4 Authors:

Ben Moins  University of Antwerp, Belgium

Cyril France University of Antwerp, Belgium

Wim Van den bergh University of Antwerp, Belgium

Amaryllis Audenaert University of Antwerp, Belgium

*Corresponding author

E-mail address: ben.moins@uantwerpen.be (Ben MOINS)

 

Abstract

Life cycle cost analysis (LCCA) has received notable attention and application within the road industry. As one of the three pillars in sustainability assessment, LCCA offers an empirical framework to assess costs over the entire lifespan of road projects. To incorporate the agency and user cost for all different life cycle phases, a robust framework is needed. Thus, it is vital to gain insight into the application and limitations of LCCA in road projects. Reviewing the existing economic models and frameworks, with a particular focus on road projects, will be the first step in providing a robust and uniform model. The goal of this paper is to provide a state-of-the-art review of existing methodologies in the wider field of LCCA for road projects. Hence, it can highlight critical processes and identify hotspots so the robustness of LCCA frameworks can be increased. It is concluded that agency costs related to the end of life (EOL) phase, transport and road user costs are often excluded, despite having a substantial impact. However, with sustainability in mind, these aspects are important and should always be incorporated. Modelling the EOL enables the user to include the effect of recycling, hence lowering the economic impact of raw material extraction. Additionally, road user costs are closely related to the social aspect of sustainability assessment. Finally, this paper presents the inconsistent use of modelling parameters,

e.g. applied discount rate and analysis period, which supports the conclusion of a missing conclusive and robust framework.

 

 

Highlights

-          If an LCCA wants to transform into a sustainability assessment, it should always incorporate road user costs.

-          If pavements are compared, an analysis period of 40 years will probably be sufficient. However, when full structures are compared, an analysis period of 40 years will not be long enough.

-          It is concluded that transport between extraction, production, construction and waste processing sites are often excluded or unknown.

 

 

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

1.  Introduction

With the increasing focus on sustainability, the road industry is confronted with the challenge of considering sustainable practices [1]. However, enabling sustainable transition, in combination with ensuring durability, requires high levels of investments [2,3]. Because agencies are often constrained by inadequate funds for investments, the assessment of future costs over longer periods has gained attention [4–6]. Considering budgetary constraints for projects, agencies need to use rigorous decision-making methodologies that provide insights about long-term economic viability of investments. One of those methodologies is life cycle cost analysis (LCCA), which evaluates the economic burden of an asset’s life cycle while still focussing on its durability [7–12].   

LCCA requires thorough identification of all materials and processes used at every phase. However, the practical application of LCCA depends on several factors. Examples include the availability of credible supporting data, insights in structural characteristics and material properties, modelling deterioration to predict the pavement condition [13–15], and the availability of guidelines to estimate user costs [16–18]. Additionally, it is indicated that different LCCA models are used interchangeably while they cannot be linked due to the lack of consistency in system boundaries and input parameters [5,19]. Some authors suggest a hybridized eco-efficiency analysis, which combines life cycle assessment (LCA) and LCCA [19,20]. Others propose an LCCA based on optimizing performance and available funds [14]. Another type of analysis that is proposed, is a new probabilistic simulation- optimization LCCA that takes into account the uncertainty of the input parameters [5,21,22].

 Therefore, the difference in technique and their parallelism should be analysed, and in addition, how they are applied in roads projects. Hence, this study contains two primary aims that are achieved by answering five sub-questions:

 1.       To provide a state-of-the-art review and analysis of the existing methodologies in the wider field of LCCA for road projects.

a.       How is LCCA for road projects defined and how was it developed?

b.      What are the life cycle phases and related cost components of a road?

2.       To highlight the critical processes and differences between the models, and to identify the shortcomings so the robustness of an LCCA framework for road projects can be increased.

a.       Which cost components are currently (not) considered?

b.      What are the most commonly applied economic models and corresponding analysis parameters?

c.       How is sensitivity analysis being performed?

 

In order to answer these, 44 case studies from the past decade were analysed. In this way, practitioners can access valuable information to gain insights into LCCA in practical terms. A full discussion of price ranges and the impact of individual materials on the life cycle cost (LCC) lies beyond the scope of this study as it would distract attention from the aforementioned gaps.

 

2.  LCCA development and definitions

The concept of cost comparison in road engineering was first introduced by William Mitchell Gillespie in 1847. He stated: “The road which is truly cheapest is not the one which has cost the least money, but the one which makes the most profitable returns in proportion to the amount which has been expended upon it” (p.65 [23]). This concept was not widely used until the 1950s and 1960s. In 1960, the American Association of State Highway and Transportation Officials (AASHTO) published an informational guide on project procedures [24]. According to AASHTO, cost computations should reflect original investments, anticipated lifespans, maintenance and salvage values. They highlighted the importance of cost comparisons based on service life of a pavement structure because maintenance would seriously affect the cost comparison. However, it should only be implemented when based on accurately kept long-term records.

 

Initially developed by the US Department of Defence to enhance its cost-effectiveness in awarding competitive bids, LCCA have gained relevance in other sectors that seek to make decisions for sustainable development. AASHTO introduced LCC in their road design guide in 1972 and reused it in their 1983 and 1993 guides. According to AASHTO, LCC consists of all costs and benefits involved in the provision of pavements throughout their complete life cycle. This should include costs related to construction, maintenance, rehabilitation and recycling for the highway agency and costs related to travel time, vehicle operation (VOC), accidents (AC) and time delay during initial construction, maintenance or rehabilitation for the road user. Since these costs do not occur at the same time, an interest rate or time value of money became important. Hence, the terms net present value (NPV) and equivalent uniform annual cost (EUAC) were introduced [25]. The National Cooperative Highway Research Program (NCHRP) adopted these definitions and published in 1985 one of the first manuals to perform an LCCA of pavements [26]. In 1988, they revised this work, which resulted in recommendations regarding methods of cost-effectiveness analysis for highway projects [27].

 

Later in 1995, the Federal Highway Administration (FHWA) published a designation act with requirements to conduct an LCCA for all project segments on the national highway system with a cost of $25 million or more [28]. In 1996, they wrote a policy statement where they highlighted the importance of LCCA and encouraged the implementation in projects with a cost lower than $25 million [29]. FHWA published a manual for LCCA in pavement design in 1998 that addressed broad fundamental principles as well as detailed procedures. This document advocated the use of probabilistic approaches to incorporate risk analysis to consider uncertainty, which was typically hidden in the traditional deterministic approaches [30]. Therefore, it became the foundation for later FHWA LCCA guidance and tools.

 

LCCA for European road projects was introduced in 1997 during the Forum of European Highway Research Laboratories [4]. This led to the establishment of the PAV-ECO (economic evaluation of pavement maintenance) project in 1999 with European agencies from Finland, Denmark, Germany, France, Switzerland and United Kingdom [31]. Recently, efforts have been made by several researchers to define the term LCCA. Santos et al. [3] stressed the importance of long-term effects: “LCCA is an analytical methodology that uses economic principles to evaluate long-term alternative investment options in infrastructure management processes in order to select optimal strategies” (p.1 [3]). Also, Lee et al. focussed on long-term effects: “LCCA is an analytical technique that uses economic principles to evaluate long-term alternative investment options for highway construction” (p.2 [32]). Abdelaty et al. focussed on the fact that LCCA should be performed during the design stage of projects. According to them, LCCA is “a set of procedures used to evaluate the economic value of different design alternatives at the design stage of the project development process” (p.724 [33]).
 

Guo et al. combined both aspects in the following definition: “LCCA is a way to evaluate the long-term cost-effectiveness of different pavement designs or treatment actions” (p.389 [5]). Hasan et al. did not focus on the timeframe or period when an LCCA should be performed, but on a clear difference between initial costs and costs that incur in the future: “LCCA is the conventional procedure for the evaluation of the financial benefits and returns from any investment by analysing its future expenditures along with the initial costs” (p.542 [34]).

 

All the previous definitions are somewhat different. However, there are some keywords that are recurring: analytical technique, economic principles, long-term, a period of analysis, design alternatives, initial costs and future costs. The lack of a clear and conclusive definition for LCCA in road engineering makes it is imperative to have a working definition. In the context of road engineering, this paper defines LCCA as a systematic or analytical methodology that uses economic techniques to evaluate long-term life cycle costs of alternatives by calculating the initial costs and discounting all future costs incurred, throughout the road’s lifespan, over a predefined period of analysis.

 

3.  The life cycle of a road

The general life cycle of a road is presented in Fig. 1. This diagram differentiates two approaches for material flows. The first approach is linear, represented by phase 1 raw materials extraction until phase 5B landfill or 5C energy recovery after incineration, which considers material flows where new virgin materials are required after the end of life (EOL) of the previous system. The second approach is circular, represented by phase 1 raw materials extraction until phase 5A recycling, which has the goal to utilize products at the highest value of all time and takes into account material flows where primary materials are saved due to recycling and reuse of waste products after the EOL of the previous system [7,35,36].

 

Whether recycling or reuse is preferred over the use of new materials depends on the impact of multiple factors. For example, the difference in transportation distances, the impact of the recycled or reused materials on the durability of the new product and the impact of the recycling or disposal processes. Hence, including all the different life cycle phases is important for the sustainability assessment of alternatives [37–39]. Therefore, the following part of the paper will describe all the different life cycle phases of a road and important cost components which should be considered.

 

3.1.  Raw materials extraction

A holistic consideration of the LCCA of a road’s life cycle commences with the cost of raw materials extraction, acquisition and/or production of primary materials such as aggregates, fillers, sand, bitumen, cement, water and additives [40,41]. This should include all the processes involved in acquiring (e.g. extracting and mining), processing (e.g. refining and crushing) and transporting road materials [1]. The costs associated with this phase arise from the use of equipment, materials, energy, labour, and transportation of materials to the production plant. In most cases, all the cost components are grouped into a material unit price. However, distances between materials sources and production plants can significantly affect the cost of materials, especially when secondary waste materials are being recycled and compared with primary materials [42–44]. In addition, the transportation type can also have a substantial effect on the cost of the materials [44]. This makes it difficult to differentiate costs if a unit price is used when other transport distances or types are considered for other alternatives. Therefore, for a detailed LCCA, the cost of transportation should be settled separately from the material cost.

 

3.2.  Mixture production

The production of bitumen or cement-bound mixtures is according to several authors one of the main contributors to the total LCC due to the high amount of energy that is needed for drying, heating and mixing the different materials [45–49]. The energy consumption during production also has a strong link with the impact of the material extraction phase. For example, when primary aggregates for asphalt production are being saved due to the use of reclaimed asphalt pavement (RAP), a second burner often needs to be installed for drying RAP. Consequently, the energy cost during the production will increase, while the material cost decreases. Vice versa, when the energy consumption is lowered, for example, when the temperature is decreased to produce warm mix asphalt instead of hot mix asphalt, an additive is used. This additive is an extra material which increases the material cost but decreases the energy cost during production. Other cost components that are associated with this phase are the costs for handling the materials at the production plant, equipment, emission permits, labour, taxes, licensing and operation permits [3,50].

 

3.3.  Road construction

Road construction comprises diverse processes and associated equipment requirements for initial construction works [40,41]. First, the unbound materials, bitumen or cement-bound mixtures and equipment must be transported to the construction site. Afterwards, construction can begin, which can include (but is not limited to) the following aspects: clearing of the site, excavating, treating the base or foundation with cement or lime and compacting it, constructing and compacting the different road layers, and integrating the different ancillary road facilities (e.g. lighting and signs) [40]. The associated costs include transportation, safety measurements, construction machinery or equipment (i.e. fuel consumption, mobilization, demobilization, insurance, taxes, interest depreciation and licenses), storage on site and labour (i.e. wages, contractual benefits) [3]. Additionally, road users incur extra costs due to construction. These costs can be significant and should be included in the analysis. They include VOC, a higher possibility for AC due to more narrow lanes and work zone delay costs (WZDC) due to time loss [1].

3.4.  Use phase and maintenance

The use phase accounts for the interactions between vehicles, the pavement surface condition and the pavement performance over its service life [40]. The road user costs associated with this stage consist of fuel consumption, tire wear, vehicle maintenance and vehicle depreciation due to mileage [1,3,40]. Road durability and performance is affected by climatic factors [51,52] (e.g. temperature) and human- induced activities (e.g. traffic load and initial design), consequently causing deterioration. Therefore, maintenance is needed to lower the road’s deterioration speed. Maintenance and rehabilitation (M&R) are a set of preventive and reactive works that are performed at different periods throughout the service life of the road to enhance the overall performance and maintain its serviceability [41]. If there is excessive deterioration, rehabilitation is required to restore the pavement to its former condition.

 

In general, three types of M&R can be identified [53–57]. Preservation is applied when the road is still in good condition. This treatment extends the road’s service life without increasing its structural capacity, which makes this treatment relatively simple and inexpensive. However, preservation treatments must be applied before deterioration starts which results in a more frequent application. Examples of preservation treatments are crack filling, patching, slurry seals, chip seals, micro-surfacing and diamond grinding. Maintenance delays future deterioration and improves the road’s condition without substantially increasing its structural capacity. Examples of maintenance treatments are ultra- thin and thin asphalt overlays, stress absorbing membrane interlayers for concrete pavements, hot in- place recycling and cold in-place recycling. Finally, when the pavement condition and structural capacity is too poor, with a high risk of structural failure, service life must be extended by applying rehabilitation treatments. Rehabilitation has the highest cost because it involves the milling or demolishing of the existent road and the reconstruction of a new one. Similar to the construction phase, the associated cost components for M&R activities may include materials, construction machinery or equipment, labour and transportation [3].

 

3.5.  End of Life: road deconstruction and waste processing

The EOL phase includes the final disposal, processing or recycling of the road at the end of its service life [1,41]. Generally, three waste streams are generated when a road is milled or demolished and transported to a processing site [3,40]. Firstly, see 5A in Fig. 1, waste can be processed (crushed and sieved) into secondary materials for production, e.g. RAP or reclaimed concrete aggregate. Secondly, it can be disposed at a landfill site, as presented by 5B in Fig. 1, when the material is contaminated with and cannot be reused directly as construction material, e.g. asphalt with steel fibres. However, this is not preferred because it is a linear stream that results in the extraction of new virgin materials. Additionally, waste should only be landfilled when it is not harmful to the surrounding environment. Finally, if the materials are harmful to the surrounding environment, such as RAP containing tar, they are incinerated so that energy can be recovered, see 5C in Fig. 1. The associated cost components are fees for waste generation and disposal, transportation, milling/demolishing and sweeping during deconstructing, waste processing and labour.

 

4.  LCCA in road engineering – Process & Steps

 

4.1.  Define goal & scope and alternative design strategies

The first step in performing an LCCA is to define the goal & scope and design alternatives. This helps to understand how design choices may impact the future cost of the road user, initial construction, M&R and EOL [58,59]. It is important to keep in mind that all alternatives have to follow all necessary

standards so that mechanical performance and durability are ensured during the specified period of analysis [60]. The analysis period is regarded as the time horizon over which future costs are evaluated. However, the BS ISO 15686-5 standard [61], which describes the methodology for performing LCCA for buildings and constructed assets, does not specify an analysis period. According to the standard, it is the researcher who determines the analysis period.

 

Walls and Smith [59] advocate for an analysis period longer than the road’s lifespan, except for instances in which there is an extremely long-lived road, e.g. roads with perpetual pavements. In that case, the analysis period should be equal to the road’s lifespan. Other authors, such as Caltrans [62] and Li et al. [44], recommend an adequately long analysis period to reflect long-term costs associated with practical design strategies, including M&R and EOL. They propose analysis periods of 35 and 55 years. The analysis period of 35 years is suggested if a preventive maintenance program for 20-year design alternatives is compared. The analysis period of 55 years should be used when 20-year and 40- year design alternatives are compared. Analysis periods shorter than 10 years should only be used when the road or pavement is a temporary solution which needs significant reconstruction within this analysis period. Generally, when alternatives with different design lives are compared, the analysis period should be based on the alternative with the longest design life. Additionally, as a rule of thumb, the selected analysis period should be the same for all alternatives under consideration so that the long-term costs for these alternatives can be determined and compared with each other.

 

The absence of a clear guideline for a predefined analysis period is demonstrated in Fig. 2, which shows the analysis periods used in the recent case studies from Fout! Verwijzingsbron niet gevonden. and Fout! Verwijzingsbron niet gevonden.. The graph shows a wide variety of applied analysis periods. This is an important observation as it has implications for the comparison of results from different authors. Studies with different analysis periods should never be compared because the LCC is highly dependent on the analysis period used. The large spread of the results suggests that only a few studies can be compared with each other and that results should never be compared one-to-one without considering the analysis periods.

30,00%

 

25,00%

 

20,00%

 

15,00%

 

10,00%

 

5,00%

 

0,00%

n/a       10        20       23,5      25        30        35        40        45        50        58        60        75        90

Analysis period [years]

 Asphalt pavement  Concrete pavement  Flexible structure  Rigid structure  Semi-flexible structure

 

Fig. 2 Frequency of analysis period used in recent case studies per road type (n=56)

However, some observations can be made despite the variance in analysis period. Generally, it can be concluded that studies which are focusing on the life cycle of rigid structures use longer analysis periods. In addition, case studies that are only focusing on flexible and/or semi-flexible structures use shorter analysis periods. As mentioned before, it is important to use the same analysis period for all alternatives. Therefore, when studies combine rigid structures with other types, the analysis period also tends to be higher. This is likely to be related with the higher lifespan of concrete compared to asphalt.

  

General information					Analysis information				User costs			Agency costs
												Initial Constr.				Maintenance				EOL
Author	Year	Country	Variants	Focus	Method	AnP	DR	SA	VOC	WZDC	AC	ME	MP	RC	T	P	M	R	T	RDC	WP	T	SV	RV
Robinette et al. [63]	2010	USA	12	FS, SFS	CBA	-	-	-	-	-	-	x	x	x	?	x	x	x	?	x	-	-	-	-
Leng et al. [64]	2011	USA	3	AP	CBA	-	-	-	-	x	-	x	x	x	?	-	-	-	-	-	-	-	-	-
Lee et al. [65]	2011	USA	3	FS, RS	NPV	60	4%	-	-	x	-	x	x	x	?	-	x	x	?	x	-	-	-	-
Amini et al. [66]	2012	Iran	2	FS	NPV	40	2,3%	OFAT	x	x	-	x	x	x	x	x	x	x	?	x	-	-	-	-
Cheng et al. [67]	2012	USA	2	AP	NPV, EUAC	20	4%	MC	x	x	-	x	x	x	?	x	x	x	?	x	-	-	x	-
Santos et al. [68]	2012	Portugal	16	FS	NPV	40	3%	-	?	?	?	x	x	x	?	x	x	x	?	x	-	-	-	x
Gschösser et al. [47]	2013	Switzerland	3	FS, SFS, RS	NPV	75	2%	-	-	-	-	x	x	x	x	-	x	x	x	x	x	x	-	-
Li et al. [45]	2013	USA	3	FS, SFS	NPV	23,5	3%	-	-	-	-	x	x	x	x	-	-	x	x	x	-	-	x	x
Aurangzeb et al. [69]	2014	USA	4	AP	NPV	45	4%	OFAT	?	x	?	x	x	x	?	x	x	x	?	-	-	-	-	x
Mirzadeh et al. [46]	2014	Sweden	1	AP	EUAC	25	4%	OFAT	-	x	-	x	x	x	x	-	-	x	x	x	-	-	-	-
Son et al. [70]	2014	USA	6	AP	CBA	-	-	MC	?	x	-	-	-	-	-	-	x	-	?	-	-	-	-	-
Zaumanis et al. [71]	2014	?	1	AP	SPPA	-	-	OFAT	-	-	-	x	x	-	?	-	-	-	-	-	x	-	-	-
Santos et al. [3]	2015	USA	3	AP	NPV	50	2,3%	OFAT	x	x	-	x	x	x	x	-	x	x	x	x	-	-	-	x
Yang et al. [72]	2015	USA	11	AP	CBA	-	-	-	-	-	-	x	x	x	x	-	-	-	-	-	-	-	-	-
Yu et al. [53]	2015	?	1	AP	NPV	40	4%	-	x	-	-	-	-	-	-	x	x	x	?	x	-	-	-	x
Abdelaty et al. [33]	2016	USA	3	AP	EUAC	25	4%	MC, OFAT	-	-	-	-	-	-	-	x	x	x	?	x	-	-	-	-
Nazzal et al. [73]	2016	USA	2	AP	NPV	10	?	-	-	-	-	x	x	x	?	?	x	?	?	?	-	-	-	-
Souliman et al. [74]	2016	Sweden	3	AP	CBA	-	-	OFAT	-	-	-	x	x	x	?	-	-	-	-	-	-	-	-	-
Sultan et al. [75]	2016	China	7	FS, SFS	CBA	-	-	-	-	x	-	x	x	?	?	-	-	-	-	-	-	-	-	-
Sultan et al. [76]	2016	China	3	SFS	NPV	40	?	-	-	x	-	x	x	x	?	-	x	x	?	x	x	?	x	-
Wennström et al. [56]	2016	Sweden	12	AP	NPV, EUAC	40	4%	OFAT	x	x	x	x	x	x	?	-	-	x	?	x	-	-	-	x
Yepes et al. [77]	2016	Chile	4	AP, CP	NPV	25	4%	-	-	-	-	-	-	-	-	x	x	x	?	x	-	-	-	-

 

AnP = Analysis Period, DR = Discount Rate, SA = Sensitivity Analysis, VOC = Vehicle Operation Costs, WZDC = Work Zone Delay Costs, AC = Accident Costs, ME = Material Extraction, MP = Mixture Production, RC = Road Construction, T = Transport, P = Preservation, M = Maintenance, R = Rehabilitation, RDC = Road Deconstruction, WP = Waste Processing, SV = Salvage Value, RV = Residual Value, AP = Asphalt Pavement, CP = Concrete Pavement, FS = Flexible Structure, SFS = Semi-Flexible Structure, RS = Rigid Structure, CBA = Cost-Benefit Analysis, NPV = Net Present Value, EAUC = Equivalent Annual Uniform Cost, SPPA = Simple Payback Period Analysis, OFAT = One Factor At-a-Time, MC = Monte Carlo simulation, BA = Bayesian Analysis, - = Excluded, x = Included, ? = Unknown

General information					Analysis information				User costs			Agency costs
												Initial Constr.				Maintenance				EOL
Author	Year	Country	Variants	Focus	Method	AnP	DR	SA	VOC	WZDC	AC	ME	MP	RC	T	P	M	R	T	RDC	WP	T	SV	RV
Akbarian et al. [21]	2017	USA	6	AP, CP	NPV	35	2%	MC	x	x	-	x	x	x	?	x	x	x	?	x	-	-	-	x
Wu et al. [78]	2017	China	3	CP	NPV	25	3,5%	MC, OFAT	x	x	-	-	-	-	-	x	x	x	?	x	-	-	-	-
Liu et al. [79]	2017	UK	10	AP	CBA	-	-	-	-	-	-	-	x	-	-	-	-	-	-	-	-	-	-	-
Underwood et al. [51]	2017	USA	3	AP	NPV	30	4%	-	-	-	-	x	x	x	?	-	-	x	?	x	-	-	-	x
Santos et al. [80]	2017	USA	14	AP	NPV	50	2,3%	OFAT	x	x	-	x	x	x	x	x	x	x	x	x	-	-	-	x
Ozer et al. [81]	2017	USA	8	SFS	CBA	-	-	-	-	-	-	x	x	x	x	-	-	-	-	-	-	-	-	-
Batouli et al. [58]	2017	USA	3	FS, RS	NPV	50	5%	MC, OFAT	x	x	x	x	x	x	?	?	x	x	?	x	-	-	-	-
Torres-Machi et al. [54]	2017	Chile	20	AP, CP	NPV	25	6%	OFAT	-	-	-	-	-	-	-	x	x	x	?	x	-	-	-	-
Coleri et al. [82]	2018	USA	3	AP	NPV	50	?	OFAT	-	-	-	x	-	x	?	-	-	x	?	x	-	-	-	x
Lee et al. [32]	2018	USA	2	AP, CP	NPV	60	4%	-	-	x	-	x	x	x	?	-	x	x	?	x	-	-	-	-
Diependaele [83]	2018	USA	2	FS, RS	NPV	90	3%	OFAT	-	-	-	x	x	x	?	x	x	x	?	x	-	-	-	-
Diependaele [83]	2018	Belgium	2	AP, CP	NPV	inf.	4%	OFAT	-	-	-	x	x	x	?	x	x	x	?	x	-	-	-	-
Qiao et al. [84]	2019	USA	3	FS	NPV	20	4,6%	OFAT	x	x	-	x	x	-	?	-	x	-	?	-	-	-	-	-
Qiao et al. [52]	2019	USA	2	FS	NPV	20	2%	OFAT	x	-	-	x	x	x	x	x	x	-	x	x	x	x	x	-
Wang et al. [85]	2019	USA	2	AP	EUAC	-	4%	BA, OFAT	-	-	-	-	-	-	-	-	x	x	?	-	-	-	-	-
Chen et al. [86]	2019	Taiwan	3	AP	NPV	40	4%	MC, OFAT	x	x	-	x	x	x	x	-	x	x	x	x	-		-	-
Guo et al. [5]	2019	USA	2	FS, RS	NPV	30	4%	MC, OFAT	x	-	-	x	x	x	?	-	x	-	?	-	-	-	-	-
Okte et al. [55]	2019	USA	1	AP	NPV	58	3%	OFAT	x	x	x	x	x	x	?	-	x	-	?	-	-	-	-	-
Nahvi et al. [87]	2019	USA	4	AP	EUAC	-	3%	MC	-	-	-	-	-	-	-	x	-	-	?	-	-	-	-	-
Choi [50]	2019	Korea	3	AP	NPV	40	4,5%	-	x	x	x	-	-	-	-	-	x	x	?	x	-	-	-	-
Nahvi et al. [12]	2019	USA	2	AP	EUAC	-	3%	MC	-	-	-	-	-	-	-	x	-	-	?	-	-	-	-	-
Gu et al. [17]	2019	USA	1	AP	NPV	40	2,1%	MC, OFAT	x	x	-	x	x	x	?	-	x	x	?	x	-	-	-	x
Jahanbakhsh et al. [48]	2020	Iran	9	AP	CBA	-	-	-	-	-	-	x	x	-	?	-	-	-	-	-	x	-	-	-

 

AnP = Analysis Period, DR = Discount Rate, SA = Sensitivity Analysis, VOC = Vehicle Operation Costs, WZDC = Work Zone Delay Costs, AC = Accident Costs, ME = Material Extraction, MP = Mixture Production, RC = Road Construction, T = Transport, P = Preservation, M = Maintenance, R = Rehabilitation, RDC = Road Deconstruction, WP = Waste Processing, SV = Salvage Value, RV = Residual Value, AP = Asphalt Pavement, CP = Concrete Pavement, FS = Flexible Structure, SFS = Semi-Flexible Structure, RS = Rigid Structure, CBA = Cost-Benefit Analysis, NPV = Net Present Value, EAUC = Equivalent Annual Uniform Cost, SPPA = Simple Payback Period Analysis, OFAT = One Factor At-a-Time, MC = Monte Carlo simulation, BA = Bayesian Analysis, - = Excluded, x = Included, ? = Unknown

Fig. 3 presents the analysis period used per set of system boundaries. The analysis shows that 29,5% of the case studies did not use or specify an analysis period (n/a). This includes nine cases that performed a cost-benefit analysis (CBA), one case that performed a simple payback period analysis (SPPA) and three cases that calculated the EUAC of maintenance activities for asphalt pavements with service lives under 15 years. In total, five cases used an analysis period of 25 years. Four of these are only focussing on M&R and do not consider initial construction and EOL. Additionally, there are two CBAs that only focus on M&R and do not consider other costs. Hence, it can be concluded that studies which are focussing on M&R often use no analysis period or analysis periods shorter than 25 years. Finally, it can be concluded that CBAs don’t use an analysis period because all activities are done in the base year.

 

30,00%

25,00%

20,00%

15,00%

10,00%

5,00%

0,00%

n/a       10        20       23,5      25        30        35        40        45        50        58        60        75        90

Analysis period [years]

 Initial construction (individually)                         M&R (individually)

 Initial construction -> M&R                                  Initial construction -> EOL

Fig. 3 Frequency of analysis period used in recent case studies per set of system boundaries (n=44)

 

 

 

4.2.  Determine performance periods & M&R activities

After the design alternatives and analysis period have been determined, it is important to predict the initial service lives and M&R to account for future costs [59,88]. Yang et al. [89] describe the initial period as the average time in years for a newly constructed road to reach an agency’s threshold for the first rehabilitation, while the rehabilitation period is the length of time after this first threshold to reach another rehabilitation threshold. Because M&R may be postponed due to budgetary restrictions, the actual rehabilitation does not necessarily occur at the same time which a pavement has reached the performance threshold. Hence, inaccurate estimations of the performance periods directly affects the frequency of agency intervention, consequently affecting agency costs and user costs during construction and M&R [88].

Therefore, the use of pavement performance data in a pavement management system (PMS), in combination with a pavement performance prediction model (PPPM), is critical [90,91]. These are vital techniques as they predict the optimal timing for M&R based on different factors that cause road deterioration, such as material ageing, traffic characteristics and climatic effects [92]. Hence, a PPPM helps to allocate funds for road projects efficiently and decreases the cost of M&R [75,91,93]. Fig. 4 demonstrates the operation of a typical PPPM [9,59,88]. To significantly increase the pavement condition, when excessive distress has accumulated, high maintenance budgets are often needed for rehabilitation. Therefore, in some cases, it is better to use preventive maintenance strategies, which have lower costs than rehabilitation, but need to be applied more frequently. To select the optimal strategy, it is important to determine the impact on the overall LCC using available budgets and the level of acceptance of distress.

 The PPPM uses a condition indicator, often based on measurements of surface characteristics, to predict the pavement condition. Over the years, different pavement condition indicators have been designed for PPPMs and used to characterize the physical condition of roads and to allow accurate M&R decision-making. Some of the most commonly used indicators include the Pavement Serviceability Index, Pavement Condition Index, International Roughness Index, Pavement Condition, Present Serviceability Rating and Pavement Surface Condition Index [94–97]. The choice and application of a PPPM is location specific because the type and accumulation of distress over time often depends on the climate and traffic characteristics [98]. Once all activities have been determined for the analysis period, a corresponding expenditure stream diagram can be prepared, see Fig. 4. The graphical representation of an expenditure diagram in LCCA is important because it presents a visual overview of all cash flows for all alternatives with regards to its initial construction, M&R and EOL considerations throughout its life cycle [10,99]. 

4.3.  Estimate costs

The costs presented in section 3 are often categorized in several groups. According to the ISO standard

[61] and Santos et al. [3], costs can be categorized as being either fixed or variable. Costs that are fixed remain the same regardless of the amount of material production, e.g. costs of insurances, depreciation, licensing and permits. Costs that are influenced by the amount of produced mixture are variable costs. Materials can be, for example, variable because the unit price may vary depending on the purchased quantity. Others, such as Estevan et al. [100] and Wong [101], categorize costs under direct or indirect. According to them costs that are directly associated with road construction e.g. acquisition, production, construction, M&R and EOL are direct costs and external costs like environmental costs are indirect costs.
 

A third and related distinction can be made between tangible or intangible costs [44,99]. Tangible costs are “measurable” out-of-pocket costs that are considered as the project’s expenditures and are estimated based on their available market values. Hence, these are all costs related to the construction, M&R and EOL phase of a road incurred by contractors and road agencies. Intangible costs are the costs encountered as a result of implementing the project but are incurred indirectly and are out-of-pocket. Intangible costs are, for example, incurred by road users because of the time delay due to road construction. Finally, costs can be categorized as either present costs or future costs [46]. In LCCA, all the aforementioned categories are incurred either by the road agency or by the road user [99]. Therefore, all the cost components of the case studies were categorized under user and agency costs as presented in Fig. 5.

 

4.3.1.  Agency costs

Agency costs denotes all costs incurred directly by the road agency throughout the road’s lifespan [58]. These costs should include preliminary engineering, contract administration, production, construction, M&R, transportation of materials and equipment and EOL [30,32,93]. Furthermore, agency costs are estimated for the entire length and lifespan of the design alternatives under consideration. Therefore, these costs are often subdivided into three categories: costs associated with initial construction in the beginning of the road project, M&R during the use phase and EOL at the end of the project.

 

In LCCA, comparisons are only made between competing alternatives that are mutually exclusive, reflecting differential costs between alternatives. Therefore, costs that are equal for all alternatives are often excluded in the analysis. In the past, traffic control costs were not included by many agencies because authors assumed that construction durations for all alternatives would be the same. However, it is possible that alternatives have different construction periods. Hence, the alternative with the longest construction period will have the highest traffic control cost. Therefore, Jackson et al. [102],

strongly recommend considering traffic management costs in comparing alternative design costs as they may have a significant effect. Another example of a cost component that is often excluded, is the planning of a road project, e.g. costs related to structural and/or mixture design and preliminary testing. Fig. 5 demonstrates that most of the other cost components, related to initial construction, are included in LCCA. What stands out in Fig. 5 is that transportation is either unknown or excluded in 77% of the cases. When CBA or SPPA is performed, it is highly likely that the analysis will focus on the initial construction as it takes place in the base year. However, in some LCCAs initial construction is left out of the analysis because the project is focussing on M&R.

 

Although preservation costs are generally low, thus negligible compared to other costs, preservation has to be performed on a regular basis and thus should not be underestimated [102]. It is observed that these especially have a significant effect on the LCCA when rapid fluctuations are observed in future M&R and material prices [53]. However, it is apparent from Fig. 5 that very few studies consider preservation and that transportation costs during M&R are often excluded or unknown. The percentage of studies that include M&R is lower compared to initial construction as CBAs and SPPAs often exclude this phase.

 

The final subcategory of the agency cost is the EOL cost related to the deconstruction of the existing road or pavement, the transportation of waste, waste processing and the salvage or residual value [44,50,82,88]. However, Fig. 5 demonstrates that EOL in most cases is excluded. The only cost component that the cases are considering is the cost associated with the deconstruction of the road, hence milling. In addition, it was found that salvage and residual value are being used interchangeably despite the difference in modelling. Therefore, clear and conclusive definitions are introduced for the use in road design. This paper defines residual value as the value of a road when its service life reaches beyond the end of the analysis period and salvage value as the value after the recycling of materials when a road has reached the end of its service life, so primary materials are saved.

 

According to Gu et al. [17], the salvage value of a material is calculated based on the price of the primary materials minus the processing cost of the recycled material. RAP, for example, has two possibilities. Firstly, it can be recycled as an unbound or cement-bound material when the bitumen is not recycled. Therefore, the salvage value is equal to the material and transport cost of primary aggregates minus the cost of transport and processing RAP. Secondly, it can be recycled into new bitumen-bound layers. Hence, the salvage value is equal to the binder percentage of RAP multiplied by the price (transport included) of virgin bitumen plus the aggregate percentage of the RAP multiplied by the price (transport included) of primary aggregates minus the cost of transport and processing RAP. Hence, Eq. 1 is proposed to determine the salvage value:

 

𝑛 𝑆𝑉 = (∑ 𝑥𝑖 (𝑈𝑀𝐸𝑖 + 𝑈𝑇𝑖 ∗ 𝐿𝑖)) − 𝑈𝑊𝑃𝑟𝑒𝑐. − 𝑈𝑇𝑟𝑒𝑐. ∗ 𝐿𝑟𝑒𝑐. 𝑖=1

Eq. 1

 

Where SV is the salvage value after recycling in €/ton, n is the number of materials that are being saved, xi is the mass percentage of material i in the recycled content, UMEi is the unit price of material i in €/ton, UTi is the unit price for transporting material i in €/(ton.km), Li is the transport distance of material i in km, UWPrec. is the unit cost for processing the recycled material in €/ton, UTrec. is the unit price for transporting the recycled material in €/(ton.km) and Lrec. is the transport distance of the recycled material i in km.

 

The residual value on the other hand can be determined in a simple fashion using Eq. 2 [17,51,53,56]:

 

𝑆 𝑅𝑉 = 𝐶𝐶 (1 −  ) 𝑇

Eq. 2

Where RV is the residual value after the end of the analysis period, CC is the construction cost of the latest activity (initial construction or M&R), S is the expected service life until the end of analysis and T is the total expected service life.

 

However, according to Santos et al., it is better to link the residual value of a pavement with its remaining surface condition [68]. As discussed before, several pavement condition indicators exist. Therefore, Eq. 3 presents the residual value in terms of a general indicator that can be replaced by any specific indicator.

 

𝑅𝑉 = 𝐶𝐶 ∗        𝑃𝐶𝐼 − 𝑃𝐶𝐼𝑡ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑 𝑃𝐶𝐼𝑖𝑛𝑖𝑡𝑖𝑎𝑙 − 𝑃𝐶𝐼𝑡ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑

Eq. 3

Where RV is the residual value after the end of the analysis period, CC is the construction cost of the latest activity (initial construction or M&R), PCI is the measured value of the applied pavement condition indicator after the analysis period is finished, PCIthreshold is the threshold limit for M&R for the pavement condition indicator and PCIinitial is the initial value of the pavement condition indicator of a new pavement or after M&R activities.

 

4.3.2.  User costs

User costs are mainly incurred by the public through the use and operation of vehicles as well as their travel time. The road user cost (RUC) is often grouped as an aggregation of WZDCs, VOCs and ACs [58,102,103]. Fig. 5 indicates that RUC is often excluded in LCCAs for road projects. A possible explanation for this might be that CBAs or SPPAs are focussing on the initial construction and are often only interested in the agency cost. There are, however, other possible explanations. Firstly, RUC is often excluded due to its complexity and challenges that are associated with quantifying the cost components based on unreliable data. Secondly, LCCAs often only include significant differences between alternatives. When projects assume no difference in construction time, there will be no difference in RUC, hence RUC is left out of the analysis. However, there are two important reasons why RUCs should be included in LCCAs for road projects. Firstly, the RUC has a strong connection with the social aspect of sustainability assessments of road projects as the RUC is paid by society. Secondly, although the construction period for several alternatives might not differ, the planning of M&R can differ. Hence, due to discounting, there will be a difference in RUC. Therefore, the following part of this section will describe how the components of the RUC can be determined.

 

4.3.2.1.   Vehicle operation cost (VOC)

VOC models are used to quantify the cost related to vehicle operation and changes in traffic flow conditions. These models can be very complex as they consider several parameters including vehicle category, pavement condition, fuel consumption, oil consumption, tire wear, vehicle M&R, depreciation and time related adjustment factors. Some of the most applied VOC models in road engineering are: NCHRP’s report 133 method [104], FHWA’s HERS-ST model [103], EPA’s Moves model [105], World Bank’s HDM-4 model [106], Australian Road Research Board’s Road Fuel Consumption

model [107] and NCHRP’s MicroBENCOST model [108]. Three of these models will be discussed to present the difference in complexity and show which parameters can be included in VOC models.

 

The model applied by Yu et al. [53] is a simplified version of the aforementioned models. Yu et al. are focussing on the VOC by only considering the fuel consumption of five different vehicle categories and the road condition using Eq. 4:

 

5 𝑉𝑂𝐶 = 𝑈𝐶𝑓 ∑(𝐹𝐶𝑏𝑎𝑠𝑒,𝑖 ∗ 𝐴𝐹𝑃𝐶,𝑖) 𝑖=1

Eq. 4

Where VOC is the vehicle operation cost per km for five different vehicles, UCf is the unit cost of fuel in €/l, FCbase,i is the fuel consumption for vehicle type i in ml/km and AFPC,i is an adjusting factor based on the pavement condition as presented in Table 3.

 

Table 3 Effects of roughness on fuel consumption for a speed of 112 km/h applied by Yu et al. [53]

 

Vehicle class	Fuel consumption		Adjusting factors
International Roughness Index (m/km)
	1	2	3	4	5	6
Medium Car	107.85 ml/km	1.02	1.05	1.07	1.09	1.12
Van	128.96 ml/km	1.01	1.02	1.03	1.03	1.04
SUV	140.49 ml/km	1.02	1.04	1.06	1.08	1.10
Light truck	251.41 ml/km	1.01	1.02	1.02	1.03	1.04
Articulated truck	656.11 ml/km	1.01	1.02	1.04	1.05	1.06

 

The MicroBENCOST model includes, in addition to the fuel consumption, the following parameters: oil consumption, tire consumption, vehicle M&R and vehicle depreciation. The model calculates the VOC by applying equations that include facility length, traffic volume, three vehicle categories and the relevant cost components. Afterwards individual VOCs are calculated and multiplied with their unit costs and a pavement condition factor. Finally, the total VOC is calculated by taking the sum of the several components as presented by Eq. 5.

 

3 𝑉𝑂𝐶 = ∑(𝑈𝑉𝑂𝐶𝑓𝑢𝑒𝑙,𝑖 + 𝑈𝑉𝑂𝐶𝑜𝑖𝑙,𝑖 + 𝑈𝑉𝑂𝐶𝑡𝑖𝑟𝑒,𝑖 + 𝑈𝑉𝑂𝐶𝑀&𝑅,𝑖 + 𝑈𝑉𝑂𝐶𝑑𝑒𝑝,𝑖) 𝑖=1

Eq. 5

Where VOC is the vehicle operation cost per km for 3 different vehicles, UVOCfuel,i is the fuel-related unit VOC per km for vehicle category i, UVOCoil,i is the oil-related unit VOC per km for vehicle category i, UVOCtire,i is the tire-related unit VOC per km for vehicle category i, UVOCM&R,i is the M&R-related unit VOC per km for vehicle category i and UVOCdep,i is the depreciation-related unit VOC per km for vehicle category i.

Table 4 HERS-ST unit costs for VOC resource components in 2004 dollars [103]

 

Cost Component	Small auto’s	Medium/large auto	4-tire truck	6-tire truck	3+ axle single	3-4 axle combo	5+ axle combo
Fuel ($/gal)	1.93	1.93	1.93	1.93	1.84	1.84	1.84
Oil ($/quart)	4.48	4.48	4.48	1.79	1.79	1.79	1.79
Tires ($/tire)	45.89	72.55	79.96	193	477.90	477.90	477.90
M&R ($/1000 miles)	103.50	125.60	159.60	298.70	422.50	437.60	437.60
Depreciation  ($/vehicle)	19717	23255	25061	37448	82386	95432	103767

 

However, the previous models are both using constant speeds and do not consider speed change cycles due to work zones or curvatures of the road. FHWA has developed the HERS model to analyse these aspects and uses a fairly complicated method to compute the VOCs of seven vehicle types as a function of fuel, oil, tires, M&R and depreciation as presented in Table 4. In addition, the model calculates individual VOC components for:

1.       Constant speed operating based on the vehicle category, average speed, average consumption and pavement condition;

2.       Extra operating due to speed change cycles;

3.       Extra operating due to road curvature; Afterwards the total VOC is given using Eq. 6:

7 𝑉𝑂𝐶 = ∑(𝑈𝑉𝑂𝐶𝐶𝑆,𝑖 + 𝑈𝑉𝑂𝐶𝑆𝐶,𝑖 + 𝑈𝑉𝑂𝐶𝑅𝐶,𝑖) 𝑖=1

Eq. 6

Where VOC is the vehicle operation cost per km for seven different vehicles, UVOCCS,i is the constant speed-related unit VOC per km for vehicle category i, UVOCSC,i is the speed change-related unit VOC per km for vehicle category i and UVOCRC,i is the road curvature-related unit VOC per km for vehicle category i.

 

Hence, it can be concluded that the computation of VOCs has several levels of complexity. Although it is often neglected, it is an important parameter which should be considered during the computation of the road user cost. Especially, because the impact of vehicle operation has such a strong connection with the social and environmental aspects of sustainability assessment.

 

4.3.2.2.   Work zone delay cost (WZDC)

WZDCs are calculated using the delay time due to work zones, the value of time (VOT) and the number of vehicles that are affected by the construction zone [55,56,58,103]. Several models exist, however almost all of them are based on the same parameters: speed flow, traffic demand, capacity analysis, queue length and queue speed. Batouli et al. [58] proposed a framework using Eq. 7 - Eq. 10:

 

𝑡      = 𝐿𝑊𝑍 − 𝐿𝑊𝑍 𝑆𝑅        𝑣𝑊𝑍            𝑣0

Eq. 7

 

𝐿𝑄 𝑡𝑄 = 𝑣 𝑄

Eq. 8

 

𝑡𝑊𝑍𝐷 = 𝑡𝑆𝑅 + 𝑡𝑄

Eq. 9

 

𝑊𝑍𝐷𝐶 = 𝑡𝑊𝑍𝐷 ∗ 𝐴𝐴𝐷𝑇 ∗ 𝑡𝑊𝑍 ∗ 𝑉𝑂𝑇

Eq. 10

Where WZDC is the total work zone delay cost, tsr is the speed reduction delay in hours, LWZ is the work zone length in km, vWZ is the maximum speed in the construction zone in km/h, v0 is the upstream speed in km/h, tQ is the queue delay in h, LQ is the average queue length in km, vQ is the queue speed in km/h, tWZD is the work zone delay time in h, AADT is the annual average daily traffic, tWZ is the work zone duration in days and VOT is the value of time in €/h.

 

The abovementioned framework is in line with the framework proposed by FHWA [103]. However, FHWA highlights that the VOT differs for alternative traffic categories. There should be, for example, a difference between the VOT of passenger cars for personal travel, business travel or trucks. Therefore, Eq. 10 should be transformed in Eq. 11 where i stands for the traffic category and n stands for the total amount of considered traffic categories.

 

𝑛 𝑊𝑍𝐷𝐶 = 𝑡𝑊𝑍𝐷 ∗ 𝑡𝑊𝑍 ∗ ∑(𝑉𝑂𝑇𝑖 ∗ 𝐴𝐴𝐷𝑇𝑖) 𝑖=1

Eq. 11

4.3.2.3.  Accident cost (AC)

ACs are all costs incurred by road users resulting from an increase in accidents in work zones due to lane closure and more narrow lanes [50,56,58,103]. According to the framework proposed by FHWA [103], the first step is to determine the pre-construction crash rate (CRPC). In case no data is available, Eq. 12 can be used to estimate CRPC:

 

𝐴 ∗ 106 𝐶𝑅𝑃𝐶 = 𝐿       ∗ 365 ∗ ∑𝐴𝑛𝑃 𝐴𝐴𝐷𝑇 𝑊𝑍                               𝑖=1                      𝑖

Eq. 12

Where CRPC is the pre-construction crash rate per million vehicle km of travel, AnP in the analysis period in years, A is the number of crashes along the project for the analysis period, LWZ is the length of the work zone in km and AADTi is the annual average daily traffic in year i of the analysis period.

 

Afterwards, the accident cost can be determined using Eq. 13:

 

𝐴𝐶 = 𝐶𝑅𝑃𝐶 ∗ 𝐶𝑀𝐹𝐿𝐶 ∗ 𝐶𝑀𝐹𝑆𝑀 ∗ 𝑡𝑊𝑍 ∗ 𝐴𝐴𝐷𝑇 ∗ 𝐿𝑊𝑍 ∗ 𝑈𝐴𝐶

Eq. 13

Where AC is the accident cost related to road construction, CRPC is the pre-construction crash rate per km, CMFLC is a crash modification factor (CMF) due to an increase in crashes after lane closure (as presented, for example, in Table 5), CMFSM is a CMF due to a decrease in crashes after safety measures, twz is the work zone duration in days, AADT is the annual average daily traffic, LWZ is the length of the work zone in km and UAC is the unit cost of accidents.