Integration of Non-Fuel Coproducts into the GREET Model

Using this approach, a GREET-compatible external tool was developed to calculate the life-cycle inventory of GTL coproducts to determine the life-cycl...
0 downloads 0 Views 704KB Size
Download PDF
Article pubs.acs.org/est

Integration of Non-Fuel Coproducts into the GREET Model Grant S. Forman*,† and Stefan Unnasch‡ †

Sasol North America, 900 Threadneedle, Suite 100, Houston, Texas 77079-2990, United States Life Cycle Associates, LLC., Unit A11, 884 Portola Road, Portola Valley, California 94028, United States

‡

S Supporting Information *

ABSTRACT: The life-cycle greenhouse gas (GHG) emissions of alternative fuels that are capable of replacing conventional, petroleum-derived gasoline and diesel continue to be scrutinized for policy implementation. These alternative fuel technologies can also produce a number of value-adding nonfuel coproducts that require thorough and rigorous assessment in order to achieve an accurate life-cycle GHG emissions value. By using the gas to liquids (GTL) diesel pathway as a proxy for other alternative fuel pathways with coproducts, this paper examines how integration of coproduct analysis using the substitution method is possible within the existing framework and functionality of the GREET model. Using this approach, a GREETcompatible external tool was developed to calculate the life-cycle inventory of GTL coproducts to determine the life-cycle GHG emissions of GTL diesel using the substitution method. In addition to having built-in regional scenarios, this tool allows the user the flexibility to configure a given GTL product slate and to calculate the life-cycle GHG emissions of GTL diesel based on a given product composition. Using this protocol, the life-cycle GHG emissions of GTL diesel can range from 71.7 to 95.7 gCO2e/MJ on a well to wheel basis, with the range in carbon intensity being dependent on the mix of coproducts. These results highlight a weakly understood relationship between fuel and chemical products in LCA models. The coproduct integration approach described herein could potentially be incorporated into fuel LCA models, such as GREET, to allow users to further understand the potential environmental benefits of alternative fuel pathways, such as GTL.

■

INTRODUCTION In the future, transportation fuel sold in the United States (U.S.) could come from a more diverse mix of feedstock sources other than crude oil.1 Due to heightened environmental and energy security concerns, regulations continue to be developed in the U.S. to improve energy efficiency and promote environmentally benign alternative fuels in order to reduce petroleum consumption, particularly in the transportation sector, where the majority of petroleum products are consumed. The Renewable Fuel Standard (RFS) program regulations were developed with the intention of ensuring transportation fuel sold in the U.S. contains a minimum volume of renewable fuel.2 In the State of California, the Low Carbon Fuel Standard (LCFS) was implemented with the intention of reducing carbon intensity in transportation fuels as compared to conventional petroleum fuels, such as gasoline and diesel.3 In addition, the recent rapid development of natural gas production from extremely low-permeability shale formations has generated interest in expanding U.S. natural gas usage beyond the realm of electricity generation to transportation.4 These alternative transportation fuel pathways, which are growing in number and complexity,5 must be critically analyzed to determine their greenhouse gas (GHG) emission impacts while concurrently considering scalability and security of supply issues. A common element of many alternative fuel conversion technologies is the production of valuable chemical, agricultural, and electricity coproducts in addition to hydrocarbon © 2015 American Chemical Society

fuels. In the U.S., an efficient supply chain and close vicinity to end markets ensures that these coproducts are an integral component of sustaining the alternative fuel facilities profitability. For example, processing of corn and sugar cane yields distillers’ grain solubles (DGS) and bagasse, which are used as livestock feed and fuel for heat and electricity generation.6 Soybean-derived biodiesel synthesis also coproduces soy meal as animal feed and glycerin.7 Algae production also generates glycerin, residual biomass for animal feed, and omega-3 fatty acids for nutraceuticals.8 Depending on market need, the Gas to Liquids (GTL) process can be tailored to produce a number of products other than GTL diesel, GTL naphtha,9 and Liquefied Petroleum Gas (LPG). Potential additional chemical products include GTL lubricant base oils,10 GTL waxes, GTL normal paraffin,11 and electricity. In many cases, the quality of the coproduct produced from an alternative fuel facility can be superior to analogous materials produced from conventional petroleum-derived crude oil.12 For example, both GTL fuel and chemical products share the common qualities of very low levels of sulfur and aromatics while at the same time being rich in hydrogen relative to crude oil derived analogues. Taking into account these nonfuel coproducts becomes critical in achieving an accurate life-cycle GHG emissions Received: Revised: Accepted: Published: 4372

December 10, 2014 February 19, 2015 February 24, 2015 February 24, 2015 DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology

that have employed only energy allocation methodology have afforded lifecycle GHG emissions slightly higher than current petroleum-derived conventional diesel baselines, while LCAs of GTL diesel that have used the systems boundary expansion method have generally produced lifecycle GHG emissions equivalent to current conventional diesel baselines (see the Supporting Information, SI).25 The GREET model26 provides the flexibility for both allocation and substitution methods and is configured with displacement credits for many fuel pathways, such as distiller’s grains from corn ethanol and electric power from Fischer− Tropsch (FT) diesel, cellulosic ethanol, and sugar cane ethanol. In the case of FT and renewable diesel production, diesel, jet fuel, and naphtha are treated as primary fuel products. The GREET model currently treats coproducts from FT diesel as fuel products and allocates energy and emissions based on the energy content of these products. Energy inputs and emissions are allocated between diesel and naphtha and in some scenarios electric power is treated as a coproduct with a displacement credit for grid power. In reality, however, the GREET model currently oversimplifies the handling of coproducts from FT processes, such as GTL. Currently, the GREET model does not accommodate value-adding chemical coproducts from GTL, including GTL waxes, GTL lubricant base oils and GTL normal paraffin for Linear Alkyl Benzene (LAB) production. In practice, these nonfuel coproducts are likely to be part of a GTL facility that is located near large, accessible markets for these value-added chemicals, such as the U.S. Thus, the yield of the primary GTL product, GTL diesel, can vary significantly between 40 and 80%, depending on the amount of nonfuel coproducts produced from a GTL facility. Even if the GTL facility is configured to produce primarily GTL diesel, it will also coproduce GTL naphtha, which is more likely to be used as a feedstock for chemical production, rather as a gasoline blendstock. Due to its chemical composition (almost exclusively saturates with a high percentage of normal paraffins), GTL naphtha has a particularly poor octane number. The exact number is difficult to measure directly as it is below the bottom range of octane measurement (at 40 octane) and can thus only be determined in a blend with higher octane components, subject to various assumptions. The resulting octane number is somewhere between 15 and 30, in contrast to the octane numbers (Antiknock Index, AKI) of typical gasoline in the range of 87 to 93. Significant additional processing of GTL naphtha in combination with import of other blend components would be required to meet the gasoline specification.27 Consequently, GTL naphtha is utilized as a feedstock for steam cracking, rather than as a gasoline blendstock, which is the current assumption in the GREET model. GTL Naphtha is highly paraffinic and has an essentially zero aromatics content, resulting in greater olefin yields and reduced coking of furnace tubes and catalyst relative to cracking of conventional naphtha, allowing for extended run durations.9 These properties result in GTL naphtha being utilized as a feedstock for steam cracking, rather than as a gasoline blendstock, which is the current incorrect assumption in the GREET model. Using the substitution method, these impacts would create a coproduct credit, however currently in the GREET model GTL naphtha is only considered as a gasoline blend component and emissions are allocated to it as if it were a fuel product.

assessment of a particular alternative fuel pathway. As the number and complexity of alternative fuel pathways increase, the need to equitably assess coproducts in Life-Cycle Analysis (LCA) becomes more acute. In the context of policy frameworks to promote fuels with lower carbon intensity, the financial implications of incorrect transportation fuel life-cycle GHG emissions values can be potentially high.13 Consequently, recently LCA practitioners and policy makers have begun to apply a greater degree of granularity in determining upstream14 and production phase15,16 emissions to ensure greater certainty to baseline gasoline and diesel pathways. This trend toward greater LCA granularity is also extending to alternative fuel pathways as understanding of these processes and downstream coproduct impacts matures. How GHG emissions are distributed among alternative fuels and these coproducts differs on which allocation approach is employed. Allocation and substitution (also known as the “system expansion” method) are two common methods to accomplish the division between the alternative fuel functional unit and its coproducts.17 When the allocation method is applied, impacts belonging to a particular process are allocated to coproducts proportional to the amount of each coproduct produced.18 This amount is commonly measured in mass, monetary, or energetic units, although other units are also possible.19 Substitution requires utilizing data for comparative alternate coproduct processes typically outside the systems boundary that is used in the allocation method. Impacts from these external processes are then applied to the original multioutput process, leaving the impacts that are embodied in the functional unit.20 When the substitution approach is applied, all of the emissions for feedstock and processing are assigned to fuels, and the coproducts are accounted for by displacement. This latter approach results in higher emissions for the feedstock and production phase, which are offset by the coproduct displacement credit, thereby more accurately representing the impact of the coproducts in the overall system. Substitution is the preferred approach for the distribution of emissions to multiple products, as recommended by the ISO 14040 standards.21 This approach is largely followed by the fuel pathways approved under the Environmental protection Agency (EPA) Renewable Fuel Standard (RFS2)2 and California Low Carbon Fuel Standard (LCFS).3 Despite this, allocation has long been a subject of debate in the LCA community.22 Even among policy makers, the choice of partitioning criterion necessary to perform allocation is an entirely subjective decision. Often, the issue of gearing is used as a reason to avoid substitution. Several authors23 have noted that the substitution method24 can generate distorted LCA results if the coproducts in alternative fuels pathways are actually main products (for example, biodiesel, and renewable diesel from soybean fuel pathways). In addition, data requirements for substitution can also prove prohibitive for LCA practitioners who do not have access to industry data. These issues can lead to LCA practitioners’ employing hybrid allocation approaches or developing sensitivity analysis for different allocation scenarios. Not surprisingly, these divergent approaches to coproduct analyses can result in wide differences in life-cycle GHG emissions results for alternative fuels. Although numerous LCAs of the GTL pathway exist in the literature, most of these studies have employed only energy allocation methodology. This is most likely due to a lack of data associated with product quality, composition, and information on downstream applications. LCAs of the GTL diesel pathway 4373

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology

Figure 1. Systems boundary diagram for Gas-to-Liquids (GTL) conversion to fuels and coproducts.

The functional unit for the analysis is one megajoule (MJ) of product fuel energy, which includes GTL diesel, jet, and naphtha for fuel blending. Energy consumption is expressed as a unit-less ratio of total energy input (including fuel energy) per unit of fuel product output (J/MJ). GHG emissions are expressed as grams of CO2 equivalent per unit of fuel energy (gCO2e/MJ), commonly referred to as the carbon intensity (CI). The GHG emissions considered in this analysis are CO2, N2O, CH4, CO, and volatile organic compounds (VOC). GHG emissions from the tank to wheel (TTW, the emissions from vehicle use) portion of the analysis arise from the carbon content of the fuel (gC/MJ fuel) converted to CO2 (44.0 gCO2/12.0 × gC) plus vehicle emissions of CH4 and N2O. Global warming potentials (GWP, gCO2e/g constituent) for CH4 and N2O are based on estimates from the Intergovernmental Panel on Climate Change (IPCC)29 for a 100 year time horizon, while CO and VOC are counted as fully oxidized CO2 in the atmosphere and thus have a GWP of 1 when expressed as CO2.30 These analyses exclude the climate impacts of secondary and higher order atmospheric species that arise from direct emissions, including ozone, oxides of nitrogen (NOx), and aerosols.31 The analysis presented herein assigns all of the emissions to fuel products and coproduct credits are calculated for each of the nonfuel products based on the production-phase and downstream GHG impact of the displaced product. The coproduct credit is calculated from the life-cycle of the substitute products and all of the emissions from feedstock production and FT processing are assigned to fuel products (see SI for more details). The life-cycle GHG emissions correspond to the sum of all the inputs including feedstock collection and transport, GTL process, GTL fuel transport, and GTL diesel vehicle operation. Feedstock and GTL processing emissions and coproduct credits are allocated among fuels. This calculation approach follows the methods used in the GREET model for FT diesel from various feedstocks. Various natural gas resources are potential feedstocks for GTL diesel production. The energy and emissions intensity depends on the processing steps, transport, and fugitive methane emissions. The life-cycle GHG emissions are

The discussion above serves to underscore the need to have the necessary functionality and flexibility in the existing framework of the GREET model to include a more granular analysis of all possible GTL nonfuel coproducts using substitution methodology. Using the GTL process as a proxy for other GREET alternative fuel pathways, the objective of this study was to consider the displacement impacts of potential coproducts from the GTL process to calculate the life-cycle energy and Greenhouse Gas (GHG) emissions of GTL diesel. These analyses follow the calculation steps, functionality and assumptions applied in the CA_GREET model28 (version 1.8b) used by the California Air Resources Board (CARB) to calculate the GHG emissions for fuel pathways. The calculations take into account the energy impacts and GHG emissions from GTL diesel production coproducts for a range of natural gas based FT diesel pathways. By using a number of different regional, product slate and natural gas sourcing scenarios, this paper examines the impacts of nonfuel coproducts on the carbon intensity of transportation fuels from the GTL production processes. A variety of in-built coproduct scenarios are examined using the substitution method for several natural gas based GTL diesel scenarios and an input interface enables the user to develop a number of “what-if” scenarios to determine the life-cycle GHG emissions of GTL diesel based different product slates and regional assumptions.

■

METHODS AND DATA The systems boundary diagram for GTL diesel production from natural gas is shown in Figure 1. The fuel pathway includes natural gas production, conversion through the FT process to GTL fuels and coproducts, fuel transport, and vehicle end use. GTL coproducts include LPG, GTL naphtha, GTL lubricant base oils, GTL normal paraffin, GTL waxes, and electric power. Using the substitution method, the energy and materials required to produce these coproducts, along with any downstream impacts, are subtracted from similar burdens associated with conventional, petroleum-derived crude-derived analogues. 4374

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology

emissions (EFC) for natural gas include the upstream emissions to produce the gas (ENG) and upstream natural gas emissions associated with fuel production (ENGF). These additional upstream fuel cycle emissions (ENGF) are calculated according to eq 4, based on the FT process efficiency (η) and loss factor (L). The loss factor reflects FT diesel product lost in the distribution system;

calculated in the CA_GREET model (version 1.8b) based on extraction and production efficiency, fugitive leakage, and transport distance. LCI data for natural gas is constructed from aggregate extraction and processing efficiencies and fugitive emissions data.32,33 The GREET model represents data on gas production and recovery as efficiencies. Using eq 1, for gas recovery based on 100% natural gas for processing fuel, the emissions array for gas recovery (ENGRec) was calculated in the following manner: E NGRec = (1/ηR − 1) × (EFNGICE + E NG)

E FC = (E NG + E NGF) × L = (E NG + E NG × (1/η − 1)) × L

(1)

For GTL scenarios that use otherwise flared gas, the avoided life-cycle emissions from flared gas combustion include the avoided emissions from flaring (EFNGFL), such that the upstream fuel cycle emissions from flared gas are represented by eq 5:

where ηR represents the efficiency for natural gas extraction; EFNGICE represents the emissions factor for natural gas combustion equipment; and ENG represents the upstream fuel cycle emissions for natural gas. A recovery efficiency of 97.2% corresponds to 31 144 J of natural gas input per MJ of natural gas that is recovered plus additional fugitive losses. Small amounts of other fuels are also inputs for gas processing in the GREET model. It is noteworthy that a wide range of GHG emissions estimates are associated with oil and gas production due to differences in methodology, aggregation, and data assumptions relating in particular to well establishment, flaring, and methane fugitive emissions.34,35 This study uses the CA_GREET (version 1.8b) LCI data for natural gas, which is based on the average inventory for oil and gas production in the U.S. Currently, there is significant uncertainty and controversy relating to the environmental impacts associated with shale gas production. In particular, there is currently a wide variation in the estimates of GHG emissions associated with shale gas production due to differences in methodology and data assumptions relating in particular to methane fugitive emissions.36 However, it remains unclear how these assumptions also affect oil production, particularly tight oil from hydraulic fracking. Therefore, the parameters used by regulators (ARB) in fuel LCA are the basis for the current study.27 As shown in eq 2, natural gas processing emissions (ENGP) were calculated from and aggregate processing efficiency (ηP) and the emission factor for natural gas boilers (EFNGB), such that, E NGP = (1/η P − 1) × (EFNGB + E NG)

E FCFL = (E NG − EFNGFL) × 1/η × L

(5)

When grouped according to the GREET model steps, the equation is rearranged into eq 6: E FCFL = (E NG + E NG × (1/η − 1) × L − EFNGFL × 1/η × L

(6)

The GREET model treats the GTL process emissions as tail gas combustion combined using an energy balance.30 This approach approximates the operations in the GTL facility in an aggregate manner. As illustrated in eq 7, the non-CO2 GTL process emissions (EFTP) are modeled in proportion to the difference between input natural gas and GTL products, combined with an emission factor for tail gas combustion (EFNGFL): E FTP = (1/η − 1) × EFNGFL

(7)

The process CO2 emissions (EFTP,CO2) are determined by carbon balance. The carbon in feedstock is either converted to GTL products or carbon-containing species including CO2, CO, VOCs, or CH4 (eq 8): E FTP,CO2 = NGCO2 × 1/η × ((1 − ηC)

(2)

− minor species) × L

The upstream emissions for natural production are given by the recursive calculation (eq 3): E NG = E NGRec × L + E NGP

(4)

(8)

As shown in eq 9, the carbon factor for natural gas (NGCO2) is based on its lower heating value (LHV) and carbon content (CNG = 72.4%) represented on a CO2 basis such that,

(3)

where L represents the losses from processing and transmission. Upstream fuel cycle emissions for natural gas include the steps from extraction to delivery to the GTL facility. The GREET model is configured with the flexibility for imported power and light hydrocarbons as feedstocks for the GTL facility, however, in practice, world scale GTL facilities use only natural gas and generate power on site. The natural gas is either pipeline grade with natural gas liquids removed or raw gas, with the natural gas liquids still in the gas. Light ends are further removed at the GTL facility to provide a purified methane stream for reforming. In this study, the GREET inputs for supplemental electric power and butane were set to zero to reflect the 100% natural gas feedstock in operating GTL facilities. The GREET model groups the upstream fuel cycle emissions from FT fuels into two separate steps. Upstream fuel cycle

NGCO2 = C NG/LHV × 44.0/12.0 = 55.1 g CO2 /MJ

(9)

Equation 10 shows that the carbon efficiency (ηC) for GTL production corresponds to the ratio of the carbon in products to natural gas feedstock such that, ηC =

C NG × LHVNG) ∑ (Si × Ci/LHV)/( i

(10)

where S, C, and LHV represent the energy fraction, carbon content, and lower heating value for each GTL product. In the current study, the life-cycle GHG emissions from upstream production and GTL processing are assigned to GTL fuels based on the energy fraction of GTL diesel (SD), GTL aviation fuel (Saviation), and GTL naphtha (SNfuel), respectively. As shown in eq 11, The upstream and GTL production emissions (EFCi + EFTD) for GTL diesel, GTL aviation fuel, and GTL naphtha are the same per unit of energy since the 4375

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology Table 1. Distribution of Fuels and Coproducts from FT Diesel (U.S. GTL with Electricity Export Case) distribution (J/MJ out) GTL products

application

energy share (Si) (%)

products

diesel GTL naphtha GTL lubricant base oil GTL waxes GTL paraffin LPG total electricity export feedstock

fuel steam cracking engine oil waxes LAB production merchant LPG

45.6 22.7 8.3 6.9 12.7 3.8 100.0 2.6

456 000 227 000 83 000 69 000 127 000 38 000 1 000 000 26 070 1 587 302

a

fuels

coproduct (J/MJ fuel)a

carbon share (gCO2e/MJ)b

497 807 182 018 151 316 278 509 83 333

32.29 15.78 6.28 4.88 8.83 2.45 70.51

456 000

456 000 57 172 3 480 925

87.53

b

Coproduct displacement credit is assigned in proportion to Si/(SD + SAviation + SNfuel). Carbon in products is determined from carbon content and heating value Si × Ci/LHVi × 44.0/12.0. Carbon efficiency = sum of carbon shares/feedstock carbon share = 70.5/87.6 = 80.6%.

vehicle fuel economy with Group III+ lubricating oils without any change in oil drain interval.38 The life-cycle impact of petroleum wax production is calculated using a crude oil to petroleum fuel pathway in GREET with a refinery thermal efficiency of 83%. The GHG emission intensity for lubricant base oils and waxes are consistent with values reported in the literature.39

emissions are allocated by energy, so the upstream and process emissions for GTL diesel correspond to (eq 11): E FCD + E FTD = (E FC + E FTP)/(SD + Saviation + S Nfuel) (11)

The displacement credit for coproducts are calculated based on the functional unit of one MJ of fuel. The displacement credit (Di) for each coproduct (i) proportional becomes (eq 12); Di = Si /(SD + Saviation + S Nfuel) × Ei

■

LIFE-CYCLE GHG MODEL The analysis presented herein uses LCI data from CA_GREET (version 1.8b) in a separate spreadsheet workbook to calculate disaggregated fuel cycle emissions in gCO2e/MJ fuel produced. LCI data from CA_GREET for natural gas, electric power, and other energy carriers are combined with inputs presented herein to calculate the life-cycle GHG emissions for a range of GTL scenarios. Data arrays are assembled in an off-model matrix format that uses the specific energy process step or coproduct to calculate emissions for each step of the fuel pathway. GHG emissions are summed using the same energy accounting system as used in the GREET model and these results provide exact agreement with the CA_GREET model for reference cases (see SI). This model is configured with an interface input sheet that allows the user to modify the GTL product slate. In addition, the model contains a number of regional-based built-in GTL scenarios (U.S, Middle East, and Global GTL) with different GTL coproduct compositions and includes scenarios for electricity export, Carbon Capture and Storage (CCS) and use of otherwise flared gas (see SI for more information). Disaggregated GHG emissions are calculated following the GREET formulas for each feedstock scenario. These results are then distributed to GTL fuels and coproduct credits and are calculated by substitution. A copy of this model is available for download in SI.

(12)

where Ei represent the life-cycle GHG emissions for each coproduct, i. The calculation approach for each GTL nonfuel coproduct depends on the upstream LCI data for the displaced product as well as downstream use phase effects, such that the functional output of the system boundary remains the same. The energy impacts of the use phase for all of the GTL products, except LPG differs from that of conventional, petroleum-derived analogues. Both GTL naphtha and petroleum-derived naphtha are feedstocks for steam crackers, which produce olefins for a variety of chemical applications. The displacement credit for GTL naphtha is based on the relative olefin yields relative to a given amount of petroleum-derived naphtha feedstock.9 The life-cycle impact of LAB production from normal paraffin is represented by the energy difference between petroleumderived LAB production and GTL-based LAB production per unit of paraffin feed.11 The energy inputs associated with the production of LAB from petroleum-derived kerosene are based on existing industrial LAB operating emissions data. The GTL process can also be configured to produce high quality Group III+ lubricant base oils. In the U.S., these lubricant base oils are expected to displace petroleum-derived Group II lubricant base oils.10 The displacement credit for GTL lubricant base oils depends upon the displaced quantity of petroleum base oil, the relative energy burdens of each pathway and downstream effects, such as vehicle fuel efficiency and lubricant base oil recovery and recycling. In the current study, the credit for GTL lubricant base oils is calculated from the lubricant base oil used over the life of the passenger vehicle.37 A number of variables in this analysis include vehicle fuel use, lubricant base oil recycling and potentially increased oil drain intervals relative to displaced conventional, petroleum-derived Group II lubricant base oils. The coproduct impacts are based on the net change in petroleum-derived lubricant base oil, recycling inputs, and vehicle fuel savings per kg of GTL lubricant base oil. The baseline analysis presented herein reflects a 0.85% increase in

■

RESULTS Table 1 shows the accounting of carbon for the U.S. GTL product mix. The total emissions of GTL products (gCO2/MJ) are shown in the right column along with the carbon in the natural gas feedstock. A notable feature of this scenario is the diversification of the GTL product slate to include GTL paraffin, GTL waxes, and GTL lubricant base oils in addition to GTL diesel, GTL naphtha, and LPG. In this example, these three nonfuel coproducts account for 28% of the total GTL product energy share. 4376

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology Table 2. Fuel Cycle Greenhouse Gas (GHG) Emissions for Different GTL Pathways U.S. GTL pathway step flared gas credit, EFCFL upstream, ENG NG processing in GTL phase, ENGF production phase, EFTP fuel transport fuel combustion GTL coproducts (Si x Ei) GTL naphtha for steam cracker GTL lubricant base oil GTL wax GTL paraffin for LAB production LPG electricity for export GTL diesel for refinery blending total (gCO2/MJ)

U.S. GTL + power export

middle east GTL

global GTL

diesel/naphtha GTL

diesel/naphtha GTL: U.S. flared gas

CARB diesel

16.81 9.87

16.81 9.87

10.42 6.11

10.88 6.39

8.14 4.78

−89.74 6.90 4.05

10.25

37.32 0.72 71.49

37.32 0.72 71.49

28.60 1.22 71.49

30.11 1.22 71.49

22.93 1.22 71.49

22.93 0.72 71.49

11.41 1.47 74.74

−13.51

−13.51

−10.65

−10.98

−8.00

−8.00

−24.60 −3.62 −11.47

−24.60 −3.62 −11.47

−26.75 0.00 −2.67

−22.40 −0.98 −4.62

0.00 0.00 0.00

0.00 0.00 0.00

−0.91 0.00 −2.73

−0.91 −7.66 −2.73

−0.56 0.00 −2.00

−0.64 −2.05 −2.46

−0.42 0.00 −4.49

0.00 0.00 −4.49

79.4

71.7

75.2

76.0

95.7

3.4

98.03

Figure 2. Waterfall plot of the life-cycle greenhouse gas (GHG) emissions (USGTL electricity export case).

Table 2 shows the life-cycle GHG emissions results for a range of regional, product mix and feedstock resource scenarios for the GTL pathway compared with petroleum-derived conventional diesel. For conventional natural gas resources, the life-cycle GHG emissions of GTL diesel range from 71.7 to 95.7 gCO2e/MJ. The life-cycle GHG emissions of GTL diesel are lowest for product mixes with the highest fractions of GTL lubricant base oil and GTL paraffin for LAB production. The natural gas production emissions vary both with resource type as well as the fraction of GTL fuel. Note that in the case where

GTL diesel and GTL naphtha are the predominant products from the GTL process, 75% of the product slate (by energy) is fuel, compared to only 45.6% for the U.S. GTL case (see SI). This gives rise to proportionally lower natural gas processing emissions in the former case because more emissions are allocated to fuel. In a similar fashion, the GTL processing emissions are also lower for the GTL diesel/GTL naphtha case relative to the U.S. GTL case. In addition, compared to the U.S. GTL case, the coproduct impacts are lower for the GTL diesel/ GTL naphtha case, since the only coproduct is GTL naphtha 4377

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology for steam cracking. These results also suggest that the coproduction of GTL lubricant base oils is the most significant coproduct credit due in the most part to the impact of an incremental increase in passenger vehicle fuel economy. The net GHG emissions are comparable among all of the GTL options with a broad mix of nonfuel coproducts. Generation of excess electricity can also provide a significant GHG emissions credit, particularly in situations where distribution and transport losses are avoided. GTL diesel transport and distribution emissions appear to correspond to only a relatively small fraction of the total life-cycle GHG emissions value and do not affect the relative ranking among scenarios significantly. Scenarios that include CCS have the capacity to further lower the life-cycle GHG emissions of the GTL pathway, although it is worthy to note that the magnitude of GTL coproduct credits can potentially be larger than any benefit associated with CCS alone. GTL pathway scenarios that utilize avoided flared gas result in low life-cycle GHG emissions values. Using diverted flared gas as a feedstock results in a significant displacement credit that is similar in magnitude to the combined GHG emissions of CO2 from the vehicle plus the GTL facility. The relevancy of utilizing GTL as a scalable flared gas mitigation option is supported by Escravos GTL in Nigeria,40 which converts around 325 million cubic feet of natural gas per day into 33 000 bbl of GTL naphtha and GTL diesel. In principle, smaller scale GTL facilities could also capture gas from offshore deep-water−oil wells and shale plays, while targeting zero flare emissions and making economic stranded small gas fields. The impact of integrating GTL coproducts into the GREET GTL pathway is illustrated by Figure 2. In this case, (USGTL with electricity export) GTL coproduct credits offset much of the GTL production phase emissions. Notably, in this instance, the total life-cycle GHG emissions (71.7 gCO2e/MJ) for GTL diesel are 37% lower than CARB diesel (98.0 gCO2/MJ).41 Compared to previous life-cycle GHG emissions estimates of the GTL pathway (see SI),25 the substitution method described herein results in higher emissions for the feedstock and production phase, which are offset by coproduct displacement credits.

Several areas suggest future work. Utilization of substitution methodology often includes market-related analysis of the nature in which the coproduct displaces a conventional product. These market-related assumptions can be regional in nature and could have a large impact on the magnitude of a coproduct credit. For example, in the current study a North American based market displacement analysis was applied to the penetration of GTL lubricating base oils in the U.S. Specifically, in this study it was conservatively assumed that Group III+ GTL lubricant base oils would displace Group II lubricant base oils in the U.S., affording an average 0.85% fuel economy gain in passenger vehicles that drive the higher quality lubricant base oil. In different regions these assumptions may not be valid. For example, if a GTL lubricant oil were to replace a Group I lubricant base oil (currently used in significant quantity in China), then the fuel efficiency gains could be 3% or higher.10 In addition, although this study draws upon extensive existing coproduct analysis, further work to assess the benefit of GTL coproducts is still required. For example, the environmental benefits, including downstream GHG emissions benefits, of GTL waxes are currently being investigated in detail, and these results will be published shortly. Environmental management and GHG emissions reduction need to be exercised in order for alternative fuels such as GTL diesel to be produced sustainably. This analysis provides some insight on the importance of accounting for coproducts in alternative fuelcycle analysis. The calculation approach described herein represents a case study where a complex alternative fuel coproduct slate has been completely integrated into the existing framework of the GREET model, which is used for policy analysis. In addition to adding to the coproduct analysis functionality within GREET, this model could be used by regulators and stakeholders for GTL pathway development. As alternative transportation fuel pathways continue to be critically analyzed to determine their GHG emission impacts, the coproduct integration approach described herein could potentially be incorporated into new versions of the GREET model to enable users to better understand the potential environmental benefits of these pathways.

■

■

DISCUSSION These results show that by careful and systematic integration of coproduct analysis into the GREET model, alternative fuel pathways such as GTL can be more accurately modeled using the substitution method. The results presented herein suggest that the coproduction of nonfuel products in alternative fuel facilities can potentially positively impact the life-cycle GHG emissions of the particular fuel, resulting in environmental impacts that can be significantly different from conventional, petroleum-derived fuels. By using the substitution method, the composition of the product slate of the GTL pathway appears to have an influence on the final life-cycle GHG emissions of GTL diesel. In broad terms, the life-cycle GHG emissions of GTL diesel are lower in cases when the product slate is diversified to produce high quality chemical products that displace conventional, petroleum-derived analogues. Importantly, this implies that the environmental advantages of FT fuels, including GTL, could potentially be optimized with changes to the composition of the GTL product slate. Interestingly, FT technology, such as GTL, offers a level of flexibility in terms of production slate tuning that is not possible for other alternative fuel pathways.

ASSOCIATED CONTENT

* Supporting Information S

Data and details of methods, coproduct allocation assumptions, analyses, and results. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +1 288 588 3445; fax: 1 281 920 7123; e-mail: grant. [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Annual Energy Outlook 2014 Early Release Overview; Report Number: DOE/EIA-0383ER; United States Energy Information Agency: Washington, DC, 2014; http://www.eia.gov/forecasts/aeo/ er/early_elecgen.cfm?src=Renewable-b1 (last accessed February 2015). (2) Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis; United States Environmental Protection Agency; Report Number: EPA-420-R-10-006; Washington, DC, 2010; http://www.

4378

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology epa.gov/otaq/renewablefuels/420r10006.pdf (last accessed February 2015). (3) A Low Carbon Fuel Standard for California, Part 1: Technical Analysis; California Air Resources Board: Sacramento, CA. 2007; http://www.arb.ca.gov/fuels/lcfs/lcfs_uc_p1.pdf (last accessed February 2015). (4) Burnham, A.; Han, J.; Clark, C. E.; Wang, M.; Dunn, J. B.; PalouRivera, I. Life-cycle greenhouse gas emissions of shale gas, natural gas, coal, and petroleum. Environ. Sci. Technol. 2012, 46, 619−627. (5) Alternative Fuels Data Center, U.S. Department of Energy, Energy Efficiency & Renewable Energy, http://www.afdc.energy.gov/ (last accessed February 2015). (6) Wang, M.; Han, J.; Dunn, J. B.; Cai, H.; Elgowainy, A. Well-towheels energy use and greenhouse gas emissions of ethanol from corn, sugarcane and cellulosic biomass for US use. Environ. Res. Lett. 2012, 7, 045905. (7) Wang, M.; Huo, H.; Arora, S. Methods of dealing with coproducts of biofuels in life-cycle analysis and consequent results in the U.S. context. Energy Policy 2010, 39 (10), 726−736. (8) Batan, L.; Quinn, J.; Willson, B.; Bradley, T. Net energy and greenhouse gas emission evaluation of biodiesel derived from microalgae. Environ. Sci. Technol. 2010, 44, 7975−7980. (9) Dancuart, L.; Mayer, J.; Tallman, M. J.; Adams, J. Performance of the sasol SPD naphtha as steam cracking feedstock. Pet. Chem. Div. Preprints 2003, 48 (2), 132. (10) Forman, G. S.; Hahn, T. E.; Jensen, S. C. Greenhouse gas emission evaluation of the GTL pathway. Environ. Sci. Technol. 2011, 45, 9084−9092. (11) Forman, G. S.; Hahn; Hauser, A. B.; Adda, S. M. Life-cycle analysis of gas-to-liquids (GTL) derived linear alkyl benzene. J. Clean Prod. 2014, 80, 30−37. (12) Tsujimura, T.; Goto, S.; Matsubara, H. A study on PM emission characteristics of diesel vehicle fueled with GTL. Proc. J. SAE Annu. Congr. 2006, 54 (06), 1−6. (13) California Air Resources Board (CARB), Monthly LCFS Credit Trading Activity Reports, http://www.arb.ca.gov/fuels/lcfs/ lrtmonthlycreditreports.htm (last accessed February 2015). (14) El-Houjeiri, H. M.; Brandt, A. M.; Duffy, J. E. Open-source LCA tool for estimating greenhouse gas emissions from crude oil production using field characteristics. Environ. Sci. Technol. 2013, 47, 5998−6006. (15) Elgowainy, A.; Han, J.; Cai, H.; Wang, M.; Forman, G. S.; DiVita, V. B. Energy efficiency and greenhouse gas emissions intensity of petroleum products at U.S. refineries. Environ. Sci. Technol. 2014, 48 (13), 7612−7624. (16) Forman, G. S.; Divita, V. B.; Han, J.; Cai, H.; Elgowainy, A.; Wang, M. Q. U.S. Refinery efficiency: Impacts analysis and implications for fuel carbon policy implementation. Environ. Sci. Technol. 2014, 48 (13), 7625−7633. (17) Ekvall, T.; Finnveden, G. Allocation in ISO 1404A critical review. J. Clean Prod. 2001, 9 (3), 197−208. (18) Curran, M. A. Studying the effect on system preference by varying coproduct allocation in creating life-cycle inventory. Environ. Sci. Technol. 2007, 41, 7145−7151. (19) Rosen, M. A. Allocating carbon dioxide emissions from cogeneration systems: Descriptions Of selected output-based methods. J.Clean Prod. 2008, 16 (2), 171−177. (20) Weidema, B. Avoiding coproduct allocation in life-cycle assessment. J. Ind. Ecol. 2000, 4 (3), 11−33. (21) Life-Cycle Assessment: Requirements and Guidelines; International Organization for Standardization; ISO 14044 First ed. 2006−07−01 Environmental management; International Organization for Standardization: Geneva, Switzerland, 2006; http://www.iso.org/iso/home/ store/catalogue_tc/catalogue_detail.htm?csnumber=38498. (last accessed February 2015). (22) Klöpffer, W.; Rebitzer, G. A report on LCA activities at the SETACThird World Congress, 21−25 May 2000 in Brighton, U.K. Int. J. Life Cycle Ass. 2000, 5 (4), 249.

(23) Stratton, R. W.; Wong, H. M.; Hileman, J. I. Quantifying variability in life-cycle greenhouse gas inventories of alternative middle distillate transportation fuels. Environ. Sci. Technol. 2011, 10, 4637− 4644. (24) Kim, S.; Dale, B. E. Allocation procedure in ethanol production system from corn grain. Int. J. LCA 2002, 1−7. (25) (a) Norton, P.; Vertin K.; Bailey, B. (National Renewable Energy Lab, NREL), Clark, N. N.; Lyons, D. W. (Wed Virginia Univ., WVU), Goguen, S.; Eberhardt, J. (U.S. Department of Energy, DOE). Emissions from Trucks using Fischer−Tropsch Diesel Fuel International Fall Fuels and Lubricants Meeting and Exposition San Francisco, California, October 19−22,1998 (SAE Technical Paper Series 982526). (b) Greene, D. L. (Oak Ridge National Laboratory), An Assessment of Energy and Environmental Issues Related to the Use of Gasto-Liquids Fuels in Transportation, Center for Transportation Analysis, November 1999. (c) Bagajewicz, M.; Sujo, D.; Martinez, D.; Savelski, M. Driving Without Petroleum?A Comparative Guide to Biofuels, Gasto-Liquids and Coal-to-Liquids as Fuels for Transportation, May 2007. (c) Johnson, E. (SRI Consulting) Carbon Footprints of Biofuels & Petrofuels, October 2007. (d) Jaramillo, P.; Griffin, M.; Matthews, S. Comparative Analysis of the Production Costs and Life-Cycle GHG Emissions of FT Liquid Fuels from Coal and Natural Gas. Environ. Sci. Technol. 2008, 42 (20), 7559−7565. (e) Oak Ridge National Laboratory for US DOE, An assessment of Energy and Environmental Issues Related to the Use of Gas-To-Liquid Fuels in Transportation, 1999. (f) Kavalov, B. Techno-economic analysis of Natural Gas application as an energy source for road transport in the EU, Joint Research Council, January 2004. (g) Edwards R. Well-to-wheels analysis of future automotive fuels and powertrains in the European context, Concawe, Eucar, European Commissions, well-to-wheels report, version 2a, December 2005. (h) Beer, T.; Grant, T.; Morgan, G.; Lapszewicz, J.; Anyon, P.; Edwards, P.; Nelson, P.; Watson, H.; Williams, D. (Australian Greenhouse Office, CSIRO) Comparison of Transport Fuels -Life-cycle Emissions Analysis of Alternative Fuels for Heavy Vehicles, 2006. (i) Marano, J. J.; Ciferno, J. P. (DOE) Life-Cycle Greenhouse-Gas Emissions Inventory for Fischer−Tropsch Fuels, June 2001. (j) General Motors Well-to-Wheel Energy Use and Greenhouse Gas Emissions of Advanced Fuel/Vehicle Systems − North American Analysis, June 2001. (k) Wang, M.; Wu, M.; Huo, H. (Center for Transportation Research, Argonne National Laboratory) Life-Cycle Energy and Greenhouse Gas Results of Fischer−Tropsch Diesel Produced from Natural Gas, Coal, and Biomass, May 2007. (l) Brandt, A. R.; Farrell, A. E. Scraping the bottom of the barrel: greenhouse gas emission consequences of a transition to low-quality and synthetic petroleum resources. Clim. Change 2007, 84, 241−263. (m) Van Vliet, O. P. R.; Faaij, A. P. C.; Turkenburg, W. C. Fischer−Tropsch diesel production in a well-to-wheel perspective: A carbon, energy flow and cost analysis. Energy Convers. Manage. 2009, 50, 855−876. (n) Stratton, R. W.; Wong, H. M.; Hileman, J. I. Quantifying variability in life cycle greenhouse gas inventories of alternative middle distillate transportation fuels. Environ. Sci. Technol. 2011, 10, 4637−4644. (o) Hao, H.; Wang, H.; Song, L.; Li, X.; Ouyang, M. Energy consumption and GHG emissions of GTL fuel by LCA: Results from eight demonstration transit buses in Beijing. Applied Energy 2010, 87, 3212−3217. (p) Bengtsson, S.; Andersson, K.; Fridell, E. A comparative life cycle assessment of marine fuels: Liquefied natural gas and three other fossil fuels. Proc. IMechE 2011, 225 (2), 97−11. (26) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model. http://greet.es.anl.gov/. last accessed February 2015). (27) Analysis of Natural Gas-to Liquid Transportation Fuels via Fischer−Tropsch; National Energy Technology Laboratory (NETL), September 13, 2013, DOE/NETL-2013/1597. (28) California GREET model. http://www.arb.ca.gov/fuels/lcfs/ software.htm (last accessed February 2015). (29) Fourth Assessment Report: Climate Change 2007, Intergovernmental Panel on Climate Change: Geneva, Switzerland, 2007; http://www.ipcc.ch/publications_and_data/publications_and_data_ reports.htm#1 (last accessed July 2014). 4379

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380

Article

Environmental Science & Technology (30) Methane emissions in GREET are not counted toward the fully oxidized carbon as CO2. The secondary conversion of methane to CO2 in the atmosphere is intended to be reflected in the GWP of methane. (31) Lee, D. S.; Pitari, G.; Grewe, V.; Gierens, K.; Penner, J. E.; Petzold, A.; Prather, M. J.; Schumann, U.; Bais, A.; Berntsen, T.; Iachetti, D.; Lim, R.; Sausen, L. L. Transport impacts on atmosphere and climate. J. Atmos. Environ. 2010, 44 (37), 4678−4734. (32) GREET 1.5: Transportation Fuel Cycle Model. Methodology, Development, Use, and Results; Argonne National Laboratory Report ANL/ESD-39, Chicago, IL, 1999. (33) Full Fuel Cycle Assessment Well to Tank Energy Inputs, Emissions and Water Impacts, California Energy Commission Report: CEC-6002007-002-D; Sacramento, CA. 2007. (34) Updated Fugitive Greenhouse Gas Emissions for Natural Gas Pathways in the GREET Model; Argonne National Laboratory: Chicago, IL, 2013. (35) Life-Cycle Greenhouse Gas Inventory of Natural Gas Extraction, Delivery and Electricity Production; U.S. Department of Energy: Washington, DC, 2011. (36) Howarth, R. W.; Santoro, R.; Ingraffea, A. Methane and the greenhouse-gas footprint of natural gas from shale formations. Clim. Change 2011, 106, 679−690. (37) EMFAC mobile emissions vehicle model Website; http://www. arb.ca.gov/msei/modeling.htm. (38) The Effect of OW-20 Low Viscosity Engine oil on Fuel Economy, SAE Series Technical Paper Series, 1999−01−3468; Honda Research and Development Co., Ltd: Warrendale, PA. 1999. (39) Boughton, B.; Horvath, A. Environmental assessment of used oil management methods. Environ. Sci. Technol. 2004, 38 (2), 353−358. (40) http://www.chevron.com/deliveringenergy/gastoliquids/ (last accessed February 2015). (41) Detailed California-Modified GREET Pathway for Ultra Low Sulfur Diesel (ULSD) from Average Crude Refined in California; California Environmental Protection Agency Air Resources Board Stationary Source Division, September 12, 2012. http://www.arb.ca. gov/regact/2011/lcfs2011/ulsd.pdf (last accessed February 2015).

4380

DOI: 10.1021/es505994w Environ. Sci. Technol. 2015, 49, 4372−4380