- Email: [email protected]
Catalytic cracking of bio-oil to organic liquid product (OLP)
Bioresource Technology 101 (2010) 8855–8858
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Catalytic cracking of bio-oil to organic liquid product (OLP) K.L. Hew a, A.M. Tamidi a, S. Yusup a,*, K.T. Lee b, M.M. Ahmad a a b
Chemical Engineering Department, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia School of Chemical Engineering, Universiti Sains Malaysia, Engineering Campus, Seri Ampangan, 14300 Nibong Tebal, Pulau Pinang, Malaysia
a r t i c l e
i n f o
Article history: Received 25 February 2010 Received in revised form 11 May 2010 Accepted 17 May 2010
Keywords: Bio-oil Empty fruit bunch Gasoline Catalytic cracking Taguchi Method
a b s t r a c t The main objective of this paper is to find the optimum operating condition to upgrade the EFB-derived pyrolysis oil (bio-oil) to liquid fuel, mainly gasoline using Taguchi Method. From the analysis that has been done, it is found that the optimum operating condition for heterogeneous catalytic cracking process is at 400 °C, 15 min of reaction time using 30 g of catalyst weight where operating at this condition produced the highest yield of gasoline fraction which is 91.67 wt.%. This observation proves that EFB-derived pyrolysis oil could be upgraded via heterogeneous catalytic cracking to produce gasoline. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Global warming, energy crisis and depletion of world crude oil resources are some of the many reasons why more environmentally friendly solutions must be considered to satisfy the current energy consumption. One of the most promising and potentially least expensive fuels is bio-fuel produced from biomass. One of the ways is by using thermo chemical conversion process such as pyrolysis to convert the solid EFB into liquid form pyrolysis oil (bio-oil). The potential for direct substitution of bio-oil for petroleum and chemical feedstock may be limited due to the high viscosity, high oxygen content and especially the thermal instability of bio-oil. Consequently, upgrading of the bio-oil before use is desirable to give a liquid product that can be used in a wider variety of applications. Furthermore, upgrading process, which is necessary to improve the quality of bio-oil through reduction of oxygenates normally involves process such as catalytic cracking, hydrogenation and steam reforming (Zhu et al., 2005; Zhang et al., 2006; Guo et al., 2004). There are two approaches to catalytic cracking i.e. off line catalytic cracking that utilizes bio-oil as raw material and online catalytic cracking which utilizes pyrolysis vapours as raw material (Bao et al., 2006; Guo and Yan, 2006; Sharma and Bakhshi, 1993a,b; Hyun et al., 2006; Adam et al., 2006; Nokkosmaki et al., 2000). Proof-of-principle studies demonstrated that bio-oil could be converted to gasoline with reasonable hydrogen consumption
* Corresponding author. Tel.: +60 5 3687642; fax: +60 5 3688151. E-mail address: [email protected] (S. Yusup). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.05.036
using mild hydrotreating followed by hydrocracking (Babu and Chaurasia, 2002). Twaiq et al. (2003) have used catalyst such as a composite of microporous HZSM-5 coated with layer of siliceous mesoporous crystalline material MCM-41 to convert palm oil to gasoline. The palm oil conversion is found to be 95.4%, gasoline yield of 47 wt.% and the aromatics composition in the OLP was about 52.2%. In addition direct conversion of biomass derived fast pyrolysis vapours through in situ catalytic upgrading was investigated by Judit et al. (2006). Their findings show that, the hydrocarbon and phenol yields increased in the organic phase while the carbonyl and acid yield decreased. Mesoporous catalysts were used to convert the pyrolysis vapours of spruce wood in order to obtain improved bio-oil properties. Limited study has been conducted to explore the potential of converting bio-oil derived from empty fruit bunch to gasoline or organic liquid vapour. In this paper, bio-oil obtained from the pyrolysis of empty fruit bunch is subsequently subjected to off line heterogeneous catalytic cracking process as an approach to upgrade the bio-oil following Taguchi L9 method for the design of experiment. The usage of Taguchi Method is significant in order to find the optimum operating condition for upgrading the bio-oil. 2. Methods 2.1. Upgrading of bio-oil The bio-oil derived from empty fruit bunch obtained from local supplier was upgraded through catalytic cracking using a High Pressure Reactor. This unit is operated manually and main operating parameters can be adjusted at any time. The catalyst is circu-
8856
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858
Table 1 L9 orthogonal array for catalytic cracking process. Trial
Temperature (°C)
Reaction time (min)
Weight of catalyst (g)
1 2 3 4 5 6 7 8 9
350 350 350 400 400 400 450 450 450
0.5 15 30 0.5 15 30 0.5 15 30
10 20 30 20 30 10 30 10 20
alcohol and carboxylic chains. In this work the product of interest is gasoline and only the gasoline fraction was measured and recorded. The distillate samples were then analyzed using gas chromatography mass spectrometry (GC-MS) analyzer (Shimadzu 210, FID, TCD mass spectrometer) in order to confirm the product quality as compared to the commercial gasoline. The equations used to calculate the yield of OLP and the yield of gasoline are as follow:
YieldOLP ¼
WeightOLP 100% Weightbio-oil
Yieldgasoline ¼ lated using nitrogen as the carrier gas with purity of 99.9% from MOX-Linde Gases Sdn. Bhd. Taguchi L9 method was adapted for experimental approach of the catalytic cracking process (Ross, 1996; Roy, 2001). Three factors with three levels were selected and the selected orthogonal array for the process of upgrading the bio-oil is shown in Table 1. These factors and their respective levels were chosen based on previous studies (Tamunaidu and Bhatia, 2007). The maximum temperature was set at 450 °C because above 500 °C, the liquid will be converted to gas and below 350 °C cracking of bio-oil to liquid product is ineffective. Reaction time is also a significant factor in ensuring effective catalytic process to take place, Zhang et al. (2004). If the reaction time exceeds the optimum time, the majority of the liquid will be further crack to other product such as kerosene and gas. 250 ml of bio-oil was filled into the liner of the high pressure reactor. Zeolite catalyst, ZSM-5 was obtained from Euramco (M) Sdn Bhd, was weighed according to values indicated in Table 1 and mixed together with the bio-oil in the liner. The reactor was securely covered and purged for 5 min with nitrogen gas. The reactor was heated up until the desired temperature was reached. All experiments of bio-oil cracking were conducted for a specified period with stabilized nitrogen gas flow rate at 1L/hour. The reactants were left to react for the certain period of reaction time with constant stirring at 7 Hz. The product was then cooled down before being removed from the reactor. The liquid product (organic and aqueous fractions) was collected in a liquid sampler while the gaseous product was collected in a gas sampling beg but the amount is too small to be quantified. The aqueous phase was separated from condensed liquid products using a syringe. The OLP produced from the catalytic upgrading process was then weighed to determine the yield of OLP. About 12 g of the OLP was weighted and distilled using a micro lab scale distillation unit at atmospheric pressures with boiling range of 60– 120 °C for an hour. It is anticipated that heavy liquid product which contains kerosene and diesel would be presence if the boiling range is increased from 170–250 °C (Ahget and Bodiuszynski, 1994). The aqueous fraction is expected to contain mainly water and some organic components that are soluble in water such as
ð1Þ
Weightgasoline 100% 12 g
ð2Þ
2.2. Characterization of gasoline fraction The fraction of gasoline produced from distillation was analyzed in order to obtain the necessary physical and chemical properties and the result was compared with the commercial gasoline as a standard (RON 95). The density was measured using a calibrated pycnometer (Jaytech, UK) following standard test method for liquid density. The chemical composition of the gasoline fraction, mainly carbon (C), oxygen (O) and hydrogen (H) was determined using Carbon, Hydrogen, Nitrogen and Sulfur (CHNS) Analyzer, LECO (Model 932) CHNS Analyzer. 3. Result and discussion Table 2 summarized the results obtained pertaining to the yield of OLP and gasoline for all the trials based on Taguchi L9 orthogonal array as outlined in Table 1. For the OLP yield the standard deviations vary between 0.35 to 2.57% depending upon the trials. Comparatively the standard deviations for gasoline yield vary between 0–13.55% depending upon trial runs. The highest yield of OLP is obtained in Trial 1 where the optimum operating conditions are at 350 °C, 30s of reaction time and 10 g of catalyst weight. For gasoline, the highest yield is obtained in Trial 5, in which the optimum operating conditions are at 400 °C, 15 min of reaction time and 30 g of catalyst utilized. Further analysis using Signal to Noise Ratio (S/N) is conducted to obtain the exact optimum operating condition to yield both OLP and gasoline considering the noise effect. Two ways ANOVA with replication was used for statistical analysis of the data. The results were summarized in the ANOVA tables as shown in Table 3(a) and (b). From the F Tables (Ranjit, 1990) at 99% confidence, the obtained F value is 7.21. The F columns in Table 3(a) and (b), show that reaction temperature has higher calculated F value than the F value obtained from the F Tables. This shows that with 99% confidence, temperature is the main contributor that influences the yield of OLP and gasoline.
Table 2 Yield of OLP and gasoline. Trial
1 2 3 4 5 6 7 8 9
Yield of OLP (%)
Yield of gasoline (%)
Run 1
Run 2
Avg values
Standard deviation
Run 1
Run 2
Avg values
Standard deviation
92.3 88.6 85.63 80.28 79.53 78.6 76.43 77.8 72.18
90.1 89.2 82 83.4 81.07 79.1 74.9 76.98 82.37
91.20 88.90 83.82 81.84 80.30 78.85 75.66 77.39 77.28
1.55 0.42 2.57 2.21 1.09 0.35 1.08 0.58 7.21
31.67 38.33 48.33 46.67 91.67 95 40.83 32.5 30.83
34.17 42.5 38.75 46.67 90 75.83 37.5 34.17 35
32.92 40.42 43.54 46.67 90.84 85.42 39.17 33.34 32.92
1.77 2.95 6.78 0 1.18 13.55 2.36 1.18 2.95
8857
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858 Table 3 ANOVA (a) Yield of OLP (b) Yield of gasoline. Source of variance
Degree of freedom (f)
Sum of square (S)
Variance (V)
Variance ratio (F)
Pure sum of square (S’)
Percent contribution (P)
(a) Temperature (A) Reaction time (B) Catalyst weight (C) Error (e) Total
2 2 2 11 17
394.84 27.31 27.59 81.47 531.20
197.42 13.66 13.79 7.41
26.66 1.84 1.86 1.00
380.02 12.50 12.77 125.90
71.54 2.35 2.40 23.70 100
(b) Temperature (A) Reaction time (B) Catalyst weight (C) Error (e) Total
2 2 2 11 17
18.87 3.28 4.13 2.89 29.16
9.43 1.64 2.06 0.26
35.89 6.24 7.85 1.00
18.34 2.75 3.60 4.47
62.89 9.44 12.35 15.32 100
Table 3 (a) shows that the percentage yield of OLP is influenced largely by temperature in which the percentage of contribution of this factor is 71.54%. An increase in temperature will induce vapourization of liquid product to gas which results in decrease yield of OLP as shown in Table 2. In addition the catalyst activity decreased when subjected to higher temperature and this may contribute to decrease in the OLP yield. The collapses of primary structure of the catalyst and changes in catalytic properties might occur which resulted in decreased in cracking activity of the catalyst and thus decreased in the value of yield obtained (Liu et al., 1991). Similar effect of temperature on gasoline yield was observed as shown in Table 3(b). Temperature has the strong influence on the yield of gasoline obtained in which the percentage of contribution is 62.89%. Twaiq et al. (2003) in their study on catalytic cracking of palm oil to liquid fuels using HZSM-5 found that the optimum yield of OLP obtained is 91.6% and gasoline is 40.5% in which the quantity obtained is comparable to this work. The optimum OLP and gasoline yields obtained through this study are 91.20% and 46.67% respectively. From Table 3a and b, the contribution of reaction time towards OLP and gasoline yield is less than 10%. But the OLP yield is observed to decrease with reaction time as shown in Table 3a. It is anticipated that conversion of liquid to gas occurs due to secondary cracking process. Contrary, to OLP yield, the gasoline yield is observed to increase when reaction time is increased from 0.5 min to 15 min and decrease slightly when reaction time is further increased to 30 min. More OLP is converted to gasoline as reaction progresses. In addition catalyst weight is determined to be an important factor that contributes to increase in gasoline yield. It contributes about 12.35% as highlighted in Table 3(b). Optimum amount of catalyst is crucial to ensure optimum OLP and gasoline yield. If the quantity of the catalyst is higher than the optimum value in the case for OLP, the yield is observed to decrease due to further secondary cracking reaction that occurs as shown in Trial 3 and 7 in
Table 2. Gasoline yield is higher when 30 g of catalyst is employed. More OLP is converted to gasoline. Signal to Noise Ratio (S/N) is a variance index that allow us to identify the most optimum condition for the process by selecting the highest S/N value. This is because, the greater the value, the smaller the product variance to the target value (Ranjit, 1990). The equation to calculate S/N is:
S ¼ 10 Log 10 ðMSDÞ N
where, (MSD) stands for ‘‘Mean Squared Deviation” from the target value of the quality characteristic. In this paper, both of the determining factors which are percentage yield of OLP and percentage yield of gasoline were assigned to the quality characteristic of ‘‘The bigger the better”. For this quality characteristic, MSD is defined as:
MSD ¼
*
Temperature (°C)
Time (min)
Catalyst weight (g)
1 2 3 4 5 6 7 8 9
350 350 350 400 400 400 450 450 450
0.5 15 30 0.5 15 30 0.5 15 30
10 20 30 20 30 10 30 10 20
Optimum condition for S/N OLP and gasoline yield.
Gasoline
39.20* 38.98 38.47 38.25 38.09 37.94 37.58 37.77 37.70
4.10 6.06 7.09 7.81 14.25* 13.80 7.46 5.95 5.44
n
ð4Þ
Table 5 Analysis of gasoline fraction (a) Comparison of retention time between commercial gasoline (RON 95) and produced gasoline (b) Comparison of physical and chemical properties. Peak No.
Signal to Noise Ratio (S/N) OLP
1 1 1 þ þ þ ... y21 y22 y23
where n = number of repetition, y1, y2, etc. results of experiments. *Source, Ranjit, 1990.
Table 4 Signal to noise ratio of yield of OLP and yield of gasoline. Trial
ð3Þ
*
Retention time (min) Commercial gasoline (RON 97)
Produced gasoline
(a) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
1.695 1.869 2.014 2.231 2.355 2.735 3.259 3.397 4.248 4.645 5.007 5.007 5.428 6.663 6.663 8.795 10.992 13.011
1.707 1.872 2.006 2.24 2.318 2.646 3.271 3.388 4.199 4.678 5.088 5.213 5.383 6.542 6.672 8.718 10.972 12.96
(b) Properties
Commercial gasoline*
Gasoline fraction
Composition C H O Density @ 15 °C Boiling point
80–88 wt.% 12–15 wt.% 0 wt.% 719.7–779.68 (kg m3) 60–120 °C
60–75 wt.% 18–23 wt.% 0.5–1.2 wt.% 815.66–851.65 (kg m3) 80–120 °C
Commercial gasoline source: http://www.eere.energy.gov.
8858
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858
Table 4 shows the calculated S/N for yield of OLP and yield of gasoline. From Table 4, it is observed that the highest S/N ratio for the yield of OLP was achieved at 350 °C, 0.5 min reaction time using 10 g of catalyst while for the yield of gasoline; the highest S/ N ratio was achieved at 400 °C, 15 min reaction time using 30 g of catalyst. Both of these conditions are the optimum conditions for the catalytic cracking of the bio-oil. These results confirmed the analysis obtained from the main effect shown in Table 2. Table 5(a) shows the GC-MS analysis for the optimum condition where the highest yield of gasoline is produced. The analysis shows that the gasoline fraction produced contained mainly saturated hydrocarbon (peak 1–peak 11) and alkyl-benzenes (peak 12–peak 17). The retention time produced for each peak was then compared with the retention time of commercial gasoline (RON 97). From Table 5(a), it is observed that the difference between retention time of commercial gasoline and the produced gasoline fraction is insignificant. Thus, it is concluded that the gasoline fraction produced is comparable to the commercial gasoline (RON 97). Table 5(b) shows the physical and chemical properties of commercial gasoline and gasoline fraction produced from catalytic cracking. Based on these results, both fuels have almost similar properties. 4. Conclusion Zeolite ZSM-5 catalyst is a potential catalyst for conversion of bio-oil to gasoline through the catalytic cracking process. Proper control of experimental condition is essential in achieving highest yield of gasoline. To achieve this, Taguchi design of experiment has been utilized and from the analysis it is found that the optimum operating condition for the catalytic cracking process is at 400 °C, reaction time of 15 min and utilizing catalyst weight of 30 g that results in the highest yield of gasoline fraction obtained, which is 91.67%. In addition, side product is also obtained which has the potential to be upgraded to other types of fuel such as kerosene and diesel. Acknowledgement The authors would like to acknowledge Universiti Teknologi PETRONAS for the support given in conducting this research.
References Adam, J., Antonakou, E., Lappas, A., Stocker, M., Nilsen, M.H., Bouzga, A., Hustad, J.E., Gisle, O., 2006. In situ catalytic upgrading of biomass derived fast pyrolysis vapours in a fixed bed reactor using mesoporous materials. Microporous and Mesoporous Materials 96 (1–3), 93–101. Ahget, K.H., Bodiuszynski, M.M.B., 1994. Composition and Analysis of Heavy Petroleum Fraction. Chemical Industries Series. 54. New York, Marcel Dekker. Babu, B.V., Chaurasia, A.S., 2002. Modeling for pyrolysis: Kinetic and heat transfer effects. Energy Conversion and Management 44, 2251–2275. Bao, W.R., Xue, X.L., Cao, Q., Lu, J.J., Lu, Y.K., 2006. Study on biomass pyrolytic liquid products with MCM-41/SBA-15 as catalyst. Journal of Fuel Chemistry and Technology 34 (6), 675–679. Guo, X.Y., Yan, Y.J., 2006. Analysis of the coke precursor on bio-oil refining catalyst and the regeneration of the deactivated catalyst. Journal of Chemical Engineering of Chinese Universities 20 (2), 222–226. Guo, X.Y., Yan, Y.J., Li, T.C., 2004. Influence of catalyst type and regeneration on upgrading of crude bio-oil through catalytical thermal cracking. The Chinese Journal of Process Engineering 4 (1), 53–58. http://www.eere.energy.gov (Accessed April 2010).. Hyun, J.P., Young-Kwon, P., Joo-Sik, K., Jong-ki, J., Kyung-Seun, Y., Jin-Heong, Y., Jinh, J., Jung, M.S., 2006. Bio-oil upgrading over Ga modified zeolites in a bubbling fluidized bed reactor. Studies in Surface Science and Catalysis 159, 553–556. Judit, A., Eleni, A., Angelos, L., Michael, S., Meret, H.N., Aud, B., Johan, E.H., Gisle, O., 2006. In situ catalytic upgrading of biomass derived fast pyrolysis vapours in fixed bed reactors using mesoporous materials. Microporous and Mesoporous Materials 96, 93–101. Liu, S.B., Wu, J.F., Ma, L.J., Tsai, T.C., Wang, I., 1991. On the thermal stability of zeolite beta. Journal of Catalysis 132, 432–439. Nokkosmaki, M.I., Kuoppala, E.T., Leppamaki, E.A., Krause, A.O.I., 2000. Catalytic conversion of biomass pyrolysis vapours with zinc oxide. Journal of Analytical and Applied Pyrolysis 55 (1), 119–131. Ranjit R, 1990. A Primer on the Taguchi Method. Society of Manufacturing Engineer. Ross, P.J., 1996. Taguchi Techniques for Quality Engineering. McGraw-Hill Companies, Inc.. Roy, R.K., 2001. Design of experiments using the Taguchi approach: 16 steps to product and process improvement. John Wiley and Sons, Inc.. Sharma, R.K., Bakhshi, N.N., 1993a. Catalytic conversion of fast pyrolysis oil to hydrocarbon fuels over HZSM-5 in a dual reactor system. Biomass and Bioenergy 5 (6), 445–455. Sharma, R.K., Bakhshi, N.N., 1993b. Conversion of non-phenolic fraction of biomassderived pyrolysis oil to hydrocarbon fuels over HZSM-5 using a dual reactor system. Bioresource Technology 45 (3), 195–203. Tamunaidu, P., Bhatia, S., 2007. Catalytic cracking of palm oil for the production of biofuel. Bioresource Energy 98, 3593–3601. Twaiq, F.A., Zabidi, N.A.M., Mohamed, A.R., Bathia, S., 2003. Catalytic conversion of palm oil over mesoporous aluminosilicate MCM-41 for the production of liquid hydrocarbon fuels. Fuel Processing Technology 1660, 1–16. Zhang, Q., Chang, J., Wang, T.J., Xu, Y., 2006. Progress on research of properties and upgrading of bio-oil. Petrochemical Technology 35 (5), 493–498. Zhang, S., Yan, Y., Li, T., Ren, Z., 2004. Upgrading of liquid fuel from the pyrolysis of biomass. Bioresource Technology 96, 545–550. Zhu, X.F., Zheng, J.L., Guo, Q.X., Zhu, Q.S., 2005. Upgrading and utilization of bio-oil from biomass. Engineering Science 7 (9), 83–88.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Catalytic cracking of bio-oil to organic liquid product (OLP) K.L. Hew a, A.M. Tamidi a, S. Yusup a,*, K.T. Lee b, M.M. Ahmad a a b
Chemical Engineering Department, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia School of Chemical Engineering, Universiti Sains Malaysia, Engineering Campus, Seri Ampangan, 14300 Nibong Tebal, Pulau Pinang, Malaysia
a r t i c l e
i n f o
Article history: Received 25 February 2010 Received in revised form 11 May 2010 Accepted 17 May 2010
Keywords: Bio-oil Empty fruit bunch Gasoline Catalytic cracking Taguchi Method
a b s t r a c t The main objective of this paper is to find the optimum operating condition to upgrade the EFB-derived pyrolysis oil (bio-oil) to liquid fuel, mainly gasoline using Taguchi Method. From the analysis that has been done, it is found that the optimum operating condition for heterogeneous catalytic cracking process is at 400 °C, 15 min of reaction time using 30 g of catalyst weight where operating at this condition produced the highest yield of gasoline fraction which is 91.67 wt.%. This observation proves that EFB-derived pyrolysis oil could be upgraded via heterogeneous catalytic cracking to produce gasoline. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Global warming, energy crisis and depletion of world crude oil resources are some of the many reasons why more environmentally friendly solutions must be considered to satisfy the current energy consumption. One of the most promising and potentially least expensive fuels is bio-fuel produced from biomass. One of the ways is by using thermo chemical conversion process such as pyrolysis to convert the solid EFB into liquid form pyrolysis oil (bio-oil). The potential for direct substitution of bio-oil for petroleum and chemical feedstock may be limited due to the high viscosity, high oxygen content and especially the thermal instability of bio-oil. Consequently, upgrading of the bio-oil before use is desirable to give a liquid product that can be used in a wider variety of applications. Furthermore, upgrading process, which is necessary to improve the quality of bio-oil through reduction of oxygenates normally involves process such as catalytic cracking, hydrogenation and steam reforming (Zhu et al., 2005; Zhang et al., 2006; Guo et al., 2004). There are two approaches to catalytic cracking i.e. off line catalytic cracking that utilizes bio-oil as raw material and online catalytic cracking which utilizes pyrolysis vapours as raw material (Bao et al., 2006; Guo and Yan, 2006; Sharma and Bakhshi, 1993a,b; Hyun et al., 2006; Adam et al., 2006; Nokkosmaki et al., 2000). Proof-of-principle studies demonstrated that bio-oil could be converted to gasoline with reasonable hydrogen consumption
* Corresponding author. Tel.: +60 5 3687642; fax: +60 5 3688151. E-mail address: [email protected] (S. Yusup). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.05.036
using mild hydrotreating followed by hydrocracking (Babu and Chaurasia, 2002). Twaiq et al. (2003) have used catalyst such as a composite of microporous HZSM-5 coated with layer of siliceous mesoporous crystalline material MCM-41 to convert palm oil to gasoline. The palm oil conversion is found to be 95.4%, gasoline yield of 47 wt.% and the aromatics composition in the OLP was about 52.2%. In addition direct conversion of biomass derived fast pyrolysis vapours through in situ catalytic upgrading was investigated by Judit et al. (2006). Their findings show that, the hydrocarbon and phenol yields increased in the organic phase while the carbonyl and acid yield decreased. Mesoporous catalysts were used to convert the pyrolysis vapours of spruce wood in order to obtain improved bio-oil properties. Limited study has been conducted to explore the potential of converting bio-oil derived from empty fruit bunch to gasoline or organic liquid vapour. In this paper, bio-oil obtained from the pyrolysis of empty fruit bunch is subsequently subjected to off line heterogeneous catalytic cracking process as an approach to upgrade the bio-oil following Taguchi L9 method for the design of experiment. The usage of Taguchi Method is significant in order to find the optimum operating condition for upgrading the bio-oil. 2. Methods 2.1. Upgrading of bio-oil The bio-oil derived from empty fruit bunch obtained from local supplier was upgraded through catalytic cracking using a High Pressure Reactor. This unit is operated manually and main operating parameters can be adjusted at any time. The catalyst is circu-
8856
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858
Table 1 L9 orthogonal array for catalytic cracking process. Trial
Temperature (°C)
Reaction time (min)
Weight of catalyst (g)
1 2 3 4 5 6 7 8 9
350 350 350 400 400 400 450 450 450
0.5 15 30 0.5 15 30 0.5 15 30
10 20 30 20 30 10 30 10 20
alcohol and carboxylic chains. In this work the product of interest is gasoline and only the gasoline fraction was measured and recorded. The distillate samples were then analyzed using gas chromatography mass spectrometry (GC-MS) analyzer (Shimadzu 210, FID, TCD mass spectrometer) in order to confirm the product quality as compared to the commercial gasoline. The equations used to calculate the yield of OLP and the yield of gasoline are as follow:
YieldOLP ¼
WeightOLP 100% Weightbio-oil
Yieldgasoline ¼ lated using nitrogen as the carrier gas with purity of 99.9% from MOX-Linde Gases Sdn. Bhd. Taguchi L9 method was adapted for experimental approach of the catalytic cracking process (Ross, 1996; Roy, 2001). Three factors with three levels were selected and the selected orthogonal array for the process of upgrading the bio-oil is shown in Table 1. These factors and their respective levels were chosen based on previous studies (Tamunaidu and Bhatia, 2007). The maximum temperature was set at 450 °C because above 500 °C, the liquid will be converted to gas and below 350 °C cracking of bio-oil to liquid product is ineffective. Reaction time is also a significant factor in ensuring effective catalytic process to take place, Zhang et al. (2004). If the reaction time exceeds the optimum time, the majority of the liquid will be further crack to other product such as kerosene and gas. 250 ml of bio-oil was filled into the liner of the high pressure reactor. Zeolite catalyst, ZSM-5 was obtained from Euramco (M) Sdn Bhd, was weighed according to values indicated in Table 1 and mixed together with the bio-oil in the liner. The reactor was securely covered and purged for 5 min with nitrogen gas. The reactor was heated up until the desired temperature was reached. All experiments of bio-oil cracking were conducted for a specified period with stabilized nitrogen gas flow rate at 1L/hour. The reactants were left to react for the certain period of reaction time with constant stirring at 7 Hz. The product was then cooled down before being removed from the reactor. The liquid product (organic and aqueous fractions) was collected in a liquid sampler while the gaseous product was collected in a gas sampling beg but the amount is too small to be quantified. The aqueous phase was separated from condensed liquid products using a syringe. The OLP produced from the catalytic upgrading process was then weighed to determine the yield of OLP. About 12 g of the OLP was weighted and distilled using a micro lab scale distillation unit at atmospheric pressures with boiling range of 60– 120 °C for an hour. It is anticipated that heavy liquid product which contains kerosene and diesel would be presence if the boiling range is increased from 170–250 °C (Ahget and Bodiuszynski, 1994). The aqueous fraction is expected to contain mainly water and some organic components that are soluble in water such as
ð1Þ
Weightgasoline 100% 12 g
ð2Þ
2.2. Characterization of gasoline fraction The fraction of gasoline produced from distillation was analyzed in order to obtain the necessary physical and chemical properties and the result was compared with the commercial gasoline as a standard (RON 95). The density was measured using a calibrated pycnometer (Jaytech, UK) following standard test method for liquid density. The chemical composition of the gasoline fraction, mainly carbon (C), oxygen (O) and hydrogen (H) was determined using Carbon, Hydrogen, Nitrogen and Sulfur (CHNS) Analyzer, LECO (Model 932) CHNS Analyzer. 3. Result and discussion Table 2 summarized the results obtained pertaining to the yield of OLP and gasoline for all the trials based on Taguchi L9 orthogonal array as outlined in Table 1. For the OLP yield the standard deviations vary between 0.35 to 2.57% depending upon the trials. Comparatively the standard deviations for gasoline yield vary between 0–13.55% depending upon trial runs. The highest yield of OLP is obtained in Trial 1 where the optimum operating conditions are at 350 °C, 30s of reaction time and 10 g of catalyst weight. For gasoline, the highest yield is obtained in Trial 5, in which the optimum operating conditions are at 400 °C, 15 min of reaction time and 30 g of catalyst utilized. Further analysis using Signal to Noise Ratio (S/N) is conducted to obtain the exact optimum operating condition to yield both OLP and gasoline considering the noise effect. Two ways ANOVA with replication was used for statistical analysis of the data. The results were summarized in the ANOVA tables as shown in Table 3(a) and (b). From the F Tables (Ranjit, 1990) at 99% confidence, the obtained F value is 7.21. The F columns in Table 3(a) and (b), show that reaction temperature has higher calculated F value than the F value obtained from the F Tables. This shows that with 99% confidence, temperature is the main contributor that influences the yield of OLP and gasoline.
Table 2 Yield of OLP and gasoline. Trial
1 2 3 4 5 6 7 8 9
Yield of OLP (%)
Yield of gasoline (%)
Run 1
Run 2
Avg values
Standard deviation
Run 1
Run 2
Avg values
Standard deviation
92.3 88.6 85.63 80.28 79.53 78.6 76.43 77.8 72.18
90.1 89.2 82 83.4 81.07 79.1 74.9 76.98 82.37
91.20 88.90 83.82 81.84 80.30 78.85 75.66 77.39 77.28
1.55 0.42 2.57 2.21 1.09 0.35 1.08 0.58 7.21
31.67 38.33 48.33 46.67 91.67 95 40.83 32.5 30.83
34.17 42.5 38.75 46.67 90 75.83 37.5 34.17 35
32.92 40.42 43.54 46.67 90.84 85.42 39.17 33.34 32.92
1.77 2.95 6.78 0 1.18 13.55 2.36 1.18 2.95
8857
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858 Table 3 ANOVA (a) Yield of OLP (b) Yield of gasoline. Source of variance
Degree of freedom (f)
Sum of square (S)
Variance (V)
Variance ratio (F)
Pure sum of square (S’)
Percent contribution (P)
(a) Temperature (A) Reaction time (B) Catalyst weight (C) Error (e) Total
2 2 2 11 17
394.84 27.31 27.59 81.47 531.20
197.42 13.66 13.79 7.41
26.66 1.84 1.86 1.00
380.02 12.50 12.77 125.90
71.54 2.35 2.40 23.70 100
(b) Temperature (A) Reaction time (B) Catalyst weight (C) Error (e) Total
2 2 2 11 17
18.87 3.28 4.13 2.89 29.16
9.43 1.64 2.06 0.26
35.89 6.24 7.85 1.00
18.34 2.75 3.60 4.47
62.89 9.44 12.35 15.32 100
Table 3 (a) shows that the percentage yield of OLP is influenced largely by temperature in which the percentage of contribution of this factor is 71.54%. An increase in temperature will induce vapourization of liquid product to gas which results in decrease yield of OLP as shown in Table 2. In addition the catalyst activity decreased when subjected to higher temperature and this may contribute to decrease in the OLP yield. The collapses of primary structure of the catalyst and changes in catalytic properties might occur which resulted in decreased in cracking activity of the catalyst and thus decreased in the value of yield obtained (Liu et al., 1991). Similar effect of temperature on gasoline yield was observed as shown in Table 3(b). Temperature has the strong influence on the yield of gasoline obtained in which the percentage of contribution is 62.89%. Twaiq et al. (2003) in their study on catalytic cracking of palm oil to liquid fuels using HZSM-5 found that the optimum yield of OLP obtained is 91.6% and gasoline is 40.5% in which the quantity obtained is comparable to this work. The optimum OLP and gasoline yields obtained through this study are 91.20% and 46.67% respectively. From Table 3a and b, the contribution of reaction time towards OLP and gasoline yield is less than 10%. But the OLP yield is observed to decrease with reaction time as shown in Table 3a. It is anticipated that conversion of liquid to gas occurs due to secondary cracking process. Contrary, to OLP yield, the gasoline yield is observed to increase when reaction time is increased from 0.5 min to 15 min and decrease slightly when reaction time is further increased to 30 min. More OLP is converted to gasoline as reaction progresses. In addition catalyst weight is determined to be an important factor that contributes to increase in gasoline yield. It contributes about 12.35% as highlighted in Table 3(b). Optimum amount of catalyst is crucial to ensure optimum OLP and gasoline yield. If the quantity of the catalyst is higher than the optimum value in the case for OLP, the yield is observed to decrease due to further secondary cracking reaction that occurs as shown in Trial 3 and 7 in
Table 2. Gasoline yield is higher when 30 g of catalyst is employed. More OLP is converted to gasoline. Signal to Noise Ratio (S/N) is a variance index that allow us to identify the most optimum condition for the process by selecting the highest S/N value. This is because, the greater the value, the smaller the product variance to the target value (Ranjit, 1990). The equation to calculate S/N is:
S ¼ 10 Log 10 ðMSDÞ N
where, (MSD) stands for ‘‘Mean Squared Deviation” from the target value of the quality characteristic. In this paper, both of the determining factors which are percentage yield of OLP and percentage yield of gasoline were assigned to the quality characteristic of ‘‘The bigger the better”. For this quality characteristic, MSD is defined as:
MSD ¼
*
Temperature (°C)
Time (min)
Catalyst weight (g)
1 2 3 4 5 6 7 8 9
350 350 350 400 400 400 450 450 450
0.5 15 30 0.5 15 30 0.5 15 30
10 20 30 20 30 10 30 10 20
Optimum condition for S/N OLP and gasoline yield.
Gasoline
39.20* 38.98 38.47 38.25 38.09 37.94 37.58 37.77 37.70
4.10 6.06 7.09 7.81 14.25* 13.80 7.46 5.95 5.44
n
ð4Þ
Table 5 Analysis of gasoline fraction (a) Comparison of retention time between commercial gasoline (RON 95) and produced gasoline (b) Comparison of physical and chemical properties. Peak No.
Signal to Noise Ratio (S/N) OLP
1 1 1 þ þ þ ... y21 y22 y23
where n = number of repetition, y1, y2, etc. results of experiments. *Source, Ranjit, 1990.
Table 4 Signal to noise ratio of yield of OLP and yield of gasoline. Trial
ð3Þ
*
Retention time (min) Commercial gasoline (RON 97)
Produced gasoline
(a) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
1.695 1.869 2.014 2.231 2.355 2.735 3.259 3.397 4.248 4.645 5.007 5.007 5.428 6.663 6.663 8.795 10.992 13.011
1.707 1.872 2.006 2.24 2.318 2.646 3.271 3.388 4.199 4.678 5.088 5.213 5.383 6.542 6.672 8.718 10.972 12.96
(b) Properties
Commercial gasoline*
Gasoline fraction
Composition C H O Density @ 15 °C Boiling point
80–88 wt.% 12–15 wt.% 0 wt.% 719.7–779.68 (kg m3) 60–120 °C
60–75 wt.% 18–23 wt.% 0.5–1.2 wt.% 815.66–851.65 (kg m3) 80–120 °C
Commercial gasoline source: http://www.eere.energy.gov.
8858
K.L. Hew et al. / Bioresource Technology 101 (2010) 8855–8858
Table 4 shows the calculated S/N for yield of OLP and yield of gasoline. From Table 4, it is observed that the highest S/N ratio for the yield of OLP was achieved at 350 °C, 0.5 min reaction time using 10 g of catalyst while for the yield of gasoline; the highest S/ N ratio was achieved at 400 °C, 15 min reaction time using 30 g of catalyst. Both of these conditions are the optimum conditions for the catalytic cracking of the bio-oil. These results confirmed the analysis obtained from the main effect shown in Table 2. Table 5(a) shows the GC-MS analysis for the optimum condition where the highest yield of gasoline is produced. The analysis shows that the gasoline fraction produced contained mainly saturated hydrocarbon (peak 1–peak 11) and alkyl-benzenes (peak 12–peak 17). The retention time produced for each peak was then compared with the retention time of commercial gasoline (RON 97). From Table 5(a), it is observed that the difference between retention time of commercial gasoline and the produced gasoline fraction is insignificant. Thus, it is concluded that the gasoline fraction produced is comparable to the commercial gasoline (RON 97). Table 5(b) shows the physical and chemical properties of commercial gasoline and gasoline fraction produced from catalytic cracking. Based on these results, both fuels have almost similar properties. 4. Conclusion Zeolite ZSM-5 catalyst is a potential catalyst for conversion of bio-oil to gasoline through the catalytic cracking process. Proper control of experimental condition is essential in achieving highest yield of gasoline. To achieve this, Taguchi design of experiment has been utilized and from the analysis it is found that the optimum operating condition for the catalytic cracking process is at 400 °C, reaction time of 15 min and utilizing catalyst weight of 30 g that results in the highest yield of gasoline fraction obtained, which is 91.67%. In addition, side product is also obtained which has the potential to be upgraded to other types of fuel such as kerosene and diesel. Acknowledgement The authors would like to acknowledge Universiti Teknologi PETRONAS for the support given in conducting this research.
References Adam, J., Antonakou, E., Lappas, A., Stocker, M., Nilsen, M.H., Bouzga, A., Hustad, J.E., Gisle, O., 2006. In situ catalytic upgrading of biomass derived fast pyrolysis vapours in a fixed bed reactor using mesoporous materials. Microporous and Mesoporous Materials 96 (1–3), 93–101. Ahget, K.H., Bodiuszynski, M.M.B., 1994. Composition and Analysis of Heavy Petroleum Fraction. Chemical Industries Series. 54. New York, Marcel Dekker. Babu, B.V., Chaurasia, A.S., 2002. Modeling for pyrolysis: Kinetic and heat transfer effects. Energy Conversion and Management 44, 2251–2275. Bao, W.R., Xue, X.L., Cao, Q., Lu, J.J., Lu, Y.K., 2006. Study on biomass pyrolytic liquid products with MCM-41/SBA-15 as catalyst. Journal of Fuel Chemistry and Technology 34 (6), 675–679. Guo, X.Y., Yan, Y.J., 2006. Analysis of the coke precursor on bio-oil refining catalyst and the regeneration of the deactivated catalyst. Journal of Chemical Engineering of Chinese Universities 20 (2), 222–226. Guo, X.Y., Yan, Y.J., Li, T.C., 2004. Influence of catalyst type and regeneration on upgrading of crude bio-oil through catalytical thermal cracking. The Chinese Journal of Process Engineering 4 (1), 53–58. http://www.eere.energy.gov (Accessed April 2010).. Hyun, J.P., Young-Kwon, P., Joo-Sik, K., Jong-ki, J., Kyung-Seun, Y., Jin-Heong, Y., Jinh, J., Jung, M.S., 2006. Bio-oil upgrading over Ga modified zeolites in a bubbling fluidized bed reactor. Studies in Surface Science and Catalysis 159, 553–556. Judit, A., Eleni, A., Angelos, L., Michael, S., Meret, H.N., Aud, B., Johan, E.H., Gisle, O., 2006. In situ catalytic upgrading of biomass derived fast pyrolysis vapours in fixed bed reactors using mesoporous materials. Microporous and Mesoporous Materials 96, 93–101. Liu, S.B., Wu, J.F., Ma, L.J., Tsai, T.C., Wang, I., 1991. On the thermal stability of zeolite beta. Journal of Catalysis 132, 432–439. Nokkosmaki, M.I., Kuoppala, E.T., Leppamaki, E.A., Krause, A.O.I., 2000. Catalytic conversion of biomass pyrolysis vapours with zinc oxide. Journal of Analytical and Applied Pyrolysis 55 (1), 119–131. Ranjit R, 1990. A Primer on the Taguchi Method. Society of Manufacturing Engineer. Ross, P.J., 1996. Taguchi Techniques for Quality Engineering. McGraw-Hill Companies, Inc.. Roy, R.K., 2001. Design of experiments using the Taguchi approach: 16 steps to product and process improvement. John Wiley and Sons, Inc.. Sharma, R.K., Bakhshi, N.N., 1993a. Catalytic conversion of fast pyrolysis oil to hydrocarbon fuels over HZSM-5 in a dual reactor system. Biomass and Bioenergy 5 (6), 445–455. Sharma, R.K., Bakhshi, N.N., 1993b. Conversion of non-phenolic fraction of biomassderived pyrolysis oil to hydrocarbon fuels over HZSM-5 using a dual reactor system. Bioresource Technology 45 (3), 195–203. Tamunaidu, P., Bhatia, S., 2007. Catalytic cracking of palm oil for the production of biofuel. Bioresource Energy 98, 3593–3601. Twaiq, F.A., Zabidi, N.A.M., Mohamed, A.R., Bathia, S., 2003. Catalytic conversion of palm oil over mesoporous aluminosilicate MCM-41 for the production of liquid hydrocarbon fuels. Fuel Processing Technology 1660, 1–16. Zhang, Q., Chang, J., Wang, T.J., Xu, Y., 2006. Progress on research of properties and upgrading of bio-oil. Petrochemical Technology 35 (5), 493–498. Zhang, S., Yan, Y., Li, T., Ren, Z., 2004. Upgrading of liquid fuel from the pyrolysis of biomass. Bioresource Technology 96, 545–550. Zhu, X.F., Zheng, J.L., Guo, Q.X., Zhu, Q.S., 2005. Upgrading and utilization of bio-oil from biomass. Engineering Science 7 (9), 83–88.