Fouling of metal heat transfer surfaces used in crude oil refining operations wastes millions of dollars yearly as it causes costly shut downs for cleaning and maintenance. Fouling is also costly since it reduces thermal efficiency and reduces process throughput. Additionally, if fouling is effectively controlled it can significantly reduced greenhouse gas emissions by increasing the efficiency of heating processes.
Crude oil fouling can occur through several different mechanisms such as through asphaltenes/organic, autooxidation, inorganics, corrosion, or combinations thereof. In general, for most well desalted crude oils, fouling of heat exchangers downstream of the desalter is largely affected by the stability of the asphaltenes. Asphaltenes are the largest most polar and aromatic surface active fraction of the oils. They can be precipitated with aliphatic solvents but dissolved in aromatic solvents. They also have a tendency to adsorb at surfaces and stabilize emulsions. Several novel characterization methods developed at WRI can be used to predict the fouling propensity of crude oils.
In this study two very different crude oils were analyzed for their propensity to foul at heated surfaces. One crude oil was heavy with a large amount of very stable asphaltenes and the other crude oil was light with a low amount of unstable asphaltenes. The oils were characterized by a variety of methods which all indicated that the much lighter oil should contribute significantly to fouling despite having very few asphaltenes.
To conduct the fouling tests, a new fouling measurement technique—based upon the hot wire fouling methodology—was used. This system has significantly better thermal accuracy and fouling sensitivity than current standard industry fouling test methodologies. Improved thermal accuracy is achieved with thermal microscopy calibration. These measurements show that sub micrometer fouling buildup sensitivity is possible with a significantly reduced thermal mass of a microwire.
Results and Discussion
In this study, the fouling properties of two desalted oils are compared: a heavy oil with a high concentration of stable asphaltenes (°API = 16.4, 11.2 wt% asphaltenes) and a light oil with a low concentration of unstable asphaltenes (°API = 37.0, 0.5 wt% asphaltenes). Optical microscopy shows that the light oil is a self-incompatible oil, meaning it is unable to dissolve a portion of its own asphaltenes at ambient conditions. The suspended asphaltenes in the light oil can be dissolved by adding a small amount of nitrobenzene to the oil (Figure 1).
Figure 1. Optical microscopy (40X) showing the self-incompatible light crude oil with suspended asphaltenes (left), and the same oil with the asphaltenes dissolved in nitrobenzene (right, 3:1 wt:v, respectively).
Despite the fact that the light oil has a low amount of asphaltenes, they are not well dissolved. It is well known that self-incompatible oils are highly unstable and problematic and will contribute significantly to fouling.1 By using the Asphaltene Determinator, it was shown that the asphaltene solubility profile of the oil-insoluble asphaltenes were significantly less soluble than all the asphaltenes precipitated by heptane.
Additional oil characterization using a flocculation method and a modified colloidal instability index also clearly indicate that the light oil would be significantly less stable and contribute much more to fouling at heat transfer surfaces than the heavy oil.2,3 The automated flocculation titration (AFT) method (ASTM D6703) was adapted to analyze the lighter crude containing a small amount of asphaltenes by using nitrobenzene as the solvent and isooctane as the precipitant. The AFT results show the light oil has significantly less overall stability with a P value of 1.26 relative to the whole oil which has a P value of 2.5. Other Hiethaus parameters also show that the solubility of the asphaltenes (Pa) is much less in the light oil than the heavy oil (0.785 vs. 0.661, respectively), and that the solvation power of the oil matrix (Po) for the light oil is also much less than the heavy oil indicating that the maltenes of the light oil are significantly more aliphatic.
AFT results were supported by distilling both oils and determining the amount of ≈430 °C distillate and measuring the aromaticity of the distillate by proton nuclear magnetic resonance spectroscopy (1H NMR). The results showed that aromaticity was similar for both distillates but that the light oil contained 71 wt% of mostly aliphatic distillate in the maltenes and the heavy oil contained 35 wt% of aliphatic distillate in the maltenes. The much higher amount of mostly aliphatic distillate in the lighter oil is expected to have a larger precipitating effect on the asphaltenes—consistent with previous observations. Saturates, aromatics, resins-Asphaltene Determinator (SAR-AD) analysis of the distillation residues also showed that the light oil residue was significantly less stable than the heavy oil residue. Colloidal instability index values, (asphaltenes+saturates)/(aromatics+resins), were calculated for the residues which showed that the value for the light oil was significantly higher (2.25) than the heavy oil (0.77). Previous fouling studies have shown that oils with a colloidal instability index greater than 0.9 have a high fouling propensity.4 By combining the amount of distillate with its aromaticity and the amount of residue with its colloidal instability index, a modified colloidal instability index was calculated to account for all of the fractions of the crude oil.3 The modified colloidal instability index for the light oil was 1.26 and 0.79 for the heavy oil, again showing that the whole light oil will contribute significantly to fouling.
The results from the various characterization methods showed that the light oil was very unstable and should cause significant fouling at heat transfer surfaces compared to the heavy oil. This was verified in a series of fouling tests. The oils were tested in a sealed 12.4 mL autoclave charged with 300 psi of dry Argon. Fouling tests were conducted for 10 hours with the wire temperature at 300°C and the oil temperature set to 204°C. The repeats of these tests are shown in Figure 2 with the fouling factor plotted vs. test time. These results clearly show that the light oil has a much higher fouling rate and over time continues to foul at a much greater extent than the heavy oil.
Figure 2. Fouling test results of Light Oil and Heavy Oil compared to a baseline low fouling paraffin oil.
The relationship between the temperature of the fine wire and its resistivity is fundamental to the measurement of fouling using the hot wire method. This relationship must be determined for a given diameter and type of wire which then allows a calibration to be performed for each test wire.
The determination of the relationship of wire resistance to temperature is a relatively simple process: measure the resistivity of a wire directly with an ohmmeter over a temperature range. The accurate determination of temperature, however, is less simple.
The temperature can be measured directly with a thermal imaging microscope. All surfaces emit black-body radiation according to their temperature, which can be measured with the use of an infrared microscope in the temperature ranges in which we are interested.
A calibration was completed using thermal imaging of a painted wire as it was heated from room temperature to 315°C while the resistance was electrically measured. These measurements created our calibration curve shown in Figure 3.
Figure 3. Calibration obtained using thermal imaging to correlate electrical resistance of the micro wire to the temperature of the wire.
The morphology of the foulant is different for each of the oils tested and high resolution elemental mapping shows the differences in composition of the foulant for each oil. The grain structure of the metal surface is correlated to the fouling morphology showing the effect of grain boundaries on foulant formation and their spatial correlation with sulfide plumes in the foulant.
(1) Wiehe, I. A. J. Dispersion Sci. Tech., 2004, 25, 333-339.
Adams, J. J.; Schabron, J. F. Proceeding of the 15th International Conference on Petroleum Phase Behavior and Fouling; Galveston, TX, June 8-12, 2014.
Adams, J. J.; Schabron, J. F.; Boysen, R. Energy Fuels, 2015, 29, 2774-2784.
Asomaning, S. Pet. Sci. Technol., 2003, 21, 581-590.