-
CorrosionPredictionModellingA guide to the use ofcorrosion
predictionmodels for risk assessmentin oil and gas productionand
transportationfacilitiesA J McMahon, D M E Paisley
Sunbury Report No. ESR.96.ER.066dated November 1997
Main CDContents
-
Contents
Summary
Acknowledgements
Introduction 1
"Cassandra 98" Corrosion Prediction Spreadsheetby A J
McMahon
Introduction 5Quick Start 6Limitations of Corrosion Prediction
Models 8Detailed Description of the Spreadsheet 11Comparing Output
from the "Cassandra 98" Model with Field Data 27Appendix 1: Henry's
Law Constans for CO2 Dissolved in Brine 29
The Use of Corrosion Prediction Models During Designby D E
Paisley
Introduction 31Important Factors not Covered by the Corrosion
Model 35Effect of Corrosion Inhibitors 42Predicting the
Effectiveness of Corrosion Inhibitors - 48'The Inhibitor
Availability Model'Recommended Values for use in the Inhibitor
Availability Model 51Comparisons of the Inhibitor Availability
Model with BP's Previous Model 62Corrosion Rates of Low Alloy
Steels 64Preferential Weld Corrosion 65Effects of Pitting
66Choosing an Optimum Corrosion Allowance 67Applying Models to
Different Flow Regimes 69Applying Models to Transportation
Equipment 72Applying Models to Process Equipment 86Flow Velocities
in Process Pipework 89Economic Tools to Use During Materials
Selection 92
References 95
Installation of the Cassandra 98 Excel Workbook 97
Page
-
Summary
This document decribes BP's current approach to Corrosion
Prediction and itsuse during the design of pipelines and
facilities. It is divided into two sections.
The first section introduces a new prediction spreadsheet called
Cassandra 98*which is BP's implementation of the CO2 prediction
models published by deWaard et al. It builds on these models to
include BP's experience of such systems.The pocket inside the front
cover of this report contains a floppy disc whichcontains the
necessary programs and spreadsheets to run it together with a setof
installation instructions.
The second section discusses how the prediction model may be
used for designpurposes and it introduces several improvements from
previous guidelines.These include the use of the probabilistic
approach to corrosion prediction andthe use of corrosion inhibitor
availabilities instead of efficiencies. It also discussesthe use of
"corrosion risk categories" as a way of quantifying the corrosion
riskat the design stage. The floppy disc also contains a
spreadsheet for calculatingthe risk category.
To illustrate the points made examples have been obtained from
many BP assetsworldwide. Where financial data are shown it is from
1997.
Since this subject is continually changing it is anticipated
that these guidelineswill be updated in future years and so any
comments or suggestions regardingeither the content or appearance
of them would be very welcome.
*In Greek mythology Cassandra was the daughter of Priam and
Hecuba. She was endowed withthe gift of prophecy but fated never to
be believed. She is generally regarded as the prophet
ofdisaster........especially when disregarded.
-
Acknowledgements
The authors would like to thank the following BP staff for
theircontributions to these guidelines.
Jim CorballyLaurence CowieMike FielderDon HarropBill HedgesWill
McDonaldTracy SmithSimon WebsterRichard Woollam
-
1Introduction
Carbon dioxide corrosion represents the greatest risk to the
integrity ofcarbon steel equipment in a production environment.
Compared with theincidences of fatigue, erosion, stress corrosion
cracking or over-pressurisation, the incidences of CO2 related
damage are far more common.Unfortunately, the engineering solutions
to eradicating the CO2 corrosion riskrequire high capital
investments in corrosion resistant materials. As Figure 1shows,
providing a corrosion allowance of 8 mm to carbon steel
flowlinescosts a significant sum at circa US$1 million per 5 km but
even this isinsignificant in terms of the costs of the various
corrosion resistant flowlineoptions.
Similar relative costs are incurred when specifying corrosion
resistantmaterials downhole or in facilities. This is rarely
justified. For this reason, CO2corrosion of carbon steel will
always be a problem that BPX has to deal with.Managing CO2
corrosion therefore becomes a priority and it can becomeexpensive.
The replacement of the original Forties MOL and the severedamage to
the Beatrice MOL are two examples of high costs that BPX
haveincurred in recent years due to unpredicted corrosion rates.
Successfulmanagement of CO2 corrosion starts off with the
identification of risks andcontinues with the provision of suitable
controls and the review of thesuccess of the controls via
monitoring - as illustrated in Figure 2.
Figure 1: FullyInstalled Costs forVarious FlowlineMaterials
Options inColombia (1997)
0
5
10
15
20
25
30
35
6 8 10 12 14 16 18 20 22 24 26 28 30
Nominal Flowline Diameter - Inches
Cost per
5 Km ($mil l)
Carbon steel 8mm ca
Duplex SS
13%Cr
Bi-metal 13Cr liner
Carbon steel, no ca
-
2INTRODUCTION
This document sets out BPs approach to the quantification of CO2
corrosionrisk through the use of predictive models. In doing so, it
also discusses thereliance that can be placed on corrosion
inhibition as the only viable controlmeasure for carbon steel and
the importance of suitable corrosionmonitoring. To put the
importance of this into context, corrosion costs BPX8.3% of its
capex budget and increases lifting costs by 14%, an average ofover
8 cents per barrel. Figure 3 shows that the costs are distributed
acrossthe entire range of facilities.
Apply ControlsMonitor Effectiveness
Quantify Risk
Figure 2: The FeedbackLoop that is Required forSuccessful
Managementof CO2 Corrosion
-
3INTRODUCTION
The quantification of corrosion risk is required at several
stages during an assetslife. The most obvious period is during the
project phase when the originalmaterials of construction are being
selected. This process must be repeatedduring the life of the asset
if failures or expansions require the procurement ofadditional
facilities. Quantifying the corrosion risk is also important in
tailoringinspection strategies. Risk based inspection is now widely
adopted and, as CO2corrosion represents one of the most important
factors governing the probabilityof failure for much equipment, a
reasoned approach should be taken. It isimportant that this
approach is theoretically sound but also reflects
pastexperience.
This version of the BP CO2 prediction model is the first to be
published since1993/4 when the guidelines on multiphase and wet gas
transport respectivelywere issued. The new guidelines incorporate
changes by the authors to the semi-empirical model used in the
original guideline as well as comprehensiveguidance on how to use
the spreadsheet included with this version. The newmodel also
includes the ability to predict the affects of changing flow
velocitieson uninhibited corrosion rates.
Downhole13%
Subsea59%
Chemicals4%
Topsides23%
Personnel1%
Figure 3: TheDistribution of Costs ofCorrosion Across TenBPX
North Sea Assets,1990 to 1994.
-
4INTRODUCTION
The new guidelines also consider the probabilistic approach to
predicting CO2corrosion. Probabilistic approach to design in
general is becoming morewidespread and offers several advantages
over the traditional deterministicapproach. The probabilistic
approach is neither endorsed nor disallowed but isdiscussed as, in
some cases, it may be more appropriate than a
deterministicapproach.
The approach to designing for the use of corrosion inhibitors
has been changedsignificantly. The previous approach described the
affects of an inhibitorthrough the use of an efficiency factor,
such as 90%. This does not reflect BPXsrecent field data generated
under severe conditions which showed inhibitorscan be more
effective than predicted. "Inhibitor efficiencies" have
thereforebeen replaced with "inhibitor availabilities" that more
closely reflect fieldexperience. There is a general move in the
industry towards this methodologyand it offers several
advantages.
However, it has become clear that for inhibitors to work
effectively thecorrosion management system must be highly
organised. Recommendations aretherefore included on methods to
ensure that the inhibitor availabilitiesassumed at the design stage
occur during the operational stage.
-
5"Cassandra 98" Corrosion Prediction Spreadsheetby A J
McMahon
"Cassandra 98 is BP's new implementation of the 1991, 1993 and
1995 CO2corrosion prediction models published by De Waard et al.
The pocket inside thefront cover of this report contains a floppy
disc with the programme togetherwith a set of installation
instructions.
The 1991 and 1993 De Waard models are already widely used within
BP andelsewhere in a variety of customised forms. This report
describes the newCassandra 98 spreadsheet for Microsoft Excel. It
is based primarily on the 1993De Waard model, incorporates some
equations from the 1991 model, and usesthe 1995 model to assess
velocity effects. The spreadsheet is intended to captureall the
best features of the 1991, 1993 and 1995 models [1,2,3]. Certain
extrafeatures from outside the De Waard papers, based on standard
physicalchemistry, have also been included. The source, background
and limitations ofall the assumptions and equations in the
spreadsheet are fully documented inthese guidance notes.
The Cassandra 98 spreadsheet is written in a simple and
accessible format withinMicrosoft Excel (version 7.0). It avoids
the use of macros or special techniquesso that the logic and the
calculations are as transparent as possible. Thisapproach also
ensures that the spreadsheet is immediately compatible with
newversions of Excel.
The Excel add-in module "CRYSTAL BALL" (from Decisioneering Ltd,
1380Lawrence Street, Suite 520, Denver, Colorado 80204, USA. Tel:
+1 303 292 2291.Cost ~100) enables probability distributions to be
set for each input cell and itthen uses Monte-Carlo simulation to
combine these into a probability distributionfor the resulting
corrosion rate. You must buy "CRYSTAL BALL" separately foryour
Excel environment. It can't be bundled with this spreadsheet. The
detaileduse of CRYSTAL BALL is well covered in the manufacturer's
handbook andtherefore is not repeated in these guidelines.
Care is required when comparing the output of any existing
in-house version ofthe De Waard models against this new Cassandra
98 spreadsheet. It is very easyfor errors and untested assumptions
to be entered into a spreadsheet whichmight then perhaps be passed
on from user to user and often compounded withother assumptions.
Cassandra 98 has been written from scratch with a
detailedre-evaluation of all assumptions, all of which are
presented. Cassandra 98 isintended to be a standard, reference
version of the De Waard approach for usewithin BP and its partners,
until such time that a more consistent approach tocorrosion
modelling becomes established within the oil industry. The
activities ofthe NORSOK industry forum in Norway are making helpful
moves in thisdirection.
INTRODUCTION
-
6This section gives enough information to allow experienced
modellers to makea start. The subsequent section gives a more
detailed description of all the inputand output parameters. The
spreadsheets themselves also carry frequent "cellnotes". These are
marked by a red dot in the top right hand corner of thosecells.
Just double click on the cell to read the contents.
To carry out a basic calculation enter the following input
values into the cellswith a white background:
Only the inputs in the preceding Table are needed for a
straightforwardnumeric calculation. Some further information is
required in order to carry outa probabilistic calculation using
CRYSTAL BALL. The spreadsheet can easily becustomised by individual
users to permit more extensive handling ofprobabilities:
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
QUICK START
Input Parameters
Probabilistsic Inputs
P total gas pressure bar F7%CO2 CO2 in gas mole % (NB = v/v%)
F8%H2S H2S in gas mole % (NB = v/v%) M8water composition ion ppm
values ppm (NB = mg/ltr) A15-L15brine pH enter known value, F17
or enter "d", "o", or "x" to accept one of the calculated values
shown in F18-F20(see Page 17)
T System temperature oC F24Ts Scaling temperature, enter
oC F25the calculated scalingtemperature, given in cellF26, or
another known orpreferred value
d hydraulic diameter m M24U velocity m/s M25
Parameter Comments Units CellTable 1: Inputparameters for
anumeric calculation
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
7
P F7 use a uniform distribution; set F7 as the maximum; set G7
as the minimum
%CO2 F8 use a normal distribution; adjust standard deviation as
necessary
brine pH F17 must enter a known or a calculated value; use a
normal distribution; adjust standard deviation as necessary
T F24 use a uniform distribution; set F24 asthe maximum; set G24
as the minimum
d M24 use a uniform distribution; set M24 as the maximum; set
N24 as the minimum
U M25 use a uniform distribution; set M25 as the maximum; set
N25 as the minimum
Output Parameters
1993 basic Vcor E32 the uncorrected corrosion rate for static
conditions
1993 correction factors G32-K32 correction factors for pH,
fugacity,scaling, and glycol
1993 corrosion rate G34 the corrosion rate for static conditions
corrected for pH
1995 corrosion rate G39 the corrosion rate for dynamic
conditions calculated from the components Vr and Vm in G37 and
G38
93/95 merged corrosion G41 the average of the 93 and 95 rate
corrosion rates; this cell enables
"CRYSTAL BALL" to combine the93 and 95 probability
distributions
Parameter Cell Comment
Parameter Cell Comments
The resulting output parameters are described in Table 3. See
p23 for a moredetailed description of how to interpret and use
these values. Briefly, the 1993rate should be regarded as the
minimum. Velocity effects may increase thisminimum rate as shown by
the 1995 rate. Hence, the 1993 and 1995 rates willnormally give the
lower and upper bounds on the expected corrosion rate. The1995
model is not accurate at low velocities and so it should be
ignoredwhenever it falls below the 1993 value.
Table 2: AdditionalInput Parameters for
aProbabilisticCalculation
Table 3: OutputParameters
-
8"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
The use of simple equations and the precision of the spreadsheet
environmentcan lead one to think that the De Waard corrosion models
are equally precise.However, this is not the case. The models are
only valid over a certain range ofconditions, and even within this
range a certain amount of data has beenignored if it doesn't fit
the main trends. Each model appears to be constructedby obtaining a
large number of corrosion rates over a range of conditions andthen
finding an equation which draws a line passing close to the
majority of thiscloud of points. The equations appear to be freely
adjusted in order to give thebest fit to the data. The primary
concern is to obtain a good fit to the data, ratherthan obtaining
mechanistically rigorous equations. These are empiricalengineering
models rather than scientific theories.
Neither the 1991 or 1993 De Waard papers give many precise
details about therange of validity of the models. The 1995 paper
does give a more thorough setof figures (see below) but still omits
important features such as the type of brineused in the tests, and
the elapsed time when the corrosion rates were measured.De Waard's
very early work used a 0.1% NaCl solution [4] and this may wellhave
been used in all the subsequent studies because his main focus has
alwaysbeen low salinity water in gas lines. Table 4 shows the
approximate ranges ofvalidity for the different parameters in the
Cassandra 98 spreadsheet.
LIMITATIONS OF CORROSION PREDICTION MODELS
Table 4 : Range ofValidity of De WaardModels
P
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
9
The spreadsheet gives freedom to enter any value for most
parameters. Whenthe input value is outside the approximate range of
the 1991 and 1993 De Waardmodels then the text will turn RED in the
cell as a warning. The predictedcorrosion rate may still be useful
but the user must accept the additional risk ofgoing beyond the
known limits of the correlations.
To develop the 1995 model [3] corrosion rates were obtained on
the IFE flowloop (Kjeller, Norway) using a radiochemical technique
to measure corrosionrates. Tests were carried out over 2-3 days but
there is no information about thecorrosion rate profile over this
time or when the final data point was taken. Datawere obtained for
the following conditions.
- St-52 DIN 17100 steel (Cr 0.08%, C 0.18%) which is similar to
ASTM A537Gr1
- 0.1, 3.1, 8.5, 13 m/s flow velocity- 20 - 90oC- 0.3 - 20 bara
CO2
Certain inconsistencies in the data set were eliminated prior to
developing themodel. These included:
- 0.1 m/s excluded- 13 m/s excluded when corrosion rate less
than at 8.5 m/s- 90oC excluded- CO2 >6.5 bar excluded
Eventually 221 data points were used in the main correlation
(Figure 2 ref 3).The main equations are specific to St-52 steel
because, "The equations obtainedfor St-52 showed a complete lack of
correlation for the other steels". The 15other steels were
normalised steels and quench-and-tempered (Q&T) low
alloysteels. These were examined over the following conditions to
produce somemodified equations which take account of steel
composition.
- 3.1, 8.5, 13 m/s flow velocity- 60oC- ca 2 bar CO2- pH
4,5,6
Limits of the 1995Model
-
10
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
For normalised steels a "Cr correction" and a "C correction" can
be calculatedseparately and together. For Q&T steels the "C
correction" has no effect and onlythe "Cr correction" is relevant.
The Cassandra 98 spreadsheet does not includethe steel composition
equations due to the poor correlations obtained whenfitted to the
model.
Errors in matching equations to data points are defined in the
1995 paper by"coefficients of determination". This is a complicated
statistical function rangingfrom 0 (poor correlation) to 1 (perfect
correlation). It is not the same as the"correlation co-efficient"
in regression analysis which scales from -1 to 1. The"co-efficients
of determination" in the paper are 0.91 for the main St-52equations
(after excluding the data that doesn't fit), 0.83 for the
normalisedsteels, and 0.80 for the Q&T steels. For the main
St-52 correlation thiscorresponds to a standard deviation of 25% on
the predicted corrosion rate. Thisis the error given in this
spreadsheet. Because of this error the predictedcorrosion rates are
only shown to one decimal place. A "CRYSTAL BALL"probabilistic
analysis gives a more realistic impression of the error on
eachprediction.
The De Waard models were all developed using water-only systems
in thelaboratory. The 1993 model is intended for nearly static,
aqueous conditions andso for all but the lowest velocities (see
page 77) it can be regarded as theminimum corrosion rate of a water
wet region in a gas/water, water/oil, or awater/oil/gas system. Due
to the different hydrodynamics in these field casessome assumptions
are required in order to apply the 1995 model effectively.These
assumptions will only affect the diameter and velocity values used
asinputs in the model. The other inputs will be unaffected. Table 5
gives somesuggested assumptions. However, users are free to develop
their ownapproaches to meet the demands of their own particular
circumstances. Someof the issues involved in extrapolating the
models to the field are discussed inmore detail on pages 27-28.
Errors on CorrosionRates
APPLYING THE MODEL TO DIFFERENT FIELD SITUATIONS
-
Units are specified for each parameter listed in this section.
The same units areassumed in all the equations given below and
throughout the Cassandra 98spreadsheet. The spreadsheet has a
"units conversion box" at cell P5. The UNITSspreadsheet allows
conversions between a wider range of units. The SALTSspreadsheet
enables conversion between an ionic analysis of brine and the
saltsrequired to prepare a synthetic analogue. The FUGACITY
spreadsheet is a data-base used to calculate fugacity corrections
at high total pressures.
P...total gas pressure (bara, i.e. bar absolute) INPUT cells F7
and G7
For a multiphase system this is simply the prevailing local P in
the gas. For aliquid only system it is the P in the last gas phase
which was in equilibrium withthe liquid, e.g. the separator gas in
the case of a crude oil export line. For adownhole liquid
pressurised above the bubble point then use the bubble
pointpressure (Figure 4).
For a simple numeric calculation, enter the P value into cell
F7. Cell G7 is thenunused. For a probabilistic calculation using
"CRYSTAL BALL", set up a uniformdistribution for P with F7 set as
the maximum and G7 as the minimum.
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
11
DETAILED DESCRIPTION OF THE SPREADSHEET
Total Pressure
Water only use pipe diameter and water velocity
Liquid/Gas use hydraulic diameter (see p 21)use true liquid
velocity rather than nominal velocity(see p 22)
Water/Oil use pipe diameter and total liquid velocity(n.b. this
ignores the possibility of water drop out orstratification which
could lead to the water phase moving more slowly than the oil
phase)
Water/Oil/Gas use a specialist multiphase program to calculate
the wall shear stress or the "C factor" for the pipe system,then
choose diameter and velocity inputs whichreproduce this
hydrodynamic value.
Field Situation Recommended ApproachTable 5: Applying the1995 De
Waard Modelto Field Situation
-
12
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
Figure 4: SchematicDiagram of an OilProduction System(downhole,
separator,export)
%CO2...CO2 in gas (mole%, which is same as v/v%) INPUT cell
F8
For a multiphase system this is simply the prevailing local %CO2
in the gas. Fora liquid only system it is the %CO2 in the last gas
phase which was inequilibrium with the liquid, e.g. the separator
gas in the case of a crude oilexport line. For a downhole liquid
use the %CO2 in the gas formed at thebubble point. If this gas
analysis is not available then use the CO2 dissolved inthe brine,
the Henry's constant, and the bubble point pressure to
back-calculatethe "effective %CO2" which would be required in the
bubble point gas in orderto sustain the known level of dissolved
CO2 (see box at cell P19). Indeed, thisprocedure can be followed
for any region where the CO2 dissolved in the brineis known, but
the gas analysis is unknown.
There may be occasions when it is helpful to apply parts of the
Cassandramodel to a water which is in equilibrium with ambient air
(e.g. for pHp redictions). The appropriate atmospheric inputs are P
= 1 bara and%CO2=0.035 mole%. Remember that under these conditions
the corrosionprediction from the model will only relate to the
dissolved CO2 component andnot the dissolved O2.
For a probabilistic calculation using "CRYSTAL BALL", set up a
normaldistribution for %CO2 using an appropriate standard
deviation.
%CO2
-
13
pCO2...partial pressure of CO2 (bara) OUTPUT cell F9
fCO2...fugacity of CO2 (bar) OUTPUT cell F10
The non-ideality of gases means that at high total pressures the
partial pressureis not an accurate description of the activity of a
gas component. The fugacityis the true activity of the gas
component. The 1991 and 1993 models use pCO2in the main corrosion
prediction equations and then at the end apply a fugacitycorrection
factor (Ffug) to account for fugacity effects. In Cassandra 98
theequations from the 1991 and 1993 models use fCO2 directly,
therefore there isno need to use a fugacity correction factor
(Ffug). The equations from the 1995model in Cassandra 98 also use
fCO2 directly - instead of pCO2. Hence, inCassandra 98, it is fCO2
which is used as the primary parameter for all theequations which
consider CO2 as an input.
Fugacity data from the work of R H Newton [5] are tabulated in
theFUGACITY.XLS spreadsheet in the workbook. The Cassandra 98
spreadsheetuses the input values of temperature and total pressure
to look-up the correctvalue of the fugacity co-efficient (g) in the
FUGACITY spreadsheet,
fCO2 = pCO2 g
The R H Newton data are generally applicable to many pure gases.
The datashow fugacity co-efficients as a function of "reduced
temperature" and "reducedpressure",
where Tr is reduced temperature (dimensionless)T is the
prevailing local temperature (oC)Tc is the critical temperature for
the gas (from tables) (oC)
pCO2
fCO2
Tr =TTc
pCO2 =P.%CO2
100
-
14
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
where Pr is reduced pressure (dimensionless)P is the total
pressure (bar)Pc is the critical pressure for the gas (bar)
Oilfields produce gas mixtures rather than pure gases. Hence, a
difficulty arisesin deciding whether it is the Tc and Pc for
methane or for CO2 that one shoulduse. In the Cassandra 98
spreadsheet, empirical values of Tc and Pc are assumedwhich allow
the Newton model to agree with the CO2/methane mixed gasfugacity
data in Figure 5 of the 1993 De Waard paper to 10%. In other
wordsthe De Waard data are used to calibrate the Newton model.
The De Waard calibration data are valid up to 140oC and 250 bar.
The Newtondata extends beyond these levels up to 300oC and 400 bar.
The general trendsin the data will be accurate under these extreme
conditions, however, theabsolute values are unchecked. For accurate
work it will be necessary tocalculate or obtain the correct value
of fugacity from elsewhere and thenmanipulate %CO2 in cell F8 by
trial and error in order to obtain the correctfugacity in cell
F10.
%H2S...H2S in gas (mole%, which is same as v/v%) INPUT cell
M8
H2S is not included in any of the De Waard models. It is only
used in theCassandra 98 spreadsheet in the calculation of solution
pH by XLpH (seebelow). It can be ignored completely simply by
entering zero.
Pr =PPc
CO2 31 73methane -82 45.8
empirical values used to correlate with De Waard data -37
56.7
%H2S
Tc Pc(oC) (bar)
Table 6: ReducedTemperature andReduced PressureValues for CO2
andMethane
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
15
It is by lowering the solution pH that H2S can potentially
increase the corrosionrate, often in synergy with CO2. In practise,
H2S tends to promote FeS surfacefilms which reduce the observed
general corrosion rate but which increase thelikelihood of
localised corrosion whenever the film fails. The CO2
generalcorrosion rate is often assumed as the worst-case localised
corrosion rate for theregions with no FeS film.
An alternative approximate approach for handling the presence of
H2S is toassume that every 1 mole% H2S has the same corrosivity as
0.01 mole% CO2.This rule of thumb assumes that 1 ppm dissolved CO2
and 200 ppm dissolvedH2S give roughly equal corrosion rates [6],
and that H2S is roughly twice assoluble in water as CO2 for a given
partial pressure [7].
pH2S ...partial pressure of H2S (bar) OUTPUT cell M9
pH2S = P . %H2S
water chemistry ..ion concentrations (ppm, same as mg/ltr) INPUT
cells A15-L15
The water chemistry is used to calculate the solution pH (see
below). Enter ppmvalues for Na+, K+, Ca2+, Mg2+, Ba2+, Sr2+, Cl-,
HCO3-, SO42-, Fe2+, acetate.(NB enter the sum of all organic acids
as acetate). Enter the %v/v value for glycolin cell L15. Use the
SALTS spreadsheet to check that the total positive andnegative
charges of the ions are roughly balanced. Any significant
misbalance(e.g. >10%) may invalidate the pH calculation. Note
that ion charges are handledin general chemistry by using the term
"equivalents": 1 mole of positive chargesis equal to one
equivalent; in other words 0.7 mole of Ca2+ ions is equal to
1.4equivalents of positive charge. Some further aspects of the
acetate entry arediscussed on p.19.
T D S...total dissolved solids in water phase (ppm, same as
mg/ltr) OUTPUT cell M17
pH2S
LIQUID PARAMETERS
Water Chemistry
Total DissolvedSolids
-
16
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
This is the sum of all the individual dissolved ions
concentrations. TDS and[HCO3-] are used in the Oddo & Tomson pH
calculation. TDS is also used toestimate the "salting-out" of CO2
as salinity increases. This will tend to reducethe concentration of
dissolved CO2 and thereby reduce the corrosion rate [8].The box at
X19 shows how to apply the salting-out correction. The
procedureuses "Henry's Law" to calculate the solubility of a gas in
a liquid.
pCO2 = KH XCO2
where KH is Henry's constant (bar/mole fraction)XCO2 is mole
fraction of CO2 dissolved in brine.
The Henry's constant from the De Waard paper is only valid for a
low salinitybrine (ca 0.1% NaCl). Therefore, by calculating the
true Henry's constant for aspecific brine it is possible to apply a
salinity correction to the De Waardcorrosion rate.
The salt-correction procedure first calculates the Henry's
constant used by theDe Waard model (equation 28 from the 1993
paper- which is used in thederivation of equation 13 in the 1993
paper),
where KH is Henry's constant (mole/ltr bar)
Note that this KH equation from the De Waard paper has different
units(mole/ltr bar) from those given earlier (bar/mole fraction).
Much of theconfusion over Henry's constants arises from the wide
and sometimes awkwardrange of units which can be used to express
the parameter. For consistency inthis report the De Waard equation
for an aqueous solution can be rewritten inorder to maintain KH in
units of (bar/mole fraction)..
where KH is Henry's constant (bar/mole fraction)
log10 KH =1088.76T + 273
- 5.113
log10 KH = -1088.76T + 273
- 5.113
181000
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
17
The true Henry's constant is a function of both salinity and
temperature(Appendix 1) so that,
Therefore, the salt-correction factor, Fsalt, is,
The best way to use Fsalt is to apply it to fCO2 to give an
"effective CO2 fugacity".This "effective fCO2" will give the
correct dissolved CO2 concentration whenused with the other
equations in the Cassandra 98 model. The salt correctioneffect only
becomes significant for TDS > 10% w/v.
pH ...brine pH control parameter INPUT cell F17
Enter the known pH value, or else enter a letter to accept one
of the calculatedpH values given in cells F18, F19, or F20
m "d" or "D" will accept the De Waard distilled water pH
m "o" or "O" will accept the Oddo & Tomson brine pH
m "x" or "X" will accept the BP XLpH calculated value.
The accepted value is displayed in cell F21 for
confirmation.
KHtrue (for 0 - 125 C) = (1.77 T + 47.1)
TDS10000
+ (45.2 T + 559)
KHtrue (for 125 - 200 C) = 250
TDS10000
+ 6500
Fsalt =KH
De Waard
KHtrue
Brine pH
-
18
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
When doing a probabilistic calculation using CRYSTAL BALL then a
numericvalue of pH (either known or calculated) must be entered.
Use a normaldistribution for the probability adjusting the standard
deviation so as to coverappropriate minima and maxima.
pH(CO 2)...pH of distilled water containing CO2 OUTPUT cell
F18
Equation (8) from the 1995 paper...
pH(CO2) = 3.82 + 0.000384 T - 0.5 log10 (fCO2)
fCO2 is used here rather than the pCO2 quoted in the original
paper. Theequation is valid over 10-80oC. It gives the pH for pure
water containingdissolved CO2 at the prevailing temperature and
fCO2.
pH(act, Oddo) ..Oddo & Tomson calculated pH in brine OUTPUT
cell F19
An empirical equation from reference 9...
+0.000000458 (T * 9/5 * 32)2 - 0.0000307 (P * 14.5)...
fCO2 is used here rather than the pCO2 quoted in the original
paper. Theequation is valid up to 200oC and 1200 bar, but is
inaccurate for low values of[HCO3
-]. The Cassandra 98 spreadsheet is set to give an error for
pH(act, Oddo)if [HCO3
-] < 50 ppm.
pH(CO2)
pH(act)
- 0.477TDS
58500
1 / 2
+ 0.193TDS
58500
pH = log10HCO3
-[ ]fCO2 *14.5 *61000
+ 8.68 + 0.00405(T * 9 / 5 * 32)...
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
19
pH(act, XLpH) ...XLpH calculated pH in brine OUTPUT cell F20
XLpH is an Excel add-in function for calculating both pure water
and brine pHswith no restrictions on salinities or component
concentrations. It was developedby XTP, Sunbury using well
documented code published by the US GeologicalSurvey (the "PHREEQ"
model). The original version of XLpH [10] has since beenupdated to
include pH2S as an input parameter. XLpH has been validated
againstother pH models such as in CORMED and also against
literature and recentlaboratory values.
XLpH uses the individual ion concentrations in cells A15-L15.
The positive andnegative charges must be approximately balanced
(see "water chemistry", p15,above). XLpH will automatically
compensate for any small misbalances by addingNa+ or Cl- ions.
Enter the sum of all organic acids as acetate. Note that the pH
of CO2-containing-brine will differ depending on whether the
acetate is added in the form of sodiumacetate salt or acetic
acid...
pH of 0.5 M NaCl / 300 ppm NaHCO3, 1 bar CO2, 25oC plus...
no acetate 6.8 mM Na acetate 6.8 mM acetic acid(i.e. 571 ppm)
(i.e. 422 ppm)
5.53 5.41 4.17
XLpH assumes that the acetate value entered in cell K15 is
acetic acid, becausethis is the worst case. If one wishes to assume
Na acetate then zero should beentered for Ac and the molar
equivalent of Na acetate should be added to the Naand Cl entries.
Unfortunately a field water analysis will not directly
revealwhether Na acetate or acetic acid should be used to simulate
the water chemistry.This can only be established by making
laboratory pH measurements under CO2saturation and comparing the
results with the XLpH model.
Inclusion of the organic acid concentration will always improve
the reliability ofa prediction. However, when organic acid data is
not available it is possible tomake some rule-of-thumb
approximations in order to aid progress. Organic acidsare typically
present in formation water at 150ppm, the presence of organic acids
is likely to make little difference to thecalculated pH and
therefore corrosion rate. In such cases, an API water
analysis(which omits organic acids) will often suffice. If the
formation water is low inbicarbonate (
-
20
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
accepted pH ...confirmation of selected pH OUTPUT cell F21
This is confirmation of the pH value which has been accepted for
the corrosionprediction equations.
T...temperature (oC) INPUT cell F24
The prevailing local temperature. When doing a probabilistic
calculation usingCRYSTAL BALL then use a uniform distribution for
the temperature : set F24 asthe maximum and G24 as the minimum.
Ts...selected scaling T (oC) INPUT cell F25
Enter a preferred value for the scaling temperature or enter "a"
(or "A") to acceptthe calculated value shown in cell F26.
Researchers are still actively investigating the issue of what
happens tocorrosion rates at temperatures above the scaling
temperature. Previous workhas shown that sometimes the scale films
are protective and can reduce thecorrosion rate, whereas sometimes
the films are non-protective so that thecorrosion rate continues to
increase. Choosing one or other of these optionscould on the one
hand lead to significant under-design, and on the other handto
significant over-design. Therefore, until the matter is fully
resolved BPprefers to choose a middle course for design purposes.
BP assumes that thecorrosion rate reaches a peak at the scaling
temperature and remains on aplateau at the same value for higher
temperatures. The Cassandra 98spreadsheet follows this approach. In
order to achieve this outcome both fCO2and pH are set to a plateau
for T > Ts.
T
Scaling T
IFE, Norwaydata
BP approach
De Waardapproach
Ts
Corrosion Rate
Temperature
Accepted pH
Figure 5: The PossibleEffects of HighTemperature Scaling onthe
Corrosion Rate
-
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
21
Ts...De Waard calculated scaling T (oC) OUTPUT cell F26
Equation (13) from the 1995 paper,
This is obtained by setting log10 Fscale = 0 (i.e. Fscale = 1)
in equation (13) in the1995 paper. Note that the equation above is
expressed in oC and uses fCO2 ratherthan the oF and pCO2 used in
the paper. The 1993 paper gives a similar equationto the 1995 paper
but uses a factor of 0.67 in front of the log term instead of
0.44.
d...hydraulic diameter (m) INPUT cell M24
A diameter input value is only required for the velocity
equations in the 1995model. It is not needed for the 1993 model.
The 1995 paper actually uses"hydraulic diameter" rather than a
simple pipeline diameter. Let Dp be pipelinediameter, and let Dh be
hydraulic diameter, then,
..for gas/liquid pipelines, Dh < Dp
Dh = 4 A / S
..where A is the cross-sectional area of the liquid in the pipeS
is the cross-sectional perimeter length of the liquid region (i.e.
liquid/pipe + liquid/gas interfaces, see Figure 6)
..therefore for a pipeline full of liquid, Dh = Dp
De Waard CalculatedScaling Temperature
Ts =2400
6.7 - 0.44log10fCO2
- 273
Diameter
-
22
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
There is a box at cell P39 for calculating hydraulic diameters
in gas/liquid lines.The ratio of the liquid and gas cross-sectional
areas, or the ratio of the liquiddepth to the pipe radius, is
required as an input parameter. Calculation of thisparameter is
outside the scope of the Cassandra 98 spreadsheet.
When doing a probabilistic calculation using CRYSTAL BALL then
use a uniformdistribution for the hydraulic diameter : set M24 as
the maximum and N24 asthe minimum.
U...flow velocity (m/s) INPUT cell M25
A flow velocity input value is only required for the velocity
equations in the1995 model. It is not needed for the 1993 model.
There is a box at cell P5 whichenables calculation of flow velocity
from pipe diameter and flow in liquid onlylines. The calculation is
more complicated for the liquid phase in gas/liquidlines,
therefore, the box at cell P39 should be used.
When doing a probabilistic calculation using CRYSTAL BALL then
use a uniformdistribution for the flow velocity : set M25 as the
maximum and N25 as theminimum.
cross-sectional perimeter lengthof the liquid region
Flow Velocity
Figure 6: Explanationof Parameter "S" in aGas/Liquid System
-
23
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
Vcor ...basic corrosion rate (mm/yr.) OUTPUT cell E32
Equation (13) from the 1993 paper,
The basic corrosion rate is adjusted by multiplying with the pH
and occasionallythe glycol correction factors (FpH and Fglyc
respectively). The application of eachof these is discussed
below.
For the basic corrosion rate and the correction factors, the
values reached at thescaling temperature are set to remain the same
at higher temperatures. This isto ensure that the corrosion rate
reaches a peak at the scaling temperature andthen remains on a
plateau at the same value for higher temperatures (see Tssection
above). Hence, the BP approach does take account of scaling at
hightemperatures but doesn't use the De Waard scaling factor,
Fscale, directly.
FpH ...pH correction factor OUTPUT cell G32
Equations (9) and (10) from the 1991 paper,
log10 FpH = 0.32 (pHCO2 - pHact)
for pHCO2 > pHact
where ...pHact is the actual pH of the brine which wets the
pipewall...pHCO2 is the pH under the same conditions but in
pure,
salt-free water
log10 FpH = - 0.13 (pHact - pHCO2)1.6
for pHCO2 < pHact
These equations show that as pHact rises, FpH will get smaller
and so thecorrosion rate will fall.
Outputs : 1993 De Waard Model
Vcor...BasicCorrosion Rate
log10 Vcor = 7.96 -1710
T- 0.67 log10 (fCO2)
pH CorrectionFactor
-
24
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
These equations use pHCO2 instead of the "pHsat" used in the De
Waard paper.pHsat is the pH at which a brine first becomes
saturated with either FeCO3 orFe3O4 as a result of the steel
corroding and building up dissolved Fe2+ in thesolution. The
problem with pHsat is that it is difficult to define. Even the
DeWaard paper only gives some approximate expressions for one
particular brinecomposition (10% NaCl). Furthermore, there is
serious doubt over the wholeconcept of a fixed saturation pH due to
the observation of massivesupersaturation effects by IFE (Norway)
and also within BP. Dissolved Fe2+
concentrations can often reach hundreds of ppm and can exceed
the theoreticalsaturation values by orders of magnitude. Hence,
pHsat is not a reliableconcept.
Until the pHsat issue is resolved BP prefer to use pHCO2 as an
alternativereference point. It has the advantage that it is well
defined and is valid over awide range of conditions. Therefore, a
pure water system will give pHact =pHCO2 and so FpH = 1 in the BP
approach. Certain conditions can make pHact< pHCO2 (e.g. high
salinity, zero bicarbonate) and so FpH > 1. The presence
ofbicarbonate will tend to make pHact > pHCO2 and so FpH <
1.
One way of reconciling these divergent approaches is to say that
the direct DeWaard approach uses Fph to derive the initial
corrosion rate in a brine beforecorrosion products build up and
gradually reduce the corrosion rate until itreaches a steady state.
This is the issue discussed in the 1993 De Waard paper.The BP
approach on the other hand does not deal with initial corrosion
ratesat all. It deals only with steady state corrosion rates and
uses Fph to express theeffect of water composition on the steady
state rate. This effect is not coveredin the direct De Waard
approach. In essence BP have taken an equation fromthe direct De
Waard approach and then adapted it for another purpose.
Hence,overall, the two approaches are different but consistent.
Ffug ...fugacity correction factor OUTPUT cell J32
Equation (3) from the 1991 paper,
Fugacity CorrectionFactor
log10 Ffug = 0.67 0.0031 -1.4
T + 273
P
-
25
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
Ffug is not required in the BP approach because fCO2 is used in
preference topCO2 throughout the calculation and so fugacity has
already been accounted for.
Fscale ...scaling correction factor OUTPUT cell K32
Equation (16) from the 1993 paper,
where ... T > Ts otherwise Fscale = 1... Tscale is scaling
temperature (defined above)
This factor is not used directly in the BP approach. It is
included in thespreadsheet only for completeness.
Fglyc ...glycol correction factor OUTPUT CELL H32
Equation (20) from the 1993 paper,
log10 Fglyc = A (log10 W - 2)
where ... A is a constant = 1.6 to a first approximation... W is
water content (%) of water/glycol mixture
BP only use this factor for cases without corrosion inhibitor.
When a corrosioninhibitor chemical is used or is planned then BP
assume that any effect of glycolis included within the corrosion
inhibitor efficiency (normally 90%, but seediscussion on pages
42-48).
V'cor ...corrected corrosion rate (mm/yr.) OUTPUT cell G34
This is BP's preferred output from the 1993 DeWaard model. It is
the basecorrosion rate multiplied by the FpH correction factor.
Note that for the basiccorrosion rate and the correction factor,
the values reached at the scalingtemperature are set to remain the
same at higher temperatures. This is to ensurethat the corrosion
rate reaches a peak at the scaling temperature and thenremains on a
plateau at the same value for higher temperatures (see T(s)
sectionabove). Hence, the BP approach does take account of scaling
effects at hightemperatures but doesn't use the De Waard scaling
factor, Fscale, directly.
log10 Fscale = 24001
T + 273-
1Tscale + 273
Glycol CorrectionFactor
CorrectedCorrosion Rate
Scaling CorrectionFactor
-
26
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
The 1995 De Waard model is derived in a different fashion from
the 1993 model,in particular it does not use the idea of correction
factors applied to a basecorrosion rate. Instead, the overall
corrosion rate is calculated from twocomponents : the reaction rate
Vr and the mass transfer rate Vm.
Vr ...reaction rate (mm/yr.) OUTPUT cell G37
Equation (11) from the 1995 paper,
Vm ...mass transfer rate (mm/yr.) OUTPUT cell G38
Equation (10b) from the 1995 paper,
Vcor ...corrosion rate (mm/yr.) OUTPUT cell G39
Equation (2) from the 1995 paper,
where Vcor is overall corrosion rateVr is reaction rateVm is
mass transfer rate
Outputs : 1995 De Waard Model
Reaction Rate
log10 Vr = 6.23 -1119
T + 273+ 0.0013 T + 0.41log10 (fCO2 ) - 0.34pH act
Mass Transfer Rate
Vm = 2.45U0.8
d0.2fCO2
Overall CorrosionRate
1Vcor
= 1Vr+
1Vm
-
27
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
Vcor ...merged corrosion rate (mm/yr.) OUTPUT G41
The merged rate simply takes the average of the 1993 and 1995
values. Thisallows CRYSTAL BALL to combine the probability
distributions for the 1993 and1995 rates so that one can see the
lower and upper bounds on the expectedcorrosion rate.
The 1993 rate is regarded as the minimum. Velocity effects may
increase thisminimum rate as given by the 1995 value. The 1995
model is not accurate at lowvelocities so it is ignored whenever it
falls below the 1993 value, and then themerged rate is the same as
the 1993 rate.
The validity of any corrosion prediction model depends on how
well it agreeswith the measured corrosion rates in the field.
However, the comparison is notalways straightforward. This is
because the models are developed from wellcharacterised, clean and
stable systems in the laboratory, and they are beingapplied to
partially characterised, dirty, and variable systems in the field
wherethe full operating history is not always known. This is no
criticism of fieldactivities. It is simply a fact of life of
operations where the aim is to producehydrocarbons, not to generate
completely rigorous corrosion data.
The discrepancies between the models and r eal field corrosion
data which doexist arise because there are parameters in the field
which the model can not takeaccount of effectively, or at all, e.g.
surface coatings (scales, corrosion products,biomass), crude oil
wetting, local hydrodynamics, weld metallurgy.
The industry generally regards the De Waard model as
conservative compared tothe field, i.e. it over-estimates the field
corrosion rate. Much of this opinion isbased on anecdotal and
semi-quantitative evidence - often not published in theopen
literature - but it is confirmed by the occasional formal
presentation [12].
1993 & 1995Merged CorrosionRate
Vcormerged =
Vcor1993 + Vcor
1995
2
COMPARING OUTPUT FROM THE Cassandra 98 MODEL WITH FIELD DATA
-
28
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
BP is currently compiling a database of field corrosion data
from a variety ofsources which will be used to assess the Cassandra
98 spreadsheet presentedhere.
In the meantime Table 7 gives a comparison of the Cassandra 98
spreadsheetagainst new laboratory data; data which were not used in
compiling themodel. The final column shows whether the observed
corrosion rate falls within15% of the range encompassed by the 1993
and 1995 models and there is someagreement. However, the
discrepancies show the pitfalls in trying to push theaccuracy of
the model too far. It is best used to gain order of
magnitudeestimates of corrosive situations rather than absolute
corrosion rates to severaldecimal places.
Table 7: Comparisonof Model Predictionswith Laboratory Data
BP 1993 0.1% NaCl, 3 litre flow loop (15 mm ID)25 1.9 1 5 1.1
5.8 yes25 1.9 0.27 2.2 0.5 1.9 yes35 1.9 0.27 3.4 0.7 2 noBP 1992
Forties brine, beaker test and 5 litre flow loop (15 mm ID)50 0
0.88 2.5 1.5 0.1 no50 1.2 0.88 2.5 1.5 3.2 yesCAPCIS Flow Project
Forties brine, flow loop (10 mm ID)25 3.2 1 1.8 0.6 3.3 yes50 1.1
0.88 3.8 1.5 3.2 yes50 1.7 0.88 4.1 1.5 3.9 yes50 2.5 0.88 2.5 1.5
4.4 yes50 3.2 0.88 4 1.5 4.7 yesCAPCIS Flow Project 3% NaCl, flow
loop (10 mm ID)25 3.2 1 6 1.2 7.7 yes50 3.2 0.88 12.1 3.1 9.2 no70
3.2 0.88 17.4 5.3 8.4 no50 1.1 0.88 6.8 3.1 4.8 no50 1.7 0.88 7.3
3.1 6.4 yes50 2.5 0.88 8.6 3.1 8.1 yes
corrosion rate (mm/yr.)T U fCO2 observed 93 95 correct?(oC)
(m/s) (bar) model model
-
29
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
"Henry's Law" describes the solubility of a gas in a liquid,
pCO2 = KH XCO2
where KH is Henry's constant (bar/mole fraction)XCO2 is mole
fraction of CO2 dissolved in liquid
Henry's constants are dependant on both temperature and salinity
and they areeasily found for CO2 dissolved in pure water [e.g. 13].
The data for brines is lessextensive [14-16]. Figure 7 is compiled
using data from all these sources. Thereduced number of points at
higher salinity are still sufficient to show that thedata in the
0-10% region can be reliably extrapolated up to ca 30% NaCl.
Notethat the 16 and 31% data at 75 and 100oC are actually for MgCl2
in the originalpaper but have been plotted in Figure 7 at the
equivalent ionic strength of NaCl.
APPENDIX 1 : "Henry's Law" Constants for CO2 Dissolved in
Brine
0
2000
4000
6000
8000
10000
12000
14000
0 5 10 15 20 25 30 35
[NaCl] %w/w
Kh (
bar/
mol fr
ac)
20017515012510075503010
T (oC)
The lines in this figure can be represented by the following
equations (to within15%),
Figure 7: Henry'sLaw Constants as aFunction of Salinity
-
30
"CASSANDRA 98" CORROSION PREDICTION SPREADSHEET
where KH is Henry's constant (bar/mole fraction)
Cell AD31 in the spreadsheet uses these equations to calculate
the true Henry'sconstant for the input values of T and TDS.
KH (for 0 - 125 C) = (1.77 T + 47.1)TDS
10000+ (45.2T + 559 )
KH (for 125 - 200 C)250TDS
10000+ 6500
-
31
The Use of Corrosion PredictionModels During Design by D M E
Paisley
The value and purpose of predictive corrosion rate models should
be neitheroverlooked nor exaggerated. The models (of which CO2
models are oneexample) are tools for the Materials Engineer to use
during materials selectionstudies. The models help to quantify the
corrosion risk and to help assess theimpact of various process or
production scenarios. However, corrosion rateprediction models
should always be used in conjunction with other tools such aslife
cycle costing as well as previous operational experience if the
final materialsselection is to offer the optimal balance between
cost and reliability. As eachproject will have unique economic
factors, materials selection should reflect theseand the economic
assessment will be as important as the corrosion modelling inthe
selection of the final materials. In-depth coverage of techniques
such as lifecycle costing and estimating values are beyond the
scope of this document butboth techniques are briefly covered in a
previous publication [17].
Over the past few years, several design guidelines have been
issued by BP fordealing with CO2 corrosion risks. Each document
deals with a specificapplication. This more general document
summarises all previous guidelines butcan not deal with the
specific issues to the level of detail possible in theindividual
guidelines. The previously issued guidelines are listed in Table
8.
Introduction
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
32
Table 8: PreviouslyIssued DesignGuidelines
A corrosion philosophy for the transport of wet oil and
multiphasefluids containing CO2
This was the first undertaking in recent years to document a BP
approach todefining internal corrosion risks and the basic approach
is still followed. Itrecommended the use of the de Waard and
Milliams model to predict in-situcorrosion rates along with a 90%
corrosion inhibitor efficiency. Much of thework is still valid but
it is in the areas of high temperature scaling, corrosioninhibitor
efficiencies and impact of various flow regimes that the new
guidelines
Report Title Authors Report Number Issue Date
A corrosion philosophy for the I D Parker ESR.93.ER.013
1/3/93transport of wet oil and J Pattinsonmultiphase fluids
containing A S Green.CO2
A corrosion philosophy for I D Parker ESR.94.ER.016 28/8/94the
transport of wet J Pattinsonhydrocarbon gas containing A S
Green.CO2
Assessment of a top of line D Paisley Branch Report
5/10/92versus bottom of line corrosion J Pattinson No 124 421ratio
for use in the design of S Websterwet natural gas pipelines
The application of pH D Paisley ESR.95.ER.042 10/4/95moderation
as a means of corrosion control for wet gas pipelines
The effects of low levels of D Paisley ESR.95.ER.073
22/6/95hydrogen sulphide on carbon R Gourdindioxide corrosion: A
review of industry practice and a guideto predicting corrosion
rates
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
33
differ. Most of the recommendations made in these guidelines
have beenreproduced or superseded in the present document and
therefore the originalguidelines are redundant.
A corrosion philosophy for the transport of wet hydrocarbon
gascontaining CO2
This was a companion document to the guidelines on wet oil and
multiphasesystems. The basic approach was similar but this document
dealt with thespecific wet gas application. Most of the
recommendations made in theseguidelines have been reproduced or
superseded in the present document andtherefore the original
guidelines are redundant.
Assessment of a top of line versus bottom of line corrosion
ratio for usein the design of wet natural gas pipelines
Wet natural gas pipelines operating under stratified flow have
two distinctcorrosion environments : (a) the bottom of line which
is continually wetted bycondensed water, hydrate inhibitor and
hydrocarbons, and (b) the top of linewhich is wetted intermittently
by condensing liquids. The corrosion rate at thetop of the line is
lower than that at the bottom due to the more limited exposureto
corrosive species. Predicting this rate is done by predicting the
bottom of linerate using models in the normal way and applying a
moderating factor for thetop of line rate. Up to 1992, BP used a
factor of 0.3, i.e. the top of line corrosionrate was predicted to
be 30% of the bottom of line rate. When inhibitors areused to
control the bottom of line rate, the top of line corrosion rate
becomesthe limiting rate as inhibitors are assumed not to protect
against condensingcorrosion. This report reviewed the top of line
factor and recommended theadoption of a moderating factor of 0.1.
For inhibitor efficiencies up to 90%, thetop of line corrosion rate
is therefore not the limiting rate. This approach is nolonger valid
since BP have moved away from the direct use of
inhibitorefficiencies, as described later in this report. However,
the assumption that topof line rates are 1/10th of the predicted
uninhibited bottom of line rates can stillbe used. For applications
were the 'top of line' corrosion rate is the faster rate(using the
0.1 moderating factor) then a more detailed evaluation should
becarried out. Such a scenario does not lend itself to the use of
simplifiedguidelines.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
34
The application of pH moderation as a means of corrosion control
forwet gas pipelines
This technique is not widely applicable but may find niche
applications inhighly corrosive wet gas lines utilising recycled
glycol for hydrate control. It iscovered in more detail on p75 but
if this technique is of interest the fullguideline document should
be reviewed.
The effects of low levels of hydrogen sulphide on carbon
dioxidecorrosion: A review of industry practice and a guide to
predictingcorrosion rates
This document summarised how low levels of H2S influence
corrosion ratesdominated by CO2. The conclusion was that H2S at
levels below the NACEcriteria for sulphide stress corrosion
cracking (ref MR0175, NACE Publications)reduces general metal loss
rates but can promote pitting. The pitting proceedsat a rate
determined by the CO2 partial pressure and therefore
CO2-basedmodels are still applicable at low levels of H2S. Where
the H2S concentration isgreater or equal to the CO2 value, or
greater than 1 mole%, the corrosionmechanism may not be controlled
by the CO2 and therefore CO2 based modelsmay not be
appropriate.
Summary of Previous Guidelines
In summary, the old guidelines are generally still applicable.
What has changedis BPs views on the reliability and performance of
corrosion inhibitors as wellas the availability of updated models
incorporating flow affects. The oldguidelines defined a corrosion
inhibitor efficiency of 90% with no scope forvariation. There were
also stringent velocity restrictions for use undermultiphase
conditions which restricted the energy of slug flow to below 20
Pa,later raised to 100 Pa. In light of favourable field data, this
approach is nowseen as too pedantic and inhibitor availabilities
are seen as a better way ofdescribing the role of inhibitors. These
differences in approach are covered inmore detail in the following
sections. Furthermore, the corrosion rateprediction model (p5-30)
does not cover some aspects that are important duringdesign and
these are covered in the next section.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
35
The modelling approach outlined in this document deals with all
the inputs(mole% CO2, temperature etc.) on a deterministic basis.
However, each inputwill have a level of uncertainty associated with
it and this can have importanteffects on the outcome. One way to
deal with this it to calculate a range ofoutput values, (in this
case the predicted corrosion rate) across the whole rangeof input
values. Where the model is dealing with several inputs
(temperature,pressure, CO2 mole %, pH, scaling factor), this can be
time consuming. Also, thevalue of these inputs will not all vary in
a uniform manner. Some will behaveuniformly while others may behave
in a normal or log-normal manner.
Calculating the impact of all these variables is time consuming,
unless aprogramme such as Crystal Ball is used. This is an add-in
to Excel and handlesthe variability by performing a Monte Carlo
analysis. Any number of iterationscan be performed and the output
is displayed in terms of a probability, ratherthan as a discreet
value. In general, a minimum of 1,000 iterations, involving tensof
thousands of individual calculations are required to show the
effects of thevariability in input data. A modern PC can perform
such a task in a minute ortwo.
The important factors to consider are the range and type of
distribution assumedfor each variable. If process data are
available, this will form an ideal basis fordetermining the range
and type of distribution but if this is lacking, someassumptions
will have to be made.
Using distributions to define variables in a predictive model
can have significanteffects on the outcome.
Engineering design traditionally uses worst case inputs so that
the final designwill be safe under all foreseeable combinations of
events. This approach hasalso been adopted when predicting
corrosion rates, where pressure andtemperature etc. are used as
inputs to the models. In the past this approach wasthe only viable
one as predicting the enormous range of possible outcomes forall
variables would have been too time consuming but it can result in
substantialover-design. Metal loss corrosion processes do not lead
to sudden failure due toa combination of variables over short time
periods (unlike high pressure whichcan lead to an instantaneous
failure) but rather reflect a combination of varying
Worst Case Design
Important Factors not Covered by the Corrosion Model
The ProbabilisticApproach toPredictive Modelling
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
36
conditions over a longer time period. Using the worst case
values is thereforenot a sensible approach, if a range of more
realistic values can be handled.
In defining a range of likely operating variables such as
temperatures andpressures, the design values will form the maximum
for the respectivedistributions but lower values should be
included. Defining this range willrequire inputs from the Process
and Reservoir Engineers. Due to the nature ofthe uncertainty, such
that all values within the range are as likely as each
other,Uniform distributions are probably the most appropriate for
these variables.
The yield strength and wall thickness of linepipe are other
examples of the typeof variables that can be treated in this
manner. The linepipe properties areimportant if using corrosion
models to calculate mean time to failure. Ratherthan using the
minimum values for each, based on the specified material andthe
variation allowed within the specification, typical distributions
can bedefined for each value. Such variables tend to be distributed
normally arounda mean with the specified minimum properties
defining a lower bound.
Many variables in corrosion rate predictions, such as the level
of CO2 in the gasphase, are based on best guess or on limited well
test data. No attempt ismade to define the uncertainty in these
data and this is a major limitation ofdeterministic modelling. In
defining the distributions of such variables, themean value should
be based on the best guess or well test data in a similar wayto the
deterministic approach. However, a range of possible values should
beconsidered. In the absence of any other information, the
distribution of valuesis likely to be symmetrical around the mean
with the greatest probabilityassociated with values close to the
mean. The Normal distribution is a familiarexample of this type and
should be used.
It should be noted that using a symmetrical distribution, such
as a Normaldistibution, does not correspond to using a single value
equal to the mean ifthe variable under consideration has a
non-linear relationship with the outcome.For example, the corrosion
rate prediction model used by BP states that:
Non-LinearRelationships
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
37
Therefore, the corrosion rates associated with the CO2 partial
pressure values inthe Normal distribution that are greater than the
mean value are closer to themean corrosion rate than those
associated with the values below the mean CO2partial pressure. In
other words, defining symmetrical distributions for variableswhose
influence is described by a power < 1 produces a
non-symmetricaldistribution of outcomes (predicted corrosion
rates). The mean value of thisdistribution will be lower than the
single value calculated using the mean of theinput variable.
The same applies to all symmetrical distributions, including
Uniformdistributions. In the previous section on 'worst case
design', the uncertaintiesregarding operating temperature and
pressure were discussed. In both cases,Uniform Distributions were
used to define the range of possible values. Incorrosion rate
modelling, both these inputs have non-linear relationships withthe
outcome (predicted corrosion rate). The effect of pressure is
moderated bya fugacity coefficient related to the non-ideality of
CO2. Therefore, consideringa range of pressures distributed
symmetrically around a mean value will tend toreduce the predicted
corrosion rate.
The effect of temperature on predicted corrosion rates is
strongly non-linear. Athigher temperatures, the role of protective
corrosion products or scales can beimportant. There is a great deal
of uncertainty in the effects of these scales butthe bounds of the
expected values can be defined using existing models. Oneapproach
would be to use a log normal distribution, defined as follows:
1. The de Waard & Milliams unscaled rate (upper bound), 2.
The de Waard & Milliams fully scaled rate (lower bound), 3. A
modal value equivalent to the standard BP approach that uses the
scaling
temperature to calculate the corrosion rate for all temperatures
above this.
Again, the outcome of considering a range of temperatures
symmetricallydistributed around a mean will tend to be a lower
corrosion rate estimation thanfound by calculating a single value
at the mean temperature.
Each input into a corrosion rate prediction should be considered
and a range ofpossible outcomes defined. By consideration of the
way in which the value mayvary in practice, a distribution function
can also be defined. This may have tobe done subjectively but the
following basic rules offer some guidance. In thefollowing
examples, distributions are shown that have been used in the
CrystalBall software.
Summary of Inputsto a Monte CarloAnalysis
Corrosion Rate CO2 partialpressure( ) 0.67
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
38
1. Where variations would be due to nature, such as the
difference in CO2levels around the field, a Normal Distribution
should be used with a meanequivalent to the best guess. Figure 7
shows an example of a NormalDistribution describing the expected
variation in CO2 levels, centredaround a mean of 5%.
Figure 7: An Exampleof a NormalDistribution for theconcentration
of CO2in a gas. The MeanValue is 5 mole% with arange of 3 to 7
mole%.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
39
2. Where an input may vary over a wide range but would be
expected to beskewed around the 'best guess' or predicted value, a
Log NormalDistribution should be used. The effects of high
temperature scalingwould be an example of this type of
distribution, or the pit depth at whichinhibitors fail to control
corrosion. Figure 8 shows the Log NormalDistribution used to
describe the critical pit depth with a modal value of 8mm and a
range of 5 to 12mm.
Figure 8: An Exampleof a Log NormalDistribution describingthe
critical pit depth.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
40
Figure 9: An Exampleof a UniformDistribution Describingthe
Flowline OperatingPressure
3. Where a value may occur equally often within the defined
range e.gflowline operating pressure, a Uniform Distribution should
be used, i.e.all values are equally likely to occur. Figure 9 shows
how a range offlowline operating pressures can be described. In
this case the range of1,000 to 1,200 psi has been used.
Table 9 summarises the assumptions used in a recent
probabilistic study intomean time to failure, based on CO2
corrosion risks. As the study looked atfailure mechanisms as well
as corrosion rates, some of the factors apply to thelinepipe steel
while others apply to the CO2 prediction model. The 'StandardValue'
corresponds to the value that would be used in a deterministic
study.The Table does not attempt to fully define the distributions
in a statistical sensebut more information is available from the
authors if required.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
41
Linepipe Wall thickness e.g. 0.75" Mean = 0.75" NormalSD =
0.01
Linepipe Yield Stress SMYS Mean = 70 ksi Normale.g. 65 ksi SD =
2.5 ksi
Linepipe Flow Stress - - - - 1.15 x Yield Stress Normal
Fluids CO2 Content 5 mole% Mean = 5% NormalSD = 0.72
Fluids Temperature 110oC 85 - 110oC UniformFluids Pressure 1,200
psi 1,000 - 1,200 psi Uniform
Corrosion Water pH Cormed * Cormed * Normalmodel prediction 0.25
unitsCorrosion Corrosion rate >Rate at scaling Unscaled to Log
Normalmodel scaling ToC temperature fully scaled
Inhibitor Inhibitor 90% 65 - 95% Log Normalefficiency
availabilityInhibitor Critical pit depth 8 mm 5 - 12 mm Log
NormalefficiencyInhibitor Inhib. effic. > 0% 0 - 90%
Uniformefficiency critical pit depth
Table 9: Summary ofVariables Modelled,the Values that wouldbe
Assigned Using aStandard Approach,and the Range ofValues Used in
theExample Study
Component Variable 'Standard Range Used Distributionin study
Value'
Note * Cormed is a software programme which can predict in-situ
pH values ofoilfield brines.
Figure 10 shows the output from a Monte Carlo simulation, using
20,000iterations to determine the distribution in outcomes
(predicted corrosion rate)due to the variation in inputs detailed
above. The most likely corrosion rate iscirca 1 mm/yr. While there
is a possibility that higher or lower rates occur, theprobability
of such rates decreases the further they are from the most
likelyoutcome.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
42
This section represents a significant shift from previous BP
recommendationsand therefore is covered in some detail.
The guidelines on the reliance to be placed on corrosion
inhibitors presentedhere have been based on experience gained with
continuous injection systems.The success of batch treatments with
corrosion inhibitor is less welldocumented and generally this
approach to corrosion control is less reliable.These guidelines
should therefore not be used when designing systems that willbe
protected with batch treatments - this effectively rules out their
use forthe majority of downhole applications. Instead, it is
recommended thatrelevant operational experience with batch
treatments is sought beforedesigning on the basis of batch
inhibition. The authors will be able to assist insourcing relevant
operational experience.
Previous BP guidelines have dealt with the affect of corrosion
inhibitors on CO2corrosion by assigning a corrosion inhibitor
efficiency. This described theextent to which an inhibitor reduced
the predicted rate and a figure of 85% wasoriginally used, later
raised to 90%. This was despite laboratory observationsthat showed
inhibitors could reduce corrosion rates by 95% or more. However,it
was accepted that in the field, inhibitor is not delivered at the
recommendeddose rate for 100% of the time and therefore a degree of
conservatism isnecessary when estimating the benefits of
inhibitors.
Frequency Chart
mm/yr
.000
.028
.057
.085
.113
0
565
2260
0.00 1.13 2.25 3.38 4.50
20,000 Trials 313 Outliers
Forecast: Predicted Corrosion RatesFigure 10: TypicalOutcome of
the BPCorrosion Rate ModelRun Using aProbabilistic Approach
Effect of Corrosion Inhibitors
Inhibited CorrosionRates
Applicability of theGuidelines
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
43
One major limitation with inhibitor efficiencies is that it
allows no considerationof the effects due to increased dose rates
or the development of better chemicals.It is well known that
increasing the dose rates of corrosion inhibitors up to acertain
level reduces the corrosion rate. Figure 11 shows the
relationshipbetween dose rate of inhibitor and corrosion rate on
corrosion coupons atPrudhoe Bay. Clearly, the inhibitor efficiency
is not a constant value andincreasing the inhibitor concentration
(or changing the chemical for a moreefficient one) enables lower
corrosion rates to be achieved.
A second major limitation with using a single value for
corrosion inhibitorefficiencies is that they are unlikely to be
constant across the whole range of fieldconditions. CO2 corrosion
models can handle input values across a wide rangeand moderation
factors have been developed over the years to reduce
theconservatism due to the extrapolation of the data set used to
develop the model.However, no such moderation factors have been
developed for corrosioninhibitor efficiencies and by applying a
blanket efficiency, it is assumed they areconstant across the range
of applications.
BP is fortunate in having one of the more corrosive fields in
Prudhoe Bay. Thisfield also lends itself to effective corrosion
monitoring due to the use of above-ground flowlines and there is a
great deal of data on inhibited corrosion rates.There is a good
relationship between observed corrosion rate and
inhibitorconcentration, as shown in Figure 12. In this Figure, the
effect of the increaseddose rate of chemical between January 1994
and September 1996 can be seen inthe increased efficiency of the
chemical, based on the predicted corrosion ratesusing BPs CO2
corrosion rate prediction model.
1
10
100
40 50 60 70 80 90 100 110 120 130 140
Corrosion Inhibitor Concentration - ppm
1/co
rros
ion
rat
e (y
ears
per
mm
)
Figure 11: TheImprovement inPerformance of aCorrosion
Inhibitorwith IncreasingConcentration
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
44
In Figure 12 all efficiency values lie within the range 98.6%
and 99.7%,apparently extremely good performance but in January 1994
only 40% of theflowlines at PBU had acceptable rates of corrosion,
defined as corrosion ratesunder 2 mpy (0.05 mm/yr.) based on
corrosion probes - see Figure 13. Theimprovement in performance
from January 1994 onwards correlates with theincrease in average
dose rates shown in Figure 12.
0
20
40
60
80
100
120
140
Jan-94 May-94 Sep-94 Jan-95 May-95 Sep-95 Jan-96 May-96
Sep-96
Date
Ave
rage
Cor
rosi
on I
nh
ibit
or C
once
ntr
atio
n -
pp
m
98.20%
98.40%
98.60%
98.80%
99.00%
99.20%
99.40%
99.60%
99.80%
'Ave
rage
' C
orro
sion
In
hib
itor
Eff
icie
ncy
Corrosion inhibitor concentration
Corrosion inhibitor efficiency,defined using BP's model
Figure 12: TheRelationship BetweenCorrosion Inhibitor DoseRate
and ObservedEfficiency at PrudhoeBay
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
45
Prudhoe Bay was constructed before the development of the
earlier BPguidelines on CO2 corrosion, but if their flowlines were
to be constructed todayusing the same materials and corrosion
allowances, it would infer a corrosioninhibitor efficiency of
approximately 98%. As PBU have now demonstrated thatcorrosion
control of their system is possible it is clear that inhibitors can
beeffective under highly corrosive conditions. This in turn
indicates that either:
m Higher inhibitor efficiencies can be assumed in more
aggressiveconditions, or
m Corrosion inhibitor efficiencies are not the correct way to
describe the roleof inhibitors in corrosive service.
The former premise does not lend itself to design as it would
require a slidingscale of inhibitor efficiencies and the field data
is not available to allow this to beproduced. The latter is the
belief of several oil companies who do not useinhibitor
efficiencies, preferring to use a design corrosion rate for
inhibitedsystems in the range 0.1 to 0.3 mm/year. For mildly
corrosive conditions(~1.0mm/year) the use of an efficiency of 90%
generally works well. However,for highly corrosive conditions
(~10mm/year) it would result in a conservativeestimate of the
inhibited corrosion rate. This adds weight to the argument thatthe
role of corrosion inhibitors can not be described by
efficiencies.
Percentage of Production Lines with Corrosion Under Control
-100%
-80%
-60%
-40%
-20%
0%
20%
40%
60%
80%
100%
Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96
2 < CR < 5 CR >5 mpy1 < CR < 2 CR < 1 mpy<
2 mpy by Qtr
Note
Covers 3 phase productionlines >6" in diameter with
WLCsincluding LDFs, LP, HP andGHX.
Figure 13: HistoricalRecord of CorrosionRates in PBU
FlowlinesShowing ImprovingPerformance SinceJanuary 1994
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
46
BPs data indicate that inhibited corrosion rates of 0.1 mm/year
are possibleunder optimum conditions of high inhibitor dose rates
and optimised chemicals.This is confirmed with inspection data from
PBU where flowlines which havebeen effectively inhibited have
pipewall corrosion rates of less than 0.1 mm/yr.
In general, inhibitors require free and regular access to the
steel surface to beeffective. Anything that interferes with this
will reduce their effectiveness to lowor negligible levels.
Examples of low or stagnant flow situations are vessels,instrument
and drain piping and tanks. Historically, inhibitors have not
beenassumed to work well in these environments and other corrosion
controlmeasures are used, such as coatings and/or cathodic
protection.
Inhibitors also perform poorly in low velocity pipework and
pipelines,particularly if the fluids contain solids such as wax,
scale or sand. Under suchcircumstances, deposits inevitably form at
the 6 oclock position, preventingtransportation of the inhibitor to
the metal surface. Flow velocities belowapproximately 1.0 m/s
should be avoided if inhibitors are to provide
satisfactoryprotection and this will be critical in lines
containing solids.
The figure of 1.0 m/s is a rule-of-thumb which has been used in
the industryfor many years. However, it is now possible to
calculate the velocity moreaccurately, using an approach developed
by the 'Corrosion in MultiphaseSystems Centre' at Ohio University
[18]. The work agrees with the rule of thumbfor most black oil
systems but allows more accurate quantification if theminimum
velocity is restrictive.
The costs associated with corrosion inhibition are driven by the
volume ofchemical used per annum and the chemical cost. There may
be some incidentalcosts associated with the provision and
maintenance of injection equipment butincreasingly this is being
handled by the chemical suppliers and is thereforecovered by the
chemical cost.
In general, inhibitors are most attractive when protecting long
lengths ofpipeline while they are rarely cost effective when
protecting short runs ofprocess piping. The dose rates required are
dependent on factors such as liquidthroughput, CO2 partial
pressure, pH and flow regime. Dose rates are notdependent on the
length of pipeline or pipework being treated and
Operating CostsAssociated WithCorrosion Inhibition
Applications WhereInhibitors Are LessThan Fully Effective
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
47
Beatrice 40Brae 10Bruce * 46Forties Pipeline * 26Magnus 20Miller
* 35Nelson Enterprise * 17Scott Amerada Hess * 9AVERAGE 25
Table 10: Dose Ratesof Corrosion Inhibitorsinto Several North
SeaExport Pipelines,Based on Total FluidVolumes
Field Dose Rate (ppm)
Note * - These fields deploy concentrated corrosion inhibitors
to improvelogistics offshore. The quoted dose rates correspond to
the standard product,manufactured by the same supplier.
At Prudhoe Bay the field-wide average corrosion inhibitor
injection rate is 110ppm, with maximum rates of 250 ppm in certain
flowlines, based on waterproduction (typical water cuts are 50%).
These rates reflect the rapid corrosionexperienced in some PBU
flowlines in recent years.
The determination of dosage rates in gas systems is not as
straightforward as forliquid filled lines. The three methods which
are commonly used to do this are:
1. Based on Gas Flow. This is the most commonly used method and
a commonrule of thumb is to apply 1 pint of inhibitor to every 1
million standard cubicfeet of gas (1 pint/MMscf). Actual values are
found to vary enormously in therange of 2 and 0.05 pints/MMscf of
gas.
2. Based on the Water Content in the Pipe Line. This is the
method favouredby corrosion engineers as it usually indicates a
very low requirement forinhibitor. It is common to assume a dosage
of 200 ppm of chemical in thewater. This method will often give
erroneously low values, especially when the
therefore the same operating cost is incurred in protecting 10
metres of pipeworkas is required to protect 20 km of flowline.
Corrosion resistant materials arelikely to offer lower life cycle
costs for pipework while carbon steel plusinhibition tends to be
the cheapest method of constructing and operatingflowlines
[19].
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
48
water content is very low and/or the pipeline is very long. This
is becausethe volume predicted will be too low to allow a film to
be build up over theentire surface of the pipe.
3. Based on the Formation of a Protective Film. This is probably
the leastused method but one whch provides a good check on the
values obtainedfrom the first two methods. Typically it is the
volume required to form a0.05mil (1 micron) film over the entire
internal surface of the pipe. Thisvolume is then applied
continuously on a daily basis. If the product is to beapplied as a
batch treatment the volume is increased by a factor of ten
(x10).
In practice it is sensible to do all three calculations and to
use the greatestvolume as the starting point. This should hopefully
be the most conservativevolume re q u i red. Again, highly
corrosive duties associated with hightemperatures or CO2 partial
pressures will tend to require dose rates towardsthe upper end of
this scale.
Chemical costs vary from supplier to supplier and may be tied in
with theprovision of other services such as corrosion monitoring.
However, for thepurposes of life cycle costing a chemical cost of
US$8 per US gallon isreasonable. On this basis, corrosion inhibitor
costs 0.84 cents to 8.4 cents perbarrel at inhibitor dose rates of
25 to 250 ppm. There will also be costsassociated with monitoring
and inspection. These aspects are beyond the scopeof this document
but are covered in detail in SELECTING MATERIALS FORWEALTH
CREATION: A Material Selection Philosophy Based On Life CycleCosts
[17].
Due to the limitations of corrosion inhibitor efficiencies as a
design tool, theinhibitor availability model has been adopted. This
approach can be used todefine a corrosion allowance as follows:
C At o t a l = CAi n h i b i t e d (x years @ 0.1 mm/yr.) + CAu
n i n h i b i t e d (y years @ uninhibited rate)
This approach assumes that the inhibited corrosion rate is
unrelated to theuninhibited corrosivity of the system and all
systems can be inhibited to 0.1mm/year. The approach also
acknowledges that corrosion inhibitor is notavailable 100% of the
time and therefore corrosion will proceed at theuninhibited rate
for some periods.
Predicting the Effectiveness of Corrosion Inhibitors - The
Inhibitor AvailabilityModel
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
49
In the context of this model, corrosion inhibitor availability
infers the presenceof a suitable corrosion inhibitor at sufficient
concentration to reduce thecorrosion rate to 0.1 mm/yr. The factors
that lead to inhibitor availability below100% are:
m Inhibitor injection equipment is not available on Day 1 of
operations.m Injection equipment requires maintenance and repairs.m
Operators set the dose rate incorrectly.m Chemical is not available
when required.m Chemical dose rate is less than optimum. This can
be due to a variety of
reasons including lack of response to increases in throughput,
or water cutor sand rate.
m Well stimulation fluids such as hydrochloric acid are produced
along withthe crude oil and reduce corrosion inhibitor
effectiveness.
m The corrosion inhibitor injection facilities are used for
delivery of otheroilfield chemicals such as demulsifiers or
combined products such as scaleand corrosion inhibitors.
m Inhibitors are deployed via large bore pipework (instead of
via injectionquills) and are not dispersed in the flow stream for
some distance, providingpoor protection.
All of these factors and others not listed have lead to less
than optimal deliveryof corrosion inhibitor into production
equipment in BPX. No asset is immune tosuch problems and therefore
the maximum inhibitor availability that should beassumed is 95%. In
many instances, a lower availability should be assumed;
see,'Recommended Values For Use in the Inhibitor Availability
Model, pp 51.'
Words of Caution
Production data from Cusiana shows that their 12 inhibitor
injection skidsaveraged 99.2 % availability over the second half of
1996, an identical figure tothat generated at a new gas treatment
plant in the Middle East. This is probablyclose to the maximum that
inhibitor injection units can be available, bearing inmind the
requirements for chemical feedstock, power and the reliability of
thepumps. However, this should not be used as a basis for assuming
an inhibitoravailability of greater than 95%. Figure 14 shows the
delivery of corrosioninhibitor against the target rate for a North
Sea platform. There was only oneinstance when the inhibitor
injection system was not delivering chemical - duringMarch 1993 -
but there were also only 3 short periods where the chemical
wasfully available with respect to the target dose rate.
-
THE USE OF PREDICTIVE MODELS DURING DESIGN
50
At the project stage, it is difficult to determine the
availability of inhibitor infuture years but relatively easy to
ensure inhibitor is available on day one. Theprovision of chemical
injection equipment is often outside the scope of EPICcontracts and
therefore assets are brought on-stream without the
necessaryfacilities to inhibit valuable equipment. In previous
projects, this has taken upto 2 years to correct and therefore the
best inhibitor availability that can beachieved will be 90%,
assuming a 20 year design life. If the provision ofchemical
injection equipment is brought inside the scope of the EPIC
contract,measures can be taken to ensure inhibitor is available on
day 1 of operations.
Achieving good inhibitor availability during operations is
partly down to systemdesign and partly due to management of the
changing corrosion risk. Inhibitorinjection systems are simple
systems and lend themselves to high levels ofmechanical
availability. This can be improved further through the use of
lowlevel warning devices on the storage tanks and dose rate gauges
such as thesight glass or more complicated dose rate monitoring
systems. Together, thesetwo simple measures will help to ensure
that the target dose rate is achievedfor a high proportion of the
time.
Ensuring the target dose rate is correct is more difficult and
requires thatconstant changes to the target are made to reflect
changes in production rate,water cut etc In extreme cases, this may
require weekly tailoring of the targetdose rate. This is where
corrosion control programmes can fail and thereforeit is important
that the materials or corrosion engineer concentrates on
thisaspect.
100
80
60
40
20
0
January1993
March1993
May1993
July1993
September1993
November1993
January1994
March1994
May1994
Target = 50ppm
Figure 14: TheAvailability ofCorrosion Inhibitorinto a
Main-Oil-Lineover an 18 MonthPeriod
-
THE USE OF PREDICTIVE MODELS DURING