Life-Cycle Cost Analysis for Infrastructure Project Pavement Design A. Whyte*, and A. Pham* * Civil Engineering Dept., Curtin University, GPO Box U1987, Perth, Australia (E-mail: [email protected])

ABSTRACT Asset management and decision-support tools at the planning phase and throughout the life-cycle of civil infrastructure projects are essential for stakeholders charged with the determination of costeffective design-solutions over an asset’s useful life. The need for a life-cycle cost analysis (LCCA) to guide economic decision-making when comparing competing alternatives is explored by casestudy that juxtaposes the pavement design options of: concrete; and, asphalt. Discussion below outlines briefly the processes involved in developing a new infrastructure LCCA model. Specific design alternatives of: continually reinforced concrete pavement (CRCP); and, thin asphalt surfaced unbound granular pavement (AC) were input for direct whole-life cost comparison. Using life-span periods of 30, 60 and 100 years, with a discount rate calculated at 8%, (incorporating sensitivity analyses) the findings suggest that the asphalt pavement alternative is the cheapest in all cases, with a clearance of up to 34%. Overall, the outcomes of this project validate and vindicate the need for an LCCA for infrastructure projects, while specifically recommending that an asphalt pavement is a cheaper alternative than a concrete pavement in this particular environments locally.

KEYWORDS Infrastructure Asset-Management, Life-cycle Costing, Engineering Economics INTRODUCTION Asset management and the need for decision-support tools both at the planning phase and throughout the life-cycle of civil infrastructure is a key issue for the Australian construction industry; it remains a challenge (for civil engineers) to reduce the costs involved, of a resource or asset, over the duration of its useful design life. In particular, pavement engineers are often faced with alternative specifications to choose from. Comparison of the available alternatives may find that one option is clearly the cheapest according to the initial construction costs. However upon inspection of the costs induced over the life-time of the (pavement) component, it may well be that the maintenance and repair costs outweigh the costs initially saved at construction. Life-cycle cost analysis (LCCA) give a means and method to determine which alternative gives the most cost benefit over the long term. Component comparison methodology is presented in Section 3; this paper gives a practical investigation of LCCA, via development of a LCCA spreadsheet model and application of a casestudy that best represents civil infrastructure. To demonstrate the benefits of applying a LCCA, the research conducted analysis of the alternatives of concrete pavement and asphalt pavement for a high volume highway. The subject was deemed appropriate, with the potential to influence greatly the civil infrastructure industry, given that roads make up a large proportion of total civil infrastructure in Australia. In addition, highways are major schemes that are owned by the State Government for their whole life, as opposed to smaller roads which may be built by developers and owned thereafter by the local council. Thus, due to the sole ownership, the whole-of-life costs would be applicable to the one Government transportation agency.

In particular, the alternatives for the case study were chosen based on an apparent/anecdotal preference for bitumen pavement in Western Australia, rather than a concrete alternative. Concrete pavement is adopted in other places such as parts of the United States and Europe, as well as having gained popularity in the Eastern States of Australia. Why is concrete not readily adopted in Western Australia? Purely on a cost assertion, it would seem that asphalt pavements are favoured due to their cheaper acquisition costs. However, with low initial costs one could assume hefty maintenance costs thereafter. This hypothesis is supported by several sources such as the American Concrete Pavement Association (ACPA, 2002), which states, ‘concrete roadways cost as little as one-third the total cost of building, maintaining and repairing asphalt roadways’. Similarly Sharp (70) claims that ‘it is certain that many local authorities would, if they kept and analysed their records accurately, find that concrete roads are considerably cheaper than any alternative form of construction on an annual cost basis.’ The objective then was to conduct a LCCA to determine whether this claim was true over the asset’s whole- life.

LIFE-CYCLE COST ANALYSIS LCCA is defined by Kirk and Dell’Isola (95) as ‘an economic assessment of competing design alternatives, considering all significant costs of ownership over the economic life of each alternative, expressed in equivalent dollars.’ For pavement designs, life cycle costing can be more specifically described as, ‘an economic assessment of road infrastructure and its use during its life’ (BTCE’90). LCCA takes into account all costs that occur over the effective life of the resource. It begins with the analyst defining a schedule that encompasses all the activities and associated costs involved over the entire project life for each design alternative. The technique of discounting is then applied so that all costs induced over several time periods are converted into present dollars and summated into a net present value (NPV) for each alternative, allowing the analyst to establish and recommend the best alternative in terms of cost. In many cases after LCCA is performed on the alternatives, it can be seen that the long term costs of operation and maintenance in fact are much greater than the initial costs of construction. LCCA is therefore an effective tool and should be applied as early on in the development of the project as possible to yield the best outcomes and allow for possible changes to occur during planning (Demos ‘06). Figure 1 shows how the cost of making changes increases, while the opportunity for savings decreases as time progresses.

Figure 1. Potential Savings And Cost Relationship; Nsw Treasury (‘04) The time value of money refers to the changing value of the dollar over time, due to the effects of inflation and interest, and is taken into account by a discount rate. ‘Inflation is the general increase in the price of goods and services over time’, whereas, Interest is somewhat the opposite and represents a ‘return on an investment’ comments the American Concrete Pavement Association (2002). The analysis period is the time frame in which the costs are compared, and is not necessarily equal to the design or service life of the pavement. Net present value (NPV) is ‘the present value of proceeds minus present value of outlays’ states Barringer (‘03), so that an effective

comparison may be made and where effectively the greatest NPV (where costs are considered to be negative) is considered the option with the competitive edge. The NPV for each alternative can be calculated using equation (1) as follows: NPV=Initial cost+∑(Future cost*(1+r)^(-N) ) (1) Of particular importance is the determination of the discount rate for the analysis, as variation of this can lead to differing results. The discount rate is a measure of the time value of money and is measured as a percentage per annum. This is often defined as the ‘actual rate of increase in the time value of money’ (Kirk ‘95), which takes into account both the interest and inflation rates. Extensive literature has been written on the determination of the discount rate and whilst much debate exists, it seems no one method has been universally accepted by construction stakeholders and civil engineers. The discount rate employed by many analysts depends largely on personal judgment along with economic and political factors, argues the BTCE (‘90). In some transportation agencies the discount rate is simply taken as an acceptable and familiar value when conducting a LCCA. Among these different agencies, the discount rate used can be seen to vary, from the Eastern state Victoria ‘VicRoads’ opting for a value of 7%, to the National Road Transport Commission (NRTC) (2009)of Australia recommending 5%. In this research, the discount rate was calculated and based on a derivation of equations which takes into account the factors of treasury-bond rate of return, inflation rate and average equity return rate. The resulting discount rate is outlined in Section 4.1. For pavements, the costs induced over the considered analysis period could be categorized as initial construction costs, maintenance costs, rehabilitation and reconstruction costs and salvage (residual) value. Initial construction costs are directly related to the construction of the pavement, including costs relating to ‘subgrade preparation; base; subbase, and surface materials; labour; equipment; drainage and the like’ (ACPA, 02). Maintenance costs can be defined as ‘actions taken to restore a system or piece of equipment to its original capacity, efficiency, or capability’ (Vanier ‘01). Typically for pavements these costs include ‘contracts, materials and equipment, staff salaries, and the like for the maintenance of the pavement surface, shoulders, striping, and drainage’ (ACPA ,02). To determine the maintenance costs, previous projects should be referred to but altered as appropriate for the proposed project at hand. When the serviceability reaches the minimum acceptable level and user satisfaction is compromised, rehabilitation works are required to increase such serviceability to an acceptable level. The need for rehabilitation works as the pavement deteriorates is shown in an example case for competing alternatives in Figure 2. Rehabilitation incorporates the costs made in the future for maintaining a pavement at a serviceable condition. Rehabilitation for concrete pavements includes methods such as full-depth repair, slab stabilization or joint and crack resealing (ACPA, 02). Rehabilitation to do with asphalt pavement may involve activities such ‘leveling and resurfacing, performed in conjunction with widening’ (Wallace 67). Reconstruction is similar to rehabilitation but relates to more extreme works similar to the initial construction and is usually implemented at the end of the service life. Residual value (or salvage value) is the estimated value of the pavement once its useful life is deemed to be complete. If parts can be recovered at the end of the analysis period, this value is considered to be positive, otherwise if demolition is required the costs of demolishing and disposal of such wastage is included as a negative value. For pavements, the residual value often bears insignificant value in the evaluation of the pavement’s worth. It is generally perceived as a small value once discounted to present value, considering that over the life of the alternative, the costs incurred would be much greater than the salvageable portions at the end of its useful life, argues BTCE (‘90). Finally, the intangible user costs are argued by Smith & Walls (’98) to be the ‘costs incurred by the highway user over the life of the project’. These include user

delay costs, vehicle operating costs, and accident or crash costs (ACPA ‘02). Although applicable to pavement engineering these are often excluded in the analysis as they are difficult to prepare to a dollar value.

Figure 2. Rehabilitation Works For Competing Alternatives; Smith (98) The requirement for a LCCA methodology comes as a result of the prevalent issue that designers often face regarding their recommendations to the client for the best, cost-effective solution in the long term. Providing an estimate of the initial cost of purchase followed by the associated follow-on costs can be time-consuming and costly, unless a reliable and efficient methodology is followed. Thus, LCCA is a technique that allows all these costs over the analysis period to be considered in an organised manner that ultimately allows comparison of competing alternatives.

METHOD In addition to the extensive background research pertaining to LCCA, the research presented here developed a spreadsheet to reinforce the value of applying LCCA in today’s Australian pavement engineering industry. A mixed methods approach was used to carry out the study, encompassing mainly quantitative research with elements of qualitative analysis. Qualitative measures in the form of objective industry comment were required for an understanding of the applicability of LCCA for local pavements. A more quantitative approach was necessary in the development of the LCCA spreadsheet itself, especially in the collection of cost data. In addition to this data acquisition, quantitative methods, as stated by Creswell (05), entailed the explanation of ‘how one variable affects another’ and addressed the need to relate respective cost variables in order to ultimately calculate a net present value for each pavement alternative, therefore allowing direct comparison. A new, unique LCCA model/spreadsheet was developed using Microsoft Excel for this West Australian infrastructure case-study (off-the-shelf LCC software was deemed too indistinct for the specific in-depth study proposed here, not least due to an inability of existing software to calculate the discount rate). The framework developed by the authors enabled each component of the analysis to be tied together to generate a desired outcome of which alternative would prevail as the most cost effective. Development of such a model required the inclusion of the theories of time value of money by the use of the discount rate (and appropriate sensitivity analyses) to achieve reliable results. Formulae depicting the effect of the time value of money based upon secondary research were programmed into the spreadsheet accordingly. The design of the spreadsheet allowed different cost components over the useful life of the pavement to be entered under respective appropriate times within the life cycle, whereby an automated analysis consequently revealed the desired, cost effective alternative. The user-friendly model developed here, is deemed a tool to allow replicable analysis.

In order to conduct the study several primary information sources were sought; quantitative data was obtained not only from industry standards/building-cost-information-services (BCIS) but also from specific contacts within the Roads and Traffic Authority (RTA) New South Wales, for typical cost values for both alternatives of asphalt surfaced (thin asphalt surfaced unbound granular pavement (AC)) and concrete pavement (continually reinforced concrete pavement (CRCP)) designs. In addition to quantitative data acquisition, qualitative research methodology (via semistructure stakeholder interviews) clarified, maintained, confirmed and validated the specification(s), cost and maintenance data generated. As discussed above a key component required for this research was the assessment of an appropriate discount rate, deemed to play a significant role in the outcomes of the life cycle cost analysis. Components of inflation rate, treasury-bond rate, and average equity return rates were necessary for specific discount rate calculations (section 4.1 below). By adopting a mixed methods approach, the methodology for this study is argued as suitable to achieve successfully the objectives of: justifying the use of LCCA in the civil engineering industry; as well as, providing a specific recommendation to pavement design engineers in relation to the competing alternatives of asphalt compared with concrete pavements. RESULTS AND DISCUSSION 1) Discount Rate: The discount rate was calculated as 8.0% which necessarily takes into account the factors of inflation and interest and: a 5.44% treasury-bond rate of return, based on a reasonable 10 year yield (Bloomerberg ‘08); the inflation rate as depicted by the Reserve Bank of Australia’s website of 1.5%, and confirmed by Words (07); and, the average equity return rate assessed as 13.6%, taken from Investment Wise (‘09). These values allowed the discount rate to be calculated as 8.0%, when inserted into equation 2 to equation 4 below, where risk is assumed to be half that of the average risk premium discount rate. The resultant discount rate calculation is argued as robustly accurate. It is noteworthy that the overall outcome of the LCCA was not jeopardized when a sensitivity analysis was applied to this discount rate. The following logical applies: No Risk Return = Treasury Bond Rate of Return – Inflation; Average Risk Premium Discount Rate = Average Equity Return – Treasury Bond; and, Discount Rate = No Risk Return + 0.5*Average Risk Premium Discount Rate. 2) Overall Comparison: In the overall comparison of net present values of asphalt and concrete pavement alternatives (thin asphalt surfaced unbound granular pavement (AC) and, concrete pavement continually reinforced concrete pavement (CRCP)) considered over the analysis periods of 30, 60 and 100 years, the asphalt alternative prevailed as the obvious cheaper alternative in all cases when considering the time value of money. The summary of these results are below in Table 1. Table 1. Summary Of Results Option / Type of Pavement

Construction Cost ($/m2)

Maintenance Rehab/Recon 2 Costs ($/m ) costs ($/m2) 30 year Analysis Period 1 AC1 74.20 14.52 0 2 CRCP2 133.00 2.45 0 60 year Analysis Period 1 AC 74.20 15.33 6.04 2 CRCP 133.00 2.87 1.48 100 year Analysis Period 1 AC 74.20 15.46 6.31 2 CRCP 133.00 2.91 1.51 1 thin asphalt surfaced unbound granular pavement (AC) 2 continually reinforced concrete pavement (CRCP)

Salvage Value ($/m2)

NPV ($/m2)

0 0

88.72 135.45

2.65 -0.74

98.22 136.61

2.77 -0.75

98.74 136.67

The remainder of the results discussion focuses on the analysis period of 60 years as this was deemed sufficient to include the design lives of both concrete and asphalt pavements. Moreover, the results for the 30 year and 100 year analysis periods ultimately achieved the same outcome. The overall net present value (NPV) at the end of the 60 year period revealed the cheaper alternative to be the asphalt pavement option, as depicted in Figure 3. Figures and tables should appear in numerical order, be described in the body of the text and be positioned close to where they are first cited. Make sure all figures and tables will fit inside the text area. Please ensure all text inside figures is legible; minimum of 8pt equivalent is required.

Cumulative Cost,  $/m^2

160 140

concrete

120

asphalt

100 80

Asphalt Concrete

60 40 20 0 0

5

10

15

20

25

30

35

40

45

50

55

60

Years

Figure 3. Cumulative Discount Costs (NPV) For Analysis Period Of 60 Years 3) Cost Input Data: In the overall comparison of net present values, the pie-charts for the 60 year analysis period shown in Figure 4 below, show significant differences in the cost components for each alternative. The initial cost for the concrete pavement accounted for half of the total costs, while for asphalt the initial costs were only 20% of the entire life cycle costs. Maintenance for asphalt conversely was double that of concrete’s (this was a sub-hypothesis assumed at the beginning of the study). The results confirm that here, higher costs at the construction phase of the pavement lead to less maintenance costs over the life of the specification in question (in the case of the concrete pavement, higher costs lead to less maintenance works over the life of the pavement). The reconstruction was more expensive for asphalt at $132/m2 compared to the concrete reconstruction price of $70/m2 reflecting the higher involvements in the removal and replacement of the more extensive layers associated with the asphalt alternative.

residual

Initial cost

recon maint residual

Initial cost

maint

Figure 4. Asphalt V Concrete Total Costs For Analysis Period Of 60 Years 3.1) Initial Costs: The initial costs for the asphalt and concrete alternatives consisted of the costs involved to construct each option and showed a (somewhat anticipated) large variation. The costs for the concrete alternative were seen to be significantly greater, by 44%. The costs of concrete initial construction almost doubled that of asphalt and therefore can be presumed to be a major factor that deters pavement agencies to choose this option Table 2. Initial Costs. Initial Costs ASPHALT Description 175 limestone sub-base 230 base-course Two coat emulsion seal Tack coat 30 dense graded asphalt 30 open graded asphalt

TOTAL

Rate ($/m2) 13.77 30.84 6.76 0.34 13.59 8.90

74.2

CONCRETE Description 150mm sub-base LMC including finishing, curing/provision of typical quants of sub-grade beams for accesses & intersection 240mm continuously reinforced concrete pavement: supply & place concrete, & longitudinal joints and slab anchors Finish cure and texture base Supply and place steel reinforcement TOTAL

Rate ($/m2) 35

60

3 35 133

3.2) Maintenance Costs: The maintenance costs formed a bulk of the overall costs over the life of each alternative, as depicted in the bar graphs from Figure 5. Results show that the asphalt alternative (represented by blue) is more regular with higher costs involvements. This is supported by the fact that asphalt pavement options historically require more routine repairs to fix deterioration such as potholes and other structural failures. Concrete on the other hand with its higher durability and strength characteristics can withstand the elements and the same amount of traffic loading without the extensive restoration techniques required for asphalt pavement. Peak maintenance costs occur at year 15 and 55 where asphalt resurfacing at $28/m2 is required due to construction issues that mainly affect the base. The event occurs also at year 30, this time costing $14/ m2 due to fact that the previous major asphalt resurfacing would have enhanced the quality of the pavement. The regular maintenance works evident every 5 years are the heavy patching works to restore the surface failures. For the concrete alternative the maintenance works are far more spread between works. For the first 25 years, the maintenance works only occur at years 10 and 20 at a rate of $2.25/m2 and $4.50/ m2 respectively for base replacement due to construction activities. The relatively small rates are

reasonable since the concrete has not yet experienced significant deterioration, given its characteristics of long life. As time progresses, the concrete deteriorates and requires more stringent maintenance strategies, where at year 49 nearing the end of the concrete’s life expectancy, the concrete pavement receives its final maintenance strategy at $5.85/m2 in preparation of the reconstruction which occurs at year 50.

Figure 5: Maintenance Costs For Analysis Period of: 1-20; 21-40; 41-60 Years

3.3) Reconstruction Costs: The reconstruction cost for the concrete pavement was $70/m2 occurring at year 50, at the end of its design life cycle. The reconstruction cost for the asphalt pavement conversely carried a much higher rate at around $132/m2. Higher costs are associated with the

asphalt pavement after a long period of time in which high amounts of deterioration deem the pavement no longer appropriate to simply apply routine maintenance works. 3.4) Residual Costs: The asphalt residual value was calculated as approximately -$58/m2, representing an addition to the overall costs for the asphalt total life cycle cost; the cost of reconstruction was significantly higher than the initial costs of construction (in Western Australia opportunity/ability to recycle/re-use asphalt is limited). For the concrete pavement the salvage value was assumed to be $35/m2 occurring at the time of reconstruction, which indicates that the total costs were reduced by this amount. This was based on RTA’s advice that the concrete pavement’s reconstruction involved re-using part of the existing pavement for the sub base (RTA, personal communication August 31, 2009).

CONCLUSIONS LCCA provides guidance for design engineers (and construction stakeholders) to make objective decisions when faced with opposing alternative specifications. The degree of current local uncertainty around asset management and the need for building professionals to also act as (predictive) financial advisors gives due reason for a framework to enable identification of the most cost effective design-solution. The easy to use LCCA framework developed here for the Western Australian infrastructure industry does indeed enable whole-of-life costs to be systematically input, and allows ultimately the summation of these into a single net present value at a base year, for comparison and hence identification of the cheaper alternative from competing options. Two major benefits were attained by completing the research study described here. The first was to prove that LCCA is a valid means for the assessment of alternative component specifications in terms of whole-cost. A second key achievement is argued to be the contribution made in the objective recommendation of a particular pavement type from assessed alternatives, for Western Australian roads. More particularly the research conducted and described here shows that implementation of the developed LCCA model recommends that an asphalt pavement is cheaper than a concrete pavement for heavy use/loading highways. This ultimately may be viewed as an important finding for transport agencies in Western Australia; in other words there is now sufficient objective evidence to support the continued use of asphalt pavement types on highways locally.

REFERENCES American Concrete Pavement Association ACPA. (2002) Life Cycle Cost Analysis: A guide for Comparing Alternative Designs. Volume 1. Illinois: American Concrete Pavement Association. Barringer, H. P. (2003). A Life Cycle Cost Summary. International Conference of Maintenance Societies, Perth. Bloomerberg, L. D., and M. Volpe. (2008). Completing Your Qualitative Dissertation: A Roadmap From Beginning to End. California: Sage Publications. BTCE. 1990. Pavement Management: Development of a Life Cycle Costing Technique, edited by B. o. T. a. C. Economics. Canberra. Australian Government Publishing Service. Creswell, J. W. (2005). Educational Research: Planning, Conducting, and Evaluating Quantitative and Qualitative Research. Second Ed. New Jersey: Pearson Education.

Demos, G. P. (2006). Life Cycle Cost Analysis and Discount Rate on Pavements for the Colorado Department of Transportation. Colorado: Colorado Department of Transportation. http://www.dot.state.co.us/publications/PDFFiles/discountrate.pdf (accessed May 15, 2009). Investment Wise. (2009). (1), http://allianceinvestments.com.au/files/16KH315JV7/Investement%20Wise%20Vol%208%Issue% 201.pdf (accessed August 22, 2009). Kirk, J., and A. J. Dell'Isola. (1995). Life Cycle Costing for Design Professionals. Second ed. New York: McGraw-Hill. Life Cycle Costing (1996). National AustStab Guidelines. http://www.auststab.com.au/pdf/nat02.pdf (accessed August 22, 2009). New South Wales Treasury, (2004). Total Asset Management: Life Cycle Costing Guideline. First ed. New South Wales Treasury. Sharp, D. R. (1970). Concrete in Highway Engineering. Sydney: A. Wheaton and Co. Smith, M.R., and J. Walls / Federal Highway Administration. (1998). Life Cycle Cost Analysis in Pavement Design: In Search of Better Investment Decisions. Volume 1. Washington: US Department of Transportation. Vanier, D. J. (2001). Why Industry Need Asset Management Tools. Institute for Research in Construction. http://irc.nrc-cnrc.gc.ca/pubs/fulltext/nrcc44702/nrcc44702.pdf (accessed July 14, 2009). Wallace, H. A., and J. R. Martin. (1967). Asphalt Pavement Engineering. New York: McGraw-Hill Book Company. Words, F. (2007). Consumer Price Index (CPI) - Australia. http://www.fxwords.com/c/consumerprice-index-cpi-australia.html (access August 22, 2009).