Ted RedmondSurface-engineered tubes developed by Westaim Surface Engineered Products Inc. (SEP) were successfully tested by Union Texas Petrochemicals (UTP) in a partial ethylene-furnace evaluation. The UTP plant cracks ethane feedstock in Geismar, La.
Westaim Surface Engineered Products Inc.
Michael P. Bergeron
Union Texas Petrochemicals Corp.
The first surface-engineered application, trademarked CoatAlloy, is a coating that resurfaces the inside of tubes and fittings so they resist carburization and coke buildup in pyrolysis furnaces.
Four coated tubes were tested in one of UTP's short residence time furnaces. Pressure drops of these four tubes were compared with that of uncoated tubes in the evaluation. After 1 year of operation, one coated tube and one uncoated tube were extracted from the furnace for metallurgical analysis.
Data gathered through tests show that CoatAlloy products significantly reduce coke buildup in ethylene pyrolysis furnaces and exhibit strong carburization resistance.
In addition, the tests prove that CoatAlloy-engineered surfaces have the mechanical properties necessary for sustained operation under the severe conditions of ethylene furnaces, including hot erosion resistance, ductility, thermal stability, and thermal shock resistance.
On the basis of strong results from the partial furnace test in 1998, UTP replaced all tubes and fittings of an existing sister furnace with fully coated CoatAlloy products.
Coking problemMotivation to reduce coking in furnaces stems from the industry's drive to wrest the greatest performance from the ethylene-production process. The design and operation of high severity pyrolysis furnaces have had major advances in recent years, resulting in increased productivity.
Existing process technology has begun, however, to reach its limits. New technologies address process optimization challenges posed by carburization and coke formation.
Carburization is a particularly severe problem for ethylene producers pursuing higher operating severities. In fact, carburization nearly doubles for every 100° F. (55.6° C.) increase in tube-metal temperature.
Carbon pickup increases metal volume and the coefficient of expansion, resulting in strong internal stresses that can cause permanent tube failures. The loss of the chromium layer due to carburization leads to voids and degeneration of tube surfaces. When carburization reaches 30-50% of wall thickness, it becomes the most frequent cause of ethylene-tube failures.
Formation of a stable, compact, protective oxide film is needed to minimize the effects of carburization. Higher chromium-tube construction is effective in oxide formation. Unfortunately, these alloys do not impart a perfect chrome-oxide layer. Also, evaporation of the oxide can occur at temperatures above 1,900° F. (1,038° C.) constraining obtainable severity of furnace tubes.
During pyrolysis, ethylene furnace tubes gradually become coated with an internal layer of coke, causing an increase in tube-metal temperature and an elevated pressure drop. Pressure drops cause production to decline, and output is further restrained as furnaces are taken off line to burn off the coke deposits.
The cumulative effects associated with coking have been a problem for the chemical industry since producers began cracking feedstock to create ethylene. Today, ethylene producers lose about $2 billion/year in production as a result of downtime for decoking. A similar magnitude of ethylene production is lost because of the need to operate furnaces at lower than desired severity to slow the rate of coke buildup.
Effectively managing this problem could save individual plants $5-20 million/year.
Shutting down furnaces for decoking can be an expensive proposition. The longer a plant can run, the more efficiently it can operate. But ethylene producers are most interested in turning up the heat to squeeze out more ethylene per pound of feedstock, while increasing plant margins.
Operating temperatures have been creeping up for years. Coil-outlet temperatures now run in the 1,544-1,625° F. (840-885° C.) range and tube-metal temperatures approach 2,000° F. Not surprisingly, the industry is eager to push the temperature barrier even higher with consistent severities to 2,012° F. (1,100° C.) and above.
Coatings technologyThe coating technology offered by SEP's CoatAlloy product can meet ethylene producers' run times or higher conversion rates.
Fig. 1a [105,424 bytes] shows that a lower coking rate allows longer run lengths. Alternatively, a producer may choose to run at the same run length but have a higher conversion (Fig. 1b).
The CoatAlloy product is a permanent engineered surface that can be applied to tubes and fittings of most any size and shape.
The technology is a three-layered system that creates a stable oxide to prevent catalytic coking, mitigate the accumulation of amorphous coke, and prevent carbon from penetrating into the tube and causing carburization. This is accomplished without sacrificing the mechanical properties of strength, ductility, and creep resistance in the base alloy.
Fig. 2a [166,416 bytes] shows a sample of a tube coated with a CoatAlloy surface. The structure of the coating is clearly visible.
Starting from the bottom of the figure, the first layer is the unaltered base tube; above that is the diffusion barrier; above that is the enrichment pool; above that is the engineered surface (i.e., anticoking surface).
The diffusion barrier serves to isolate the enrichment pool from the nickel and iron in the base alloy. The enrichment pool is the source of atoms that generate and regenerate the engineered surface.
CoatAlloy-engineeered surfaces offer a permanent barrier through surface alloy enrichment. The technology offers advantages over other anticoking products and processes, including the most prevalent among existing anticoking solutions: antifoulants.
Antifoulants involve injecting additives into the feed stream that react with the interior environment to resist the formation of coking during the cracking process. Antifoulants, however, are maintenance intensive. Plant engineers must constantly monitor additive levels during the operation.
In addition, because antifoulant additives are added at the front end of the furnace cycle, their application can be inconsistent along the full length of tubing.
While antifoulants work within the feed stream, other anticoking approaches are based on tube metallurgies. Elemental species such as nickel and iron on furnace tubes and fittings can provide catalytic sites creating filamentous coke, which can trap and hold amorphous coke during the cracking process.
Micro alloys, in particular high chrome-content alloys, can create surface oxides that reduce catalytic coking. Under normal operating conditions, however, the chromium oxide can spall, depleting the chromium content of the alloy.
Commercial testsSince the autumn of 1998, CoatAlloy products have been available for commercial application. They have been in qualifying tests with a number of producers for more than 2 years.
Recently, tests completed in conjunction with UTP evaluated the performance of four CoatAlloy-surfaced tubes in a commercial setting.
The results of tests at UTP's Geismar, La., plant have been positive. CoatAlloy-engineered surfaces significantly reduced coke buildup, exhibited strong carburization resistance, and allowed for sustained operation under severe conditions.
UTP operates an ethane/propane feed ethylene plant with a capacity of 1.25 billion lb/year. The plant has 13 pyrolysis furnaces, including three short residence-time furnaces.
One of these millisecond furnaces, BA100, was retubed in December of 1997. Tubes of 35Cr/45Ni were installed to replace the original tubes of similar metallurgy, which had been in service for 7 years. As part of an ongoing development program, four of the new 35Cr/45Ni tubes and fittings were coated with the SEP's CoatAlloy product.
Following several runs in the first quarter of 1998 to stabilize the operations, four runs were conducted to assess the performance of the SEP-coated tubes compared to uncoated tubes.
The furnace operated during the entire period on a feedstock that was nominally 95% ethane, 2.5% methane, and 2.5% propane. Conversion was held at 70% of design feed rates at a steam-to-hydrocarbon ratio of 0.35. Run lengths ranged for 25-30 days for the four tests. Radiant tube inlet and outlet pressures were recorded once per day.
The effectiveness of CoatAlloy-coated tubes in reducing coke buildup is measured through relative pressure-drop increase over the course of four runs. High rates of pressure drop signify coke build up.
Fig. 3 [65,899 bytes] presents the relative pressure-drop increase in the end of run (EOR) for each of the four runs. Performance of the uncoated tubes (maximum and average) and the CoatAlloy tubes are ranked according to the average performance of the CoatAlloy tubes.
These figures indicate that in each of the runs the CoatAlloy tubes experienced significantly lower coke buildup than the uncoated tubes.
In fact, as can be seen in Fig. 4 [92,834 bytes], the CoatAlloy tubes consistently had a lower-than-average pressure drop over the course of one run. Data from other runs show similar superior results for the coated tubes vs. the uncoated tubes.
Metallurgical testsThe relative stabilization of pressure drop from the above tests provides insight into the operational benefits of CoatAlloy-engineered surfaces.
To get a greater perspective on the anticoking performance of the technology, an extensive metallurgical-testing program was carried out on samples of two tubes which were pulled from the furnace after 1 year of service: a CoatAlloy tube and an uncoated tube.
These tests examined the microstructure and elemental content before and after 1 year of operation in the UTP furnace. The purpose of these tests was to determine the effect of 1 year of operation on coated and uncoated 35Cr/45Ni alloy in a high-severity ethane cracker.
Tests included examination of samples using optical microscopy (OM), scanning electron microscopy (SEM), and electron diffusion spectrometry (EDS) as well as etching tests to determine the extent of carburization.
Fig. 2b shows a sample of an uncoated tube after 1 year of service. The sample has significant chromium depletion in the near-surface region. This chromium depletion is signified by the circular voids apparent in the figure.
The depth of chromium depletion varied from 0 mm at samples taken at the inlet of the furnace to 0.15 mm in the middle of the tube to a maximum of 0.2 mm in samples taken roughly 75% of the distance from the inlet to the outlet.
Below the near-surface region of the uncoated tubes there is evidence of the onset of carburization, indicated by the densely packed chromium carbides in the lower part of Fig. 2b. These carbides were not present in the uncoated tubes prior to service. Etch tests indicated that carburization occurred to a depth of 0.5-1.5 mm (out of a total tube thickness of 7 mm) in the middle section of the uncoated tube.
The carburization in the uncoated tube can be ascribed to the chromium depletion near the surface of the pipe. The chromium depletion prevented regeneration of a protective chromium oxide layer. Loss of the chromium layer exposed nickel and iron sites, which catalyzed coke formation; this coke then penetrated the tube, causing carburization.
The reason for the chromium depletion is that chromium oxide is known to be mechanically unstable and prone to spalling under thermal cycling. After the chromium oxide spalls off, the remaining chromium near the tube surface regenerates the oxide until there is insufficient chromium left to regenerate the surface.
Fig. 2c shows a typical sample from the middle section of a coated tube after 1 year of service in the UTP furnace. Examination of the sample by optical microscopy indicated that the coating was in place and serving its intended function.
The protective oxide layer covers the entire surface, and there is no evidence of chromium carbides. Etch testing revealed no carburization. The areas of surface roughness are areas where the original surface has been depleted and the enrichment pool has regenerated the engineered surface.
The absence of densely packed chromium carbides, as present in the uncoated tube (Fig. 2b), confirms that the continuous engineered surface prevented carbon from penetrating the tube.
Also observable in Fig. 2c are a few minor voids in the diffusion barrier, which are not expected to impact the ongoing performance of the coating.
Field performance and metallurgical testing indicate that after 1 year in a high severity ethylene-pyrolysis furnace, the coating was performing well. This corroborates field data from other SEP-coated furnace installations, which indicate that after a similar period of service, coated tubes exhibit excellent anticoking performance.
Pilot plant resultsResults similar to those seen at UTP were also observed in pilot plant testing by Kellogg Brown & Root Inc. and presented last year at the 10th Annual Ethylene Producers Conference in New Orleans.
As SEP detailed at the conference, extensive lab and pilot plant testing demonstrated the coating technology's thermal stability, thermal shock resistance, hot erosion resistance, carburization resistance, ductility, and creep resistance.
Fig. 5 [91,594 bytes] shows the pressure drop over time for three different runs using a 20Cr/32Ni alloy tube. After an initial conditioning run, the rate of pressure increase in Runs 2 and 3 of the coated tubes is exceptionally low.
It is also noteworthy that even when conversion was increased to 80% near the end of Run 3, no increase in coking rate was observed. Likewise, sulfur levels were taken from 25 ppm to 0 ppm, and no increase in coking rate was observed.
Targeting new marketsThe performance of the SEP CoatAlloy tubes in commercial and pilot testing was sufficiently promising for UTP to place an order to coat the tubes of a sister furnace, BA99, for a retubing project in the first quarter of 1999.
At the end of February 1999, CoatAlloy-coated tubes had been installed in five partial and three full furnace retubes at the plants of five ethylene producers.
Through additional commercial trials similar to the UTP study, SEP continues to learn lessons of operational parameters for CoatAlloy products. In the process, SEP is establishing the operational performance for various gas and liquid feedstocks and pursuing new coating applications in process industries.
Ted Redmond is president of Westaim Surface Engineered Products Inc. He has more than 16 years' experience in new product development and commercialization. Prior to joining SEP in 1998, he was a senior manager with Boston Consulting Group, where he developed and implemented growth and competitive enhancement strategies for corporate clients in a number of industries including refining and petrochemicals. Redmond holds an MBA from the Stanford Graduate School of Business, a masters degree in electrical engineering from the University of Toronto, and a BS in computer engineering from the University of Alberta.
Michael P. Bergeron is a process improvement engineer at Union Texas Petrochemicals. He joined UTP in 1995 where he has worked to improve furnace operations and to take advantage of increased furnace conversions. Bergeron has 22 years of experience in the petrochemical industry. Prior to UTP, he worked for Exxon Chemical Co. in the pyrolysis group at the Baton Rouge petrochemical plant. His experience also includes the areas of reactive distillation, high-purity separation processes, and catalyst research and development. Bergeron holds a BS in chemical engineering from Louisiana State University in Baton Rouge.
Copyright 1999 Oil & Gas Journal. All Rights Reserved.