A Well-to-Wheels Lifecycle Assessment of Used Vegetable Oil Biodiesel Produced On MIT Campus Alexander Pak Biodiesel@MIT Summer 2007 8/17/07 1 Executive Summary ..............................................................................................3 Methodology and Assumptions .............................................................................4 2.1 Scope of the Lifecycle Analysis ...................................................................4 2.1.1 Process Scope......................................................................................4 2.1.2 Geographic Scope ................................................................................5 2.2 Petroleum Diesel Model ..............................................................................6 2.2.1 Crude Oil Extraction..............................................................................6 2.2.2 Crude Oil Transportation.......................................................................7 2.2.3 Crude Oil Refining ................................................................................7 2.2.4 Diesel Fuel Transportation....................................................................9 2.2.5 Diesel Fuel Use/Combustion ................................................................9 2.3 UVO-Biodiesel Model ................................................................................10 2.3.1 UVO Collection/Transportation ...........................................................10 2.3.2 UVO Transesterification......................................................................12 2.3.3 Biodiesel Transportation/Fueling ........................................................14 2.3.4 Biodiesel Use/Combustion..................................................................14 Data and Analysis ...............................................................................................16 3.1 Petroleum Diesel Results ..........................................................................16 3.1.1 Energy Lifecycle .................................................................................16 3.1.2 Emissions Lifecycle ............................................................................17 3.2 Biodiesel Results.......................................................................................18 3.2.1 Energy Lifecycle .................................................................................18 3.2.2 Emissions Lifecycle ............................................................................18 3.3 Comparison of Petro-Diesel and Biodiesel ................................................19 Conclusions ........................................................................................................22 Appendix 1: Crude Oil Extraction ........................................................................23 Conventional Onshore Extraction....................................................................23 Conventional Offshore Extraction....................................................................23 Advanced Steam Injection...............................................................................24 Advanced CO2 Injection ..................................................................................25 Crude Oil Extraction Lifecycle Inventory..........................................................26 Appendix 2: Transportation Models for Petro-Diesel and Biodiesel ....................27 Crude Oil Transportation .................................................................................27 Diesel Fuel Transportation ..............................................................................29 UVO Collection/Transportation........................................................................30 Biodiesel Fuel Transportation..........................................................................34 Appendix 3: Crude Oil Refinery...........................................................................36 Appendix 4: Transesterification Reactor .............................................................37 Appendix 5: Engine Combustion of Petro-Diesel and Biodiesel..........................38 Diesel Fuel Combustion ..................................................................................38 Biodiesel Fuel Combustion..............................................................................38 References .........................................................................................................40 2 Executive Summary Biodiesel@MIT, a student group at MIT, has been working to install on campus a batch reactor that recycles used vegetable oil (UVO) from dining halls and converts it into biodiesel fuel. The group has determined that MIT has the potential to produce 3000 gallons of biodiesel every year, enough to run the MIT Fleet on B20 (20% biodiesel, 80% petroleum diesel blend) fuel. Biodiesel is theoretically a cleaner alternative energy source as it comes from an agricultural feedstock and thus, requires less fossil fuel in order to produce it. However, many other energy and material inputs are required to produce biodiesel instead of petroleum diesel. Besides the actual production of biodiesel, MIT also has to store, collect, and filter UVO and then distribute the biodiesel. The purpose of this study is to do a comprehensive audit of the energy input and emissions of two possible scenarios: MIT’s continued use of petroleum diesel or a switch to a biodiesel blend fuel source. In the first scenario, the comprehensive audit considered all of the material and energy inputs/outputs involved in petroleum diesel use, from oil extraction to enduse combustion. This assessment is known as a well-to-wheels analysis, a type of lifecycle analysis (LCA). An extensive lifecycle analysis incorporates factors such as the construction and decommissioning of the equipment used. Those factors, however, were not included within the scope of this study. In the second scenario, both petroleum diesel and biodiesel blend together in the end product. The audit, then, must also consider the material and energy inputs/outputs involved in the production of UVO-based biodiesel, from the gathering of UVO to end-use combustion. To this end, this study integrates different lifecycle studies that have been done on both petroleum diesel and biodiesel with varying assumptions. Following their methodologies, this study has developed a model that simulates the energy and emissions flows in petroleum and biodiesel use specific to MIT Campus. Based on the study’s proposed model, biodiesel is a cleaner, environmentally friendlier alternative to petroleum diesel. Lifecycle energy use, carbon dioxide emissions and other air emissions such as NOx are all reduced in biodiesel use. The following table summarizes the benefits of biodiesel. Type Energy Use Fossil CO2 NOx SOx CO PM10 All HC % Reduction in B100 81.3 91.7 0.4 38.2 44.2 79.6 66.7 3 % Reduction in B20 16.3 18.3 0.1 7.7 8.8 15.9 13.3 Methodology and Assumptions The purpose of this study is to document and determine the environmental feasibility of introducing a UVO-based biodiesel blend into MIT’s Fleet. Furthermore, this study also aims to provide a detailed lifecycle analysis methodology for future studies to use and improve in their own LCA’s. The well-to-wheels assessment of petroleum diesel mainly follows the methodology of the 1998 report by the National Renewable Energy Laboratory entitled “Life Cycle Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus.” The assessment of biodiesel is based on a custom model this study developed as well as a model developed by Argonne National Laboratory, “Greenhouse Gases, Regulated Emissions, Energy Used in Transportation” (GREET). The study emulates the life cycle inventory (LCI) method done in the NREL report. The LCI quantifies and compares the energy and material flows involved over the entire lifecycle of both petroleum diesel and biodiesel. The report, and this study, then compares the two fuels on the basis of a brakehorsepower-hour (bhp-h). This comparison is necessary as fuel chemistry affects engine performance. 2.1 Scope of the Lifecycle Analysis 2.1.1 Process Scope The two scenarios this study is comparing are the usage of petroleum diesel versus the usage of UVO-biodiesel blend in MIT’s fleet. In the LCA, the study considers all of the processes involved in the production and use of petroleum diesel. These processes are as follows: • • • • • Extraction of crude oil Transportation of crude oil to oil refineries Refining of crude oil into diesel fuel Transportation of diesel fuel to site Engine use of diesel fuel For each of these steps, an LCI was done to compare the overall energy inputs and emissions with those of the second scenario. However, rather than considering the entire lifecycle of the biodiesel fuel, the scope of the study was limited to only the processes that would change with the installation of the biodiesel reactor. For the purposes of this study, the models will follow the energy inputs and emissions for B100. However, when the study compares the petroleum diesel to the biodiesel blend, this study’s results will be converted to reflect those of B20. These processes are as follows: • Collection and transportation of the used vegetable oil to the reactor 4 • • • Conversion of UVO to biodiesel Transportation and fueling of the fleet with B100 Engine use of B100 The second scenario disregards the feedstock agriculture, transportation and conversion to oil of the biodiesel as both scenarios include the usage of vegetable oil on campus dining halls. 2.1.2 Geographic Scope Most of the existing literature in biodiesel LCA’s focus their analyses on a national average. This study, however, tries to specifically base its analysis on the location of MIT’s campus. Most of the crude oil in Massachusetts comes from foreign sources. Therefore, the study includes international data in its assessment. Using Petroleum Administration for Defense District (PADD) data and Massachusetts’ Energy Profile, the LCA shapes the transportation models to reflect Massachusetts’ specific means of receiving and refining both crude oil and diesel fuel. Unfortunately, not enough information could be found to study crude oil extraction and refining on a geographic specific level. Instead, this study uses the 1998 NREL report’s data based on international and national averages respectively. The following two tables summarize the geographic scope of this study. Table 1 – Geographic Scope of Petroleum Diesel Processes Process Crude Oil Extraction Scope International averages based on MA’s crude oil sources PADD 1 averages National averages Massachusetts averages National averages Crude Oil Transportation Crude Oil Refining Diesel Fuel Transportation Diesel Fuel Combustion Table 2 – Geographic Scope of Biodiesel Processes Process UVO Collection/Transportation UVO Conversion to Biodiesel Biodiesel Fueling/Transportation Biodiesel Fuel Combustion Scope MIT campus averages Reactor specific MIT campus averages National averages 5 2.2 Petroleum Diesel Model 2.2.1 Crude Oil Extraction To model petroleum diesel, the study created a LCI based on the LCI’s used in the 1998 NREL report’s model of both domestic and foreign crude oil extraction. Three main processes were used to encompass the entirety of crude oil extraction- onshore production, offshore production, and enhanced recovery. Enhanced recovery describes the process of injecting steam or CO2 into the ground to recover crude oil. The following table describes the domestic and foreign allotment of each process in the 1998 NREL report: Table 3 – Domestic and Foreign Crude Oil Extraction by Process Process Type Conventional Onshore Conventional Offshore Steam Injection CO2 Injection Domestic Allotment (%) 69 20 6.93 4.07 Foreign Allotment (%) 77 20 1.89 1.11 The report also provides LCI’s for domestic crude oil extraction and foreign crude oil extraction. These LCI’s measure raw material inputs, energy inputs, and air emissions in a kilogram of petroleum diesel. The 1998 NREL report, however, allocates 50% of all crude oil in the US to domestic production, while the other 50% is due to foreign imports. This study only recognized crude oil profiles in Massachusetts’ district, PADD 1. This district receives 99% of its crude oil from foreign sources1: Table 4 – PADD 1 Crude Oil Profile (Thousand Barrels) YR2005 Source Domestic Crude Oil Foreign Crude Oil Barrels Received 8299 1316138 % of Total Barrels 0.63 99.37 In order to calculate the energy required and emissions for diesel fuel from crude oil extraction, this study assigned relative contributions of the energy and emissions from domestic and foreign extraction. The 1998 NREL study uses mass % allocation to determine how much of the LCA is attributed to diesel fuel once the crude oil is refined. Therefore, we only need to employ one addition formula, equation (1), to calculate the relative contributions: Relative Contribution of Crude Oil Extraction = ηNREL * %Source / 100 1 “Petroleum Navigator.” EIA 6 (1) The variable ηNREL refers to the numbers in the 1998 NREL report’s LCI while the variable %Source refers to the percentages in Table 4. For an extensive description of the crude oil extraction model, see Appendix 1. 2.2.2 Crude Oil Transportation The crude oil transportation model in the 1998 NREL report again categorized crude oil by source, mode of transportation, and PADD. The modes of transportation considered were by pipeline, tanker, barge, railcar, and truck. Using average contributions of each mode of transportation over all 5 PADDs, the NREL report composed LCI’s for a kilogram of domestic and foreign crude oil transportation. In this study, the LCI was changed to meet the assumptions pertaining to MIT, particularly the change to PADD 1. The following table shows the composition of crude oil transportation in PADD 1 based on mode: Table 5 – Domestic and Foreign Crude Oil Transportation by Mode for PADD 1 YR2005 Mode Pipeline Tanker Barge Railcar Truck Domestic Barrels (Thousands) 2481 0 3171 1260 3334 % of Dom. Total 24.21 0 30.95 12.30 32.54 Foreign Barrels (Thousands) 23816 478191 82539 0 0 % of Foreign Total 4.07 81.81 14.12 0 0 Table 5 depicts the total number of barrels transported by a specific mode of transportation to and within PADD 1. Using these numbers, it’s possible to calculate the relative contributions of each mode of transportation with regards to energy use and emissions by using equation (2): R.C. of Transportation by Mode = ηNREL / (%NREL / 100) * (%2005-Trans / 100) (2) The variable ηNREL refers to the numbers in the NREL report’s LCI, while %NREL refers to the percentage breakdown of transportation modes assumed in the report (found in Appendix 2) and %2005-Trans refers to the numbers in Table 5. Equation (3) then calculates the relative contribution of domestic and foreign crude oil to the entirety of crude oil transportation: R.C. of Crude Oil Transportation = (R.C. of Trans. by Mode) * %Source / 100 (3) For more information, see Appendix 2. 2.2.3 Crude Oil Refining Due to the lack of data, the crude oil refining model discards the geographic PADD separation and instead basis its analysis on a national average of data 7 available from US refineries. Figure 1 describes the model used by the 1998 NREL report. A mass allocation by % was used to determine the energy use and emissions specific to the production of diesel fuel. When the NREL study was conducted, crude oil refineries in the US produced 13.4 weight % of the mass of their total products in diesel fuel with less than 0.05% sulfur content. Therefore, 13.4% of total energy use and emissions were allocated to the production of diesel fuel in that report. Since 1998, refinery technologies have developed and improved the efficiency of their systems. Energy use and emissions are decreased for every kilogram of processed crude oil. However, diesel fuel production has also increased to 18.1 weight % as of 20052. Considering that the diesel fuel this report is assessing, diesel with 15ppm and under sulfur content, is 82 weight % of the total crude oil processed as diesel fuel with less than 0.05% sulfur content3, crude oil refineries have increased efficiency and increased production of diesel fuel, the study determined that the LCI of the 1998 report would best reflect a net estimate of the refineries’ energy use and emissions today. Figure 1 – The Crude Oil Refinery Model used by the 1998 NREL report For more information, see Appendix 3. 2 3 “Petroleum Navigator.” EIA “Petroleum Navigator.” EIA 8 2.2.4 Diesel Fuel Transportation In the 1998 NREL report, the diesel fuel transportation model follows two routes: transportation via pipeline to truck and local transportation via truck. Using data from the Association of Oil Pipelines, their model determined that the average pipeline ships diesel fuel 595 miles while trucks travel an average of 100 miles. The model used in this study incorporates the LCI done by the NREL report and modifies it to meet the diesel fuel transportation profile for Massachusetts. Diesel fuel is assumed to be transported via three main modes: pipeline, truck and barge. Massachusetts receives most of its land-shipped diesel fuel from two diesel storage facilities. These facilities are located in New Haven, CT and Providence, RI. Diesel is transported via pipeline from these sites to Springfield, MA. The diesel fuel is then shipped by truck to its local surroundings, including MIT. Diesel fuel is also transported by barge from NYC’s harbors to Boston’s harbors. The following table describes the different modes of transportation: Table 6 – Diesel Fuel Transportation by Mode YR2005 From New Haven, CT Providence, RI Springfield, MA NYC Harbor To Springfield, MA Springfield, MA MIT Boston Harbor Mileage 65 86 90 334 Mode Pipeline Pipeline Truck Barge The LCI for diesel fuel transportation in the NREL report only includes pipeline and truck transportation. However, since transportation of crude oil and transportation of diesel fuel are relatively similar, it’s possible to use the LCI for the transportation by barge of crude oil. In the 1998 NREL report, it assumes that domestic barges travel an average of 200 km. The study then calculated the relative contributions of each mode to transport a kilogram of diesel fuel by using equation (4): Relative Contribution by Mode = ηNREL / (DNREL / 100) * (D2005-Trans / 100) (4) The variable ηNREL refers to the LCI numbers in the 1998 NREL report. The variable DNREL refers to the distances assumed in the NREL report while D2005Trans refers to the distances in the current model. For more information, see Appendix 2. 2.2.5 Diesel Fuel Use/Combustion The combustion of diesel fuel in the engine is when the energy content of the fuel is converted into work and tail-pipe emissions are released. This study assumes 9 that the average fuel economy of the trucks used is 7250 Btu/bhp-h. Since the lower heating value for diesel is 43.5 MJ/kg, 0.172 kg of fuel delivers 1 bhp-h. It is important to convert the energy inputs and emissions from the previous LCI’s into MJ/bhp-h and g/bhp-h respectively as that will be the standard this study compares petroleum diesel to biodiesel. End-use combustion of the diesel is when the most emissions are released. Many studies have been performed to measure engine emissions using diesel fuel. This study bases engine emissions of total hydrocarbons excluding CH4 (THC), nitrous oxides (NOx), carbon monoxide (CO) and particulate matter less than 10 microns in diameter (PM) on Society of Automotive Engineers (SAE) studies4. Carbon dioxide emissions were taken from the 1998 NREL report. CH4 emissions are based on a study done by the Aerodyne Mobile Laboratory conducted in 2004. Sulfur dioxide emissions are estimated based on the sulfur content of the fuel. During combustion, all of the sulfur is converted into sulfur dioxide. Therefore, sulfur dioxide emissions for a kilogram of diesel fuel can be calculated with the following equation: Sulfur Dioxide Emissions = wt% of sulfur * conversion factor of sulfur to SO2 (5) The conversion factor is given by the molar mass ratio of SO2 to S. For more information, see Appendix 5. 2.3 UVO-Biodiesel Model 2.3.1 UVO Collection/Transportation MIT’s campus dining facility currently disposes UVO through Baker Commodities, a company that collects and then disposes of the waste oil. In this study’s proposed scenario, collection and transportation of the UVO would be transferred to MIT Facilities. This study uses Baker Commodities collection protocol in order to model the energy use and emissions of UVO collection and transportation. The proposal suggests that UVO be collected from seven dining locations. Using the ruler function in Google Earth, the study measured the estimated mileage a truck would travel to pick up UVO from each location and transport it to W92, the proposed reactor location. Table 7 describes the UVO collection/transportation model based on information from Ward Ganger, Head of MIT’s Campus Dining: 4 “Comparison of Chassis Dynamometer In-Use Emissions with Engine Dynamometer FTP Emissions in Three Heavy Duty Diesel Vehicles” and “Transient Emissions from #2 Diesel and Biodiesel Blends in a DDC Series 60 Engine.” 10 Table 7 – UVO Collection/Transportation Model Site Faculty Club Lobdell Baker McCormick Next House Simmons Hotel@MIT W92 to Garage Miles per Trip 2.35 1.21 1.22 1.22 1.22 0.87 0.87 0.84 UVO per Trip (gal) 55 110 55 55 55 55 55 - Trips per Year 3 12 3 2 2 4 10 36 UVO per Year (gal) 165 1320 165 110 110 220 550 - Miles per Year 7.05 14.52 3.66 2.44 2.44 3.48 8.70 30.24 Using this model, the study calculated the energy required and emissions for the transportation of UVO. The energy required is based on the energy content of petroleum diesel, 146.4 MJ/gal, and an estimate of its fuel economy. The study adopts an estimate of 4 MPG for a heavy duty truck5. Using equation (6), the study then calculated the total energy required to transport UVO in a year: Energy to Transport UVO = DMPY / F.E.Diesel * E.C.Diesel (6) The variable DMPY refers to the total mileage trucks must travel in a year to collect UVO. The variable F.E.Diesel refers to the fuel economy of diesel and E.C.Diesel refers to the energy content of diesel. Next, the study found the amount (in kg) of biodiesel MIT could produce in a year. Based on results in MIT’s laboratories, 80 weight % of the UVO is converted into biodiesel. Using equation (7), the study calculated the amount of biodiesel produced yearly from UVO: Biodiesel Produced Yearly = VGPY * ρUVO * % Conversion to Biodiesel (7) The variable VGPY refers to the total gallons of UVO collected in a year, while ρUVO refers to the density of UVO, 3.33 kg/gal. Finally, the study calculated the average energy used to transport a kilogram of biodiesel: Energy per kg Biodiesel: Energy to Transport Yearly / Biodiesel Yearly (8) Emissions data is based on The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET) created by Argonne National Laboratory. The GREET model has an extensive database of energy use and emissions of different vehicle types with various fuels. This study uses the emissions values for CIDI (compression-ignition, direct-injection) diesel engine vehicles in order to calculate total emissions per kilogram of biodiesel with the following equation: Emissions by Transportation = ηGREET * DMPY / Biodiesel Produced Yearly (9) 5 1998 NREL report. 11 The variable ηGREET refers to the emissions numbers from the GREET model. For more information, see Appendix 2. 2.3.2 UVO Transesterification Biodiesel@MIT plans to install a batch reactor from MBP, Bioenergy, LLC to convert collected UVO into biodiesel. The reactor has its own patented technology and doesn’t require steam. Therefore, the three main sources of energy use and emissions are the electricity to run the reactor, methanol (MeOH) production and potassium hydroxide (KOH) production. In this study, the energy use and emissions were calculated separately for each source. The energy required to power the reactor were estimated based on conservative estimates of the pumps and heating of the UVO. Based on these calculations, the study estimates 52.6 kWh / tonne of UVO6. Emissions produced from electricity usage were calculated using Tiffany Groode’s methodology for assessing MIT’s greenhouse gas inventory. Electricity is generated from various sources. Table 8 describes Massachusetts’ electricity profile based on source: Table 8 – Average Breakdown of Massachusetts Electricity by Source YR2005-2007 Source Coal Natural Gas Petroleum Hydroelectric Nuclear Other Renewables % of Massachusetts Electricity 27.5 44.6 11.7 2.2 10.8 3.7 Assuming an average of 8% energy loss through distribution and transmission losses in the Massachusetts grid, the study then calculated the energy delivered from the utilities by source for a kilogram of biodiesel: Energy from Util. by Source = Ereactor / (1-%loss/100) * %Source/100 * %Conv/100 (10) The variable Ereactor refers to the electricity required per kilogram of UVO. The variable %Source refers to the energy profile of Massachusetts electricity by source and %Conv refers to the percent weight conversion of UVO to biodiesel. Each source of electricity has its own efficiency. Table 9 describes the efficiencies of each of the sources: 6 “Comparative Life Cycle Analysis of Diesel, Commercial Diesel and Biodiesel produced from WVO.” Joseph Roy-Mayhew. 12 Table 9 – Efficiency of Sources of Electricity Source Coal Natural Gas Petroleum Hydroelectric Nuclear Other Renewables % Efficiency 34 41.2 34.2 35 34 35 Using these values, equation (10) calculates the total energy generated at the utilities by source for a kilogram of biodiesel: Total Energy by Source = Energy from Util. by Source / (%eff / 100) (10) Once the total energy used had been calculated, the study found the emissions associated with that energy separated by source. Using the conversion factors found in Table 10, the following equation was used to calculate carbon, NOx and CH4 emissions per kilogram of biodiesel: Emissions by Source = Tot. Energy by Source (MMBtu) * Conv. Factor / 100 (11) Table 10 – Emissions Factors for Energy Sources Source Coal Natural Gas Petroleum Hydroelectric Nuclear Other Renewables C Emissions (tonne C/MMBtu) 0.027 0.01633 0.0225 0 0 0 NOx Emissions (g/MMBtu) 0.75 1.1 0.91 0 0 0 CH4 Emissions (g/MMBtu) 0.298 1.1 0.36 0 0 0 Finally, to calculate CO2 emissions, the study assumes 99% combustion efficiency and converts the carbon content into CO2 using the following equation: CO2 = C Emissions * %combust/100 * Conversion Factor of C to CO2 (12) The conversion factor is equal to the molar mass ratio of CO2 to C. Other types of air emissions are excluded since emissions from MeOH will contribute over 98% of the total emissions for those types. Similarly, the energy input and emissions can be calculated for KOH production. To do this, the study assigns an electricity-equivalent energy content for the KOH (MJ/kg). Since 1 wt% of the UVO is the required amount of KOH, the study can calculate the electricity needed for the KOH necessary for a kilogram of UVO. Again, using the same methodology previously mentioned, the study can 13 calculate the total energy input and emissions for KOH production (see Appendix 4). Energy use and emissions for methanol production are based on the LCI for the Soybean Oil Conversion process in the 1998 NREL study. In the MBP reactor, an excess 6:1 molar ratio of MeOH is needed to maximize conversion to biodiesel. Therefore, 0.22 kg of MeOH is needed for every kilogram of UVO in the reactor. In the NREL report, an average of 0.08628 kg of MeOH is used for every kilogram of vegetable oil. By the following equation, the study calculated the energy use and emissions from MeOH production for every kilogram of biodiesel: Methanol LCI = ηNREL * %Conv * mMeOH / mNREL (13) The variable ηNREL refers to the LCI numbers for MeOH production from the NREL study. The variable mMeOH refers to the mass of MeOH needed in MIT’s reactor for a kilogram of UVO while mNREL refers to the mass of MeOH used to convert a kilogram of vegetable oil. For further information, see Appendix 4. 2.3.3 Biodiesel Transportation/Fueling The transportation model for the newly converted biodiesel closely follows the transportation model used for UVO collection and transportation. Essentially, trucks will drive to W92 to fill their tanks (25 gallons) with the new biodiesel blend and then drive back to the parking garage, NW62. The model differs from the UVO collection/transportation in that the gallons collected now refer to gallons of biodiesel filled. For the purpose of comparing pure biodiesel to petroleum diesel, the study assumes that the tank will be filled with B100. The study will account for the B20 blend when it compares the petroleum diesel scenario to the B20 scenario; mixing the B20 blend requires no additional energy and releases zero emissions. The same methodology to calculate energy use and emissions in the UVO collection/transportation step can be used in this step as well. However, there is no need to include the percent conversion to biodiesel as it has already been converted. For more information, see Appendix 2. 2.3.4 Biodiesel Use/Combustion End-use combustion of biodiesel is where most of the emissions in the LCA are produced. Just as the study did for end-use combustion of petroleum diesel, the study compiled data from various tests done on B100 emissions and B20 14 emissions. As a means of comparison, emissions data is given per bhp-h. For biodiesel, 0.203 kg of B100 is needed to produce 1 bhp-h7. The engine use of biodiesel is the step in which this study incorporates the CO2 released from biomass. Biodiesel is derived from a feedstock which is grown through agriculture. During agriculture, the feedstock uptakes CO2 as the plants undergo photosynthesis. Therefore, a percentage of the total carbon content of the biodiesel is actually recycled biomass carbon and should not be considered in the net emissions of CO2. On average, pure biodiesel is 77.2% carbon. In the biodiesel, the carbon contained in the fatty acids reacted with methanol is considered non-biomass carbon while the carbon in the fatty acid portion of the methyl ester is considered biomass carbon. Of the total carbon content, 73.2% of it is attributed to biomass carbon8. It is also assumed that biodiesel contains zero sulfur content. As a result, biodiesel has zero SOx emissions from end-use combustion. Emissions data for THC including CH4, CO, NOx and PM10 were taken from various research papers9 that have tested the emissions of biodiesel blends including B100 and B20. CO2 emissions were taken from the 1998 NREL report. For more information, see Appendix 5. 7 1998 NREL study. 1998 NREL study. 9 “Biodiesel Comprehensive Handbook” and “Transient Emissions of #2 Diesel and Biodiesel Blends in DDC Series 60 Engines.” 8 15 Data and Analysis In this study, LCI’s are presented for pure biodiesel, petroleum diesel and B20 blends. These LCI’s are useful for comparing energy inputs and air emissions of the different fuel types. However, the two fuel types should not be compared by mass as they will burn differently in an engine. Instead, the total energy use and emissions of the two fuels should be compared based on the work they deliver on the engine. In the study’s final analysis, the two fuels will be compared on their lifecycle energy use and emissions in delivering 1 bhp-h in a standard diesel engine. It should be noted that this lifecycle analysis does not incorporate cold flow properties of biodiesel. In cold ambient temperature, biodiesel may drastically decrease in efficiency. Other studies are currently being done to determine the best way to offset these temperature changes. 3.1 Petroleum Diesel Results In a standard diesel engine, 0.172 kg of petroleum diesel is required for 1 bhph10. 3.1.1 Energy Lifecycle The energy inputs for each step of petroleum diesel’s lifecycle are summarized in the following figure: Figure 2 – Lifecycle Energy Audit of Petroleum Diesel 10 1998 NREL report. 16 Crude oil extraction is the most energy intensive of the steps involved in petroleum diesel production. Of the total energy input, crude oil extraction makes up 89.7% of it. Foreign crude oil extraction makes up 99.3% of the total energy required for extraction since almost all oil in Massachusetts comes from foreign sources. Reducing dependence on foreign oil will greatly reduce energy use. After crude oil extraction, crude oil refining is the second most energy intensive, making up 5.7% of the total energy input. 3.1.2 Emissions Lifecycle The CO2 emissions for each step in petroleum diesel’s lifecycle are summarized in the following figure: Figure 3 – Lifecycle CO2 Emissions Audit of Petroleum Diesel Tailpipe emissions of CO2 contribute 85.5% of the total emissions of CO2. Following engine emissions, crude oil refining is the second most CO2 emissions intensive, contributing 9.8% of the total CO2 emissions. The following table shows the lifecycle air emissions of other greenhouse gases over the lifecycle of petroleum diesel: Table 11 – Lifecycle Air Emissions of Petroleum Diesel (g / bhp-h) Process NOx SOx CO Crude Oil Extraction 0.03 0.16 0.02 Crude Oil Transportation 0.02 0.19 0.00 Crude Oil Refining 0.13 0.44 0.04 Diesel Fuel Transportation 0.02 0.00 0.01 Diesel Fuel Combustion 5.01 0.05 2.52 Total 5.21 0.84 2.60 PM10 0.03 0.01 0.08 0.00 0.18 0.30 All HC 0.21 0.13 0.00 0.01 0.21 0.56 Again, combustion is where most emissions are released into the environment. 17 3.2 Biodiesel Results In a standard diesel engine, 0.203 kg of pure biodiesel is required for 1 bhp-h11. 3.2.1 Energy Lifecycle The energy inputs for each step in biodiesel’s lifecycle are summarized in the following figure: Figure 4 – Lifecycle Energy Audit of Biodiesel The batch reactor is the most energy intensive of all the steps in UVO-biodiesel production. The reactor process requires 94.9% of the total energy input. Of the reactor process, methanol production takes up 92.3% of the total energy required to convert UVO into biodiesel. 3.2.2 Emissions Lifecycle The CO2 emissions for each step in biodiesel’s lifecycle are summarized in Figure 5. Just as it was with the petroleum diesel, the greatest amount of CO2 is released during the combustion of biodiesel. Combustion emissions contribute 57.5% of the total emissions of CO2. The reactor process is the second greatest contributor of CO2. The process emits 39.0% of the total emissions of CO2. 11 1998 NREL report. 18 Figure 5 – Lifecycle CO2 Emissions Audit of UVO Biodiesel 3.3 Comparison of Petro-Diesel and Biodiesel The total energy inputs over the lifecycle of petroleum diesel, pure UVO-biodiesel and B20 are summarized in the following figure: Figure 6 – Comparison of Lifecycle Energy Input of Petro-Diesel, B100 and B20 Pure UVO-biodiesel produced on MIT campus requires 81.3% less energy in its lifecycle that petroleum diesel. In the proposed scenario, however, petroleum diesel will be replaced with B20. Petroleum diesel and B100 can be mixed in three ways: splash blending, in-tank blending and layered blending. To splash blend, petro-diesel is put into the vehicle’s tank followed by biodiesel. The mixing 19 from turning, going over bumps and changing elevations is enough to blend the fuels. In-tank blending is where petro-diesel and B100 are pumped into the tanks at the same time. Finally, layered blending is a passive form of blending in which biodiesel is layered over petroleum diesel. The B100 will sink and blend with the petro-diesel since it is denser. Therefore, no additional energy inputs are required to produce the end result B20 blend. The study calculated the total energy input required for B20 by using the following equation: B20 LCA Energy = 0.80 * EPD + 0.20 * EBD (14) The variable EPD refers to the lifecycle energy input of petro-diesel while the variable EBD refers to the lifecycle energy input of biodiesel. Based on these calculations, this study has found that B20 reduces energy use by 16.3%. The total CO2 emissions over the lifecycle of petroleum diesel, pure UVObiodiesel and B20 are summarized in the following figure: Figure 7- Comparison of Lifecycle CO2 Emissions of Petro-Diesel, B100 and B20 The B100 blend reduces CO2 emissions by 91.7%. Using the same methodology stated above, the study also calculated the CO2 emissions of the B20 blend. Based on these calculations, the study found that B20 reduced CO2 emissions by 18.3%. The lifecycle results of other air emissions for the B20 blend are summarized in the following table: 20 Table 12 – Lifecycle Air Emissions for B20 Air Emission NOx SOx CO PM10 All HC B20 Emissions (g / bhp-h) 5.21 0.78 2.37 0.26 0.49 Based on these results, the study calculated that B20 reduced all emissions, including NOx. Although biodiesel emits greater amounts of NOx during engineuse combustion, its total NOx emissions over its lifecycle still amount to fewer emissions from the lifecycle of petroleum diesel. This is mainly due to the fact that if MIT produces biodiesel, it will be recycling the already used vegetable oil instead of disposing of it. Therefore, NOx emissions from processes before UVO collection are excluded from the LCA. Table 13 summarizes the results for these air emissions: Table 13 – Comparison of Lifecycle Air Emission Reduction for B100 and B20 GHG Type NOx SOx CO PM10 All HC % Reduction in B100 0.4 38.2 44.2 79.6 66.7 % Reduction in B20 0.1 7.7 8.8 15.9 13.3 It should be noted that all SOx emissions in B100’s lifecycle come from its upstream processes rather than from combustion. 21 Conclusions Biodiesel@MIT’s proposal to implement a UVO-biodiesel production program promises to reduce MIT Fleet’s energy use and emissions by a significant amount. The lifecycle energy used to supply the diesel currently fueling MIT’s fleet can be reduced by 16.3% with the usage of B20. CO2 emissions can also be reduced by 18.3% and other GHG emissions can be reduced by as much as 15.9%. However, other factors still need to be considered such as the temperature during the winter time. Certain technologies may have to be implemented to ensure effective use of B20 during the winter. As such, these additions will affect the LCA of biodiesel. Unfortunately, the scope of this study did not include a sensitivity analysis of the data and results. Future LCA studies of MIT produced biodiesel should look into how different assumptions affect the change in results. One example of such factor is the model of the UVO collection/transportation. Collection trucks may change paths or make trips more frequently or less frequently. Future studies should also incorporate more data sets in order to accurately represent the lifecycle energy use and emissions of both petroleum diesel and biodiesel as a range of possible numbers. One example of a data set this study lacked was information on crude oil refinery and extraction technologies in PADD 1. 22 Appendix 1: Crude Oil Extraction Conventional Onshore Extraction The following figure depicts the system boundaries and flows considered in the onshore extraction model in the 1998 NREL study: Figure 8 – Conventional Onshore Crude Oil Extraction In this study, however, solid waste, water effluents and VOC emissions were not considered within its scope. Energy used to construct and decommission equipment used in extraction was excluded in both studies. The only raw materials considered were crude oil and natural gas. Energy requirements were based on the electricity used for pumping and the natural gas used for oil recovery. Due to lack of information, these assumptions were used for both domestic and foreign oil extraction. Mass allocation was used to determine how much of the energy use and emissions were due to the crude oil product versus the natural gas product. Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. Conventional Offshore Extraction Figure 9 depicts the system boundaries and flows considered in the offshore extraction model in the 1998 NREL report. 23 Solid waste, water effluents and VOC emissions as well as energy used for equipment production/destruction were excluded from this study’s model. The only raw materials considered were natural gas and crude oil. Energy requirements were based on natural gas as offshore rigs use natural gas for pumping, recovery and electricity production. These assumptions apply to both domestic and foreign oil extraction. Mass allocation was used to determine emissions for the crude oil product. Figure 9 – Conventional Offshore Crude Oil Extraction Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. Advanced Steam Injection Figure 10 depicts the system boundaries and flows considered in the 1998 NREL report for advanced steam injection. In this study, however, solid waste and VOC emissions are not considered. The raw materials considered were crude oil, natural gas, and water for steam generation. Energy requirements were based on electricity for pumping and natural gas for recovery and steam generation. It is assumed that all of the natural gas was burned for steam generation so all emissions and energy use are allocated to the crude oil product. 24 Figure 10 – Advanced Onshore (Steam Injection) Crude Oil Extraction Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. Advanced CO2 Injection The following figure depicts the system boundaries and flows considered in the 1998 NREL report on advanced CO2 injection: Figure 11 – Advanced Onshore (CO2 Injection) Crude Oil Extraction In this study, solid waste and VOC emissions were not considered within its scope. Energy used to construct and decommission equipment used in extraction 25 was excluded in both studies. The only raw materials considered were crude oil and CO2. Energy requirements were based on the electricity used for pumping and preparation of the CO2. Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. Crude Oil Extraction Lifecycle Inventory The following figures depicts the modified LCI used in this study: Figure 12 – Domestic Crude Oil Extraction LCI Figure 13 – Foreign Crude Oil Extraction LCI 26 Appendix 2: Transportation Models for Petro-Diesel and Biodiesel Crude Oil Transportation The following figure depicts the breakdown of transportation modes used in the 1999 NREL study: Figure 14 – Crude Oil Transportation Breakdown by Mode The numbers in Figure 14 are used in the equations stated in the methodology for crude oil transportation. The following figure depicts the model used for loading and unloading: 27 Figure 15 – Crude Oil Transportation Loading/Unloading 28 The following figures depict the LCI for crude oil transportation used in this study: Figure 16 – Foreign Crude Oil Transportation LCI Figure 17 – Domestic Crude Oil Transportation LCI Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. Diesel Fuel Transportation The 1998 NREL report uses the system boundaries depicted in Figure 18 to model diesel fuel transportation. Loading is modeled as it is for crude oil transportation. 29 Figure 18 – Diesel Fuel Transportation Figure 19 shows the LCI used for diesel fuel transportation in this study. Figure 19 – Diesel Fuel Transportation LCI Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. UVO Collection/Transportation The following figures depict the paths this study assumed trucks would take to collect UVO from all of the sites on campus: 30 Figure 20 – The assumed path the transportation truck would take to transport UVO from the Faculty Club to W92. Figure 21 – The assumed path the transportation truck would take to transport UVO from Lobdell to W92. 31 Figure 22 – The assumed path the transportation truck would take to transport UVO from McCormick, Baker and Next House to W92. Figure 23 – The assumed path the transportation truck would take to transport UVO from Simmons to W92. 32 Figure 24 – The assumed path the transportation truck would take to transport UVO from Hotel@MIT to W92. Figure 25 – The assumed path the transportation truck would take to get back to the garage facility from W92. The following figures depict the LCI model for UVO transportation based on the GREET model: 33 Figure 26 – UVO Transportation LCI Emissions Based on GREET Figure 27 – UVO Transportation LCI Energy Biodiesel Fuel Transportation The following depict the LCI of biodiesel fuel transportation based on the GREET model: Figure 28 – Biodiesel Transportation LCI Emissions Based on GREET 34 Figure 29 – Biodiesel Transportation LCI Energy 35 Appendix 3: Crude Oil Refinery The following figure depicts the LCI model used in this study for crude oil refining based on the 1998 NREL report: Figure 30 – Crude Oil Refining LCI Please see Life Cycle Inventory of Petroleum Diesel and Biodiesel for Use in an Urban Bus, the 1998 NREL report, for a complete description of the model. 36 Appendix 4: Transesterification Reactor The methodology for calculating energy and emissions of KOH production follow the same methodology for doing the same with electricity usage in the reactor. The following table summarizes the data pertaining to KOH that this study used: Table 14 – Electricity Equivalent Energy Required for KOH Electricity Equiv. Energy Content KOH’s Required Wt% of UVO Equivalent Elec. Needed for KOH 1.87 1 0.0052 MJ / kg KOH % kWh / kg UVO The following figures represent the LCI’s used in this study for the batch reactor: Figure 31 – Methanol Production LCI Figure 32 – UVO Conversion to Biodiesel LCI 37 Appendix 5: Engine Combustion of Petro-Diesel and Biodiesel Diesel Fuel Combustion The following figures represent emissions data used in this study: Figure 33 – Average Emissions in SAE Papers on Diesel Engine Emissions Figure 34 – Emissions Data for Diesel (SOx and CH4) Biodiesel Fuel Combustion The following figures represent emissions data used in this study for B100: Figure 35 – Average Emissions Data for B100 Combustion 38 Figure 36 – Average Emissions Data for B20 Combustion 39 References Ahlvik, Peter. Alternative Diesel Fuels. DieselNet Technology Guide. Ecopoint Inc., 2003. Argonne National Laboratory (2004) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model. Transportation Technology R&D Center, US Department of Energy, Argonne, IL. A Comprehensive Analysis of Biodiesel Impacts on Exhaust Emissions. Air and Radiation. Environmental Protection Agency, 2002. Biodiesel Emissions Database. US Environmental Protection Agency. <http://www.epa.gov/otaq/models/biodsl.htm>. Dewulf, Jo, and Herman V. Langenhove. Renewables-Based Technology Sustainability Assessment. West Sussex, England: Wiley, 2006. 168-243. “Energy Density of Methanol.” <http://hypertextbook.com/facts/2005/JennyHua.shtml>. Ganger, Ward, and Derreck Barber. Information on MIT Campus Dining. Email. Graboski, M. S., J. D. Ross, and R. L. McCormick. Transient Emissions From No. 2 Diesel and Biodiesel Blends in a DDC Series 60 Engine. Colorado Institute for Fuels and High-Altitude Engine Research. Warrendale, PA: Society of Automotive Engineers, Inc., 1996. Groode, Tiffany A., and John B. Heywood. A Methodology for Assessing MIT’s Energy Use and Greenhouse Gas Emissions. Laboratory for Energy and the Environment. Massachusetts Institute of Technology, 2004. Herndon, S.C., Shorter, J.H., Zahniser, M.S., Wormhoudt, J., Nelson, D.D., Demerjian, K.L., and Kolb, C.E. Real-Time Measurements of SO2, H2CO, and CH4 Emissions from In-Use Curbside Passenger Buses in New York City Using a Chase Vehicle. Environ. Sci. Technol., 39, 20, 7984 - 7990, 2005, 10.1021/es0482942. Kulkarni, G. Mangesh, and Ajay K. Dalai. Waste Cooking Oil- An Economical Source for Biodiesel: A Review. Catalysis and Chemical Reaction Engineering Laboratories, University of Saskatchewan. Industrial & Engineering Chemistry Research, 2006. Majewski, W. Addy. Biodiesel. DieselNet Technology Guide. Ecopoint Inc., 2005. 40 “Massachusetts State Energy Profile.” Energy Information Administration. <http://tonto.eia.doe.gov/state/state_energy_profiles.cfm?sid=MA>. McCormick, Robert L., Lisa Ryan, Teresa L. Daniels, Janet Yanowitz, and Michael S. Graboski. Comparison of Chassis Dynamometer in-Use Emissions with Engine Dynamometer FTP Emissions for Three HeavyDuty Diesel Vehicles. Colorado School of Mines. Warrendale, PA: Society of Automotive Engineers, Inc., 1998. Mittelbach, Martin, and Claudia Remshmidt. Biodiesel: the Comprehensive Handbook. Graz, Austria: Martin Mittelbach, 2004. Niederl, Anneliese, and Michael Narodoslawsky. "Ecological Evaluation of Processes Based on by-Products or Waste From Agriculture: Life Cycle Assessment of Biodiesel From Tallow and Used Vegetable Oil." Feedstocks for the Future: Renewables for the Production of Chemicals and Materials. Washington DC: American Chemical Society, 2006. 239251. The Opening of the Cape Code Canal. Bulletin of the American Geographical Society. 1914. 832-834. "Petroleum Navigator." Chart. Energy Information Administration. <http://tonto.eia.doe.gov/dnav/pet/pet_sum_top.asp>. Products of the Office of Coal, Nuclear, Electric and Alternate Fuels. Energy Information Administration. <http://www.eia.doe.gov/cneaf/>. Roy-Mayhew, Joseph. Comparative Life Cycle Analysis of Diesel, Commercial Biodiesel, and Biodiesel produced from WVO. Biodiesel@MIT. 2006. Sheehan, John, Vince Camobreco, James Duffield, Michael Graboski, and Housein Shapouri. Life Cycle Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus. US Department of Energy and US Department of Agriculture. Golden, Colorado: National Renewable Energy Laboratory, 1998. 41
© Copyright 2026 Paperzz