ALGAL-00003; No of Pages 13

Algal Research xxx (2012) xxxxxx

Contents lists available at SciVerse ScienceDirect

Algal Research

Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate
Harrison B. Bucy, Marc E. Baumgardner, Anthony J. Marchese
Colorado State University, Fort Collins, CO 80523, United States

a r t i c l e

i n f o

a b s t r a c t
Microalgae are currently receiving strong consideration as an advanced biofuel feedstock because of their theoretically high yield (gal/acre/year) in comparison to terrestrial vegetable oil feedstocks. Microalgal lipids can be readily converted into a variety of biofuels including fatty acid methyl esters (i.e. biodiesel) via transesterication or alkanes via hydroprocessing. In contrast to parafnic fuels whose properties can be tailored for a specic application, the properties of algal methyl ester biodiesel are directly related to the fatty acid composition of the algal lipids. Several microalgae species that are suitable for large scale cultivation such as those in the genus Nannochloropsis produce lipids that contain long chain-polyunsaturated fatty acids (LC-PUFA) such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). These constituents have high value as co-products but are problematic in terms of biodiesel properties such as ignition quality and oxidative stability. The objective of this study was to examine the effect of varying levels of EPA and DHA on algal methyl ester fuel properties. Oxidative stability, Cetane Number, density, viscosity, bulk modulus, cloud point and cold lter plugging point were measured for algal methyl esters produced from various microalgae feedstocks as well as model algal methyl ester compounds formulated to match the fatty acid composition of Nannochloropsis sp., Nannochloropsis oculata and Isochrysis galbana subjected to varying levels of removal of EPA and DHA. The results suggest that removal of 50 to 80% of the LC-PUFA from Nannochloropsis-based methyl esters would be sufcient for meeting existing specications for oxidative stability. However, higher levels of LC-PUFA removal from Nannochloropsis-based methyl esters would be required to produce fuels with acceptable Cetane Number. The removal of EPA and DHA was shown to have a detrimental effect on cold ow properties since the algal methyl esters are also high in fully saturated fatty acid content. 2012 Published by Elsevier B.V.

Article history: Received 30 November 2011 Received in revised form 1 February 2012 Accepted 7 February 2012 Available online xxxx Keywords: Algal biofuels Biodiesel Oxidative stability Ignition quality Cold ow properties Fatty acid composition

1. Introduction Depletion of fossil fuels and climate change from anthropogenic greenhouse gas emissions arguably represent the rst civilizationscale challenges ever faced by the human race [1,2]. While there is no single technological solution for these omnipresent challenges, biofuels have the potential to play an important role in displacing fossil fuel consumption (particularly for applications that require high energy density liquid fuels such as aviation) and mitigating anthropogenic greenhouse gas emissions. However, many of today's rst generation biofuels (e.g., corn ethanol and soy biodiesel) are limited in their capability to substantially impact global fossil fuel consumption and greenhouse gas emissions because of high energy requirements for production and/or poor yield (gallons/acre/year). For example, studies suggest that for every 1 unit of energy contained in corn
Corresponding author at: Department of Mechanical Engineering, Colorado State University, 1374 Campus Delivery, Fort Collins, CO 80523-1374, United States. Tel.: +1 970 491 2328; fax: +1 970 491 3827. E-mail address: marchese@colostate.edu (A.J. Marchese). 2211-9264/$ see front matter 2012 Published by Elsevier B.V. doi:10.1016/j.algal.2012.02.001

ethanol, the equivalent of 0.8 units of fossil fuel energy input are required [3]. Conversely, while soy biodiesel production is less energy intensive, the low yield of 60 gal/acre/year [4] prevents soy biodiesel from achieving the scale required to displace a substantial quantity of liquid fossil fuel consumption. Moreover, both soy biodiesel and corn ethanol compete with the global food supply. Accordingly, a critical need exists to develop advanced biofuels that minimize life cycle greenhouse gas emissions and can be cultivated at a scale that is commensurate with a substantial fraction of liquid fossil fuel consumption (i.e. greater than 10%). In recent years, the U.S. government has enacted policies aimed at increasing domestic production of biofuels toward a level that is consistent with approximately 10% of U.S. liquid fossil fuel consumption. Specically, the 2007 Energy Independence and Security Act (EISA) contains a mandate of 36 billion gallons of renewable fuel per year by 2022 [5]. However, the EISA places a cap of 15 billion gallons of corn ethanol per year, thereby requiring 21 billion gallons of advanced biofuels by 2022. When the EISA was enacted, it was anticipated that the advanced biofuel requirement would be predominantly met by cellulosic ethanol production but the EISA advanced biofuels mandate may

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

also be met by any non-corn starch-derived biofuel [5]. Therefore, one approach for meeting the advanced biofuels standard domestically would be to increase biodiesel and/or renewable diesel production from alternative lipid feedstocks such as microalgae. Microalgae derived biofuels have potential as scalable, environmentally benign fossil fuel replacements. Microalgae derived lipids and biomass can be converted into alcohols, methyl esters and alkanes for use in spark ignited engines, compression ignition engines and aircraft gas turbine engines. As described below, this paper focuses on the physical and chemical properties of microalgae derived fatty acid methyl esters (FAME). Specically, the goal of this study was to examine the oxidative stability, ignition quality, viscosity, density, bulk modulus, cloud point and cold lter plugging point of microalgae-based FAME. 1.1. The case for algal biofuels Depending on the species, growing conditions and growth stage, microalgae have been shown to produce various types of lipids including triglycerides, phospholipids, glycolipids and betaine lipids [6]. Since nitrogen and phosphorus are detrimental to problematic for processing systems, the triglycerides are of interest for direct conversion into biodiesel via the transesterication process or conversion into renewable diesel or synthetic parafnic aviation kerosene via hydroprocessing. Microalgae have many potential cultivation benets in comparison to terrestrial vegetable oil feedstocks. Microalgae are the fastest growing photosynthesizing organisms [7] and under certain cultivation conditions have been shown to be nature's most efcient biological oil producer [8]. Microalgae require less land for cultivation than terrestrial crops, can grow in non-potable water, and do not displace food crops [9,10]. The potential for increased productivity from microalgae-derived biofuels in comparison to current terrestrial vegetable oil derived biofuels cannot be understated. Microalgae commonly double their biomass within 24 h and biomass doubling times during exponential growth can be as short as 3.5 h. And, under specic cultivation conditions, their oil content can exceed 50% by weight of dry biomass [11]. According to the most comprehensive models, the productivity of algae derived biofuels is predicted to be on the order of 5000 gal/ acre/year, which is approximately two orders of magnitude greater than the yield from terrestrial oilseed crops such as soybeans [7,12]. If such productivity levels were achieved, it is conceivable that less than 3% of the total existing U.S. crop area would be sufcient for producing algal biomass to satisfy 50% of all the transport fuel needs for the United States [11]. In the short term, an inux of microalgae oil could also help the U.S. to better utilize the existing production capacity of the U.S. biodiesel industry, which is currently feedstock limited. In 2009, only 506 million gallons of biodiesel were produced, which was well below the 2 billion gallon/year production capacity [13]. 1.2. Algal methyl ester biodiesel Microalgal lipids, whole algal biomass and lipid extracted algae (LEA) can be used as feedstocks to produce a myriad of advanced biofuels. For example, microalgal lipids can be converted to biodiesel, renewable diesel, renewable gasoline or synthetic parafnic aviation kerosene. The latter three fuels are composed of mixtures of straight and branched alkanes with composition tailored to produce the required properties for the specic application. Conversely, the properties of algae derived biodiesel (i.e. alkyl esters produced directly from algal triglycerides via a transesterication reaction with an alcohol) are directly related to the fatty acid composition of the algal triglycerides. This study focuses on fuel properties of microalgae derived methyl ester biodiesel. The transesterication process consists of the reaction of triglyceride molecules with alcohol in the presence of a catalyst to produce glycerol and mono-alkyl fatty acid esters. If the triglycerides are derived from vegetable oil or animal fat, the mixture of mono-alkyl

esters is what is dened as biodiesel [14]. Biodiesel is typically transesteried using methanol and therefore the fatty acid alkyl esters that are produced are fatty acid methyl esters (FAME). FAME are the most prevalent alkyl esters in the current biodiesel market because of the price and availability of methanol compared to other alcohols [15]. Although the transesterication process reduces the viscosity of FAME compared to the parent oil, the fatty acid composition is not altered. This attribute of biodiesel is critical because the chain length and number of double bonds in the hydrocarbon chain of the FAME determine the characteristics of some critical parameters of biodiesel such as oxidative stability, ignition quality, cold ow properties and pollutant emissions. If algal methyl ester biodiesel is to capture a substantial percentage of the diesel fuel market, it must qualify as t for purpose as a blend stock for existing compression ignition engine technologies. To ensure that algal biodiesel is compliant, it must meet the ASTM Biodiesel Standard D6751 in the U.S. and the EN 14214 in the European Union (EU) for B100 use in diesel engines. One major difference between algal methyl esters and most common vegetable oil derived FAME is that some algal lipids contain a substantial quantity of long chain polyunsaturated fatty acids (LC-PUFA) including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [11]. The EPA fatty acid has a carbon chain length of 20 with 5 double bonds (C20:5) and the DHA fatty acid has a carbon chain length of 22 with 6 double bonds (C22:6). Table 1 contains the fatty acid composition of common biodiesel feedstocks in comparison with that of algal lipids produced from several different species [10,14,16]. The algal lipids listed in Table 1 clearly have greater quantities of LC-PUFA compared to typical feedstocks. It should also be noted that the algal lipids also have higher quantities of fully saturated fatty acids (C14:0, C16:0 and C18:0) in comparison to soy and canola. As discussed below, the higher quantities of LC-PUFA and fully saturated fatty acids both have implications in terms of fuel properties. It is well understood that the fatty acid composition (carbon chain length and degree of unsaturation) of FAME has a major effect on fuel properties. The most important characteristics affected by the level of unsaturation are oxidative stability, ignition quality (i.e. Cetane Number), and cold ow properties [14,15,17]. For example, fully saturated methyl esters have high oxidative stability and a high Cetane Number, but suffer from poor cold ow properties. Conversely, methyl esters with a higher degree of unsaturation have better cold ow properties but decreased oxidative stability and decreased Cetane Number. 1.3. Oxidative stability of methyl esters Oxidative stability is critical issue in the storage, handling and use of methyl ester biodiesel [18]. In terms of storage and handling,
Table 1 Comparison of fatty acid proles for biodiesel sources by percent weight. The fatty acids are denoted by number of carbons: number of double bonds. Fatty acid Feedstock Soy [14] Canola [14] Nannochloropsis salina [This study] Nannochloropsis oculata [16] Isochrysis galbana [10]

8:0 10:0 12:0 14:0 16:0 16:1 18:0 18:1 18:2 18:3 20:4 20:5 22:6

11 4 24 53 8

4 2 61 19 10

3 30 39 1 8 1 1 3 11

2 15 16 2 10 4 3 6 21 3

23 14 3 1 14 5 7 5 14

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

oxidative stability is an issue because decomposition byproducts can form deposits in tanks, fuel systems, and lters. During use, biodiesel oxidation products can form insoluble sediments that are capable of plugging lters, fouling injectors, and interfering with engine performance [19,20]. The oxidation process can also lead to an increase in viscosity [21] and, in extreme cases, can result in separation of the fuel into two phases thereby causing fuel pump and injector issues [22]. Factors that promote oxidation are the presence of air, light, elevated temperatures, impurities in the fuel [23], and the fatty acid composition. Specically, it has been shown that even low concentrations of polyunsaturated fatty esters can have a disproportionately large effect on oxidative stability [19]. For biodiesel, oxidative stability is determined using the EN 14112 method, which measures an induction period for oxidation when the fuel is subjected to an elevated temperature. Both the EU and the U.S. have standards for a minimum induction period at elevated temperature as prescribed by the EN 14112 method. The U.S. requires a minimum of 3 h (ASTM D6751) and the EU requires a minimum of 6 h (EN 14214). McCormick and coworkers have determined that if the induction time is near or below the three hour limit, that B100 blends will no longer meet ASTM specications after four months of storage [24]. It should also be noted that the EN is currently considering extending the minimum oxidative stability limit to 8 h. Since the oxidative stability of fatty acid methyl esters has been shown to decrease with increasing level of unsaturation, the EU biodiesel standard also contains specications for iodine value (IV), which relates to the total number of double bonds in a fat or oil. A major drawback of the IV is that it does not distinguish between the structural differences in fatty compounds such as the position of the double bonds [18,23]. Specically, it has been shown that the bisallylic sites in the fatty acid hydrocarbon chain are critical to oxidative stability because they are the location most susceptible to the attack from molecular oxygen. The allylic site is a methylene group (CH2) adjacent to a double bond and the bis-allylic site is a methylene group adjacent to two double bonds [25]. Fig. 1 shows the location of the allylic sites and bis-allylic site on an unsaturated fatty acid methyl ester. The gure shows that there are always two allylic sites whenever there are one or more double bonds and one less bis-allylic site than the number of double bonds present. It should also be noted that the bond structure in which two single bonds separate each double bond is found in all naturally occurring vegetable oils and animal fats that contain two or more double bonds. To more accurately account for the chemical structure of unsaturated fatty acids, two new saturation indices have been recently developed by Knothe [25] called the bis-allylic position equivalents (BAPE) and the allylic position equivalents (APE). These new indices have been shown to correlate more accurately with oxidative stability than the IV [18,21,24,2630] and are dened as [25]: APE
n X i1 n X i1

api Aci

BAPE

bpi Aci

where api is the number of allylic positions in a specic fatty acid methyl ester, bpi the number of bis-allylic positions in a specic fatty acid methyl ester, Aci the mass-percent of each fatty acid methyl ester in the mixture and n the total number of fatty acid methyl esters in the mixture. The bis-allylic and allylic sites react with oxygen via the autoxidation mechanism with the classical radical chain reaction steps of initiation, propagation, and termination [28]. The initiation step is the removal of a hydrogen atom from a bis-allylic or allylic site, which results in a radical site that subsequently reacts directly with molecular oxygen to form a peroxy radical. The peroxy radical site then abstracts a hydrogen atom from a nearby C\H bond, thereby producing an additional reactive site that is susceptible to oxygen addition. The most reactive position for the initial radical site formation is the bis-allylic position, whereas the allylic position is much less reactive. Therefore, autoxidation of unsaturated fatty compounds proceeds at different rates depending on the number and position of double bonds [22,23]. This attribute is the reason that small amounts of highly unsaturated fatty acids containing bis-allylic carbons have been observed to have a disproportionately strong effect on oxidative stability [18]. One solution to improve the fuel quality of algal FAME would be to remove the troublesome LC-PUFA to produce a product with a fatty acid composition similar to that of commonly used feedstocks such as soy or rapeseed. This process would improve the oxidative stability and the Cetane Number. Another advantage to removal of LC-PUFA from algae is that many LC-PUFA, such as EPA and DHA, have a very high economic value for cosmetic, pharmaceutical, nutritional, food additive and aquaculture applications [31]. However, removal of the LC-PUFA has a detrimental effect on the cold ow properties since most algal oils are also very high in fully saturated fatty acids such as C16:0. The LC-PUFA such as EPA and DHA are considered pharmacologically important for dietetics and therapeutics. They have been used for treatments of chronic inammation such as rheumatism, skin diseases, and inammation of the gastrointestinal tract. They are also believed to have a positive effect on cardio-circulatory diseases, coronary heart diseases, atherosclerosis, hypertension, cholesterol, and cancer treatment [31,32]. There are already several successful companies that specialize in the production of LC-PUFA from algal lipids for nutritional additives. For example, Martek Biosciences Corporation (Columbia, MD) supplies algal oil rich in DHA as an omega-3 fatty acid additive for milk, cooking oil, and infant formula [33]. Because of the high value of EPA and DHA, combined with the fact that some algae species that accumulate EPA and/or DHA in substantial quantities (i.e. Nannochloropsis) also have high productivity in terms of growth rate and total lipid accumulation, it is expected that conditions will be economically favorable for the removal of some percentage of EPA and DHA from algal lipids prior to conversion of the remaining lipids into biofuels. Such an approach would improve the quality of the fuel feedstock and decrease the consumer costs for the high nutritional products. The U.S. Department of Energy is also suggesting the approach of pursuing high value co-products to offset algal biofuel production costs in the short term [34]. It should be noted, however, that if algal biofuels obtain annual production rates on the order of the EISA Advanced Biofuels standards (10 billion gal/year) then the EPA and DHA may no longer retain a high economic value.

allyic sites

bis-allylic sites
Fig. 1. Schematic diagram showing the location of allylic and bis-allylic sites in the hydrocarbon chain for EPA methyl ester (C20:5).

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

The use of oxidative stability additives have been widely employed during biodiesel production and may be a solution to enhance algal methyl ester biodiesel fuel properties. A recent study by Karavalakis and coworkers [35] determined that the use of synthetic phenolic antioxidant additives at a concentration of 1000 mg/kg improved oxidative stability for each additive tested. The additives tested, listed in order of effectiveness in B100, were propyl gallate (PG), pyrogallol (PA), tert-butyl hydroquinone (TBHQ), butylated hydroxytoluene (BHT), and butylated hydroxyanisole (BHA). The phenolic antioxidants generally improve stability by providing protons that inhibit the formation of free radicals or interrupt the propagation of free radicals through their active hydroxyl group [40]. As discussed below, an oxidative stability additive containing TBHQ was shown to be highly effective in enhancing the oxidative stability of algal methyl ester model compounds with high levels of EPA and DHA. 1.4. Ignition quality of methyl esters The fatty acid composition of the feedstock used for biodiesel production also affects the ignition quality of the fuel as measured by its Cetane Number (CN). The Cetane Number (CN) is a dimensionless number related to the ignition delay period that the fuel experiences upon injection into the cylinder of a diesel engine with higher CN corresponding to shorter ignition delay periods. The CN of a diesel fuel must be within acceptable limits for proper operation of a compression ignition engine. If the CN of the fuel is too high, combustion can occur before the fuel and air are properly mixed resulting in incomplete combustion and increased soot formation; conversely, low CN leads to misring, slower engine warm-up, and incomplete combustion [15]. Moderately high CN (50) are desired for proper operation of a compression ignition engine and also help to ensure good cold start properties [17]. The ASTM D6751 standard for biodiesel requires a minimum CN of 47 and the EN 14214 requires a minimum CN of 51. The carbon chain length and the number of double bonds in a hydrocarbon both have an effect on the CN. Specically, for long chain hydrocarbons such as the methyl esters produced from vegetable oils or animal fats, CN increases with increasing carbon chain length, but decreases with increasing number of double bonds in the hydrocarbon chain. As discussed below, the high level of unsaturation in the EPA and DHA methyl esters results in a decreased CN for algal methyl esters that contain these compounds. 2. Methods and materials The goal of this study was to experimentally determine the oxidative stability, Cetane Number, viscosity, density, bulk modulus, cloud point and cold lter plugging point of algae-based FAME compounds subjected to varying levels of removal of EPA and DHA. In this section the experimental techniques and test protocols are described. 2.1. Oxidative stability measurements The oxidative stability tests were conducted using a Metrohm 743 Rancimat instrument. The 743 Rancimat is an approved instrument
Table 2 Oxidative stability and ignition quality test parameters [37,38] . Instrument Metrohm 743 Rancimat Method EN 14112 Standard D6751 EN 14214 Standard

for measurement the oxidative stability induction period of 100% biodiesel (B100) as specied by the ASTM D6751 and EN 14214 Standards [36,37]. Both standards refer to the EN 14112 test method for determination of the oxidative stability induction period. Metrohm has developed the 743 Rancimat system as an automatic variant to the complex Active Oxygen Method for determining the induction period of fats and oils. The Rancimat test parameters and specications are listed in Table 2. The Rancimat system accelerates the autoxidation process that normally occurs slowly at ambient temperatures to study the oxidative decay of vegetable oil and animal fat. The Rancimat method allows the sample to be exposed to a variable air ow at a constant temperature ranging between 50 and 220 C, but tests were conducted herein at 10 L/h and 110 C as per the EN 14112 method. The Rancimat internal pump supplies the air ow through uorinated ethylene propylene (FEP) tubing by rst passing the ambient air through absorbent molecular sieves. A glass reaction vessel, which holds a 3 gram fuel sample, sits in a heating block and the air is bubbled through the sample via a glass tube attached to the FEP tubing. The volatile compounds that are released from the liquid sample then ow up through a silicone tube to the measuring vessel. The air is bubbled through a distilled water solution via a polytetrauoroethylene (PTFE) tube attached to silicone tubing. An electrode in the measuring vessel cell measures the electrical conductivity of the solution, which is recorded as a function of time by the Rancimat data acquisition system. When the airow carries highly volatile secondary oxidation products (mainly formic acid) to the measuring vessel where they are absorbed in the distilled water, the conductivity of the solution rapidly increases. The time period from the beginning of the test to the rapid increase in thermal conductivity of the solution is dened as the oxidative stability induction period. The Rancimat computer program uses an algorithm to determine the oxidative stability induction period automatically from the measured Rancimat curve, which is the measured electrical conductivity in microSiemens per centimeter of the solution as a function of time in hours. The 743 Rancimat was calibrated for conductivity and temperature. The conductivity was calibrated using 0.01 M potassium chloride (KCl) conductivity solution from Sigma Aldrich (St. Louis, MO). The KCl was diluted to a 0.001 M solution resulting in 147 S/cm conductivity at 25 C. The temperature was calibrated using silicone oil from Sigma Aldrich (St. Louis, MO) resulting in a 0.7 C and 0.8 C temperature offset for heating block A and heating block B, respectively. 2.2. Ignition quality measurements Ignition quality tests were conducted using a Waukesha Fuels Ignition Tester (FIT), which measures the Derived Cetane Number (DCN) in accordance with the ASTM D7170 method [38]. For biodiesel, the ASTM D6751 Standard requires a minimum Cetane Number of 47. Waukesha developed the FIT for DCN measurements as a substitute to the conventional CFR F5 Cetane Method Diesel Rating Unit for Cetane Number because the FIT method requires much less fuel and a smaller testing footprint. The FIT method of determining a DCN uses a constant volume combustion chamber. The constant volume combustion chamber is

Specication 3 h minimum 6 h minimum Specication

Test parameters 10 L/h air ow 110 C 3 g sample

Instrument

Method

Test parameters # of injections Injection period 5.00 0.25 ms Fuel temperature 35 2 C

Waukesha FIT

D7170

D6751

47 minimum

25

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

heated to a specied wall temperature and pressurized to 2.40 0.02 MPa. The FIT determines the DCN of a given fuel sample by measuring the average ignition delay period of a series of 25 fuel injections. Since the Cetane Number describes the propensity of a fuel to autoignite when exposed to air at elevated temperature and pressure, the ignition delay period is measured and recorded by the FIT, which is then used to calculate the DCN according to the empirical relationship: DCN 171 ID

2.4. Density and speed of sound measurements The density and speed of sound measurements were conducted at 40.00 C using an Anton Paar DSA 5000M. The DSA 5000M uses an oscillating U-tube method to simultaneously determine two physically independent properties from one sample injection. The sample is injected manually into a U-shaped borosilicate glass tube that is electronically excited at its characteristic frequency. This characteristic frequency changes depending on the density of the sample injected. Through determination of the characteristic frequency and a mathematical formula, the density of the sample is calculated based on the equation = KA Q2 f1 - KB f2, where KA and KB are instrument constants, Q the quotient of the period of oscillation of the U-tube divided by the period of oscillation of the reference oscillator, and f1 and f2 correction terms for temperature, viscosity, and nonlinearity. The sound velocity is measured with an ultrasonic transmitter located on one side of the measuring cell and a receiver on the other side. The transmitter sends sound waves of a known period through the sample and the sound velocity is determined by calculating the period of the received sound waves according to the equation: a L 1 1:6e5T   Ps 512 f 3

where ID is the average measured ignition delay period in milliseconds. To control the liquid fuel temperature, the FIT uses a circulating bath with 5050 antifreeze and coolant mixture of ethylene glycol and water adjustable between 20 and 100 C to one tenth of a degree using an integrated PID controller and circulating pump. The high pressure combustion air is delivered from bottled high purity air containing 20.9+/1.0% oxygen, less than 0.0003% hydrocarbons, and less than 0.025% water. The combustion air is provided to the FIT via a two stage regulator adjusted to 415 psi and a high pressure line with ttings. The low pressure air, which is used for the injection actuator, is provided from 120 psi shop air. The FIT was calibrated by adjusting the combustion chamber wall temperature set point using n-heptane and methylcyclohexane (MCH) from Sigma Aldrich (St. Louis, MO). The wall temperature was adjusted until all the ignition parameters were met and the ignition delay period for three consecutive n-heptane tests was 3.15 0.02 ms. The FIT was then deemed calibrated when the measured ignition delay for MCH was 10.1 0.5 ms for two consecutive tests. The FIT was checked for quality assurance before each operating period with n-heptane. Table 2 contains the test parameters and associated standards for the ignition quality tests. 2.3. Viscosity measurements Dynamic and kinematic viscosity measurements were conducted using an Anton Paar SVM 3000 Stabinger Viscometer instrument. The measurements followed the ASTM D7042 Standard Test Method [39], which requires three consecutive measurements of the sample at 40.00 C to determine the validity of the dynamic viscosity and density measurements. The SVM 3000 measures the dynamic viscosity and density simultaneously by using a rotational coaxial cylinder measuring system and a U-shaped oscillating sample measuring tube. Both measuring cells are held at the same temperature by copper blocks surrounding the cells. A thermoelectric heating and cooling system ensures that the copper blocks are maintained at temperatures within 0.005 C. Since dynamic viscosity describes the resistance of a uid to ow or deform under external shear forces, the dynamic viscosity is measured with the rotational coaxial system. A motor drives the outer cylinder at a known rotational speed and the inner cylinder is held in position by the sample and a magnetic iron ring. The inner lowdensity cylinder rotational speed is measured with a Hall Effect sensor by counting the frequency of the rotating magnetic eld when equilibrium of the driving torque, viscous forces, and eddy current torque are reached. The density is measured with the U-shaped oscillating sample tube by exciting the tube at its characteristic frequency. This characteristic frequency changes depending on the density of the sample injected and the density is determined from a digital analyzer. The kinematic viscosity is calculated from the equation = /, where is the kinematic viscosity, the dynamic viscosity, and the density. Note that the density measurements reported in this study were made using the Anton Paar DSA5000M (see Section 2.4) since the DSA 5000M measures density to six decimal places whereas the SVM 3000 only measures density to four decimal places.

where L is the path length of the sound waves, T the temperature deviation to 20 C, Ps the oscillation period of the received sound waves, the instrument constant, and f3 a correction term for temperature. The DSA 5000M was calibrated using pure water at 20 C. The oscillation period is measured by optical pickups and two integrated Pt 100 platinum thermometers with Peltier elements provide the precise thermostating of the sample. Since gas bubbles greatly affect the precision measurement, an automatic lling check detection warns of gas bubbles in the measuring cell and a built in camera allows the user to visually inspect the U-tube for gas bubbles on the integrated display and controller screen. 2.5. Cloud point and cold lter plugging point measurements A Lawler DR4-14L automated cold lter plugging point (CFPP) and cloud point (CP) analyzer was used to measure the CFPP and CP of the formulated algal fuels. The Lawler DR4-14L conforms to ASTM standards D6371 and D2500 for measuring the CFPP and CP, respectively [40,41]. The results of these two standards are also required to be reported per the US and European biodiesel standards. For any given fuel the CFPP and CP tests can be run simultaneously. Each sample container consists of a glass cylinder encased in a test jacket which is maintained at a constant temperature by a hermetically sealed refrigeration system. This arrangement allows the sample to achieve temperatures as low as 70 C. The CFPP is measured by drawing a 20 mL sample through a 45 m lter via a vacuum system over the course of 60 s. If the sample can be drawn and returned within the specied time, then the sample temperature is lowered by 1 C. This process is repeated until the test fails due to solidication of the sample at which point the CFPP is reported. To test for CP, the sample is cooled at a given rate and tested for cloudiness at each nominal temperature by a ber optic probe immersed within the sample. The probe measures a difference in light transmission between each reading and, once the light transmission attenuation is above a given value, the sample is determined to have reached its CP. It is important to note that, since acceptable cold temperature properties depend on the local climate, specic limits for CFPP or CP are not specied by either the US or EU standard. Rather, these standards only require that one of these values is reported. However,

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

some European countries do require biodiesel (and petroleum diesel) to meet a minimum CFPP, which represents a stricter standard than that specied in the EU 14214 standard. 2.6. Fuel formulation The algal methyl esters tested in this study were obtained and produced from a variety of sources. Unrened algal crude oil was provided by Eldorado Biofuels (Santa Fe, NM), Solix BioSystems (Fort Collins, CO) and Inventure Chemical (Tuscaloosa, AL). The Eldorado algal oil consisted primarily of triglycerides and was readily transesteried in-house into a high quality methyl ester fuel. The Eldorado sample did not contain any EPA or DHA as would be found in algal lipids produced from the genus Nannochloropsis and therefore the El Dorado methyl ester had moderate BAPE value of 51 from the bisallylic sites present in 18:2 and 18:3. The unrened Inventure algal crude oil was produced by a single step methanol extraction and transesterication process, which resulted in a product that was only 43% FAME. For the tests presented herein, the Inventure FAME was further rened via a distillation process at Pacic Northwest National Laboratories. The Inventure product was very high in DHA, resulting in a calculated BAPE of 254. The Solix BioSystems crude oil was extracted from Nannochloropsis salina and contained 11% EPA. The Solix crude oil had a high free fatty acid content and therefore required a glycerolysis step prior to transesterication, which successfully reduced the free fatty acid content from 11% to less than 2%. Despite the presence of 11% EPA, the resulting methyl ester from Solix also had a moderate BAPE value of 56 since it had very little 18:2 or 18:3. The fatty acid proles of the three algal methyl ester samples are shown in Table 3. The fatty acid proles were obtained via Gas Chromatography using an Agilent 7890 GC (Agilent Technologies, USA) with a ame ionization detector and a Restek (Restek Corporation, USA) FAMEWAX column. The real algal lipids and algal FAME samples were in limited supply and only the Solix BioSystems sample represented an algal monoculture of known origin (N. salina). Therefore, another methyl ester source was necessary to determine the effect of EPA and DHA removal on algal methyl ester fuel properties. Accordingly, using a technique similar to that employed in a previous study by the authors [42], model algal methyl ester compounds were formulated using mixtures of transesteried sh oil and various vegetable oils. Using this technique, it was possible to reproduce the fatty acid proles of methyl esters produced from three different algal triglycerides subjected to varying amounts of removal of the LC-PUFAs EPA and DHA. Pharmaceutical grade sh oil high in EPA and DHA was purchased from Jedwards International Inc. (Quincy, MA) and vegetable oils were purchased locally. The vegetable oils used for formulation were Great Value pure corn oil, Great Value vegetable oil (100% soybean), and Wesson pure canola oil. A sample of methyl laurate (C12:0; CE1295) from Procter & Gamble's chemical division (Cincinnati, OH)
Table 3 Measured fatty acid proles and calculated BAPE values of algal methyl ester samples. Inventure distilled FAME 10:0 12:0 14:0 16:0 18:0 16:1 18:1 18:2 18:3 20:4 20:5 22:6 BAPE 2 5 10 14 2 1 14 1 Solix Nannochloropsis salina Eldorado biofuels

was also used as a formulation ingredient. The fatty acid prole of each ingredient oil is listed in Table 4. The sh and vegetable oils were transesteried on the bench-scale in 500 mL batches using potassium hydroxide (KOH) pellets from Fisher Scientic (Fairlawn, NJ) and methanol from Sigma-Aldrich USA (St. Louis, MO). The Eldorado and Solix algal lipid samples were transesteried in a similar fashion. All of the fuels used were stored in ambercolored glass bottles at 14 C immediately after transesterication. Model algal methyl ester compounds were formulated to obtain fatty acid proles similar to the storage lipids produced by three different algae species that produce lipids that are high in LC-PUFA. The fatty acid composition from these algae species were obtained from the literature. The rst alga chosen was Nannochloropsis oculata as proled during a logarithmic growth stage [16] because it is high in LC-PUFAs containing over 21 and 3% EPA and DHA, respectively. The second alga chosen was Nannochloropsis sp. because it was cited to contain 26.5% EPA [43]. The last alga modeled was Isochrysis galbana [10] because it contains 14% DHA and 5% EPA. The full fatty acid proles for each of these algae species as provided in [10,16,43] are shown in Table 5. To accurately model methyl esters produced from lipids from these algae species, the BAPE and APE of the model compounds were matched to those values that would be produced from the fatty acid compositions of the algal lipids in Table 5. The BAPE and APE were matched since the bis-allylic carbon site is the point of initial oxidation as previously discussed. Even though the allylic site is not known as a strong participator in initial oxidation reactions, both BAPE, APE, and BAPE + APE were matched to yield a more complete and accurate representation of the total double bond count and structure of the fatty acids. To examine the effect on algal fuel properties subjected to removal of varying levels of LC-PUFA, the removal of EPA and DHA were simulated in amounts of roughly 0, 25, 50, 75, and 100%. The BAPE and APE of the three algal methyl esters subjected to varying levels of LC-PUFA removal were calculated for each removal percentage and then model algal methyl ester compounds were formulated by matching the BAPE, APE, and BAPE + APE with the appropriate amounts of sh, soy, canola, corn, and CE1295 methyl esters. The overall formulation mixtures that were tested are shown in Table 6.

Table 4 Fatty acid proles of vegetable and sh oils used to produce model algal methyl ester compounds. Soy 8:0 10:0 12:0 14:0 15:0 16:0 17:0 18:0 20:0 22:0 24:0 14:1 16:1 17:1 18:1 20:1 22:1 24:1 18:2 20:2 18:3 20:4 20:5 22:5 22:6 Canola Corn CE1295 0.1 0.3 98.2 1.1 0.1 Fish oil [42]

0.2 10.6 4.1 4.2 1.9 0.7 12.6 2.2

6.5 0.5 15.1 2.0 3.2 0.8

3 30 1 39 8 1 1 3 11 56

24.2

7 2 47 36 7

60.7 1.5

31.1 0.1

52.5 7.9

19.3 9.9

51.0 1.0

0.6 9.2 0.5 13.1 2.2 1.1 0.6 1.1 1.2 1.1 17.8 2.1 11.5

51 254

51

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx Table 5 Fatty acid proles of the three species considered for the production of model algal methyl ester compounds [10,16,43].
BAPE - Actual APE - Actual BAPE + APE - Actual BAPE - Formulation APE - Formulation BAPE + APE - Formulation

300 250 200 150 100 50

12:0 14:0 15:0 16:0 18:0 20:0 14:1 16:1 17:1 18:1 20:1 22:1 18:2 20:2 22:2 18:3 20:3 18:4 20:4 20:5 22:6 Unknown Total

2.4 0.7 14.5 1.8 2.0 15.7 4.1 10.1 0.9 0.7 3.6 1.6 2.5 0.5 6.0 21.5 3.2 8.2 100

2.8 3.2 9.2 2.1 3.2 19.8 3.6

23.1 1.1 14.0 1.1

3.0 14.0

BAPE/APE/BAPE+APE

Nannochloropsis oculata [16]

Nannochloropsis sp. [43]

Isochrysis galbana [10]

0
1.1 6.6 9.1 2.5 5.0 1.0 7.0 10.0 26.5 10.4 100 5.0 14.0 1.7 100

20

40

60

80

100

Percent Removal of EPA and DHA

Fig. 2. Calculated BAPE, APE and BAPE + APE for Nannochloropsis oculata methyl ester subjected to varying levels of EPA and DHA removal compared with those values calculated for the model methyl ester formulations employed herein.

With the given set of ingredient oils, a zero percent removal formulation was not accurately attained for Nannochloropsis sp. or I. galbana. Comparisons between the calculated BAPE, APE and BAPE + APE values for the actual algal methyl esters and model methyl ester compounds subjected to varying levels of removal of EPA and DHA removal are shown in Figs. 2 through 4 for N. oculata, Nannochloropsis sp., and I. galbana, respectively. These gures show that the fuel formulations accurately describe the algae species complete double bond structure.

to two trials. Twenty-four different samples were tested without additives as shown in Table 7. The methyl laurate sh methyl ester blends were rst tested to validate the BAPE theory of oxidation. Next, the model algal methyl ester formulations and ingredient FAMEs were tested, followed by the actual algae methyl esters. Fig. 5 is a plot of measured oxidative stability induction period as a function of BAPE for all of the fuels tested. The gure clearly shows that oxidative stability correlates strongly with BAPE. Specically, the oxidative stability induction period was shown to vary exponentially with BAPE and an exponential curve was tted to the data with an R-squared value of 0.91. Based on the data, the following empirical correlation is recommended: 19:238e
0:023BAPE

3. Results and discussion 3.1. Oxidative stability of model algal methyl ester compounds For the oxidative stability tests, the Metrohm 743 Rancimat was operated following the ASTM D6751 B100 Standard Specication which calls for the EN 14112 test method. The EN 14112 requirements specify a three gram sample heated at 110 C subjected to an air ow rate of 10 L/h. Each three gram sample was weighed to the nearest hundredth of a gram. The automatic induction period determination was used for all samples and each sample was subjected
Table 6 Model algal methyl ester formulations. Formulation % EPA + DHA Ingredient oil percentage removed Fish Soy Canola Corn CE oil 1295 100% 75% 50% 25% 0% 100% 75% 50% 25% 100% 75% 50% 25% 5 25 45 70 75 10 40 60 75 35 50 70 80 25 15 15 10 0 10 0 5 0 0 15 5 5 35 20 5 5 25 50 45 30 25 30 10 5 5 5 20 25 15 0 0 0 0 0 0 0 5 0 30 20 10 0 0 30 15 5 0 35 25 15 10

where is the oxidative stability induction period in hours and BAPE is the bis-allylic position equivalents for the methyl ester. According to this empirical correlation, algal methyl ester fuels with BAPE values of approximately of 80 and 50 would be sufcient to pass the ASTM and EN oxidative stability specications, respectively. As discussed further below, the results for each of the three real algal methyl formulations also fell along the same exponential oxidative stability vs. BAPE curve.
BAPE - Actual APE - Actual BAPE + APE - Actual BAPE - Formulation APE - Formulation BAPE + APE - Formulation

300 250 200 150 100 50 0

Total 100 100 100 100 100 100 100 100 100 100 100 100 100

Nannochloropsis oculata

Nannochloropsis sp.

BAPE/APE/BAPE+APE

20

40

60

80

100

Isochrysis galbana

Percent Removal of EPA and DHA

Fig. 3. Calculated BAPE, APE and BAPE + APE for Nannochloropsis sp. methyl ester subjected to varying levels of EPA and DHA removal compared with those values calculated for the model methyl ester formulations employed herein.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

BAPE - Actual APE - Actual BAPE + APE - Actual BAPE - Formulation APE - Formulation BAPE + APE - Formulation

280 240 200 160 120 80 40

25

BAPE/APE/BAPE+APE

20

Induction Period (hr)

15

Methyl Laurate-Fish Methyl Ester Blends Nannochloropsis sp. Formulations Nannochloropsis oculata Formulations Isochrysis galbana Formulations Soy Methyl Ester Canola Methyl Ester Corn Methyl Ester Eldorado Algal Methyl Ester Solix Algal Methyl Ester Inventure Algal Methyl Ester 3 Hour ASTM Limit 6 Hour EN Limit

10

20

40

60

80

100

120

0 0 50 100 150 200 250

Percent Removal of EPA and DHA

Fig. 4. Calculated BAPE, APE and BAPE + APE for Isochrysis galbana methyl ester subjected to varying levels of EPA and DHA removal compared with those values calculated for the model methyl ester formulations employed herein.

BAPE
Fig. 5. Oxidative stability induction period as a function of BAPE for all methyl esters.

To more clearly show the effect of EPA/DHA removal from the three model methyl ester species considered in this study, Figs. 6 through 8 contain plots of oxidative stability induction period plotted as a function of the percentage of EPA/DHA removal for N. oculata, Nannochloropsis sp. and I. galbana, respectively. The N. oculata formulation oxidative stability results are shown in Fig. 6. The results suggest that for this species, the removal of approximately 50% of the EPA/DHA (BAPE 90) would be sufcient for passing the ASTM three hour induction period limit, whereas approximately 80% removal of EPA/DHA (BAPE 65) would be required to pass the EN standard. The Nannochloropsis sp. formulation oxidative stability results are shown in Fig. 7. For these formulations, only the maximum removal rate of 100% EPA/DHA passed the ASTM three hour induction period limit and EN six hour induction period limit. The results suggest
Table 7 Test matrix for oxidative stability, ignition quality, viscosity, density, speed of sound, cloud point and cold lter plugging point tests. Note that ignition quality tests were not performed on the real algal methyl ester samples because of limited fuel quantities. Methyl esters Fish Soy Canola Corn Methyl laurate sh methyl ester blends 0% Fish 25% Fish 50% Fish 75% Fish Algal methyl ester samples Inventure Solix Eldorado Model algal methyl ester formulations Nannochloropsis oculata EPA + DHA removal 15.6% 19.0% 52.0% 77.8% 98.8% 28.7% 44.3% 71.0% 100% 8.7% 31.0% 68.1% 99.4%

that for this species, the removal of approximately 75% of the EPA/ DHA (BAPE 80) would be sufcient for passing the ASTM three hour induction period limit, whereas approximately 90% removal of EPA/DHA (BAPE 60) would be required to pass the EN standard. The oxidative stability results for the I. galbana formulations are shown in Fig. 8. For these formulations, none of the simulated EPA/ DHA removal percentages passed either the ASTM or EN specication. Referring back to Fig. 4, the minimum BAPE obtained via removal of 100% EPA/DHA from I. galbana would be greater than 60, which is insufcient to meet ASTM oxidative stability standards.

3.2. Oxidative stability of real algal methyl esters The oxidative stability results for the three real algal methyl esters are shown in Table 8. The Eldorado algal methyl ester, which had a BAPE value of 50.6 and contained no EPA or DHA, passed both the ASTM and EN oxidative stability standards with an induction period of 8.43+/0.77 h. The Solix algal methyl ester, which had a BAPE value of 50.6 but contained 11% EPA, had an induction period of 4.13+/0.85 h, which passes only the methyl ester ASTM limit. The distilled Inventure algal FAME performed very poorly with regard to oxidative stability with an induction period of less than 5 min. This result was expected due to the fact that the Inventure FAME was 50% DHA leading to the highest BAPE of all samples tested at 254. It should be noted that the distilled Inventure algal FAME is not intended for use as nished fuel product, but rather as a feedstock for renewable diesel production via hydroprocessing.
14 12
Nannochloropsis oculata Formulations 3 Hour ASTM Limit 6 Hour EN Limit
Curve Fit: y=1.0373exp(0.0232x) R =0.9343
2

Induction Period (hr)

10 8 6 4 2 0 0 20

Nannochloropsis sp.

40

60

80

100

Isochrysis galbana

Percent Removal of EPA and DHA

Fig. 6. Oxidative stability induction period as a function of the percent EPA and DHA removed for the Nannochloropsis oculata model compound formulations.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

14 12
Nannochloropsis sp. Formulations 3 Hour ASTM Limit 6 Hour EN Limit
Curve Fit: y=0.5098exp(0.0282x) R =0.9328
2

Table 8 Oxidative stability test results for three algal methyl esters. Sample Inventure distilled algal methyl ester Solix algal methyl ester Eldorado algal methyl ester Test 1 2 1 2 1 2 Induction period (h) 0.06 0.04 3.70 4.55 8.97 7.88 Average (h) 0.05 4.13 8.43 Standard deviation 0.01 0.85 0.77 BAPE 254.3 56.0 50.6

Induction Period (hr)

10 8 6 4 2 0 20

40

60

80

100

Table 9 Test matrix for Nannochloropsis sp. methyl ester model compounds with Vitablend Bioprotect 350 (30% TBHQ) oxidative stability additive. Formulation EPA + DHA removal 28.7% 44.3% 71.0% 100% Additive amount tested 0.10% 0.10% 0.10% 0.10% 0.15% 0.15% 0.15% 0.15% 0.20% 0.20% 0.20% 0.20% 0.33% 0.33% 0.33% 0.33%

Percent Removal of EPA and DHA

Fig. 7. Oxidative stability induction period as a function of the percent EPA and DHA removed for the Nannochloropsis sp. model compound formulations. Nannochloropsis sp.

3.3. The effect of oxidative stability additives To determine the effectiveness of oxidative stability additives on the oxidative stability induction period of algal methyl esters subjected to varying levels of EPA/DHA removal, the same Metrohm 743 Rancimat test protocol was repeated for the Nannochloropsis sp. compounds with varying levels of oxidative stability additive. The additive used for testing was Bioprotect 350 from Vitablend (Wolvega, Netherlands), which is a formulation containing 30% tert-butylhydroquinone (TBHQ). The Nannochloropsis sp. formulations were chosen for testing the additive effects because they yielded both passing and failing oxidative stability results. Table 9 contains the oxidative stability/additive test matrix wherein the additive percentage is tabulated on a mass basis. As indicated in the table, tests were conducted with 0.10 to 0.33% Bioprotect 350, which corresponds to 0.03 to 0.1% TBHQ, respectively. Figs. 9 and 10 contain plots of oxidative stability induction period as a function of BAPE and percent EPA/DHA removal, respectively, for the model Nannochloropsis sp. methyl ester formulations, with the addition of 0 to 0.33% Bioprotect 350 additive. The gures clearly show that for Nannochloropsis sp. any negative effect of EPA and DHA can be offset with the TBHQ fuel additive. Without the addition of an oxidative stability additive, the Nannochloropsis sp. methyl ester formulations passed the ASTM and EN oxidative stability standards only when 100% EPA/DHA were removed. The addition of only 0.1% additive (0.03% TBHQ) resulted in all of the formulations passing the ASTM 3 hour oxidative stability standard. The addition of 0.2% additive (0.06% TBHQ) was sufcient for all of the formulations to pass
3.2 2.8

the EN 6 hour oxidative stability standard. Note that it was not possible to formulate a model Nannochloropsis sp. methyl ester with 100% EPA/DHA removal, but based on the oxidative stability induction period vs. BAPE results of Fig. 9, it is expected that 0.33% additive would be sufcient to pass the EN 6 hour oxidative stability standard. This level of additive would not be cost prohibitive as an addition of 0.33% Bioprotect 350 to Nannochloropsis sp. methyl ester would cost approximately $0.08/gal of fuel. 3.4. Ignition quality test results Ignition quality tests were conducted on all of the model algal methyl ester compounds listed in Table 7. Note that ignition quality tests were not performed on the real algal methyl ester samples because of limited fuel quantities. Ignition quality was characterized via the ASTM D7170 Derived Cetane Number (DCN) test method [38] using the FIT instrument as described above. Each model algal methyl ester sample was tested twice. The DCN results for the model algal methyl ester compounds are plotted in Figs. 11 and 12 as a function of BAPE and percent EPA/ DHA removal, respectively. As expected, the results show that, as the level of unsaturation increases (increasing BAPE) in the algae methyl ester formulations, the DCN decreases. Indeed, Fig. 11 suggests that there is a linear relationship between DCN and BAPE for the model algal methyl ester compounds tested in this study.
30

Induction Period (hr)

2.4 2.0 1.6 1.2 0.8

0 20

Curve Fit: y = 1.234e0.0069x R = 0.976

Induction Period (hr)

Isochrysis galbana Formulations 3 Hour ASTM Limit

25 20 15 10 5 0 40

No Additive 0.1% Additive = 0.03% TBHQ 0.15% Additive = 0.045% TBHQ 0.2% Additive = 0.06% TBHQ 0.33% Additive = 0.1% TBHQ 3 Hour ASTM Limit 6 Hour EN Limit

40

60

80

100

60

80

100

120

Percent Removal of EPA and DHA

Fig. 8. Oxidative stability induction period as a function of the percent EPA and DHA removed for the Isochrysis galbana model compound formulations.

BAPE
Fig. 9. Oxidative stability induction period as a function of BAPE for Nannochloropsis sp. formulations with TBHQ additive.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

10

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

25

Induction Time (hr)

20

No Additive 0.1% Additive = 0.03% TBHQ 0.15% Additive = 0.045% TBHQ 0.2% Additive = 0.06% TBHQ 0.33% Additive = 0.1% TBHQ 3 Hour ASTM Limit 6 Hour EN Limit

Table 10 Density, viscosity, speed of sound and bulk modulus results for model algal methyl ester compounds and real algal methyl esters. Sample name BAPE Viscosity Dynamic (mPa s) Nannochloropsis oculata formulations 41 65 90 117 118 41 75 101 118 62 86 108 120 254 56 51 3.943 3.141 3.475 3.896 4.522 2.854 3.227 3.882 4.505 2.754 3.034 3.362 3.271 4.023 3.794 3.983 Kinematic (mm2/s) 4.528 3.618 3.984 4.443 5.140 3.304 3.714 4.438 5.123 3.174 3.482 3.840 3.737 4.540 4.356 4.594 Density (g/cm3) Speed of sound (m/s) 1325.3 1326.9 1339.3 1351.8 1356.4 1315.1 1331.7 1346.2 1356.0 1314.2 1325.8 1337.9 1340.8 1352.2 1349.9 1344.0 Bulk modulus (MPa) 1531.4 1530.4 1566.8 1603.2 1621.0 1496.6 1541.9 1586.5 1618.4 1500.9 1532.9 1568.7 1575.8 1620.3 1587.2 1568.4

15

10

5
Nannochloropsis sp. formulations

0 20 40 60 80 100
Isochrysis galbana formulations

Percent Removal of EPA and DHA

Fig. 10. Oxidative stability induction Period as a function of percent EPA and DHA removed for Nannochloropsis sp. formulations with TBHQ additive.

50 48
Nannochloropsis sp. Nannochloropsis oculata Isochrysis galbana

Derived Cetane Number

Inventure algal methyl ester Solix algal methyl ester Eldorado algal methyl ester

0.8719 0.8692 0.8736 0.8773 0.8811 0.8653 0.8695 0.8755 0.8803 0.8691 0.8721 0.8764 0.8765 0.8861 0.871 0.8682

46 44 42 40 38 36 34 20 40 60 80

double bonds present in the EPA and DHA have a stronger effect on CN than chain length. 3.5. Density, viscosity and speed of sound test results The density, dynamic viscosity, kinematic viscosity and speed of sound were measured for the model algal methyl ester formulations, ingredient methyl esters along with the Eldorado, Inventure and Solix methyl esters using the Anton Paar SVM 3000 and DSA 5000M. The results are tabulated in Table 10. The kinematic viscosity, density and speed of sound are plotted as a function of BAPE in Figs. 13 through 15. All of the algal methyl ester formulations and real algal methyl esters met the ASTM B100 Standard for kinematic viscosity of 1.9 b b 6.0 mm 2/s where is the kinematic viscosity. As shown in Fig. 13, the kinematic viscosity does not correlate strongly with BAPE. Fig. 14, however, shows that the density of the model algal methyl ester compounds varied linearly with BAPE. As shown in Fig. 15, the speed of sound of the algal methyl ester compounds also varied linearly with BAPE. Fig. 16 contains a plot of bulk modulus for the model algal methyl ester formulations. The bulk modulus is a measure of the resistance to compressibility of the liquid fuel, which has been shown to affect NOx emissions in biodiesel. Specically, studies have shown that the
5.5

100

120

140

BAPE
Fig. 11. Derived Cetane Number as a function of BAPE for model algal methyl ester compounds.

The minimum Cetane Number required to meet the B100 ASTM D6751 Standard is 47. None of the I. galbana formulations met this standard and only when all of the EPA and DHA are removed from the Nannochloropsis sp. and N. oculata formulations was this minimum Cetane Number value obtained. It is well established that for long chain hydrocarbons Cetane Number increases with increasing hydrocarbon chain length, but decreases with increasing number of double bonds [44]. The results presented herein indicate that the

50

Kinemataic Viscosity (mm2/s)

48

Derived Cetane Number

46 44 42 40 38 36 34 0

Nannochloropsis sp. Nannochloropsis oculata Isochrysis galbana

5.0 4.5 4.0 3.5 3.0 2.5 2.0

Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

20

40

60

80

100

120

20

40

60

80

100

120

140

Percent Removal of EPA and DHA

Fig. 12. Derived Cetane Number as a function of percent EPA/DHA removal for model algal methyl ester compounds.

BAPE
Fig. 13. Kinematic viscosity of model algal methyl ester compounds subjected to varying levels of EPA and DHA removal.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

11

0.885

0.880

Percent Saturated Methyl Esters

48 44 40 36 32 28 24 20 0

Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana Nannochloropsis salina

Density (g/cm3)

0.875

0.870
Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

0.865

0.860 40 60 80 100 120

BAPE
Fig. 14. Density of model algal methyl ester compounds subjected to varying levels of EPA and DHA removal. Dark line is a linear regression and dashed lines represent 99% condence interval.

20

40

60

80

100

Percent Removal of EPA and DHA

Fig. 17. Calculated percentage of fully saturated methyl esters as a function of percentage EPA/DHA removal from Nannochloropsis oculata, Nannochloropsis sp., Isochrysis galbana and Nannochloropsis salina methyl esters.

1360
Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

Speed of Sound (m/s)

1350

1340

in increased NOx emissions in comparison with petroleum diesel. However, a recent study on engine emissions from high BAPE model algal methyl ester compounds by the authors [42] indicated that, while these fuels did exhibit advanced injection timing, the model algal methyl esters resulted in decreased NOx emissions in comparison to petroleum diesel. 3.6. Cold ow properties of model algal methyl ester formulations As indicated in Tables 1 and 5, the fatty acid proles for algal methyl esters considered herein are not only high in LC-PUFA, but are also high in fully saturated fatty acids such as C16:0 and C18:0. Accordingly, as EPA and DHA are removed from these algal methyl esters to enhance oxidative stability and ignition quality, the cold ow properties worsen. Fig. 17 illustrates how the percentage of fully saturated methyl esters increases as EPA and DHA are removed from N. oculata, Nannochloropsis sp., I. galbana and N. salina methyl esters. As shown in the Figure, the algal methyl esters are initially high in fully saturated methyl ester content (20 to 40% saturated) and the percentage of saturated methyl esters increases to levels of 27 to 50%, depending on the species. By way of comparison, soy methyl ester and canola methyl ester contain only 11% and 4% fully saturated methyl esters, respectively. The percentage of fully saturated methyl esters in the algal methyl ester species subjected to removal of 100% EPA/DHA as considered herein are much closer to that of beef tallow, which typically contains approximately 50% saturated methyl esters
0

1330

1320

1310 40 60 80 100 120

BAPE
Fig. 15. Speed of sound of model algal methyl ester compounds subjected to varying levels of EPA and DHA removal.

higher bulk modulus of biodiesel in comparison to petroleum diesel results in advanced injection timing, which can lead to increased NOx formation in the cylinder [45]. The bulk modulus can be calculated directly from the measured speed of sound and density data from the relationship a2 where is the bulk modulus, a the speed of sound and the density. Fig. 16 shows that the bulk modulus also increases linearly with BAPE for the model algal methyl ester formulations. This result might suggest that algal methyl esters would result
1640

Bulk Modulus (MPa)

1600 1580 1560 1540 1520 1500 1480 40

Cold Filter Plugging Point [oC]

1620

Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

-4

-8
Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

-12

60

80

100

120

-16 20

22

24

26

28

30

BAPE
Fig. 16. Bulk Modulus of model algal methyl ester compounds subjected to varying levels of EPA and DHA removal.

Percent Combined C16:0 and C18:0

Fig. 18. Cold lter plugging point as a function of percentage of C16:0 and C18:0 combined for the model algal methyl ester compounds.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

12

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx

-4

-8
Nannochloropsis oculata Nannochloropsis sp. Isochrysis galbana

-12

-16 20 22 24 26 28 30

Percent Combined C16:0 and C18:0

Fig. 19. Cloud point as a function of percentage of C16:0 and C18:0 combined for the model algal methyl ester compounds.

(predominately C16:0 and C18:0) and is known to have very poor cold ow properties [14]. As expected, for the model algal methyl ester compounds considered herein, the cold lter plugging point (CFPP) and cloud point (CP), were found to vary strongly with the percentage of EPA/DHA removal. Figs. 18 and 19 are plots of cold lter plugging point (CFPP) and cloud point (CP) as a function of percentage of C16:0 and C18:0 combined for the model algal methyl ester compounds.

specication, 0.06% of TBHQ would need to be added to Nannochloropsis sp. methyl ester if no EPA or DHA was to be removed. In addition to the model algal methyl ester compounds, three real algal methyl esters were also tested to determine their oxidative stability, ignition quality, viscosity, density, speed of sound and cold ow properties. All three of the real algal methyl esters proved to perform consistently with the model algal methyl ester formulation results. The Eldorado algal methyl ester, which contained no EPA or DHA, passed both the ASTM and EN Standards for oxidative stability without the addition of any fuel additives. Conversely, the Inventure sample, which consisted of nearly 50% DHA exhibited very poor oxidative stability. The Solix BioSystems N. salina methyl ester had an oxidative stability induction period that passed the ASTM standard, but did not pass the EN standard. The cloud point and cold lter plugging point were also found to be sensitive to the level of EPA and DHA present in the algal methyl esters since the balance of the fatty acid prole for algal methyl esters consist of a very high percentage of fully saturated methyl esters. After removal of all of the EPA and DHA from the algal methyl esters considered herein, the remaining product could contain up to 50% saturated methyl esters, which is similar to the saturated content of beef tallow methyl ester, a fuel that is known to have poor cold ow properties. Acknowledgments The authors would like to acknowledge funding of this work by the US Department of Energy under contract DE-EE0003046 awarded to the National Alliance for Advance Biofuels and Bioproducts. References
[1] R.L. Hirsch, R. Bezdek, R. Wendling, Peaking of world oil production: impacts, mitigation and risk management, US Dept. Energy/National Energy Technology Laboratory, 2005. [2] B. McKibben, A deeper shade of green, National Geographic 210 (2) (2006) 3741. [3] M.S. Shapouri, J.A. Dufeld, M. Wang, The energy balance of corn ethanol: an update, United States Department of Agriculture, Agricultural Economic Report Number 814, 2002. [4] J. Van Gerpen, Biodiesel processing and production, Fuel Processing Technology 86 (2005) 10971107. [5] Congress, 110th United States, Energy Independence and Security Act of 2007, House Bill H.R.6, 2007. [6] H. Greenwell, L. Laurens, R. Shields, R. Lovitt, K. Flynn, Placing microalgae on the biofuels priority list: a review of the technological challenges, Journal of the Royal Society, Interface (2010) 703726. [7] A. Demirbas, Importance of biodiesel as transportation fuel, Energy Policy (2007) 46614670. [8] L. Gouveia, A. Oliveira, Microalgae as a raw material for biofuels production, Journal of Industrial Microbiology & Biotechnology (2009) 269274. [9] L. Rodol, G.C. Zittelli, N. Bassi, G. Padovani, N. Biondi, G. Bonini, R. Tredici, Microalgae for oil: strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-cost photobioreactor, Biotechnology and Bioengineering (2009) 100112. [10] Q. Hu, M. Sommerfeld, E. Jarvis, M. Ghirardi, M. Posewitz, M. Seibert, A. Darzins, Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances, The Plant Journal (2008) 621639. [11] Y. Chisti, Biodiesel from microalgae, Biotechnology Advances (2007) 294306. [12] K.M. Weyer, D.R. Bush, A. Darzins, B.D. Willson, Theoretical maximum algal oil production, Bioenergy Research 3 (2009) 204213. [13] Monthly Biodiesel Production Report, U.S. Energy Information Administration, DOE/EIA 0642(2009/12), Release Date: October 2010, December 2009. [14] M. Graboski, R. McCormick, Combustion of fat and vegetable oil derived fuels in diesel engines, Progress in Energy and Combustion Science (1998) 125164. [15] G. Knothe, R. Dunn, M. Bagby, Biodiesel: the use of vegetable oils and their derivatives as alternative diesel fuels, Fuels and Chemicals From Biomass (1997) 172208. [16] A. Roncarati, A. Meluzzi, S. Acciarri, N. Tallarico, P. Melotti, Fatty acid composition of different microalgae strains (Nannochloropsis sp., Nannochloropsis oculata (Droop) Hibberd, Nannochloris atomus Butcher and Isochrysis sp.) According to the culture phase and the carbon dioxide concentration, Journal of the World Aquaculture Society (2004) 401411. [17] M. Ramos, C. Fernandez, A. Casas, L. Rodriguez, A. Perez, Inuence of fatty acid composition of raw materials on biodiesel properties, Bioresource Technology (2009) 261268.

4. Summary and conclusions The objective of this study was to evaluate the effects of LC-PUFA such as EPA and DHA on oxidative stability, ignition quality, density, viscosity, speed of sound, bulk modulus and cold ow properties of algal methyl ester biodiesel. The study was accomplished by applying a variety of ASTM test methods on a series of model algal methyl ester compounds along with several real algal methyl esters that were derived from algal lipids from several different sources. The model algal methyl ester compounds were formulated based on fatty acid compositions of specic algae species from the literature. The model compounds were then reformulated to represent varying amounts of removal of EPA and DHA. A major outcome of this study was to determine the amount of EPA and DHA removal from algal methyl esters that would be required to meet the ASTM and EN Standards for oxidative stability and Derived Cetane Number. It was found that the oxidative stability induction period varies exponentially to the calculated BAPE value of the algal methyl esters. The results of this study suggest that approximately 50% of the EPA and DHA would need to be removed from the N. oculata methyl ester to meet the 3 hour ASTM oxidative stability specication and approximately 80% would need to be removed to meet the EN 6 hour specication. For I. galbana, removal of 100% of the EPA and DHA was insufcient for passing either the ASTM or EN oxidative stability requirements. The effect of the presence of EPA and DHA on ignition quality was also found to be quite substantial. Specically, it was found that nearly 100% of the EPA and DHA would need to be removed from the algal methyl esters considered herein to comply with the ASTM D6571 standard for minimum Cetane Number of 47 for biodiesel. Other fuel properties such as viscosity and density were well within specications for all the algal methyl ester formulations tested herein. A fuel additive containing TBHQ was found to be very effective in increasing the oxidative stability of the algal methyl ester model compounds. Specically, it was found that addition of only 0.03% of TBHQ is sufcient for Nannochloropsis sp. methyl esters to pass the ASTM specication even if no EPA or DHA were removed. To pass the EN

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001

Cloud Point [oC]

H.B. Bucy et al. / Algal Research xxx (2012) xxxxxx [18] G. Knothe, R. Dunn, Dependence of oil stability index of fatty compounds on their structure and concentration and presence of metals, Journal of the American Oil Chemists' Society (2003) 10211026. [19] T. Durrett, C. Benning, J. Ohlrogge, Plant triacylglycerols as feedstocks for the production of biofuels, The Plant Journal (2008) 593607. [20] G. Karavalakis, D. Karonis, S. Stournas, Evaluation of the oxidation stability of diesel/biodiesel blends using the modied Rancimat method, SAE 2009-01-1828, 2009. [21] G. Knothe, Some aspects of biodiesel oxidative stability, Fuel Processing Technology (2007) 669677. [22] R. McCormick, M. Ratcliff, L. Moens, R. Lawrence, Several factors affecting the stability of biodiesel in standard accelerated tests, Fuel Processing Technology (2007) 651657. [23] G. Knothe, Biodiesel: current trends and properties, Topics in Catalysis (2010) 714720. [24] R. McCormick, S. Westbrook, Storage stability of biodiesel and biodiesel blends, Energy & Fuels (2010) 690698. [25] G. Knothe, Structure indices in FA chemistry: how relevant is the iodine value? Journal of the American Oil Chemists' Society (2002) 847854. [26] D. Berthiaume, A. Tremblay, Study of Rancimat Test Method in Measuring the Oxidation Stability of Biodiesel Ester and Blends, Oleotek, Inc., 2006 [27] S. Jain, M. Sharma, Stability of biodiesel and its blends: a review, Renewable and Sustainable Energy Reviews (2010) 667678. [28] J.A. Waynick, Characterization of Biodiesel Oxidation and Oxidation Products, NREL, CRC Project No. AVFL-2b, 2005. [29] K. Arisoy, Oxidative and thermal instability of biodiesel, Energy Sources, Part A: Recovery, Utilization, and Environmental Effects (2008) 5161522. [30] J. Pahgova, L. Jorikova, J. Cvengros, Study of FAME stability, Energy & Fuels (2008) 19911996. [31] T. Mata, A. Martins, N. Caetano, Microalgae for biodiesel production and other applications, Renewable and Sustainable Energy Reviews (2010) 217232.

13

[32] D. Williard, T. Kaduce, S. Harmon, A. Spector, Conversion of eicosapentaenoic acid to chain-shortened omega-3 fatty acid metabolites by peroxisomal oxidation, Journal of Lipid Research (1998) 978986. [33] Martek Biosciences Corporation, http://www.lifesdha.com/Finding-lifesDHA-/ Partner-Products/tabid/129/Default.aspx 2009. [34] U.S. Department of Energy, National Algal Biofuels Technology Roadmap, DOE Biomass Program, 2010. [35] G. Karavalakis, S. Stournas, Impact of antioxidant additives on the oxidation stability of diesel/biodiesel blends, Energy Fuels 24 (6) (2010) 36823686. [36] ASTM D6751, Standard Specication for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels, ASTM International, 2009. [37] EN 14112, Determination of Oxidation Stability (Accelerated Oxidation Test), European Standard, 2003. [38] ASTM D7170, Standard Test Method for Determination of Derived Cetane Number (DCN) of Diesel Fuel Oils Fixed Range Injection Period, Constant Volume Combustion Chamber Method, ASTM International, 2009. [39] ASTM D7042, Standard Test Method for Dynamic Viscosity and Density of Liquids by Stabinger Viscometer (and the Calculation of Kinematic Viscosity), ASTM International, 2010. [40] ASTM D6371-05, Standard Test Method for Cold Filter Plugging Point of Diesel and Heating Fuels, 2010. [41] ASTM D2500-11 Standard Test Method for Cloud Point of Petroleum Products, 2011 [42] B.C. Fisher, A.J. Marchese, J. Volckens, T. Lee, J. Collett, Measurement of gaseous and particulate emissions from algae-based fatty acid methyl esters, SAE International Journal of Fuels and Lubricants 3 (2010) 292321. [43] A. Benamotz, T. Tornabene, W. Thomas, Chemical prole of selected species of microalgae with emphasis on lipids, Journal of Phycology 21.1 (1985) 7281. [44] M.J. Murphy, J.D. Taylor, R.L. McCormick, Compendium of Experimental Cetane Number Data. NREL/SR-540-36805, 2004. [45] A. Boehman, D. Morris, J. Szybist, The impact of the bulk modulus of diesel fuels on fuel injection timing, Energy & Fuels 18 (2004) 18771882.

Please cite this article as: H.B. Bucy, et al., Chemical and physical properties of algal methyl ester biodiesel containing varying levels of methyl eicosapentaenoate and methyl docosahexaenoate, Algal Research (2012), doi:10.1016/j.algal.2012.02.001