Refinery equipment – such as tanks, vessels, valves and piping – is often worn away, microscopic bits at a time, through corrosion and erosion. These two phenomena attack invisibly from the inside out, but the damage is very real. If left unaddressed, leaks and even catastrophic failures are the result. Corrosion is a growing problem in refineries for several reasons.
First, crude oil contains some level of corrosive contaminants. Some are natural, while others are added during extraction. A major 2019 incident involving Gomeltransneft in Belarus started when the pipeline began shipping heavily contaminated crude oil to European processors. As Reuters reported, “The oil was contaminated with organic chlorides, compounds used in the industry to boost extraction from oilfields by cleaning oil wells and accelerating the flow of crude. The compounds must be removed before oil enters pipelines as they can destroy refining equipment or, at high temperatures, create the poisonous gas chlorine.”1 During the incident, organic chloride content was 15 to 30 times more than normally allowable levels and the crude was shipped all over Europe before it was discovered. Every situation is not this dramatic. Some refiners purchase discounted ‘opportunity crudes’ in the knowledge that it may well contain higher levels of corrosive contaminants and other undesirable elements.
Second, the equipment in a refinery and even within a single unit typically differs by age, quality, and design. For example, a crude unit was originally conceived and built to process specific amounts of crude oil on a daily basis from a specific range of sources. Piping metallurgy and schedule reflect the expected service life under the anticipated conditions. Those assumptions likely reflect industry practices from 10 or maybe 40 years ago, but not today’s realities. Outputs are probably higher than originally planned, operating conditions will have been adjusted, and higher levels of contaminants may be incompatible with the materials making up the fixed equipment of a refinery. Every year, the refinery’s fixed equipment becomes another year older and experiences another year of corrosion attack from the inside.
Third, operations today across the industry are leaner than ever, with less staff and smaller budgets due to retirements and cutbacks. For many refineries, the most experienced people are largely gone, leaving little opportunity for the few technicians available to monitor conditions manually.
These situations produce a triple-threat of demands for higher production using lower-quality feedstocks while running with often limited knowledge of equipment condition.
Working with the cost picture
So how do these realities work within the CAPEX and OPEX framework? Refining can be considered an ultra-competitive industry, so every refinery is under pressure to improve financial performance. Those that cannot are often sold or shut down, sometimes leaving just the tank farms for use as terminals. Each faces a list of problems and pressures, including variable feedstock quality (Figure 1). This presents a very real operating problem of equipment corrosion, with impacts on asset health.
Figure 1. Operating without detailed knowledge of plant health can lead to incidents and poor profitability.
Under these circumstances, a plant wanting to achieve top-quartile financial performance can follow two strategies, or some of both.
Invest in plant or unit upgrades, replacing the most critical piping and equipment with higher-grade alloys and heavier schedules able to withstand corrosive feedstocks and process conditions over a long service life. This is obviously an expensive solution and takes time to implement. Planners will have to decide how far to carry the upgrade but, once completely done, it solves the problem without creating ongoing costs, assuming the corrosiveness of feedstocks does not increase even more in the future. It should be kept in mind that even with an unlimited budget, there is no such thing as a perfect material capable of withstanding attacks from all of the tens of different corrosive elements that it will likely face over the range of refinery process conditions.
Attempt to reduce the metal loss caused by corrosive process fluids by adding corrosion inhibitor additives. This can be very effective but presents its own challenges. First, the inhibitor must be matched to the corrosive agent(s) in the crude and the concentration. An ineffective combination does not solve anything and wastes the additive, so the effectiveness must be optimised and reevaluated whenever there is a feedstock source change or other process adjustment. Under the best circumstances, this can be a viable solution, but it carries an ongoing cost.
Given the situation with many refineries today, the question for plant managers is: how long can the situation be maintained before something breaks?
How bad is corrosion?
Determining the extent to which corrosion is a problem also becomes a CAPEX and OPEX question. Refinery personnel need to know both the immediate situation (how corrosive is the process fluid being run right now?) and the cumulative effect on the equipment (how thick is the pipe wall right here at this elbow? Is it still in spec or about to break through?)
The OPEX approach calls for technicians to perform non-destructive testing (NDT) tasks at strategic points around the unit, shutting down the operation and looking inside the pipes where possible. Failing that, the next choice is using hand-held ultrasonic thickness gauges. Manual inspection carried out by skilled technicians is excellent for providing a snapshot of equipment health at the locations that can be inspected, but in practice it is difficult to get sufficient consistency to historise and graph metal loss in a specific area over time. Given operational requirements and the personnel situation at most plants, this solution is not particularly practical for the purposes of monitoring, since physical access to the metal is necessary, requiring staging and insulation removal every time a measurement is taken.
The CAPEX approach uses probes installed in various locations to monitor conditions continuously (Figure 2). There are several technologies available, each with its own capabilities:
- Coupons, using the same material as the piping, can be inserted into the process fluid to mimic the pipe wall. Monitoring metal loss from the coupon manually over time can be extrapolated to the whole system. This is useful as far as it goes, but it provides no real-time information, nor does it reflect any particular part of the system. It cannot answer the elbow wall-thickness question.
- Electrical resistance (ER) and electrochemical probes indicate the presence of corrosive agents in the process fluid in real-time. Installing these requires process penetrations, but the data is highly useful since it provides immediate feedback as to the corrosiveness of a new batch of crude, and therefore warns when conditions are conducive to high metal loss.
- Ultrasonic thickness (UT) sensors mounted in strategic positions to measure metal thickness continuously. UT sensors that use the latest signal processing techniques are capable of repeatability of around 10 μm (0.4 mils), making it possible to detect even very small changes in metal thickness over time. Data can be collected and historised to monitor metal loss for root cause analysis and predictive maintenance. Sensors are available to cover the complete operating temperature range encountered in a refinery, up to 600°C (1100°F).
Figure 2. The best solution will likely use multiple measurement techniques to determine the corrosiveness of process fluids and their impact on actual metal thickness.
Using a combination of UT probes with ER or electrochemical probes provides the best combination for data collection and analysis, since it is possible to quantify the corrosiveness of the process fluid in real-time while measuring the extent of metal loss as it is being processed. This provides the most telling indication of what is happening for the most cost-effective immediate and predictive maintenance.
A systemic approach
One or two strategically placed inline probes can monitor the corrosive characteristics of all the feedstock coming into a unit. On the other hand, monitoring pipe and vessel wall thickness at all the critical points around a unit could call for 10 to perhaps a few dozen individual UT sensors.
Traditionally, installing this many devices would be a major undertaking due to the cost of cabling within hazardous areas on the process plant, but devices using WirelessHART® require no cabling for either power or communication. Their internal power module operates for up to 10 years, making the devices effectively maintenance-free once installed. In addition, since they require no process penetration, they can be installed any time without a shutdown.
Once devices are installed and communicating via a self-organising WirelessHART network, it is critical to ensure the data, both real-time and historical, gets to the right people at the right time. Emerson’s PlantwebTM digital ecosystem (Figure 3) can interface with WirelessHART to ensure reliable, robust, and secure data retrieval from the UT sensors, the inline probes and any other pervasive sensing devices in the plant via a gateway, the device connecting the wireless world in the plant to the existing IT infrastructure.
Figure 3. UT sensors and wireless-enabled inline probes share WirelessHART infrastructure to deliver actionable information to the right personnel worldwide.
The digital transformation solution is completed with data visualisation and analytics tools built into software to enable the operator or engineer to see how the corrosion risk is changing and how corrosion effects are impacting plant health. Such review and analysis can happen in the plant control room, office building, remotely, or anywhere that the company networks reach. Reliability and production optimisation teams can decide how to operate the plant more effectively, safely driving it to its maximum capability.
It becomes a simple matter to observe how a particular batch of crude or other process change affects the piping. Graphing thickness data during a run of, say, two weeks, will provide a detailed picture of metal loss across the monitored locations. The graph curves can be extrapolated to suggest how the effect will compound if it continues. This data can support better-informed decision-making, a primary goal of any monitoring system or digital transformation solution.
Improving data presentation and analysis
Figure 4 presents data spanning about six months from a single UT sensor, with wall thickness on the Y axis and time on the X axis. The data is continuous, high-quality, and repeatable. Using the latest ultrasonic signal processing methods, it is easy to see varying corrosion rates observed in this location, ranging from very low to very high metal loss.
Figure 4. Data from UT sensors is very detailed and accurate, making it easy to see historical metal loss variations and project forwards for predictive maintenance.
An engineer performing root cause analysis will undoubtedly try to understand why the corrosion rate changed so quickly on a certain date and why it may have stopped on another date. A logical next step is to look at data from ER or electrochemical probes to see how those values changed at the same time, or correlating the trends against other recorded process variables to determine the causes of the elevated corrosion rates. This information about what the plant is capable of handling in terms of corrosion risk enables optimised operational decision-making in the future by, for example, validating and optimising integrity operating windows (IOWs) for a particular area of the plant.
This historical information then informs determination of where the unit might be headed: is it reasonable to think present trends will continue, or can the process be adjusted to reduce the degree of metal loss? When should required piping replacement be anticipated?
Fortunately, effective monitoring systems can deliver the data automatically to those who need it. Sophisticated visualisation and analytical tools reduce the time needed for analysis and interpretation. One presentation mechanism is a single graphic, displaying data from dozens of UT sensors arranged vertically by their proximity in the plant or unit.
The resulting ‘heat map’ represents thousands of individual data points, with the colour intensity representing the rate of metal loss. With a glance, it is easy to identify where and when the most damage was taking place. This can be compared to corresponding operating conditions to determine what factors contributed to the problem, and what might prevent it from happening again.
The reverse is also true: planners can see situations where the plant is not being driven as hard as it could be (without causing excessive damage), or that a specific source of opportunity crude is not as destructive as initially thought. Questions such as ‘Are we going too far?’ or ‘Should we push harder?’ can be answered using real data rather than guesswork.
Return on investment
Understanding what is happening in relation to erosion and corrosion in a plant or unit, and acting correctly in response, can generate returns in multiple ways:
- Safety improves since incidents related to leaks and product releases decline.
- Routine maintenance spend is optimised to replace only equipment that needs to be replaced due to more detailed condition data.
- Turnarounds can be less frequent and shorter, focusing on just the equipment requiring attention.
- Corrosion inhibitors and other additives can be optimised since their effect is easier to see.
- Plant availability increases since there are fewer unplanned outages.
- Opportunity crudes and other profit-driving operational decisions can be pursued more aggressively, since it is easier to see their effects and validate that corrosion is not damaging the plant too much.
All of these factors add up and contribute to higher profitability for the plant. Monitoring corrosion risks and the impact they have on asset health is essential for safe, reliable, and profitable operations. Using probes to track corrosion risk and thickness monitoring, in combination with effective data retention and analysis, ensures all assets are driven to their maximum capability while maintaining safety.
Case study: crude unit overhead corrosion
A refinery’s crude distillation unit (CDU) overhead system experienced pipe corrosion, causing leaks over many years. The refinery tried various solutions, including pipe clamps as reinforcements and a filming amine corrosion inhibitor, all to no avail. The plant installed 29 UT sensors to determine what was happening. After observing the data for several weeks, various efforts were undertaken to reduce the corrosion rate being experienced. After several failed efforts, the corrosion monitoring system made it clear that changing the chemical inhibitors injection was the solution, and the thickness tracking and the corrosion rate levelled out immediately.
The thickness data showed that, with the treatment changes, the pipe still had enough life to allow the replacement of the piping – which had already undergone substantial damage – to be scheduled far in advance in the next planned turnaround. Avoiding having to pay for rush replacement piping fabrication as well as avoiding an emergency shutdown saved several million dollars in rush charges and lost income from an unscheduled outage.
Internal corrosion in process plants, such as refineries or chemical plants, is a real, complex and generally poorly understood threat to the plant’s ability to function effectively and profitably. The corrosion risk experienced in the plant varies dramatically over time and this, in turn, alters the rate of damage experienced by the plant equipment.
Employing a corrosion monitoring solution that monitors for both corrosion risk (fluid corrosivity) and corrosion impact (by monitoring the remaining thickness of the fixed equipment) can deliver clarity to this problem, enabling operators to drive their plant to its maximum capability.
The cost of ownership, including the complexity of installation and use, of these monitoring solutions has been dramatically reduced via the use of technologies such as wireless, battery powered, non-intrusive sensing devices, along with data visualisation and analytics software. Operators adopting these solutions are able to drive substantial additional profit with return on investment coming from a large range of sources, delivering payback typically within single-digit months.
- YAGOVA, O., GORODYANKIN, G., and ZHDANNIKOV, D., ‘How Russia contaminated $2.7 billion of oil exports to Europe’, https://www.reuters.com/article/us-russia-oil-insight/how-russia-contaminated-27-billion-of-oil-exports-to-europe-idUSKCN1S61YM
Written by Jake Davies, Emerson, UK.