Oct-2013

High-acid crude processing enabled by unique use of computational fluid dynamics

A methodology for applying CFD enables faster identification of pipe elements that are the most vulnerable to attack by high-acid crudes

Dr COLLIN CROSS
GE Water & Process Technologies

Article Summary

In today’s refining marketplace, the potential to process naphthenic acid-laden crudes often appears very attractive from an economic perspective. This is because such crudes are usually priced at a significant discount with respect to more mainstream crudes. Many of these naphthenic acid crudes are extremely corrosive, however, and can cause substantial damage to refinery equipment and/or have adverse impacts on reliability and safety if not processed correctly. Unfortunately, estimations of the possible problems involved with processing these crudes are often fraught with uncertainty. Rules of thumb, 
experiential-based strategies and reactionary tactics have dominated the industry in its approach to managing the associated risks. Due to the complexities involved with these estimations, many refiners adopt a conservative stance regarding the rate at which they utilise acidic crudes, or depend upon capital-intensive upgrades to equipment.

In order to gain the most benefit from acidic crudes, new strategies and tactics based on data-driven decisions are needed. By better understanding the interactions between a particular blend of naphthenic crude, a particular crude unit and a particular operating envelope, it is possible to more accurately estimate the feasibility of successfully processing high-acid crudes. Additionally, the refiner is able to deploy a host of proactive measures, allowing high-acid crudes to be processed more confidently and reliably.

Background
Naphthenic acid corrosion was first recognised as a corrosive influence in certain crudes as early as the 1920s.1 Since that time, the impact of naphthenic acid corrosion has been typified by its detrimental effects, as well as its difficulty to predict. While high-acid crudes (HACs, crudes with substantial amounts of naphthenic acid) have been available for quite some time, it has only been the last several decades where their processing has become more widespread. In fact, the last 10 years have seen many more refiners begin to explore the possibilities of processing HACs due to increasing pressures on refinery profit margins. While refinery economics provide a substantial incentive to process these acidic crudes, it remains difficult to manage the long-term corrosive impact of HACs.

Naphthenic acid corrosion, in contrast to more traditional forms of refinery corrosion, happens in sections of the crude unit having temperatures generally greater than 450°F (232°C). In these regions, liquid water cannot exist and there are often strong contributions from sulphidic corrosion. Due to this situation, traditional overhead inhibitor chemistries, monitoring tactics and concepts are no longer effective. Overall, the best strategy to deal with the ongoing processing of HACs is to upgrade the metallurgy of the refinery to high-grade stainless steels, such as SS317 or SS316. However, this is typically a long-term and capital-intensive effort. Due to the high cost of advanced metallurgies, shorter- term strategies to deal with the impact of HACs often revolve around controlling the blended total acid number (TAN) of feed to the unit and/or the TAN values in selected side cuts of the crude fractionators. Additionally, effective high-temperature corrosion inhibitors have been used in a widespread fashion to mitigate the impact of naphthenic acid corrosion.

By adopting a crude blending strategy, in concert with the use of high-temperature inhibitors and appropriate monitoring practices, refineries can begin to benefit from the attractive raw material costs of HACs while reducing their potential impact on refinery operations. Unfortunately, this strategy suffers from several disadvantages, which stem from two primary categories. The first is that TAN is well known to correlate very poorly with actual manifested corrosion rates in a refinery. Consequently, running various crudes at similar targeted blended feed TAN levels can give rise to very different corrosion rates within the crude unit. The second is that naphthenic acid corrosion from a particular crude blend is typically highly localised. This high degree of localisation, and the variability of localisation from crude to crude, prevents traditional inspection and monitoring techniques from being effective.

The two factors mentioned above make it challenging to understand the actual impact of long-term processing of HACs either before or during their processing.2-4 To further exacerbate these problems, significant non-linear variances for actual corrosion rates are found at equivalent TAN levels and caused by blended crude interactions.5 Therefore, blending the same HAC at the same feed TAN level with different base crudes can lead to significant variance both in terms of corrosion rates and patterns.

Many attempts have been made to predict actual refinery corrosion rates based on individual and/or blended crude properties.4 In large, these techniques are inadequate for real-world applications due to 
the highly localised nature of 
naphthenic acid corrosion. Examples abound in which a single circuit in a refinery side-cut circulation system will experience multiple corrosion rates over a relatively short time period in varied locations. Measured corrosion rate variances within a side circuit, such as an HVGO circuit, often occur typically within a range of two to five times. For instance, measured wall corrosion rates may be 10 mpy in one section of piping, but could be 40 mpy a few pipe diameters downstream. Such behaviour makes the application of any model providing a single corrosion rate from a set of fluid properties practically inadequate for understanding the true impact of long-term processing of HACs. These conditions expose the refinery processing HACs to significant risk in terms of both reliability and safety. This situation is largely because there is usually not enough practical information available that can be used to proactively measure and control active corrosion, as it is occurring, during a given processing run.

In order to combat these challenges, inspection departments often use risk-based inspection programs in which the frequency of inspection is based upon historical corrosion rates. Corrosion probes are also frequently placed into available valve locations to measure fluid corrosion potential. These techniques are not sufficient for the reliable processing of HACs. Corrosion probes placed in the wrong location will not provide a realistic view of active corrosion at the pipe wall. Additionally, it is not economically feasible to perform wall thickness measurements over an adequate number of locations with enough frequency to proactively identify and react to active corrosion. If the locations where active corrosion is likely to occur could be identified in some fashion, inspection efforts could be focused on these regions with much higher frequencies.