· Evaluation of alternative technical solutions to the collection of road tolls at designated locations in Greater Vancouver,

· Project appraisal in an urban corridor,

· Policy evaluation for low-traffic-volume and medium-traffic-volume rural roads, and

· Urban transportation system plan evaluation to meet provincial and regional policy objectives in Greater Vancouver.
 
 

Costs are broken down into (a) road user plus agency costs (vehicle operation, infrastructure and parking, travel time and accident costs), and (b) environmental impacts. The analyses agreed on the total user plus agency costs because the estimation methodologies are firmly established by ministry procedures and the unit cost data is "hard" and available (Bein et al. 1994). The differences in vehicle operation costs and accident costs between the two analyses reflect a different approach to accounting for auto insurance premiums. Bein’s estimate included them in the total social cost of traffic accidents, while KPMG included them in the expenses of vehicle users.

The difference in estimates of the costs of environmental impacts arose because each study applied different approaches to the monetization of externalities. KPMG’s estimates of air pollution, global warming and noise impacts were obtained using the control cost approach, whereas Bein used damage cost methods. (A discussion of these methods is in Chapter 2). The reasons for the different cost estimates for local air pollution, global warming, ozone depletion, noise and barrier effects are reviewed below.

Bein's 1993 estimate of local air pollution was based on the California experience, which showed that fine particulates rather than ground-level ozone cause the dominant damage in urban areas and adjoining regions (Hall et al. 1992). Hall et al. found that the PM10 damage costs were 2.7 times higher than the ground-level ozone damage costs. Bein calculated ground-level ozone costs in the Lower Mainland ($153 million), based on Pace University Center (1993) and G.E. Bridges and Associates (1991) unit costs for ground-level ozone precursors (NOx and VOC) and found the PM10 damage costs ($153 million times 2.7 = $413 million). Bein's 1993 local air pollution calculation also included road dust costs of $100 million, based on Pace University Center (1993) and Sælensminde (1992) unit cost data for visibility loss and cleaning costs, respectively. The local air pollution impacts thus summed up to $0.8 billion total. Bein's 1996 re-estimate used costs specific to the Lower Mainland for PM2.5 (Chapter 4.1.1) and ground-level ozone damage (Chapter 4.1.4), yielding $0.35 billion total. Visibility and road dust cleaning costs were not included, as no local cost data was available.

KPMG (1993) lumped local, regional and global air pollution costs into one figure ($0.4 billion). Bein's 1993 estimate of global warming damage costs was based on a simplistic precautionary calculation of a unit cost of $700/t CO2 for a long-term warming by several degrees causing damage of the order of 20% of world GDP. Bein's 1996 estimate used a refined value of $1 000/t CO2 (Chapter 4.1.2) and made a 44% allowance for the greenhouse gas emissions from upstream energy use (Chapter 3.4.1).

By comparison, KPMG’s estimate of global warming costs was based on a control cost of sequestration of carbon by tree planting in British Columbia. The control cost used as a shadow price does not incorporate a number of components included in the damage cost approach, such as the irreversibility of global climate change, residual damage remaining after the trees have sequestered the small fraction of emissions, the international equity question, and the greenhouse gases embedded in vehicles and in the added infrastructure required by automobile-dependent urban sprawl.

According to KPMG, traffic noise and barrier effects are not a problem in the study area. By contrast, Bein’s estimate based on Norwegian urban externality costing (Sælensminde 1992) shows these cost categories to be comparable to the local air quality problem. Only the arterials were assumed to have the necessary combination of traffic volumes and speeds to produce noise levels in excess of 50 dB(A). Based on traffic noise curves in the MoTH (1992) manual, Bein (1993) assumed that traffic noise along the arterials is on average 14 dB(A) higher than the 50 dB(A) threshold, and that each excess 1 dB(A) causes a loss of property value of 0.6%. The properties behind the ones exposed on each side of the arterials were included. The loss of value of the properties affected by the excess noise was amortized over 10 years and the equivalent annual value was calculated and multiplied by the factor of six from Sælensminde (1992) to arrive at the traffic noise damage cost of $720 million per year.
 

Table 5.1 demonstrates that different approaches to monetization can produce significant disparities in estimates of the environmental costs of transportation, and may lead to differences in policy. KPMG, in using control costs to monetize externalities, followed prevailing consulting practice. Bein’s estimates embrace a wider range of costs, in an effort to ensure that ecological and societal costs were more fully accounted for in transportation planning. Noting that neither study included some significant costs, such as most land use impacts and the degradation of biodiversity, it is possible to argue that the more precautionary approach may better serve the aim of achieving a sustainable transportation system.
 

The evaluation considered capital and operating agency costs of each option, road user costs (travel time and vehicle operation), and global, regional and local air emission costs. Traffic noise was not considered since the collection of tolls was examined only for bridge and tunnel sites in Greater Vancouver that are isolated from more developed areas. The possible increased frequency of mostly minor traffic accidents was also ignored. The impacts of travel time differences between options on traffic redistribution in the network were not considered.
 
 
 

Table 5.2 demonstrates the small impact of monetized damage due to NOx and HC, relative to monetized global warming damage. Neither of the estimates fully accounts for loss of species, degradation of biodiversity and for other intangibles. The results, however, clearly show that options A and B are inferior to the fully automated option C.

It is important to realize that the above conclusion would be reached even if environmental costs were not considered. Because the bulk of incremental costs is in road user travel time, and because emissions in this case are correlated with travel time delays, the addition of emission costs to NPV does not change the ranking of the options.
 
 

Agency, user, noise and emission costs were included (Table 5.3). The impact on parking requirements for the destinations of traffic going through the facility and the relative impact of the corridor improvement on land use were evaluated qualitatively. The cost of particulate damage was based on an aggregate estimate by SENES (1994) for the region, distributed per kilometre travelled by light and heavy vehicles. The heavy vehicle emission rate of particulates was assumed to be one order of magnitude larger than for light vehicles. Greenhouse gas emissions were calculated by a method similar to that in Section 5.2 and a unit cost of $700 per tonne CO2-equivalent was used.
 

Travel time savings dominate the benefits, and the project would still be viable in the absence of environmental benefits. This is often the case in urban transportation improvements involving alleviation of peak traffic vehicle volumes by diversion of travellers to more efficient modes, such as buses and HOV in this case. The capacity is actually larger than in the base case, although the throughput of vehicles per peak hour is not changed much. This is an example of making better use of an existing facility.

Cost benefit analyses of transportation projects, especially in urban areas, are sometimes criticized for overstating the travel time value. In the Hastings Street case, the travel time savings are robust. The project would still be viable if all the environmental benefits and vehicle operating savings were not considered, and if the value of time was reduced from $6 per hour to less than $3 per hour for passengers. Adding the environmental benefits would enhance the viability of the project.

A number of non-monetized benefits contributr to project justification. If more motorists leave their vehicles at home, the demand for parking space at their destinations will lessen, while local traffic noise and pollution are also reduced, if not in absolute terms, then per traveller. Because the improvement does not significantly change long-distance commuting time, it is not believed to significantly increase urban sprawl.
 
 

In 1992, approximately 52% of all traffic did not experience any delays. The traffic was free-flowing to semi-dense, with volumes less than 500 vehicles per hour. With a typical traffic growth rate of 3% per annum, and in the absence of geometric, safety and capacity improvements (which were not costed), the average operating speed of traffic would drop by about 5 km/h between 1998 and 2007, affecting an increasingly larger number of vehicles. Between 1999 and 2007, the proportion of vehicles forced to stop and go would increase from 0.9% to 7.5%. This is in contrast to high-volume roads in urban areas where the majority of traffic operates under impeded flow conditions.

The consequence of the reduction in operating speed would be the following annual losses per kilometre of highway in both directions combined during the period 1999-2007 (expressed in 1991 dollars):

These costs add up to a total annual cost of $713 000 per kilometre, not counting the additional fuel consumed, other vehicle operating costs and increased local air pollution in communities traversed by the Trans Canada Highway. Calculated over the eight-year analysis period with an 8% discount rate, the present value of these costs amount to approximately $145 million. The extra costs of increased CO2 emissions amount to only about 2% of the total.

In summary, if capacity on a typical rural two-lane highway is not improved, the level of service on the rural highway system will deteriorate due to impeded flow conditions, and additional annual road user and environmental costs of $713 000 per kilometre will arise. This additional cost is over and above any user and environmental costs experienced currently. If aggregated over the whole provincial network, deteriorated service due to congestion on low-volume sections of rural highway could incur significant costs. However, the question of whether government decides to maintain the road infrastructure at its current or designated level of service will have to weigh the capital costs against the potential savings in the road user and environmental costs.
 
 

Table 5.4 shows two of the policies analyzed. Policy 1 involves more frequent HIPR than Policy 2. Both policies save agency costs when compared to the typical mill-fill-overlay. Policy 1 saves more in construction costs relative to the base case. Policy 1 produces near-zero road user cost savings, because it interferes with traffic more than Policy 2. For the same reason, Policy 1 produces more emissions than Policy 2.

The choice of policy depends on the level of traffic volume on the road. For lower volume roads, Policy 2 is more economical since it has a higher net present value. For higher volume roads, Policy 1 is more viable. For both policies at the higher traffic volumes, the emission costs accruing from traffic impediments lowers the NPV. Thus, even though HIPR saves on the construction materials and disposal costs associated with mill-fill-overlay, there are substantial traffic-related environmental costs which must be considered.
 
 

For the multiple account evaluation, the quantifiable criteria were as follows:

The analysis quantified the following environmental aspects:

Traffic speed representation in EMME is deficient for the determination of fuel consumption, vehicle operating costs and emissions, all of which strongly depend on vehicle speeds. These accounts would be inaccurate unless the speeds predicted by EMME were corrected. A method was developed for deriving traffic speeds from traffic volume outputs of EMME, based on speed-flow relationships observed on typical facilities in the study area. For each type of road and mode, a speed-flow curve was derived from empirical data (Kawczynski 1996). EMME traffic volume output is fed into an appropriate curve and the corresponding vehicle speed is read out for given type of facility. This method provides more reliable data than untreated speed output from EMME.

The MAE framework, process and analysis details are described by Kawczynski and Bein (1996). The following examples (Bein and Kawczynski 1996) illustrate the use of MAE in appraisals of:

One of the facilities being considered, the East-West Connector, would provide a missing link in the regional road network. From a larger number of feasible solutions, four options were short-listed and evaluated against a "do minimum" scenario in the year 2021:

The capital and operating costs are those of the facility only, while revenues are system-wide. The protected areas account also refers to the facility only. All accounts, except capital and operating costs, protected areas and the qualitative accounts were quantified for the a.m. peak hour only. To allow cost-effectiveness comparisons, the financial accounts and the economic value of protected areas were transformed to a.m. peak hour equivalents.. The results of the quantitative accounts are summarized in Table 5.5. For comparison, the accounts are also shown for 1994, representing present conditions.

Total user costs and revenues are much higher than the monetized total environmental costs. On purely economic grounds, the environment cannot "compete". The environmental costs, however, are likely understated and do not include non-use values and intangibles. Unless the decision-making framework takes this into account, the environment does not have a chance if compared dollar for dollar with the financial and user cost accounts. Options C and D, for example, encroach on protected areas, but their total environmental costs are not significantly different from Option A.

One striking conclusion from the financial accounts is that equivalent hourly cost of the facility is a small fraction of the hourly revenues expected system-wide from just one morning peak hour in 2021. If an option was more expensive than others in terms of financial costs, but provided better user and environmental benefits, it would be worth financing through increased user or environmental charges.

Evaluation of System Investment Options

Different combinations of major road facilities were assembled together with other committed projects. Each combination forms one highway investment option, and each one was examined regarding performance of the system. Each investment option is the sum total of many projects starting in different years, each project having its own stream of capital costs over a time period. Medium-range investment horizon up to years 2006-2008 was considered, since planning longer term is too uncertain. Two highway investment options (Option 1 and 2) and two other were analyzed as follows:

The system-level accounts are similar to the facility-level analysis. Land Use Goal is a qualitatively describes the risk of not attaining compact metropolitan area objectives. Non-attainment might be possible if the SOV mode continues to get infrastructure and fiscal policy support. Protected Areas account was not considered since facility options entering the investment options were already screened for this criteria at the facility level. Transportation Energy was not accounted for, either, since it is strongly correlated with Greenhouse Gas Emissions account. All system performance results were expanded from the morning peak hour model results to 24-hour system operation, since neither the afternoon peak nor the off-peak period are modelled by EMME.
 
 

Table 5.6 summarizes the results. The MAE reveals key aspects of the transportation system performance, allowing informed decisions on trade-offs between accounts and between options, even though the information produced by EMME model is imperfect. All options show considerable user benefits over the Base Case and all have revenues exceeding financial costs–largely due to the TDM policies. The beneficial impacts of the transit and TDM options are clear. Higher road investments generate more user benefits, but the largest user savings go with the transit investment Option 3. A large part of benefits is due to TDM, clearly demonstrating the importance of demand-side management.

The bulk of monetized environmental benefits is from greenhouse gas emissions avoided. Total environmental benefits (monetized benefits only) tend to increase with the road infrastructure capital costs as well. Option 1, the most costly infrastructure investment, produces almost $1 billion in environmental benefits, while Option 2 (the least costly investment option)–about $0.75 billion. However, investment in transit only (Option 3), produces $2.2 billion in environmental benefits. Doubling the investment in transit, if accompanied by attracting SOV users to transit, could bring the environmental benefits to the level of user benefits. When the Land Use Goal is included in the MAE with a high weight, it further supports Option 3 as the preferred one.

As in the facility-level MAE, the monetized environmental account has little chance when competing dollar for dollar with the total user benefits. Both account values, however, increase with the road investment. All investment put in transit generates the highest overall monetized benefits, albeit at a higher cost. Evaluations in which the user and environmental accounts do not reinforce each other, should not lump the monetized values together. In any case, the non-monetized environmental accounts should be described to make informed judgements possible.