## Sustainable Bridge Procurement Considering LCC, LCA, Lifespan, User-Cost and Aesthetics

# Case Study

Trafikverket is currently planning to build a wildlife crossing bridge (total width 35 m, total length 64 m) over the European route E6 in Gothenburg. In addition to allowing animals to freely pass over the E6, it will connect or reconnect habitats and help avoid collisions between vehicles and animals. At the bridge location the E6 has two lanes in each direction carrying 55,000 vehicles/day, 5% of which are trucks. Trafikverket is planning to procure this bridge through a D-B contract. Thus, they commissioned a consultant to prepare a conceptual design for this bridge (Proposal 1, Figure 1), to attach to the tender documents. The following analysis investigates the LCC-effectiveness of the consultant’s solution, clarifies how this bridge could be procured through the holistic approach proposed here and addresses the roles of the agency and the contractors in the approach.

## Life-Cycle Cost Analysis

### Step 1: Exploration of technically feasible proposals and their anticipated INV costs

There are not many wildlife overpass bridges in Sweden, which complicates attempts to conceive and tabulate technically feasible proposals. However, information and drawings related to existing road bridges with similar dimensions extracted BaTMan using its navigation tool “WebHybris”, together with drawings and photos of bridges with similar dimensions in other countries, assisted this work. Based on this information, two other proposals, in addition to Proposal 1, are considered technically feasible (Figure 1).

Figure 1. Technically feasible proposals and their outlines

In an early investment phase, it might be difficult to accurately assess the INV cost of the technically feasible proposals. Table 1 presents the unit cost (Swedish Krona “SEK”/m^{2}, 1SEK is equal to 0.109 Euro, 0.136 US Dollar in November 2014) and anticipated INV cost of each proposal in 2013, based on statistical analysis of cost records related to existing road bridges registered in BaTMan with similar dimensions. The real INV cost is expected to be lower, since wildlife overpass bridges are designed to carry considerably lower loads than road bridges. The use of cost records is intended to roughly anticipate the INV costs of new similar designs rather than compare the real INV costs that would never be known unless the contractors’ bids are received. The INV cost of a bridge design could substantially be varied depending on many parameters such as the construction method, design details, profit margin, conditions at the bridge location, etc. The uncertainty in INV costs will not significantly affect the LCC benchmarks considering the proposed approach as it will be clarified later.

Table 1. The proposals’ anticipated initial investment costs and associated target quantities

### Step 2: Preliminary quantification of the proposals’ BSMs

The BSMs and elements required for each proposal (based on an intensive review of drawings of similar bridges in BaTMan) that would incur LCM costs during the bridge’s life-cycle are also listed in Table 1. These quantities are preliminary. Therefore, contractors under the proposed approach are requested to supplement their bids with more precise BOQs quantifying the BSMs and elements comprising their proposed designs.

### Step 3: Formulation of a life-cycle plan for each proposal

Based on information given by Mattsson (2008) on technical and real lifespans of bridge types in Sweden, Proposals 1 and 2 were assigned a technical service lifespan of 100 years and Proposal 3 a lifespan of 80 years. Table 2 presents the LCMs commonly performed on BSMs comprising bridges under similar conditions in Sweden with those lifespans, and associated unit costs. The bridges in BaTMan are broken down into 14 BSMs (Trafikverket 2004). Different bridge types have different numbers and quantities of those BSMs. The LCMs’ costs Table 2 were extracted from BaTMan’s latest updated bridge LCMs price list (Trafikverket (2013), in which 14 bridges’ structural members are listed together with comprehensive compilations of actions applied to them and their average unit costs. The fixed costs stated in Table 2 are quantity-independent and mainly related to the establishment and traffic control costs prior to implementation of the LCMs. The probability of a LCM’s necessity refers to the likelihood that it will be required. Thus, the NPV of a LCM equals to present values of that LCM multiplied by its necessity probability. The actions’ times and necessity probabilities in Table 2 are based on regression analysis and analysis of variance of intensive historical data related to similar actions performed on BSMs composed on similar bridges extracted from BaTMan using the Webhybris tool, taking conditions at the bridge location into account. Intensive repair records related to all BSMs comprising all bridges constructed in Sweden after 1950 were collected and analyzed. Detailed information about the analyses of those records and the recommended lifecycle plans of the various BSMs comprising new Swedish bridges are presented by Safi (2013).

It is important to acknowledge that the lifecycle plans generated considering statistical treatment of historical records will inherently be uncertain and do not necessarily represent the actual LCMs that will be implemented during the life-cycle of new bridges. New BSMs designed to meet new standards may have greater durability than old BSMs. Thus, the LCMs performed on old BSMs might not necessarily be needed for new BSMs. However, that uncertainty will not significantly affect the analysis results considering the proposed procurement approach, partly because the NPV of LCMs cost is a small portion of the total LCC as it will be highlighted in the analysis below, and lifecycle plans are applied to BSMs or target-quantities regardless the type of the proposals compressing them as shown in Table 2. The major factors affecting the LCCA results are the numbers and quantities of the BSMs required for each proposal (Table 1).

Table 2. Life-cycle measures applied in the life cycle plans

### Step 4: Preliminary LCCA comparison

The LCCA was based on the proposals’ BOQs and LCMs (Table 1 and Table 2, respectively). Figure 2 presents the proposals’ anticipated INV costs and the NPV of their LCM cost at real discount rates of 2% and 4%. The proposals’ LCC shown in Figure 2 and the below comparisons are preliminary since they are based on uncertain INV costs. Since the proposals have unequal lifespans, the EAC was used to identify the anticipated most LCC-efficient proposal.

Figure 2. The proposals’ anticipated INV costs, LCM costs, LCC and EAC at indicated real discount rates

Figure 2 also presents results of a sensitivity analysis addressing the impact of varying the discount rate on the preliminary LCC-effectiveness of the proposals. Their total LCC NPV and LCC EAC are shown by the dashed curves (see right y axis and upper x axis for values). According to the EAC curves, regardless of the discount rate, Proposal 3 is expected to be the most LCC-efficient and Proposal 2 the least LCC-efficient. Theoretically, applying a high discount rate will favor investment alternatives with low INV costs, short lifespans and high LCM costs, and vice versa. A discount rate of 4% should be considered in such LCCA according to Trafikverket’s 2013 rules. However, intensive review of an array of annual price lists for bridges’ LCM costs published by Trafikverket detected an average inflation rate of 2%, which should be included in the analysis. One way to do this is to consider the real costs of the LCMs at their implementation times instead of the present costs. The other practical way is to deduct the inflation rate from the discount rate. Thus, a real discount rate of 2% is used here.

At a real discount rate of 2%, the anticipated net savings resulting from implementing Proposal 3 rather than Proposals 2 and 1 are 17.62 and 11.23 million SEK, respectively for a lifespan of 80 years, and the opportunity loss resulting from implementing Proposal 1 rather than Proposal 3 is 12.18 million SEK for a lifespan of 100 years.

### Step 5: Interpolation of results and preparation of LCC-efficient benchmarks

Since the preliminary LCCA (step 4) showed that Proposal 3 is expected to be the most LCC-efficient at a 2% real discount rate, Trafikverket could procure it through a traditional fixed-design approach. However, this procurement method is strongly not recommended as it may often result in selection of a sub-optimal design, for three main reasons. Firstly, the INV costs included in the preliminary LCCA analysis are considerably uncertain, reflected by the prediction intervals in Table 1. Secondly, conceiving all the bridge designs that could provide feasible solutions in a given location is difficult, so the fixed design stated in the tender documents might not be the most cost-efficient. Thirdly, detailed drawings and specifications are required, which are costly, time-consuming and beyond the capability of many agencies to produce. Therefore, the method we proposes is to allow contractors to offer the designs they prefer while meeting the functional demands and design standard requirements, but incorporate each design with an LCC added-value.

In this step, two levels of LCC-efficient benchmarks are computed: proposal-level and BSM-level LCC added-values. Since the proposals have unequal lifespans, their LCC added-values account for differences in both their LCM costs and lifespans. The dashed curves in Figure 3 depict the LCC added-values of Proposals 1 and 2 relative to Proposal 3 at different discount rates (with values indicated by the right y axis and upper x axis). As shown in this figure, the LCC added-values are inversely related to the discount rate. Figure 3 also presents the classification of the LCC added-values of Proposals 1 and 2 relative to Proposal 3 at different discount rates. At a real discount rate of 2%, the LCC added-value of Proposal 1 relative to Proposal 3, expressed as (1,3) in this figure, is 1.11 Million SEK includes a +0.41 Million SEK LCM cost component and -1.54 Million SEK lifespan component. This implies that the cost equivalent of the last 20 years of Proposal 1’s lifespan, after Proposal 3’s 80-year lifespan, is only 1.54 million SEK. However, the cost equivalent of these 20 years will be equal to 2.46 Million SEK at a real discount rate of 2% if Proposal 1 is used as the reference proposal instead of Proposal 3. Therefore, it is highly recommended to consider the most LCC-efficient proposal identified in Step 4 as the reference design in such analysis. The lifespan contribution to the added-values of Proposal 1 and 2 relative to Proposal 3 is close to zero at an 8% real discount rate (Figure 3).

Figure 3. The proposals’ LCC added-values relative to Proposal 3 at indicated real discount rates and their lifespans, and LCM cost added-values contributions at these rates

### Structural-members’ LCC added-values

The LCC added-values of the BSMs can be computed using equation (2). In order for contractors to compute the lifespan-adjusted added values of BSMs for their proposals using this equation, the technical service lifespans of the various bridge types have to be clarified by agencies. This has been discussed by Mattsson (2008) in relation to Swedish bridges’ conditions. The anticipated INV cost of the reference proposal should also be stated in the tender documents so the contractors can determine the second term in equation (2). Contractors can then calculate the LCC added-values associated with their proposals by multiplying the BSMs’ added-values given in the tender documents by the respective quantities in their proposals. In our case study, assume that a contractor wants to propose a different design, Proposal 4, similar to Proposal 1 but with two simply supported spans composed of prefabricated concrete beams. A technical lifespan of 100 year is considered for this option. Assume that this proposed design has no integrated breast walls. Hence, expansion joints at both ends are included, and double the number of bearings at the intermediate support, in that proposal. Table 3 presents the added-values associated with each of the BSMs at a real discount rate of 2% that should be stated in the tender documents, as both unit and fixed costs. The fixed cost is quantity-independent and mainly related to the establishment of work-zones and traffic-control costs prior to implementation of the LCMs. Table 3 also presents the LCC added-value computed by a contractor for Proposal 1 using the LCC added-values given in Table 3, which is the same as the one computed by the agency (Figure 3), and LCC added-values computed for Proposal 4.

Table 3. Structural members’ LCC added-values at real discount rate of 2%, and Proposals 1 and 4’s LCC added-values relative to Proposal 3 computed by a contractor using added-values for structural members given by an agency in the tender documents.

## Bridge User Costs

Since it is an animal overpass bridge, only the underpass traffic is considered in the following WZUC analysis. A preliminary construction method and TCP based on standard bridge construction methods in Sweden were formulated for each of the proposals presented in Figure 1. However, contractors under the proposed approach are required to describe the construction methods that would be used to deliver their designs, the time needed for construction, and plans to control the underpass traffic during construction. The supports of Proposals 1 and 2 could be built without disturbing the underpass traffic. The steel superstructure of Proposal 2 could be built on one side of the bridge location then launched from that side without disturbing the underpass traffic. Scaffolding is needed to construct the reinforced concrete superstructure of Proposal 1. The traffic in one direction could be temporarily directed to two temporary lanes built beside the two lanes carrying traffic in the other direction. The estimated cost of these temporary lanes is 1.5 million SEK, which should be included in the INV cost of Proposal 1. Falseworks, traveling gantries or girders could possibly be used to avoid the need for temporary lanes. The superstructure of Proposal 1 would take an estimated 60 days to construct, in average conditions, if the concrete is cast on-site. Alternatively, it could be formed by prefabricated concrete members, which could be lifted by cranes, thereby cutting the construction time and allowing traffic to pass under the bridge during construction. The underpass traffic would be disturbed for an estimated three weeks during the construction phase of Proposal 3. The cost of providing two temporary lanes in each direction should be included in its INV cost. For Proposal 2, painting the steel superstructure (in years 25, 50 and 75), which would require establishment of a movable work-zone for two weeks on each occasion, is the only LCM assumed to disturb the traffic. Based on the formulated TCPs for each proposal, the underpass traffic should flow freely under work-zone conditions, and the estimated delay for each vehicle passing the bridge location, relative to the normal travel time (estimated using the InfraLCC program) would be 21 seconds on average. Anticipated WZUC based on these assumptions are presented in Figure 4.

The estimated WZUC incurred by a one-day traffic disturbance during construction is 61,000 SEK (Figure 4), but it could be much higher if there is no extra space in the bridge location to build the two temporary lanes when closing one direction. However, most new bridges in Sweden are built as parts of new road or rail corridors, thus no WZUC are usually incurred during the construction phase. The dotted and dashed curves in Figure 4 (see the right y axis and upper x axis for values) respectively depict the anticipated NPV of the WZUC and the EAC of the LCC associated with the various proposals at different discount rates, based on the construction methods and TCPs mentioned above. Thus, those results might differ during the bids evaluation process depending on the actual construction methods and TCPs described in the contractors’ bids. The discount rate variation does not affect the WZUC NPV of Proposals 1 and 3 since the LCMs of these two proposals are considered to have no effect on the underpass traffic. The anticipated LCC EACs of the proposals presented in Figure 4 include their WZUC, INV, LCM and TCP costs. As shown in Figure 4, at a real discount rate 2%, although Proposal 3 has a greater WZUC NPV than Proposal 2 it remains more LCC-efficient than Proposal 2 since it has a lower LCC EAC.

Figure 4. The anticipated NPV of the proposals’ WZUC and total EAC at indicated real discount rates and the WZUC caused by a one-day traffic disturbance during their construction

The WZUC incurred by a one-day traffic disturbance during construction (Figure 4) and user-cost EAC of the reference proposal should be clearly stated in the tender documents, thereby allowing contractors to use Equation (3) to determine their respective proposals’ user-cost added-values relative to the reference proposal (Figure 6). This will stimulate contractors to minimize the WZUC by selecting designs, construction methods and construction staging that minimize traffic disruption.

## Integration and Evaluation of Aesthetic Aspects

For Trafikverket preliminary investigation (data not shown) indicates that a general aesthetics-demand factor of 10% to 50% could be acceptable, thus 30% is used in the case study below. Accordingly, the most aesthetically pleasing design could ultimately be chosen even if it is 30% more expensive than the aesthetically-worst alternative. The lower section of Table 5 presents results of a sensitivity analysis addressing the impact of varying this parameter on the LCC-effectiveness of the proposals.

Four items are considered in the evaluation of the bridge site class, shown in Table 4. A bridge site class is assigned to each item based on the corresponding argument shown in Table 4, while the overall site class is determined as the average of those four classes. Thus, the bridge site is assigned to Class II, “Demanding”, corresponding to a 67% site-class factor.

Table 4 The process used to evaluate the bridge site class

Consequently, the aesthetics WTPE computed using equation 4, would be 3.66 Million SEK, considering a value of 0.30 for Trafikverket’s general aesthetics-demand factor. The aesthetics WTPE should be stated in the tender documents. The items to be evaluated and their weight factors (Table 5) are based on recommendations in several bridge aesthetics guidelines (ETSI Project 2009; NSW RTA 2003). The total weight of the considered items has been set here at 100, but the user could change these values, neglect some items or add others, as appropriate for a particular case.

Table 5. Items considered for aesthetic evaluation, their weight factors and evaluation results

Generally bridges seem more pleasing if they have a simple form, thin deck (relative to the span), continuous structural lines and structural members with shapes reflecting the forces acting on them. The authors of this paper have individually evaluated each proposal considering the outlines given in Figure 1. However, contractors are encouraged to supplement their bids with more detailed drawings and possibly 3D models for the evaluators to thoroughly rank their aesthetic merit. The evaluation points in Table 5 are averages of four evaluators’ scores, which have been used, with equation (5), to compute an aesthetic coefficient for each proposal and CEAM values using equation (6). The results indicate that Proposal 3 is the best aesthetically (Table 5). Since the evaluators regard all the proposals as generally beautiful, none of them would consume the full amount of the aesthetics WTEP, and the difference in the CEAM of Proposals 3 and 2 is just 2.09 million SEK. The full amount of the aesthetics WTEP would be consumed only if all items considered for evaluation are assigned +2 for the aesthetically-best alternative and -2 for the aesthetically-worst alternative, which is a very rare case. This shows that attractive bridges are not necessarily much more expensive than ugly bridges.

Considering the sensitivity analysis shown in the lower section of Table 5, regardless of the general aesthetics-demand factor, Proposal 3 is the most cost-efficient and Proposal 2 the least cost-efficient. The variation of the aesthetics-demand factor has a more significant influence on Proposal 3 and 1 than Proposal 2. This is because of the intermediate aesthetic coefficient value (almost zero) of Proposal 2. Increasing the value of the aesthetics-demand factor has only increased the cost-gap between the proposals. Thus, in this case study, the aesthetics-demand factor does not have a considerable impact on the proposals’ cost-effectiveness. However, it will have a more significant impact in other bridge cases in which the bridge site is assigned to Class I “very demanding” and the proposals’ aesthetic coefficients defer considerably.

To strengthen the proposed approach, agencies should establish lists of items to be evaluated and their weight factors, include them in tender documents, and also standard lists assigning specific points to various shapes or designs of BSMs, so contractors can pre-estimate the aesthetic coefficients of their proposals.

## LCA Evaluation and Integration

### Life cycle impact assessment (LCIA)

The LCA in this case study follows the framework proposed by Du (2012), Du and Karoumi (2014) and applies the Matlab-based tool developed in the cited study. Since the bridge is in an early planning phase, the analysis covers only the manufacture of materials listed in the lower section of Table 1, traffic delays during construction and maintenance activities. The life-cycle inventory data used are related to average European conditions, and were extracted from the Swiss center for life cycle inventories (Ecoinvent v2.2), European reference Life cycle database (ELCD) and world steel association databases. The environmental releases from construction materials have been quantified by the life-cycle inventory aggregated data, which cover all the raw materials, upstream processing, and energy utilized in each process. Traffic delays are based on assumed extra distance equivalents due to speed reductions for 16-32 t trucks and petrol-fueled passenger cars. Table 6 presents 12 mid-point environmental impact categories of related atmospheric, water and solid releases, which are required for full characterization of proposals’ environmental impact: global warming (GWP), Ozone depletion (ODP), Human toxicity (HTP), Photochemical oxidant formation (POFP), Particulate matter formation (PMFP), Ionizing radiation (IRP), Terrestrial acidification (TAP), Freshwater eutrophication (FEP), Marine eutrophication (MEP), Terrestrial ecotoxicity (TETP), Freshwater ecotoxicity (FETP), Marine ecotoxicity (METP), where the P in each case refers to potential.

### The characterized environmental impact

Table 6 also gives the monetary weighting of each proposal based on their characterized environmental impacts, while Figure 5 depicts environmental impact contributions of their material components. The results show that Proposal 3 provides the best environmental performance, e.g. 46, 44, 35 and 23% lower GWP, HTP, POFP and PMFP, respectively, than the least environmentally friendly proposal (1), mainly because of the latter’s larger steel reinforcement and concrete requirements and longer traffic delays. The concrete, steel reinforcement and steel structure requirements for both Proposals 1 and 2 contribute most (82% on average) to the investigated impact categories, followed by traffic delays (37% for Proposal 1 and 20% for proposal 2) in the POFP category. The steel structure and filling material components of Proposal 3 respectively account for 46% and 38% of the total impact in the investigated categories.

Figure 5. Proportional contributions of indicated components for each proposed bridge design to their total environmental impact in the indicated categories

The results also indicate that minor structural components and work (bearings, painting and asphalt) have negligible impacts, and that GWP, HTP, POFP and PMFP are the main impact categories (Table 6). Since the proposals have unequal lifespans, the environmental impact coefficient has been based on the lifespan of the reference proposal (80 years).

Table 6. Impact categories, the proposals’ total impact and integration results

The environmental results presented in Table 6 have been aggregated by Ecovalue08 monetary weightings (Göran et al. 2013) and updated Ecovalue12 values for global warming, particulate matter formation and exotoxicological impact categories (Markus and Finkbeiner 2011). As yet no monetary weighting factors for the ODP, IRP, TETP and FETP categories are available, so they have been excluded from the total monetary impact cost estimates of the proposals presented in Table 6. The impact categories considered, their monetary weighting factors and the environmental WTPE should be stated in the tender documents, allowing contractors to estimate the CEEI of their proposals. The presented LCA results are not perfect since they are based on rough assessments of the proposals’ BOQs (Table 1). Therefore, under the proposed approach contractors should supplement their bids with more precise BOQs specifying the BSMs and elements required for their proposed designs.

## Bid Evaluation and Preliminary Results

Agencies should update the analysis after the contractors’ bids are received in which the designs’ actual INV cost, BOQs, TCPs and external shape are identified. Considering the preliminary assessment of those parameters, Figure 6 presents the anticipated INV costs, LCC added-values, user-cost added-values and cost equivalents of the aesthetic merit and environmental impact of the considered proposals. The cost of the two temporary lanes required during the construction of Proposals 1 and 3 have been added to their anticipated INV costs. The figure also presents the anticipated net equivalent LCC of the proposals computed using equation (1). Proposal 3 seems to be the most Sustainable (life-cycle efficient), if the anticipated INV cost matches the real INV cost offered in a contractor’s bid. As shown in Figure 6, the proposals’ INV costs offered in the contractors’ bids remain the main parameter guiding the final decision. Based on Figure 6 and equation (1), a contractor who submitted a proposal identical to Proposal 3 would be awarded the contract even if the price of his bid reached 35.54 Million SEK instead of 18.29 Million SEK, provided that other contractors offered higher prices for either of the other two proposals presented in Figure 6. Furthermore, the real INV costs of Proposals 2 and 1 must be less than 16.7 and 17.32 million SEK to be regarded as the most LCC-efficient. Professional contractors should be aware, as far as possible, of the bid evaluation criteria to allow them to perform such analysis before preparing and submitting bids.

Figure 6. Life-cycle aspects’ contributions and net equivalent LCC costs of Proposals