cathodic protection marshall e. parker.pdf

8/10/2019 cathodic Protection Marshall E. Parker.pdf

1/177


2/177

Pipe Line

Corrosion

Cathodic

Protection

THIRD

EDITION


3/177

Gulf

Professional Publishing

an imprint of

But terworth-Heinem ann


4/177

PipeLine

Corrosion

^ ^and ^

Cathodic

Protection

A

practical manual

lor

corrosion engineers,

technicians, and

field

personnel

THIRD

EDITION

MarshallE.Parker

Edward G.Peattie


5/177

Gulf

Professional Publishing

and Butterworth-Heinem ann are

imp rints

of

ElsevierScience.

Copyright

1999

by

Elsevier Science (USA).

All

rights

reserved.

Originally

published

b y

G ulf Publishing Compa ny, Houston,

TX.

No part

of

this publication

may be

reprodu ced, stored

in a

retrieval system,

or

transmitted in any

form

or by anymeans, electronic, m echa nical,

photocopying, recording, or otherwise, w itho ut the prior written perm ission of

the

publisher.

Permissions m ay be sought directlyfromElsevier's Science & Technology

Rights Department

inOxford, UK:

phon e: (+44) 1865

843830,

fa x: (+44)

1865

853333,

e-mail: p [email protected]. Y ou m ay also complete

your request on-line via the Elsevier Science ho me page

(http://www.elsevier.com), by selecting 'CustomerSupport'and then

'Obtaining

Permissions'.

This book is printed on acid-free pape r.

Libraryof C ongressCataloging-in-Publ icationData

Parker, M arshall E.

Pipe line corrosion and cathodic protection.

Includes index.

ISBN

0-87201-149-6

1.

PipelinesCorrosion

2.

Pipel inesCathodic

protection.

I.

Peattie,

EdwardG. II.Title

TJ930.P33 1984

621.8'672

83-22630

The publisher offers special discounts on bulk orders of th is book.

For information,pleasecontact:

Managero fSpecialSales

Elsevier Science

200

Wheeler Ro ad

Burlington,

MA01803

Tel:781-313-4700

Fax:781-313-4802

For

information

on all

Gulf

publications

available, contact

our

World Wide

W eb

hom epage at

http://www.bh.com/gulf

10

9 8 7 6

Printedin theU nited Stateso fAmerica.

iv


6/177

C o n t e n t s

Preface to Third Edition vii

Preface to

First

Edition ix

1

Soil Resistivity Su rveys

1

Soil ResistivityUnits. Two -Terminal Resistivity Determina tion. Four-

Terminal

Resistivity Determination. Other Methods. Locating

H ot

Spots on Bare Lines. Surveys fo r Ground Beds. Area Surveys. Loga-

rithmic Resistivity Ranges. Sum m ary.

2 Potential Surveys 16

Pipe-to-Soil Potentials: Electrodes. Electrode Placement. Pipe Line

Connection. Surface Potential Survey

fo rCorrosion. Pipe-to-Soil Po-

tential as a Criterion of Cathodic Protection. Other Applications of

Pipe-to-Soil Potentials. Other Criteria. Summary.

3

Line Currents

31

Measurement of Line Current in Test Section. Stray-Current Studies.

Long-Line

Currents. Cathodic Protection Tests. Coating Conductance

Measurement. Summary.

4

Current Requ irement Surveys

36

The Problem: Coated Lines. Principles ofC urren t RequirementTest.

Current

Sources

for

Tests. Temporary Ground Beds. Special Condi-

tions. Summary.

5 Rectifier Systems forCoated Lines 45

General Design Principles. A ttenu ation Curves. LineTermination. An-

odeProxim ityEffects. Atte nu atio n w ith M ultiple Drain Points. Design

Procedure. An Alternate Method. The Economic Balance. Summary.

v


7/177

6

Ground

BedDesignand Installation 59

Design Principles. D isturb ing Factors. Field M od ification . Installatio n

Methods. H igh-Sil icon Iron Anod es. Steel An odes. H orizontal G raph -

ite Anodes.

Deep

Anodes. A Typical G round B ed Instal la t ion . Sum -

mary.

7 G alvanic Anodeson Coated Lines 78

When

and Where to Use Magn esium . W hen and W here to Use Zinc.

General System Design. Insta llation Procedure. Polarization and Final

Adjustment.

M odif icat ion of

Process with

Experience. Sample Prob-

lems. Summary.

8 HotSpot Protection 88

What Hot

Spots

Are.Hot Spot Protection. LocatingHotSpots.An-

ode Selection and Spa cing. Field In stallatio n. Field D esign. Zinc An-

odesin HotSpot Protection. Installation Details. Su pervisionandCon-

trol . Summary.

9 Stray-Cu rrent E lectrolysis 1

Stray CurrentCorrosion. Sources of Stray Cu rren ts. Detection of Stray

Curren t . Remedial Measures. Negat ive Bu s Bon ding. Exposure Areas.

Potential Surveys. Secondary Exposure. Summary.

10

Interference

in Cathodic

Protection

109

The Problem. Basic Solutions. Design. Crossing Bonds. Calculation of

Bond Resistance. M ult iple Bo nds.

Auxi l ia ry

Drainage. Parallel Lines.

Radial-Flow Interference. Foreign Lines w ith Insu latin g Join ts. Sum -

mary .

11 Operation

and

Maintenance

125

ImportanceofAdequate Su pervision. Fai luresin Rect i f ier-GroundB ed

Systems.

Failures

in

Ma gnesium Anode Systems. M inim um Inspection

Schedule

for

Rect i fier System. M inim um Inspect ion Schedule

for An-

odes. Monitor System.

Troubleshooting.

S u m m a ry .

12 Coating InspectionandTesting 136

Construction

Inspect ion. Evaluat ion

of

Coat ing

in

Place. Coating

or

Leakage Conductance. Pearson Surveys. Accelerated Coating Tests.

Summary.

Appendix

A

Fundam entals

of

Corrosion

146

Appendix B Cathodic Protection of Steel inSoil 150

Appendix

C

Corrosion

of

Steel

in

Soil

153

Appendix D Attenuation Eq uations 160

Index 163

vi


8/177

P r e f a c e to T h i r d E d i t i o n

A pipe line buried in the earth represents achallenge. It is made of

steela strong,but chemicallyunstable, ma terial and isplaced in an

environment

which

is

nonuniform, nonprotective,

and

n onyielding.

It is

the

duty

of the

corrosion engineer

to

study

the

properties

of

th is system

to ensure that the pipe line will not deteriorate.

In 1955, when I was first working on cathodic protection fo r pipe

lines

in

Saudi Arabia,

the

firstedition

of

this

bookby

Marshall Parker

was jus t a year old. Fortunately, the company library contained M r.

Parker's book. I found its simplicity and directness preferable in ap-

proaching a

complex subject.

During

the 30

years since

its

p ublica tion, generations

of

pipe line

en-

gineers and technicians have used this book as their first exposure to

corrosion control. M an y bookson thesubject have been pu blished since

1954,but theParker book isstillthe best introductionto the fundamen-

tals. Newtechnologyhasbeen developed, yet theprinciplesofcathodic

protection

are

still

th e

same.

The

result

is

that

we

have more sophisti-

cated

instrum ents

to

use,

bu t the

measurements have

not changed.

Con-

sequently,Ihave retained thestill-valid m aterialo f theoriginal Marshall

text

and

m ade changes o nly whe n better

and

shorter methods

are

avail-

able.

The

first task

of the

pipe line corrosion engineer

is to

study

the

prop-

erties

of the

earthen environment. First,

we

shall learn

how to

measure

th e

re sistivity

of the

soil

as a

prepa ration

fo r

further

work.

In the

second

vii


9/177

chapter we shall study another electrical measurement: potential differ-

ence.

To do

this,

we

shall learn about

the

standard electrode

for

this

measurement. In Chapter 3 we go on to line currents and extend these

measurements to current requirement surveys in Chapter 4. Thus, the

current

necessary

to design a cathodic protection

system will

be

calcu-

lated.

Chapters5 and 6showthecomputations requiredfor thedesignof an

impressed current cathodic protection system. Then, Chapter

7

shows

the

designprocedure

for a

sacrificial cathodic protection system.

The

problem

of

partial cathodic protection

of a

pipe line

by

concen-

trating

on

hot spots

is

discussed

in

Chapter

8. A

problem

in

Chapter

9,

which

doesnot

usually occur nowadays,

is

that

of

stray-current cor-

rosion; however,

a

corrosion engineer should still

be

able

toidentify

this

phenomenon and

find

its

source.

Interference (Chapter 10) is a problem

corrosion engineers face every

day and are

still learning about.

The last two chapters of the book (Chapters 11and 12)show the op-

eration and maintenance of a cathodic protection system, and how to

evaluate

the

coatings system

in

place.

The

Appendixes containmaterial

basic

to the

knowledge

of all

corrosion engineers.

Atthe end ofthis

book,

the

reader

w i ll

bewellon the way tobeinga

capable corrosion engineer.

Edward

G.

Peattie

Professorof

Petroleum Engineering

Mississippi State University

vlii


10/177

P r e f a c e

t o F i r s t E d i t i o n

Howdo

people become pipe line corrosion engineers?

N ot by

obtain-

ing a degree in the subject, for no such degree is offered. Many corro-

sion engineers hold degreesin chemistry, chemical engineering, elec-

trical

engin eering, or any one of several others. M an y others either have

no

degree, orhave studied in some field apparently or actually remote

fromcorrosion. All of these men became corrosion engineers by on-the-

job

experience

and by

individual

studysom eby trial-and-error

meth-

ods, having been assigned

the

responsibility

of

protecting

some

struc-

turefrom

corrosion; some by workingwith people already experienced.

That this should be the case is not surprising, particularly when the

protection of un dergro und stru cture s against corrosive attack is still as

much

an art as it is ascience. Say, rather, tha tit is atechnology; mosto f

the design procedures used areeither em pirical, or, atbest, arebasedon

empirically

modified theory. Almost every cathodic protection system

installedhas to be adju sted , by tria l, to do its job properly. It is com m on

experience thatno twojobs are alike; every new project contains some

surprises, some conditions

not

previously encountered.

These things are true for two main reasons. First, we do not as yet

knowenough aboutth e subject; there is

still

room for the development

of more powerful an alytic al m ethods. Second, the soil is a bewilder-

inglycomplex environ m ent,

an d

structures placed thereinaffect

one an-

other in

very complicated ways. Almost never

can we

m easure directly

a

single qu antity wh ich

weseek;

what

our

meters usually indicate

is the

ix


11/177

result of w hat we are looking for and a n um ber of par t ia l ly un kn o w n

disturbing factors.

Thecorrosion

engineer, then, must

be to a

large extent self-trained.

And he can

never expect

tocomplete

that t ra in ing . This m anual

is an

attempt

to

present

some

workable methods

of

doing some

of the

th ings

th e pipe line corrosion engineer is called upon to do. In no case is the

method given necessarily th e method; it is merely one method. If, by

use of

this

manual,some

corrosion

engineers

avoid

someof the

pitfalls

intowhichIhave fa l len ,andsave tim eand

effort

inobtaininga ndinter-

preting field data, then I w ill be happy tha t m y tim e has been w ell spent

setting down some

of the

experiencea

smal l par t

of

which

has

been

his own,

a

very large share

of

w hich

is that of

m a n y

othersand

thus

providing,

in a

way,

a

shoulder

to

look over.

Marshall

E.

Parker

Houston, Texas

January

18,

1954

x


12/177

1

S o i l R e s is t iv i ty S u r v e y s

Pipe linesmay be cathodically protected ineither earthor water.To

determine if cathodic protection may be used to prevent corrosion of

steel pipe lines,

we

must first learn

how to

measure

the

resistivity

of

these en vironm ents .

SoilResistivity Units

Theuni tofsoilresistivity is the

oh m-centimeter,

usually abbreviated

ohm-cm;the resistivity of a given soil is nu m erically equal to the resis-

tance

of a

cube

of the

soil

one

centimeter

in

dim ensions,

as

measured

from opposite faces (Figure

1-1).

The resistance of a rectangular solid

other

than

a

cube

is

given

by

whereW , L, and D are the dim ension s in centimeters, as show n in Fig-

ure

1-1,and p is the resistivity; the un it m ust be ohm -cm inorderfor the

equation

to be consistent. The resistance between any two terminals, of

any

size

and

shape,

in

contact with

a

body

of

soil,

of any

extent ,

is de-

termined

by the size and spacing relationships and by the resistivity of

th esoil.

For

simp le cases,

th e

resistance

can be

computed,

but the

math-

ematical complexities

are

often very great.

1


13/177

Figure

1-1.(a)Resistivity @ )in ohm-cm isnumerically

equal

to the re-

sistance R)inohmsbetweenopposite facesof acubeone cm on the

side,(b)

Resistance

of arectangularsolid,(c)Soilbox,inwhich is ob-

tained

bymeasuringthe resistance

between

theplanesof the two po-

tential

pins.

Two-Terminal

Resistivity

Determination

An

apparatus similar

to

that illustrated

in

Figure

1-1 may be

used

to

determine

th e

resistivity

of

soil samples;

in

corrosion investigations,

however, it is much more

useful

to measureth e soil in place.

If

tw o

term inals

are

placed

in the

soil,then

itwillbe

possible

to

deter-

mine its

resistivity

by

measuring

the

resistance between

the two

elec-

trodes.If

they

are

placed close together,

it

w ill

be

necessary

to

m aintain

aknown

and

fixed distance between them ;

if

they

are far

enough apart ,

surprisingly

enough ,

the

indicated value

isvirtually

independent

of the

distance. Either AC or DC may be used, and the resistance may be de-

2

Pipeline Corrosion

and Cathodic

Protection


14/177

Soil Resistivity Surveys 3

Figure 1-2.

Shepard canes

for

soil resistivity measurement. Current

from

a

three-volt battery (two flashlight cells)

is

passed through

the

soil

between two

iron electrodes mounted

on

insulatingrods. Current flow

is

measuredby adouble range milliammeter (0-25 and0-100)graduated

to

read directly

in

ohm-cm

(10,000-400and500-100).

Cathode

is

made

largertoavoid

polarization;

accuracyisabout 6o when tipsaresepa-

rated

by 8inchesormore. Meter,

batteries,

andswitcharemountedon

anode

rod.

termined

by

measuring current

and

potential,

or by a

bridge

circuit,in

whichthe unknown resistance is compared to one which is within the

instrument.

Figure 1-2illustrates Shepard canes, which use DC impressed be-

tween

two

ironelectrodes.This apparatus

is

rapid

in

use,

but

gives

the

valuecorresponding to a small sample of soil, in the immediate neigh-

borhood

of the

electrodes. Figure

1-3

shows

an AC

bridge-type appa-

ratus; this

deviceis

alsovery rapid

in

use,

but it,

too, measures only

a

smallsample. Numerous variations of the latter are on the market.

Four-Terminal

Resistivity

Determination

To include a larger sample of the soil and measure the resistivity at

greater depths, the four-terminal or Wenner method may be used. The

basic circuit is shown in Figure

1-4;

the following figures (Figures 1-5,

1-6, and 1-7) show some of the specific methods based on this tech-

nique. Detailsofoperationvary withtheparticular instrument

used,

but


15/177

4 Pipeline Corrosion

and

Cathodic Protection

Figure 1 3.

AC soil rod. Current from an AC source (usually a battery,

buzzer,and

condenser)

is

passed through

the

soil between

a

steel

rod

andan

insulated

steel tip.Theslidewire isthen adjusteduntilno

signal

is

heard

in the

headphones;

at thepointof

balance

Inpractice,theslidewire R

R

B

isgraduated to readRdirectly; then

= R - C, whereC is aconstantfor theparticular rod, determinedby

calibrationin a solution of known resistivity.

Figure 1 4.Four-terminal Wenner)measurementofsoil resistivity. Dis-

tance (b), depth of electrode, must be small compared to (a). The basic

formula is

Theresistivity

is averaged to a

depth approximately equal

to the

elec-

trode spacing (a).


16/177

So/7Re sist ivity Surveys

5

Figure 1 5. Ammeter-voltmeter measurement of resistivity.Astorage

battery of drycellsmay be used forcurrent. Coppersulfateelectrodes

will

help avoid

polarizationerrors.The voltmeter used must behighre-

sistance, or a potentiometer may be

employed.

If the

distance

(a) is

made

5 feet,

2

A >

inches,thenthe

formula

simplifiesto

Directionof current

should

be reversed, andreadingsaveraged, to bal-

ance

out

extraneouscurrents

or

potentials

in the

soil.

the principle is common to all. It should be noted that the resistance

measured

is

that between

the two

inner

or

po tentia l electrodes;

the

outer pair serve

to

introduce

the

current into

the

soil.

The value obtained is an

average

value to a depth approximately

equal

to theelectrode

spacing; actually

it is

influenced

to

some degree

by

soil lyin g at even greater depths. There is no sharp divid ing line but

the

influence

ofsoil at

greater

and

greater depths becomes smaller

and

smaller.

Other

Methods

There are at least three possible indirect methods of measuring soil

resistivity in which no contact is made with the soil at all. An induction-

type

pipe locater gives an indication of the resistivity below it and may

be calibrated to yield fairly acc urate resu lts. In general this method

may

bevery

useful

inscoutingout low -resistivity areas w hich canthen


17/177


and

Cathodic Protection

Figure

1-6. Vibroground

(simplified

diagram).Vibrators

V ,and

V

2

are

synchronized;

Vi

converts the

battery's

DC

into

AC; the

slidewire

S is

adjusteduntiltheIRdrop

across

R,justbucks out the current in the po-

tential

circuit,

asindicatedby azerodeflection on the

galvanometer

G;

R

is

then

readdirectly on the

slidewire,

and isobtained

from

=

2xaR

be explored more precisely by one of the methods already described.

The v alue and phase angle of the impedance (high-frequ ency) be-

tween

tw o

parallel conductors lying

on or

parallel

to the

surface

is a

function of thesoil resistivity, and thu s a device could be b ui lt w hich

would

employ this principle. It is also true thatthe tiltof theradiation

field of a remote radio transmitterat the surface of the earth is a func-

tion

of the

soil resistivity. Both

of

these methods offer some promise

of

development into very rapid and possibly automatic systems for the re-

cording of resistivity, but the development has not yet been done.

Locating Hot Spots on BareLines

Bare lines, particularly gathering lines, are often so situated that the

point of

m ax im um economic return

for cathodic

protection

is not the

preventionof all leaks, but the preven tion of85-95%of them . This can


18/177

Soil Resistivity Surveys

7

often be done at acost as low as 15% of that of full protection by the

application

of

galvanic

anodesusual ly

magnesiumto

the hot

spots''

or

areas

o f

m ax im um corrosiveness. This

has

been shown

to

cor-

relate highly with resistivity so that a system of locating areas of

low,

and

of

relatively

low,

resistivity,

is the

indicated technique. Such

apro-

cedure is known as a soil resistivity survey, and consists of a series of

mea suremen ts taken along the line (or right-of-w ay for a planned lin e),

using any of the methods described. The four-point method is most

commonly used.

Readings should

be

ta ken according

to a

systematic procedure; there

is

some conflicthere between the demands of science and those of prac-

ticality. If

readings

are

taken

atuniform

spacing,

th e

statistical ana lysis

of the

entire

set

will

be

more representative

of the

corrosive conditions

along

this line,

as

compared

to

other lines.

In

other w ords, such

a

proce-

dure w ill give

th e

m a x i m u m

of

general information.

The

m a x i m u m

of

useful

inform ation, however useful

in the

design

of a hot

spot protective

Figure 1-7. Soil resistivity Megger (simplified diagram).

DC

from

the

hand-cranked generator

G

isconvertedto AC by thecommutatorC

c

,

and enters thesoilthroughC

1

and C

2

, after passing through the current

coil;

the AC potential picked up between

P

1

and

P

2

is converted to DC by

commutatorC

p

and is fed to the potential

coil.

The needle comes to rest,

under the opposing forces from the twocoils,at a point which indicates

the

resistanceR.

is

then obtained from

=

2iraR


19/177

8

Pipeline Corrosion

and

Cathod ic Protection

system,

is notobtained in

this way,

butratherby

some such method

as

th e

following:

1. A

reading

to be

taken

at

least every

400

feet.

2. A reading to be taken w herever there is a visible chan ge in soil

characteristics.

3. Two successive readings should not

differ

by aratio greater tha n

2:1

(when

a

reading

differs

from

th e

preceding

by

m ore than this,

then it is necessary to go back and insert a rea ding; this should be

repeated

unt i l

condition 3 is met).

4. As anexceptionto the last

rule,

no two

readings

need

be

taken

closer than

25

feet.

5.

As another exception to the 2:1 rule, it does not apply when the

lower

o f the two

readings

is

greater tha n

20,000

ohm-cm.

The

fig-

ures given in this set of criteria may be varied for-specific cases;

theforegoingset hasbeen

found

useful andeconom icalin thesur-

vey of over a mill ion

feet

of gathering lines.

For

this type

o f

survey, readings w ould

be

taken w hic h correspond

to

th e

soil at pipe depth. This means that holes will have to be drilled or

punched

for thesingle-rod

apparatus.

The

four-terminal

system

should

be used witha spacing of about 1

'/2

times pipe depth; very often a 5-

foot, 2'/2-inch

spacing (which makes the multiplier

2ira

just equal to

1000) is well suited for this purpose.

Data obtained

in the

survey

are

plotted

on a

diagram representing

th e

length of the line, as illustrated in Figure 1-8. The vertical resistivity

scale

is

logarithmic, since resistivity ratios

are of

interest, rather

than

their

differences. From such

a

diagram

the

hot spots

can

readily

be

located. This technique

will

be

discussed more

fully in a

subsequent

chapter.

Surveys forGround Beds

The resistivityof the soil in which a ground bed is to be installedis

one of the primary design quantities involved. It is most importantto

get as low a total resistance as possible in order to obtain the necessary

current

output with

the

m i n i m u m a m o u n t

o f

power. Unlike

the

survey

to

determine

corrosiveness,

as in the

preceding topic,

in

this case

th e


20/177

Figure 1-8. Soil resistivity profile. Measurements taken

by any of the

methods described

are

plotted

as

ordinates(using

alogarithmicscale),

with distances along

the

pipeline

as

abscissas.

resistivity

to

great depth

is

imp or t an t.

For

this reason,

the

four - te rm inal

method is the only one

useful

here. It is customary to take readings at

5-,

10-,and

20-foot

pin

spacing (actual values used

are 5

feet,

2

l

/2

inches;

10 fee t , 5 inches; and 20 fee t , 10 inches ; g iv ing mul t ip l ie r s of

1000,

2000,

and4000, respectively).

The prospective site should be marked off in a rectangular grid, with

spacings

of 50 or 100

feet,

and

readings taken

at

each grid intersection;

where differences

are too

great, intermediate readings should

be

taken.

Soil Resistivity Surveys 9


21/177


and

Cathodic Protection

These

diagrams should alw ays be plotted in the

field

so that a comp lete

picture

will

be

obtained. Contours

can be

drawn,

as

indicated

in

Fig-

ure

1-9.

If

colors

are used, the readings at all three pin spacings can be

put

on the same diagram; only one is shown

here.

Note that the values

chosen for contours are logarithmic, as were the vertical scales in the

line resistivity survey.

Area Surveys

Where a set ofpiping and other underground structures is spread out

over an area, as in a refinery or chemical plant, a survey similar to the

one

jus t described w ill

furnish

information

useful

both

in the

prediction

ofcorrosiveness

and in the design of protective systems. The spacing

used between readings

is

u sua lly greater

than for

ground

bed

design.

A

Figure 1-9.

Resistivity

plat.

Resistivity readings taken

at

spaced inter-

vals, with

additional

readings

at

criticalpoints,

are

indicated

by the

small

figures.

The

curves represent

linesof

equal resistivity. Such plats

are of

value

in

locating sites

for

groundbeds

and in

their design.


22/177

SoilResistivity Surveys

11

Figure 1 10.

Resistivity

histogram. Equally spaced resistivity readings

taken

over

an

area

or

along

a

pipelineroute

are

plotted according

to the

frequency ofoccurrenceofreadingsin thedifferent ranges. Note that

successive

ranges cover equal resistivity ratios, rather than arithmetic

differences.

pre l iminary

spacing

of 400

feet

is

often used, with intermediate posi-

tions used where necessary

to get a

clear picture.

As in the

ground

bed

survey,

readings

are

usually displayed

on a

resistivity contour

map of

th e

plant site. Usually only one pin spacing isused, and the most fre-

quent

value is the nominal five-foot spacing already mentioned.

In

addition to thegeographical distribution shown on the resistivity

plat, it is useful to

prepare

a

histogram,

or

frequency distribution,

as

indicated in

Figure 1-10. This shows

the

number

of

readings found

in

various classes or

ranges

of

resistivity; these ranges should

beloga-

r i thmic,as are the

contour intervals

on the

plat already described.

Logarithmic

Resistivity Ranges

Three examples have already been given

of

logarithmic treatment

of

resistivities. There

is

sound scientific basis

for

this approach, aside

from

the

intuitive observation that ratios

are

more important than differ-

ences. It hasbeen found that large numbersofresistivity readings, when

expressed

as

logarithms, tend

to fall

into

standard

distributions.

The


23/177

12 Pipeline Corrosion and Cathodic Protection

standard distribution is a well-known and extremely useful concept in

statistical analysis.

A consequence of this is that two extensive structures (such as pipe

lines) each exposed

to a

wide variety

of

soils

can be

compared

as to

their

corrosion exposure most accurately

by

comparing their

logarithmic

mean

resistivities.The logarithm ic mea n resistivity of a set of resistivi-

ties is the value whose logarithm is the average of all the logarithms of

the measured values. These can most easily be averaged by grouping

them into

classes,

as shown in Figure 1-11,where the logarithmic m ean

of 115

different

readings is

found

to be

14,900

ohm-cm.

It is almost universal practice to

choose

a ratio of

2:1

for the succes-

sive values used in any loga rithm ic system, for contours, h istog ram s, or

the calculation of a me an. Three

different

systems are in widespead use:

thetw o

already illustrated

(A)

based

on 100

ohm-cm,

as

used

in

Figures

1-9 and

1-10;

(B ) based on

1000 ohm-cm,

as

used

in

F igure

1-11;

and (C)

which

is not

truly logarithmic,

in

that

th e

ratio

is not al-

ways the same. All three are shown here for comparison:

A) B) C)

50

62.5

50

100 125 100

200 250 200

400 500 500

800

1,000

1,000

1,600

2,000 2,000

3,200 4,000 5,000

6,400 8,000 10,000

Scale

(C) has the

advantage

in

that

all of the

values

are

round

num bers whereas

if

either

(A) or (B) is

extended

fa r

enough

in

either

direction, the numbers tend to become awkward. On the other hand,

when a

mathematical analysis

is

made

of the

data,

the

irregular ratios

involved in scale (C) introduce complications. Note,

when

(C) is used

for

a

histogram,

the

horizontal widths

of the

cells should

be

made

proportional

to the

logarithms

of the

ratios.

Log 2 =

0.301,

log 2.5 =

0.398;

so the

ratio

of the two spaces

should

be

about 3:4.

The corrosion technician now has vario us methods at his disposal to

measure the resistivity of the soil in which the pipe line is to be in-


24/177

Soil Resistivity Surveys

13

Figure

1-11.Summary

of

soilresistivity measurements.

stalled. The question may now be asked: What if the pipe line is to be

installed

in

waterwould

the

same methods

of

measurement work?

The answer is

yes,

but w ith reservations. Ac tually, a single sam ple of

the water

is

enough ,

and it can be

measured

in a

soilbox,

as

shown

in

Figure 1-12. The countless va riations of resistivity w hich w e have

shown in

Figure

1-9 for

soil

do not

exist

in the

water environment ,

and

so onere sistivity valu einohm -centimetersisusual ly enoughfor the en-

gineer

touse. For soils, as we have seen, it is more complicated.


25/177

14 Pipeline Corrosion and Cathodic

Protection

Figure 1-12.Typical connections

for use of

soil

box

with varioustypes

ofinstruments. (Courtesy

of M. C.

Miller Co., Inc.)


26/177

So/7 Resistivity Surveys

15

Summary

The

resistivity

of

soil

is

usually

measured

by the

four-pin

method

with

theVibroground(Figure

1-6),

possibly

the

easiest

instrument

to

use.It is

possible

to

attach

a set of

pins

andleads

which will minimize

th e

tim e to get soil resistivities, and a com plete surve y may be done in a

relativelyshort tim e. The new est examp le of the soil resistance m eter is

shown in

Figure 1-13

and has

been available since

1981.

Its

advantage

is

that there

are no

moving parts

and no

vibrators

to

change. Although

called a

four-pin

soil resistance meter,

it is

also excellent

for

m easuring

water samples

in a

soilbox.

Figure1-13.Four-pinsoilresistance meter.


27/177

2

Potential

S u r v e y s

Pipe-to-Soil

Potentials:Electrodes

The potential difference between aburied pipeand thesoilis ofcon-

siderable importance, either ininvestigatingth e corrosive conditionsor

in

evaluatingtheextentofcathodic protection being applied. This qu an -

tity actuallyis measured by connectingan instrumen t betweenthepipe

itself

(direct metallic contact)

and a

special electrode placed

in

contact

with the soil. Thiselectrode is also called a half cell .

The most common typeof electrode in use is that which employs a

m etal-to-electrolyte ju n ctio n consistingofcopper incontact

with

asatu-

rated solutionofcopper sulfate; this particular com binationis madeof

easily

available m aterials and is very stable. Figure 2-1 shows two typi-

cal electrodes. Other types in use

are:

1. The hydrogen electrode, used

only

in laboratory investigations.

2. The calomel electrode, used often in fresh water or saltwater.

3. The

lead/lead

chloride electrode, frequently employed in studyin g

the corrosion of lead cable.

4. The silver/silver chloride electrode, used in seaw ater because it is

not subject tocontam inationby salt incursio n.

5. The pu re zinc electrode (in packaged b ac kfi ll). This is suita ble as

ape rm an ently installed reference electrode. (Note: all these listed

electrodes m ust be corrected to give readings sim ilar to thecop-

per/copper sulfate

halfcell.

See Appendix C.)

16


28/177

PotentialSurveys 17

Figure 2-1. Copper sulfate electrodes. The essential parts are: 1) a

piece of copper, to which the instrument is connected; 2) a saturated

solution

of

copper sulfate,

in

contact w ith

the

copper;

3) a

porous mem-

ber,placed incontact withthesoil.It isessential thatnometal touchthe

copper sulfate solution

except

copper.

Excess

c rystals areadded,to en-

sure

that the solution always

will

be saturated.

Voltmeters

One choice for the instrum en t to be used in m easuring

this

potential is

the voltmeter,

as

shown

in

Figure 2-2; there

are two

possible sources

of

error

in its

use.First,w hi le

a

voltm etercorrectly indicates

the

potential

difference across it sterminals,

the current which flows through the in-

s t rument (a voltmeter is essentially a

mi l l iammeter )

introduces an IR

drop in the rest of the circuit, which is not included in the reading. For

example,

if the

resistance

of the

external circuit (pipe, lead wires, soil,

and

electrode)

is

one-fourth

the

resistance

of themeter,

then

the

meter

wil l

read only four-fifths

of the

total potential;

the

error

wil l

be

20%.


29/177


Figure 2-2.

Voltmeter measurement ofpipe-to-soil

potential.

Connec-

tionto the

pipe

may be by

means

of aweldedor

soldered leadwire

as

illustrated),by a contact

bar,

or bymechanical

connection

to a

valve

or

fitting

above the surface. Thelatter

should

be used withcaution,as

flangedfittings

sometimes

introduceextraneouspotentials.If thesoilis

dry,a

little

watermay beaddedtolowerthe contact resistance; it isusu-

ally

adequate

to dig a

littlebelow

the surface.

A second source

of

error

is

that

the

passage

of

current through

the

circuitmay polarize the electrode or even the pipeitself, and so change

the potential we are trying to measure. Both of these errors may be

minimized by the use of a high-resistance voltmeter; not less than

50,000ohms per volt is recomm ended, and this is often too low. Read-

ings

should always be taken on two

different

scales (introducin g two

different values of meter resistance); if they are in substantial agree-

ment ,

th e

value

may be

accepted.

In

th enex t section, methods w illbegivenforobtainingthecorrected

value when

the

error

due to

voltmeter resistance

is not too

high (see

Equations 3-2 and 3-3).

There

is

also available

a

special dual sen sitivity voltm eter w hich

offers

two diffe ren t resistance va lues on the same

scale.

If the reading is

the same for these two (the change is m ade by m erely p ush ing a button),

then no correction is needed. If pushing the button makes a small

change in the reading , then a very simp le correction can be m ade. An d,

finally,

if the change is large, an equation com parable to those cited pre-

viously can be used. This meter isvery useful for m akin g rapid poten-

tial

readings.


30/177

Potential Surveys

19

Figure 2-3.

Slide-wire

potentiometer, a)

Basiccircuit.When

the

slide-

wire S is adjusted so that the galvanometer G shows no deflection, then

the unknown potential connected

to the

terminals

can be

read directly

on

the slide-wire

scale.

This simple

circuit

could be used only if the bat-

tery

voltage w ere

constant,

b) Modified

circuit,

with provision for adjust-

ing

battery voltage with resistor

R

B

,

using standard

cellSC for

calibra-

tion.

ResistorR

0

protects the galvanometer against excess currents; it

isshorted out by the switch when the instrument is almost perfectly bal-

anced. Mostinstrumentsfor

corrosion

usehavetwo ormorescales,in-

stead of the one shown here.

Slide-Wire

Potentiometer

These problem s largely

are

overcom e

by the use of the

potentiometer,

an instrum ent which draws no current from th e circuit at the pointof

balance.

Itshouldberemem bered,

however,

thatcurrentisd rawn while

balancing is in progress and

that

some polarization m ay occur. Figure

2-3 shows this instrument and indicatesitsuse.

Potentiometer-Voltmeter

Another

approach

to the

problem

is the potentiometer-voltmeter, one

form

of

which

is

shown

in

Figure 2-4. Both

of

theseusefulinstrum ents

suffer

a

disadvantage w henfluctuating potentials

a re

encountered,

as in

a stray-curren t area, since it is impossible to follow rapid fluctuations

with them.


31/177

20 Pipeline Corrosion and

Cathodic Protection

Figure 2-4. The potentiometer-voltmeter, a) Shows the basic circuit.

When the galvanometer deflection iszero,the external potential con-

nected to the terminals is the same as that portion of the battery voltage

which is indicated onvoltmeter V ;

this,

then, is adirect reading of the

unknownpotential.In (b)several refinements have been added:(1)w ith

switch

$

open,ahighrangeisadded; 2) the tworesistorsR ,andR

2

,of

different values, provide a fine and coarse adjustment, and 3) switch S

2

controlsaprotective resistor, similar tothatinF igure 2-3. More thantwo

ranges are possible, and it is also possible to use the same instrument

asanammeterby

building

inshunts.

Vacuum-Tube Voltmeter

This instrument (not illustrated) has, through

the use of

transistors

and

other recent developments, now reached a stage of reliabilityand

stability

w hich makes

it

quite

useful in the

field. Battery-poweredunits

are

availablewith input im pedances

as

high

as 10

m egohms,

so the

cir-

cuit

resistance e rror ceases to be a problem . They po ssess an ad ditiona l

advantage

over

th e

various poten tiom eter types

of

ins truments

in

that

theydo not

have

to be

balanced,

and

hence

may be

used

to

follow

fluc-

tuating potentials w ithout

difficulty.

Solid-State

Voltmeter

This

instrument

(Figure 2-5)

is the

successor

to thevacuum-tube

volt-

meter

and

differs

onlyin the substitutionofsolid-state transistorsfor the

more

vulnerable

vacuum

tubes.

T he

resistance problems have been vir-

tual lye l imina ted ,and accurateresults m ay beobtainedby

no ntechnical


32/177

Potential Surveys 21

personnel .

A

p ar t icu lar model

has

selectable

inpu t

resistances

of

1,

10,

25, 50,

100,

and 200

megohms. This al lows

theuserto

determine pipe-

to-soil

potent ia lsin such

diff icul t

env i ronmen tsasci ty pavem ents . The

range in reading varies from0.1 m v to 20 m v. This is a rugged ins tru -

ment

for the

technic ian

who has to

take many readings .

Figure 2-5.

Solid-state DC voltmeter. Courtesy of M. C.

MillerCo.,

Inc.)

MulticombinationMeter

This

is the

ins trument most often used

by

corrosion

engineers

be-

cause

of its

versatility.

It is a

developm ent

of the

potentiometer-voltme-

ter

already described

but has the

following particular characteristics,

as

shown

by the

latest 1983 model

in

Figure 2-6.

The

right meter

has a

l iquid crystal display with5 ranges from - 20 mv to 200 v andw i th

selectable i n pu t resistances

of

1-200 m ego hm s.

The

left meter also

has

a liquid

c rystal display

an d

serves both

as a

vol tmeter

and an

ammeter .

The

four vol tage ranges include- 20 mv to 20 v, and theamm eter range

is from

20

ma to 20 a. As an ohmmete r , it reads from

2 0

o h m s to

2000 ohms . Th i s ins t rumen t

is

sti l l ava i l ab le w i th

the

convent ional

m o v i n g

pointer dials , which are not as easy to read as are the l iquid


33/177


34/177

Potential Surveys 23

Figure

2-7. Theelectronic potential

meter.

(Courtesy of M. C.

Miller

Co.,

Inc.)

whe n

the

electrode

is

placed directly over

the

pipe.

The only way in

whichthe

theoretically

correct

reading could

be

taken would

be toplace

th eelectrode

adjacent

to thepipe,or to use a complicated nullcircuit

involving

two or more

electrodes,

in

which

the

effect

of the IRdrop is

canceled out. Both of these are quite t ime-consuming, and thus have

gained little favor.


35/177

24 Pipeline Corrosion and Cathodic

Protection

Figure 2 8. Potential meter with optional accessory electrode exten-

sion.

Courtesy

of M. C.

Miller

Co., Inc.)

Pipe Line

Connection

As

shown in

Figure

2-2,

connection

to the

pipe

mustbe

made

by an

insulated lead wire

or by an

insulated contact

bar or,

best,

by

permanent

test stations installed along

the

pipe line

at

regular intervals.

The

advantage of the

test

station is that, once it is installed, itwillnot be

necessaryto damage the

pipeline

coating to

make

contact

with

the

line.

Figure2-10shows a typical installation. It can be

seen

that

aportionof

the

coating

isscraped off ,and thewireisthen attachedto thelineby a

process known

as cadwelding.

This

is a

small thermite unit using


36/177

Potential Surveys

25

Figure 2-9. Current

and

potential field around

a

pipe.

The

dotted

lines

represent

the

flow

of

current

to or

from

a

pipe lying

in

soil

of

uniform

resistivity; thedashed

lines

represent equal differences ofpotential. It

can

be seen that no

point

on the surface is at the samepotentialas that

immediately adjacent

to the

pipe,

but an

electrodeplaced

at A

repre-

sentsa closer approximation thanoneplacedat B.

a l umi num

powder to produce enough heat to bond the copper wire to

th e

steel. Then

the

connection

and the

surrou ndin g bare metal

are

coated

with a

coating similar

to the

pipe line coating

and

usually f in-

ished with tape. The lead wire isthen brou ghtto thesurfaceand isusu-

ally

raised to a heightof

five

feet, where a box containinga lead will

make connection of the testing meter very convenient. When the test

station isinstalled, it isusually convenient to use it as one end of a test

section for

m easuring l ine current ,

as

shown

in the

next chapter.

SurfacePotentialSurveyfor Corrosion

Figure 2-11 represents the current and potential fields around a sec-

tionof p ipe line w ith a single active corroding (ano dic) area. Th e distri-

bution ofpotential along th e surfaceof theearth above th epipe clearly

indicates th e locationof theactive area;thus asurveyo fsurface poten-


37/177


Figure

2-11.Potential field surroundingasingle anodic point.Aniso-

latedanodic pointon aburied pipe linecan bedetected easilybysur-

facepotentialmeasurements,provided that readings

are

taken

at

suffi-

cientlyclose spacing.


38/177

Potential Surveys 2 7

tials

along

th e

length

of a

pipe lin e should

be of

value

in

locating

corro-

sion. Figure 2-12 diagrams th e procedure for making such a survey,

while

Figure

2-13

shows a typical pipe-to-soil potential of a protected

line.

Variations in

soil resistivity introduce potential differences which

m ay

mask those sought, so the problem is by no means as simple as it

might

appear. Another

drawback to

this

methodand a

very serious

oneis

that

i twil lnot

disclose

a

hig hly localized

cell,

where

the

anode

andcathodeare

very

close

together,as for

exam ple,

the

common con-

centrat ion cell where

th e

anode

is on the

bottom

of the

pipe

and the

cathode

on top, at the same location.

Pipe-to-Soil Potential

as a

Criterion

of

Cathodic Protection

It

is almo st un iver sally accepted that a steel structure unde r cathodic

protection isfully protected if thepotentialis atleast0.85-voltneg ative,

referred

to a

standard copper-saturated copper

sulfate

electrode placed

in

th e electrolyte immediately adjacent to the metal surface. The entire

structure is

fully

protected,

of

course,

onlyif this criterion is met at

ev-

erypoint on the surface. This is not a m inimu m requirement ; it is a

m a x i m u m .

In

other words,

it is

almost certainly true that protection

is

complete when this is achieved. It is also possible to have complete

freedom from

active

corrosion

at lower potentials.

As m en tion ed, it is imp racticably slow to place the electrode imm edi-

ately adjacent

to the

pipe surface;

it is

completely impossible

to

take

a

reading at every point of the surface. As a consequence of these two

difficulties, it is generally accepted that th eelectrode is to be placed in

the

soil im m ediately over

th e

l ine;

it is not

difficult

to

show that,

for a

well-coated line, this is as good as nex t to the pipe. F or a bare lin e, this

is not

t rue ,

but

bare lines

are

rarely

fully

protected

in any

event.

The

other

difficulty

isavoided byassumingace rtain degree ofcontinuityin

th e

potential gradient along a line. This assumption is almost always

valid,

a nd , itturns ou t , when itbreaks dow n because of discontinuities

in soil resistivity,

it

fails only

in

areas which

are

virtual ly

noncorrosive

without

protect ion.

It

is

cus tom ary , then ,

and relat ively

safe ,

to

consider

a

coated line

to

be

fully protected

if a

survey along

i ts

length yields

a

reasonably smooth


39/177


Figure 2-12.

Surface

potential survey.


40/177

Potential

S urveys 29

Figure 2-13.

Pipe-to-soil potential of protected

line.

Typical distribution

ofpotential alongaline with cathodic protection unitsat A, B, and C.

Points

X and Y are thecritical points; if thepotential remains satisfac-

toryatthese points,theentire line isprobably adequately protected.

potent ia l

curve which does not dip below the value of 0.85 volt; all

readingsbeing takenare

wi th

respect to astandard copper sulfate elec-

trode.

Oncea detailedsurvey

over

the system hasestablishedthatthelineis

protected,

then

periodic

checks

at the critical points

(low potential

points betweenunits)canmake itpossibleto know that protection is

still being afforded. Thissubject

is

discussedmore fully later.

Other

Applicationsof

Pipe-to-Soil

Potentials

This measurement forms a part of the technique used in several types

ofstudies, most of which will be described later. Among them may be

ment ioned:(1)

interference studies,

(2)

stray current

investigations,(3)

coatingconductance measurements, and (4) cathodic protection tests.

Other Criteria

In

addition to use of the

-0.85-volt

standard for adequacy of a ca-

thodic protection system, two other methods are in use:

1.

Test coupon. Burying corrosion test coupons along

the

pipe line.

These should

be

weighed

and

connected

to the

line

at

locations

possibly

subject

to

extreme corrosion conditions.

After a

given

exposure time, they should

be

removed

and

reweighed.

A ny

loss


41/177

30 Pipeline CorrosionandCathodic Protection

in

weight

may be

used

to

calculate

th e

degree

of

protection neces-

sary.

2.

Potentialchange.

Another

criterion(called doubtful

by

Uh lig*

but

oftenused in

cathodic protection

design)is to use a

potential

change of 0.3 volt when

current

is applied as a measure of ca-

thodic protection. Thisinvolves an interruptedcathodic protec-

tion

current measurement, as will be

shown

in Chapter 4. Also,

thisconceptm ay beused in primary designcalculations, as de-

scribed in

Chapter

5.

Summary

We have thus seen that, although there are several reference elec-

trodes (half cells) available for potential difference measurements, only

the copper/copper sulf ate electrode is used for mo st of the rea ding s the

corrosion

engineer will take

on

land. Also,

for

precision,

the

poten-

tiometer, particularlyin the

form

of acombinat ion ins t rum ent ,isprefer-

able

to the

low-resistance voltmeters.

The

placement

of the

reference

electrode

is

best made

on the

surface

of the

earth directly

over the

bur-

ied

pipe

line.

Sometimes,th e

soil

should bem oistened, although prefer-

ably

nothing should

be

done

to

affect

th e

envi ronment .

For

contact with

the

pipe

line, it isadvisabletohave test stations installed at regular in -

tervals along

all

major pipe l ines. This will el iminate

the

problems

of

direct

metal

contact

with

th e

pipe

itself,

which will

be

necessary

if the

permanent leads

are not

available. Although

the leap-frog

system

maybe

useful when there

is a

distance between test stations,

it

would

be

advantageous

to

have

a

reel

of

wire available

to

extend

th e

measure-

ments along

the

pipe l ine , part icularly

in

regions near

th e

m idpoints

b e-

tween rectifiers. Modern methods of pipe-to-soil measurement use re-

cording vol tmeters, which

may be

used

to

complete

a

series

of

pipe-to-soil potentials along

a

pipe line

in a

very short time.

* Uhlig ,H. H.,

Corrosion and Corrosion Control: An

Introduction

to

Corrosion Science

andEngineering, 2 ndEdition, 1971, John Wiley and Sons, Inc. , New York, p. 225.


42/177

3

L i n e C u r r e n t s

Now that

w e

know

the

most comm on pipe l ine m easu rem ent, pipe-to-

soil potential difference, we shall s tudy some other measurements nec-

essary for corrosion control. The n ex t is l ine curren t m easurem ent.

MeasurementofLineCurrent in TestSection

Atest section consists of a len gth of uni nt err up ted pipe l in e (no joints

other than welds, no f i t t ings) between two points to which an ins tru-

ment may be connected (see Figure 3-1). The connection points usu ally

take the form of a pair of permanently attached lead wires; the use of

contact bars is

di f f icul t

and general ly unsat is factory.The resistance of

the pipe

itself

(not the lead wires) between the two points must be

k n o w n ;

it

wi l l usual ly

be on the

order

o f

0.001 oh m ,

or may be ex-

pressed in conductance un i ts , as 1.00 am pe re/m il l ivo l t . To o btain th is

va lue , the lengthof the section must vary according to thepipe weight ;

values between 50 and 400 feet are common.

Voltmeter

Line

Current

Measurement

The difference in potential between the two leads, mul t ip l ied by the

conductance of the section,

wi l l

give the value of thecur ren t f lowin gin

the l ine . This difference

in

potent ial

may be

de termined wi th

a

mi l -

31


43/177

32

Pipeline Corrosion

and

Cathodic Protection

Figure3-1.Measurement

of

IRdrop

in

line section. R

w

in

Equation

3-1

istotal resistanceof

7 ^,

T

2

,

L

1(

L

2

,

andL

3

,unlessL

1

andL

2

are apairof

"calibrated" leads matched

to and

supplied with

the

instrument,

in

which case

R

w

= T

t

+ 7

2

+ ^ -3 only-

Figure 3-2.Calibration

of

test section.

The

conductance

of the

section,

in amperes per

millivolt,

is given by

Ifthesetwosectionsare notequal, then

livoltmeter,

as

shown

in

Figure

3-2.For

currents occurring naturally,

a

meter with a full-scale

deflection

as low as one millivolt may be re-

quired; the currents used in cathodic protection give rise to larger

val-

ues.

An

instrument with

a

range such

as

this will have

a low

internal

resistance, perhaps as low as one

ohm;

the resistance of the lead wires

will

thus be of

considerable

importance. If the value of the lead wire

resistance is known, the

corrected

reading may be obtained from


44/177


45/177

34

Pipeline Corrosion and Cathodic Protection

CalibrationofTestSection

The conductance of a test section may be approximated

from

the for-

mula

This

is

only

a

rough approximation;

th e

factor

m ay

vary

from 3.5 to

as

high

as

5.0,

and

further

variation

is

introduced

by the

weight toler-

ance

for

lin e pipe.

By far the

better method

is

that

o f

direct calibration ,

asshown inFigure 3-2. Tw oconn ections areusedforcurrent inp ut ,an d

tw o

others (w hich w ill rem ain

as

perm ane nt test leads)

for the

voltmeter

connection. Not all of the current

f lowing

throughthe ammeter , how-

ever,

f lows

through

th e

test section; some

of it f lows

through

a

remote

earth path. This effect is quite variable, depending upon coating con-

ductance and earth resistivity; to correct for it, an additional pair of

leads is required.

Stray-CurrentStudies

W henever fluc tua ting currents or p otentials are observed, stray-cur-

rent

effects

are

suspected.

Useful

clues

to the

source

of

such strays

can

be had

from

24-hour recordings of line current potentials. If theactual

current values are desired, corrections may be made by Equation

3-1;

or resistance

can be

inserted into

the

circuit

so as to

makeR

w

=

R

m

\

the

correction factor

will

then

be

exa ctly

2 and may be

made

by

selection

of

theproper cha rt, so as to record corrected m illivo lts. Co m pariso n of the

charts with records

of

operation

of

suspected equipment

will often lo -

cate

the

source beyond question.

Long-Line

Currents

A use of

line current measurement more

in

vogue

in the

past than

at

present is thetracingof long-line currents and the location of the areas

ofdischarge to earth. If cur ren ts of this type were the

sole

cause of pipe

line corrosion, all of the anodic areas could thus be found ;

unfor tu-

nately, however, th e most common attackonbu ried pipes isassociated


46/177

Line Currents

35

with corroding

cells

of very small dimensions; very often the anode is

on the bottom and the cathode on top of the pipe only a few inches

away;

line current studies cannot locate such a condition.

Cathodic Protection

Tests

The current flows associated with cathodic protection systems, either

permanent

or

temporary test equipment,

are

typically long-line cur-

rents; a very valu able part of such testing is in the m easurem ent of these

currents; the application

wil l

be discussed in a later chapter.

Coating Conductance Measurement

The same thingis true of the determ ination of coatingcondu ctance

a

type

of

measurement which will grow

in

impo rtance

to the

corrosion

engineer in the

future .

Line current measurements are muchless sub-

ject to disturbing factors than are potential measurements and

offer

th e

engineer a valuable approach to this

difficult

subject; this

will

also be

discussed later in greater detail.

Summary

As

has already been shown, if test stations along the pipelines have

been installed, it is very easy to construct a test section for line current

measurement . Al though the lengths m ay vary for these sections, it is

advisable to have them at least 100 ft long for 8-inch schedule 40 pipe

and

as

much

as 400 ft for

30-inch

OD

pipe. Again,

th e

multipurpose

corrosion meter can be used, since values as low as one mill ivolt may

have

to be measured. Also, the test section may beused tom easure the

coating resistance once the conductance of the pipe itself has been cali-

brated.


47/177

4

C u r r e n t

Requirement

S u r v e y s

The

Problem:Coated

Lines

The simplest approach to the

cathodic

protection of coated lines is to

aim at

obtaining

a

pipe-to-soil potential

of

0.85 volt

(to a

CuSO

4

elec-

trode)

throughout

the line.

Meticulous investigations

m ay

show that

protection can be had at lower figures, but only on bare or very poorly

coated

structures

are the

savin gs likely

to be

great enough

tojustify th e

additional engineering work required.

In

determining

th e

amount

and

distribution

of

drainage current

to

produce

the

desired potential,

tw o

basic approaches

are

possible:

(1) the

complete

protectionof the

line with tempo rary

installations,

these being

varied and shifted untila satisfactory combination is

found;

and (2) the

determination

of the

electrical characteristics

of the

l ine,

and the

calcu-

lation of the system from these data.

The

first

method

is

almost impossibly complicated

in

execution

and

is

seldom even attempted. The second is theoretically possible, al-

though inpractice a certain amountof trial-and-error procedure isusu-

ally

n ecessary. Acompromise techn ique,inwh ichthe trialanderrorof

testing the

performance

of

various combinations

is

done

on

paper,

rather

than

in the

field,

is

recommended.

36


48/177

Current Requirement Surveys 37

Principlesof

Current

RequirementTest

Current

is to be

drained

from

the

line

to a

temporary ground bed,

and

th e

effects determined

for as

great

a

distance

as

they

are

m easurable (far

beyond

the

point

of

complete

protection);

from

one or

moresuch

tests,

the

behavior

of the

l ine with respect

to

cur ren t d ra inage

can be

determined with sufficient accuracy

tocarryout the

design

of a

com-

plete protective system . The e nd resu lt of this design w ill be the specifi-

cation

of

drainage units

at

certain p oints, drain ing specified quan tities

ofcurrent; the detail design of anode beds to accomplish this is another

matter.

Field

Procedure

The

following steps

are

taken ,

in

order:

1. Static

potentials. A pipe-to-soil potential survey is made over the

entire line; permanent test leads

are

recommended,

but bar

con-

tacts may be necessary. Spacing m ay vary from a few hundred

feet

on poor coating to several m iles on good coating; often road

crossings willbe adequate. Readings should be taken at all cas-

ings

(reading po tentials

of

b oth casing

and

pipe),

and on all

cross-

ing lines and other structures w hich m ight either be shorted or be a

source

of

interference

in

either direction. Such structures include ,

in addition to casings, bridges, A-frames, supports of all kinds ,

and

sometimes even such things as metal fences and guy wires.

2. Polarization run. At a chosen location, preferablyone whichap-

pears

to be

suitable

for a

rectifier installation,

a

temporary dra in-

agepoint isestablished and a steady current drained for aperiod

of from

one to three hours. The pipe-to-soil potential of a nearby

point

(within

a few

miles ,

on a

well-coated line;

a few

hundred

feet on a

poorly coated one)

is

observed du ring this inte rval

to

fol-

lo w

th e

course

of

polarizat ion.

The

va lue

of

current

to be

used

should

be

large enough

to

cover

a usefully

long portion

of the

l ineperhaps

all of

i tbut should

not be too

large.

The

pipe-to-

soil

potential near

the

drain

point should not,

in

general, exceed

three volts.


49/177

38

Pipeline Corrosion and

Cathodic Protection

3.

Potential survey.

An interrupter, either manual or automatic, is

now

inserted in the circuit and placed in operation. The schedule

should consist

of

unequal

on

and

off

periods

of

predeter-

mined duration;

40

seconds

on and 20

seconds

off is a

good

combination touse.W ith this operating, asurvey should be made

of

pipe-to-soil potentials

at all of the

points covered

in 1. Static

potentials.It is

im portant that

the

same electrode position

be

used

as was used for the static potentia l at each loca tion . At each po in t,

both the on and off

potentials should

be

read;

the on

reading should

be

taken just before interruption,

and the off

reading as soon as possible after interruption. Care should be

taken

that

these

two

readings

are

correctly

identified; the

length

of the

cycle

is

used

fo r

that purpose.

It is

not safe

to

assume that

the

higher (more negative) readingisalwaysthe

on

readingas

this is not thecasewitha properly insulated lateral orcasing. It is

recommended that a graph of these values be ma de in the

field

as

th e

readings are taken in order to detect any peculiarities which

might call for additional readings. Figure

4-1

shows a drawin g of

an

automatic interrupter suitable

for use on the

above test.

Figure 4-1.

Current interrupter. (Courtesy

of M. C.

Miller Co., Inc.)


50/177

Current

Requirement Surveys

39

4. Pola rization and cu rrent

check.

The

series

of

po tential readings

at

a point near thedrain point, aswell as a

series

of current values,

should be

continued during

th e

potential survey.

It is

desirable

to

keep

th e

current constant throughout

th e

test

if

this

can be

done

easily. Where this

is not

possible,

it is

better

for it to

change

smo othly over a considerable range of values than to change often

and abruptly. If the current change is not great, then only one or

tw o

values need

to be

taken during

the

potential survey. Success-

ful survey s ha ve often been made w ith only

the

value

at the

start

and

finish (together with those taken during

the

polarization ru n).

These values should also be plotted in the field.

5. Line current survey. If

spaced p airs

of

test leads

are

available,

th e

line current should

be

determined

oneach

side

of the

drainage

point ,and a t aremote locationineach dire ction . Interm ediate val-

ues may

also

be useful ,

part icular ly

if it

turns

out

that there

are

some short-circuited structures on the line. Like the potential

readings,

these

should consist

of on - off pairs.

6.

Repeat as necessary.

New drainage points are chosen, and the

outlined

process repeated

unti l the

entire line

has

been covered

with

values large enou ghto beusable.It is by nomeans necessary

to

achieve complete protection over

th e

whole line,

or

even over

any

portion thereof.

Graphical

Presentationof Data

Allof the

info rm ation gathered

in the

survey should then

be

prepared

in

the

form

of a set of curves, as follows:

1. Current

and

polarization.For

each test run,

th e

value

o f

drainage

current and the

potential

at the

check point should

be

plotted

against

time (see Figure 4-2). The curve will have many points

close together duringtheearly polarization ru n ,andperhaps only

one or two

thereafter,

but

should always

be

plotted

as a

complete

curve.This

wil l

makei tpossible tointerpolateandth us determine

th e

value

of the

current f low ing

at any

particular time,

tocorrelate

with the

various potential readings.

2. Longitudinal distributioncurves.On a

single cu rve,

the following

should

beplotted against distance alongth e entire line : staticpo-


51/177

Figure

4-2.

Longitudinaldistributioncurves.

Distributionof

pipe-to-soil

potentialalongtheline;thevery obviousdifference betweenthe two

ends of theline

illustrated

are

attributable

to ashort-circuitedbare cas-

ing on the

right-hand

section. The

horizontal

unitis 1000

feet.

tential, off potential, on potential,

and

line current.

There

will

be only one static curve, but there will be one of each of the

others for each test run. Normally, these will overlap, and colors

may

be required for clear separation. These curves will present

most

of the

data taken

in the field in

graphic

form.

3.

Attenuation

curve.

Again

on a

single sheet (Figure

4-3),

but

using

semilogarithmic paper,

th e

following

are

plotted against distance

(using the samehorizontal scale as

above):

a. Polarization potential, AV>; this is the difference between the

off potential and the static potential, at each point.

b. Driving voltage,

AE;

this is thedifference between the on

potential and the off potential, at each point.

c. Line current,A/;again th evalueto beplotted is thedifference

in

th e

on value

and the

off value;

due

attention must

be

paid

to thedirections,an d thealgebraic

difference taken. That

is, if the

test shows

40

m illiamperes flow ing east during off conditions

and300 milliamperes flowing west during the on part of the

cycle, then A / is 340 milliamperes, the net change produced by

the test current.


and Cathodic

Protection


52/177

Current Requirement Surveys

41

Figure 4 3.Attenuation curves. Driving

voltage AE) and

polarization

potential

AV

P

)

plottedonsemilog paper against distance alongtheline.

For

an

infinitely long

line,

such curves should

be

straight lines;

for a fi-

nitelineendingin aninsulated joint, concave upward,as on theleft-

handsection;and for a"grounded"section, concave downward,as on

the right-hand section.Thetest point numberson allthree figures corre-

spond.

4. Polarization chart (Figure 4-4). Thisisusually plottedonlog-log

paper, although

the

results

are

often

as

good when ordinary linear

paper

is

used. This

is a

plot

of the

value

of

A V / > plotted against

A, and, where several tests have been

run on the

same line,

or

even where there have been repeat tests

on the

same section with

thesameo r

different

current values, allpairs

from

alltests can be

combined. In general, it wi l l not beadvisable to combine values

taken

on

different lines

or on

different sections

if it is

known that

pipe size, line age,

or

type

of

coating

is

markedly different. This

chart wil l usually

be a scatter

diagram rather than

a

smooth

curve; often it isadvisable todraw limiting lines rather thanasin-

gle

curve through

the

points.

The

chart shown

in

Figure

4-4 is

bet-

ter

than most

in the

degree

of

conformity

to a

single curve.


53/177

Figure4-4.Polarization chart.Fromthisplot,the functionalrelationship

betweendriving voltageandpolarization potential may beobtained.

Test point 2 is

anomalous.

Theexplanation is found in the short-cir-

cuited

casing

at

this

location,

which

minimized

polarization. The

anom-

aly atpoint8 is due to

proximity

to the anode; the apparent value of the

drivingvoltage

is toohigh.

Current

Sources for Tests

Whereverylargecurrents20-100amperesarerequired, aw eld-

in g

machine

is the

usual source

of DC

power

for

current requirement

tests. Smaller values can be supplied by storage batteries. It is possible

to

connect

any

number

of

such batteries

inseriesto

obtain voltages high

enough

for medium-to-high

resistance ground beds.

As

high

as 120

volts (ten 12-volt batteries)

has

been used with success.

Very

small currents, u p to perhaps two am peres, can be suppliedwith

ordinary

dry

cells, often

at a

considerable saving

in

cost,

if no

storage

battery

is

rea

cathodic protection marshall e. parker.pdf

Documents