A Methodology for a Pavement Resurfacing Strategy to Minimize Life-cycle Costs and Greenhouse Gas Emissions
- Author(s): Lidicker, Jeffrey Roger
- Advisor(s): Madanat, Samer
- Horvath, Arpad
- et al.
In recent decades pavement management optimization has been designed with the objective of minimizing user and agency life-cycle costs. However, pavement management decisions also have significant impacts on life-cycle energy use and environmental emissions from pavement management activity and user vehicles. This study expands beyond optimizing pavement rehabilitation strategy for minimization of life-cycle costs to also include greenhouse gas (GHG) emissions. We extend previous work on the single-facility, continuous-state, continuous-time optimal pavement resurfacing problem to solve the multi-criteria optimization problem with the two objectives of minimizing costs and GHG emissions.
The balance between the potentially two different optimal rehabilitation policies is found through the use of a Pareto frontier, which exists in the span between the cost- and emission- optimal strategies. The Pareto frontier provides decision makers with the dollars per tonne of GHG emissions saved due to a change in rehabilitation strategy. Results using California data indicate that there is a tradeoff between costs and emissions when developing a pavement resurfacing strategy, providing a range of GHG emissions reduction cost-effectiveness options.
Case studies for a two-lane arterial and a ten-lane major highway in California are presented, where traditional hot-mix asphalt overlays are applied. The 2011 case studies are particular to California by traffic loadings and pavement durability. However, the user and agency emission and cost estimations are based on national data. Thus, generalizing the case study results should be subject to these caveats. An ordinary medium-volume metropolitan state-designed road and an extremely heavily traveled highway bearing commuters into and from San Francisco are optimized as representative situations.
Results for a one-kilometer segment of Interstate 80, in Berkeley California, with ten lanes and 273,000 light-duty vehicles and 13,100 heavy-duty vehicles per day, indicate that the life-cycle cost minimum occurs when asphalt overlays are applied every 15 years or equivalently when the pavement roughness reaches an international roughness index (IRI) of 2.7 m/km. Coincidentally, this is the same roughness Caltrans uses to decide when to apply an overlay for the state's roads. However, where any of the conditions or characteristics for any pavement segments are different, the coincidence may cease to exist. The minimum life-cycle cost at this optimal pavement rehabilitation strategy is approximately $490,000 per kilometer per year, at which point resurfacing activity and user vehicles would emit approximately 220 tonnes of CO2 equivalents per kilometer per year. The GHG emissions minimum corresponds to an overlay interval of 22 years or the equivalent threshold roughness IRI of 3.4 m/km. The minimum GHG emissions (user emissions plus agency emissions) are approximately 200 tonnes of CO2 equivalents per kilometer per year, with life-cycle costs (user costs plus agency costs) at approximately $520,000 per kilometer per year. Agency and user emission (and cost) estimates each change in opposing directions when overlay intervals change. Thus, when each of the like agency and user attributes are added together, a minimum is guaranteed to exist.
Any pavement rehabilitation strategy that makes use of overlay intervals outside of this sub-interval defined by the life-cycle cost and GHG emissions optima are trivial in that any strategy change designed to reduce costs also reduces emissions. However, inside this special sub-interval, any change in strategy that reduces costs will increase emissions and vice versa. Thus, this special sub-interval constitutes a Pareto frontier of optimal solutions where tradeoffs are associated with each change. For example, if Caltrans is currently operating at the life-cycle cost minimum by applying an overlay interval every 15 years but decides to reduce emissions by changing the interval to every 18 years, there will be a reduction in emissions. However, it will come at a total life-cycle cost of approximately $500 per tonne of CO2 equivalents. Of course different pavement rehabilitation strategy changes will present different cost-effectiveness ratios. If the change spans points outside the Pareto frontier, the costs may be minimal or even negative. However, within the Pareto frontier, attempts to save even more emissions will increase costs per unit of CO2 equivalents saved.
If a market value for CO2 exists, then a unique optimal pavement rehabilitation strategy is defined. On the Pareto frontier is every possible market value as the negative of the slope of the tangent line to each point on the curve represents a market value starting with zero dollars per tonne of emissions (cost minimum) to infinite dollars per tonne of emissions (emissions minimum).
Results for a two-lane arterial road segment, also in Berkeley, which has only 25,000 light-duty vehicles and 480 heavy-duty vehicles per day, indicate that similar pavement rehabilitation strategy overlay intervals are optimal. For the life-cycle cost minimum, a 16-year overlay interval is optimal, which corresponds to a threshold roughness IRI of 2.1 m/km. For the GHG emissions minimum, an overlay interval of 25 years and its associated threshold roughness IRI of 2.5 m/km are optimal. Although the overlay intervals are not that different from the larger ten-lane interstate highway case, the threshold roughness values are more favorable in the two-lane case. The life-cycle cost minimum occurs at approximately $80,000 per kilometer per year and is associated with approximately 51 tonnes of CO2 equivalents per kilometer per year. The GHG emissions minimum occurs at approximately 47 tonnes of CO2 equivalents per kilometer per year and is associated with approximately $86,000 per kilometer per year.
A sensitivity analysis on model input parameters revealed which parameters required the best accuracy and shed light on policy decisions. Pavement deterioration rate, within a 20% variation, had a relatively little effect on outcomes. This indicates that uncertainty around the pavement deterioration rate is not very important. However, a small change in vehicle miles traveled had a large effect on outcomes. Other results highlighted the contrast between strategy decisions for various pavement and vehicle technologies. For example, it is found that, in both case studies, improving vehicle fleet fuel economy will save total (tailpipe plus pavement) emissions. The two-lane case showed a larger percentage of relative reduction in emissions but the ten-lane case was found to have a larger total reduction in emissions. However, an improved fuel economy for the vehicle fleets means that the effect of roughness on fuel consumption is less. Thus, the GHG emissions associated with pavement management become a larger share of the total emissions. This means that at the emissions optimal, the pavements are allowed to become rougher before being rehabilitated again. Thus, to counteract the expected fuel economy improvements of the future, the use of new technologies that reduce emissions associated with pavement overlay activity, but also reduce roughness at optimality, is paramount. For the same reason, technologies that provide more durable pavements are also encouraged.