- Email: [email protected]
Cleaning of lubricant-oil-contaminated plastic using liquid carbon dioxide
G Model
JIEC-2748; No. of Pages 8 Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Cleaning of lubricant-oil-contaminated plastic using liquid carbon dioxide Manop Charoenchaitrakool a,b, Supaporn Tungkasatan a,b, Terdthai Vatanatham a,b, Sunun Limtrakul a,b,* a
Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand Center of Excellence on Petrochemical and Materials Technology, Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand
b
A R T I C L E I N F O
Article history: Received 26 June 2015 Received in revised form 2 October 2015 Accepted 1 December 2015 Available online xxx Keywords: Waste plastic Lubricating oil Cleaning Carbon dioxide Simulation
A B S T R A C T
The feasibility of using liquid carbon dioxide for cleaning contaminated plastics was evaluated. Flow sheets for the CO2 cleaning processes have been conceptualized and the material and energy balances were conducted using ASPEN Plus. The average % lubricant oil in waste plastics was found to be 2.49%. Although the CO2 cleaning process with hexane as a co-solvent resulted in a higher cleaning efficiency than the process without co-solvent, it consumed 2.2 times more energy. Based on cleaning 30 kg of plastics, the operating cost of the process with hexane was 5.7 times higher than that of the process without hexane. ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Plastics are used in a wide range of products due to their various beneficial properties such as versatility, durability and relatively inexpensive to produce. However, an increase in plastics waste generation, especially lubricating-oil containers, is unavoidable due to the rapid growth in industrial development and automobile industries. Since plastics are not biodegradable, they are not suitable for land filling. Their destruction via an incineration process can also cause serious air pollution problem due to the release of airborne particles and carbon dioxide into the atmosphere [1–3]. In addition, due to the high disposal amount of lubricating-oil containers each year, it is a challenge to minimize the disposal quantity. Recycling of these lubricating-oil containers seems to be an appropriate solution because it avoids accumulation in landfills, and clean plastics can be reused in the recycling process for making new containers. In the recycling of lubricating-oil containers, the post-consumer plastics are inspected for quality and washed to remove any
* Corresponding author at: Kasetsart University, Chemical Engineering Department, 50 Ngamwongwan Road, Jatujak, Bangkok 10900, Thailand. Tel.: +66 2 797 0999x1210; fax: +66 2 561 4621. E-mail address: [email protected] (S. Limtrakul).
residual impurities. Then, they are ground into pieces, dried and processed into pellets or flakes. Finally, the processed materials, in either flake or pelletized form, become feedstock in the manufacture of new products. It is important to note that thermal, mechanical and impact properties of the recycled plastics should be close to the virgin material to ensure the quality of the final products [4,5]. In general, these lubricating-oil containers are washed by using either detergents or organic solvents. However, there are many drawbacks to this conventional cleaning technique. These include the production of large quantities of contaminated wash solution which must be handled as hazardous waste, low cleaning efficiency and the use of non environment-friendly organic solvents. A supercritical fluid (SCF) has been established as a good alternative solvent. The use of supercritical fluids in the area of extraction has been well documented over the last few decades [6–11]. The adjustable solvent strength and gas-like transport properties make supercritical fluids efficient solvents for the extraction process. In addition, the increased scrutiny of industrial solvents by governments and awareness of pollution control have motivated the use of supercritical fluid as a cleaning solvent [12]. Recently, many researchers have used supercritical carbon dioxide for precision cleaning application, for examples, removing lubricating oil from metallic contacts [13], removing contaminants for remanufacturing industry [14,15], cleaning of rollers for
http://dx.doi.org/10.1016/j.jiec.2015.12.009 1226-086X/ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 2
M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
printing industry [16] and degreasing process in the leather industry [17]. Low critical point fluids are typically preferred since they can be used in the processing of heat sensitive compounds such as pharmaceuticals, and the inherent expense of high pressure equipment is not significant. Carbon dioxide is the most common fluid employed in supercritical fluid applications since it is non-flammable, non-toxic, non corrosive, readily available and inexpensive and has a relatively low critical point (Tc = 31.8 8C and Pc = 7.3 MPa) [18]. Although the supercritical fluid cleaning has proved to be efficient and can be considered as a green sustainable process, the main drawback of this technology for up-scaling is the high operating pressure. A high pressure pump and high pressure resistant piping are mandatory in the cleaning process, resulting in an unavoidable high capital cost. In addition, operating under high pressure contributes to a higher risk and thus more safety precaution is needed. In order to compensate the drawback of supercritical fluid cleaning but still maintain its advantages, the use of liquid carbon dioxide for lubricating-oil removal was investigated in this study. The objective of this work was to study the feasibility of lubricating-oil removal from plastic containers using liquid carbon dioxide. The effects of operating temperature, pressure, co-solvent and time of contact between oil and plastic on the removal efficiency were investigated. The original and the processed plastics were then analyzed using a Melt Flow Index Tester in order to monitor the quality of the extruded or injection-molded thermoplastics. Furthermore, flow sheets for the cleaning process of 30 kg/day of contaminated plastics (with and without hexane as a co-solvent), along with the material and energy balances, were simulated using the software ASPEN Plus version 7.1. Materials and methods Materials Used lubricating-oil containers were purchased from a local market. The lubricating oil used in this research was Performa semi-synthetic SAE10W-40 API SM/CF from PTT Thailand (a mixture of long-chain hydrocarbons with a viscosity of 14.9 cSt at 100 8C and a flash point of 232 8C [19]). The material for the contaminated plastic and unprocessed plastic was high density poly ethylene (HDPE). Carbon dioxide (high purity grade, TIG) was used as an extracting or cleaning solvent. Hexane (commercial grade) was used as a co-solvent.
%cleaning efficiency ¼
the top, middle and bottom, and then cut again into small pieces. Two different techniques were used to determine the % lubricating oil in the plastics in this study. The first method was to find the % lubricating oil in different sections and then calculate the average value. Note that for each section, five specimens were selected based on systematic random sampling. The second method was to randomly select 5 specimens from all three sections and determine the overall % lubricating oil. Each specimen was weighed and washed with hexane. In order to ensure the complete removal of the lubricating oil, the specimen was immersed in hexane, in which the ratio of hexane to contaminated plastics was 5:1. The system was shaken for 1 min and the specimen was soaked in hexane for 3 h. The cleaning process was repeated three times with the use of new hexane. After completing the cleaning process, the specimen was air-dried to remove hexane and the final constant weight of clean plastic was then recorded. The %lubricating oil in the contaminated plastic was calculated based on the average of five selected specimens for each section. Plastic cleaning process Using new plastics contacted with lubricating oil In order to control the amount of lubricating oil in the plastic sample, a brand new HDPE plastic container was used as a starting material. The container was cut into 15 cm 10 cm pieces and then was placed to contact with the lubricating oil only on one side for 7, 15 and 30 days. After a certain period of contacting time, the piece of plastic was removed and the excess oil was drained out. The contaminated plastic was then cut into 0.5 cm 0.5 cm pieces and used as samples for the cleaning process. Cleaning plastics with liquid CO2 was carried out using the experimental setup as shown in Fig. 1. Approximately 5.6 g of contaminated plastics was packed in a 500 cm3 cylindrical extractor. The system temperature was controlled to within 0.1 8C using a recirculation heater (Thermoline Unistat 130). Liquid CO2 was directed from a dip tube cylinder to the extractor using a high pressure CO2 cylinder. The system was maintained at the desired pressure and temperature for at least 30 min prior to commencing the cleaning process. The cleaning was initiated by opening valve V-6 and controlling the flow rate of liquid CO2 at 4 mL/min. The cleaning process was fixed for 3 h. After the cleaning process was complete, the system was depressurized and the CO2 was released to the atmosphere. All cleaned plastics were then removed from the sample cylinder for further analysis. The % cleaning efficiency was determined by the gravimetric method as follows:
mass of plastics before washingmass of plastics after washing mass of plastics before washingmass of dried plastics
Methods Determination of %lubricating oil in the waste plastics Prior to performing the lubricating-oil removal using liquid CO2, it is necessary to determine the amount of oil left in the waste plastics. The obtained % lubricating oil in the waste plastic is important data that is required for determining the % cleaning efficiency and for scaling up the process. Twenty lubricating-oil containers from various places were used as representative samples for determining the amount of lubricating oil left in the contaminated plastics. The labels on the containers were removed and the containers were cut into three different sections, namely
(1)
Apart from the continuous cleaning process as mentioned earlier, batch cleaning processes with or without adding hexane as a co-solvent were also conducted in order to compare the cleaning efficiency. In the batch process, the extractor loaded with the contaminated plastics was filled with liquid CO2 at 6.5 MPa and 5 8C. The system was maintained at the desired pressure and temperature for 3 h. After that, the liquid CO2 was separated from the plastics. The system was then depressurized and the CO2 was released to atmospheric pressure. In the case of adding hexane as a co-solvent, the contaminated plastics were loaded into the extractor, followed by adding hexane. The ratio of hexane to liquid CO2 feed was set at 1:10 by volume.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Fig. 1. Schematic diagram of plastic cleaning equipment using liquid carbon dioxide.
3
plastics/day. Two scenarios of cleaning, with or without hexane as a co-solvent, were studied. Flow sheets for both processes were conceptualized and the material and energy balances were simulated. The following assumptions were made for the simulation. A mixture of C19H40 and C30H62 in the weight ratio of 1:1 was used to represent the lubricating oil. The cleaning process was in the continuous mode. The Peng–Robinson equation of state was used for the property method. The operating conditions in the extractor were set to 6.5 MPa and 5 8C. In the separator unit, a Flash2 icon was selected from the model library to separate the extracted oil and CO2. In the case of neat CO2, the conditions in the separator were set to 5 MPa and 60 8C. However, in the case of CO2 with hexane as a co-solvent, the temperature in the separator unit was increased to 120 8C in order to maximize the amount of hexane in the vapor stream and at the same time maximize the amount of extracted oil in the liquid stream. The operating hours for the cleaning process were fixed at 8 h/day. Results and discussion
In this study, conventional plastic cleanings with hexane or dish detergent were also performed at room temperature and atmospheric pressure. In the case of using hexane, the ratio of hexane to contaminated plastics was fixed at 5:1. The plastics were soaked in liquid hexane for 3 h at atmospheric pressure. For the cleaning with dish detergent, the amount of detergent used was 10 g/100 cm3 of water. Using the commercial lubricating-oil containers The cleaning process for the commercial lubricating-oil containers was the same as that given in the above section. Since the actual amount of lubricating oil in the plastics is not known, this value was then calculated from the product of %lubricating oil in the contaminated plastic (as previously determined in the Determination of %lubricating oil in the waste plastics section) and weight of waste plastics. The % cleaning efficiency for the commercial lubricating-oil container can then be calculated as follows:
As mentioned in the Determination of %lubricating oil in the waste plastics section, it is necessary to determine the amount of oil left in the waste plastics. Fig. 2 shows the % lubricating oil in waste plastics for various sections of the container. It was found that each section of the container had a different amount of lubricating oil in the plastic. The top section had the highest % lubricating oil in the waste plastic; this could be due to the fact that the oil container was turned upside down to drain out the fluid when it was in use. In addition, lubricating oil could be accumulated easily in the neck and handle of the container in the top section. For the middle and bottom sections, there was no significant difference in % lubricating oil in the plastics. Since the surface for the middle and bottom sections was smooth and had no dead zone, less chance of lubricating oil could accumulate on the surface. It was also found that thickness of the middle section and
mass of plastics before washingmass of plastics after washing %oil in waste plasticsmass of waste plastics
Melt flow index (MFI) The flow properties of the new HDPE plastic container and plastics after washing were analyzed using Melt Flow Index (MFI) Tester (Dynisco, PPC, Bangkok, Thailand) under ASTM D 1238. A small amount of the sample (approx. 4.2 g) was packed properly inside the barrel. The sample was then preheated for 5 min at 190 8C. After that, a fixed load of 2.16 kg was introduced onto the piston to cause the extrusion of the molten polymer. The extruded polymer was collected and weighed after a desired period of time. MFI is expressed as grams of polymer/10 min of total time of the test. For each sample, the MFI tests were repeated five times, and the average value was reported. Process design and simulation using ASPEN Plus Based on the lab-scale experiments of the previous section, the concept of plastic cleaning using liquid CO2 was then further extended for pilot-plant scale. In this section, the software ASPEN Plus version 7.1 was used for process design of liquid CO2 cleaning. The process simulation was undertaken to handle 30 kg of waste
(2)
that of the bottom section were greater than that of the top section by 15% and 23%, respectively. Assuming the ability of lubricating oil that can diffuse into the plastic was the same for all sections, the thicker plastic in the middle and bottom sections would then have the lower % lubricating oil in the plastics compared to the top section. From Fig. 2, it is also interesting to note that the average
% lubricating oil in plastics
%cleaning efficiency ¼
% Lubricating oil in the waste plastics
3.5
3.08%
3
2.30%
2.5
2.49%
2.34%
2.05%
2 1.5 1 0.5 0
Top section
Middle section
Bottom section
Average
Overall
Fig. 2. % Lubricating oil in plastics for various sections of the lubricant oil container.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
4
value of % lubricating oil in plastics from three different sections was approximately close to the overall value. Plastic cleaning process Before scaling up the plastic cleaning process, it is necessary to understand the effect of operating parameters on the cleaning efficiency. In this study, the plastic in the middle section was used as a representative sample. In the case of using new plastics contacted with the lubricating oil, the effects of temperature, pressure, time of contact between oil and plastic and co-solvent on the cleaning efficiency were investigated. In the case of using waste plastics, only the effects of temperature, pressure and cosolvent on the cleaning efficiency were studied. Fig. 3 shows the % cleaning efficiency at various operating conditions for both contaminated new plastics and waste plastics. In general, it was found that as the temperature of the system was increased isobarically, the % cleaning efficiency was decreased. This is due to the fact that carbon dioxide density decreases as the temperature is increased, resulting in a lower solvating power. Fig. 3(a) also illustrates that for the constant pressure of 5.5 MPa, the % cleaning efficiency dramatically decreased when the operating temperature was greater than 15 8C. This is attributed to the fact that at the pressure of 5.5 MPa the boiling point of CO2 is 17 8C. Therefore, CO2 was no longer in a liquid state when cleaning plastic at temperatures higher than 17 8C and at 5.5 MPa. Similarly, for the operating pressures of 6.0 and 6.5 MPa the % cleaning efficiency decreased significantly when the cleaning process was conducted at 25 8C. This is because the boiling points of CO2 at 6.0 MPa and 6.5 MPa are 20 8C and 25 8C, respectively. Fig. 3 also illustrates the % cleaning efficiency at various pressures for both contaminated new plastics and waste plastics. It was found that the % cleaning efficiency increased as the pressure was increased. This is due to the fact that as pressure is increased,
carbon dioxide density increases and the intermolecular mean distance of carbon dioxide molecules decreases; thus the specific interaction between the solute and solvent molecules is increased [20]. The effect of contact time between oil and plastic on the cleaning efficiency can be observed in Fig. 3(a–c). It was found that the cleaning efficiency decreased slightly as the contact time between oil and plastic was increased from 7 days to 30 days. The longer contact time resulted in a deeper penetration of oil into the plastic matrix. As a result, it was more difficult to remove the oil from the interior of the plastics. Viguera et al. [13] reported that paraffinic heavy fractions and some additives in the lubricating oil also caused a chemisorbed oil layer on the surface of metallic contacts used in the electronics industry. As a result, lubricating oil bound strongly to the metallic surface, and the removal rate of highly viscous lubricating oil from metallic contacts was decreased. Thus, the longer contact time of the lubricating oil with the HDPE plastics may cause more chemisorption of lubricating oil on the surface. The effect of co-solvent on the % cleaning efficiency was also investigated for both contaminated new plastics and waste plastics. Hexane was selected as a co-solvent since it is non-polar and commonly used for oil extraction. In this study, hexane was added to the extractor and the cleaning process was carried out in a batch mode. The cleaning process was conducted at 6.5 MPa and 5 8C. For the contaminated new plastics, the plastics contacted with oil for 7 days were used as a representative sample. Fig. 4(a) shows the effect of co-solvent on the % cleaning efficiency for contaminated new plastics and waste plastics. The introduction of hexane in the system clearly enhanced the cleaning efficiency for both types of plastics. Note that the cleaning efficiencies of the contaminated new plastics were higher than those waste plastics. This was due to the fact that the contact time between oil and plastics for the contaminated new plastics was much shorter than that of the waste plastics.
Fig. 3. Effect of temperature on the % cleaning efficiency using liquid CO2 and contaminated new plastics with various contact times between oil and plastics (a) 7 days, (b) 15 days, (c) 30 days and (d) waste plastics.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Further investigations on the plastic cleaning using pure hexane and dish detergent were also conducted. Fig. 4(b) shows the cleaning efficiency for the conventional cleaning compared to the liquid CO2 operated in the continuous mode. It was found that using the dish detergent gave the lowest cleaning efficiency. In addition, large amounts of waste water generated from plastic cleaning using dish detergent can also cause environmental problems. The highest cleaning efficiency (83.2%) was obtained when using pure hexane. Since hexane is a non-polar organic solvent and has similar dielectric constant to the lubricating oil, it is a good solvent for oil removal [21]. However, the use of organic solvents is starting to be abandoned due to the increased scrutiny of industrial solvents by government and due to awareness of pollution control. Moreover, cleaning processes using pure hexane can be expensive due to the high cost associated with recovering the solvent, usually by distillation [13]. Although the cleaning efficiency using liquid CO2 in the continuous mode was approximately half that obtained using pure hexane, cleaning efficiency using liquid CO2 can be further enhanced by the addition of co-solvent or using a longer period of washing. In addition, CO2 can be separated from the extracted lubricating oil by simply depressurizing, liquefying and recycling it back to the washing process. Alternatively, as in this study, a flash drum or knock-out drum can be used to separate the extracted oil and CO2 under high-pressure condition in order to save the operating cost for re-pressurizing CO2. Melt flow index (MFI) The melt flow index (MFI) is one of the physical properties quoted for measuring the rate of extrusion of thermoplastics through an orifice at a prescribed temperature of 190 8C with a load of 2.16 kg. The MFI value can be used to analyze the extent of plastics degradation. The degraded plastics would flow fluently and display deterioration in physical properties. The residual oil in HDPE will increase the MFI of the plastics, resulting in more plastic flow compared to new plastic at the same operating conditions in extrusion processes. In this study, the MFI of new HDPE plastics, waste plastics and cleaned plastics from various cleaning methods were determined and reported in Table 1. Note that for HDPE plastics the standard MFI is in the range of 0.15–0.4 [22]. Therefore, the MFI value can be used as an indication to monitor whether the cleaning process is efficient and the cleaned plastics can be recycled for the manufacture of new containers. From Table 1, the MFI of the new HDPE plastics used for making the lubricant oil container was found to be 0.39, which is in a good agreement with the standard value. The MFI values for waste plastics before
Table 1 Melt flow index of new HDPE plastics, waste plastics and cleaned plastics at various cleaning conditions. Type of plastic
Conditions
Time of contacting with oil (days) Before washing 1. New HDPE plastics 2. Waste plastics Top section Middle section Bottom section After washing 3. Contaminated new plastics after cleaning with CO2
Pressure (MPa)
Melt flow index (g/10 min) Temp.(8C)
– –
– –
– –
0.39 0.46
– –
– –
– –
0.47 0.46
7 7 7 30 30 30
6.5 6.5 6.5 6.5 6.5 6.5
5 10 15 5 10 15
0.37 0.42 0.44 0.38 0.41 0.44
4. Waste plastics after cleaning with CO2
– – –
6.5 6.5 6.5
5 10 15
0.40 0.47 0.47
5. Waste plastics after cleaning with hexane 6. Waste plastics after cleaning with detergent
–
0.1
30
0.35
–
0.1
30
0.47
cleaning at different sections were found be in the range of 0.46– 0.47. Although the top section had the highest % lubricating oil in the waste plastic (3.08%), there was no significant difference in the MFI values for all sections. In the case of using contaminated new plastics, the cleaning process with liquid CO2 operated at 6.5 MPa and 5 8C resulted in the lowest MFI values, which were in the range of the standard MFI for HDPE plastics. This obtained result was also consistent with the highest cleaning efficiency when using liquid CO2 at 5 8C. As the temperature of CO2 was increased from 5 to 15 8C, the MFI values were also increased. It is interesting to note that CO2 cleaning processes conducted at 10 and 15 8C were not effective since the MFI values were higher than the range of the accepted standard values. As mentioned earlier, an increase in the contact time between oil and plastic from 7 days to 30 days resulted in a slight reduction in the cleaning efficiency. However, an increase in the contact time did not have any significant effect on the MFI values of the cleaned plastic. When comparing the
(a)
(b) 74.3%
90
70
Contaminated new plastics
60.1% 60
Waste plastics
50
37.4%
40
26.5%
30 20 10
% cleaning efficiency
% cleaning efficiency
80
5
83.2%
80 70 60
44.8%
50 40
29.2%
30 20 10 0
0
without hexane
with hexane
without hexane
with hexane
hexane
CO2 (continuous)
detergent
Fig. 4. (a) Effect of co-solvent on the % cleaning efficiency using liquid CO2 in the batch mode and (b) the cleaning efficiency for the conventional cleaning compared to the liquid CO2 operated in the continuous mode.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
6
5 °C,
3.78°C, 5 MPa
6.5 MPa
Vapor frac. = 0
CO2 H 2O Oil
Extractor
Liquid
Expansion Valve
Heater 1
60°C, 5 MPa Vapor frac. = 0.997 60 °C,
5 °C, 6.5 MPa
5 MPa
Separator
Liquid
Liquid 60 °C, 5 MPa Vapor frac. = 1 17.58°C, Cooler 2
6.5 MPa
14 °C, 5 MPa
Pump
Liquid
Liquid
Cooler 1
Fig. 5. Simplified block diagram of the proposed CO2 cleaning process without using hexane.
contaminated new plastics and waste plastics, it was found that the MFI values of the cleaned plastics from contaminated new plastics were slightly lower than those from the waste plastics. This could be due to the fact that for the waste plastic container, there were some lubricating oils which had penetrated into the plastics and were not removed by CO2 cleaning. However, cleaning the waste plastic using liquid CO2 at 5 8C was found to be effective since the MFI of the cleaned plastic was within the range of the accepted standard values. The MFI values of waste plastics cleaned by using pure hexane and dish detergent were also determined and are listed in Table 1. Cleaning with hexane gave the lowest MFI value in this study, whereas cleaning with detergent was found to be ineffective. Although the % cleaning efficiency by hexane was almost twice as high as that obtained when cleaning by liquid CO2 operated at 5 8C and 6.5 MPa (Fig. 4(b)), the MFI values of the cleaned plastics by these two methods were within the range of the standard values. These results implied that cleaning plastic by liquid CO2 operated at 5 8C and 6.5 MPa was effective and the
cleaned plastics can be recycled and used as raw material for making new containers. Process design and simulation using ASPEN Plus As discussed earlier, using liquid CO2 at 6.5 MPa and 5 8C can effectively remove the lubricant oil from waste plastics. In this section, the software ASPEN Plus version 7.1 was used for process design of liquid CO2 cleaning to handle 30 kg of waste plastics/day. Two scenarios of cleaning, with or without hexane as a co-solvent, were simulated. The % cleaning efficiency, CO2 loss from the system and energy consumption for each scenario were then compared in order to find a suitable cleaning process. Case I: Using liquid CO2 without hexane as a co-solvent The simplified block diagram of the process for removal of lubricant oil from waste plastic is shown in Fig. 5. The material balance of the proposed process is shown in Fig. 6. The main
CO2 = 37.601 kg/hr
Expansion
CO2 = 37.601 kg/hr
H2O = 0.0003 kg/hr
H2O = 0.0188 kg/hr
Valve
H2O = 0.0188 kg/hr
Oil = 0.3741 kg/hr
Oil = 0.375 kg/hr
CO2 = 0.0696 kg/hr
Extractor
Heater 1
Oil = 0.375 kg/hr CO2 = 37.601 kg/hr H2O = 0.0188 kg/hr Oil = 0.375 kg/hr
CO2 = 37.5314 kg/hr H2O = 0.0185kg/hr
Separator
CO2 = 0.0696 kg/hr H2O = 0.0003 kg/hr
Oil = 0.0009 kg/hr CO2 = 37.5314 kg/hr
Oil = 0.3741 kg/hr
H2O = 0.0185kg/hr Oil = 0.0009 kg/hr
Cooler 2
CO2 = 37.5314 kg/hr H2O = 0.0185kg/hr Oil = 0.0009 kg/hr
CO2 = 37.5314 kg/hr Pump
H2O = 0.0185kg/hr
Cooler 1
Oil = 0.0009 kg/hr
Fig. 6. Material balance for the proposed CO2 cleaning process without using hexane.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
CO2
3.86°C, 5 MPa
5 °C,
H 2O
Extractor
Oil
7
6.5 MPa
Expansion
Liquid
Valve
Vapor frac. = 0 Heater 1
C6H14 120°C, 5 MPa Vapor frac. = 0.995 5 °C, 6.5 MPa
120°C, 5 MPa
Liquid Separator
Liquid
120°C, 5 MPa Vapor frac. = 1 15.41°C, Cooler 2
6.5 MPa
Pump
10°C, 5 MPa Liquid
Liquid
Cooler 1
Fig. 7. Simplified block diagram of the proposed CO2 cleaning process with the use of hexane.
equipment in this process comprised one extractor, one separator, one heater, two coolers and one pump. The operating conditions in the extractor were set to 6.5 MPa and 5 8C. The conditions in the separator were set to 5 MPa and 60 8C. The cleaning efficiency of this process was found to be 69.93% compared to the oil feed to the extractor. CO2 loss from this process was found to be 0.185% based on the CO2 fed to the extractor. Case II: Using liquid CO2 with hexane as a co-solvent As mentioned earlier, the cleaning efficiency can be improved by adding hexane as a co-solvent. The simplified block diagram of the process for removal of lubricant oil from waste plastics with the use of 10 mol% hexane is shown in Fig. 7. The material balance
CO2 = 0.0931 kg/hr H2O = 0.0003 kg/hr Oil = 0.3625 kg/hr
Extractor
of the proposed process is shown in Fig. 8. The main equipment in this process is the same as that in Case I. However, the conditions in the separator were set to 5 MPa and 120 8C in order to separate hexane out of the liquid stream and recycle it back to the process. The cleaning efficiency of this process was found to be 91.74% compared to the oil feed to the extractor. CO2 and hexane losses from this process were calculated to be 0.180% and 2.63%, respectively. Due to the high temperature required in the separator (120 8C) in the Case II, the overall energy consumption for this process was found to be 56.4 kW/30 kg of waste plastics, which was 2.2 times higher than that in Case I. As shown in Fig. 8, the cleaning process had a loss of 0.2945 kg/h of hexane. It was assumed that the operating hours for the cleaning process were 8 h/day. Therefore, the
CO2 = 51.6259 kg/hr
CO2 = 51.6259 kg/hr
H2O = 0.0258 kg/hr
H2O = 0.0258 kg/hr
Oil = 0.375 kg/hr C6H14 = 11.2117 kg/hr
C6H14 = 0.2945 kg/hr
Expansion Valve
Oil = 0.375 kg/hr
Heater 1
C6H14 = 11.2117 kg/hr
CO2 = 51.6259 kg/hr H2O = 0.0258 kg/hr Oil = 0.375 kg/hr C6H14 = 11.2117 kg/hr
CO2 = 51.5328 kg/hr H2O = 0.0256 kg/hr
CO2 = 0.0931 kg/hr
Oil = 0.0124 kg/hr Separator
C6H14 = 10.9172 kg/hr
H2O = 0.0003 kg/hr Oil = 0.3625 kg/hr
CO2 = 51.5328 kg/hr
C6H14 = 0.2945 kg/hr
H2O = 0.0256 kg/hr Oil = 0.0124 kg/hr C6H14 = 10.9172 kg/hr
CO2 = 51.5328 kg/hr Cooler 2
H2O = 0.0256 kg/hr
Pump
CO2 = 51.5328 kg/hr H2O = 0.0256 kg/hr
Oil = 0.0124 kg/hr
Oil = 0.0124 kg/hr
C6H14 = 10.9172 kg/hr
C6H14 = 10.9172 kg/hr
Cooler 1
Fig. 8. Material balance for the proposed CO2 cleaning process with the use of hexane.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
8
Table 2 Estimated operating cost for cleaning 30 kg of waste plastics. Process
CO2 loss (US$)
Hexane loss (US$)
Electricity (US$)
Total operating cost (US$)
LCO2 LCO2 + hexane
0.74 0.99
8.25
1.43 3.12
2.17 12.36
amount of hexane loss was 2.356 kg/day. Based on the price of hexane in Thailand (US$ 3.5/kg), the cost of hexane loss was calculated to be US$ 8.25. Table 2 shows the estimated operating cost for cleaning 30 kg of waste plastics. It was found that the operating cost in Case II was approximately 5.7 times greater than that in Case I. The major contribution of this high cost is the hexane loss. Conclusions The feasibility of lubricating-oil removal from HDPE plastic containers using liquid carbon dioxide was investigated. Prior to the plastic cleaning process, the % lubricating oil in waste plastic from various sections of the container was determined, with the average value being found to be 2.49%. An increase in the operating pressure from 5.5 to 6.5 MPa resulted in a higher % cleaning efficiency. However, as the temperature or the contact time between oil and container increased, a slight reduction in % cleaning efficiency was observed. With the use of liquid CO2 operated at 6.5 MPa and 5 8C, the highest % cleaning efficiencies were achieved, i.e., 91.0% for contaminated new plastic and 44.8% for waste plastic. In addition, it was found that cleaning waste plastic using liquid CO2 operated at 6.5 MPa and 5 8C was approximately half as efficient as cleaning with neat hexane. The cleaning efficiency was enhanced by adding hexane as a co-solvent. Based on the MFI tests, plastics cleaned by liquid CO2 at 6.5 MPa and 5 8C exhibited the MFI values within the range of the accepted standard values. This promising result shows that the cleaned plastics can be recycled as a starting material for new containers. Acknowledgements The authors gratefully acknowledge the financial supports of the Kasetsart University Research and Development Institute
(KURDI), Faculty of Engineering Kasetsart University, the Thailand Research Fund (TRF), and the Center of Excellence on Petrochemical and Materials Technology. References [1] J.N. Sahu, K.K. Mahalik, H.K. Nam, T.Y. Ling, T.S. Woon, M.S.B.A. Rahman, Y.K. Mohanty, N.S. Jayakumar, S.S. Jamuar, Environ. Prog. Sustainable Energy 33 (2014) 298. [2] L. Saeed, R. Zevenhoven, Energy Sources 24 (2002) 41. [3] S.H. Hamid, M.B. Amin, J. Appl. Polym. Sci. 55 (1995) 1385. [4] S.M. Al-Salem, P. Lettieri, J. Baeyens, Waste Manage. 29 (2009) 2625. [5] M. Sa´nchez-Soto, A. Rossa, A.J. Sa´nchez, J. Ga´mez-Pe´rez, Waste Manage. 28 (2008) 2565. [6] Zˇ. Knez, E. Markocˇicˇ, M. Leitgeb, M. Primozˇicˇ, M. Knez Hrncˇicˇ, M. Sˇkerget, Energy 77 (2014) 235. [7] J. Martı´nez, A.C. de Aguiar, Curr. Anal. Chem. 10 (2014) 67. [8] B.A.S. Machado, C.G. Pereira, S.B. Nunes, F.F. Padilha, M.A. Umsza-Guez, Sep. Sci. Technol. 48 (2013) 2741. [9] H. Valde´s, R. Sepu´lveda, J. Romero, F. Valenzuela, J. Sa´nchez, Chem. Eng. Process.: Process Intensification 65 (2013) 58. [10] L. Fiori, Chem. Eng. Process.: Process Intensification 49 (2010) 866. [11] X. Zhang, B. Han, Clean 35 (2007) 223. [12] M.J.E. van Roosmalen, G.F. Woerlee, G.J. Witkamp, J. Supercrit. Fluids 27 (2003) 337. [13] M. Viguera, J.M. Go´mez-Salazar, M.I. Barrena, L. Calvo, J. Supercrit. Fluids 73 (2013) 51. [14] W.W. Liu, B. Zhang, Y.Z. Li, Y.M. He, H.C. Zhang, Appl. Surf. Sci. 292 (2014) 142. [15] W.W. Liu, M.Z. Li, T. Short, X.C. Qing, Y.M. He, Y.Z. Li, L.H. Liu, H. Zhang, H.C. Zhang, J. Cleaner Prod. 93 (2015) 339. [16] G. Della Porta, M.C. Volpe, E. Reverchon, J. Supercrit. Fluids 37 (2006) 409. [17] A. Marsal, P.J. Celma, J. Cot, M. Cequier, J. Supercrit. Fluids 16 (2000) 217. [18] E. Ramsey, Q. Sun, Z. Zhang, C. Zhang, W. Gou, J. Environ. Sci. 21 (2009) 720. [19] PTT, Product Data Sheet, 2015 hhttp://pttweb2.pttplc.com/weblub/Files/Lube/ attach/6_tPDS.pdfi (Last accessed: 30 September 2015). [20] G.S. Gurdial, N.R. Foster, Ind. Eng. Chem. Res. 30 (1991) 575. [21] A.A. Carey, The dielectric constant of lubrication oils, Computational Systems Incorporated, Knoxville, TN, 1998. [22] H.G. Lester, The Complete Corrugated Polyethylene Pipe Design Manual and Installation Guide, Plastic Pipe Institute, Irving, TX, 2011.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
JIEC-2748; No. of Pages 8 Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Cleaning of lubricant-oil-contaminated plastic using liquid carbon dioxide Manop Charoenchaitrakool a,b, Supaporn Tungkasatan a,b, Terdthai Vatanatham a,b, Sunun Limtrakul a,b,* a
Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand Center of Excellence on Petrochemical and Materials Technology, Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand
b
A R T I C L E I N F O
Article history: Received 26 June 2015 Received in revised form 2 October 2015 Accepted 1 December 2015 Available online xxx Keywords: Waste plastic Lubricating oil Cleaning Carbon dioxide Simulation
A B S T R A C T
The feasibility of using liquid carbon dioxide for cleaning contaminated plastics was evaluated. Flow sheets for the CO2 cleaning processes have been conceptualized and the material and energy balances were conducted using ASPEN Plus. The average % lubricant oil in waste plastics was found to be 2.49%. Although the CO2 cleaning process with hexane as a co-solvent resulted in a higher cleaning efficiency than the process without co-solvent, it consumed 2.2 times more energy. Based on cleaning 30 kg of plastics, the operating cost of the process with hexane was 5.7 times higher than that of the process without hexane. ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Plastics are used in a wide range of products due to their various beneficial properties such as versatility, durability and relatively inexpensive to produce. However, an increase in plastics waste generation, especially lubricating-oil containers, is unavoidable due to the rapid growth in industrial development and automobile industries. Since plastics are not biodegradable, they are not suitable for land filling. Their destruction via an incineration process can also cause serious air pollution problem due to the release of airborne particles and carbon dioxide into the atmosphere [1–3]. In addition, due to the high disposal amount of lubricating-oil containers each year, it is a challenge to minimize the disposal quantity. Recycling of these lubricating-oil containers seems to be an appropriate solution because it avoids accumulation in landfills, and clean plastics can be reused in the recycling process for making new containers. In the recycling of lubricating-oil containers, the post-consumer plastics are inspected for quality and washed to remove any
* Corresponding author at: Kasetsart University, Chemical Engineering Department, 50 Ngamwongwan Road, Jatujak, Bangkok 10900, Thailand. Tel.: +66 2 797 0999x1210; fax: +66 2 561 4621. E-mail address: [email protected] (S. Limtrakul).
residual impurities. Then, they are ground into pieces, dried and processed into pellets or flakes. Finally, the processed materials, in either flake or pelletized form, become feedstock in the manufacture of new products. It is important to note that thermal, mechanical and impact properties of the recycled plastics should be close to the virgin material to ensure the quality of the final products [4,5]. In general, these lubricating-oil containers are washed by using either detergents or organic solvents. However, there are many drawbacks to this conventional cleaning technique. These include the production of large quantities of contaminated wash solution which must be handled as hazardous waste, low cleaning efficiency and the use of non environment-friendly organic solvents. A supercritical fluid (SCF) has been established as a good alternative solvent. The use of supercritical fluids in the area of extraction has been well documented over the last few decades [6–11]. The adjustable solvent strength and gas-like transport properties make supercritical fluids efficient solvents for the extraction process. In addition, the increased scrutiny of industrial solvents by governments and awareness of pollution control have motivated the use of supercritical fluid as a cleaning solvent [12]. Recently, many researchers have used supercritical carbon dioxide for precision cleaning application, for examples, removing lubricating oil from metallic contacts [13], removing contaminants for remanufacturing industry [14,15], cleaning of rollers for
http://dx.doi.org/10.1016/j.jiec.2015.12.009 1226-086X/ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 2
M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
printing industry [16] and degreasing process in the leather industry [17]. Low critical point fluids are typically preferred since they can be used in the processing of heat sensitive compounds such as pharmaceuticals, and the inherent expense of high pressure equipment is not significant. Carbon dioxide is the most common fluid employed in supercritical fluid applications since it is non-flammable, non-toxic, non corrosive, readily available and inexpensive and has a relatively low critical point (Tc = 31.8 8C and Pc = 7.3 MPa) [18]. Although the supercritical fluid cleaning has proved to be efficient and can be considered as a green sustainable process, the main drawback of this technology for up-scaling is the high operating pressure. A high pressure pump and high pressure resistant piping are mandatory in the cleaning process, resulting in an unavoidable high capital cost. In addition, operating under high pressure contributes to a higher risk and thus more safety precaution is needed. In order to compensate the drawback of supercritical fluid cleaning but still maintain its advantages, the use of liquid carbon dioxide for lubricating-oil removal was investigated in this study. The objective of this work was to study the feasibility of lubricating-oil removal from plastic containers using liquid carbon dioxide. The effects of operating temperature, pressure, co-solvent and time of contact between oil and plastic on the removal efficiency were investigated. The original and the processed plastics were then analyzed using a Melt Flow Index Tester in order to monitor the quality of the extruded or injection-molded thermoplastics. Furthermore, flow sheets for the cleaning process of 30 kg/day of contaminated plastics (with and without hexane as a co-solvent), along with the material and energy balances, were simulated using the software ASPEN Plus version 7.1. Materials and methods Materials Used lubricating-oil containers were purchased from a local market. The lubricating oil used in this research was Performa semi-synthetic SAE10W-40 API SM/CF from PTT Thailand (a mixture of long-chain hydrocarbons with a viscosity of 14.9 cSt at 100 8C and a flash point of 232 8C [19]). The material for the contaminated plastic and unprocessed plastic was high density poly ethylene (HDPE). Carbon dioxide (high purity grade, TIG) was used as an extracting or cleaning solvent. Hexane (commercial grade) was used as a co-solvent.
%cleaning efficiency ¼
the top, middle and bottom, and then cut again into small pieces. Two different techniques were used to determine the % lubricating oil in the plastics in this study. The first method was to find the % lubricating oil in different sections and then calculate the average value. Note that for each section, five specimens were selected based on systematic random sampling. The second method was to randomly select 5 specimens from all three sections and determine the overall % lubricating oil. Each specimen was weighed and washed with hexane. In order to ensure the complete removal of the lubricating oil, the specimen was immersed in hexane, in which the ratio of hexane to contaminated plastics was 5:1. The system was shaken for 1 min and the specimen was soaked in hexane for 3 h. The cleaning process was repeated three times with the use of new hexane. After completing the cleaning process, the specimen was air-dried to remove hexane and the final constant weight of clean plastic was then recorded. The %lubricating oil in the contaminated plastic was calculated based on the average of five selected specimens for each section. Plastic cleaning process Using new plastics contacted with lubricating oil In order to control the amount of lubricating oil in the plastic sample, a brand new HDPE plastic container was used as a starting material. The container was cut into 15 cm 10 cm pieces and then was placed to contact with the lubricating oil only on one side for 7, 15 and 30 days. After a certain period of contacting time, the piece of plastic was removed and the excess oil was drained out. The contaminated plastic was then cut into 0.5 cm 0.5 cm pieces and used as samples for the cleaning process. Cleaning plastics with liquid CO2 was carried out using the experimental setup as shown in Fig. 1. Approximately 5.6 g of contaminated plastics was packed in a 500 cm3 cylindrical extractor. The system temperature was controlled to within 0.1 8C using a recirculation heater (Thermoline Unistat 130). Liquid CO2 was directed from a dip tube cylinder to the extractor using a high pressure CO2 cylinder. The system was maintained at the desired pressure and temperature for at least 30 min prior to commencing the cleaning process. The cleaning was initiated by opening valve V-6 and controlling the flow rate of liquid CO2 at 4 mL/min. The cleaning process was fixed for 3 h. After the cleaning process was complete, the system was depressurized and the CO2 was released to the atmosphere. All cleaned plastics were then removed from the sample cylinder for further analysis. The % cleaning efficiency was determined by the gravimetric method as follows:
mass of plastics before washingmass of plastics after washing mass of plastics before washingmass of dried plastics
Methods Determination of %lubricating oil in the waste plastics Prior to performing the lubricating-oil removal using liquid CO2, it is necessary to determine the amount of oil left in the waste plastics. The obtained % lubricating oil in the waste plastic is important data that is required for determining the % cleaning efficiency and for scaling up the process. Twenty lubricating-oil containers from various places were used as representative samples for determining the amount of lubricating oil left in the contaminated plastics. The labels on the containers were removed and the containers were cut into three different sections, namely
(1)
Apart from the continuous cleaning process as mentioned earlier, batch cleaning processes with or without adding hexane as a co-solvent were also conducted in order to compare the cleaning efficiency. In the batch process, the extractor loaded with the contaminated plastics was filled with liquid CO2 at 6.5 MPa and 5 8C. The system was maintained at the desired pressure and temperature for 3 h. After that, the liquid CO2 was separated from the plastics. The system was then depressurized and the CO2 was released to atmospheric pressure. In the case of adding hexane as a co-solvent, the contaminated plastics were loaded into the extractor, followed by adding hexane. The ratio of hexane to liquid CO2 feed was set at 1:10 by volume.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Fig. 1. Schematic diagram of plastic cleaning equipment using liquid carbon dioxide.
3
plastics/day. Two scenarios of cleaning, with or without hexane as a co-solvent, were studied. Flow sheets for both processes were conceptualized and the material and energy balances were simulated. The following assumptions were made for the simulation. A mixture of C19H40 and C30H62 in the weight ratio of 1:1 was used to represent the lubricating oil. The cleaning process was in the continuous mode. The Peng–Robinson equation of state was used for the property method. The operating conditions in the extractor were set to 6.5 MPa and 5 8C. In the separator unit, a Flash2 icon was selected from the model library to separate the extracted oil and CO2. In the case of neat CO2, the conditions in the separator were set to 5 MPa and 60 8C. However, in the case of CO2 with hexane as a co-solvent, the temperature in the separator unit was increased to 120 8C in order to maximize the amount of hexane in the vapor stream and at the same time maximize the amount of extracted oil in the liquid stream. The operating hours for the cleaning process were fixed at 8 h/day. Results and discussion
In this study, conventional plastic cleanings with hexane or dish detergent were also performed at room temperature and atmospheric pressure. In the case of using hexane, the ratio of hexane to contaminated plastics was fixed at 5:1. The plastics were soaked in liquid hexane for 3 h at atmospheric pressure. For the cleaning with dish detergent, the amount of detergent used was 10 g/100 cm3 of water. Using the commercial lubricating-oil containers The cleaning process for the commercial lubricating-oil containers was the same as that given in the above section. Since the actual amount of lubricating oil in the plastics is not known, this value was then calculated from the product of %lubricating oil in the contaminated plastic (as previously determined in the Determination of %lubricating oil in the waste plastics section) and weight of waste plastics. The % cleaning efficiency for the commercial lubricating-oil container can then be calculated as follows:
As mentioned in the Determination of %lubricating oil in the waste plastics section, it is necessary to determine the amount of oil left in the waste plastics. Fig. 2 shows the % lubricating oil in waste plastics for various sections of the container. It was found that each section of the container had a different amount of lubricating oil in the plastic. The top section had the highest % lubricating oil in the waste plastic; this could be due to the fact that the oil container was turned upside down to drain out the fluid when it was in use. In addition, lubricating oil could be accumulated easily in the neck and handle of the container in the top section. For the middle and bottom sections, there was no significant difference in % lubricating oil in the plastics. Since the surface for the middle and bottom sections was smooth and had no dead zone, less chance of lubricating oil could accumulate on the surface. It was also found that thickness of the middle section and
mass of plastics before washingmass of plastics after washing %oil in waste plasticsmass of waste plastics
Melt flow index (MFI) The flow properties of the new HDPE plastic container and plastics after washing were analyzed using Melt Flow Index (MFI) Tester (Dynisco, PPC, Bangkok, Thailand) under ASTM D 1238. A small amount of the sample (approx. 4.2 g) was packed properly inside the barrel. The sample was then preheated for 5 min at 190 8C. After that, a fixed load of 2.16 kg was introduced onto the piston to cause the extrusion of the molten polymer. The extruded polymer was collected and weighed after a desired period of time. MFI is expressed as grams of polymer/10 min of total time of the test. For each sample, the MFI tests were repeated five times, and the average value was reported. Process design and simulation using ASPEN Plus Based on the lab-scale experiments of the previous section, the concept of plastic cleaning using liquid CO2 was then further extended for pilot-plant scale. In this section, the software ASPEN Plus version 7.1 was used for process design of liquid CO2 cleaning. The process simulation was undertaken to handle 30 kg of waste
(2)
that of the bottom section were greater than that of the top section by 15% and 23%, respectively. Assuming the ability of lubricating oil that can diffuse into the plastic was the same for all sections, the thicker plastic in the middle and bottom sections would then have the lower % lubricating oil in the plastics compared to the top section. From Fig. 2, it is also interesting to note that the average
% lubricating oil in plastics
%cleaning efficiency ¼
% Lubricating oil in the waste plastics
3.5
3.08%
3
2.30%
2.5
2.49%
2.34%
2.05%
2 1.5 1 0.5 0
Top section
Middle section
Bottom section
Average
Overall
Fig. 2. % Lubricating oil in plastics for various sections of the lubricant oil container.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
4
value of % lubricating oil in plastics from three different sections was approximately close to the overall value. Plastic cleaning process Before scaling up the plastic cleaning process, it is necessary to understand the effect of operating parameters on the cleaning efficiency. In this study, the plastic in the middle section was used as a representative sample. In the case of using new plastics contacted with the lubricating oil, the effects of temperature, pressure, time of contact between oil and plastic and co-solvent on the cleaning efficiency were investigated. In the case of using waste plastics, only the effects of temperature, pressure and cosolvent on the cleaning efficiency were studied. Fig. 3 shows the % cleaning efficiency at various operating conditions for both contaminated new plastics and waste plastics. In general, it was found that as the temperature of the system was increased isobarically, the % cleaning efficiency was decreased. This is due to the fact that carbon dioxide density decreases as the temperature is increased, resulting in a lower solvating power. Fig. 3(a) also illustrates that for the constant pressure of 5.5 MPa, the % cleaning efficiency dramatically decreased when the operating temperature was greater than 15 8C. This is attributed to the fact that at the pressure of 5.5 MPa the boiling point of CO2 is 17 8C. Therefore, CO2 was no longer in a liquid state when cleaning plastic at temperatures higher than 17 8C and at 5.5 MPa. Similarly, for the operating pressures of 6.0 and 6.5 MPa the % cleaning efficiency decreased significantly when the cleaning process was conducted at 25 8C. This is because the boiling points of CO2 at 6.0 MPa and 6.5 MPa are 20 8C and 25 8C, respectively. Fig. 3 also illustrates the % cleaning efficiency at various pressures for both contaminated new plastics and waste plastics. It was found that the % cleaning efficiency increased as the pressure was increased. This is due to the fact that as pressure is increased,
carbon dioxide density increases and the intermolecular mean distance of carbon dioxide molecules decreases; thus the specific interaction between the solute and solvent molecules is increased [20]. The effect of contact time between oil and plastic on the cleaning efficiency can be observed in Fig. 3(a–c). It was found that the cleaning efficiency decreased slightly as the contact time between oil and plastic was increased from 7 days to 30 days. The longer contact time resulted in a deeper penetration of oil into the plastic matrix. As a result, it was more difficult to remove the oil from the interior of the plastics. Viguera et al. [13] reported that paraffinic heavy fractions and some additives in the lubricating oil also caused a chemisorbed oil layer on the surface of metallic contacts used in the electronics industry. As a result, lubricating oil bound strongly to the metallic surface, and the removal rate of highly viscous lubricating oil from metallic contacts was decreased. Thus, the longer contact time of the lubricating oil with the HDPE plastics may cause more chemisorption of lubricating oil on the surface. The effect of co-solvent on the % cleaning efficiency was also investigated for both contaminated new plastics and waste plastics. Hexane was selected as a co-solvent since it is non-polar and commonly used for oil extraction. In this study, hexane was added to the extractor and the cleaning process was carried out in a batch mode. The cleaning process was conducted at 6.5 MPa and 5 8C. For the contaminated new plastics, the plastics contacted with oil for 7 days were used as a representative sample. Fig. 4(a) shows the effect of co-solvent on the % cleaning efficiency for contaminated new plastics and waste plastics. The introduction of hexane in the system clearly enhanced the cleaning efficiency for both types of plastics. Note that the cleaning efficiencies of the contaminated new plastics were higher than those waste plastics. This was due to the fact that the contact time between oil and plastics for the contaminated new plastics was much shorter than that of the waste plastics.
Fig. 3. Effect of temperature on the % cleaning efficiency using liquid CO2 and contaminated new plastics with various contact times between oil and plastics (a) 7 days, (b) 15 days, (c) 30 days and (d) waste plastics.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
Further investigations on the plastic cleaning using pure hexane and dish detergent were also conducted. Fig. 4(b) shows the cleaning efficiency for the conventional cleaning compared to the liquid CO2 operated in the continuous mode. It was found that using the dish detergent gave the lowest cleaning efficiency. In addition, large amounts of waste water generated from plastic cleaning using dish detergent can also cause environmental problems. The highest cleaning efficiency (83.2%) was obtained when using pure hexane. Since hexane is a non-polar organic solvent and has similar dielectric constant to the lubricating oil, it is a good solvent for oil removal [21]. However, the use of organic solvents is starting to be abandoned due to the increased scrutiny of industrial solvents by government and due to awareness of pollution control. Moreover, cleaning processes using pure hexane can be expensive due to the high cost associated with recovering the solvent, usually by distillation [13]. Although the cleaning efficiency using liquid CO2 in the continuous mode was approximately half that obtained using pure hexane, cleaning efficiency using liquid CO2 can be further enhanced by the addition of co-solvent or using a longer period of washing. In addition, CO2 can be separated from the extracted lubricating oil by simply depressurizing, liquefying and recycling it back to the washing process. Alternatively, as in this study, a flash drum or knock-out drum can be used to separate the extracted oil and CO2 under high-pressure condition in order to save the operating cost for re-pressurizing CO2. Melt flow index (MFI) The melt flow index (MFI) is one of the physical properties quoted for measuring the rate of extrusion of thermoplastics through an orifice at a prescribed temperature of 190 8C with a load of 2.16 kg. The MFI value can be used to analyze the extent of plastics degradation. The degraded plastics would flow fluently and display deterioration in physical properties. The residual oil in HDPE will increase the MFI of the plastics, resulting in more plastic flow compared to new plastic at the same operating conditions in extrusion processes. In this study, the MFI of new HDPE plastics, waste plastics and cleaned plastics from various cleaning methods were determined and reported in Table 1. Note that for HDPE plastics the standard MFI is in the range of 0.15–0.4 [22]. Therefore, the MFI value can be used as an indication to monitor whether the cleaning process is efficient and the cleaned plastics can be recycled for the manufacture of new containers. From Table 1, the MFI of the new HDPE plastics used for making the lubricant oil container was found to be 0.39, which is in a good agreement with the standard value. The MFI values for waste plastics before
Table 1 Melt flow index of new HDPE plastics, waste plastics and cleaned plastics at various cleaning conditions. Type of plastic
Conditions
Time of contacting with oil (days) Before washing 1. New HDPE plastics 2. Waste plastics Top section Middle section Bottom section After washing 3. Contaminated new plastics after cleaning with CO2
Pressure (MPa)
Melt flow index (g/10 min) Temp.(8C)
– –
– –
– –
0.39 0.46
– –
– –
– –
0.47 0.46
7 7 7 30 30 30
6.5 6.5 6.5 6.5 6.5 6.5
5 10 15 5 10 15
0.37 0.42 0.44 0.38 0.41 0.44
4. Waste plastics after cleaning with CO2
– – –
6.5 6.5 6.5
5 10 15
0.40 0.47 0.47
5. Waste plastics after cleaning with hexane 6. Waste plastics after cleaning with detergent
–
0.1
30
0.35
–
0.1
30
0.47
cleaning at different sections were found be in the range of 0.46– 0.47. Although the top section had the highest % lubricating oil in the waste plastic (3.08%), there was no significant difference in the MFI values for all sections. In the case of using contaminated new plastics, the cleaning process with liquid CO2 operated at 6.5 MPa and 5 8C resulted in the lowest MFI values, which were in the range of the standard MFI for HDPE plastics. This obtained result was also consistent with the highest cleaning efficiency when using liquid CO2 at 5 8C. As the temperature of CO2 was increased from 5 to 15 8C, the MFI values were also increased. It is interesting to note that CO2 cleaning processes conducted at 10 and 15 8C were not effective since the MFI values were higher than the range of the accepted standard values. As mentioned earlier, an increase in the contact time between oil and plastic from 7 days to 30 days resulted in a slight reduction in the cleaning efficiency. However, an increase in the contact time did not have any significant effect on the MFI values of the cleaned plastic. When comparing the
(a)
(b) 74.3%
90
70
Contaminated new plastics
60.1% 60
Waste plastics
50
37.4%
40
26.5%
30 20 10
% cleaning efficiency
% cleaning efficiency
80
5
83.2%
80 70 60
44.8%
50 40
29.2%
30 20 10 0
0
without hexane
with hexane
without hexane
with hexane
hexane
CO2 (continuous)
detergent
Fig. 4. (a) Effect of co-solvent on the % cleaning efficiency using liquid CO2 in the batch mode and (b) the cleaning efficiency for the conventional cleaning compared to the liquid CO2 operated in the continuous mode.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
6
5 °C,
3.78°C, 5 MPa
6.5 MPa
Vapor frac. = 0
CO2 H 2O Oil
Extractor
Liquid
Expansion Valve
Heater 1
60°C, 5 MPa Vapor frac. = 0.997 60 °C,
5 °C, 6.5 MPa
5 MPa
Separator
Liquid
Liquid 60 °C, 5 MPa Vapor frac. = 1 17.58°C, Cooler 2
6.5 MPa
14 °C, 5 MPa
Pump
Liquid
Liquid
Cooler 1
Fig. 5. Simplified block diagram of the proposed CO2 cleaning process without using hexane.
contaminated new plastics and waste plastics, it was found that the MFI values of the cleaned plastics from contaminated new plastics were slightly lower than those from the waste plastics. This could be due to the fact that for the waste plastic container, there were some lubricating oils which had penetrated into the plastics and were not removed by CO2 cleaning. However, cleaning the waste plastic using liquid CO2 at 5 8C was found to be effective since the MFI of the cleaned plastic was within the range of the accepted standard values. The MFI values of waste plastics cleaned by using pure hexane and dish detergent were also determined and are listed in Table 1. Cleaning with hexane gave the lowest MFI value in this study, whereas cleaning with detergent was found to be ineffective. Although the % cleaning efficiency by hexane was almost twice as high as that obtained when cleaning by liquid CO2 operated at 5 8C and 6.5 MPa (Fig. 4(b)), the MFI values of the cleaned plastics by these two methods were within the range of the standard values. These results implied that cleaning plastic by liquid CO2 operated at 5 8C and 6.5 MPa was effective and the
cleaned plastics can be recycled and used as raw material for making new containers. Process design and simulation using ASPEN Plus As discussed earlier, using liquid CO2 at 6.5 MPa and 5 8C can effectively remove the lubricant oil from waste plastics. In this section, the software ASPEN Plus version 7.1 was used for process design of liquid CO2 cleaning to handle 30 kg of waste plastics/day. Two scenarios of cleaning, with or without hexane as a co-solvent, were simulated. The % cleaning efficiency, CO2 loss from the system and energy consumption for each scenario were then compared in order to find a suitable cleaning process. Case I: Using liquid CO2 without hexane as a co-solvent The simplified block diagram of the process for removal of lubricant oil from waste plastic is shown in Fig. 5. The material balance of the proposed process is shown in Fig. 6. The main
CO2 = 37.601 kg/hr
Expansion
CO2 = 37.601 kg/hr
H2O = 0.0003 kg/hr
H2O = 0.0188 kg/hr
Valve
H2O = 0.0188 kg/hr
Oil = 0.3741 kg/hr
Oil = 0.375 kg/hr
CO2 = 0.0696 kg/hr
Extractor
Heater 1
Oil = 0.375 kg/hr CO2 = 37.601 kg/hr H2O = 0.0188 kg/hr Oil = 0.375 kg/hr
CO2 = 37.5314 kg/hr H2O = 0.0185kg/hr
Separator
CO2 = 0.0696 kg/hr H2O = 0.0003 kg/hr
Oil = 0.0009 kg/hr CO2 = 37.5314 kg/hr
Oil = 0.3741 kg/hr
H2O = 0.0185kg/hr Oil = 0.0009 kg/hr
Cooler 2
CO2 = 37.5314 kg/hr H2O = 0.0185kg/hr Oil = 0.0009 kg/hr
CO2 = 37.5314 kg/hr Pump
H2O = 0.0185kg/hr
Cooler 1
Oil = 0.0009 kg/hr
Fig. 6. Material balance for the proposed CO2 cleaning process without using hexane.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
CO2
3.86°C, 5 MPa
5 °C,
H 2O
Extractor
Oil
7
6.5 MPa
Expansion
Liquid
Valve
Vapor frac. = 0 Heater 1
C6H14 120°C, 5 MPa Vapor frac. = 0.995 5 °C, 6.5 MPa
120°C, 5 MPa
Liquid Separator
Liquid
120°C, 5 MPa Vapor frac. = 1 15.41°C, Cooler 2
6.5 MPa
Pump
10°C, 5 MPa Liquid
Liquid
Cooler 1
Fig. 7. Simplified block diagram of the proposed CO2 cleaning process with the use of hexane.
equipment in this process comprised one extractor, one separator, one heater, two coolers and one pump. The operating conditions in the extractor were set to 6.5 MPa and 5 8C. The conditions in the separator were set to 5 MPa and 60 8C. The cleaning efficiency of this process was found to be 69.93% compared to the oil feed to the extractor. CO2 loss from this process was found to be 0.185% based on the CO2 fed to the extractor. Case II: Using liquid CO2 with hexane as a co-solvent As mentioned earlier, the cleaning efficiency can be improved by adding hexane as a co-solvent. The simplified block diagram of the process for removal of lubricant oil from waste plastics with the use of 10 mol% hexane is shown in Fig. 7. The material balance
CO2 = 0.0931 kg/hr H2O = 0.0003 kg/hr Oil = 0.3625 kg/hr
Extractor
of the proposed process is shown in Fig. 8. The main equipment in this process is the same as that in Case I. However, the conditions in the separator were set to 5 MPa and 120 8C in order to separate hexane out of the liquid stream and recycle it back to the process. The cleaning efficiency of this process was found to be 91.74% compared to the oil feed to the extractor. CO2 and hexane losses from this process were calculated to be 0.180% and 2.63%, respectively. Due to the high temperature required in the separator (120 8C) in the Case II, the overall energy consumption for this process was found to be 56.4 kW/30 kg of waste plastics, which was 2.2 times higher than that in Case I. As shown in Fig. 8, the cleaning process had a loss of 0.2945 kg/h of hexane. It was assumed that the operating hours for the cleaning process were 8 h/day. Therefore, the
CO2 = 51.6259 kg/hr
CO2 = 51.6259 kg/hr
H2O = 0.0258 kg/hr
H2O = 0.0258 kg/hr
Oil = 0.375 kg/hr C6H14 = 11.2117 kg/hr
C6H14 = 0.2945 kg/hr
Expansion Valve
Oil = 0.375 kg/hr
Heater 1
C6H14 = 11.2117 kg/hr
CO2 = 51.6259 kg/hr H2O = 0.0258 kg/hr Oil = 0.375 kg/hr C6H14 = 11.2117 kg/hr
CO2 = 51.5328 kg/hr H2O = 0.0256 kg/hr
CO2 = 0.0931 kg/hr
Oil = 0.0124 kg/hr Separator
C6H14 = 10.9172 kg/hr
H2O = 0.0003 kg/hr Oil = 0.3625 kg/hr
CO2 = 51.5328 kg/hr
C6H14 = 0.2945 kg/hr
H2O = 0.0256 kg/hr Oil = 0.0124 kg/hr C6H14 = 10.9172 kg/hr
CO2 = 51.5328 kg/hr Cooler 2
H2O = 0.0256 kg/hr
Pump
CO2 = 51.5328 kg/hr H2O = 0.0256 kg/hr
Oil = 0.0124 kg/hr
Oil = 0.0124 kg/hr
C6H14 = 10.9172 kg/hr
C6H14 = 10.9172 kg/hr
Cooler 1
Fig. 8. Material balance for the proposed CO2 cleaning process with the use of hexane.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009
G Model
JIEC-2748; No. of Pages 8 M. Charoenchaitrakool et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx
8
Table 2 Estimated operating cost for cleaning 30 kg of waste plastics. Process
CO2 loss (US$)
Hexane loss (US$)
Electricity (US$)
Total operating cost (US$)
LCO2 LCO2 + hexane
0.74 0.99
8.25
1.43 3.12
2.17 12.36
amount of hexane loss was 2.356 kg/day. Based on the price of hexane in Thailand (US$ 3.5/kg), the cost of hexane loss was calculated to be US$ 8.25. Table 2 shows the estimated operating cost for cleaning 30 kg of waste plastics. It was found that the operating cost in Case II was approximately 5.7 times greater than that in Case I. The major contribution of this high cost is the hexane loss. Conclusions The feasibility of lubricating-oil removal from HDPE plastic containers using liquid carbon dioxide was investigated. Prior to the plastic cleaning process, the % lubricating oil in waste plastic from various sections of the container was determined, with the average value being found to be 2.49%. An increase in the operating pressure from 5.5 to 6.5 MPa resulted in a higher % cleaning efficiency. However, as the temperature or the contact time between oil and container increased, a slight reduction in % cleaning efficiency was observed. With the use of liquid CO2 operated at 6.5 MPa and 5 8C, the highest % cleaning efficiencies were achieved, i.e., 91.0% for contaminated new plastic and 44.8% for waste plastic. In addition, it was found that cleaning waste plastic using liquid CO2 operated at 6.5 MPa and 5 8C was approximately half as efficient as cleaning with neat hexane. The cleaning efficiency was enhanced by adding hexane as a co-solvent. Based on the MFI tests, plastics cleaned by liquid CO2 at 6.5 MPa and 5 8C exhibited the MFI values within the range of the accepted standard values. This promising result shows that the cleaned plastics can be recycled as a starting material for new containers. Acknowledgements The authors gratefully acknowledge the financial supports of the Kasetsart University Research and Development Institute
(KURDI), Faculty of Engineering Kasetsart University, the Thailand Research Fund (TRF), and the Center of Excellence on Petrochemical and Materials Technology. References [1] J.N. Sahu, K.K. Mahalik, H.K. Nam, T.Y. Ling, T.S. Woon, M.S.B.A. Rahman, Y.K. Mohanty, N.S. Jayakumar, S.S. Jamuar, Environ. Prog. Sustainable Energy 33 (2014) 298. [2] L. Saeed, R. Zevenhoven, Energy Sources 24 (2002) 41. [3] S.H. Hamid, M.B. Amin, J. Appl. Polym. Sci. 55 (1995) 1385. [4] S.M. Al-Salem, P. Lettieri, J. Baeyens, Waste Manage. 29 (2009) 2625. [5] M. Sa´nchez-Soto, A. Rossa, A.J. Sa´nchez, J. Ga´mez-Pe´rez, Waste Manage. 28 (2008) 2565. [6] Zˇ. Knez, E. Markocˇicˇ, M. Leitgeb, M. Primozˇicˇ, M. Knez Hrncˇicˇ, M. Sˇkerget, Energy 77 (2014) 235. [7] J. Martı´nez, A.C. de Aguiar, Curr. Anal. Chem. 10 (2014) 67. [8] B.A.S. Machado, C.G. Pereira, S.B. Nunes, F.F. Padilha, M.A. Umsza-Guez, Sep. Sci. Technol. 48 (2013) 2741. [9] H. Valde´s, R. Sepu´lveda, J. Romero, F. Valenzuela, J. Sa´nchez, Chem. Eng. Process.: Process Intensification 65 (2013) 58. [10] L. Fiori, Chem. Eng. Process.: Process Intensification 49 (2010) 866. [11] X. Zhang, B. Han, Clean 35 (2007) 223. [12] M.J.E. van Roosmalen, G.F. Woerlee, G.J. Witkamp, J. Supercrit. Fluids 27 (2003) 337. [13] M. Viguera, J.M. Go´mez-Salazar, M.I. Barrena, L. Calvo, J. Supercrit. Fluids 73 (2013) 51. [14] W.W. Liu, B. Zhang, Y.Z. Li, Y.M. He, H.C. Zhang, Appl. Surf. Sci. 292 (2014) 142. [15] W.W. Liu, M.Z. Li, T. Short, X.C. Qing, Y.M. He, Y.Z. Li, L.H. Liu, H. Zhang, H.C. Zhang, J. Cleaner Prod. 93 (2015) 339. [16] G. Della Porta, M.C. Volpe, E. Reverchon, J. Supercrit. Fluids 37 (2006) 409. [17] A. Marsal, P.J. Celma, J. Cot, M. Cequier, J. Supercrit. Fluids 16 (2000) 217. [18] E. Ramsey, Q. Sun, Z. Zhang, C. Zhang, W. Gou, J. Environ. Sci. 21 (2009) 720. [19] PTT, Product Data Sheet, 2015 hhttp://pttweb2.pttplc.com/weblub/Files/Lube/ attach/6_tPDS.pdfi (Last accessed: 30 September 2015). [20] G.S. Gurdial, N.R. Foster, Ind. Eng. Chem. Res. 30 (1991) 575. [21] A.A. Carey, The dielectric constant of lubrication oils, Computational Systems Incorporated, Knoxville, TN, 1998. [22] H.G. Lester, The Complete Corrugated Polyethylene Pipe Design Manual and Installation Guide, Plastic Pipe Institute, Irving, TX, 2011.
Please cite this article in press as: M. Charoenchaitrakool, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.12.009