See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271714349

A Strategic Approach to Corrosion Monitoring and Corrosion Management

Article in Procedia Engineering · December 2014

DOI: 10.1016/j.proeng.2014.11.082

CITATIONS READS

14 931

1 author:

William Cox
Corrosion Management Ltd.
20 PUBLICATIONS 128 CITATIONS

SEE PROFILE

All content following this page was uploaded by William Cox on 07 January 2016.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com

ScienceDirect
Procedia Engineering 86 (2014) 567 – 575

1st International Conference on Structural Integrity, ICONS-2014

A Strategic Approach to Corrosion Monitoring and Corrosion

Management
William M. Cox
Corrosion Management Ltd, Rugby, CV22 6HL, United Kingdom
E-mail ID: [email protected]

Abstract

The introduction of modern corrosion surveillance instrumentation has opened the pathway for conventional practices to be
updated. The application of a risk-based approach to the siting of corrosion sensors, the selection of modern instrumentation that
is capable of detecting the onset and propagation of localized attack as well as uniform corrosion and instantaneous correlation of
on-line corrosion with process chemistry data as well as with periodic inspection results, presents the opportunity for both the
severity and the duration of attack to be minimized. This paper considers each of these aspects in detail and illustrates the future
prospect of plant-wide corrosion and condition management.
© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2014 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research.
Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research
Keywords: corrosion monitoring, inspection, corrosion management

1. Introduction

Established corrosion monitoring practice is based largely on the use of weight loss coupons, electrical
resistance (ER) probes and linear polarization (LPR) methods [1, 2]. There is an implicit assumption that these are
the only techniques available to evaluate metal loss due to corrosive conditions and that such instrumentation is the
‘standard’ monitoring equipment. Although the use of these techniques is well established, the design of off-the-
shelf commercial instrumentation has been dominated by their use in the oil and gas production industry, primarily
to demonstrate the effectiveness of chemical treatment products in as short a time as possible. Probe design was
constrained as much by the necessity to utilize high pressure (‘Cosasco-type’) access fittings as by the technical
requirements of the monitoring technique to be applied [3]. Furthermore, the technology is targeted primarily on the
assessment of uniform corrosion and plain carbon steel.

1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Peer-review under responsibility of the Indira Gandhi Centre for Atomic Research
doi:10.1016/j.proeng.2014.11.082
568 William M. Cox / Procedia Engineering 86 (2014) 567 – 575

In consequence, little, if any, thought is invested in thinking through why any specific location might be
monitored and what the objective of monitoring might be in the first place. Coupon probes and monitoring sensors
are scattered throughout the plant in the vague hope that they might detect damage. As the majority of probes
provide only retrospective data, the focus of monitoring is primarily on providing an indication of how much
damage has already been sustained, rather than in preventing it. Furthermore, it has been reported that
approximately 70% of corrosion sustained by process plant is localized corrosion (pitting, crevice attack, stress
corrosion cracking, corrosion fatigue, under-deposit attack, filiform corrosion, or fretting damage) [4].
Conventional corrosion instrumentation is incapable of detecting these phenomena, let alone providing a real-time
indication of the onset or propagation of damage. More recently, the introduction of techniques such as
electrochemical impedance spectroscopy (EIS), harmonic analysis (HA) and electrochemical noise (EN)
measurements have revolutionized the field of on-line corrosion monitoring. Although EN analysis is arguably the
most powerful of these, the technique has yet to make a full impact on practical real-time corrosion rate assessment,
not least because the method of application and the quality of available instrumentation has been inconsistent. The
difficulty in obtaining similar (let alone identical) results from different sources of instrumentation and/or from
different users has served to undermine confidence in what should be viewed as the likely foundation of real-time
condition management in static plant equipment.

2. Monitoring or Inspection?

There is considerable confusion about which activities constitute ‘monitoring’ and what is really
‘inspection’. On numerous occasions, plant personnel will report “we have a corrosion issue and we are monitoring
it”, whereas what is really happening is some form of inspection – normally poorly planned and inadequately
implemented. In the absence of a clear understanding of the objective, it is hardly surprising that the outcome is
unsatisfactory. Inspection is the periodic evaluation of plant condition, normally by a skilled technician, to obtain a
limited number of high-quality measurements. So, for example, if the plant is suffering from an excessively high
rate of attack, the frequency of inspection might be increased to once every six months, where formerly the item was
examined once a year, or even once every five years. By contrast, monitoring is a low-cost method of obtaining a
large volume of moderate-quality measurements. The primary purpose of monitoring, therefore, should be to obtain
a near real-time indication of the approximate rate of corrosion attack, with the objective of identifying transient
operating conditions that may lead eventually to cumulative serious damage.

In the past, some plants have employed an ‘inspection only’ strategy, on the grounds that corrosion rate
indications from on-line monitoring instrumentation are too approximate. In this situation, the plant initially obtains
a very accurate indication of the actual condition of the plant item. However, it is then effectively operated blindly
until the next inspection date, at which point a further high-quality indication is obtained to establish the new plant
condition. In the event that no corrosion has occurred, it is assumed that operational conditions are satisfactory and
the period between inspections can be extended. There are virtually no grounds for the basis of such a decision,
however, as small changes in the service environment may cause a catastrophic change in the rate of attack being
sustained by the plant. Conversely, it may be discovered that catastrophic damage has already been sustained. In
this case, the only response is to increase the frequency of inspection, in the hope of detecting damage already being
sustained before the mechanism has sufficient time to result in an unscheduled outage. It is evident, therefore, that
although inspection provides a very reliable indication of actual plant condition, it does not offer any control
whatsoever on the degree of damage being sustained by the plant.

The alternative is to turn to monitoring. In this case, the instrumentation should operate continuously with
a minimum of operator effort to provide a continuous indication of plant condition. Typically, the accuracy of
corrosion monitoring predictions of damage rates initially may not be very precise. However, the continuous flow
of information enables the operator to track the approximate rate of attack. This means that it is possible to take
immediate action in response to a sudden increase in corrosion rate that typically is related to an unexpected
excursion in the service environment. Interestingly, minor changes in the service condition can result in a dramatic
change in the rate of corrosion. However, equally minor changes to take it out of the risk condition just as rapidly.
Prompt response enables both the severity and the duration of attack to be reduced, thereby delivering a significant
increase in the service life of the plant with minimum interference by the operator. Furthermore, it has been shown
in chemical process industry applications that timely control of this nature not only reduces the rate of corrosion
 William M. Cox / Procedia Engineering 86 (2014) 567 – 575 569

attack sustained by the plant, but also reduces the degree of product contamination that results from corrosive
degradation of the plant construction materials [5].

Perhaps the best parallel to the use of a sensitive real-time corrosion indication is to the experience of
driving a car with a real-time indication of fuel consumption. Any person who has driven such a vehicle has surely
been appalled by the drastic decrease in the indicated fuel consumption rate as soon as they press the accelerator. At
the same time, lifting the foot from the accelerator when the vehicle is slowing down immediately produces a
similarly amazing improvement in the consumption rate. Practically, the driver does not expect either indication to
be particularly accurate. However, on checking the actual fuel consumption the next time the tank is refilled, there
is no doubt that if the driver has tried to avoid harsh use of the accelerator the fuel consumption will be reasonable,
whereas if he has been clumsy with the accelerator the evidence is clearly reflected by the speed that money
disappears from his wallet.

Thus, ‘inspection’ and ‘monitoring’ are not alternatives; they are complementary techniques that should be
applied in tandem to obtain maximum effect. Monitoring provides an easy low-cost but instantaneous prediction of
the likely overall corrosion rate. It can be used to evaluate the probable effectiveness of alternative plant operating
strategies to limit or avoid corrosion damage. Inspection is an intermittent method of obtaining a very accurate but
expensive indication of the actual condition of process equipment. It does not provide any control, but it does
enable the accuracy of the predicted rate to be improved with experience. As confidence increases in the reliability
of the monitored rate, there is a solid basis for extending the intervals between inspections without increasing the
risk of unexpected failure and consequent unplanned downtime. This issue is considered later in the context of risk
based inspection (RBI).

3. Equipment and Instrumentation

Corrosion monitoring equipment and instrumentation can be divided into two broad categories:

• Conventional Instrumentation
• Advanced Instrumentation

3.1‘Conventional’ Instrumentation

It was reported historically that approximately 50% of expenditure on corrosion monitoring equipment was
for access fittings. Of the other 50%, half was for weight loss coupons and half for instrumentation. Of the 25%
allocated for ‘instrumentation’, two-thirds was for ER equipment and one-third for LPR equipment (and of the one-
third related to LPR equipment, at least half was not applied correctly or was being used in an inappropriate
circumstance and that could not yield any reliable data. It should be noted that these findings related primarily to the
oil and gas production and refining industries, but the volume of sales outside those sectors was sufficiently small
that for all practical purposes it could be ignored. It should be noted that these techniques apply primarily to the
monitoring of uniform corrosion. Thus, bearing in mind the findings of the DuPont survey [4], this means that a
minimum of 70% of the corrosion monitoring market currently is not addressed at all.

It should not be concluded, however, that conventional instrumentation is necessarily outdated or of no

practical value. On the contrary, corrosion coupons are an inexpensive and simple method of obtaining a visual
indication of corrosion conditions within a vessel or pipeline. It may be questionable whether this indication is more
or less valuable than a UT (ultrasonic testing) wall thickness survey, but the advantage of a corrosion coupon is that
a clean freshly-prepared surface is being exposed in the service environment on a regular basis and, provided that
the corrosion rate is not especially severe, the data so obtained serve to verify that the corrosion rate of the plant
itself is not excessive.

In locations where coupon exposures indicate that the corrosion rate is higher than desired, further
investigation is called for. It is at these points that on-line instrumentation is applied. The simplest type is ER
instrumentation. The ER technique originally was patented by Standard Oil [1], but was exploited initially by Baker
to provide a relatively rapid indication of the efficiency of its oilfield corrosion inhibitors and treatment chemicals.
570 William M. Cox / Procedia Engineering 86 (2014) 567 – 575

The monitoring group was devolved from Standard Oil and became Rohrback Instruments, and that company is still
well-recognized in the corrosion monitoring field. The main competitor to Baker at that time was Petrolite, and as
the patented ER technology was licensed to Baker and Rohrback, Petrolite needed to identify alternative technology
to demonstrate the effectiveness of their oil field treatment chemicals. In consequence, Petrolite licensed and
developed LPR instrumentation [2]. However, whereas ER probes would operate satisfactorily in low-conductivity
hydrocarbon conditions typical of oil and gas production systems, LPR instrumentation had to be used in moderate
or highly-conductive environments. This was not a problem in cooling water systems or oil field installations with a
high water cut, but LPR sensors did not perform well in sour systems where the sulfide deposits tend to bridge the
sensor elements, or in low conductivity systems where there was insufficient moisture to provide a conductive path.

ER probes are the basic tool for oil field monitoring even today, and the technique is simple and well-
understood. Nevertheless, they too have significant disadvantages. For example, in order to have sufficient
sensitivity to obtain a reasonably prompt response, the sensor element must be relatively thin. Unfortunately, this
means that if a short-term excursion in the rate of corrosion takes place, the service life of the ER sensor is quickly
used up. In addition, any incidence of pitting attack will result in excessively high apparent corrosion rates because
the sensor element is being thinned at a specific location, whereas in practice the degree of pitting may not be
particularly serious. ER probes also have poor tolerance to fluctuations in operating temperature, which interferes
significantly with the accuracy of the readings. For this reason it is not possible to produce a reliable temperature-
controlled ER sensor and this again limits its applicability for any heat-transfer related application.

LPR instrumentation is based on an electrochemical monitoring approach. In this case, instead of

monitoring the apparent change in the resistance of an exposed wire or tube, the change in resistance across an
electrochemical cell is evaluated and changes in the polarization resistance value correlated with metal loss as
calculated using the Stern Geary relationship [6] and Faraday’s Law. There is an implicit assumption when using
the LPR technique that the “Polarization Resistance” value is directly related to the corrosion rate. In practice, this
value is the sum of the solution resistance (i.e. the conductivity) of the electrolyte in the corrosion cell and the
charge transfer resistance, i.e. the value of the resistance that controls the transfer of charge at the corrosion interface
due to the anodic dissolution of metal to form metal ions. In moderately conductive and highly conductive
environments, this assumption results in a small but constant error on the calculated corrosion rate, but the error is
not problematic when comparing corrosion behavior in a suitable corrosion environment (such as in brine or treated
cooling water, for example).

However, in low conductivity environments such as hydrocarbon streams containing small concentrations
of moisture, or condensing environments with thin electrolyte layers, or high purity water systems, the resistance of
the bulk environment introduces a large and possibly variable error component that it is impossible to compensate
for using LPR instrumentation. This limitation is unfortunate, because electrochemical monitoring instrumentation
has the particular advantage of proving a truly real-time response to changes in the corrosion interface.

3.2 ‘Advanced’ Instrumentation

For want of a better title, instrumentation that has been developed after the commercialization of ER and
LPR technology can be grouped under the heading of ‘advanced’ instrumentation. Although it might be argued that
techniques such as inductive determinations of (ER) corrosion rates, Cion (a type of specialized EN approach), and
even potential mapping (FSM) might be considered to be ‘advanced’, these are perhaps best considered as
proprietary attempts to claim a market advantage. By contrast, electrochemical impedance spectroscopy (EIS),
harmonic analysis (HA) and electrochemical noise (EN) measurements are substantial developments of the
precursor DC polarization method.

EIS was the first of these, and offered some advantage compared to DC polarization in that an alternating
signal was applied to the corrosion cell, enabling the relative contributions of the solution resistance and the charge
transfer resistance components to be delineated. EIS was/is particularly useful in that it can permit the contributions
of various components in a corroding system to be characterized by their differing time constants, and in
consequence the technique has found particular use in the research laboratory. However, it is arguable that the
response of the corrosion cell is influenced directly by the application of the monitoring technique itself. For
 William M. Cox / Procedia Engineering 86 (2014) 567 – 575 571

example, the fact that the measurements are made under sequential perturbations of progressively longer amplitude
means that the response of the electrochemical cell is disturbed to an increasing degree by the progressively longer
and slower polarization cycles. This is an artifact of the measurement system, as a naturally-corroding system is not
subjected to such external perturbations or, in many cases, to any external polarization whatsoever. In consequence,
it is debatable whether corrosion rates calculated on the basis of impedance results are ‘accurate’, though the
capability to measure and discount errors resulting from changes in solution conductivity and/or diffusion effects is
undoubtedly useful.

Harmonic Analysis (HA) is essentially a specialized determination of the Tafel coefficients identified by
Stern and Geary to be pivotal in controlling the rate of corrosion in an electrochemical system. The technique
therefore is closely associated with EIS measurements though it is distinctly different and the results obtained relate
more to the estimation of corrosion rate from polarization measurements than to EIM analysis as such. Again, it
should not be inferred from the above comments that HA has no value as there is every justification for concluding
that estimates of the Tafel coefficients obtained from HA can reflect short-term changes in the rate of corrosion
indicated from LPR or EIS measurements. Thus, such determinations can improve the precision of the calculated
corrosion rate. However, this still leaves open the question as to whether or not gross perturbation from the
spontaneous corrosion condition will still yield ‘accurate’ corrosion rates. A better conclusion might be, for
example, that the data so obtained provide more insight into the corrosion mechanism and can be used to improve
the precision of corrosion rate estimates based on DC polarization or AC impedance approaches.

The advantage of the electrochemical noise (EN) monitoring approach is that no external perturbation of
the electrochemical interface. Potential and current transients generated by the corrosion process itself provide the
primary inputs to the measurement system [7, 8]. However, the EN technique has long been shrouded in mystery
and skepticism, not least because for years it was believed impossible to monitor corrosion directly and hence it was
‘necessary’ to resort to some kind of external polarization in order to obtain a reliable estimate of corrosion activity.
This assumption is ridiculous and there has been a constant progression of EN applications development, often
driven by the desire to better understand complex corrosion phenomena that were not amenable to evaluation using
‘conventional corrosion instrumentation [9-12], which in due course will be exploited in plant-wide
corrosion/condition management. In real terms, EN monitoring is the technical equivalent of vibration monitoring
on rotating and reciprocating equipment. Who would countenance the idea of stopping a turbine or pump, putting
on a vibration sensor, and then artificially shaking the machine in the hope that fluctuations in the output of the
sensor could be used to determine the condition of the equipment? Yet that, in essence, is precisely what is done
when using external polarization, whether DC or AC, to obtain and indication of corrosion rate. In so doing, the
characteristic nuances of the electrochemical potential and current responses are eradicated in the tide of much
larger external applied transients.

The key advantage of vibration monitoring technology is that analysis of the signals spontaneously
generated by the passive sensor can not only characterize the degree of vibration on the machine, but the vibration
‘signatures’ can be used also to identify the source of the signature – i.e. misalignment, imbalance, broken teeth in a
gearbox, loose footings on a drive. Furthermore, the rate at which such signatures develop can provide the engineer
with a prediction of the time to failure, allowing scheduled outages to be arranged during which the affected
component can be repaired or replaced.

The predictive capability is especially valuable. Mechanical maintenance is a given requirement in the
operation of rotating and reciprocating equipment. The ability to anticipate a probable failure, even to the extent
that its time to failure can be projected, is highly advantageous in that it allows embarrassing, inconvenient and
expensive unplanned outages to be avoided altogether. It is for this reason that vibration monitoring has become the
key component of so-called plant condition monitoring. Unfortunately, the main disadvantage of vibration
monitoring is that it does not work well on static plant. As the majority of plant equipment is not supposed to spin
or shake, this limits the applicability of ‘condition monitoring’ to only a few sections of the plant.
572 William M. Cox / Procedia Engineering 86 (2014) 567 – 575

Electrochemical noise monitoring is the equivalent technology to vibration monitoring, but it can be
applied throughout the static plant equipment – i.e. to heat exchangers, reaction vessels, and piping. When properly
applied, EN also has the capacity to provide an accurate real-time indication of corrosion rate. Furthermore, EN is
able to provide that real-time information in the specific low-conductivity and/or condensing environments that
preclude the use of conventional LPR instrumentation. Additionally, EN can be used to identify the mechanism of
attack – whether uniform, pitting, crevice, SCC, stress-assisted IGA, CF, cavitation, under-deposit attack, droplet or
bubble impingement, fretting or even filiform attack. Each of these mechanisms exhibits a different and absolutely
exclusive potential and current ‘fingerprint’ response generated by the fundamental electrochemical interface
condition in precisely the same way that vibration monitoring is able to capture and characterize differing
mechanical phenomena. Sadly, consistent shortfall in the quality of EN instrumentation and in the efficacy of its
application has resulted in a persistent lack of confidence that the capability claims can be substantiated.
Nevertheless, the fundamental technology is without question and there is no doubt, therefore, that in due course
EN-based sensor technology will become the key to advanced plant-wide condition management.

The term ‘management’ is emphasized here because ‘monitoring’ just adds cost. It calls for more
equipment, more software, more time, more analysis, and more things to install and maintain. There is no advantage
in monitoring for its own sake. Monitoring only makes sense if the results of the monitoring activity are used as a
basis to change the way the plant is operated. Whereas monitoring simply adds cost, management brings cost-
benefit. Being in a position to manage the plant condition to avoid corrosion, improve product quality, reduce
inspection activities, increase safety, enhance reliability and increase capacity increases the productivity and
profitability of plant operations in a way that simple monitoring – counting what has been lost – can never aspire to
achieve.

4. Strategic Application

With the above objectives in mind, how then can the breadth of corrosion instrumentation, whether
conventional or advanced, be used to best advantage in the short term, while retaining the longer-term objective of
plant-wide corrosion and condition management?

First of all, it is important to appreciate that ‘advanced’ instrumentation is not invariably needed.
Conventional monitoring approaches may be more cost-effective and can provide perfectly satisfactory results as
part of an overall corrosion management system provided that they are appropriately selected and applied.
However, it is unhelpful to hope that randomly-installed corrosion monitoring equipment will provide the basis of
an effective corrosion management system unless there has been a systematic appraisal of where such sensors
should be installed, what type of attack the sensors are to identify, and the response that should be prompted when
higher rates of attack are detected. Thus, it is useful to employ a risk-based appraisal strategy when designing the
corrosion monitoring installation, tracking each stream from source to exit in precisely the same way as might be
undertaken when implementing an RBI (risk based inspection) approach. In essence, this approach comprises a
consideration of the plant material, service environment chemistry, operating temperature, pH, flow condition, etc.,
to decide when, where, and what type of corrosion might take place. In an existing plant, it may not be necessary to
conduct a theoretical appraisal of such conditions; operating experience and maintenance records will provide a
direct indication of where corrosion is taking place and, from the standpoint of the present strategy, any degree of
corrosion greater than virtual zero is unacceptable.

The next criterion to consider is any aspect of the plant condition that might cause there to be some kind of
change or excursion in the corrosion environment at that location. This might be the introduction of another process
stream, a change in the operating temperature or pressure, a change in the plant construction material, a flow-related
phenomenon such as turbulence, cavitation or droplet/particulate impingement. When the cause of such a change is
identified, an appropriate sensor should be positioned at the location of greatest attack. This requirement is very
important. It is not useful to position a sensor designed to detect impingement attack or cavitation damaged
upstream or downstream of the location at which the damage is taking place. Very localized phenomena are causing
the damage condition. Similarly, it is a waste of time installing a sensor that is capable of monitoring only uniform
attack if the primary cause of failure is a localized corrosion mechanism.
 William M. Cox / Procedia Engineering 86 (2014) 567 – 575 573

Having said this, some compromises are permissible. For example, if droplet impingement or erosion-
corrosion (i.e. flow-assisted corrosion) is the fundamental mechanism of attack, it is possible to build a flow-cell that
will generate equivalent impingement or erosion conditions in a side-stream loop that will mirror or even exaggerate
the fault condition in the plant. That being the case, a suitable sensor design and choice of monitoring technique
will enable an instantaneous and completely reliable indication of corrosion risk to be provided to the operator
without interference in normal plant operations. Furthermore, when the service condition in the plant approaches the
fault condition, remedial actions can be formulated to avoid the risk condition and their effectiveness can be verified
in real time. It is true that designing such a facility is more expensive than installing a ‘standard’ off-the-shelf
corrosion probe. However, if that probe is positioned in the wrong location and/or an inappropriate monitoring
technique is used to interrogate it, not only is all the investment in the monitoring system wasted, but the fact that
monitoring instrumentation has been installed at all can give operators a false sense of security that the risk
condition is being avoided. This is a situation that should be avoided at all costs as not only is it potentially
dangerous but it will also undermine confidence in the entire corrosion monitoring strategy.

Ironically, confidence can be similarly undermined by installing probes where no corrosion will take place.
Consistent returns that indicate that corrosion rates are zero or negligible eventually engender a feeling that the
monitoring exercise is a ‘waste of time and money’, because the sensors are not sufficiently sensitive, or they are not
sufficiently sensitive to the mechanism of attack, or there simply isn’t any corrosion taking place at that location.
The exception to this situation is when corrosion prevention treatments or measures (i.e. control of operational
conditions or chemical treatment) are necessary in order to maintain the desired/obtained low corrosion rates. In this
case, there will invariably come a time when control of the treatment system will be less effective and then the
corrosion response will reflect it. Ideally, in such circumstances, modern corrosion instrumentation will be applied
that will provide an immediate indication in the potential corrosion risk. Not only will this engender confidence
with operations personnel that the equipment is reliable but it will demonstrate the necessity and effectiveness of
controllable corrosion prevention expenditure – a vital justification if uncontrolled remedial maintenance
expenditures, and perhaps even plant corrosion failures, are to be avoided.

5. Interface with RBI

Risk-Based Inspection (RBI) has been demonstrated to deliver substantial improvements in process plant
safety while also reducing inspection and maintenance costs. However, where intermittent excursions in process
conditions can lead to substantial unexpected increases in corrosion rate, the only strategy to address the issue is to
increase the frequency of inspection. This is expensive, especially if a plant outage is required, and may often be
unnecessary. Unfortunately, the high potential risk may dictate that a more frequent inspection regime is
mandatory. In such instances, on-line monitoring can provide a useful and cost-effective alternative to more
frequent inspections. The type, location, severity and distribution of the corrosion damage should be established and
then a suitable sensor designed that will detect the fault condition and can be placed at the location where the
corrosion damage is most extreme.

In effect, the role of the sensor is to provide a means whereby the ‘criticality estimate’ can be updated in
real time while the plant is running. All the time the sensor shows low corrosion rate, the criticality factor is low
and hence the corrosion rate assumption controlling the inspection interval is valid. When the sensor detects an
increase in the corrosion rate, this is indicative that the criticality factor is now high, and hence the inspection
interval should be decreased. However, if measures can be taken that reduce the corrosion rate again and the
decrease is confirmed by the sensor then the criticality factor is again low and therefore the originally-prescribed
inspection interval should be appropriate.

Note that the inspection interval should not be extended purely on the basis of the on-line monitoring
indication. The purpose of the on-line data is to verify on a continuous basis that the corrosion condition within the
vessel is as expected. Inspection Codes and Standards then set the criteria that specify the inspection frequency.
However, improved control afforded by on-line monitoring may well mean that future corrosion rates are lower than
historical data might suggest, and on that basis the frequency of future inspections may be decreased as confidence
in the accuracy of the real-time data is accumulated.
574 William M. Cox / Procedia Engineering 86 (2014) 567 – 575

6. Future Perspective

It is not so many years ago that engine management systems were restricted to racing cars and prestige
vehicles. Now they are a standard feature of even the smallest commuter cars. This is because even though the
initial investment may be high, the requirement for consistent performance, consistent reliability and consistently
low emissions has meant that the initial cost is immaterial as the applicability is so widespread. By comparison, in
many ways chemical, refinery and petrochemicals processing equipment, power generation combustion systems,
nuclear power generation and reprocessing plants and similar large production systems are still being operated on
the technical equivalent of a set of points and a carburetor. If the gains in efficiency and reliability are so desirable
for small vehicles, how much greater is the potential for improved management and safety for large processing
systems.

Although many organizations have adopted the term ‘management’ to describe their corrosion control
measures, in reality most are still at the ‘monitoring’ stage, with no real understanding of the targets or available
benefits from the implementation of a comprehensive corrosion and condition management strategy. The word
itself does not matter, it is the understanding of what the word means that is crucial, though a lack of appreciation of
the underlying intent can mean that the entire opportunity is squandered. The primary objective of this paper,
therefore, was to attempt to track the development of corrosion instrumentation technology in order to be able to
demonstrate clearly the future target and the means to achieve it. There is still work to do but the technology to
achieve the goal now largely is available. Miniaturization is probably the next development step but there is plenty
of advantage in the current generation of equipment for implementation to be progressed immediately.

7. Conclusions

There is an increasing acceptance that corrosion monitoring is an important tool for the measurement and
control of corrosion damage in process plant operations. However, many industries employ a scatter-gun approach
to on-line monitoring, installing corrosion probes of one sort or another randomly around the plant in the hope that
they may detect attack and provide an indication of how rapidly damage has progressed. The speed of response of
conventional instrumentation is too slow to give control and therefore the established approach to corrosion
engineering has been largely based on off-line materials selection and reactive actions on damage reports or failure
incidents.

The introduction of modern corrosion surveillance instrumentation has opened the pathway for
conventional practices to be updated. The application of a risk-based approach to the siting of corrosion sensors, the
selection of modern instrumentation that is capable of detecting the onset and propagation of localized attack as well
as uniform corrosion and instantaneous correlation of on-line corrosion with process chemistry data as well as with
periodic inspection results, presents the opportunity for both the severity and the duration of attack to be minimized,
thereby opening the way for real-time management of corrosion conditions within the plant. Future developments of
modern on-line electrochemical monitoring technologies offer the opportunity for real-time control, with its
associated benefits of improved safety, reduced operating and maintenance costs, extended service life, lower
contamination levels and higher product quality.

8. Acknowledgements

The author would like to acknowledge input and support from colleagues in Mitsubishi Chemical
Corporation, Japan; KEMA (now DNV-GL), Netherlands; Reaction Engineering International, USA, Capstone
Engineering (now Lloyd’s Register Capstone), USA; Riyadh Geotechnique and Foundations, Saudi Arabia; and The
Steel Protection Consultancy, UK.
 William M. Cox / Procedia Engineering 86 (2014) 567 – 575 575

References

1. “Conductometric corrosion measurement system”, Andrew Dravnieks Arthur J. Freedman,

Standard Oil Company, US Patent No. US3094865 A, 1959.
2. “Method and apparatus for determining corrosion rate”, James M. Kilpatrick, Petrolite
Corporation, US Patent No. US3406101 A, 1963.
3. “Adapter for handling fluid under high pressure”, Jay P. E. Gould, Grant Corporation, US Patent
No. US2752228 A, 1956.
4. Bovankovich, J. C. "On-Line Corrosion Monitoring", Materials Protection and Performance, pp
20-23, June 1973.
5. Ohtsu, T., and Miyazawa, M., “Materials Selection and Corrosion Management in a Process
Containing Halides”, CORROSION 2012, Paper No. 0001352, Houston, TX: NACE international,
2012.
6. “Electrochemical Polarization I: A Theoretical Analysis of the Shape of Polarization Curves”,
Stern, M., and Geary, A. L., J. Electrochem. Soc., Vol. 104, No. 1, pp. 56-63, 1957.
7. Hladky, K., USA Patent 455709, European Patent 084 404 A3, Canadian Patent 418938.
8. Eden, D. A., Dawson, J.L., John, D.G., USA Patent 5139627, European Patent 0302073.
9. "Corrosion Surveillance Applications for Nuclear Power Plant Systems" Roarty, D. H, Bogard, W.
T, Cox, W. M., Moore, D. C. A, and Quirk, G. P., Paper 192, NACE Corrosion 93 New Orleans,
March 1993.
10, "A Review of EPRI Projects Since 1984 that used Electrochemical Noise Instrumentation”, Syrett,
B. C., and Cox, W. M., ASTM Symposium on Electrochemical Noise Measurement for Corrosion
Applications, Montreal, Canada, May 1994. (Published in ASTM Special Technical Publication
1277: Electrochemical Noise Measurement for Corrosion Applications, J. R. Kearns, J. R. Scully,
P. R. Roberge, D. L. Reichert, and J. L. Dawson (eds.). Amer. Society for Testing and Materials,
West Conshohocken, PA, pp. 173-185 (1996).
11. “Risk Mitigation in Modern Plant Health Management”; Cox, W. M., Miyazawa, M., Foong, M.,
and Aller, J. A. Plenary Paper presented at the NACE India ‘CORCON 2000’ Conference,
Mumbai, November 2000.
12. Cox, W. M., de Jong, M., Swensen, D., “Real Time Monitoring of Corrosion and Fouling in Power
Generation Boilers and Waste to Energy Plants”, Corrosion 2012, Salt Lake City, NACE, March
2012.

View publication stats