COPTEM: A Model to Investigate the Factors Driving Crude Oil

Subscriber access provided by READING UNIV

Article

COPTEM: A Model to Investigate the Factors Driving Crude Oil Pipeline Transportation Emissions Nicolas Choquette-Levy, Margaret Zhong, Heather L MacLean, and Joule A. Bergerson Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b03398 • Publication Date (Web): 22 Nov 2017 Downloaded from http://pubs.acs.org on November 27, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Environmental Science & Technology

COPTEM: A Model to Investigate the Factors Driving Crude Oil Pipeline Transportation Emissions Nicolas Choquette-Levy†§, Margaret Zhong‡, Heather MacLean‡, Joule Bergerson†* † Department of Chemical and Petroleum Engineering, Schulich School of Engineering University of Calgary, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4 § Present address: PhD Student, Science, Technology and Environmental Policy, Robertson Hall, Princeton University, Princeton, NJ, USA 08544-1013. ‡ Department of Civil Engineering, University of Toronto, 35 St. George Street, Toronto, Ontario, Canada M5S 1A4 *Corresponding author: [email protected] Abstract Previous transportation fuel life cycle assessment studies have not fully accounted for the full variability in the crude oil transport stage – e.g. transporting a light crude through a highdiameter pipeline, vs. transporting a heavy crude through a small-diameter pipeline. We develop a first principles, fluid mechanics-based crude oil pipeline transportation emissions model (COPTEM) that calculates the greenhouse gas (GHG) emissions associated with pipeline transport as a function of crude oil parameters, pipeline dimensions, and external factors. Additionally, we estimate the emissions associated with the full life cycle of pipeline construction, maintenance, and disposal. This model is applied to an inventory of 62 major Canadian and U.S. pipelines (capacity greater than 100,000 barrels/day) to estimate the variability of GHG emissions associated with pipeline transportation. We demonstrate that pipeline GHG emissions intensities range from 0.23 – 20.3 g CO2e/(bbl*km), exhibiting considerably greater variability than data reported in other studies. A sensitivity analysis demonstrates that the linear velocity of crude transport and pipeline diameter are the most impactful parameters driving this variability. To illustrate one example of how COPTEM can be used, we develop an energy efficiency gap analysis to investigate the possibilities for more efficient pipeline transport of crude oil. 1. Introduction Crude oil pipelines in Canada and the United States comprise an essential part of the energy system, connecting petroleum resources with industrial and consumer demand. As of 2016, there were 37,400 km of crude oil transmission pipelines (i.e., those that connect petroleum resources to refineries) in Canada, and 117,000 km in the United States (1, 2). The rapid development of unconventional sources of crude oil in these countries has generated considerable debate, including questions about the future expansion of this pipeline network. Despite the increasing emphasis in regulatory settings on accurately quantifying the life cycle (full supply chain) GHG emissions intensity of transportation fuels (e.g., California’s Low Carbon Fuel Standard), the contribution of crude oil transportation by pipeline has received little attention. As these pipelines typically employ electric pumps, the transport of crude oil results in GHG emissions associated with electricity generation, transmission and distribution. This activity

ACS Paragon Plus Environment

1


48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

Page 2 of 20

has been characterized in a simplified manner in recent life cycle assessments (LCAs) (3-5) and government models. For example, Argonne National Laboratory’s Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model (6) accounts for this stage through a single greenhouse gas (GHG) emissions intensity for each mode of transportation (e.g., pipeline, rail, or ship) that is linearly scaled by the distance crude oil is transported. All of these studies estimate that GHG emissions resulting from pipeline transport of crude are only a few percent of the overall life cycle emissions (roughly 1 g CO2 equivalent (e)/MJ gasoline). This is essentially an insignificant contributor to overall life cycle (fuel production and use) emissions, which are reported to be on the order of 80-120 g CO2e/MJ gasoline (5). However, this point estimate does not differentiate based on crude oil properties, such as density and viscosity, which can vary widely (e.g., from 835 kg/m3 and 8 cSt or lower for light crude oils, to 940 kg/m3 and 350 cSt for heavy oil like oil sands-derived diluted bitumen, or dilbit). They also do not account for the wide range of dimensions (e.g., diameter and flow rate) across the Canadian and US pipeline network. Pipeline-related emissions are included in national GHG inventories (e.g., 7, 8); however, these are presented at an aggregate level and don’t provide the detail required for analyzing the effects of variations in crude oil or pipeline characteristics. Much of the recent literature on pipelines relates to specific technical elements of a pipeline system. For example, a subset of the civil engineering literature has focused on asset management of municipal water pipelines, including the identification and rehabilitation of pipeline breaks (9-12). Other studies focused on factors influencing pipeline design, including fluid viscosity, heat transfer, deposition of wax sediments, and optimizing pipeline diameters (13-15). A limited set of studies has evaluated the energy requirements and GHG emissions of transporting different substances including water, bio-oil and CO2 in pipelines (16-19). These analyses did not examine crude oil or account for the effects of variability in viscosity and density in the fluids they analyzed. Strogen (20) modeled the pipeline transport of liquid biofuels, evaluating the impact of input parameters such as ambient temperature, velocity, and diameter on energy intensity and GHG emissions. Two recent studies have begun examining the factors that drive variability in pipeline transport GHG emissions. Tarnoczi (21) conducted a life cycle-based study of selected crude oil pipeline and rail transportation routes in Canada and the United States, concluding that previous studies may have underestimated GHG emission intensities. While Tarnoczi’s model allows for inputs of different crude oil characteristics (e.g., density and kinematic viscosity) in discrete transportation pathways, the study did not analyze the impact of variation of those parameters on transport GHG emissions. As well, Tarnoczi stated that first principles fluid mechanics equations were used to calculate pipeline electricity consumption, but no formal method was presented. Nimana, B et al. (22) also compared the GHG emissions intensities of pipeline versus rail transportation of crude oil, exploring seven distinct scenarios of transporting Alberta’s bitumen resources. This study contributed a first-principles model to calculate the GHG emissions intensity of transportation through a hypothetical 3,000 km pipeline, and also explored the relative effects of changes in pipeline and crude parameters through an uncertainty analysis. However, the scope of their analysis was limited to bitumen-based crude transported from Alberta to Texas; it did not explore the full range of major Canadian and U.S. pipelines, crude oil, and their GHG emissions intensities. To our knowledge, no prior study has analyzed a large set of parameters that can affect crude oil pipeline transport GHG emissions. Consequently, existing studies are likely to have


2

Page 3 of 20

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142


underestimated the variability in GHG emissions intensity, and in turn, its potential contribution to overall life cycle emissions. To address this gap, we develop a novel model – the Crude Oil Pipeline Transportation Emissions Model, or COPTEM – that is generalizable and can be used to calculate the GHG emissions intensities of major pipelines and crudes transported in North America. COPTEM calculates the GHG emissions intensity of crude oil pipeline transport based on a broad set of parameters including: pipeline dimensions, crude oil parameters, and other external factors (e.g., elevation change, electricity grid GHG intensity); and includes both operating and non-use (pipeline materials, construction and end-of-life) components. We then apply COPTEM to an inventory of Canadian and U.S. crude oil parameters and pipeline data, in order to explore the range of crude transport GHG emissions intensities, and the current magnitude of crude pipeline transportation GHG emissions across Canada and the United States. This analysis allows us to identify the most significant parameters impacting pipeline GHG emissions, and to develop a gap analysis to inform decisions about future pipeline investments and fuel transport GHG policies, e.g. the California Low Carbon Fuel Standard. 2. Methods This section presents the methods utilized to develop COPTEM, a fluid mechanics-based, first principles pipeline transportation model. It then details the application of COPTEM to an inventory of major Canadian and U.S. crude oil pipelines, in order to assess aggregate transport GHG emissions using a set of existing infrastructure data, and to explore the variability in GHG emissions intensities throughout the network. We then describe a parametric analysis conducted to identify the most significant factors influencing the emissions intensity of pipeline transport, and explore the potential for more efficient transportation of crude oil through an energy efficiency gap analysis. 2.1 COPTEM Development 2.1.1 Crude Oil Transport Emissions COPTEM consists of a set of fundamental fluid mechanics equations. These equations are utilized to estimate the energy requirements and GHG emissions associated with transporting different qualities of crude oil through major pipelines under varying characteristics and external factors. The primary focus of the model is to estimate the head loss as crude oil is transported along the pipeline, and calculate the energy required by the pumping system to compensate for these losses. Model inputs are organized into three broad categories: crude oil properties, pipeline dimensions, and external factors. Crude oil properties include density, viscosity, and the crude’s inlet temperature; pipeline dimensions include inner diameter, roughness factor, flow rate, and pump efficiencies; and external factors include the elevation change, grid emissions factor, and ambient temperature across the pipeline’s length. A schematic of the parameters in the model is shown in Figure 1. Model outputs include the energy required to power pump stations to maintain sufficient hydraulic head, and the resulting GHG emissions intensity after applying a grid emissions factor.


3


143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168

Page 4 of 20

Figure 1. Schematic overview of COPTEM’s operating stage activities. Parameters in red represent crude oil properties, parameters in blue represent pipeline dimensions, and parameters in yellow represent external factors. Grid EF = electricity grid GHG emissions factor. The energy required for pipelines to maintain hydraulic head is estimated by dividing the pipeline into 40 representative pipeline segments ranging from 25 – 75 km in length (depending on the pipeline’s total length), and modeling the impact of parameters along each segment, approximating the portions of a pipeline between pump stations. This structure allows for a generalizable model that can be applied to actual or prospective pipelines of various lengths, though it may not be reflective of the exact locations of pump stations for any single pipeline. However, the lengths of individual pipeline segments in this model are either equivalent to or smaller than the distance between actual pump stations of most pipelines (23). As a result, COPTEM allows for an equivalent- or higher-resolution analysis of head loss than constructing a separate model for each pipeline based on the actual lengths of its individual segments. The effects of parameters that change along a pipeline’s length (e.g. crude temperature and viscosity, elevation, and grid emissions factor) are calculated based on first principles. For example, while the ambient temperature along the pipeline’s length is assumed to change linearly, the effect of temperature change on crude viscosity is non-linear and therefore modeled using the Lumped Heat Capacity Method for each segment (see Supporting Information for more detail). Data on average annual ambient temperatures from pipelines’ inlets and outlets were obtained online from The Weather Network (24) for Canadian cities, and U.S. Climate Data (25) for American cities. The overall effect of elevation change on hydraulic head remains constant regardless of the elevation profile from Point A to Point B along the pipeline; the effect of this parameter is therefore captured by assuming a linear elevation


4

Page 5 of 20

169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214


change. The grid emissions factor is assumed to change discontinuously across state and provincial boundaries, so an average grid emissions factor for the entire pipeline is calculated using state/provincial emissions factors, weighted by the length of a pipeline in each jurisdiction (7, 26). Note that at the scale of an individual pipeline segment, the assumptions regarding elevation change and grid emissions factors represent an imperfect approximation. For example, they could underrepresent the emissions for a segment that rises in elevation in a jurisdiction with a particularly GHG-intensive electricity grid. However, this imprecision is mitigated by two factors: (1) these deviations tend to regress to the mean along an entire transmission pipeline, and (2) head losses due to elevation change are generally at least one order of magnitude less than head losses due to friction.

The overall equation used for calculating GHG emissions related to pipeline transportation is: =

∑ ∗ ∗ ∗∗

[1]

Where: GHGs represent the GHG emissions associated with transporting one barrel of crude oil over the entire pipeline distance (kg CO2e/bbl crude); Δ!" represents the pressure drop across an individual pipeline segment i (kPa/segment). Note that this may vary from one pipeline segment to the next, based on changes in ambient and crude temperatures and their effects on crude viscosity (see Equation 2 below); q represents the flow rate of crude oil (m3/s). This is assumed to be constant throughout the pipeline; # represents the overall efficiency of the pump system at each pump station (derived from several individual parameters, including the pump and motor efficiencies, brake horsepower, coupling efficiency, and motor service factor). While COPTEM allows for different efficiencies to be specified for different pump stations, they are assumed to be constant across the entire pipeline for this analysis; EF represents the emissions factor of the average grid electricity mix that services the pipeline’s pump stations (kg CO2e/kWh); Q represents the flow rate of crude oil through the pipeline (bbl/s).

The friction losses associated with the transportation of oil are calculated via the DarcyWeisbach equation, in a similar manner to the model described in (22): '

Δ!$," = &" ∗ ( ∗ )" ∗

*+ ,

∗ 0.001

01 01

[2]

where Δ!$," represents the pressure loss due to friction along pipeline segment i (kiloPascals, or kPa, where 1 kPa = 1000 kg/(m*s2) ); fi represents the friction factor for fluid transport along segment i (dimensionless); L represents the length of a segment (m); D represents the inner diameter of a segment (m);


5


215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261

Page 6 of 20

)" represents the density of the fluid being transported along segment i (kg/m3); v represents the linear velocity of the transported fluid (m/s). The friction factor fi is calculated through the Colebrook-White Equations and is dependent on crude properties e.g. density ()" ) and viscosity (μi) that may change over the length of the pipeline, as well as pipeline parameters e.g. roughness factor (k), diameter (D) and flow rate (Q) that are assumed to be constant throughout the pipeline. See Supporting Information (Section 1) for additional detail.

2.1.2. Pipeline Construction/Decommissioning (Non-Use Stage) Emissions The GHG emissions associated with non-use stages of the pipeline – including mining of raw materials, construction of the pipeline, maintenance, and disposal of waste products after decommissioning – are estimated using Carnegie Mellon University’s Economic Input-Output Life Cycle Assessment (EIO-LCA) tool (27). This requires a capital cost associated with pipeline construction as an input, approximated by a total construction cost factor of 4.11 million USD per km (28). The outputs of interest from the EIO-LCA tool are the life cycle GHG emissions intensities (as a function of economic activity) associated with iron ore mining, steel manufacturing, pipeline manufacturing, maintenance and repair, and waste disposal. Further details on the calculation of these non-operating stage emissions are in the Supporting Information (Section S2).

2.2 COPTEM Application We demonstrate the utility of COPTEM by applying it to conduct three different applications: 1) an assessment of the variability in GHG emissions among the Canadian-U.S. pipeline network, 2) a parametric analysis to determine the variables with the largest impact on pipeline GHG emissions, and 3) an idealized energy efficient scenario, to illustrate potential GHG emissions reduction opportunities. 2.2.1 GHG Emissions Associated with Major U.S. and Canadian Crude Oil Pipelines We apply COPTEM to estimate the range and aggregate total of transportation GHG emissions intensities across major crude pipelines in Canada and the United States. Data on pipeline dimensions and the types of crude oil transported were collected from pipeline operator websites for the 62 major pipelines. “Major” pipelines included in the analysis are defined as those that are located in the United States or Canada, with at least 100,000 barrels per day (bpd) of crude oil capacity. This threshold does not capture gathering and feeder pipelines that transport smaller quantities of crude to trunk pipelines, whose aggregate distance is likely on the order of 100,000s km (1, 2). In cases where websites did not specify the types of crude oil transported in a particular pipeline, we assumed the average U.S. or Canadian crude oil density and viscosity, depending on the pipeline’s location (29, 30). Data on external factors – including elevation change, ambient temperature profile, and grid emissions factors – were gathered based on assuming a straight


6

Page 7 of 20

262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309


line between each pipeline’s inlet and outlet point. Further information on data collection and model assumptions is available in the Supporting Information (Section S5).

2.2.2 Parametric Analysis To illustrate the effects of pipeline dimensions and crude oil parameters on pipeline transportation GHG emissions, we present a simplified scenario that compares the transportation of two different types of crude oil (a light crude with density of 835 kg/m3, viscosity of 8 cSt; and a heavy crude with density of 940 kg/m3, viscosity of 350 cSt) through two representative pipelines (a 24” inner diameter; 398,000 bpd capacity pipeline; and a 33” inner diameter, 351,000 bpd capacity pipeline). Both examples represent existing pipelines that have the same origin and destination. Thus, pipeline length, grid emissions factors, elevation change, and temperature changes are the same for both pipelines, demonstrating the impact of pipeline diameter and crude parameters. The light and heavy crudes in this example generally represent upper and lower bounds of typical crude oil densities and viscosities that are transported through major transmission pipelines in the United States and Canada. In addition to the simplified scenario, we conduct a sensitivity analysis on all crude oil, pipeline, and external parameters in the model to identify the main drivers of pipeline transportation GHG emissions. The value of each input parameter is varied across the range of values observed for that parameter among the 62 major pipelines, while all other parameters are held constant at a “base case” value, generally reflecting the median value for each parameter (see Section S3 in the Supporting Information).

2.2.3 Energy Efficiency Gap Analysis A greater understanding of variability in pipeline transportation GHG emissions has the potential to inform future infrastructure decisions, including those focused on reducing the emissions intensity of transportation fuels. As we determine that a pipeline’s linear velocity is one of the most impactful parameters in determining crude oil transportation GHG emissions intensity (see Section 3.2), we conduct a hypothetical energy efficiency gap analysis, in which this parameter is reduced for the pipelines included in the analysis. Linear velocity is defined here as the “straight-line” rate of change in the movement of crude oil within a pipeline. It is calculated by converting a pipeline’s flow rate from bbls/day to m3/s, and dividing by the pipeline’s cross-sectional area. Specifically, the flow rate of crude oil through each pipeline in the North America network is adjusted such that the product is transported at a linear velocity of 0.60 m/s – near the lowest linear velocity we calculate among existing pipeline data. Since oil producers, transporters, and refineries all have an interest in maintaining the existing volumes of crude oil, pipelines whose flow rates must be reduced in this analysis are assumed to be “twinned”, with additional pipeline(s) sized to transport the remaining volume at the same linear velocity of 0.60 m/s. As an example, if the linear velocity of the existing Enbridge Canadian Mainline 1 pipeline (18.25 in. inner diameter, 236,500 bpd capacity) is reduced to 0.60 m/s (from 2.3 m/s), the pipeline’s flow rate would be approximately 56,000 bpd. In this scenario, the remaining 157,000 bpd would be transported at the same velocity through a new 32” pipeline that mirrors the existing pipeline’s


7


310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357

Page 8 of 20

route. While it would be impractical to twin most major pipelines in Canada and the United States, this analysis provides a basic estimate of the “efficiency gap” in the system for less GHG intensive pipeline transportation (relative to a hypothetical scenario where the pipeline system could be re-built), based on this one important parameter. 3. Results 3.1 GHG Emissions of the Existing Canada-U.S. Pipeline Network We apply COPTEM to 62 major crude oil pipelines in Canada and the United States to develop a range of pipeline transportation emissions intensities across the pipeline system, and to estimate the aggregate annual GHG emissions from this sector. The full list of pipelines analyzed, along with their parameters and GHG emissions intensities, are presented in the Supporting Information (Section 5). The diameter of these pipelines ranges from 11” to 41”, and capacities range from 100,000 bpd (the minimum capacity threshold) to a maximum of 1.2 million bpd. The estimated density of petroleum products transported along these pipelines ranges from 700 kg/m3 for a few major condensate pipelines to 940 kg/m3 for pipelines dedicated to transporting oil sands-derived diluted bitumen. We estimate that the range of GHG emissions intensities for these pipelines is from 0.23 g CO2e/(bbl*km) to 20.3 g CO2e/(bbl*km), with a mean of 3.0 g CO2e/(bbl*km) (weighted by the distance and throughput of each pipeline). Expressed in other functional units, GHG emissions range from 0.066 – 17 kg CO2e/bbl with a mean of 2.8 kg CO2e/bbl, 0.47 - 0.53 g CO2e/MJ and 18 – 2000 tonnes CO2e/(km*year) among the 62 major pipelines. We note these different functional units for comparison purposes; often, LCA practitioners report GHG emissions on a kg CO2e/bbl or tonnes/(km*year) basis. Generally, use-phase emissions outweigh non-use emissions by over an order of magnitude; the weighted mean use phase emissions among the pipelines are 2.9 g CO2e/(bbl*km), whereas the weighted mean non-use emissions are 0.1 g CO2e/(bbl*km). However, for a limited number of pipelines (five of the 62 we analyzed), the contribution from the non-use phase can be of the same order of magnitude as the use phase; these are all pipelines with a low overall GHG emissions intensity that are transporting relatively light oil. Generally, the mean intensity of 3.0 g CO2e/(bbl*km) represents a relatively small portion (approximately 1-2 percent) of the well-to-wheel GHG emissions associated with most North American transport fuels; the production and refining of crude oil and combustion of transportation fuels in vehicles are more significant sources of GHG emissions. This finding is in line with previous studies. However, our analysis also demonstrates that approximately 10 percent of the pipelines we examined have GHG emissions intensities that are more than double this mean (i.e. > 6.0 g CO2e/(bbl*km)). Figure 2 indicates that previous studies (3, 4, 6, 21, 22, 31, 32) have included only a portion of this range. As such, they may be either underestimating or overestimating the contribution of this stage to transportation fuel life cycle GHG emissions, depending on the pipeline and type of crude oil associated with a particular pathway. Note that our range includes one extreme set of conditions for a pipeline at 20.1 g CO2e/(bbl*km), which represents a Canadian pipeline transporting diluent (i.e. a light hydrocarbon product that is blended with oil sands bitumen to produce dilbit). While our study focuses on crude oil transportation, we include this estimate in our range for two reasons: (1)


8

Page 9 of 20

this particular pipeline forms part of a larger pipeline system to transport oil sands crude (33), and (2) the viscosity and density of oil sands diluent (approximately 700 kg/m3 and 1 cSt) generally fall within the range of liquid petroleum products that are transported on main crude oil trunk lines, including natural gas liquids. While we have not compiled an exhaustive list of dedicated diluent pipelines, we believe this particular case is illustrative of other diluent pipelines already in operation, or that may be built to support future oil sands production. More generally, the emissions intensities of small-diameter pipelines (less than 12 in.) transporting a high volume of heavy crude oil at a high linear velocity (4 m/s or greater) would lie near the top of our range, above most maximum intensities previously reported. Conversely, large-diameter pipelines (greater than 30 in.) transporting light to medium crude at low linear velocities (less than 1 m/s) would fall towards the bottom of our range, below the minimum intensities reported in most previous studies.

Pipeline GHG Intensity [g CO2e/(bbl*km)]

358 359 360 361 362 363 364 365 366 367 368 369 370 371


25

20

Comparison of Pipeline Transportation Emissions

15 Max Min 10

Mean Outlier

5

0 This Study

372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387

Nimana Tarnoczi IHS CERA TIAX et al. 2013 2010 2009 2017

Jacobs 2009

Hooker 1980

GREET 2014

Figure 2. Range of pipeline transportation GHG emissions intensities reported in fuel life cycle studies and models. In all cases, we convert pipeline transportation GHG or energy intensities as they are reported in other studies into the units presented here, using conversion factors from the studies themselves.

A comparison of these results with those for rail transportation of crude oil demonstrates that the GHG emissions intensities for these two modes overlap significantly. An analysis by Tarnoczi (21) of five existing and prospective rail transportation routes for crude oil found that the GHG emissions intensity ranged from 3.4 – 3.9 g CO2e/(bbl*km), and are mainly a function of train engines’ combustion of diesel. Similarly, Nimana et al. (22) found a range of 2.3 – 6.2 g CO2e/(bbl*km) for rail scenarios that transported oil sands bitumen crude from Alberta to Texas. Pipelines and rail options are not perfect substitutes. However, on a per bbl*km basis, rail emissions estimates are mostly higher than the mean pipeline intensity from this study even


9


388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407

Page 10 of 20

though the rail emissions intensities fall within the range of pipeline intensities we determine. In fact, this study demonstrates that in many cases (e.g., a small-diameter pipeline and/or high linear velocity) that include both actual operating pipelines and plausible but hypothetical conditions, transporting crude oil by pipeline can be more GHG-intensive than by rail. Overall, we estimate the annual energy use of transporting crude oil through the 62 pipelines to be 21.0 PWh/year – resulting in GHG emissions of 17.1 MT CO2e. For context, this total represents 0.23 percent of Canada’s and the United States’ combined GHG emissions in 2014 (7, 8); however, it is greater than the annual emissions of seven U.S. states (34), and is equivalent to 24 percent of the annual upstream emissions associated with oil sands extraction and upgrading in 2014 (7). Although correlations exist between the pipelines’ GHG emissions intensities and model input parameters, no one single parameter explains the distribution of GHG emissions intensities. For example, five pipelines dedicated to transporting heavier diluted bitumen rank among the top half of GHG-efficient pipelines, while some pipelines carrying lighter condensate rank among the least GHG-efficient pipelines (see Figure S1 in Supporting Information).

GHG Emissions (kg CO2e/bbl crude)

20.0 18.0

0.15 17.7

16.0 14.0 12.0 10.0

Non-Use Phase

8.0 6.0

0.15 7.1

Use Phase

4.0 0.17 3.0

2.0 0.0

0.17 1.0

Small Pipeline Small Pipeline Large Pipeline Large Pipeline (25") - Light (25") - Heavy (34") - Light (34") - Heavy Crude Crude Crude Crude

408 409 410 411 412 413 414 415

Figure 3. Simplified scenario illustrating the effects of crude parameters and pipeline dimensions while holding pipeline length (1770 km), average grid emissions factor (0.756 kg CO2e/kWh), and elevation change (-470 m) constant. The flow rate for the 24-in. inner diameter pipeline is assumed to be 398,000 bpd (linear velocity of 2.5 m/s); while the flow rate for the 33-in. pipeline is assumed to be 351,000 bpd (linear velocity of 1.2 m/s), based on observed data.


10

Page 11 of 20

416 417 418 419 420 421 422 423 424 425 426 427


3.2 Parametric Analysis Figure 3 demonstrates the variability in GHG emissions for the simplified scenario comparing the transportation of crude oil through pipelines of the same length, through the same route. Emissions vary from 1.0 – 17.7 kg CO2e/bbl crude, between transporting a light crude through a relatively large pipeline and a heavy crude through a relatively small pipeline. The results in the figure present operating emissions only. The range of operating emissions is driven both by variation in crude parameters (which increases emissions by a factor of 2-3 for the pipelines in this scenario) and in pipeline dimensions (which increases emissions by a factor of 6-7 for the crude oils presented here). Non-operating emissions total 0.14 kg CO2e/bbl crude for the 25” pipeline, and 0.16 kg CO2e/bbl crude for the 34” pipeline.

200

% Change in GHG Intensity

150

-100

Density 100 Viscosity Linear Velocity 50

Roughness Factor Crude Temperature Elevation Change

0 -50

0

50

100 Length

-50

Grid EF Diameter

-100 % Change in Parameter within Observable Range

428 429 430 431 432 433 434 435 436

Figure 4. Sensitivity of pipeline transport GHG emissions to variations in input parameters. The origin of the graph represents transporting crude oil with a density of 865 kg/m3 and viscosity of 8 cSt, through a 33 in. inner-diameter pipeline with a linear velocity of 1.35 m/s (roughly equivalent to a 390,000 bpd flow rate). These generally represent the median values observed for each parameter throughout the Canadian-U.S. pipeline network. To normalize these results based on the range of variability exhibited in each parameter, the x-axis represents the percent change in a parameter within the observed variation. That is, a point located at +/-100 percent along the x-axis represents the most extreme value of that parameter compared to its median


11


437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483

Page 12 of 20

value. The two exceptions are the roughness factor and crude temperature, as data are not readily available for these parameters. A generic range of 0.7 – 1.3 times the base case value is assumed for both of these parameters (i.e. a point at the right-hand point of the x-axis for these parameters represents 1.3 times their base case values). In addition to this scenario, a single-parameter sensitivity analysis of the COPTEM results helps identify the main drivers of pipeline transport GHG emissions. Figure 4 displays the sensitivity of pipeline transportation GHG emissions to variation in nine input parameters in the model. This analysis demonstrates that the two most important drivers of pipeline transportation GHG emissions are the pipeline’s linear velocity and diameter. The effect of these parameters can be understood intuitively: forcing fluid through a given cross-sectional area at a faster rate, or decreasing that cross-sectional area with constant flow will increase the friction, and therefore more energy is required to compensate for head losses. The increased energy requirements translate to increased GHG emissions. This concept is reflected in the underlying pipeline model equations, as the Darcy-Weisbach and Colebrook-White Equations relate head 2 345 (

, (

loss to the square of linear velocity and to the factor the pipeline.

where D is the inner diameter of

Variations in several other parameters also impact pipeline transport GHG emissions. The pipeline length and grid emissions intensity has a predictably linear effect, while variations in crude density and viscosity also have a noticeable impact on GHG emissions. Note that the effect of temperature changes along a pipeline’s length, and changes to the pipeline’s roughness factor, do not appear to have a significant effect on GHG emissions compared to these other parameters, at least over the observed range of values for these parameters. The results suggest that pipeline parameters such as diameter and linear velocity offer the most important levers for reducing GHG emissions of pipeline transport, followed by crude oil and systemic parameters such as pipeline length and the grid emissions factor.

3.3. Energy Efficient Gap Analysis The annual GHG emissions resulting from transporting the same volume of crude oil through the energy efficient conditions described in Section 2.2.3, in which all major pipeline velocities are adjusted to 0.6 m/s, amount to 3.4 MT CO2e/year (including non-operating emissions) – an 81 percent reduction in GHG emissions from the Canadian-U.S. existing infrastructure. In practice, implementing this change across the entire pipeline network would be cost-prohibitive ($490/tonne CO2e - calculated using the capital costs presented in the Pipeline Construction section divided by the CO2 emissions that are avoided by the additional efficiency gains identified in section 2.2.3). For a limited set of pipelines that have small diameters (less than 24 in.), high linear velocities (above 4 m/s), and travel short distances (less than 1,000 km) – such as sections of InterPipeline’s Polaris network in Alberta – we calculate the abatement costs are in the $60-80/tonne CO2e range. In either event, we present these results not as a GHG emissions reduction option to pursue, but as an indication of the “efficiency gap” in the existing pipeline network relative to a scenario where pipelines could be re-built from scratch, based on varying just one impactful parameter.


12

Page 13 of 20

484 485 486 487 488 489 490 491 492 493 494 495 496 497 498


While this analysis demonstrates the ability to reduce the pipeline sector’s GHG emissions through reduced linear velocities, recent trends in this industry appear to point in the opposite direction. With increased North American oil production, and economic and societal challenges to the expansion of the pipeline network, several pipeline operators appear to be transporting increased product volumes through the existing infrastructure, while pipeline capacity remains relatively constant. Increases in linear velocity of major pipelines leads to an increase in the energy needed at pump stations (or an increase in pump stations) to compensate for the head loss due to friction. The results of this study suggest that one unintended consequence of this trend may be a disproportionate increase in the emissions from transportation of crude oil. As one example, a doubling of the linear velocity of a major trunk pipeline carrying heavy oil approximately 1800 km (which has been observed in at least one such pipeline) would more than triple its GHG emissions intensity and annual aggregate emissions.


13


499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546

Page 14 of 20

4. Discussion This paper aims to contribute a greater understanding of the GHG emissions associated with pipeline transportation of crude oil, by exploring the variability in emissions intensity as a function of crude oil parameters, pipeline dimensions, and external factors. Contributions to the transportation fuels LCA literature include the development of a first principles-based pipeline transportation model, COPTEM; a range of pipeline GHG emissions intensities developed through its application to an inventory of major U.S. and Canadian crude oil pipelines; and a sensitivity analysis to identify the most significant parameters affecting pipeline GHG emissions intensity. This information helped us quantify an “energy efficiency gap” that emphasizes the need to incorporate both the integrated system and minimization of GHG emissions in future design and planning exercises related to the pipeline network. Previous LCA literature on transportation fuels either assumes a single GHG emissions factor for pipeline transport, or analyzes a few discrete case studies to demonstrate variability in this life cycle stage. By applying a model such as COPTEM to the inventory of major Canadian and U.S. pipelines, we systematically demonstrate that the emissions intensity of this stage can actually vary by two orders of magnitude [0.23 – 20.3 g CO2e/(bbl*km)]. While the mean emissions intensity [3.0 g CO2e/(bbl*km)] is in line with other studies, we find that there is a greater range of intensities than what has been previously reported. In particular, we find that there are at least eight trunk pipelines whose GHG intensities are at least twice the mean [i.e. from 6.0 – 20.3 g CO2e/(bbl*km)]. If these intensities are applied to long-distance transportation (e.g. 3000 km, roughly the distance from Alberta to the U.S. Gulf Coast), the pipeline transport stage could comprise from 4-12 percent of the well-to-wheels GHG emissions for a transportation fuel. COPTEM also allows us to assess the relative contribution of different parameters to the transportation stage emissions intensity; we find that the linear velocity of the transported crude is the most impactful parameter. As recent industry trends indicate that linear velocities along major pipelines are increasing, accounting for variability in pipeline transport GHG emissions will become increasingly important to climate policy discussions. The development of COPTEM also enables exploration of options for reducing GHG emissions across the pipeline system. For example, a pipeline operator or regulator interested in reducing the GHG intensity of an existing pipeline may choose to transport a different type of crude oil (e.g., by transporting upgraded bitumen instead of diluted bitumen), and/or to twin the pipeline to transport the same volume of product with a lower linear velocity. Either of these actions would likely result in system costs in the $1-10 billion range (incurred either by the pipeline operator, or by a company deciding to upgrade bitumen), and exhibit different potential for improving the GHG emissions intensity of pipeline transport. It is important to note that each of these actions may involve other life cycle economic and GHG emissions tradeoffs – for example, improving crude oil parameters via upgrading may reduce pipeline transport GHG emissions intensity, while still increasing overall transport fuel life cycle GHG emissions intensity (35). Additional research questions can be explored with COPTEM. For example, the model could be integrated with route-specific technical and economic analyses to identify the optimal pipeline diameter, route, and flow rate for a particular pipeline design. Integration with geographic information systems (GIS) tools could help complement the relatively “macro”, energy-focused strengths of this model with a more micro-scale analysis of specific geographic


14

Page 15 of 20

547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594


factors. On a systems level, policymakers and researchers could combine the model with dynamic economic modeling to evaluate the energy and GHG emissions impacts of potential infrastructure and policy decisions, such as allowing overseas exports of US crude oil. Furthermore, employing COPTEM in a consequential LCA that accounts for economic signals could help analyze the systems impacts of changes to the transportation fuel supply chain, including increased shale oil production, or restricted pipeline capacity relative to North American oil supply. Inclusion of other crude oil transportation pathways in the model, such as transport by rail and by ship, would also help better inform industry, regulators, and public stakeholders on the economic and GHG emissions tradeoffs of multiple transportation options, complementing current discussion around the safety aspects of these options. There are also opportunities to improve the accuracy of COPTEM – or at least, limitations that should be considered when using it. One key simplifying assumption of this analysis is the quantities of different types of crude oil present in each pipeline. Our sensitivity analysis demonstrated that crude density and viscosity have a material impact on the GHG emissions intensity of pipeline transport; however, public information is often unavailable on the precise types of crude oil and proportions that are transported by pipelines. More specific data on these properties would lead to a more accurate estimate of pipeline transport GHG emissions beyond our simplifying assumptions. Additionally, incorporating the thousands of gathering pipelines that connect individual wellheads to upstream processing facilities would likely expand the range of emissions intensities and the aggregate contribution of this sector to GHG inventories. Finally, pipeline companies may increasingly incorporate drag-reducing agents that help to reduce turbulence and friction associated with increased flow rates through a constrained pipeline system (36); this may be a valuable option to reduce the GHG emissions intensity of pipeline transport. The development of COPTEM and this analysis is intended to inform discussion on pipeline infrastructure decisions by assessing energy, GHG emissions, and (to a lesser extent) economic tradeoffs of pipeline transport options. However, other factors may be important to these decisions, including safety, protection of specific geographical features and habitats along a proposed route, sharing of economic benefits with communities along a proposed route, and broader questions surrounding the pace and scale of industrial development. This analysis cannot substitute for the broader public discussion that accompanies decisions with potentially far-reaching consequences. Nevertheless, it is our intent that the contributions of this paper will lead to more informed decision-making regarding the transportation fuels life cycle, including improved analysis of the tradeoffs involved in crude oil transportation modes, design, and technologies; decisions across the transportation fuel life cycle that affect/depend on crude oil parameters; and policymaking that intends to reduce the life cycle GHG emissions intensity of these fuels.

ASSOCIATED CONTENT AUTHOR INFORMATION Corresponding Author *Phone: 403-220-5265; e-mail: [email protected].


15


595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641

Page 16 of 20

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENTS The authors would like to thank the Natural Sciences and Engineering Research Council of Canada for financial support and insights helpful to this research. We would also particularly like to thank Dr. Anil Mehrotra and Edward Torres (University of Calgary) for their helpful feedback on this analysis, as well as the helpful feedback from three anonymous reviewers. Any opinions, findings, and recommendations expressed in this analysis are solely those of the authors. Supporting Information The Supporting Information includes a more detailed discussion of the modeling, data and sensitivity analysis. It is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.033985.

REFERENCES (1) Canadian Energy Pipeline Association. About Pipelines: Our Energy Connections, 2014. Available WWW: http://www.cepa.com/wp-content/uploads/2013/10/CEPA_Our-EnergyConnectionsE_Oct04.pdf. Last Accessed: November, 2017. (2) U.S. Energy Information Agency. U.S. Energy Mapping System, n.d. Available WWW: https://www.eia.gov/state/maps.php. Last Accessed: November, 2017. (3) Jacobs Consultancy. Life Cycle Assessment Comparison of North American and Imported Crudes. Jacobs Consultancy Life Cycle Associates: Chicago, IL, 2009. (4) TIAX. Comparison of North American and Imported Crude Oil Lifecycle GHG Emissions. TIAX LLC: Cupertino, CA, 2009. (5) Charpentier, A., Kofoworola, O., Bergerson, J.A., and MacLean, H.L. Life cycle greenhouse gas emissions of current oil sands technologies: GHOST model development and illustrative application. Environmental Science & Technology 2011, 45, 9393-9404. (6) Argonne National Laboratory. Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET), v. 1.8d1. U.S. Department of Energy: Chicago, IL, 2010; http://greet.es.anl.gov/. Last Accessed: November, 2012. (7) Environment Canada. National Inventory Report 1990-2014, 2016. Available WWW: http://unfccc.int/national_reports/annex_i_ghg_inventories/national_inventories_submissions/ items/9492.php. Last Accessed: November, 2017.


16

Page 17 of 20

642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689


(8) U.S. Environmental Protection Agency. National Inventory Report 1990-2014, 2016. Available WWW: http://unfccc.int/national_reports/annex_i_ghg_inventories/ national_inventories_submissions/items/9492.php. Last Accessed: November, 2017. (9) Zhou, W. Reliability Evaluation of Corroding Pipelines Considering Multiple Failure Modes and Time-dependent Internal Pressure. Journal of Infrastructure Systems 2011, 17, 216-224. DOI 10.1061/(ASCE)IS.1943-555X.0000063. (10) Wood, A.; Lance, B.L. Using Water Main Break Data to Improve Asset Management for Small and Medium Utilities: District of Maple Ridge, B.C. Journal of Infrastructure Systems 2009, 15, 111-119. DOI 10.1061/ASCE1076-0342200915:2111 (11) Ugarelli, R.; Venkatesh, G.; Brattebø, H.; Di Federico, V.; Saegrov, S. Asset Management for Urban Wastewater Pipeline Networks. Journal of Infrastructure Systems, 2010, 16, 112-121. DOI 10.1061/ASCEIS.1943-555X.0000011. (12) Hong, H.P.; Allouche, E.N.; Trivedi, M. Optimal Scheduling of Relacement and Rehabilitation of Water Distribution Systems. Journal of Infrastructure Systems, 2006, 12, 184-191. DOI 10.1061/(ASCE)1076-0342. (13) Bidmus, H.O.; Mehrotra, A.K. Heat-Transfer Analogy for Wax Deposition from Paraffinic Mixtures. Ind. Chem. Eng. Res. 2004, 43, 791-803. (14) Degiorgis, G.L.; Maturano, S.; Garay, M.; Galliano, G. R.; Fornes, A. Oil Mixture Viscosity Behaviour: Use in a Pipeline Design. Society of Petroleum Engineers, Proceedings from the SPE Latin American and Caribbean Petroleum Engineering Conference, 2001. (15) Ainsworth, S. Pipeline Size Optimization. Proceedings from World Water Congress, American Society of Civil Engineers, 2004. (16) Filion, Y.R., MacLean, H. L.; Karney, B. W. Life-Cycle Energy Analysis of a Water-Distribution System. Journal of Infrastructure Systems 2004, 10, 120 – 130. DOI 10.1061/ASCE1076-0342. (17) Pootakham, T.; Kumar, A. Bio-Oil Transport by Pipeline: A Techno-Economic Assessment. Bioresource Technology 2010, 101, 7137-7143. (18) Pootakham, T.; Kumar, A. A Comparison of Pipeline Versus Truck Transport of Bio-Oil. Bioresource Technology 2009, 101, 414-421. (19) Chandel, M.K; Williams, E.; Pratson, L.F. Potential Economies of Scale in CO2 Transport Through Use of a Trunk Pipeline. Energy Conversion and Management, 2010, 51, 2825-2834. DOI 10.1016/j.enconman.2010.06.020. (20) Strogen, B, Horvath, A., and Zilberman, D. Energy intensity, life-cycle greenhouse gas emissions, and economic assessment of liquid biofuel pipelines. Bioresource Technology 2013, 150, 476-485. (21) Tarnoczi, T. Life Cycle Energy and Greenhouse Gas Emissions from Transportation of


17


690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737

Page 18 of 20

Canadian Oil Sands to Future Markets. Energy Policy, 2013, 62, 107-117. (22) Nimana, B., Verma, A., Di Lullo, G., Rahman, M.M., Canter, C.E., Olateju, B., Zhang, H., and Kumar, A. Life Cycle Analysis of Bitumen Transportation to Refineries by Rail and Pipeline. Environ. Sci. Technol., 2017, 51, 680-691. (23) Enbridge Inc. Fast Facts on Pump Stations. Available WWW: http://www.enbridge.com/ projects-and-infrastructure/projects/~/media/2509CBF391B74FB9B33935264971D1C1.ashx. Last Accessed: November, 2017. (24) The Weather Network, website, 2017. Temperature Statistics. Available WWW: https://www.theweathernetwork.com/forecasts/statistics/temperature/cl3015808/caab0149. Last Accessed: November, 2017. (25) U.S. Climate Data, website, 2017. Available WW: http://www.usclimatedata.com/ (26) U.S. Energy Information Administration. State electricity profiles, website, 2011. Available WWW: http://www.eia.gov/cneaf/electricity/st_profiles/ e_profiles_sum.html. Last Accessed: November, 2017. (27) Carnegie Mellon. “Economic Input-Output Life Cycle Assessment,” website, 2016. Available WWW: http://www.eiolca.net/. Last Accessed: November, 2017. (28) Smith, C.E. “Crude Oil Pipeline Growth, Revenues Surge; Construction Costs Mount”, Oil and Gas Journal, 1 September 2014. (29) Canadian Association of Petroleum Producers. “Map of Canadian and U.S. Pipelines and Refineries”, website, 2014. Available WWW: http://www.capp.ca/canadaIndustry/oil/Pages/PipelineMap.aspx. Last Accessed: November, 2017. (30) U.S. Energy Information Agency. “Crude Oil Input Qualities,” website, 2014. Available WWW: http://www.eia.gov/dnav/pet/pet_pnp_crq_dcu_nus_m.htm. Last Accessed: November, 2017. (31) Hooker, J., Rose, A., and K. Bertram. Comparison of Operational Energy Intensities and Consumption of Pipelines versus Coastal Tankers: U.S. Gulf Coast to Northeast Coast Routes. Oak Ridge and Argonne National Laboratories, 1980. (32) IHS CERA. Oil Sands and Green House Gas Policy: A New Challenge. IHS CERA, 2010. (33) InterPipeline. Pipeline Construction Projects, 2017. Available WWW: http://www.interpipeline.com/operations/new-projects.cfm. Last Accessed: November, 2017. (34) U.S. Environmental Protection Agency. eGRID 2014 Summary Tables, website, 2017. Available WWW: https://www.epa.gov/energy/egrid-2014-summary-tables. Last Accessed: November, 2017.


18

Page 19 of 20

738 739 740 741 742 743 744 745


(35) Choquette-Levy, N; MacLean, H.; and JA Bergerson. “Should Alberta Upgrade Oil Sands Bitumen: An Integrated life cycle framework to evaluate Energy Systems Investment Tradeoffs.” Energy Policy, 2013. (36) Thomas, Laura; and Tim Burden. “Heavy Oil Drag Reducing Agent (DRA): Increasing Pipeline Deliveries of Heavy Crude Oil.” ConocoPhillips Specialty Products Presentation, 2010, Available at: http://www.coqa-inc.org/docs/default-source/meetingpresentations/06102010_Thomas_and_Burden.pdf. Last Accessed: November, 2017.


19



Page 20 of 20

COPTEM: A Model to Investigate the Factors Driving Crude Oil

Recommend Documents