This paper presents an experimental method to produce biofuels and biochemicals from canola oil mixed with a fossil-based feed in the presence of a catalyst at mild temperatures. Gaseous, liquid, and solid products from a reaction unit are quantified and characterized. Conversion and individual product yields are calculated and reported.
The work is based on a reported study which investigates the processability of canola oil (bio-feed) in the presence of bitumen-derived heavy gas oil (HGO) for production of transportation fuels through a fluid catalytic cracking (FCC) route. Cracking experiments are performed with a fully automated reaction unit at a fixed weight hourly space velocity (WHSV) of 8 hr-1, 490-530 °C, and catalyst/oil ratios of 4-12 g/g. When a feed is in contact with catalyst in the fluid-bed reactor, cracking takes place generating gaseous, liquid, and solid products. The vapor produced is condensed and collected in a liquid receiver at -15 °C. The non-condensable effluent is first directed to a vessel and is sent, after homogenization, to an on-line gas chromatograph (GC) for refinery gas analysis. The coke deposited on the catalyst is determined in situ by burning the spent catalyst in air at high temperatures. Levels of CO2 are measured quantitatively via an infrared (IR) cell, and are converted to coke yield. Liquid samples in the receivers are analyzed by GC for simulated distillation to determine the amounts in different boiling ranges, i.e., IBP-221 °C (gasoline), 221-343 °C (light cycle oil), and 343 °C+ (heavy cycle oil). Cracking of a feed containing canola oil generates water, which appears at the bottom of a liquid receiver and on its inner wall. Recovery of water on the wall is achieved through washing with methanol followed by Karl Fischer titration for water content. Basic results reported include conversion (the portion of the feed converted to gas and liquid product with a boiling point below 221 °C, coke, and water, if present) and yields of dry gas (H2-C2‘s, CO, and CO2), liquefied petroleum gas (C3-C4), gasoline, light cycle oil, heavy cycle oil, coke, and water, if present.
There is strong global interest in both the private and public sectors to find efficient and economic means to produce transportation fuels from biomass-derived feedstocks. This interest is driven by a general concern over the substantial contribution of burning petroleum fossil fuels to greenhouse gas (GHG) emissions and its associated contribution to global warming. Also, there is strong political will in North America and Europe to displace foreign-produced petroleum with renewable domestic liquid fuels. In 2008, biofuels provided 1.8% of the world's transportation fuels1. In many developed countries, it is required that biofuels replace from 6% to 10% of petroleum fuels in the near future2. In Canada, regulations require an average renewable fuel content of 5% in gasoline starting December 15, 20103. The Renewable Energy Directive (RED) in Europe has also mandated a 10% renewable energy target for the European Union transport sector by 20204.
The challenge has been to develop and demonstrate a viable economic pathway to produce fungible transportation fuels from biomass. Biological sources include triglyceride-based biomass such as vegetable oils and animal fats, as well as waste cooking oil and cellulosic biomass such as wood chips, forest wastes, and agriculture residues. Over the past two decades, research has focused on the evaluation of biomass-derived oil processing using conventional fluid catalytic cracking (FCC)5–12, a technology responsible for producing most of the gasoline in a petroleum refinery. Our novel approach in this study is to co-process canola oil mixed with oil sands bitumen-derived feedstock. Normally, bitumen must be upgraded prior to refining, producing refinery feedstocks such as synthetic crude oil (SCO)—this processing route is particularly energy intensive, accounting for 68-78% of the GHG emissions from the SCO production13 and, in 2011, constituting 2.6% of Canada's total GHG emissions14. Replacing a portion of upgraded HGO with biofeed would reduce GHG emissions, since biofuel production involves a much smaller carbon footprint. Canola oil is chosen in this work because it is abundant in Canada and the US. This feedstock possesses a density and viscosity similar to those of HGOs while the contents of sulfur, nitrogen, and metals that could affect FCC performance or product quality are negligible. Moreover, this co-processing option offers significant technological and economic advantages as it would allow utilization of the existing refinery infrastructure and, hence, would require little additional hardware or modification of the refinery. In addition, there may be potential synergy that could result in product quality improvement when co-processing a highly aromatic bitumen feed with its straight-chain biomass counterpart. However, co-processing involves important technical challenges. These include the unique physical and chemical characteristics of bio-feeds: high oxygen content, paraffinic-rich composition, compatibility with petroleum feedstocks, fouling potential, etc.
This study provides a detailed protocol for the production of biofuels at laboratory scale from canola oil through catalytic cracking. A fully automated reaction system – referred to in this work as the laboratory test unit (LTU)15 – is used for this work. Figure 1 shows schematically how this unit operates. This LTU has become the industry standard for laboratory FCC studies. The objective of this study is to test the suitability of the LTU for cracking canola oil to produce fuels and chemicals with the goal of mitigating GHG emissions.
Figure 1: Conceptual illustration of the reactor. Illustration showing flow lines of the catalyst, feed, product, and diluent. Please click here to view a larger version of this figure.
Caution: Please consult all relevant material safety data sheets (MSDS) before using the materials. Work with crude oil samples should only be done while wearing proper personal protective equipment (safety glasses, gloves, pants, closed-toe shoes, lab coat), and the opening, transfer and handling of crude samples should occur in a vented fumehood. Heated hydrocarbons can be flammable in air, and the reaction system should be carefully leak-checked prior to use with crude oil mixtures. The reactor can reach temperatures as high as 750 °C, and high-temperature safety gloves should be used when working near hot surfaces.
1. General Considerations
2. Feedstock and Catalyst Preparation
3. Test Procedure
Sample inlet T | 90 °C | Post run pressure | 30 psi | |
Injector T | 90 °C | Pressure equilibration | 10 sec | |
Run time | 300 sec | Detectors | Thermal Conductivity | |
Column pressure | 30 psi | Data acquisition rate | 50 Hz | |
Channel A | Channel B | Channel C | Channel D | |
Pre-column | PLOT-U; 30 µm × 320 µm × 3 m | PLOT-Q; 10 µm × 320 µm × 1 m | Alumina; 3 µm × 320 µm × 1 m | – |
Column | Molsieve; 12 µm × 320 µm × 10 m | PLOT-U; 30 µm × 320 µm × 8 m | Alumina; 8 µm × 320 µm × 10 m | OV1; 2 µm × 150 µm × 10 m |
Carrier gas | Argon | Helium | Helium | Helium |
Inlet mode | Backflush | Backflush | Backflush | Fixed Volume |
Column T | 100 °C | 90 °C | 130 °C | 90 °C |
Injection time | 30 msec | 120 msec | 0 msec | 100 msec |
Backflush time | 12.5 sec | 5.0 sec | 5.5 sec | – |
Table 1: GC method parameters for analysis of gas produced by the LTU.
Figure 2: Vial attachment to condenser. Photo showing the location of the glass wool plug and the attachment of a GC vial to the condenser with silicone tubing. Please click here to view a larger version of this figure.
Figure 3: Weighing of product receiver. Plastic cover for the balance to weigh the long liquid product receiver, which may stick out of the top window. Please click here to view a larger version of this figure.
Figure 4: Liquid receiver attachment. Photo showing the attachment of liquid receivers to the product line. Please click here to view a larger version of this figure.
The established protocol has been successfully applied to an oil blend of 15:85 volume ratio (i.e., 14.73:85.27 mass ratio) between canola oil and an SCO-derived HGO20. For practical reasons (cost, availability of canola oil, and possible challenges in commercial operation), the study was focused on feedstock containing 15 v% canola oil addition, although feeds with higher concentrations were also tried. The blend was catalytically cracked at 490-530 °C and 8.0 hr-1 WHSV with varying catalyst/oil ratios (in the sequence 11.25, 10, 8, 6, 4, and 11.25). For comparison, the base oil (pure HGO) was also cracked under the same conditions. Table 2 gives the conversion and yield data which have been discussed in detail previously20. In cracking the blend, assuming there is no interference between the two components, the apparent yields contributed by each can be calculated arithmetically. Table 2 demonstrates qualitatively that, upon cracking, the canola oil in the blend contributes substantially to the yields of biofuels (e.g., gasoline and diesel) and biochemicals (e.g., propane, propylene, i-butane, and butylenes in LPG). Following the above assumption, the negative calculated HCO yields contributed by canola oil (15 v% in the blend) in fact result from interference between the two components during cracking20.
Actual Yields of Base Oil (HGO) | |||||||||||||||||||
Temperature, °C | 490 | 510 | 530 | ||||||||||||||||
Catalyst-to-Oil, g/g | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | |
Conversion, mass% | 57.50 | 62.06 | 64.95 | 66.83 | 66.77 | 67.62 | 59.79 | 65.23 | 66.99 | 69.11 | 69.45 | 69.37 | 61.57 | 65.82 | 68.50 | 70.16 | 70.02 | 69.82 | |
Recovery, mass% | 99.72 | 99.35 | 99.17 | 99.27 | 99.12 | 100.10 | 99.3 | 99.9 | 99.2 | 99.2 | 99.2 | 99.95 | 99.63 | 99.66 | 99.38 | 99.54 | 98.48 | 98.38 | |
Yields, mass%: | |||||||||||||||||||
Dry Gas | 1.28 | 1.49 | 1.65 | 1.71 | 1.80 | 1.79 | 1.73 | 1.92 | 2.07 | 2.17 | 2.26 | 2.24 | 2.33 | 2.60 | 2.76 | 2.90 | 3.00 | 2.99 | |
LPG | 10.96 | 12.33 | 13.39 | 13.80 | 13.42 | 14.06 | 12.54 | 13.83 | 14.45 | 15.10 | 15.13 | 15.10 | 14.01 | 15.43 | 16.27 | 16.90 | 16.98 | 17.14 | |
Gasoline | 42.00 | 44.00 | 44.67 | 45.09 | 44.71 | 45.10 | 42.06 | 44.97 | 44.95 | 45.34 | 44.85 | 45.07 | 41.64 | 42.75 | 43.45 | 43.33 | 43.15 | 42.76 | |
LCO | 21.86 | 20.65 | 19.72 | 19.23 | 19.30 | 18.79 | 20.53 | 19.09 | 18.62 | 18.01 | 17.62 | 17.79 | 19.39 | 18.24 | 17.50 | 16.79 | 16.75 | 16.94 | |
HCO | 20.64 | 17.29 | 15.33 | 13.94 | 13.93 | 13.59 | 19.68 | 15.68 | 14.39 | 12.89 | 12.93 | 12.85 | 19.03 | 15.94 | 14.00 | 13.04 | 13.23 | 13.23 | |
Coke | 3.27 | 4.24 | 5.23 | 6.22 | 6.84 | 6.67 | 3.47 | 4.51 | 5.53 | 6.49 | 7.21 | 6.96 | 3.59 | 5.04 | 6.03 | 7.04 | 6.90 | 6.92 | |
H2O | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | |
TOTAL | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |
Actual Yields of Blends (15 v% Canola Oil in HGO) | |||||||||||||||||||
Temperature, °C | 490 | 510 | 530 | ||||||||||||||||
Catalyst-to-Oil, g/g | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | 4.02 | 6.00 | 8.04 | 10.00 | 11.25 | 11.25 | |
Conversion, mass% | 58.80 | 63.93 | 66.78 | 67.79 | 68.10 | 68.78 | 64.83 | 68.72 | 70.96 | 71.89 | 72.09 | 71.98 | 67.12 | 70.44 | 72.52 | 73.26 | 73.51 | 73.81 | |
Recovery, mass% | 98.78 | 99.46 | 99.12 | 99.13 | 99.76 | 99.53 | 99.41 | 99.18 | 99.27 | 99.21 | 99.29 | 100.07 | 99.20 | 99.44 | 99.23 | 99.89 | 99.10 | 99.19 | |
Yields, mass%: | |||||||||||||||||||
Dry Gas | 1.47 | 1.68 | 1.86 | 1.92 | 2.04 | 2.00 | 1.96 | 2.18 | 2.32 | 2.41 | 2.55 | 2.53 | 2.54 | 2.77 | 2.94 | 3.04 | 3.35 | 3.21 | |
LPG | 11.39 | 12.70 | 13.77 | 14.37 | 14.33 | 14.61 | 13.48 | 14.90 | 15.71 | 16.12 | 15.96 | 16.36 | 15.05 | 16.35 | 17.10 | 17.53 | 17.59 | 18.13 | |
Gasoline | 40.64 | 42.78 | 43.40 | 42.73 | 42.61 | 42.99 | 43.58 | 44.63 | 45.01 | 44.55 | 44.21 | 43.77 | 43.46 | 44.07 | 44.17 | 43.46 | 42.95 | 42.70 | |
LCO | 21.81 | 20.31 | 19.44 | 19.09 | 19.25 | 18.74 | 19.05 | 17.84 | 17.04 | 16.76 | 16.71 | 16.87 | 17.95 | 16.77 | 16.03 | 15.77 | 15.62 | 15.63 | |
HCO | 19.38 | 15.76 | 13.78 | 13.11 | 12.65 | 12.48 | 16.12 | 13.44 | 11.99 | 11.35 | 11.20 | 11.14 | 14.93 | 12.79 | 11.45 | 10.97 | 10.86 | 10.56 | |
Coke | 3.41 | 4.68 | 5.57 | 6.66 | 6.99 | 6.94 | 3.75 | 4.77 | 5.68 | 6.59 | 7.10 | 7.02 | 4.02 | 5.11 | 6.04 | 6.92 | 7.46 | 7.51 | |
H2O | 1.89 | 2.08 | 2.17 | 2.11 | 2.14 | 2.25 | 2.06 | 2.23 | 2.24 | 2.23 | 2.27 | 2.30 | 2.06 | 2.15 | 2.26 | 2.31 | 2.17 | 2.26 | |
TOTAL | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |
Calculated Yields (mass%) Contributed by 85 v% (85.27 mass%) HGO in the Blend | |||||||||||||||||||
Dry Gas | 1.09 | 1.27 | 1.41 | 1.46 | 1.54 | 1.53 | 1.47 | 1.64 | 1.76 | 1.85 | 1.93 | 1.91 | 1.99 | 2.22 | 2.35 | 2.47 | 2.56 | 2.55 | |
LPG | 9.35 | 10.51 | 11.42 | 11.77 | 11.44 | 11.99 | 10.69 | 11.79 | 12.32 | 12.88 | 12.90 | 12.87 | 11.95 | 13.16 | 13.87 | 14.41 | 14.48 | 14.62 | |
Gasoline | 35.81 | 37.52 | 38.09 | 38.45 | 38.12 | 38.45 | 35.86 | 38.34 | 38.33 | 38.66 | 38.24 | 38.43 | 35.51 | 36.45 | 37.05 | 36.95 | 36.79 | 36.46 | |
LCO | 18.64 | 17.61 | 16.82 | 16.40 | 16.45 | 16.02 | 17.51 | 16.28 | 15.88 | 15.35 | 15.02 | 15.17 | 16.54 | 15.55 | 14.92 | 14.32 | 14.28 | 14.45 | |
HCO | 17.60 | 14.74 | 13.07 | 11.89 | 11.88 | 11.59 | 16.78 | 13.37 | 12.27 | 10.99 | 11.03 | 10.95 | 16.23 | 13.59 | 11.94 | 11.12 | 11.28 | 11.28 | |
Coke | 2.79 | 3.62 | 4.46 | 5.31 | 5.84 | 5.69 | 2.96 | 3.85 | 4.71 | 5.53 | 6.15 | 5.93 | 3.06 | 4.30 | 5.14 | 6.00 | 5.88 | 5.90 | |
H2O | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | |
Total | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | 85.27 | |
Calculated Yields (mass%) Contributed by 15 v% (14.73 mass%) Canola Oil in the Blend | |||||||||||||||||||
Dry Gas | 0.39 | 0.41 | 0.46 | 0.46 | 0.50 | 0.47 | 0.49 | 0.55 | 0.56 | 0.56 | 0.62 | 0.62 | 0.55 | 0.55 | 0.59 | 0.57 | 0.80 | 0.66 | |
LPG | 2.05 | 2.19 | 2.36 | 2.60 | 2.89 | 2.61 | 2.79 | 3.11 | 3.39 | 3.24 | 3.06 | 3.49 | 3.10 | 3.19 | 3.23 | 3.13 | 3.11 | 3.52 | |
Gasoline | 4.82 | 5.26 | 5.31 | 4.28 | 4.49 | 4.54 | 7.72 | 6.29 | 6.68 | 5.88 | 5.97 | 5.34 | 7.95 | 7.61 | 7.12 | 6.51 | 6.16 | 6.23 | |
LCO | 3.17 | 2.70 | 2.62 | 2.69 | 2.80 | 2.72 | 1.55 | 1.56 | 1.17 | 1.41 | 1.69 | 1.71 | 1.41 | 1.21 | 1.11 | 1.45 | 1.34 | 1.19 | |
HCO | 1.78 | 1.01 | 0.71 | 1.23 | 0.77 | 0.89 | -0.66 | 0.07 | -0.28 | 0.36 | 0.17 | 0.19 | -1.30 | -0.80 | -0.49 | -0.16 | -0.41 | -0.73 | |
Coke | 0.63 | 1.07 | 1.11 | 1.35 | 1.15 | 1.26 | 0.79 | 0.92 | 0.97 | 1.05 | 0.95 | 1.09 | 0.96 | 0.81 | 0.90 | 0.92 | 1.57 | 1.61 | |
H2O | 1.89 | 2.08 | 2.17 | 2.11 | 2.14 | 2.25 | 2.06 | 2.23 | 2.24 | 2.23 | 2.27 | 2.30 | 2.06 | 2.15 | 2.26 | 2.31 | 2.17 | 2.26 | |
Total | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | 14.73 | |
Recovery, mass% | |||||||||||||||||||
Mean 99.35 | |||||||||||||||||||
Standard deviation 0.31 |
Table 2: FCC performances of the base oil and the blend and the apparent conversions and product yields of the components in the blend. Please click here to download this table as a Microsoft Excel spreadsheet.
The presence of water and CO plus CO2 (included in dry gas) as cracked products from the blend but not from HGO alone (Table 2) is a direct indication that canola oil in the blend participates in reactions. Water is produced by combination of hydrogen and oxygen and CO and CO2 are released from decarbonylation and decarboxylation of fatty acids in canola oil, respectively.
Other evidence of cracking of canola oil in the blend is presented in Figure 5, which shows the effect of process parameters on H2 and CO yields. Observations demonstrate that all yields are not very sensitive to C/O ratio changes for a feed at a given temperature. However, for a feed at a given C/O ratio, both H2 and CO yields increase with increasing temperature, which is the driving force for cracking (Note: No CO from HGO at any temperature). Comparing the two feeds, the blend gives higher CO yields but lower H2 yields than the base oil at the same severity in terms of C/O ratio and temperature. The latter observation is attributable to water formation from hydrogen and oxygen during cracking of the blend.
Figure 5: Variations of H2 and CO yields with process parameters. Color code—black for 490 °C, pink for 510 °C, blue for 530 °C; thin solid lines—H2 yield of base oil, thick solid lines—H2 yield of blend; thick dotted lines—CO yield of blend; no CO detected for base oil (0 mass% CO yield of base oil at three temperatures). Please click here to view a larger version of this figure.
Another interesting observation with canola oil is illustrated in Figure 6, which shows the reactor temperature profiles during operation. Before injection of an oil, the reactor is at a nominal temperature of 530 °C. After injection, the reactor temperature drops (due to heating, vaporization, and cracking of the oil), reaching a minimum (heat consumption is in balance with heat input from the control system once the temperature drops to a certain limit) and rising towards the initial temperature. As such, one may use the minimum temperature as a measure of the heat required for the overall process. For a given feed, the minimum temperature depends on the amount of oil injected or the C/O ratio since the catalyst weight remains constant. As C/O ratio increases, the temperature drop decreases since less oil is injected. Comparing the two feeds, the blend consistently exhibits a smaller drop by about 1.5 °C at a given C/O ratio due to the heat release from the exothermic reaction H2 (g) + ½ O2 (g) → H2O (g) (-241.8 kJ/mol at 25 °C)21. Similar phenomena are also observed at the other two reaction temperatures.
Figure 6: Reactor temperature drops before and after feed injection at 530 °C (nominal). Color code – black for base oil, pink for blend; thin lines – before feed injection; thick lines – after feed injection. Please click here to view a larger version of this figure.
The protocol described here utilizes cyclic operation of a single reactor containing a batch of fluidized catalyst particles to simulate feed oil cracking and catalyst regeneration. The oil to be cracked is preheated and fed from the top through an injector tube with its tip close to the bottom of the fluid bed. The vapor generated after catalytic cracking is condensed and collected in a receiver, and the liquid product collected is subsequently analyzed for simulated distillation to determine yields of fractions in different boiling ranges. The noncondensable gaseous product is sent to an on-line gas chromatograph for analysis to determine yields of dry gas and liquefied petroleum gas. The volume of the gaseous product is measured by the water displacement method. After suitable catalyst stripping time, the deposited coke on the catalyst is determined in situ by burning the deactivated catalyst in air at high temperatures (typically over 700 °C). Levels of CO2 are measured quantitatively via an IR cell and are converted to coke yield. Any water formed is recovered and determined by Karl Fischer titration. The total recovery (mass balance) of the feed should be in the range of 96 to 102% prior to normalization of each product yield.
One benefit to this procedure is the use of the automated sequence carried out by the LTU during the reaction process. After initiating the test sequence in step 3.2.2, the system begins with system priming during which the LTU uses the preset conditions programmed prior to the run. If there is spent catalyst remaining in the reactor from previous runs, it is discharged into the waste vessel, and fresh catalyst from the specified hopper is loaded into the reactor. The system then waits a sufficient length of time to allow the temperatures of the reactor, feed line, syringe, and feed bottle to stabilize within 5 °C of their set points. The syringe is then filled at a feed rate of 1.2 g/min for 20 sec (corresponding to two slop times), plus injection time, and the product line is purged with N2. The system then waits again, until the reactor internal and skin temperatures are within 2 °C of their set points, and the coolant temperature is within 3 °C of its set point. Finally, N2 flow to the IR gas analyzer is started and the system records the initial mass on the scale used to weigh displaced water and the pressure of the initial product gas (which should be zero).
Following priming, the syringe pump is set in motion and feed is first diverted back to the feed bottle for 10 sec (first slop time) followed by feed injection into the reactor for a preset time after switching back the three-way valve ahead of the syringe pump. Upon completion of injection, feed is diverted back again to the feed bottle for another 10 sec (second slop time). At the end of feed injection, counts begin for both liquid strip time and catalyst strip time. The former is chosen as 7 (liquid strip multiplier) times the feed injection time while the latter equals liquid strip time less 10 sec with a maximum 360 sec. The product flows are sent to the gas collection vessel through liquid receivers where high-boiling products are condensed.
Catalyst regeneration starts with valve switches at the end of the catalyst stripping cycle. Air flow begins and the reactor temperature is raised to ~715 °C. The CO2 concentration is continuously monitored by the IR gas analyzer until it is below 0.3%. The air is turned off and the N2 flow to the reactor is re-established at the end of regeneration. The scale reading (mass of displaced water) is recorded along with pressure and temperature of the gas in collection vessel followed by gas mixing and warming (to ~30 °C). At this stage, the liquid receiver for the run can be removed manually from the system for subsequent handling if desired. The line between the collection vessel and the GC is purged with product gas, and the loop is filled for the subsequent GC analysis. Cool the reactor until the skin temperatures are 50 °C below their reaction set points and save all the data for the completed run. Return the unit to the first step of the test sequence for a new run, or discharge the spent catalyst to the waste vessel if it is the last run.
The established protocol proves successful in production of transportation fuels from canola oil in a blend. Good material balances (mass recoveries) are obtained in this study with a mean of 99.35% and a standard deviation of 0.31% from 18 LTU runs (Table 2). The conversions and yields from duplicate runs at 11.25 C/O ratio for each feed at a given temperature are quite reproducible (Table 2). Several typical FCC phenomena and cracking characteristics often reported in literature were also observed in this study: (1) Catalyst poisoning by feed basic nitrogen22-24, particularly pronounced for HGO at low temperatures (490 °C in this work). The effect can be reduced at higher temperatures or C/O ratios; (2) Availability or exhaustion of crackable components in the feed and liquid product. Those in the feed are usually called "gasoline precursors", defined as the sum of saturates and monoaromatics20,25-28; (3) Accessibility of catalyst acid sites to molecules; for example, incomplete decomposition of bulky triglyceride molecules at 490 °C while their broken linear fatty acids can easily penetrate catalyst pores and be cracked6; (4) Oligomerization of olefins29 from fatty acids to form aromatics and coke; (5) Preferential skeletal isomerization of olefins to form branched compounds23,29.
The protocol is largely based on the LTU operating manual. The procedures in the manual must be strictly followed, except as noted. Critical steps within the protocol include preparation of equilibrium catalyst (must be on-size and coke-free); reactor preparation (using a feed line that yields a constant injector height, either 1.125 or 2.125 inch); CO2 analyzer calibration; preparation of the syringe (feed rate calibration) and liquid product receiver (weighing the long liquid receiver in a draft-free environment; maintenance of coolant temperature in -12 to -15 °C range); system pressure test (to ensure a leak-free environment); choices of suitable catalyst strip time and liquid strip multiplier; analyses of gases (refinery gas analysis) and liquid product (simulated distillation by ASTM D288718); assignments of molecular weight 86 (versus 89 from ASTM D7964-1430) for C5+ unresolved lump, and carbon-to-coke factor 1.0695 (versus 1.083 from ASTM D7964, which assumes that one mole of hydrogen is associated with one mole of carbon in the coke).
One modification in the protocol that deviates from the LTU operating manual and ASTM D7964-1430 is that in the final step of weighing the liquid receiver, the stopper is quickly removed from and put back in the receiver to equalize the pressure before weighing. This allows release of excess N2, which is trapped at coolant temperature. However, it may also risk the chance of losing some gaseous product. Theoretically, this step may reduce the mass balance by 2.71 mass% for a run at a C/O ratio of 10 in our study, assuming that 149 ml N2 in the liquid receiver is trapped and expanding from -15 to 25 °C at 93.5 kPa (701 mmHg) atmospheric pressure. The result agrees with experimental values (versus decreases in mass balance by 3.10 and 3.22 mass% for runs at C/O ratio of 10).
This protocol has also been extended to blends containing 50 and 100 v% canola oil in HGO. The high rapeseed oil concentration appears to be harmful to the system, requiring more frequent changes of the injector than normal, especially when pure canola oil is cracked. At low concentrations such as the one presented in this study, fouling did not occur.
In addition, the LTU cannot be applied to biomass pyrolysis oil containing emulsified water, which can evaporate at high temperatures for extended periods. In this case, an alternative test unit31 with a free but attachable syringe to deliver the feed is an option12. Also, the LTU cannot determine H2S yield quantitatively due to the water displacement method used to collect gaseous products of which H2S is partially dissolved in water. The alternative test unit modified to accommodate a gasometer consisting of two gas chambers (with pistons inside) in series was found to be satisfactory for this application22,23.
The protocol has also been employed for HGOs from paraffin-rich shale oils from oil shale and light tight oils (LTO) produced by hydraulic fracturing technology. For some of the runs using the said LTU, the simulated distillation results show significant amounts of dissolved gases in the liquid products due to the waxy nature of the feeds, resulting in overestimated gasoline yields and conversions. It is therefore recommended to crack the paraffinic feeds in the alternative test unit mentioned above, which includes a degassing step after condensation of liquid product in the receiver31. The said alternative test unit is widely used to characterize performance of FCC catalysts due to its relative simplicity, flexibility, versatility, and low cost. Over the years, the test method involved has been expanded to provide additional information such as product selectivities and qualities, and the operating variable and feedstock effects32. With adequate precaution on interpretation, test results can be used to assess commercial plant performance33.
Note that the above operational deficiencies using the LTU pertain to our existing particular model. As technology evolves, new products may overcome the problems discussed above.
The authors have nothing to disclose.
The authors wish to thank the analytical laboratory of the CanmetENERGY Technology Centre for its technical support, and Suncor Energy Inc. for supplying the synthetic crude oil. Partial funding for this study was provided by Natural Resources Canada and government of Canada's interdepartmental Program of Energy Research and Development (PERD) with project ID A22.015. Yi Zhang would like to acknowledge his Natural Sciences and Engineering Research Council (NSERC) of Canada Visiting Fellowship from January 2015 to January 2016.
Advanced Cracking Evaluation (ACE) Unit | Kayser Technology Inc. | ACE R+ 46 | Assembled by Zeton Inc. SN:505-46; consisting of (1) a reactor; (2) catalyst addition system; (3) feed delivery system; (4) liquid collection system; (5) gas collection system; (6) gas analyzing system; (7) catalyst regeneration system; (8) CO catalytic convertor; (9) coke analyzing system |
Reactor (ACE) | Kayser Technology Inc. | V-105 | A 1.6 cm ID stainless steel tube having a tapered conical bottom and with a diluent (nitrogen) flowing from the bottom to fluidize the catalyst and also serve as the stripping gas at the end of the run |
Catalyst Addition System (ACE) | Kayser Technology Inc. | Six hoppers (V-120F, with respective valves) for addition of catalyst for up to 6 runs | |
Feed Delivery System (ACE) | Kayser Technology Inc. | Consisting of feed bottle (V-100), syringe (FS-115), pump (P-100), and injector (with 1.125 inch injector height, i.e., the distance from the lowest point of the conical reactor bottom to the bottom end of the feed injector) | |
Liquid Collection System (ACE) | Kayser Technology Inc. | Six liquid receivers (V-110F) immersed in a common coolant bath (Ethylene glycol/water mixture in 50:50 mass ratio) at about –15 °C in a large tank (V-145) | |
Gas Collection System (ACE) | Kayser Technology Inc. | Based on water displacement principle; consisting of gas collection vessel (V-150) with a motor-driven stirrer (MTR-100), and a weight scale (WT-100) for weighing the displaced water collected in a beaker (V100) | |
Gas Analyzing System (ACE) | Kayser Technology Inc. | Key element being Agilent micro GC (model 3000A) with four capillary columns equipped with respective thermal conductivity detectors (TCDs) | |
Catalyst Regeneration System (ACE) | Kayser Technology Inc. | V-105 | Spent catalyst in reactor being burned in situ in air at +700 °C to ensure complete removal of carbon deposited on the catalyst |
CO Catalytic Convertor (ACE) | Kayser Technology Inc. | A reactor (V-140) with CuO as catalyst to oxidize any CO and hydrocarbons in exhausted flue gas to CO2 (to be analyzed by IR gas analyzer) and H2O (to be absorbed by a dryer) | |
Coke Analyzing System (ACE) | Kayser Technology Inc. | Servomex (Model 1440C) IR analyzer for measuring CO2 in exhausted flue gas | |
R+MM Software Suite | Kayser Technology Inc. | Including iFIX 3.5 | |
Agilent Micro GC | Agilent Technologies | 3000A | For gas analysis after cracking |
Cerity Networked Data System | Agilent Technologies | Software for Agilent Micro GC | |
CO2 Gas Analyser | Servomex Inc. | 1440C | SN: 01440C1C02/2900 |
NESLAB Refrigerated Bath | Themo Electron Corporation | RTE 740 | SN: 104300061 |
Orion Sage Syringe Pump | Themo Electron Corporation | M362 | For delivering feed oil to injector tube |
Synthetic Crude Oil (SCO) | Suncor Energy Inc. | Identified as Suncor OSA 10-4.1 | |
Catalyst P | Petro-Canada Refinery | Equilibrium catalyst | |
Balance | Mettler Toledo | AB304-S | For weighing liquid product receivers |
Balance | Mettler Toledo | XS8001S | For weighing water displaced by gas product |
Ethylene Glycol | Fisher Scientifc Inc. | CAS 107-21-1 | Mixed with distilled water as coolant (50 v% ) |
Drierite | W.A. Hammond Drierite Co. Ltd. | 24001 | For water absorption after CO catalytic converter |
Copper Oxide | LECO Corporation | 501-170 | Catalyst for conversion of CO to CO2 |
Toluene | Fisher Scientific Co. | CAS 108-88-3 | For cleaning liquid receivers |
Acetone | Fisher Scientific Co. | CAS 67-64-1 | For cleaning liquid receivers |
Micro GC Calibration Gas | Air Liquid Canada Inc. | SPG-25MX0015306 | Multicomponent standard gas |
19.8% CO2 Standard Gas | BOC Canada Ltd. | 24069890 | For calibration of IR analyzer |
Argon Gas | Linde Canada ltd. | 24001306 | Grade 5.0 Purity |
Helium Gas | Linde Canada ltd. | 24001333 | Grade 5.0 Purity |
Gas analyzer GC Module | Inficon | GCMOD-15 | Channel A |
Gas analyzer GC Module | Inficon | GCMOD-03 | Channel B |
Gas analyzer GC Module | Inficon | GCMOD-04 | Channel C |
Gas analyzer GC Module | Inficon | GCMOD-73 | Channel D |
HP 6890 GC | Hewlett-Packard Co. | G1530A | For simulated distillation |
ASTM 2887 Standard Sample | PAC L.P. | 26650.150 | For quality control in simulated distillation |
ASTM 2887 Standard Sample | PAC L.P. | 25950.200 | For calibration in simulated distillation |
Column for GC 6890 (simulated distillation) | Agilent Technologies | CP7562 | 10m x 0.53mm x 1.2µm, HP 6890 GC column |
Liquid Nitrogen | Air Liquid Canada Inc. | SPG-NIT1AC240LC | For use in simulated distillation |
Nitrogen | Air Liquid Canada Inc. | Bulk (building N2) | For use in ACE unit operation |
Isotemp Programmable Furnace | Thermo Fisher Scientifc Inc. | 10-750-126 | For calcination of catalyst |
GC Vials, Crimp Top | Chromatograghic Specialties Inc | C223682C | 2ml, for liquid product |
Seals, Crimp Top | Chromatograghic Specialties Inc | C221150 | 11 mm, for use with GC vials |
4 oz clear Boston round bottles | Fisher Scientific Co. | 02-911-784 | With PE cone lined caps, for use in feed system |
Sieve | Endecotts Ltd. | 6140269 | Aperture 38 micron |
Sieve | Endecotts Ltd. | 6146265 | Aperture 250 micron |
Shaker | Endecotts Ltd. | MIN 2737-11 | Minor-Meinzer 2 Sieve Shaker for catalyst screening |
V20 Volumetric KF Titrator | Mettler Toledo | 5131025056 | For water content analysis of the liquid product |
Hydranal Composite 5 | Sigma-Aldrich | 34805-1L-R | Reagent for Karl Fischer titration |
Methanol (extremely low water grade) | Fisher Scientific Co. | A413-4 | Mixed with toluene (40:60 w/w) for KF titration: also used to recover water in receiver |
Glass Wool | Fisher Scientific Co. | 11-388 | Placed inside the top of receiver outlet arm |