originally appeared in:

september 2007 issue, pgs 6771

Used with permission. www.Hydrocarbonprocessing.com

Article copyright 2007 by Gulf Publishing Company. All rights reserved. Not to be distributed in electronic or printed form, or posted on a Website, without express written permission of copyright holder.

Refining Developments

SpecialRepoRt

New developments in renewable fuels offer more choices

vegetable oil-based diesel can offer better integration within crude-oil refineries for fuels blending
J. Holmgren, C. gosling, r. marinangeli and T. marker, Uop LLC, Des plaines, Illinois, and g. FaraCi and C. Perego, eni S.p.A. Refining and Marketing Division, Novara, Italy

enewable fuels production is expanding worldwide due to increasing petroleum prices, government regulations and commitments to reduce greenhouse gas (GHG). Little integration exists between renewable fuels production within petroleum refineries, despite the rapidly increasing demand for renewable fuels. This segregated production increases costs, since the existing infrastructure for distribution and production of petroleum-based fuels is not utilized. Renewable fuels can have greater application in meeting the growing demand for transportation fuels if economic opportunities for blending or co-processing them within traditional petroleum refineries can be developed. A new technology can use renewable feedstocks derived from vegetable oils to produce a high-quality diesel fuel. This new diesel process has possible advantages over other renewable-fuel technologies such as fatty acid methyl ester (FAME) also known as biodiesel.

Background for renewable fuels. Biofuels production is expanding at a rapid pace due to increasing petroleum

prices, government regulations and GHG-reduction pledges. Future biofuel usage will depend on developing technology to produce high-quality transportation fuels from highly oxygenated, biologically derived, feedstocks. These new biofuels must be compatible with the existing fuel transportation and infrastructure. Existing technology to produce diesel fuel from vegetable oil has been largely centered on FAME biodiesel.1,2 While FAME has many desirable qualities such as high cetane, there are other issues associated with its usage, such as poor stability and high solvency, leading to filter plugging problems. A different processing route to convert vegetable oils into a high-quality diesel fuel or diesel blendstock that is fully compatible with petroleum-derived diesel fuel is desired. Under a collaborative research effort, a conventional hydroprocessing technology that is widely deployed in refineries and uses the existing refinery infrastructure and fuels distribution system was further developed. This new technology processes widely available vegetable-oil feedstocks to produce a high-cetane, low-

Feed Methanol Vegetable oil H2 Vegetable oil Fig. 1 Chemical structures of triglyceridesred = oxygen, white = hydrogen, green = carbon atoms. Fig. 2

Process Biodiesel

Products Biodiesel Glycerol

Hydroprocessing

Green diesel

vegetable-oil processing routes for transportation fuels.

HYDROCARBON PROCESSING SeptembeR 2007

SpecialRepoRt

Refining Developments
and can be a valuable renewable LPG product in its own right. The oxygen contained in the feed is rejected either as carbon monoxide (CO)/carbon dioxide (CO2) or water (H2O). In addition, all olefinic bonds are saturated and produces a consistent, pure-paraffin product. The chemistry used in the biodiesel or FAME route is difficult to substantiate. The stoichiometry of the reaction is: Triglycerides + Methanol ====> FAME + Glycerol 100 bbl 13 bbl NaOH 99 bbl 8 bbl About 8% of the product volume is glycerol, which is a lowvalue product in the unrefined form and has a limited market. Methanol is required as a co-feed, and feedstocks containing high concentrations of free-fatty acids can cause operational problems due to saponification reactions with the caustic present as the catalyst. For this reason, newer FAME units include costly technology to remove or treat free fatty acids. As noted previously, the new diesel process removes the oxygen by reaction with hydrogen producing a pure-paraffinic product. The volume yield is 99%. The primary co-product is propane, and the byproducts are H2O and carbon oxides. Feedstocks containing free-fatty acids can be processed in the green diesel process without the problems caused in FAME production: Green Triglycerides + H2 ====> diesel + H2O/CO2 + Propane 100 bbl Catalyst 99 bbl 9 bbl Several basic advantages over FAME production are possible with the new diesel processing technology when integrated within a refinery setting. Hydrogen required for the reaction is often readily available, and all reaction products are compounds that are found in normal refinery products and do not require any special handling. The hydrocarbon products can all be easily blended with conventional refinery products.

gravity, low-aromatics and sulfur-free diesel fuel. The cold-flow properties of the biofuel can be adjusted to meet various cloudpoint specifications in either the neat or blended fuel. The new process is targeted for commercialization by 2009.
ProCessing BiologiCally Feeds

Processing of biologically derived feedstocks is complicated as these materials contain a significant amount of oxygen. The primary feedstocks for renewable diesel are vegetable oils such as soybean, palm, jatropha or rapeseed oils. Other products such as animal fats and greases can also be used as a feedstock. Plant oils mainly consist of triglycerides with typically 1% 2% free-fatty acid content. The chemical structure of a triglyceride molecule is shown in Fig. 1.
organic fuels. Triglycerides and free-fatty acids both containing long, linear aliphatic hydrocarbon chains. The aliphatic hydrocarbon always contains an even number of carbon atoms, is generally unsaturated, and also corresponds to the carbon number range typically found in diesel fuels. There is also a three-carbon backbone in the triglyceride molecule. Table 1 summarizes the properties of various vegetable oils as compared to petroleum-derived diesel fuels. The volume yield for diesel fuel from vegetable oils is nearly 100%, whereas diesel represents only about 20% of the volume of crude oil when distilled. Two processing routes for converting vegetable oil into diesel fuel are shown in Fig. 2. The conventional processing route for diesel fuel production is via transesterification with methanol in the presence of caustic to produce FAME and glycerol as a byproduct. The produced glycerol is produced from the threecarbon backbone and has a limited market, particularly in the unrefined form. The hydroprocessing route uses hydrogen to remove oxygen from the triglyceride molecules. This is the route used in the new process to produce green diesel. The oxygen is easily removed via two competing reactions: decarboxylation and hydrodeoxygenation. The extent for each reaction depends on the catalyst and process conditions. The three-carbon backbone in the process to produce green diesel yields propane, which can be easily recovered

Vegetable oil H2 Diesel Diesel hydrotreater Green diesel Reactor Acid-gas removal CO2

Makeup hydrogen

Pretreat and co-process H2 Vegetable oil

Pretreater

Remove contaminants H2 Separator Water

Propane and light ends

Standalone unit H2 Vegetable oil Diesel

Diesel hydrotreater

Green diesel

Naphtha or jet fuel

Green diesel unit Separate unit avoids diesel hydrotreater catalyst life issues and increases exibility

Diesel product Fig. 4 the new green diesel process converts vegetable oil into fuels.

Fig. 3

Alternative vegetable-oil hydroprocessing routes to transportation fuels.

HYDROCARBON PROCESSING SeptembeR 2007

Refining Developments
While hydroprocessing was clearly the chemistry preferred for refinery applications, how to implement it in a process design is not obvious. Two options can be considered: Co-processing in an existing distillate hydroprocessing unit Building a standalone unit as shown in Fig. 3. The co-processing route was initially evaluated as an option, since existing equipment could be reused, resulting in a lower implementation cost. Co-processing was problematic after some initial evaluations were completed. Vegetable oils contain trace metal contaminants such as phosphorous, sodium, potassium and calcium. A pretreating reactor may be required to remove contaminants or an existing reactor will generally not have sufficient catalyst volume for the treating catalyst. Processing vegetable oil reactions are fairly exothermic and may require quench capabilities that may not necessarily be available. The deoxygenation products (H2O, CO, CO2) will require revamping of the recycle-gas system for removal, or using a substantial purge stream. Cold-flow properties of the combined diesel product may limit the quantity of vegetable oil that can be processed as n-paraffinthe primary product from hydrotreating vegetable oiland will impact the cloud point. In addition, it was found that the deoxygenation reactions had a tendency to compete with the primary desulfurization reactions taking place within the hydrotreating unit. This was seen to present too much risk in a fuel environment where ultra-low-sulfur diesel (ULSD) is required and rerun of off-spec products is expensive. After considering all the potential issues, it was more costeffective to build a dedicated unit that was optimized for vegetable-oil processing due to the unique nature of these feedstocks. Fig. 4 is a simplified flow scheme showing the principles of the new process. In the process shown in Fig. 4, vegetable oil is combined with hydrogen, brought to reaction temperature and then sent to the reactor where the vegetable oil is converted to the green diesel product. This product is separated from the recycle gas in the separator, and the liquid product is sent to a fractionation section. Designing the fractionation section can vary from a one-column system producing on-specification diesel and unstabilized naphtha to a three-column system producing propane, naphtha and diesel products. Most installations will be a singlecolumn design, and the lighter products will be recovered by other existing refinery process units. The recycle gas is treated in an amine system to remove CO2.
Pilot-plant work. Extensive process performance testing has neW diesel ProCess

SpecialRepoRt

evaluate other potential feedstocks including tallow and greases derived from animals. Considerable testing to confirm stability of catalyst system applied and data from an ongoing stability test evaluating rapeseed oil is shown in Fig. 5. Pilot-plant tests have shown that there is no measurable deactivation after over 2,000 hours onstream.
green diesel. The product properties are similar for all the veg-

etable oils processed. Table 2 summarizes typical product yields. Cold-flow properties can be controlled by the severity used in the process. So, the diesel yield will vary from 99 vol% to 88 vol%, which is required to achieve a 10C cloud point. The diesel yield loss is to kerosine and naphtha. In all cases, the produced product has a very high-cetane number (>80) and contains no sulfur or aromatics. Table 2 summarizes the range of hydrogen consumption for the various feedstocks. Palm oil requires substantially less hydrogen than rapeseed oil due to the lower olefin content. properties, and is very similar to diesel product that is produced via Fischer-Tropsch processes. The diesel-product properties are compared to mineral (petroleum) diesel and biodiesel in Table 3. The diesel fuel properties of FAME are not ideal for diesel blending because the cloud point is higher than EN590 requires, and the density can be an issue when the T95 is considered. The gravity is high, which can limit blending of lower-value products including hydrotreated light-cycle oil (LCO) into the refinery diesel pool. It has a narrow boiling range and can have issues in meeting stability specifications due to the inherent instability of esters. The heating value is about 12% lower than typical mineral diesel, which will reduce the fuel economy of FAME/diesel blends relative to mineral diesel or green diesel/mineral diesel blends. The green diesel, however, is a premium blending component. The boiling range is comparable to typical diesel products, with substantially higher cetane and lower density. These are very valuable properties; it allows blending low-value hydrotreated LCO into the typical refinery diesel pool and still meeting European Union (EU) diesel specifications. An example is shown in Table 4, where a straight-run diesel and kerosine blend can be produced to meet a 50-cetane ULSD product specification. No LCO can be blended without using cetane-enhancing additives. If green diesel is added to the pool, the high-cetane value and low density allow for blending a nearly equivalent volume of hydrotreated LCO into the ULSD product. Assuming a typical discount of $4.60/bbl for fuel oil relative to ULSD, this translates to a benefit of $9,200/day or over $3 million/yr that could be realized through green diesel blending.

Fuel properties. The green diesel product has excellent fuel

been done to determine optimum process conditions, catalyst stability and product properties. A range of vegetable-oil feedstocks has been processed in pilot plants including soybean, rapeseed and palm oil. An extensive program is underway to Table 1. Comparison of vegetable oil feedstocks35
Bio feedstock WTI/Brent Crude Rapeseed oil Soy oil Palm oil Jatropha oil price, $/bbl 68 89 75 62 44 Diesel yield, vol% 20 99 99 99 99 Carbon number 1122 1622 1618 1618 1618 olefins, mole% 0 94 84 58 77

green diesel process economics. Techno-economic

feasibility studies were conducted as part of the development

Table 2. Green diesel product yields

feed products Vegetable oil, % H2, % Naphtha, vol% Diesel, vol% Cetane number S, ppm 100 1.53.8 110 8898+ >80 <1

HYDROCARBON PROCESSING SeptembeR 2007

SpecialRepoRt

Refining Developments
Since vegetable oil pricing is decoupled from crude oil pricing, a sensitivity analysis was conducted to determine the crude oil price required for the green diesel process to be profitable. Three cases were evaluated: Palm oil at $420/metric ton (mton) ($52/bbl) Soybean oil at $560/mton ($67/bbl) Soybean oil with $1/gal subsidy. Results of this evaluation are shown in Fig. 6. Palm oil is profitable without any subsidy at a crude oil price of $52/bbl. Soybean oil is only profitable at a crude oil price of $67/bbl and will require a subsidy to be viable with currently published crude oil price forecasts. If the current subsidy of $1/gal for FAME is assumed to be extended to the green diesel, then the process is viable at a crude oil price of $38/bbl and has a payback of less than 1 year at a crude oil price of $60/bbl.
green diesel: more sustainable product. A life-cycle analysis (LCA) of different diesel-production routes was conducted. The LCA evaluation program is a method to determine and compare the environmental impacts of alternative products or processes, including impacts of initial resource extraction to waste disposal. In this study, the scope of the analysis was from crude-oil extraction through combustion of the refined diesel fuels in a vehicle. It was assumed that all fuels had the same performance in the vehicle. The primary focus of the analysis was on fossil energy use and GHG emissions, although other impact categories are included.

process to ensure that the new technologies are competitive with alternatives already available to refiners. As in all refinery processes, the primary factor in the green diesel process economics is the feedstock costs. Table 5 summarizes the capital and variable operating costs for two feedstocks. The variable costs in Table 5 include utilities, hydrogen, catalyst and chemical cost expenses. Hydrogen was valued at $3/Mscf for this evaluation.

110 Green diesel, vol% yield 105 100 95 90 85 80 75 70 0

500

1,000 1,500 Hours in service

2,000

2,500

Fig. 5

Catalyst stability testing in the production of green diesel from vegetable oil.

500 400 300 200 100 0 -100 -200 -300 -400 -500 30

$560 mton ($1.96/gal) w/tax incentive $560 mton ($1.96/gal) $420 mton ($1.47/gal)

Table 3. Green diesel fuel properties5 Mineral ULSD

O, % Specific gravity Sulfur content, ppm Heating value, MJ/kg Cloud point, C Distillation, C Cetane number Stability 0 0.84 <10 43 5 200 to 350 40 Good

NPV, $MM

Biodiesel, FAME
11 0.88 <1 38 5 to +15 340 to 355 50 to 65 Marginal

Green diesel
0 0.78 <1 44 10 to +20 265 to 320 70 to 90 Good

40

50 60 Petroleum crude price, $/bbl

70

80

Fig. 6

economic analysis of green diesel production costs vs. crude oil price.

Table 4. Green diesel blending benefits

300 250 Impact points 200 150 100 50 0 Petroleum Biodiesel diesel Green diesel Kg CO2/Kg fuel 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Diesel-pool components Kerosine Hydrotreated straight-run diesel Hydrotreated LCO Green diesel Blended product cetane Barrels in pool 500 7,500 2,000 2,346 Cetane index 41 52 20 74 50

Petroleum Biodiesel diesel

Green diesel

Table 5. Economic study costs

palm oil ISBL cost, $MM 33.9 6.8 5.40 OSBL cost, $MM Variable cost, $/bbl soybean oil 33.9 6.8 6.94

Fig. 7

life-cycle analysis of various green transportation fuelsimpact points and gHg emissions.

HYDROCARBON PROCESSING SeptembeR 2007

Refining Developments
Fig. 7 summarizes the results of the LCA. Both the FAME and green diesel products have much lower total environmental impact scores than mineral diesel. Both fuels have significantly lower production of climate-active CO2. However, green diesel has both a lower environmental impact and lower climate active CO2 production than FAME. The environmental impact of FAME production is higher due to the methanol-feedstock requirement, which is produced from natural gas through a very energy-intensive process with high environmental burdens.
Fuels outlook. Refiners are well positioned to play a major

SpecialRepoRt

Jennifer Holmgren is the director of UOPs Renewable Energy and Chemicals

business unit. She is responsible for the development and commercialization of new technologies and products for the processing of renewable energy sources. Dr. Holmgren received a BS degree in chemistry from Harvey Mudd College in Claremont, California, and a PhD in inorganic materials synthesis from the University of Illinois at Urbana-Champaign and an MBA from the University of Chicago. Previous assignments at UOP included the preparation and characterization of novel zeolites, molecular sieves and layered materials as, well as a number of technology delivery projects in the BTX and olefins areas.

role in renewable fuel production in the future. A new process to produce green diesel is a sustainable route to convert vegetable oil into premium-quality diesel fuel and could possibly be the first step in providing technology to enable refiners to participate in renewable fuels. The new diesel product is a superior alternative to FAME, with significantly better diesel product properties, and is fully compatible with conventional mineral diesel products. HP

Chris gosling is the senior manager of refining conversion process development for UOP and is responsible for development of FCC, hydroprocessing and renewable energy technologies. He joined UOP in 1980 after graduating from Michigan Technological University with BS and MS degrees in chemical engineering. Mr. Gosling has held various positions within UOP, working in a number of different technologies in research and development, technical service and product-line management. He is a registered professional engineer in Michigan and holds 38 patents. rich marinangeli is the manager of UOPs refining conversion development
group for the Renewable Energy & Chemicals business unit. He has 28 years of experience at UOP in process development and application of materials for catalysis and adsorption with 18 US patents. Dr. Marinangeli holds a PhD in chemical engineering from Princeton University and a BS degree in chemical engineering from the University of Notre Dame. He was a CNRS Fellow in Villeurbanne, France, prior to joining UOP.

ACKNOWLEDGEMENT Revised and updated from an earlier presentation from the 2007 NPRA Annual Meeting, March 18-20, 2007, San Antonio, Texas. LITERATURE CITED Alternative Fuels, National Renewable Energy Laboratory Website, http://www.eere.energy.gov/afdc/altfuel/whatis_biodiesel.html. 2 Radich, A. Biodiesel Performance, Costs, and Use, Energy Information Administration, 2004, http://www.eia.doe.gov/oiaf/analysispaper/biodiesel/. 3 World Bank Commodity Price Data, http://www.econ.worldbank.org, January August 2006. 4 ICIS Chem Business, Chemical Market Reporter, October 2006. 5 Alternative Fuels Comparison Chart, National Renewable Energy Laboratory website, http://www.eere.energy.gov/afdc/altfuel/fuel_comp.html.
1

Terry marker is a UOP senior engineer and team leader working with the Ecofining process for the production of green diesel. Dr. Marker holds a BS degree in chemistry from the University of Illinois and an MS degree in chemical engineering from Illinois Institute of Technology. In her 33 years of research and development experience, she has been awarded over 30 patents. giovanni Faraci is responsible for optimization process technologies for Enis
Refining and Marketing Division. He graduated in chemistry from Palermo University and has over 29 years of experience in refining process research, with a focus in the area of pilot-plant testing and catalyst evaluation.

Carlo Perego is the manager of the Eni Refining & Marketing Research Centre
in Novara, Italy. He received the Laurea in industrial chemistry in 1978 from the University of Milan, Italy. In 1980, he joined the Montedipe R&D Department. His main research activity was in the area of phenol and styrene processes. In 1989, he took the position of pilot-plant manager at Sigma 3V, a fine-chemical company. In 1990, he joined the corporate research company of Eni SpA, covering different position in the research center for catalysis and catalytic processes and in particular in the area of zeolite catalysis. He is a member of the Catalysis Commission of the International Zeolite Association.

Article copyright 2007 by Gulf Publishing Company. All rights reserved. Printed in U.S.A. Not to be distributed in electronic or printed form, or posted on a Website, without express written permission of copyright holder.