TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Transportudvalget 2019-20
TRU Alm.del Bilag 67
Offentligt

Study of dynamic effects of the FBFL

which has not been considered

in the FTC study

Report

October 2019

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Dingolfinger Straße 2

81673 München

Contact:

Dr. Markus Schubert

T +49 (89) 45 91 1127

[email protected]

Study of dynamic effects of the FBFL

which has not been considered

in the FTC study

Client

Femern A/S

Copenhagen, Denmark

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Content

Background of the study

Study approach

Analysis of existing traffic in Europe

Application of the model

Summary and conclusion

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

BACKGROUND OF THE STUDY

In the FTC

study the main demand effects for the FBFL result from

•

replacement of the existing ferry line between Rødby and Puttgarden by the FBFL:

traffic

overtake

•

traffic from other routes, mainly Great Belt, Gedser

–

Rostock and other ferry lines to the

route via FBFL:

route choice effects

•

traffic from other modes, for passenger traffic mainly air traffic between Hamburg and

Copenhagen to land based traffic via FBFL, for goods traffic mainly from road to rail:

modal-

split effects

Apart from that an "induced traffic" has been calculated for passenger traffic only:

•

additional trips respectively more frequent trips from the same origin to the same destinations

primary induced traffic).

The number of

trip generators

(inhabitants, for freight: producers/shippers) or trip attractors

(workplaces, inhabitants, tourist sites, for freight traffic: recipients/consumers)

was kept

constant

between reference case (without FBFL) and planning case (with FBFL).

The Danish transport consultant COWI recommended 2015 in the external quality assurance

the FTC study

among others an

assessment of

newly generated traffic

by the FBFL

by new

opportunities for economy, trade, tourism and housing which was not covered by the FTC study

due to the lack of effective tools for predicting the

potential dynamic effects

of the link.

Intraplan Consult GmbH and BVU Beratergruppe Verkehr + Umwelt GmbH: Fehmarnbelt Forecast 2014

- Update of the FTC-Study of 2002, on behalf of Femern A/S 2014

COWI: Ekstern kvalitetssikring af den opdaterede trafijprognose of Femern Bælt-projektet, comissioned

by the Ministry of Transport, November 2015

see footnote 1

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

STUDY APPROACH

Infrastructure projects lead among others to "induced

traffic".

Induced traffic is defined as traffic

which would not take place without the projects, neither on other routes, with other modes or to

other destinations.

There are

two main categories

of "induced traffic:

(1)

Additional or more frequent trips or transports

to existing destinations

resp. attractores

and

from existing generators

(= induced traffic in the narrower sense or

primary

induced traffic)

because travel times or travel costs ("resistance") are reduced by the

project

(2)

Additional trips or transports due to effects of the project on local/regional economy,

housing, tourist sites, logistics sites, etc. that means due to

changes in the numbers of

generators and/or attractors

(= indirect or

secondary induced traffic)

In the FTC study only (1) has been considered in the passenger traffic part.

. In the freight traffic

part no induced traffic has been calculated (because here there are doubts that this kind of

effects, primary induced traffic without changes of the generators or attractors, is existing)

. But

neither for passenger traffic nor goods traffic (2) the "secondary induced traffic" has been

considered resp. calculated as COWI has stated in its review of the FTC-forecast.

Indeed, it is conventional wisdom that transport infrastructure in general and special transport

projects have considerable effects on economy, employment, trade, tourism and settlement

consequence to a better accessibility of the regions in the influence of the project. In the case on

hand a better accessibility

between

the regions north and south of the FBFL should be relevant.

These factors, again, would generate traffic, "secondary induced traffic" or (to make a difference

to the term "induced traffic" as used in the FTC study)

"generated traffic by economic effects

from the project".

Intraplan Consult GmbH and BVU Beratergruppe Verkehr + Umwelt GmbH: Fehmarnbelt Forecast 2014

–

Update of the FTC-Study of 2002, on behalf of Femern A/S, 2014, S. 74, 118f

Intraplan Consult GmbH and BVU Beratergruppe Verkehr+Umwelt GmbH: Verkehrsverflechtungs-

prognose 2030 Los 3: Erstellung der Prognose der deutschlandweiten Verkehrsverflechtungen unter

Berücksichtigung des Luftverkehrs, Ergänzender Bericht zur Methodik, on behalf of the German

Ministry of Transport and digital Infrastructure, 21.11.2014

Copenhagen Economics Aps and Prognos AG: Economy-wide benefits Dynamic and Strategic Effects

of a Fehmarn Belt Fixed Link; Report prepared for the Ministry of Transport, Denmark, and the Federal

Ministry of Transport, Building and Housing, Germany, June 2004

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

However, it is very difficult to measure these effects in terms of values and numbers. To

measure economic effects of a single transport project and to estimate the effects on traffic by

these economic effects is quite a challenge, because accessibility is obviously an important, but

by far not the only location factor for business and settlement.

In all the available resp. substantial studies positive economic effects of important transport

projects have been found and the effects are considerable:

•

In the case of the

Channel tunnel

Ernest & Young

found a strong effect of this connection

on trade and tourism.

•

For the

Öresund Bridge

ex-post socio-economic assessments have been made

, showing a

strong effect on commuter traffic, settlement and economic growth in the neighbour regions

Copenhagen and Malmö stimulated by the new connection.

•

Also for the

Great Belt Bridge

strong "wider economic effects", here effects on commuting

and other agglomeration effects, have been found.

•

For the

transalpine rail tunnels

in Switzerland considerable regional economic effects from

better domestic and

–

even more important

–

better international accessibility caused by the

tunnels have been found.

•

In a more general and less project specific way also in

Germany

considerable "wider

economic effects" resp. stimulation of economic growth by investment in transport

infrastructure have been found.

The

results from these studies are highly significant but in our view far from being

complete and comparable

And, in none of these studies which give scarce indicators about the economic and settlement

effects of the infrastructure project there are indicators about the

after-effects

of these

on traffic

and transport,

which is the question here.

Ernst & Young: Economic footprint of the Channel Tunnel fixed link

–

An analysis of the economic value

of trade and passenger traffic travelling through the Channel Tunnel, October 2016;

less optimistic however sn the study of the University of Kent, Centre for European, Regional and

Transport Economics (Alan Hay, Kate Meredith, Roger Vickerman): The impact of the Channel Tunnel

on Kent and relationship with Nord-Pas de Calais, June 2004

M.Aa. Knudsen, J. Rich: Ex post socio-economic assessment of the Oresund-Bridge, 2012

Copenhagen Economics: Bredere økonomiske effekter af transport-investeringer, DEBATOPLÆG

udarbeijdet for Transportministeriet, Maj 2014

Schips/Hartwig (KOF at the ETH Zurich): Wachstumswirkungen und Rentabilität von

Verkehrsinfrastrukturinvestitionen

–

Stand der Forschung und wirtschaftspolitische Schlussfolgerungen,

on behalf of Schweizerische Bau-, Planungs- und Umweltdirektorenkonferenz, 2005

See for example: RWI: Verkehrsinfrastrukturinvestitionen

–

Wachstumsaspekte im Rahmen einer

gestaltenden Finanzpolitik, on behalf of German Ministry of Finance, 2010: Dependent from the

economic lifetime an investment of 1 billion € leads to a macroeconomic effect of 0,8 to 4,2 billion €.

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Given these double uncertainties, no clear or countable effects of the project on economy and

settlement and no "rules" for after-effects of the latter on traffic and transport,

we decided to

apply another approach:

•

To derive the dynamic effects by analogies from traffic and transport analyses with respect to

the

correlation between accessibility

("gravitation")

and traffic intensity

(transport science

approach) (see figure 1)

Fig. 1: Basic approach

The accessibility gains are calculated by a

gravitation model:

if travel distance and time are

reduced gravitation is growing. This leads to more traffic.

In detail the approach is outlined in figure 2. The rules which are analyzed by the approach are

applied using the FTC-model with regard to the base data, traffic and socio-economic drivers

and with regard to the gains in travel time resp. reduction of Generalized Costs caused by the

project FBFL.

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Fig. 2:

Method to estimate additional traffic caused by the economic effects of the FBFL

This approach is more substantial than other methods if reasonable "benchmark" data are

available. This is the case: There are well founded origin-destination matrices available for the

Bundesverkehrswegeplanung for the international traffic and transport between the German

regions and all foreign regions in Europe including for eight neighbouring countries.

Combining these empirical based matrices with network models and zonal socio-economic data

gravitation functions

can be derived, for passenger traffic as well as for freight transport:

•

gravitation model for international passenger traffic

Intraplan has access to well-founded data for even more country-country-pairs: from Netherlands to

Belgium/France/United Kingdom, from Austria to Italy/Czech Republic, Slovak Republic, Slovenia,

Hungary, from Switzerland to France/Italy, from Denmark to Sweden, from United Kingdom to France.

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

•

gravitation model for international freight transport

•

gravitation model for total (domestic and international) passenger traffic (for comparison)

•

gravitation model for total (domestic and international) freight traffic (for comparison)

The principle of such a gravitation function is shown in figure 3.

Fig. 3:

Gravitation model: principle

Applying these functions for the FBFL project resp. the difference between reference case

(without FBFL) and with case (with FBFL) the "theoretical" demand effects of a project can be

calculated (see fig. 3). Two more working steps are necessary in this approach:

•

calibration: if the existing traffic on Rødby

–

Puttgarden is not in line with the curve, the curve

has to be adjusted (calibrated).

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

•

deduction of the effects considered in the FTC study: in the FTC study among others

(primary) induced traffic as described above is considered. This has to be subtracted from the

results to avoid double counts.

Apart from the (primary) induced traffic (formula see footnote 13) which was applied only for

passenger traffic in the FTC study

no gravitation model

was applied due to the fact that it was a

corridor study without reference to the overall traffic (complete OD-matrices for Europe).

ANALYSIS OF EXISTING TRAFFIC IN EUROPE

In the context of the German Bundesverkehrswegeplanung detailed origin-destination-matrices

have been set-up for 2010, widely on the basis of empirical data:

•

for the passenger traffic: OD matrices using traffic counts (among others on borders), surveys

for national and international tourism on regional level, commuter statistics, surveys on

business travel, etc.

•

for goods traffic: OD-matrices based on samples of transport flows per haulage firm

The matrices for the base year 2010 are differentiated on NUTS 3 level and cover most of

Europe due to the central location of Germany.

These matrices have been combined with network models and zonal data for population and

GDP. By that analyses in the way as shown in fig. 3 were carried out.

Formula in the FTC-model to calculate (primary induced traffic

−

* min(R

) * AR

ind

max(GK

; GK

)

with

ind

induced trip per mode, purpose and OD-relation ij

Generalized Costs (GK) in the planning case

Generalized Costs (GK) in the reference case

trips in the planning case

trips in the reference case

Share of Generalized Costs (GK) on the total activity costs of the journey (dependent on

the trip purpose)

Intraplan Consult GmbH and BVU Beratergruppe Verkehr+Umwelt GmbH:

Verkehrsverflechtungsprognose 2030 Los 3: Erstellung der Prognose der deutschlandweiten

Verkehrsverflechtungen unter Berücksichtigung des Luftverkehrs, on behalf of the German Ministry of

Transport and digital Infrastructure, June 2014;

for more detail see: the same: Ergänzender Bericht zur Methodik, 21.11.2014

see footnote 14

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

For passenger traffic gravitation has been defined as a function of the population of the regions

of origin and destination and the Generalized Costs as proxy for the "resistance" resp. costs to

overcome the distance between the regions. For goods traffic the regional GDP has been

chosen as "masses" in the gravitation model.

The generalized costs are specified for passenger and freight traffic and have been derived from

the network models.

The following curves show the functional interrelationship between gravitation and traffic resp.

transport intensity.

For

total passenger traffic

the curve is shown in figure 4.

TRAFFIC INTENSITY

1) passengers/year (per 1000 inhabitants

origin

x 1000 inhabitants

destination

)

Fig. 4:

Correlation between Generalised Costs (GK) and traffic intensity (here: passenger

trips/year)

–

total passenger traffic

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

The gravitation model

with:

traffic between origin and destination

��

= ��

��

∗ ��

��

∗ ��

��

population in the zone origin (in 1000)

population in the zone destination (in 1000)

Generalized Costs between origin and destination

(in €)

gravitation exponent

has a gravitation exponent of

–

1,268 and a good regression coefficient r

with 0,87.

For

international passenger traffic

(see figure 5) the gravitation coefficient is at

–

1,462 and

the regression coefficient r

is even higher (0,95). The reason for that is the generally larger

range of distances in international traffic in Europe and therefore a higher number of (very) small

traffic intensity values compared with domestic traffic. This leads to an increase of statistical

correlation in the international traffic part. At the same time international traffic intensity is more

dependent on the transport "resistance" than domestic traffic.

TRAFFIC INTENSITY

1) passengers/year (per 1000 inhabitants

origin

x 1000 inhabitants

destination

)

Fig. 5:

Correlation between Generalised Costs (GK) and traffic intensity (here: passenger

trips/year)

–

international passenger traffic

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

For

road freight

traffic the results have a similar shape. In figure 6 the transport intensity for

road transports (domestic and international) are shown.

TRANSPORT INTENSITY

1) here: 1000 tons road/year per (gross value added

origin

(in million €)

x gross value added

destination

(in million €))

Fig. 6:

Correlation between Generalised Costs (GK) and traffic intensity (here: tons/year)

–

total road transport

The model formula is the same as in passenger traffic, with different variables

with:

��

= ��

��

∗ ��

��

∗ ��

��

transport between origin and destination in 1000 tons

gross value added in the zone

origin (in million €)

gross value added in the zone

destination (in million €)

Generalized Costs between origin and destination

(lorry traffic, in €)

gravitation exponent

The gravitation exponent is

–

0,823 with a regression coefficient r

of 0,80.

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

For

international transports

(see figure 7) the gravitation exponent is

–

1,077 with again as in

passenger traffic a higher regression coefficient r

of 0,89 and a higher dependency on transport

resistance compared to overall transports.

TRANSPORT INTENSITY

1) here: 1000 tons road/year per (gross value added

origin

(in million €)

x gross value added

destination

(in million €))

Fig. 7:

Correlation between Generalised Costs (GK) and traffic intensity (here: tons/year)

–

international road transport

APPLICATION OF THE MODEL

The model has been applied in the following way:

��

= (

with

��

��,��

��

��,��

− 1) ∗ ��

��

OD,F

generated traffic between origin and destination

model transport intensity between origin and destination in the case with FBFL

(same for freight with t

OD,F

)

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

OD,R

model transport intensity between origin and destination in the reference case

without FBFL (same for freight with t

OD,R

)

FTC

traffic between origin and destination according to the FTC model (year 2030)

Applying this model for a reference year 2030

there is a generated

passenger traffic

2.398.000 trips (see table 7). From this figure 778.000 trips have to be subtracted which have

been already considered in the FTC-study.

The remaining "secondary" induced traffic which

was not considered in the FTC study would be 1.620.000 trips or

13,2 % related to the FTC-

results.

trip purpose

FTC study

for 2030

(1000 trips)

1.604

775

874

1.372

1.085

1.704

1.368

3.227

12.009

generated

trips

(1000 trips)

324

701

396

324

288

293

2.398

thereof

already in

FTC-study

274

167

117

778

total trips

FTC +

generated

1.850

1.202

1.103

1.414

1.292

1.937

1.592

3.237

13.629

% increase

to FTC

results

15,3

55,1

26,2

3,1

19,1

13,7

16,3

0,3

13,2

Business

Day Commuter

Weekend Commuter

Shopping

Other daytrips

Visiting friends/

relatives

Short holidays

Holidays

Total

Tab. 7:

Generated trips by intensified interaction

With regard to trip purposes the biggest effects are for day and weekend commuting followed by

business and "other day trips". These results especially with regard to commuting are in line with

empirical observations in recent decades in the context of the realized projects Öresund

Bridge

and Great Belt Bridge

Complete OD-matrices and network models are not available for a later year

In the FTC study 2014 an induced traffic of only 113 thousand trips have been shown (table 6-10 on

page 140). However, this includes negative effects of walk-on-passengers due to the stop of ferry

services (see tab. 6-8). Considering that induced traffic is at 530.000 trips in the FTC study. This figure

is related to 2022, updated to 2030 (see Intraplan Consult GmbH and BVU Beratergruppe

Verkehr+Umwelt GmbH: Verkehrsprognose für eine Feste Fehmarnbeltquerung 2014

–

Aktualisierung

der FTC-Studie von 2002, im Auftrag von Femern A/S, 2016, Tab. 6-2) this figure increases to 778.000

(336.000 + 424.000).

See: M.Aa. Knudsen, J. Rich: Ex post socio-economic assessment of the Oresund-Bridge, 2012

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Altogether the highest effect would be for car trips (+ 1.226.000, see table 8) but there are also

considerable effects for rail (309.000 trips). The additional generated traffic, calculated by this

model, would lead to 772.000 additional car trips (2.115 per day) apart from around 3.000 buses

per year.

trip purpose

rail

(1000 trips)

309

bus

(1000 trips)

car

(1000 trips)

191

328

135

173

175

172

1.226

car

(1000 veh.)

164

273

772

Business

Day Commuter

Weekend Commuter

Shopping

Other daytrips

Visiting friends/

relatives

Short holidays

Holidays

Total

Tab. 8

Assignment of the generated traffic not considered in the FTC study to modes

The number of additional car trips (772.000) in relation to the additional passenger trips for this

mode is based on occupancy rate approx. of 1,6. The occupancy rate is much lower than the

overall occupancy rate in the study area (around 2,5) due to the fact that the majority of the

generated traffic is related to the trip purposes day commuter and business with relatively low

occupancy rates compared to the overall traffic which is more dominated by private purposes

including holidays.

For

freight traffic

we expect about 35.000 additional trucks on the FBFL apart from 41.000 tons

of rail freight (see table 9). The effects for freight traffic are lower due to the fact that here the

transports have much longer average distances and the time savings due to the FBFL are less

relevant for the whole transport chain and partly are compensated due to the fact that parts of

the ferry cruises could be used for the mandatory drivers rest times.

See: Copenhagen Economics: Bredere økonomiske effekter af transport-investeringer, DEBATOPLÆG

udarbeijdet for Transportministeriet, Maj 2014

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

mode

road

(1000 trucks)

rail

(1000 tons)

9.464

9.505

FTC study for 2030

generated traffic

total traffic

generated traffic in % of

FTC traffic

634

669

0,4

Tab. 9:

Generated traffic for freight traffic

It is unlikely that these "dynamic effects" will occur immediately after opening. It will take time to

develop in full scale. Therefore, the reference to the 2030 results as shown here is only fictional.

However, this is the

relationship between the "dynamic effects" not considered in the FTC

study and the traffic forecasts considered there.

SUMMARY AND CONCLUSION

The results for passenger traffic and freight traffic are summarized in table 10.

additional trips/vehicles

to the FTC study

related to 2030 (in 1000)

rail passengers

bus passengers

car passengers

total passengers

rail tons

cars

buses

lorries

total vehicles

309

1.226

1.620

772

810

additional trips in % of

the FTC traffic

28,5

6,3

12,8

13,5

0,4

20,5

8,2

5,5

18,3

Tab. 10:

Overview of the results and synthesis (related to the year 2030)

TRU, Alm.del - 2019-20 - Bilag 67: Analyse af trafikprognoserne på Femern Bælt forbindelsen, fra transportministeren

Altogether the number of vehicles would increase with 18,3 % if considering the "dynamic

effects" (induced traffic) in full scale compared to the FTC study, thereof with 20,5 % for

passenger cars.

We would consider the outcome of the chosen transport science approach to calculate potential

dynamic effects of FBFL as relevant and realistic.

The model described above is considering intensified interaction due to better accessibility.

However, there is a certain inertia with settlement and social structures. The dynamic effects

probably take some time to set in structures.

The FTC model did not consider these dynamic effects due to the fact that as stated in chapter 2

gravitation was not covered fully because not the whole traffic of the study area had been

considered, but only the traffic between Scandinavia and the continent. Thus, the induced traffic

of the FTC study did not cover the gravitational effects in full scale. The results would be the

long term view of the "generated traffic" with a considerable "ramp-up-effect" of maybe 5 to 10

years or more.

But in any case the results are considerable: Given the calculated 810 thousand (see Table 10)

additional vehicles related to the year 2030 (around 2.200 per day) and a growth rate of 2 % p.a.

we would expect a number of around 2.500 vehicles per day ten years after opening of the

FBFL.

And the results should also be a motivation for the regions along the axis Hamburg

–

Oresund

region to push regional development of economy, tourism and social interaction. FBFL opens big

chances to develop this axis to a centre of growth resp. an axis of growth between Central and

Northern Europe.