Measurement of Emissions from a Passenger Truck Fueled with Biodiesel from Different Feedstocks ()
1. Introduction
As global population increases and developing countries industrialize, energy demand around the world is increasing markedly. World energy consumption is expected to increase by 50% by 2020 [1]. According to the Energy Information Administration’s (EIA’s) International Energy Outlook, world demand for crude oil is expected to grow from 84 million barrels/day in 2005 to over 114 million barrels/day in 2030 [2].
Biodiesel has generated increased interest in the US and elsewhere recently as an alternative to petroleumderived diesel. Because it can be produced from domestic natural sources such as soybeans, canola, and recycled cooking oil, biodiesel can help reduce dependence on petroleum fuel from foreign sources, and thus foster energy independence [3]. Due to its high oxygen content, biodiesel typically burns more completely than petroleum diesel, and thus has lower emissions of hydrocarbons (HC), carbon monoxide (CO), and particulate matter. Since biodiesel is essentially free of sulfur, SO2 emissions are negligible. However, a B20 blend of biodiesel may increase or decrease nitrogen oxide (NOx) emissions (from +12% to −13%) and carbon dioxide (CO2) emissions (from +1% to −67%), depending on biodiesel feedstock, engine type, and test cycle [3-7].
Biodiesel feedstocks that reduce NOx would be beneficial to US regions facing problems with ground-level ozone, fine particulates, and/or acid precipitation. From the viewpoint of climate change, all regions should avoid biodiesel from feedstocks that increase CO2 emissions. Therefore, determining which biodiesel feedstocks increase and decrease NOx and CO2 emissions is important.
Various studies have compared NOX emissions from biodiesel from one feedstock with those from petroleum diesel, based on chassis or engine dynamometer testing [7-10]. Several additional studies of NOx emissions from one biodiesel feedstock have involved on-road or in-use testing [11-14]. Comparing emissions from various feedstocks across studies can be difficult due to different engines and testing protocols.
Various studies [5,6,8,15-20] have compared NOx emissions from multiple biodiesel feedstocks using the same engine and testing protocol with either engine or chassis dyno testing. None involved on-road testing. None of the studies looked at NOx emissions from biodiesel from animal fat, which is potentially important as a waste product source that would not compete with food crops.
For a given engine type, the amount of CO2 produced should correlate with the fuel carbon content, which would vary with feedstock. Various studies have compared CO2 emissions from biodiesel from one feedstock with those from petroleum diesel, based on chassis or engine dyno testing [7-9,21,22]. Several additional studies of CO2 emissions from one biodiesel feedstock have involved on-road testing [12,14]. As for NOx, comparing CO2 emissions from various feedstocks across studies can be difficult due to different engines and testing protocols.
A few studies [8,17,18,23] have compared CO2 emissions from biodiesel from multiple feedstocks using the same engine and testing protocol, using chassis or dyno testing. None involved on-road testing. None of the studies looked at CO2 emissions from biodiesel from animal fat.
The study described in this paper aimed to go beyond previous studies of NOx and CO2 emissions from multiple biodiesel feedstocks, in terms of test conditions and feedstocks, by: 1) involving more test states/cycles/conditions (on-road arterial, on-road highway, steady-state idling, and chassis dynamometer testing), and 2) by including feedstocks previously untested (canola for NOx and animal fat for NOx and CO2).
2. Methodology
2.1. Biodiesel Feedstocks, Blends, and Fuel Properties
Biodiesel from soybean oil, canola oil, waste cooking oil, and animal fat was tested. Soybean and canola oil were selected to represent biodiesel made from vegetable oil. Soybean oil is the primary feedstock in the US because the U.S. is the largest producer of soybean oil. Biodiesel made from non-food crops is called “second generation” biodiesel, and is increasingly emphasized so that it does not compete with food as an end-use of crops. Waste cooking oil and animal fat biodiesel were included as examples of second generation biodiesel feedstocks. B20 (biodiesel blended with 20% with ultra low sulfur diesel, or ULSD) was tested, because such blends can be used in existing engines without modification [16]. ULSD was tested as a baseline fuel.
Biodiesel from soybean, canola, and waste cooking oil were produced at the Mechanical and Aerospace Engineering Laboratory at UT Arlington. Details of the transesterification method can be found in Pala-En (2012) [24]. Animal fat biodiesel was from Houston Biodiesel. Iowa Central Fuel Testing Laboratory measured physical/chemical properties of ULSD (ASTM D 975) and B20 blends (ASTM D 7467). C, H, O, N content was measured by Intertek QTI.
2.2. Test Vehicle
The test vehicle was a 1994 diesel Chevy C/K 2500 3/4 ton pickup truck, with an extended cab and long bed, which is owned by the UT Arlington research group. Information about the vehicle is given in Table 1.
2.3. On-Road Testing Equipment
Emissions were measured as the vehicle travelled onroad using a Horiba On-Board emission measurement System OBS-1300. NOx and CO2 concentrations were measured second-by-second, along with vehicle velocity and exhaust temperature, pressure, and flow rate. The OBS-1300 system consists of a MEXA-720 NOx analyzer, a MEXA-1170 HNDIR analyzer, 2 12 V deep cycle batteries and power supply unit to convert DC to AC, a data logger PC and other accessories. The MEXA-720 NOx analyzer is a non-sampling type zirconium sensor that measures NOx concentrations, with accuracy ±3.3 ppm. The NOx sensor probe is attached to the tail pipe. The HNDIR analyzer, used to measure CO2, has a heated tube attached to the tail pipe, which takes in the sample for analysis. According to the manufacturer, accuracy for the CO2 emission measurements was ±0.3%. A GPS unit is provided log vehicle position as a function of time. Differences in vehicle position with time were used to calculate vehicle velocity; differences in vehicle velocity
Table 1. Specifications of test vehicle.
with time were used to calculate vehicle acceleration. Routine instrument calibrations and warm up were carried out each day before the start of each session of data collection. The sensor was also calibrated weekly as required by the protocol. Maintenance and diagnostic procedures were conducted as required.
2.4. On-Road Testing Procedure
For each biodiesel blend, 1.5 hours of highway testing and 1.5 hours of arterial testing were conducted during off-peak weekday hours (9 a.m.-4 p.m.). Off-peak hours were used so that traffic flow patterns would be more similar among runs. As shown in Figure 1, the arterial test route was 6.5 miles, centered on UT Arlington, and the highway test route was 50 miles, centered on the City of Arlington. In addition, 15 minutes of idling data were collected for each fuel. The driver was the same for all testing, to eliminate variability associated with different driving habits. The truck was warmed up prior to data collection to avoid cold-start conditions.
Figure 1. Routes for on-road testing (arterial route at top, highway route at bottom).
The order of testing of each fuel and route was randomly chosen to reduce systematic error and bias conditions. Each B20 blend and route was tested with 2 replications, in the order shown in Table 2. The fuel tank was drained between runs.
2.5. Dynamometer Testing Procedure
Chassis dynamometer testing was conducted at the University of Houston’s Texas Diesel Testing and Research Center using a 500 HP AC chassis dynamometer (Burke Porter, model 6356 - 6419). The Urban Dynamometer Drive Schedule (UDDS), shown in Figure 2, was selected as it simulates a combination of low speed and idling as well as high speed and high acceleration driving. The driving cycle test consisted of:
1) Vehicle fuel system purge with the fuel to be tested;
2) Vehicle warm-up;
3) Set of three repeated runs, with a 20-minute soaking period between runs.
The ambient temperature was controlled at 23˚C - 25˚C. Emissions of NOx, total hydrocarbons (THC), CO, CO2 and oxygen (O2) were analyzed by a five gas analytical bench (Horiba, MEXA 7100). PM was collected and measured using filtration capture and gravimetric method. Fuel consumption data were collected from direct gravimetric measurement, and emission carbon balance inference.
3. Results
3.1. Fuel Properties
Table 3 (at end) compares properties of biodiesel fuel from different feedstocks with ULSD fuel. Viscosity is important due to the potential of high viscosity to adversely affect fuel injection [16]. None of the viscosities