Jan-2019

Mitigating overhead corrosion

Advanced data handling and sensitivity analysis enable fuller understanding of the influence of operating changes on salt formation.

BRANDON J H PAYNE, KEYURKUMAR PATEL, COLLIN W CROSS, MATTHEW G COLLINS and PABLO A GRIECO
SUEZ Water Technologies & Solutions

Article Summary

In the pursuit of efficient and profitable refinery operation, owner-operators must continuously balance feedstock flexibility and product optimisation with mechanical availability and long/short term asset protection. Further complicating this process is the limited ability to determine effectively which operational adjustments have the greatest impact. Ideally, sufficient, timely data are available to drive refinery units effectively to their optimal balance of reliability and profitability.

One area where the ability to balance production and reliability continuously is critical is the overhead condensing system of the atmospheric fractionator. The reliability of the typical crude overhead system is often poor due to variations in crude diet, product cut points, chloride contamination levels, operational variations, tramp amine loading, and a host of other drivers. Combinations of these factors can often lead to increased levels of salt deposition in vulnerable areas of the overhead condensing system, thereby leading to mechanical failure and loss of production. The ability to avoid integrity losses in this crucial system can save millions of dollars in lost opportunity and maintenance costs.

While the crude overhead system is subject to several corrosion mechanisms typically controlled through the implementation of a variety of strategies, salt induced corrosion is one of the most common. Salt point is defined as the temperature at which the first neutralisation salt begins to precipitate from the vapour phase. Due to the severity of this commonplace mechanism, the overall success of an overhead corrosion control application is highly dependent on keeping salt point temperatures for all amines at least 15°F below the water dew point (WDP), or water wash injection temperature if a water wash is in use. Because of the difficulty in computing dynamic water wash and salt point temperatures, as well as their associated safety margins, common industry practice is to target a tower top temperature (TTT) that is at least 15-25°F above the salt point. This criterion is a way of helping to ensure that salt precipitation will occur downstream of the WDP. Keeping corrosive salts from precipitating upstream of the water’s initial condensation point (ICP) is a desired condition because they will then precipitate into a significant amount of water to dilute and wash them.

Salt formation temperatures are driven by the product of chloride and amine partial pressures. These partial pressures are, in turn, related to the measured concentrations of amines and chlorides in the overhead receiver water. This relationship links salt precipitation to both chemical and operational factors, thereby allowing many potential corrosion control measures to negatively influence product distribution and quality. To better understand how prevailing conditions can induce salt deposition and associated reliability problems, both the WDP and salt points for all amine salts need to be computed frequently and linked to their inputs via sensitivity calculations. Commonly there are multiple key amines circulating in the system with potential to show dramatic day-to-day variability. Understanding the variability of amines and their associated salt points can be key in effectively mitigating the corrosion process over time.

Frequently measuring ion concentrations and computing salt points for each amine species in a dynamically varying unit can be a tedious and time-consuming task. Historically this has led to a lack of visibility into dynamic linkages between corrosion drivers and day-to-day operational decision making.

Salt fouling and associated corrosion can be more detrimental to overhead reliability than ICP corrosion via acid attack. To gain a thorough understanding of the corrosion process, salt points for all amine chlorides in the system need to be computed frequently and salt point safety margins rigorously enforced. If this is done, then overhead operations are much more robust, stable, and optimised. However, the necessary computations have historically had limited application by primarily providing only retrospective information with negligible predictive capability on current or future conditions. Additionally, neither the granularity of data nor the sensitivity analysis techniques necessary to understand the relative influence of operational inputs on simultaneous distillation and salt formation have been available. The described situation is somewhat analogous to attempting to drive an automobile with only a rear view mirror and no ability to see the road ahead to plan a safe and effective route through driving hazards. In the world of overhead corrosion, it has long been recognised colloquially that 90% of metal loss occurs during 10% of the time in operation. Due to the reasons above, and the constantly changing tendency to form corrosive salts in an operating unit, it has remained difficult using traditional approaches to systematically identify the proverbial 10% of time where damage-causing episodes occur so that effective mitigation is possible.

An often overlooked aspect of process system corrosion is that what is typically measured as the average corrosion rate is really the sum of contributions from a series of discrete random events mostly occurring at unobserved points in time. Corrosion loss is, therefore, not usually a continuous loss of metallurgy over time. In traditional corrosion monitoring and mitigation programs, the data upon which key performance measures are based are collected infrequently. These substantial gaps in data collection produce significant blind-spots in the evaluation of asset integrity, and many discrete corrosion events remain undetected. The ability to detect individual corrosion events effectively is often most limited by the relative difficulty of obtaining frequent, timely, complete, and accurate amine speciation data and the inability to easily compute the changing impact on salt precipitation, fouling, and corrosion (see Figure 1). Historically, the low frequency sample testing and long laboratory turnaround time required to receive detailed results was inadequate for capturing and characterising salt fouling quickly enough to allow for broad based proactive corrosion mitigation. As advances in amine speciation techniques have been made, lower latency and higher frequency sampling results can be incorporated into the industry’s monitoring protocols. This is important because a higher frequency and lower latency of sampling can provide greater granularity in corrosion monitoring data with dramatically increased reaction time. Many of the salting events that were previously missed can now be captured, and effective mitigating actions can be ranked and evaluated quickly enough to allow both unit operations and the corrosion process to be simultaneously manipulated on a day-to-day basis.

With the potential for significant variability in both concentrations and types of tramp amines entering the unit, the ability to characterise a wide array of amine species is crucial to properly mitigate overhead system corrosion. If the focus of amine speciation is limited to only a few commonly problematic amines arising from the use of H2S scavengers, increased sampling frequency may not allow for sufficient breadth of coverage to effectively mitigate overhead salt fouling and corrosion in a repeatable way. SUEZ Water Technologies & Solutions has addressed this dilemma by establishing a network of satellite field laboratories in key locations. This not only substantially reduces high sample latency periods, but also allows cost effective routine screening for over 22 different amines on a bi-weekly or more frequent basis. This combination of breadth of coverage, high frequency, and low latency provides a comprehensive analytical protocol for the industry.

Unfortunately, due to the time required to process manually the amount of data generated by enhanced amine speciation, as well as the lack of available tools capable of a rigorous sensitivity analysis, even when data are occasionally available, a backward-looking view is largely what has been practically possible to date. With the introduction of the computational techniques in the SafeZone protocol, a sensitivity analysis can now be performed autonomously on the highly detailed data described above. This ability has enabled operators to evaluate a wide range of novel proactive chemical and operational actions, used on a daily basis to minimise active corrosion as it occurs. This advance represents a significant step forward in overhead corrosion mitigation diagnostics and prognostics that has not previously been available.