Summary
FLOU Consultants Report
Second Opinion on Cross-Border Freight Flow Analyses
Reference Case: The Nordic Connector (NC)
It is
laudable that an independent party was commissioned to review the NC travel
forecast. That task is rarely undertaken but should be done for all projects and
plans. Many forecasts are made by ‘black boxes’ few understand.
Finnish
Transport Infrastructure Agency and WSP consultants (FTIA/WSP)
freight forecast for Nordic Connector is deficient for two reasons: Nordic
Connector is considered just a link between Vaasa and Umeå regions. And the
choice of forecasting model systems for freight and passenger demand models
required intensive data work to obtain a trip matrix for model calibration.
Shortcomings in data assembly and calibration required “expert judgment”
corrections in freight volume forecasts.
Evaluation
of many alternatives compounds the work required to prepare the data and “run”
the models. With good scoping before deciding on the alternatives, the number
of alternatives could have been cut to two or three for thorough analysis. FTIA
is silent on project procurement, which is an important matter.
FLOU
acknowledges the importance of input data, but explanations for data “fudging”
are insufficient or absent. That concerns all data: population and employment
and the network. Description of NC as a link between Vaasa and Umeå is not
justified. Restriction on the forecast traffic flows on NC is not examined. Network
description for the models is severely limited.
FLOU recognizes that FTIA/WSP forecast contains many simplified assumptions and has questionable deficiencies. FLOU’s review treats these shortcomings lightly. FTIA/WSP forecasts 645 daily vehicles in normal circumstances, and 1350 daily vehicles if the southern ports are closed. The corresponding NCAT (NordicConnectoAnttiTalvitie) forecasts are 1550 and 1711 daily cargo units. Passenger volumes are about in the same ballpark in both FTIA/WSP and NCAT forecasts, 3500 -4900 daily vehicle trips.
Summary
of FLOU’s evaluation of the NCAT forecast
Unlike in the
FTIA/WSP forecast, NCAT forecast uses a combination of three methods, which
FLOU labels as “unsupported assumptions”. It will be shown that this
characterization is wrong.
The first and most
important reason is “reference class forecasting”, the second is traffic route
choice based on international experiences for roads, and the third is the use
of demand elasticities. Specifically, FLOU evaluates NCAT forecast
to have the following shortcomings:
1. The
forecast is a “scale analysis” for freight and elasticities for passenger
travel.
2. Use of
NC for freight is based on assumed diversion from other possible routes.
3. Freight
and passenger demand growth is based on assumed and unsupported growth rates
4. Freight
to Sweden via Haparanda and freight to Europe via Estonia are not included.
Claims
for these shortcomings are turned back with justifications and details. It is
claimed that the NCAT freight traffic forecast is theoretically sound and, if
anything, err on the downside. Travel forecasts for an important link were made
without running a complex, data-intensive, and time-consuming travel demand
model systems. The route choices are based on available data-supported modeling
evidence. NCAT forecast is well-grounded on Öresund Bridge reference-case
experiences and experiences about travel demand elasticities. There are no
“unsupported assumptions”. The FTIA/WSP model results, uncorrected or corrected
by “expert judgment”, are less than NCAT forecasts. For the most part this due
to NCAT’s different (and more realistic) understanding of the trade and
transport geography, which had led FTIA to define a transport network that
serves to limit NC’s origin/destination-reach in Finland’s foreign trade.
The passenger travel forecasts are in the same ballpark; there is much
uncertainty and underestimation is likely.
Even for the high case, no capacity problems are indicated on the NC. NCAT
forecast also identifies issues that are problematic and would benefit from a
more in-depth analysis.
FTIA/WSP
does not have a benefit-cost analysis. NCAT does; it is useful also for financial
decisions.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.