Advancing Sustainable Pavements: Life Cycle Assessment and Global Warming Potential Benchmarking for Asphalt Mixtures in Louisiana

Abstract

Transportation-related greenhouse gas (GHG) emissions in Louisiana have risen significantly, yet the environmental impacts of asphalt mixture production remain underexplored. This study conducted a cradle-to-gate environmental life cycle assessment (LCA) to quantify global warming potential (GWP) for asphalt mixtures produced in Louisiana and establish GWP benchmarks tailored to mixture types. The LCA encompassed material extraction and production, transport to plants, and asphalt mixing, using two datasets: Environmental Product Declarations from the NAPA Eco-label program (21 mixtures) and Job Mix Formulas from the LaPave database at the Louisiana Department of Transportation and Development (DOTD) (207 mixtures). GWP was evaluated using the FHWA LCA Pave tool with TRACI 2.1 factors, and benchmarks were set at the 20th, 40th, 50th percentiles, and the average. Statistical analyses assessed differences across nominal maximum aggregate sizes and traffic levels. Results showed GWP benchmarks from both datasets exceeded U.S. General Services Administration thresholds by an average of 6.5%, with significant variation among mixture types. These findings highlight the need for targeted emission reduction strategies and accurate environmental performance evaluation to promote more sustainable pavement practices and greener infrastructure in Louisiana.

1. Introduction

The transportation sector accounted for 29% of the U.S. greenhouse gas (GHG) emissions in 2022 [1]. In Louisiana, the GHG emissions of the transportation sector increased by 29.8%, from 32.1 million metric tons (MMT) of Carbon Dioxide equivalent (CO₂eq) weight in 2018 (ranked 21st among U.S. states) to 41.6 MMT CO₂eq in 2022 (ranked 13th among U.S. states) [1,2]. Carbon Dioxide (CO₂) accounts for approximately 80% of greenhouse gas (GHG) emissions in the U.S. and over 92% in Louisiana [2,3]. One of the primary contributors to GHG emissions in the transportation sector is asphalt pavement production [4]. An environmental cradle-to-gate Life Cycle Assessment (LCA) has been conducted to quantify GHG emissions during asphalt pavement production. The system boundary for the cradle-to-gate LCA encompasses raw material extraction (Phase A1), transporting materials to the asphalt facility (Phase A2), and manufacturing the asphalt mixture (Phase A3). The consumed energy was converted to GHG emissions or environmental impacts for each phase, including global warming potentials (GWP).

The GWP serves as the primary indicator for the environmental impacts of GHG emissions (referred to herein after as CO₂eq) related to asphalt mixtures [5,6]. During phase A1, GWP is relatively affected by the asphalt mixture component materials within the Job Mix Formula (JMF), including the type and proportion of asphalt binder, aggregates, reclaimed asphalt pavement (RAP), and additives. Asphalt binder makes the highest contribution to GWP due to its derivation from fossil fuel production [7]. Additives are used in relatively small quantities, yet have high GWP intensities [8]. During Phase A2, aggregate transportation represents the largest contributor to the global warming potential (GWP), particularly when local sources are unavailable and materials must be imported from other states, as observed in Louisiana [9] and Florida [6]. It is worth noting that RAP reduces the demand for virgin binder and aggregates, and reduces transportation-related emissions due to its local availability [8,10]. During phase A3, energy consumption in the mixer/burner is the primary contributor to GWP, representing around 80% [11]. It is worth noting that burner fuel consumption increases significantly with age, maintenance, and mix temperature requirements; as such, asphalt plants might have different energy consumption within the same organization [8].

The General Services Administration (GSA) has implemented material requirements to reduce embodied GHG emissions, emphasizing the use of Environmental Product Declarations (EPDs) for construction materials, including asphalt mixtures. EPDs are standardized documents that provide a transparent and comprehensive assessment of a product’s environmental impact, especially GWP throughout its LCA. Under the Emerald Eco-Label Program, the National Asphalt Pavement Association (NAPA) has published 4977 third-party verified EPDs for asphalt mixtures across various states, including 21 EPD reports from Louisiana’s asphalt plants [8]. The EPD requirements have established benchmark limits for the GWP indicator, normalized to a CO₂-equivalent unit over a 100-year timeframe [8,12]. According to ISO 21678, benchmark values are categorized into limit values (minimum acceptable performance), reference values (current practice), and target values (ambitious performance levels based on best practices). As such, the GSA has identified national GWP benchmarks for asphalt mixtures at the 20th percentile (target value), 40th percentile (reference value), and above-average performance (limit value) [13].

Several states are independently advancing the use of EPDs in their efforts toward more sustainable asphalt pavement, such as California, Oregon, Washington, Minnesota, and Colorado [14]. These distinct state-level actions highlight a growing recognition of the value of transparent environmental data in shaping more sustainable asphalt pavement across the nation. The sustainability of asphalt pavement production necessitates the establishment of GWP benchmarks for asphalt mixtures in Louisiana to enable effective mitigation of GHG emissions within the transportation sector. Despite having 21 published EPD reports under the NAPA ECO-Labeled program, Louisiana’s data remains insufficient for deriving statistically significant and regionally representative GWP benchmarks. As such, there is a critical need for localized benchmarks and standardized LCA methodologies, as emphasized by the GSA. While previous studies have mainly applied LCA as a complementary tool to support mechanical performance results, this research is dedicated solely to the LCA of asphalt mixture manufacturing. This study addresses this critical gap by conducting a project-level LCA specific to Louisiana’s asphalt mixture production to establish GWP benchmarks. By enhancing the precision of environmental impact benchmarks, this research will strengthen the reliability of guiding industry efforts to reduce GHG emissions, promote material efficiency, and ultimately support sustainable pavement objectives in Louisiana’s transportation sector.

2. Objective and Scope

The objectives of this study were to:

(A)

Conduct an LCA for asphalt mixture production in Louisiana,

(B)

Establish GWP benchmarks for asphalt mixture production in Louisiana based on two data sources: EPD reports collected from NAPA’s Eco-label Emerald program, and historical data collected from the LaPave database at DOTD and contractor surveys,

(C)

Develop GWP benchmarks tailored to specific asphalt mixture types, categorized by their mixture design level as classified by DOTD, and

(D)

Compare both approaches of GWP benchmarks to the ones proposed by GSA.

Figure 1 presents the research methodology for GWP benchmarking of asphalt mixtures in Louisiana. Two approaches were followed based on data sources: (1) published EPDs for 21 asphalt mixtures within the NAPA Eco-label Emerald Program, and (2) JMFs of 207 asphalt mixtures available at the LaPave database and contractor surveys. The cumulative distribution curves of GWP values were developed for both datasets. These curves rank GWP values from 0 to 100%, allowing for identification of mixtures below specific benchmarks, such as the 20th, 40th, 50th percentiles, and the average. Cumulative curves are useful for regulatory policy, sustainability targets, procurement strategies, and benchmarking studies [12].

3. Data Collection and Life Cycle Inventory

Two approaches were considered for GWP benchmarking based on data sources:

• DOTD (LaPave) construction materials database: A total of 207 asphalt mixtures’ JMFs, which were utilized in paving routes in Louisiana, were collected. The properties of the collected asphalt mixtures are presented in

  Table 1. Properties of asphalt mixtures investigated in this study.

  Parameter	Asphalt Mixture Type	Number of Mixtures, %
  		DOTD Database	NAPA Reports
  Mix Type	Wearing Course	77 (37)	5 (24)
  	Binder Coarse	69 (33)	-
  	Base Coarse	24 (12)	16 (76)
  	Incidental	27 (13)	-
  	Thin lift (OGFC, Coarse Mix, Dense Mix)	10 (5)	-
  NMAS	½ inch	105 (51)	13 (62)
  	¾ inch	67 (32)	4 (19)
  	1 inch	35 (17)	4 (19)
  Asphalt Binder Grades	PG 58–28	10 (5)	3 (14)
  	PG 67–22	67 (32)	3 (14)
  	PG 70–22	65 (31)	12 (58)
  	PG 76–22	65 (31)	3 (14)
  RAP Content *	0%	26 (13)	1 (5)
  	0–20%	121 (59)	10 (48)
  	20–25%	50 (24)	3 (14)
  	>25%	10 (5)	7 (33)

  Notes: NMAS: nominal Maximum Aggregate Size; RAP: Reclaimed asphalt pavement, DOTD: Department of Transportation and Development of Louisiana; NAPA: National Asphalt Pavement Association; OGFC: Open Graded Friction Course; PG: Performance Grade of asphalt binder. * RAP contents are aligned with the maximum allowable RAP content for each asphalt layer, as specified in Table 502-2 and 502-6 of the Louisiana Standard Specifications for Roads and Bridges.

  .
• The 21 EPD asphalt mixture reports of Louisiana were collected from the NAPA website [15], and the asphalt mixtures’ properties are presented in Table 1. EPD reports contain GWP for phases A1, A2, and A3. As such, there was no need to conduct an LCI or LCA for NAPA’s reports in this study; subsequently, GWP benchmarks were established directly.

The EPD tool data gathering sheet (Version 5), created under the Emerald Eco-Label EPD program [16], was distributed as a survey to asphalt mixtures’ contractors to provide details about transporting material to the asphalt plant (Phase A2). The survey includes the transportation means, distances, capacities, and fuel consumption for acquiring each component material of a studied asphalt mixture to the associated asphalt plant. In addition, the survey included details about energy consumption at the asphalt plant during asphalt mixture production (Phase A3). In this study, all asphalt plants utilized natural gas fuel for the burner/mixer.

It is worth mentioning that LCA studies inherently involve assumptions in defining system boundaries, data quality, and the allocation methods used to determine GWP intensities. These assumptions can strongly influence the outcomes and may introduce uncertainty or lead to sparse results that are difficult to interpret [17].

Table 2. Example of data collection for a studied asphalt mixture.

Phase	LCI Parameter	Quantity
A1	Coarse Aggregates, %	82.4
	Fine Aggregates, %	12.5
	Virgin unmodified asphalt binder	PG 76–22
	Virgin AC, %	5.1
A2	Coarse Aggregate 1—train distance, miles	375
	Coarse Aggregate 2—train distance, miles	603
	Fine Aggregate—train distance, miles	417
	AC—Transport distance, miles	1056
	SBS—Transport distance, miles	192
A3	Burner Energy Consumption (Natural Gas) (MFC/st)	425.2
	Electricity (kWh/st)	3.72
	Fuel (Gallons/st)	0.019

Notes: LCI: Life cycle inventory; AC: Asphalt content; SBS: Styrene-butadiene-styrene; PG: Performance grade of asphalt binder; MFC: Thousand cubic feet; st: Short ton; kWh: Kilowatt hours.

presents an example of data collection for a studied asphalt mixture. It encompasses details about component materials types and percentages, transporting materials to asphalt plants, and asphalt mixture production.

This study employed the recommended public background datasets from the American Center for Life Cycle Assessment (ACLCA) Open Standard and the FHWA LCA Pave tool to assess the potential environmental impacts. The environmental impacts used in this study were evaluated using the enhanced Environmental Protection Agency’s (EPA’s) pedigree matrix for use in the pavement LCA domain [18,19]. In the absence of an industry-wide aggregate lifecycle inventory, this study relied on publicly available Environmental Product Declarations (EPDs) from LCA Pave, which encompass the cradle-to-gate impacts of aggregate production. Transportation distances were determined using the material supplier and mixing plant locations. The environmental impacts of pavement materials were assessed using the LCA Pave tool, which applies TRACI 2.1 (Tool for the Reduction and Assessment of Chemicals and Other Environmental Impacts), an EPA-developed and ISO-recognized methodology for calculating midpoint indicators [20,21].

Table 3. Life cycle inventory for a studied asphalt mixture.

Sub-Phase	Materials/Equipment	Unit	Kg CO2eq	LCI Data Source–Year Published	Other Properties	Meta-Analysis
A1	Coarse aggregates	St	2.06	Martin Marietta—2017	EPD 4531 [22]	77.1%
	Fine aggregates	St	4.2	Martin Marietta—2017	EPD 952 [22]	77.1%
	RAP	St	1.26	Illinois Tollway LCI—2016	[23]	74.4%
	PG 76–22	St	694	Asphalt Institute, 2019	3.5% SBS [24]	79.7%
	PG 70–22	St	628	Asphalt Institute, 2019	1.5% SBS [24]	79.7%
	PG 67–22	St	578	Asphalt Institute, 2019	unmodified [24]	79.7%
A2	Train (diesel)	St-mile	0.0538	USLCI data, 2000	[25]	51.5%
	Barge (Diesel)	St-mile	0.0795	USLCI data, 2000	[26]	53.2%
	Truck (diesel)	St-mile	0.2264	USLCI data, 2000	[27]	53.2%
A3	Natural Gas	MFC	0.05	USLCI data, 2002	[28]	41.5%
	Electricity (U.S.)	kWh	0.546	National Energy Technology, 2019	[29]	68.2%
	Fuel	Diesel	10.45	USLCI data, 1998	[30]	41.5%

Note: CO₂eq: equivalent weight of Carbon Dioxide; LCI: Life cycle inventory; PG: Performance grade; St: Short ton = 907.18 Kg; RAP: Reclaimed asphalt pavement; EPD: Environmental Product Declaration; SBS: Styrene-butadiene-styrene; MFC: Thousand cubic feet; kWh: Kilowatt hour. USLCI: U.S. Life Cycle Inventory Database.

presents the life cycle inventory for the studied asphalt mixtures. It contains GWP impact factors for converting materials and energies consumed during the production of one short ton of asphalt mixture to GWP by weight of CO₂eq. It is worth mentioning that meta-analysis was reported by LCA Pave software using a scale from 1 (best) to 5 (worst) for factors, such as reliability, TRACI compatibility, representatives, data age, manufacturing technology, reviews, and completeness. However, the factor scores were averaged and weighted on a scale of 100%.

4. Environmental Impact Analysis and Life Cycle Assessment

Table 4. Example calculations for GHG emissions of a studied asphalt mixture.

Phase A1	Quantity (st)	GWP Equivalent Factor			Kg.CO2eq/st
Coarse aggregates	0.824	2.06			0.78
Fine aggregates	0.125	4.20			1.80
Asphalt binder and SBS	0.051	694			35.39
					37.62
Phase A2	Quantity (st)	Distance (mi)	Transportation Means	Equivalent CO2eq Factor/st.mi	Kg.CO2eq/st
Coarse aggregates 1	0.291	375	Train	0.0538	5.87
Coarse aggregates 2	0.532	603	Train	0.0538	17.26
Fine aggregates	0.125	417	Train	0.0538	2.80
Asphalt binder	0.051	207	Truck	0.2264	2.39
SBS (3.5% of AC)	0.002	192	Truck	0.2264	0.09
Fiber (2% of AC)	0.001	1037	Truck	0.2264	0.23
					28.65
Adding unloaded return	Multiply fuel consumption by (1.80 for Trucks—1.46 for Barge—1.50 for Trains) **				43.77
Phase A3	Quantity/st	GWP Equivalent Factor			Kg.CO2eq/st
Burner (MFC)	276.7	0.05			13.84
Electricity (kWh)	3.62	0.546			1.98
Fuel (Gallon)	0.02	10.45			0.21
					16.02
Total (A1 + A2 + A3)					97.41

Notes: st: Short ton; GWP: Global warming potential; Kg.CO₂eq: Equivalent kilo-gram of Carbon Dioxide; SBS = Styrene-butadiene-styrene; mi: Mile, AC: Asphalt content; MFC = Thousand cubic feet; kWh = Kilowatt-hour. ** Return haul factor (Table 5).

presents an example of calculations for GHG emissions of an asphalt mixture. The environmental impact analysis was conducted on the collected data (LCI) for each phase of LCA.

Table 5. Average return haul factor for fuel consumption across transportation modes in southern states.

Mode		Reference	Summery	Percent Reduction (Unloaded/Loaded)	Return Haul Factor -Considered
Truck		[31]	6.2 MPG loaded 7.3 MPG unloaded	85%	80%
		[32]	5.5 MPG loaded 7.5 MPG unloaded	73%
Barge	General, USA	[33,34]	675 TMPG —loaded	40%	46%
		[35]	270 TMPG —unloaded
	Lower Mississippi River	[31]	1290 TMPG—loaded southbound 185 TMPG (31.5%)—unloaded northbound	185 × (100/31.5)/1290 = 46%
Train		[36]	528 TMPG loaded 260 TMPG unloaded	49%	50%
		[37]	500 TMPG loaded 250 TMPG unloaded	50%
		[38]	470 TMPG loaded 235 TMPG unloaded	50%

Notes: MPG: mile per gallon; TMPG: ton mile per gallon

presents the relative fuel consumption of unloaded return trips compared to loaded trips, expressed as a return factor. This factor varies considerably by mode of transportation. Reported values are approximately 80% for trucks and 50% for trains. For barges, the return factor is more location-dependent; although a general value of 40% is often cited, a higher factor of 46% has been observed along the Lower Mississippi River, which is geographically more accurate for this study.

5. Global Warming Potential (GWP) Benchmarking

To evaluate the GWP of asphalt mixtures, collected data are fitted to a normal distribution curve. This statistical approach allows for the determination of the data’s central tendency, dispersion, and overall pattern [39]. In addition, based on the premise that asphalt mixture properties exhibit natural variability, a normal distribution was chosen for this environmental analysis to identify the average carbon emissions and potential outliers. The normal distribution, characterized by the dataset values’ mean and standard deviation, is widely used due to its simplicity and robustness in representing such variability [40]. The validity of the normal distribution assumption was assessed through the Kolmogorov–Smirnov statistical test [41].

Table 6. Statistical test (Kolmogorov–Smirnov) for normal distribution fitting.

Test	A1	A2	A3	A1–A3
p-value	0.832	0.052	<0.001	0.124
Normal distribution check (p ≥ 0.05)	pass	pass	fail	pass

presents the results of the Kolmogorov–Smirnov statistical analysis to validate the possibility of fitting the GWP results using a normal distribution curve. The normality of the data is evaluated using the p-value, a statistical measure that assesses how well the data fit a normal distribution. When the p-value is equal to or more than 0.05, it indicates that the data do not significantly deviate from normality, providing 95% confidence that the data follows a normal distribution. However, if the p-value is less than 0.05, it suggests that the data significantly differs from a normal distribution, indicating abnormality. It is worth noting that the GWP values for phases A1, A2, and A1–A3 (i.e., the combined phases A1, A2, and A3) followed a normal distribution (p-value ≥ 0.05). However, the GWP values for phase A3 did not conform to normality (p-value < 0.05). This deviation is attributed to the data source of A3; energy consumption during asphalt production (burner, electricity, and diesel use) was collected at the plant level, independent of mixture-specific components and mixing temperature. Similar uncertainty and difficulty of collecting accurate data regarding mixing production were reported by NAPA [12].

Steps of establishing normal and cumulative distribution curves:

Step 1. Calculate mean (μ) using Equation (1), and standard deviation (σ) using Equation (2):

$$\mu = {\displaystyle \frac{1}{N}}\sum _{i=1}^N{x}_i$$

(1)

$$\sigma =\sqrt{{\displaystyle \frac{1}{N-1}}\sum _{i=1}^N{(x_i-\mu )}^2}$$

(2)

where

• x_i is the value of GWP for a studied mixture.
• N is the total number of studied mixtures.

Step 2. Step size (bin width) Estimation:

To construct a well-shaped and representative frequency distribution, an optimal bin width was calculated using the Freedman–Diaconis Rule [30] as follows:

• Sort the GWP values in ascending order.
• Determine the first (Q1 = GWP at 25% of data points), second (Q2 = GWP at 50% of data points), and third (Q3 = GWP at 75% of data points) quartiles.
• Compute the Interquartile Range (IQR): IQR = Q3 − Q1

Calculate bin width (step size) using Equation (3):

$$\mathrm{Step}\mathrm{size}=(2\ast {\displaystyle \frac{IQR}{\sqrt{N}}})$$

(3)

where N is the number of data points.

Table 7. The calculations of optimal bin width for normal and cumulative curves.

Phase A1–A3	GWP, Kg.CO2eq./st
Mix 1	48.16
Mix 2	48.23
:	:
Mix 207	126.2
Mean (µ)	77.3
Standard deviation ($\sigma$)	17.9
Quartile 1 (Q1)	65.7
Quartile 2 (Q2)	75.9
Quartile 3 (Q3)	87.2
Interquartile Range IQR = Q3 − Q1	21.5
Step size = $2\mathrm{*}{\displaystyle \frac{IQR}{\sqrt{N}}}$	3.0

Notes: N: total number of studied mixtures; GWP: Global warming potential; Kg.CO₂eq./st: Kilogram of carbon dioxide emitted per short-ton of asphalt mixture.

illustrates an example of the calculations of optimal bin width for the normal and cumulative distribution for the A1–A3 phase of the studied asphalt mixture.

6. Establishing Normal and Cumulative Distribution Curves

a.

GWP values for asphalt mixtures obtained from the LaPave database

Normal Distribution Function:

The probability density function of the normal distribution curve was calculated using Equation (4) [42]:

$$f\left(x\right)={\displaystyle \frac{1}{\mathsf{\sigma }\sqrt{2\pi }}}{e}^{{-{\displaystyle \frac{1}{2}}({\displaystyle \frac{x-\mu }{\sigma }})}^2}$$

(4)

Figure 2, Figure 3 and Figure 4 shows the normal distribution curves of GWP values resulting from A1, A2, and A1–A3 phases for the studied asphalt mixtures.

Cumulative Distribution Function:

The normal distribution curve was transformed into a cumulative distribution curve, which ranks a GWP value from zero to one hundred percent. This percentile rank can identify the proportion of asphalt mixtures exhibiting GWP values below specific thresholds or benchmarks, such as the 20th, 40th, and 50th percentiles. Cumulative curves serve as valuable tools for informing regulatory policies, sustainability targets, procurement strategies, and benchmarking studies [12]. The cumulative distribution function was calculated using Equation (5) [42]:

$$F\left(x\right)= {\displaystyle \frac{1}{2}} \left[1+\mathrm{erf}({\displaystyle \frac{x-\mu }{\sigma \sqrt{2}}})\right]$$

(5)

where “erf” is the Gauss error function, approximated by the Maclaurin series using Equation (6). The Maclaurin series expansion of the error function converges for all finite values of z [30]; however, in this study, it converges up to the maximum asphalt mixture emissions level in the studied dataset.

$$\mathrm{erf} (\mathrm{z})={\displaystyle \frac{2}{\sqrt{\pi }}} (z-{\displaystyle \frac{z^3}{3}}+{\displaystyle \frac{z^5}{10}}-{\displaystyle \frac{z^7}{42}}+\dots ), \mathrm{z}={\displaystyle \frac{x-\mu }{\sigma \sqrt{2}}}$$

(6)

Figure 5, Figure 6 and Figure 7 present the resulting cumulative distribution graphs based on a proper bin width. The 20^th, 40^th, and 50^th percentiles, and the average were plotted to determine the GWP levels aligned with the GSA thresholds for GWP of asphalt mixtures.

b.

GWP values for asphalt mixtures obtained from NAPA Reports

Figure 8 illustrates the cumulative distribution curve of GWP values for phases A1–A3 for Louisiana’s reported asphalt mixtures.

Table 8. Descriptive Statistics of Phases A1, A2, A3, and A1–A3.

Statistical Parameter	DOTD-LaPave (207 Mixes)				NAPA (21 Mixes)
	A1	A2	A3	A1–A3	A1	A2	A3	A1–A3
20th percentile	29.3	17.2	12.1	62.7	26.1	12.0	25.9	71.2
40th percentile	31.6	26.7	14.5	72.5	28.1	22.5	27.3	83.7
50th percentile	32.5	28.5	14.6	77.2	31.5	28.2	27.3	86.3
Mean ± Std.	32.5 ± 3.8	29.5 ± 14.3	15.3 ± 3.7	77.3 ±17.9	30.4 ± 6.1	26.5 ± 15.4	28.4 ± 3.0	85.3 ± 14.8
Minimum	24.2	9.5	12.1	48.2	20.1	2.1	25.9	59.47
Maximum	42.4	82.2	23.4	126.2	42.1	55.8	34.9	113.83

illustrates the descriptive statistical parameters for different phases to clarify the central tendency and variation in results.

Figure 9 presents the estimated GWP benchmarks for Louisiana’s asphalt mixtures, derived from two distinct data sources: the LaPave database and NAPA published EPD reports. These benchmarks were subsequently compared against the National GSA thresholds. It is worth noting that the GWP benchmarks exhibited variation based on the data source. Specifically, the benchmarks derived from the LaPave database were consistently lower than those obtained from the NAPA EPD reports. It is worth noting that the background database and allocation methods for both analyses from NAPA and DOTD are identical and based on the TRACI V2.1 database. However, the quality of data and different percentages of high-RAP mixtures might have reduced the overall emissions of NAPA mixtures. NAPA analysis had 33% of mixtures containing more than 25% RAP compared to only 5% mixes in DOTD analysis.

However, both the LaPave and NAPA-derived GWP benchmarks for asphalt mixtures were numerically higher (17.5%) than the thresholds proposed by the GSA. This variability shows the necessity to account for regional variations in specifications, material selection, and production methodologies when establishing such environmental benchmarks. In addition, the average GWP of DOTD mixtures was numerically higher (6.5%) than the GSA average threshold, which emphasizes the need for innovative strategies to reduce asphalt-related GHG emissions in Louisiana.

7. Developing GWP Benchmarks for Asphalt Mixtures as Classified by DOTD

In the Louisiana Standard Specifications for Roads and Bridges-Table 502-6, asphalt mixtures have been categorized based on:

• Based on nominal maximum aggregate size (NMAS)
• Based on the traffic design level
• Based on average daily traffic (ADT)

To ascertain whether significant differences existed in GWP benchmarking values across various asphalt mixture categories, an ANOVA (Analysis of Variance) and Bonferroni post hoc test were conducted.

Table 9. Asphalt mixture categorized by the DOTD specification.

Classification 1: Nominal Maximum Aggregate Size
NMAS	12.5 mm	19 mm and 25 mm	p-value	Significant differences
No. of Mixes	105	102
Average GWP Value	79.4	73.7	0.017	yes
Classification 2: Traffic design level
Traffic Level	1 (less than 3 mESALs)	2 (more than 3 mESALs)	p-value	Significant differences
No. of Mixes	148	59
Average GWP Value	73.7	86.4	<0.001	Yes
Classification 3: Average daily traffic
ADT	<3500 ADT	>3500 ADT	p-value	Significant differences
No. of Mixes	61	146
Average GWP Value	69.5	80.6	<0.001	Yes

Notes: NMAS: Nominal maximum aggregate size; mESALs: Million equivalent single axle load; ADT: Average daily traffic.

presents a detailed breakdown of these asphalt mixture categories, including the results of the statistical analyses (represented by the p-value), indicating the presence or absence of significant differences among asphalt mixture types within each category.

ANOVA and the Bonferroni post hoc tests were performed to determine whether the GWP values among different categories were statistically significant. The null hypothesis for each comparison assumed that no differences existed between the categories. If the p-value is lower than 0.05, it means the difference between categories is statistically significant.

• The results showed a statistically significant difference between mixtures with an NMAS of 12.5 mm and mixtures with a larger NMAS (19 mm and 25 mm), with a p-value of 0.017, Figure 10. It is worth noting that finer-graded mixtures release higher GWP emissions due to increased surface area of finer aggregates, which necessitates higher asphalt binder content to achieve adequate coating and workability. Asphalt binder is known for producing significantly high GWP emissions.

2.

The asphalt mixture designed for traffic level 2 showed significantly higher GWP emissions (p < 0.001) than mixtures designed for traffic level 1. Mixtures designed for higher traffic levels (more than three mESALs) produce 17.2% higher emissions compared to mixtures designed for lower traffic design level (Less than three mESALs), Figure 11. This might be attributable to the structural requirements of high-traffic pavements, which often demand high-quality binder, higher binder content, and less recycled materials than low-traffic pavements.

3.

The asphalt mixtures serving less ADT than 3500 vehicle per day (vpd) showed a significant reduction in GWP emissions when compared to the ones with higher ADT, Figure 12. This might be attributable to the higher structural requirements for high volume roads, which often demand high-quality binder, higher binder content, and less recycled materials.

8. Conclusions

This study developed a global warming potential (GWP) benchmarking framework for asphalt mixtures in Louisiana using two complementary data sources: Environmental Product Declaration (EPD) reports from the NAPA Eco-label program and approved Job Mix Formulas (JMFs) from the LaPave database available at DOTD. The GWP values were assessed using the FHWA LCA Pave tool and TRACI 2.1 methodology for DOTD mixtures, while GWP values for the NAPA mixtures were directly obtained from published EPDs. Normal and cumulative distribution curves were constructed to derive percentile-based benchmarks at 20th, 40th, and 50th percentiles, which enabled comparison across categories and against national General Services Administration (GSA) thresholds. Further, statistical analyses using ANOVA and a Bonferroni post hoc test were conducted between different types of asphalt mixtures containing various nominal maximum aggregate sizes, traffic levels, and Average Daily Traffic, following the classification of DOTD. Based on the findings of this study, the following conclusions were made:

• GWP benchmarks were established for Louisiana’s asphalt mixtures at 20th, 40th, 50th percentiles, and average.
• GWP benchmarks collected from NAPA data (Average percentile = 85.3 Kg.CO₂eq./st) showed slightly higher values than those obtained from the DOTD database (Average percentile = 77.3 Kg.CO₂eq./st). This is due to the differences and uncertainties of the GWP values of phase A3.
• The average GWP benchmarks established based on DOTD data were numerically higher (6.5%) than the GWP thresholds determined by GSA. This emphasized the need for innovative strategies to reduce asphalt-related GHG emissions in Louisiana.
• Significant differences were found between GWP emissions of asphalt mixtures with different NMAS, traffic levels, and ADT as classified by DOTD. This finding necessitated an accurate, tailored GWP benchmark for asphalt mixtures to obtain more sustainable pavement and greener infrastructure.
• In this study, Phase A3 data inventory was limited to plant-average energy and emissions rather than mixture-specific data, reducing accuracy. Future work should collect mixture-level data to lower uncertainty and improve comparability across plants. In addition, this study was limited to a cradle-to-gate system boundary. Future research should extend to cradle-to-grave analyses and include sensitivity assessments of key factors to better identify emission reduction strategies and inform policy and practice.