Scheme Appraisal for Highway Projects 57the input to COBA. For example, if a choice is required between route A,
route B and neither, then the costs and benefits of neither would be subtracted
from each of the A and B valuations before the cost-benefit computation is
made.
The output from COBA is used to contribute to the following type of
decision:
To assess the need for improving a specific highway route. The improvement
could involve either the upgrading of an existing roadway or the construc-
tion of a completely new one
To determine what level of priority should be assigned to a particular scheme
by considering its economic return relative to those of the other viable
schemes in the area/region being considered by road administrators
To determine the optimal timing of the scheme in question relative to other
road schemes in the area
To aid in the presentation of viable highway options to the public within a
formal consultation process
To establish optimal junction and link designs by comparing the economic
performance of the options under consideration.
The extent to which a full COBA analysis can be undertaken for a particular
scheme depends on the stage reached in the assessment process, the data avail-
able to the decision-maker and the nature of the decision to be taken. As the
design procedure for a particular scheme advances, a more refined economic
analysis becomes possible.
Within COBA, in order to compute the three benefits accounted for within
the procedure (savings in time, vehicle operating costs and accident costs) the
program requires that the number of each category of vehicle utilising the link
under examination throughout its economic life be determined using origin and
destination data gathered from traffic surveys.
The inputs to the COBA analysis are hugely dependent on the output of the
traffic forecasting and modelling process outlined in the previous chapter. It
assumes a fixed demand matrix of trips based on knowledge of existing
flows and available traffic forecasts where travel demands in terms of origin
and destinations and modes and times of travel remain unchanged. This
assumption has the advantage of being relatively simple to apply and has been
used successfully for simple road networks. It has difficulty, however, in coping
with complex networks in urban areas or in situations where congestion is
likely to occur on links directly affecting the particular scheme being assessed.
This has a direct effect on the traffic assignment stage of the traffic modelling
process, which is of central importance to the proper working of the
COBA program. In the case of complex urban networks, where urban schemes
result in changes in travel behaviour that extend beyond simple reassignment of
trips, more complex models such as UREKA have been developed to predict
flows.