Comparison of corrosion monitoring systems
1.0
Commonly applied technology for corrosion monitoring
There are several types of instrument that have traditionally been used for monitoring corrosion in
oil refineries. Two of the most common are corrosion probes and manual ultrasonic inspection.
1.1
Corrosion (or Electrical Resistance, ER) probes
Corrosion probes have been in use since the 1960s and are a very well established technology.
They rely on an intrusive element with a sacrificial tip, which sits in the process fluid and is
(normally) made from the same material as the surrounding equipment. As the sacrificial tip
corrodes, its electrical resistivity changes, which is recorded externally (usually on a locally mounted
data logger) but these are also increasingly available wirelessly connected. The corrosion of the
sacrificial tip is used to infer the level of corrosion being experienced by the surrounding equipment.
While being simple to use, corrosion probes suffer from a number of disadvantages:




1
The centre‐line measured corrosion may not be the same as the corrosion rate at the wall, especially for corrosion mechanisms where shear velocity effects can change the corrosion rate experienced at the pipe wall. The tip often corrodes away after two to three years (or even less with "high sensitivity" applications), while many refineries are now operating 5+ years between major turnarounds. Thus, the corrosion probe tip will usually need to be replaced on‐the‐run. Very careful safety procedures and intensive technician training are required to reduce danger to personnel. In spite of this, there have been several well documented safety incidents caused by probes being ejected at high velocity under residual pressure. Several international oil companies have banned removal of intrusive probes while the plant is running, with the result that they operate 'blind', from a corrosion standpoint, for the final, and most critical, one or two years of the cycle between turnarounds. The intrusive nature of these probes means that they cannot be installed during normal operations, since they require specialist mounting flanges to be bored and welded to the piping. The intrusive probe creates a disturbance in the flow of the fluid that can induce corrosion to occur further downstream. Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672


1.2
Many of the older type, data logger based probes require an engineer to visit the equipment to download data. They therefore require physical access to the probe location and have an inherently low acquisition rate. This latter point is an important issue for crude overhead systems, such as the overhead line itself, as these are often physically remote ‐ the probe data connection then has to be cabled to a nearby platform, increasing installation costs and opening the possibility for the cable to be damaged. Coupons – another form of intrusive corrosion measurement – require the coupon to be retrieved from inside the fluid to take a measurement. These carry even higher maintenance costs and associated safety risks than ER probes since they must be retrieved to even take a measurement of the weight loss (corrosion) of the coupon. Once retrieved, they can give useful information about the type of corrosion that might be occurring in the pipe. Again, they are indirect and provide very infrequent measurements and so are not suitable for calculating reliable short term corrosion rates. Manual ultrasonic inspection
Ultrasound has been applied in the oil and gas industry for the past 50 years and is a well-established
technique for measuring metal wall thickness. The technique involves the generation of ultrasound
from a transducer that is placed directly onto the metal surface. The ultrasound is transmitted
through the metal until it is reflected off the inside metal surface (backwall). The reflected
ultrasound signal (or A-scan) is recorded and the time difference (the 'time-of-flight') between the
sending and reflected signals provides the measurement of the wall thickness. While the technique
can be reliable, completion of a full set of measurements for a medium-sized refinery with 80,000+
corrosion measurement points is very time consuming and labour intensive, such that the wall
thickness at an individual low- to medium-risk point may only be measured every 2-3 years. It is
therefore very difficult to take measurements in key locations with enough frequency to measure
corrosion rates with any confidence, or to link periods of high wall loss to specific feedstocks or
process operations (which require measurements on the time scale of days to be useful).
In addition, while being relatively simple, manual ultrasound methods have the following
disadvantages:

2
Repeatability and reproducibility errors ‐ it is highly unlikely that consecutive measurements will be taken in precisely the same location by the same NDE technician. In addition, the equipment used and the skill level of the NDE technician can vary between measurements, introducing high variability to the measurements. The chart below shows manual measurements at a single (nominal) location over time from 1984 to 2013. It is clear that different conclusions regarding wall thickness and corrosion rate can be drawn over time. From such data, it could be inferred that the accuracy of manual ultrasound is +/‐ 0.5 to 1 mm (+/‐ 20‐40 thou). Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672


2.0
Figure 1: Manual ultrasound measurements at a fixed location over time [courtesy of Chevron] High temperatures ‐ temperatures above ~100˚C (212˚F) can permanently damage the NDE equipment. There are also safety risks to NDE personnel for higher temperature locations. Physical access ‐ the inspector needs to be able to have access to the equipment at the measurement location of interest, therefore requiring scaffolding (possibly permanently installed) and stripping of insulation to expose the metal work to make the manual measurements. Overview of the Permasense technology
Permasense permanently installed, ultrasonic, wireless, wall thickness monitoring sensors are simple
and cost effective to deploy at scale. They provide a quality and frequency of data that is
otherwise unavailable and enables enhanced operational decision making to enhance profitability.
The system has sensitivity to small changes in wall thickness, robustness to extreme plant conditions
and extended battery life (enabling reliable operation over the entire cycle between turnarounds).
2.1
Resilience to high temperature
The design of the sensor incorporates a unique and patented 'waveguide' design as shown on the
following page. The waveguides are made from stainless steel, which is a poor conductor of heat,
and so the electronics are kept safely away from the hot metal surface (up to 600˚C (1100˚F)).
3
Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672
Figure 2: Effectiveness of Permasense patented waveguide technology to protect electronics from high
temperatures
The ultrasound is transmitted from the 'sending' transducer, down one waveguide and the reflection
is transmitted up the other waveguide to the 'receiving' transducer. As with manual ultrasound, the
'time-of-flight' difference between the 'surface wave' signal and the first reflection from the internal
metal surface provides the wall thickness measurement, as shown in the diagram below.
Figure 3: Signal and wavepath of the Permasense ultrasonic sensor
4
Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672
2.2
Resolution of roughness effects
Permasense have recently introduced a major advance to their technology in the proprietary AXC
(Adaptive Cross Correlation) ultrasonic signal processing method.
AXC makes use of the previous recorded waveform to improve the resilience of the measurement
when the internal metal surface morphology is very rough, where normal ultrasonic wall thickness
measurements can break down. In addition, AXC further enhances the repeatability of the
measurements, meaning that even smaller levels of corrosion or erosion can be detected in a matter
of days. AXC enables the separation of the wall thickness measurement from the onset of
roughening of the internal surface - however, the presence of roughness is now captured separately
as a colour bar, known as the Permasense Shape Indicator, or PSI. This improved processing
method makes the interpretation of the data much easier and quicker.
3.0
Local measurements and area coverage
The Permasense system is designed to have a low cost of installation, through use of wireless
communications and battery power-packs, avoiding any need for cabling with the resultant
armouring and cable tray installation. This simplicity of installation makes these sensors ideal for use
in remote locations which are only accessible during turnarounds.
Each sensor has a measurement footprint of an area of approximately 1 cm2, which is similar to
manual ultrasound inspection. Thus, the probability of detection of localised corrosion attack using a
single sensor, would be small. In order to increase the probability of detection, sensors are installed
as multi-point arrays, at the highest risk locations.
The number of sensors needed for each array is driven by the proportion of the area being
monitored that is expected to be affected by corrosion - the smaller the affected area as a
proportion of the whole equipment being monitored, the more sensors that are required to achieve
90% confidence in detecting the onset of that localised corrosion activity.
5
Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672
Confidence in detection
1.00
20%
25%
0.95
15%
0.90
10% of area
monitored exhibiting
corrosion activity
0.85
0.80
5
10
15
20
25
30
No. of sensors
Figure 4: Variation of number of sensors with area of corrosion and probability of detection
Figure 4 shows the result of mathematical analysis carried out the Department of Non-Destructive
Engineering at Imperial College in London, showing the relationship between the numbers of sensors
deployed in an area, the area of the corrosion activity as a proportion of the total area being
monitored and the resultant probability of detection. Only a modest number of sensors are
required in an area to achieve an excellent probability of detection.
4.0 Point measurement resolution and the effect of process
temperature variations
All ultrasound-based measurements are affected by process temperature variations, due to the
change in speed of sound through the metal.
Figure 5: Variation of wall thickness measurement with process temperature
6
Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672
Figure 5 shows the variation of wall thickness, measured using permanently installed ultrasonic
sensor. When zoomed in, as shown, the variation is of the order of 0.05 mm (2 thou), for process
temperature fluctuations of 20˚C (40˚F). This level of variation is not ideal for determination of
short term changes in corrosion rates, despite being orders of magnitude less than achievable with
manual inspection. The latest generation of Permasense sensors (WT210) make use of an integrated
thermocouple to measure the metal surface temperature, and can automatically compensate the wall
thickness data for process temperature variations, as demonstrated in Figure 6 for the same data
shown in Figure 5.
Figure 6: Temperature compensated wall thickness measurement
The temperature compensated data shows variation of less than 10 micrometres (0.2 thou). This
degree of precision enables detection of much smaller, shorter term corrosion rates, with
confidence. However, as importantly, the corrected data shows that corrosion was not continuous
at this location and there are two discrete corrosion events, which were masked by measurement
noise in the original data. The precision achievable with the latest sensor models and automated
data processing is comparable to that of high sensitivity intrusive probes, but without their inherent
safety problems and high installation and maintenance costs.
7.0


7
Conclusions
Market conditions are driving refiners to seek new ways to raise profitability. This includes processing more variable quality crude oils, such as the US light tight oils (LTOs). In doing so, the risk of a corrosion‐driven failure is increasing, in a cost‐constrained environment where inspection headcount and contract resource can be limited. The growing availability of light tight oils, which have their own integrity‐related processing issues, is resulting in a choice between upgrading of metallurgy and chemical inhibition/corrosion monitoring. With tight budgetary constraints in place, many oil Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672




companies are opting for chemical inhibition and tighter monitoring. Payback times from an inhibition/enhanced monitoring strategy can often be measured in the order of a few months. Intrusive corrosion probes have the required sensitivity and responsiveness, but do not measure equipment integrity and are complex to install and maintain and suffer from potential safety problems when changing sacrificial probes. They represent a single point measurement that infers the impact of the corrosiveness of the process fluid on the equipment wall. Manual ultrasound suffers from repeatability/reproducibility issues due to variations between measurements in measurement location, operator and equipment. However, manual ultrasound also requires that an inspector can gain access to the location, which is not often economically feasible or safe online. The latest generation Permasense sensors are able to provide equivalent accuracy to 'high sensitivity' intrusive probes, by using automated temperature compensation, enabling measurement of short‐term changes in corrosion rates with confidence, making them ideal for tracking corrosion from short crude processing campaigns. Installed at more than 70 refineries world‐wide that are owned by international oil companies, independents as well as national oil companies, Permasense sensors have automatically delivered more than 10 million on‐line measurements over the past 5 years to those personnel who need the data to make better informed operational and asset integrity management decisions. Contact:
For further information, about Permasense continuous corrosion and erosion monitoring solutions
please contact us by email at [email protected] or by telephone to our UK offices at +44 20
3002 3672 and +44 1224 628 258 or Houston office at +1 281 724 3774 or our Kuala Lumpur office
at +60 3 6200 0788
8
Permasense Ltd, Century House, 100 Station Road, Horsham, RH13 5UZ, UK
www.permasense.com
[email protected]
+44 20 3002 3672