected into the remaining life of the pump.

9.6.5.3.2.2 Corrosion by electrical resistance (ER)

The basic principle of this method is the measurementof the increasing electrical resistance of a metal probeas its cross section is reduced by corrosion. The probeshould be the same alloy as the pump. However,probes will typically be wrought alloys, whereas mostpump components will be cast alloys, which can result

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

6

in some small degree of error. ER probes can providea reasonable degree of accuracy for general corrosionbut they are not useful for localized forms of corrosionsuch as pitting.

9.6.5.3.2.3 Corrosion by linear polarization resistance (LPR)

This method involves measurement of a currentresponse to an applied potential through probes thatare inserted in the system. There is commerciallyavailable equipment which can do this automatically. Asmall known pulse of DC voltage is supplied to a testelectrode and the resulting current is measured. Thecurrent generated is proportional to the corrosion rate,which can be determined by electrochemical princi-ples. Although there is a time lag involved with (ER)measurements, the advantage of LPR is that it cangive instantaneous corrosion readings. To use LPR aconductive liquid is required.

9.6.5.3.2.4 Corrosion by ultrasonic thickness measurement (UTM)

Although not as accurate as the other methods, UTMcan also be used to monitor corrosion on a periodicbasis. To use this method, a baseline reading shouldbe obtained at a specific location on the pump casingor cover plate. Then a series of measurements can bemade at this same location over time and the metalloss per unit time calculated. To use UTM, the metalsurface must be bare metal (i.e., no paint). Tempera-ture extremes may influence readings. One effect oftemperature is choice of the couplant (material used toacoustically couple the sensor to the surface). Hightemperature will require a couplant other than water.

9.6.5.3.3 Corrosion monitoring frequency

The ER, LPR and UTM methods lend themselves tocontinuous monitoring or frequent checks by use ofsmall portable data acquisition devices.

The visual/dimensional checks are difficult to do on avery frequent basis because this method requires thatthe pump be shut down and opened in order to gatherthe data. The frequency of inspection by this methodwill then be determined by the importance of thepump. In the absence of experience, the first inspec-tion by this method should be no longer than 3months. The findings of these first inspections canthen determine when subsequent inspections need tobe made. Of course, data generated by this methodshould be obtained anytime a pump is opened formaintenance. Another sign as to when visual/dimen-sional inspection should be done is when a drop-off in

pump performance is noted, i.e., rate of flow and totalhead. Corrosion monitoring is also recommendedshortly after any significant process changes are madesuch as temperature, concentrations or any changesin corrosives being handled (unless experience dic-tates otherwise).

Also see Table 9.6.5.2.

9.6.5.3.4 Corrosion parameter control limits

Pump components are designed with a corrosionallowance. This corrosion allowance varies by compo-nent. Casings and rear cover plates may have up to 3mm (1/8 inch) allowance, while components such assleeves and containment shells may be as low as0.25 mm (0.01 inch). The pump manufacturer shouldbe consulted for specific corrosion allowances.

Alarm — 50% of corrosion allowance

Shutdown — 70% of corrosion allowance

9.6.5.4 Leakage detection

9.6.5.4.1 Introduction

Leakage from installed pumps is detected in a numberof ways depending on the hazard posed by the liquidbeing pumped and the surrounding environment.Leakage detection is monitored to identify the failuremode of the seal or pressure boundry. These leaksmay be in the form of liquid or vapor.

9.6.5.4.2 Means of monitoring leakage

9.6.5.4.2.1 Leakage by visual inspection

For less-hazardous liquids, leakage is often detectedvisually from joints or seal drains. Larger leaks of vola-tile light hydrocarbons such as propane may form icedeposits on the outside surface of the seal gland plate.Continued operation will cause the ice to melt and bereplaced by carbon wear debris from the seal faces.Visual monitoring is commonly used for single and dualoutboard double and tandem seals.

9.6.5.4.2.2 Leakage by sniffer inspection

Sniffers are used to detect minute leakage of volatileorganic compounds (VOCs). Typical locations moni-tored are joints, connections and seal drains. Concen-trations can be measured to determine the severity ofthe leak. The proper sniffer must be used for the com-pound pumped.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

7

All single seal installations handling VOCs must usethis method of monitoring.

9.6.5.4.2.3 Leakage by pressure buildup

Leakage through the inboard seal of a dual tandemseal arrangement may be detected by a change inpressure in the seal reservoir containing the bufferfluid. This is accomplished by blocking off the reservoirfrom the flare (vent) for at least 10 minutes and notingthe increase in pressure.

Pressure buildup in secondary containment areas ofsealless pumps may also be used to indicate leakagepast the primary containment.

9.6.5.4.2.4 Leakage by flow increase

Leakage through the inboard seal of a dual tandemseal arrangement may be detected by monitoring thegas flow from the seal to the flare system.

Leakage through the inboard seal of a dual doubleseal arrangement may be detected by measuring theloss of barrier liquid from the circulation system andreservoir.

The consumption of barrier gas through a dual doublegas-lubricated seal will vary with changes to pressure,temperature and speed.

9.6.5.4.2.5 Double-walled system

For extremely hazardous fluids such as phosgene,double-walled pipe with double sealing flange surfacesand inert purge gas arrangement is used to minimizethe chance of leakage to the atmosphere. Pumps withdouble-walled leakage protection are often used. Sniff-ers are used to detect the presence of any hazardousgas in the purge gas.

For liquids that do not flash when leaking into thelower-pressure collection areas, liquid detectors maybe used to indicate leakage past the primary boundaryinto the collection area.

Table 9.6.5.3 provides an application guideline for theabove methods.

9.6.5.4.3 Leakage monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.4.4 Leakage control limits

Allowable leakage of hazardous materials is oftenestablished by government regulations. For monitoringpurposes, the following limits are recommended:

Warning — 50% increase over baseline, or gov-ernment regulations, whichever is less

Shutdown — 100% increase over baseline, or gov-ernment regulations, whichever is less

9.6.5.4.5 Leakage control limits for gas seal consumption

Warning — three times baseline

Shutdown — five times baseline

9.6.5.5 Pressure monitoring

9.6.5.5.1 Introduction

Pump pressures can be monitored for two separatereasons. The pump or seal static pressure can bemonitored to guard against an over-pressurization of thecasing that may cause the casing joint seal or mechani-cal seal to leak. Pressure may also be monitored as an

Table 9.6.5.3 — Application guidelines for leakage monitoring systems mechanical seals

Sealarrangement

Monitoringmethod

Single

Dual seals

Double Tandem

Inboard Outboard Inboard Outboard

Visual X Xa

a Applies to units sealing a barrier fluid that is a liquid

X

Sniffer` X

Pressure buildup in seal reservoir X

Barrier fluid flow increase Xa b

b Applies to units sealing a barrier fluid that is a gas

X

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

8

indication of the operating point of the pump on theperformance curve.

9.6.5.5.1.1 Means of pressure monitoring

Monitoring for pump performance is accomplishedthrough the use of a pressure gauge, or a pressuretransducer. When monitoring pressure for hydraulicperformance, both the discharge pressure and thesuction pressure must be monitored or a differentialpressure device can be used. Follow HI StandardsANSI/HI 1.6-2000 Centrifugal Pump Tests or 2.6-2000Vertical Pump Tests for tap location and design.Mounting gauges at other locations and/or with othertap designs will provide data that can be trended, butmay have a very poor correlation to the manufacturer’spublished performance curve. This discrepancy isbecause of flow distortion and/or additional pipinglosses at the entrance and exit of the pump. (Occa-sional brief operation at shut off may be done to moni-tor the pump’s shut off head.)

The mounting location of a gauge used to measure thehydrostatic pressure of a pump is less critical andshould be on the discharge side of the pump beforeany valve. Mounting pressure gauges in the seal cavityof a pump should be discussed with the pump manu-facturer to prevent disturbances in the flow field and/orformation of a collection point for debris.

Mounting a pressure gauge to a seal pressure reser-voir should be done only after consulting with the res-ervoir manufacturer. If a pump is being used on theseal flush system to circulate the fluid through the sealcavity, the pressure gauge should be mountedbetween the discharge of the circulation pump and theseal cavity.

9.6.5.5.1.2 Pressure monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.5.1.3 Pressure control limits

Control limits will vary with the type of service that thepump is in, and the shape of the pump head versusrate-of-flow curve. Typical control limits are:

Alarm — –5% from baseline values

Shutdown — –10% from baseline values

Process requirements may dictate more or lessrestrictive limits for alarm and shutdown levels.

If pressure monitoring is being used to monitor theover-pressurization of the casing, seal pot, or seal cav-ity, a shutdown limit can be set at the manufacturer’shydrostatic limits for the unit.

9.6.5.6 Vibration monitoring

9.6.5.6.1 Introduction

Monitoring pump vibration is by far the most widelyused method to determine the condition of pumps.Presently there are many manufacturers of equipmentthat will measure the vibration of rotating equipment.However, since many different failure modes cancause an increase in the pump vibration, it is difficult topinpoint the failure mode by vibration alone. Bearingfailure, seal leakage, coupling failure, shaft breakageand hydraulic degradation are some of the failuremodes that can be detected by vibration monitoring.

9.6.5.6.2 Means of vibration monitoring

Depending on the pump bearings, there are two differ-ent vibration sensors commonly used to measurepump vibrations.

Bearing housing vibrations: Pumps that have rollingelement bearings are commonly monitored usingan accelerometer or velocity transducer. The vibra-tions are generally measured on the bearing hous-ings in the vertical, horizontal and axial positions.For rolling element bearing equipped pumps oper-ating between 500 and 5000 rpm, velocity is thepreferred unit of measure, although displacement issometimes used. If an accelerometer transducer isused, most vibration analyzers can integrate thesignal to velocity.

Shaft vibrations: On pumps (excluding seallesspumps) designed with sleeve bearings, this mea-surement is commonly taken using a proximityprobe mounted on the bearing housing. On seallesspumps where access to the shaft is not readilyavailable, bearing housing acceleration is morecommonly taken. The probe’s output is proportionalto the displacement of the shaft with respect to thebearing housing. With most pumps that operatebetween 500 to 5000 rpm and have sleeve bearings,displacement is the preferred unit of measure. Ifvelocity is required, many analyzers can take thedisplacement signal and differentiate it to obtainvelocity. Two proximity probes are used on eachsleeve bearing and are positioned 90 degrees from

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

9

each other to obtain shaft orbits. Their orientationabout the bearing is usually dependent on the bear-ing housing design. Normally they are eitherlocated in the vertical and horizontal position orequally displaced from the vertical axis (45 degreesto either side).

Vertical pumps: On vertical pumps, where thepumping element is submerged, a proximity probeis sometimes used to monitor shaft displacement. Itshould be located above grade adjacent to the shaftsealing element. Because accessibility is limited inthis area, it is common practice to monitor bearinghousing vibration just above ground level. Experi-ence has shown that both measurements provideuseful information on the pump condition.

Measurement practice: It is important that baselineand subsequent vibration measurements are takenat the same locations, using the same analysis pro-cedures, and with the pump at the same operationalconditions. Indelible ink markers, stickers, or drilleddimples can be used to identify the location of eachvibration measurement point. The vibration analysisprocedure (i.e., bandwidth, number of filters, type offilter, type of average, number of averages, pass fil-ters, etc.) should be standardized. To ensure thatthe pump is operating at the same hydraulic condi-tion, the rate of flow, speed, total head, power,NPSHA, pumpage temperature and specific gravityshould be recorded. If it is not possible to duplicateoperating conditions, measurement results mayneed to be adjusted.

9.6.5.6.3 Vibration monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.6.4 Vibration control limits

Overall unfiltered vibration levels are most indicative ofpump integrity. While acceptance levels at commis-sioning are beyond the scope of this document, the fol-lowing standards provide guidelines: API 610, ANSI/ASME B73.1M and B73.2M, and ANSI/HI 9.6.4-2000,Centrifugal and Vertical Pumps, Vibration Measure-ment and Allowable Values.

The following control limits are recommended:

9.6.5.7 Lubricant analysis

9.6.5.7.1 Introduction

Lubricant analysis is designed to measure threethings:

1) Wear rates of bearings or face seals

2) Lubricant contamination levels

3) Lubricant degradation

It is useful for monitoring the condition of a lubricant asa precursor to bearing or face seal failure.

The following sections are intended to make it easy tounderstand the basics of what the tests measure andhow to use that information.

9.6.5.7.2 Measuring metal particles from wear

Several techniques for measuring metal particles andevaluating wear mechanisms are available in industrytoday.

Generally, laboratories perform a spectrographic anal-ysis consisting of up to 21 different metallic elements.These elements cover most, if not all, inorganic addi-tives, contaminants and wear metals. A spectrographicanalysis indicates the presence of metallic elements,but not the amounts. Quantitative evaluations areobtained from particle counting analyses. A particlecounting analysis will provide information on the num-ber of particles in a sample within various size rangesand an analysis of the contaminants. A description ofthe methods and codes used can be found in ISO4402 and SAE ARP 598. Lubricant analysis laborato-ries are also a good source of information.

Some wear on metal parts, although not desirable, canbe considered to be normal. Large amounts of metalcontaminants usually indicate a serious machine prob-lem. Different machine parts are made from differentmetals, so the presence of particular metals indicates

Bearing HousingVibration

ShaftDisplacement

Alarm —30% above

baseline80% above

baseline

Shutdown —50% above

baseline150% above

baseline

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

10

which components are wearing. Common wear metalsare iron, aluminum, chromium, copper, lead, tin, nickeland silver.

9.6.5.7.2.1 Evaluating wear rates

The accurate diagnosis of wear rates can be accom-plished through the use of an ongoing lubricant analy-sis program where the data has been trended over aperiod of time, accompanied by periodic machineinspection and wear documentation.

When no historical data are available for a machineand/or there are no other like machines in the data-base, an experienced data evaluator may review thelaboratory data and make a value judgment about thewear situation of a specific sample.

When there are multiple machines of the same type,and/or when a machine has been sampled severaltimes, data can be trended. Samples that have metal-lic content that is increasing at an unusually high rateor data that are statistically out of the normal rangeindicates that there is a problem with the machine.

The lubricant analysis should also measure certainmetallic elements that are found as additives in thelubricant. The primary purpose of analyzing for theadditives is to ensure that the appropriate additives arepresent and that there are no other inorganic additivesthat indicate that cross-contamination has occurred.Performing an analysis on the fresh unused lubricantwill show which additives are there. Subsequent oilsamples can be compared to this baseline.

Common additive metals are boron, zinc, phosphorus,calcium, barium, magnesium, molybdenum and sulfur.

9.6.5.7.3 Measuring contamination of lubricant

Lubricating oil must be physically clean and chemicallysuitable to properly lubricate equipment. Contamina-tion is commonly classified as organic or inorganic.

9.6.5.7.3.1 Organic contamination of lubricant

Organic contamination can be from the by-products oflubricant degradation or from external sources. Mostoften it is from lubricant degradation. If the organiccontamination is from an external source, it is usuallyinherent to the operation and is readily identifiablesuch as chlorinated hydrocarbons, process liquids,acids, diesel fuel, glycol or freon. Organic contamina-tion can increase or decrease the viscosity of the lubri-cant. The analyses most often performed to measure

organic contamination are viscosity, acid number,infrared analysis and flash point.

9.6.5.7.3.2 Inorganic contamination of lubricant

Contaminants in the oil such as dirt, dust, weldingslag, etc., can be carried throughout the machine andcause severe abrasive wear.

Dirt and abrasives — Dirt and abrasives generallyenter the machine through poor housekeeping prac-tices or systems that are not properly sealed fromthe environment. Another source of the element sili-con can be found in sealant and anti-foam additives.

Fibers and debris — Debris in general is consideredto be an inorganic contaminate. Debris can befound in the form of fibrous materials such as ragsor filter media breakdown.

There are a variety of laboratory techniques to mea-sure “solid, inorganic” contamination. The mostcommon method for measuring the concentration ofthese contaminates without regard for the actualmake-up of the debris is an automatic particlecounting instrument.

Water — Water is one of the most common contam-inates found and in large quantities is the easiest todetect. Water can come from internal sources suchas leaks in cooling systems and condensation orexternal sources such as leaking seals. Usually thelubricant turns white in color and emulsifies.

9.6.5.7.4 Measuring lubricant degradation

Degradation characteristic of lubricating oils can bemeasured through viscosity, acidity or antioxidantlevels.

Measurable changes in viscosity or acidity are gener-ally considered to be condemning limits for a lubricant,while the antioxidant level acts as an early-warningdevice for lubricant changes. Antioxidants are addedto lubricants to protect the base oil from oxidizing.When these additives are “used up,” the base oil is leftunprotected and degrades at an accelerated rateallowing the formation of oxidation by-products harmfulto the machine. As the antioxidant nears total deple-tion, the viscosity and acid number dramaticallyincrease.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

11

9.6.5.7.5 Lubricant sampling techniques

Acquiring a representative lubricant sample isextremely important to the overall success of a lubri-cant analysis program. Sampling points should be onthe discharge side of the operating unit before the filter.

In noncirculating sumps, samples should be takenwhile the pump is operating (or immediately after shut-down) to ensure well-mixed lubricant is obtained.Samples should be withdrawn from within 12 mm (0.5inch) of the surface of an oil sump. A sample from thebottom of the sump will indicate the presence of freewater. There are a variety of sampling techniques thatcan be used. Contact the laboratory for more informa-tion on special situations.

9.6.5.7.6 Lubricant monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.7.7 Lubricant control limits

Control limits will vary with the type of lubricant andparameter being monitored. Refer to the lubricant sup-plier, or bearing or seal manufacturer for specific rec-ommendations.

9.6.5.8 Shaft position monitoring

9.6.5.8.1 Introduction

A mechanical seal is a precision device that must beinstalled in mechanically sound equipment. Oneaspect of the pump being mechanically sound is thatthere not be excessive shaft runout or shaft deflection.Guidelines for shaft runout at the face of the sealhousing and dynamic shaft deflection can be found instandards such as API 610 and ANSI/ASME B73.1.The purpose of these guidelines is to provide an envi-ronment that will maximize the life of the mechanicalseal, thus preventing seal leakage.

Excessive shaft deflection or shaft runout can causethe rotating face of the mechanical seal to wobbleagainst the stationary face. This may cause prematureseal face wear, seal spring failure or bellows fatigue, allresulting in seal leakage.

9.6.5.8.2 How to monitor shaft position

A dial indicator can be used to measure shaft runoutbetween the shaft and the seal chamber. Both theradial and axial runout should be measured.

Dynamic shaft motion/position can be measured usinga pair of proximity probes mounted 90° apart and asclose to the mechanical seal as possible. Proximityprobes are not normally used to monitor shaft runout onpumps whose shafts are supported by rolling elementbearings, but are used on shafts with sleeve bearings.

9.6.5.8.3 Shaft monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.8.4 Shaft control limits

Alarm — 50% above baseline

Shutdown — 100% above baseline

9.6.5.9 Rate of flow monitoring

9.6.5.9.1 Introduction

In some field installations, it may be difficult to accu-rately measure rate of flow. However, problems suchas plugging of flow passages, air binding, excessiveflow causing insufficient NPSHA, and increased clear-ances can be detected by flow monitoring using lessaccurate devices of sufficient sensitivity.

9.6.5.9.2 Measuring rate of flow

Monitoring of rate of flow may be accomplished byfixed in-line devices such as rotameters, turbine flowmeters, orifices, venturi meters, magnetic flow metersor nutating disc meters. Non-invasive devices such asultrasonic meters may also be used.

Those devices that cause losses or flow distortionshould be placed on the discharge side of the pump.Manufacturer’s directions should be followed to obtainthe specified accuracy of the flow measuring device.Particular attention should be given to the piping con-figuration immediately before and after the flow meter.

9.6.5.9.3 Rate of flow monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.9.4 Rate of flow control limits

Control limits will vary with the type of service thepump is in and the shape of the head versus rate-of-flow curve. Typical control limits are:

Alarm — ±10% from baseline values

Shutdown — ±20% from baseline values

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

12

9.6.5.10 Maintenance inspection

9.6.5.10.1 Introduction

Many potential causes of failure are best monitored byperiodic inspections. Examples of such failure causesare

• Coupling failure

• Shaft breakage

• Erosion

• Hydraulic performance degradation

Inspections performed to detect these impending fail-ures typically require disassembly. This section of thepractice only addresses those inspections which relateto the above failure causes. It should be recognizedthat additional inspections are often recommended bythe equipment manufacturer, and can be found in themanufacturer’s maintenance manual. Additionally,other recommended equipment checks can be foundin the reference material, ANSI/HI 1.4-2000 Centrifu-gal Pumps for Installation, Operation and Maintenanceand ANSI/HI 2.4-2000 Vertical Pumps for Installation,Operation and Maintenance.

9.6.5.10.2 Maintenance inspection practice

The failure modes and causes analysis indicates thefollowing characteristics should be considered:

Coupling degradation

Shaft breakageBending fatigueTorsional fatigueTorsional overload

Erosion

Hydraulic — Loss of head or rate of flow

Inspection practices may be different for each piece ofequipment and are determined by the specific applica-tion. An analysis of the application, equipment, andenvironment will determine the items that should beinspected. A check list for each application should bedeveloped.

Following is a discussion of the above items, which willbe helpful in developing a check list.

9.6.5.10.2.1 Maintenance inspection of keys and keyways

Inspections can provide evidence of key/keyway dam-age. Keys should be inspected for evidence of offset(potential shearing), rounding of edges, and reductionin width or height.

There should be no evidence of any of the above. Ifevidence of overload is uncovered, replace the dam-aged key with one that meets the original equipmentmanufacturer’s material specifications.

Keyways should be inspected for evidence ofdeformed groove sides and rounding of the corners.No damage is permitted. The manufacturer should beconsulted for acceptable maximum groove width orminimum key width.

9.6.5.10.2.2 Maintenance inspection of coupling flexible elements

Elastomeric elements in couplings should beinspected for evidence of loss of bond between theelastomer and any metallic component (if applicable),wear between mating parts, and permanent distortion.Evidence of a loss of bond is cause to remove the partfrom service. If wear or permanent distortion is found,the coupling manufacturer should be consulted foracceptable limits. Evidence of either is cause forreview of frequency of inspection.

Metallic elements should be inspected for evidence ofcorrosion, wear between mating surfaces, and cracks.Corrosion and wear can be determined by visualinspection or mechanical measurements. Refer to thecoupling manufacturer for recommended measure-ments and acceptable limits. Evidence of either iscause for review of frequency of inspection.

Evidence of cracks are commonly checked using dyepenetrant or magnetic particle inspection. Indication ofa crack is cause for removal of the part from service.

9.6.5.10.2.3 Maintenance inspection for shaft bending fatigue

Material failure due to bending fatigue can occur atany point on the shaft, or at any fastener. A bendingfatigue failure will initiate in the area of highest bendingstress. This usually occurs at the root of a change insection thickness, and in the area of highest bending

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

13

loads. More common locations of failure are outboardof either bearing on an overhung impeller pump, nearthe impeller of a between bearing pump, and outboardof the inboard bearing of a between bearing pump.

Attention should also be given to areas of fretting orother damage that may cause stress risers and there-fore initiation of fatigue sites.

If the impeller is secured with a thread, those threadsshould also be inspected.

Dye penetrant or magnetic particle inspection arecommon inspection methods. Any evidence of a crack iscause to remove the component from service.

9.6.5.10.2.4 Maintenance inspection for shaft torsional fatigue

A torsional fatigue failure of shafts occurs at a torsionalstress riser location. The common location is at theroot of a keyway. Dye penetrant or magnetic particleinspection are common inspection methods. Any evi-dence of a crack is cause to remove the componentfrom service.

9.6.5.10.2.5 Maintenance inspection for torsional overload

A momentary torsional overload can produce perma-nent torsional deformation of a shaft. This is normallyevident in the coupling area of the shaft. If the shafthas a keyway, it may be twisted. A friction drive cou-pling may show signs of scoring between the couplingand the shaft. Any evidence of overload is cause toremove the component from service.

9.6.5.10.2.6 Maintenance inspection for erosion

Damage from erosion occurs in regions of high veloc-ity or impingement. Components typically damaged byerosion are casing sideplates or covers, casing dis-charge lips, casing volutes, impeller vanes, impellershrouds, shafts and shaft keyways. Damage is oftenfound to be concentrated in areas where flow distur-bances occur (i.e., hand hole cover, tap connections).Operation at off-design condition can cause flow recir-culation in the impeller resulting in localized damage atthe inlet of the impeller vane. Very localized erosionmay also occur in circumferential pockets such asbetween a casing cover and the casing. Inspectionfor a known eroded area may be done by mechani-cal or ultrasonic measurement of the wall thickness.A dimensional change beyond the manufacturer’s

recommended wear limits is cause for removal of thecomponent from service.

Because erosive damage may be localized, and mayoccur in unstressed areas, it is recommended that amechanical inspection be conducted in addition to ahydrotest if the service is suspected, or known, to beerosive.

Because of the localized nature of erosive damage,other means of measurement of a closed pump, suchas electrical resistance, linear polarization resistanceor ultrasonic measurement are not effective in monitor-ing erosion.

9.6.5.10.2.7 Maintenance inspection for hydraulic performance

Pump head or rate of flow may be reduced due toincreased internal clearances or dimension changes.Increased clearances at wearing rings, or betweenopen impellers and mating surfaces, permit increasedinternal leakage and results in reduced head or rate offlow.

Clearance measurements and impeller diameter mea-surements are made mechanically. Wearing ring clear-ance measurements should be compared to themanufacturer’s recommendation. If the clearances areexcessive, the parts should be replaced.

In the case of open impellers, the manufacturer nor-mally provides a means of resetting the clearance. Ifthe wear is non-uniform, as determined using manu-facturer’s recommended methods, it may be impossi-ble to obtain a uniform clearance and the worn partsmust be retired from service.

9.6.5.10.3 Maintenance inspection frequency

The frequency of on-going maintenance inspectionsdepends on the application’s historical records, as wellas availability of the pump for inspection. Refer toTable 9.6.5.2 for recommended inspection intervals.Because service conditions may prevent removal atfrequent intervals, when inspections are done, theyshould be thorough, with results documented andtracked to predict future equipment retirement.

9.6.5.11 Speed (rpm) monitoring

9.6.5.11.1 Introduction

Pump speed is monitored to check for speed changesthat may cause loss of head or rate of flow, or to avoid

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

14

operation near a critical speed. Also, excessive speedwill cause the pump to draw more power and maycause the motor to be overloaded. There are two typesof systems which drive pumps: constant speed andvariable speed.

Constant speed systems: It is unlikely that the motorspeed will change significantly unless a major electri-cal problem has occurred. While the load of a centrifu-gal pump varies with rate of flow, the changes in speedassociated with the load changes are normally rela-tively slight (less than 2% of full load speed for NEMADesign B motors). However, changes due to high orlow voltage, or loss of power to one phase on a threephase motor may be significant.

Variable speed systems: These systems rely on speedchange to control head and rate of flow. These sys-tems can have the same problems as the constantspeed systems, but also can have unintended speedchanges. These changes can be due to a faulty driveor speed control problem. It may be necessary, whenpossible, to block out certain speed ranges to preventoperation at or near rotor or structural critical speeds.

9.6.5.11.2 Speed measurement methods

Common methods of measuring speed are strobelight, revolution counter, tachometer or electroniccounter. Any of these devices should be able to mea-sure within ±0.1%.

9.6.5.11.3 Speed monitoring frequency

Refer to Table 9.6.5.2.

9.6.5.11.4 Speed control limits

Alarm — ±2% from baseline (constant speed)

Shutdown — ±5% from baseline (variable speed)

The recommended control limits for variable speedsystems should only be considered initial guidelines.Process requirements may dictate more or lessrestrictive limits.

9.6.5.12 Bearing wear monitoring of plain bearings in sealless pumps

9.6.5.12.1 Introduction

Sealless centrifugal pumps, both magnetic driveand canned motor, use a driven internal rotatingmechanism (rotor), that may be immersed in the

pumped liquid, to rotate the pump impeller. The rotorand liquid are separated from the environment by asurrounding containment device (“shell” or “can”). Thesleeve bearings used to support the rotor are cooledand lubricated by the liquid within the containmentdevice. Excessive wear or failure of these bearingsdue to insufficient lubrication, abrasives in the lubri-cant or cavitation, could cause contact between therotor and the containment device. This contact couldresult in failure of the equipment and leakage of theprocess liquid. Providing some means to monitor thecondition of these bearings is therefore desirable.

9.6.5.12.2 Means of detecting bearing wear

9.6.5.12.2.1 Bearing materials

Typical materials used as sleeve bearings in seallesspumps are silicon carbide and carbon. Each of thesematerials exhibits different wear characteristics thatcan influence the monitoring frequency and detectionmethod.

Silicon carbideSilicon carbide bearings are more subject to suddenfailure, such as breakage due to lack of lubrication (dryrunning) and overloading, than to exhibit significantwear (“wear out”) under normal operating conditions. Aprogressive wear monitoring device may not detectsignificant wear prior to sudden failure.

CarbonCarbon bearings can exhibit wear under normal oper-ating conditions. A periodic monitoring plan using aprogressive wear indicator can be useful in detectingnormal wear prior to “wear out” or component contact.However, periodic monitoring will not detect bearingfailure or component contact between monitoringperiods.

9.6.5.12.2.2 Bearing wear detection methods

Detection of sleeve bearing wear in sealless pumpscan be accomplished by visual inspection and dimen-sional verification during pump disassembly periods orwhile the pump is in operation using instrumentationmeans.

Instrumentation means used to detect bearing wearfalls into two categories, progressive wear monitoringand detection of component contact.

Progressive monitoringProximity sensing devices can be used to monitor theposition of the rotor within the containment device.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

15

Positional changes of the rotor are then used to deter-mine the direction and amount of bearing wear. Thismethod will permit excessive wear to be detected priorto contact between the rotor and the containmentdevice or other part of the assembly (bearing holder)designed to prevent or limit rotor contact with the con-tainment device. The proximity sensing device(s)should be selected to detect both radial and axial posi-tional changes of the rotor. Rotor accessibility for prox-imity sensing, particularly in mag-drive pumps, resultin these types of devices normally being incorporatedinto the pump by the manufacturer. Canned motordesign pumps may provide accessibility to the rotatingshaft for location of non-contacting type proximity sen-sors. Progressive wear monitoring is not normallyapplied to bearings constructed from “non-wearing”materials.

Contact detectionExcessive bearing wear will allow positional changesof the rotor to cause contact (impacts or rubbing)between the rotor and the containment device or otherpart of the assembly (bearing holder) designed to pre-vent or limit rotor contact with the containment device.

Contact can be detected using a suitable acousticdetection device, power monitor, vibration sensor(accelerometer) conditioned to detect impacts andrubbing, containment shell temperature probe, conti-nuity probe or contact switch.

Application considerations for these devices are listedbelow.

Acoustic (sound) detection. An acoustic sensor canbe used to detect sound produced by contact betweenthe inner or outer magnet and the containment shell orother motion-limiting assembly. The initial contact,however, may be infrequent and the signal difficult todistinguish from other sounds, such as that producedby cavitation.

Power monitor. Rubbing of internal or outer compo-nents can cause the power required by the pump toincrease. The increase in required power can bedetected by a power monitor. However, the increase inpower prior to a significant binding can be too small todistinguish from normal operational power changes.

Vibration sensor. Rubbing of internal components canalso result in increased vibration levels detectable withvibration monitoring equipment. Initial intermittentcontact may produce impact or ringing-type vibrationsignals more easily detected with higher-frequencyvibration detection techniques.

Temperature probe. Rubbing contact between therotor and containment shell can result in an increase intemperature at the rub point. The increase in tempera-ture can be localized and minimal during the initial orinfrequent contact, making it difficult to detect.

Contact or continuity switch. In some sealless pumpdesigns, particularly canned motor pumps, directaccess to the rotating shaft is available. In thesecases, excessive wear can be detected by direct con-tact with an electrical switch or electrical continuitydetection device. The continuity detection device maybe activated by causing or interrupting a completedelectrical circuit.

9.6.5.12.3 Bearing wear monitoring frequency

Refer to Table 9.6.5.2. If a contact detection monitor-ing device is used, continuous monitoring must beperformed.

9.6.5.12.4 Bearing wear control limits

Wear limits for sleeve bearings in sealless pumps aredetermined by the pump design and internal clear-ances. The pump manufacturer or its maintainancemanuals should be consulted for allowable wear limits.

When contact type wear monitoring devices areused, indication of contact requires immediate shut-down.

9.6.5.13 Pre-installation hydrotest

9.6.5.13.1 Introduction

A pre-installation hydrotest (hydrostatic pressure test)is performed to check for leakage. This leakage maybe the result of a flaw in, or damage to, a pressure-containing housing; a damaged or improperly installedmechanical seal; o-ring; or gasket.

9.6.5.13.2 Means of conducting a pre-installation hydrotest

In order to conduct the hydrotest, it will be necessaryto block off the wet end of the pump, fill it with water orother liquid, and pressurize the enclosed wet endchamber. WARNING! It is very important, from asafety point of view, to use a liquid to perform thishydrotest, and not gas. When preparing for thehydrotest, a connection is needed to fill the chamberwith liquid, as well as a connection to vent air fromthe chamber until the chamber is totally filled with liq-uid. WARNING! All air must be vented from the

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

16

chamber. This can normally be done through taps inthe suction and discharge flange covers for end suc-tion and in-line pumps.

Hydrotest pressure should equal the pump maximumworking pressure, unless limited by other factors (e.g.,allowable system pressure and mechanical seal allow-able pressure). The pressure should be held for a min-imum of 10 minutes without any visible signs ofleakage.

For axially split case pumps and some other doublesuction pumps, the vent is normally in the top of thevolute, but it may also be necessary to vent the tops ofthe suction chambers in order to remove all the air.Note that many axially split case pumps have differentallowable hydrotest pressures for the suction chamberand the discharge chamber. In such cases, DO NOThydrotest above the lower hydrotest pressure.

For vertical double casing can type pumps, it is neces-sary to vent the suction side. Usually there is a ventplug at the top of the can. The discharge side mustalso be vented. Usually the shaft seal or packing box isthe highest elevation to allow release of entrapped air.Care should be taken not to over-pressurize the suc-tion can as it may be rated only for suction pressure.

Care must be taken when mechanical seals are usedto be sure the seal is not over-pressurized. With dou-ble mechanical seals, it may be necessary to alsopressurize the seal chamber between the two seals.To ensure proper seal installation it may be desirableto hydrotest the seal chamber between the double sealbefore and during hydrotesting the pump chamber.

9.6.5.13.3 Hydrotest monitoring frequency

A pre-installation hydrotest should be performed priorto installing the pump in the system, or after any disas-sembly. If the pump does not meet the hydrotestrequirements, corrective action must be taken. Thismay include as little as correctly tightening bolts, to asmuch as replacing a major cast component. After cor-rective measures have been completed, the hydrotestshould be repeated until the requirements are suc-cessfully met.

Subsequent tests can be performed either in the sys-tem or out of the system. When the testing is donewith the pump in the system, determine what compo-nent of the system, when pressurized, has the lowesthydrotest pressure capability, and do not exceed thatpressure during testing.

9.6.5.13.4 Hydrotest control limits

No leaks are permitted while the pressure is maintained.Pressure should be held for 10 minutes minimum.

9.6.5.14 Design review

9.6.5.14.1 Introduction

Pump applications in critical or hazardous serviceshould be formally reviewed prior to initial startup, afterany major revamp, or process change. A review shouldcover mechanical application, hydraulic application,installation and operating procedures.

The review should address the above factors asrelated to safety as a minimum. It is strongly sug-gested that the review consider reliability and function-ality issues as well.

9.6.5.14.2 Design review practice

Following is a guideline list of review items. This list isnot necessarily complete for all applications. Users arecautioned to review their application and modify the listfor their situation.

9.6.5.14.2.1 Mechanical application review

• Is wetted component metallurgy correct?

• Is the coupling selected properly?

• Is the mechanical seal of correct design?

• Is the seal properly flushed and vented?

• Is the pump design appropriate for the service?

• Do all shaft guards comply with applicable stan-dards?

9.6.5.14.2.2 Hydraulic application review

• Is the impeller sized properly?

• Are there proper lengths of straight pipe at pumpsuction?

• Are temporary suction strainers removed or pro-vided with a bypass?

• Are permanent suction strainers installed for theproper direction of flow?

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

17

• Are suction piping fittings designed to vent gases?

• Does sump design comply with ANSI/HI 9.8-1998?

• Are controls in place to prevent operation belowthe pump’s minimum allowable rate of flow?

• Are controls in place to prevent operation abovethe pump’s maximum allowable rate of flow?

• Is the motor properly sized for maximum power(over the full range of flows and specific gravity)?

• Is adequate NPSH margin provided under alloperating conditions?

9.6.5.14.2.3 Installation review

• Is piping designed and installed to meet nozzleloads requirements?

• Is piping supported?

• Is piping aligned?

• Are flange bolts properly tightened?

• Are pump and driver aligned?

• Is the baseplate grouted (if applicable)?

• Is the baseplate fully supported and leveled (if freestanding)?

• Are flush and drain lines connected?

• Are all guards in place and secure?

• Are all open ports plugged?

• Is lubricant provided to bearings?

• Are check valves dampened to prevent slam?

9.6.5.14.2.4 Operating procedures review

• Do startup procedures eliminate operation belowpump’s minimum allowable flow?

• Do startup procedures eliminate operation beyondpump’s maximum allowable rate of flow?

• Do startup procedures ensure liquid is available tothe pump?

• Do shutdown procedures eliminate operationbelow pump’s minimum allowable rate of flow?

• Do shutdown procedures eliminate operationbeyond pump’s maximum allowable rate of flow?

• Do operating procedures ensure adequate NPSHis available to the pump under all operating condi-tions?

9.6.5.14.3 Design review frequency

A design review should be conducted prior to initialcommissioning, after any revamp, or process change.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

18

Appendix A

Condition Monitoring,Failure Modes

Pumps with mechanical seals

Failure Mode Causes Indicators

Bearing Failure (rolling element)

Pitting and spalling of races Overload Bearing temperature

Insufficient lubrication Vibration (high frequency)

Cage breakage Insufficient lubrication Vibration (high frequency)

Cage wear Insufficient lubrication Vibration (high frequency)

Contaminated oil VibrationOil analysis

Seizure Overload Bearing temperature

Insufficient lubrication Vibration (high frequency)

Inner/outer race rotation Incorrect bearing fit Vibration

Seal Leakage

Face checking Overload Face temperatureSeal pot pressure/level

Lack of lubricant Face temperatureSeal pot pressure/level

Face blistering Lack of lubricant Face temperatureSeal pot pressure/level

Wrong lubricant Face temperature

Face wear Shaft movement (runout) VibrationShaft position

Contaminated lubricant Lubricant analysis

Spring failure Excessive deflection VibrationShaft position

Improper installation SniffingLeakage observation

Bellows failure Excessive deflection VibrationShaft position

Improper installation SniffingLeaking observation

Corrosion of springs, Bellows Wrong material SniffingLeaking observation

Excessive temperature Seal chamber temperature

Static seal hang up Solids in pumpage SniffingLeakage observation

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

19

Static seal corrosion Wrong material SniffingLeakage observation

Excessive temperature Seal chamber temperature

Static seal breach Cut “O” ring Pre-installation hydrotest

Improper surface finish Inspection

Coupling Failure

Hub loose (set screws) Improper installation Vibration

Worn shaft/hub Vibration

Key shearing Overloaded key Inspection

Wrong material Inspection

Worn key/keyway Inspection

Flexible element failure Excessive misalignment Vibration

Overload Power measurement

Corrosion Inspection

Runout Vibration

Incorrect selection Inspection

Shaft Breakage

Bending fatigue Excessive load Power measurement

Incorrect manufacture Vibration

Corrosion InspectionVibration

Torsional fatigue Natural frequency Vibration

Vane pass excitation Vibration

Corrosion InspectionVibration

Torsional overload Excessive load Power measurement

Material defect Inspection

Corrosion

Casing, impeller Incorrect material Corrosion monitor

Excessive temperature Corrosion monitorTemperature measurement

Fluid variation Corrosion monitor

Cover, seal components Incorrect material Corrosion monitor

Excessive temperature Corrosion monitorTemperature measurement

Fluid variation Corrosion monitor

Electrolytic cell Corrosion monitor

Erosion

Casing, cover, bowl wear rings Incorrect material Inspection

Fluid-borne abrasives Inspection

Pumps with mechanical seals (continued)

Failure Mode Causes Indicators

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

20

Impeller Incorrect material Inspection

Fluid-borne abrasives Inspection

Low flow operation InspectionPower measurement

Hydraulic

Plugging Insufficient rate of flow Rate of flow measurementPower measurement

Excessive solids Rate of flow measurement

Incorrect pump Design review

Air binding Excessive air Power measurement

Low flow operation Power measurement

Suction piping Design review

Insufficient NPSHA Excessive rate of flow Power measurementVibration

Suction piping Design review

Excessive fluidtemperature

VibrationTemperature measurement

Manufacturing error Vibration

Low flow operation Power measurement

Loss of head/flow Increased clearances Rate of flow measurementInspection

High power Changes in system Pressure measurementFlow measurement

Change in speed Speed measurementPower measurement

Impeller diameterchange

Power measurementInspection

Excessive rate of flow Rate of flow measurement

Binding Power measurement

Rubbing Power measurement

Pressure Boundry Leakage

Gasket Cut gasket Pre-installation hydrotest

“O”ring Incorrect surface finish Pre-installation hydrotest

Insufficient gasket load Leak detector

Gasket corrosion Leak detector

Cut “O” ring Pre-installation hydrotest

Casing cover Incorrect surface finish Pre-installation hydrotest

“O” ring corrosion Leak detector

Corrosion Leak detector

Pumps with mechanical seals (continued)

Failure Mode Causes Indicators

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

21

Casing cover (continued) Casting defects Pre-installation hydrotest

Fabrication defects Pre-installation hydrotest

Excessive pressure Pressure measurement

Pumps with mechanical seals (continued)

Failure Mode Causes Indicators

Sealless Pumps

Failure Mode Causes Indicators

Bearing Failure Journal Type

Worn-out bearing Insufficient lubrication Containment shell temperatureContainment shell liquid temp.Bearing wear detectorPower monitor

Cavitation VibrationBearing wear detectorPower monitor

Abrasives in lubricant Bearing wear detector

Broken bearing Insufficient lubrication Containment shell temperaturePumped liquid temperatureBearing wear detectorPower monitor

Overload Power monitor

Cavitation VibrationBearing wear detectorPower monitor

Seized bearing Overload Power monitor

Trapped solids Containment shell liquid temp.Containment shell temperature

Failed antirotation device VibrationPower monitor

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

22

Appendix B

Condition Monitoring Indicators

Indicators Cause Failure Mode

Power

High power Overload CouplingShaft breakage

Impeller rubbing Low flow or head

Impeller binding Pump seized

Oversize impeller Excessive flow or head

Increased speed Excessive flow or head

Excessive flow Impeller cavitation

Low power Low flow operation Impeller erosionShaft breakageImpeller cavitation

Excessive air/gas Low flow or headZero flow

Low power Reduced impeller diameter Low flow or head

Insufficient NPSHA Low flow or headImpeller cavitation

Reduced speed Low flow or head

Impeller clogged Low flow or head

Vibration

High-frequency vibrationrolling element bearings

Insufficient lubrication Pitting/spalling of bearing racesCage breakageCage wearSeizure

High vibration level Contaminated oil Bearing cage wear

Incorrect bearing fit Bearing race rotation

Shaft movement (runout) Mech. seal face wearFlexible coupling failure

Excessive shaft deflection Mech. seal spring failureMech. seal bellows failure

Improper installation of coupling Flexible coupling failure

Loose coupling hub Flexible coupling failure

Shaft runout Flexible coupling failure

Incorrect shaft manufacturing Shaft bending fatigue

Corrosion Shaft bending fatigueShaft torsional fatigue

Resonant frequency Shaft torsional fatigue

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

23

High vibration level (continued) Vane pass excitation Shaft torsional fatigue

Excessive flow Impeller cavitation

Excessive fluid temp Impeller cavitation

Impeller manufacturing error Impeller cavitation

Temperature

Excessive process fluidtemperature

Process upset Casing/impeller corrosionSeal component corrosion

Excessive seal chambertemperature

Loss of cooling Corrosion of seal springs/bellowsStationary seat corrosion

Seal face temperature Overload Seal face blistering

Lack of lubricant Seal face checking

Wrong lubricant Seal face blistering

Bearing temperature Overload Pitting/spalling of bearing racesBearing seizure

Corrosion

Corrosion monitor Incorrect materials of construction Corrosion of casing, impellerand/or seal

Excessive temperature Corrosion of casing, impellerand/or seal

Fluid variation Corrosion of casing, impellerand/or seal

Electrolytic cell Corrosion of casing, impellerand/or seal

Leak Detection

Leak observation at seal Improper installation Seal spring or bellows failure

Wrong material Corrosion of springs and bellowsStatic seal corrosion

Leak observation gasket joint Insufficient gasket load Gasket failure

Gasket corrosion Gasket failure

“O” ring corrosion “O” ring failure

Pressure

Casing liquid pressure Change in system Loss of head/flow

Excessive pressure Casing/cover leakage

Seal pot pressure/level Overload Seal face checking

Lack of lubricant Seal face checkingSeal face blistering

Lubricant

Lubricant analysis Contaminated lubricant Seal face wearBearing failure

Indicators Cause Failure Mode

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal and Vertical Pumps for Condition Monitoring — 2000

24

Shaft position Shaft runout Seal face wear

Excessive deflection Seal spring/bellows failure

Flow Measurement

Process flow Insufficient flow Plugged impeller

Excessive solids Plugged impeller

Increased clearances Loss in head/flow

Changes in system Loss in head/flow

Excessive flow High power

Maintenance Inspection

Inspection of coupling Overloaded key Coupling key sheared

Wrong key material Coupling key sheared

Worn key/keyway Coupling key sheared

Corrosion of coupling Coupling flex element failure

Incorrect coupling selection Coupling flex element failure

Inspection of shaft Corrosion of shaft Shaft binding fatigue

Shaft material defect Torsional overload

Inspection of liquid end parts Incorrect material Casing, cover, impeller and ringwear

Fluid-borne abrasives Casing, cover, impeller and ringwear

Low flow operation Impeller wear

Increased clearances Low head/flow

Impeller diameter change Low head/flow

Speed (RPM)

Speed measurement Low speed Low head/flow

Bearing Wear Detector

Bearing wear detector Insufficient lubricant Worn bearing

Cavitation Worn bearing

Abrasives in lubricant Worn bearing

Pre-installation Hydrotest

Assembled liquid end Cut gasket Gasket leak

Incorrect surface finish Gasket/“O”ring leak

Cut “O” ring “O” ring leak

Casing/cover defect Casing/cover leak

Design Review

Design review Incorrect pump Plugged impeller

Suction piping Air-bound impellerImpeller cavitation

Indicators Cause Failure Mode

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal Pump Design and Application Index — 2000

25

Appendix C

Index

This appendix is not part of this standard, but is presented to help the user in considering factors beyond thisstandard.

Note: an f. indicates a figure, a t. indicates a table.

Alarm limit (defined), 2

Baseline, 1Bearing failure mode causes and indicators, 18, 21t.Bearing wear monitoring, 14

acoustic detection, 15bearing materials and characteristics, 14carbon bearing wear characteristics, 14contact detection, 15contact or continuity switch, 15control limits, 15frequency, 15indicators, 24means, 14power monitor, 15silicon carbide bearing wear characteristics, 14temperature probe, 15vibration sensor, 15wear detection methods, 14

Centrifugal and vertical pumpssealed, 1sealless, 1

Control limits, 2Corrosion failure mode causes and indicators, 19t.Corrosion monitoring, 5

control limits, 6by electrical resistance, 5frequency, 6indicators, 23by linear polarization resistance, 6means, 5by ultrasonic thickness measurement, 6by visual/dimensional inspection, 5

Coupling failure mode causes and indicators, 19t.

Design review, 16frequency, 17hydraulic application review, 16indicators, 24installation review, 17mechanical application review, 16

operating procedures review, 17procedure, 16

Economic consequences of failure, 2Environmental consequences of failure, 2Erosion failure mode causes and indicators, 19t.

Failure mode causes and indicators, 1, 18t.Flow monitoring See Rate of flow monitoring

Hydraulic failure mode causes and indicators, 20t.

Indicators, 1, 22

Leak detection, 6control limits, 7double-walled protection, 7by flow increase, 7frequency, 7indicators, 23means, 6by sniffer inspection, 6by visual inspection, 6

Leakage detectionby flow increase, 7by pressure buildup, 7by sniffer inspection, 7by visual inspection, 7

Lubricant analysis, 9control limits, 11evaluating wear rates, 10indicators, 23measuring contamination of lubricant, 10measuring inorganic contamination, 10measuring lubricant degradation, 10measuring metal particles from wear, 9measuring organic contamination, 10monitoring frequency, 11sampling techniques, 11

Maintenance inspection, 12characteristics to consider, 12

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

HI Centrifugal Pump Design and Application Index — 2000

26

Maintenance inspection (continued)coupling flexible elements inspection, 12erosion inspection, 13frequency, 13hydraulic performance, 13indicators, 24key/keyways inspection, 12shaft bending fatigue inspection, 12shaft torsional fatigue inspection, 13torsional overload inspection, 13

Monitoringbaseline, 1failure mode indicators, 1, 18–21frequency, 1–2indicators, 22–24

Power monitoring, 3control limits, 3frequency, 3indicators, 22means, 3

Pre-installation hydrotest, 15axially split case pumps, 16control limits, 16double suction pumps, 16frequency, 16indicators, 24means, 15vertical double casing can type pumps, 16warnings, 15

Pressure boundary leakage failure mode causes andindicators, 20t.

Pressure monitoring, 7control limits, 8frequency, 8indicators, 23means, 8

Proximity probes, 9

Rate of flow monitoring, 11control limits, 11frequency, 11indicators, 24measuring rate of flow, 11

RPM See Speed monitoring

Safety considerations, 2Seal leakage failure mode causes and indicators, 18t.Sealless pumps

bearing wear monitoring (plain bearings), 14failure mode causes and indicators, 21t.temperature monitoring, 4

Severity level, 1–2Shaft breakage mode causes and indicators, 19t.Shaft position monitoring, 11

frequency, 11indicators, 24proximity probes, 11

Shutdown limit (defined), 2Speed monitoring, 13

constant speed systems, 14control limits, 14by electric counter, 14frequency, 14indicators, 24methods, 14by revolution counter, 14by strobe light, 14by tachometer, 14variable speed systems, 14

Temperature monitoring, 3control limits, 5frequency, 5indicators, 23liquid film bearing and seal faces temperatures, 4means, 4motor winding temperature, 4pumped liquid temperature rise, 4rolling element bearing temperatures, 4sealless pump liquid temperature, 4sealless pump temperature damage, 5temperature sensitive fluids, 4

Vibration monitoring, 8bearing housing vibrations, 8control limits, 9frequency, 9indicators, 22means, 8proximity probe, 9shaft vibrations, 8on vertical pumps, 9

VOCs See Volatile organic compoundsVolatile organic compounds, 6

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

This page intentionally blank.

Copyright © 2000 By Hydraulic Institute, All Rights Reserved.

M122