Content uploaded by Massimo Marotta
Author content
All content in this area was uploaded by Massimo Marotta on Aug 26, 2017
Content may be subject to copyright.
1
Analysis Report of the Taiwan High Speed Railway Project
1. Executive Summary
The Taiwan High Speed Railway (HSR) includes a 345 km
line connecting Taiwan's capital, Taipei, with the second
larger city, Kaohsiung, located in the south of the island.
The 35 years concession was awarded in 1997 to a group of 5
Taiwanese companies forming the Taiwan High Speed
Railway Corporation (THSRC). As the largest Built-
Operating-Transfer (BOT) in the world, the agreement also
included the commercial use of the stations & land
development. The first delay was announced in May 1999
(before the commencement of the Civil Works). THSRC
claimed the government failed to complete a number of tasks
(Cheng 1999). The planned opening date was revised to Oct.
2005 but THSRC missed it by 14 months till the 7
th
of Jan.
2007, three and half years beyond the contractual date. The 15.1B US$ project resulted in costs
overrun, for the Built Phase only, of 1.7B US$ (14%). THSRC suffered further losses during
operation due to high costs and lower then expected revenues. The analysis of the project shows also
a substantial failure of the BOT model itself. The basic concept that risks and profits shall belong to
the same party (private partner) was totally missed, with the government using public funds to
guarantee THSRC’s debt obligations, and with state-owned companies making loans to the
THSRC.(Chun, 2006)
The higher financial involvement resulted in the government taking over the THSRC management
and since 2009 the majority of the seats in the THSRC’s board of directors are occupied by
government official or state-owned company representatives. (THSRC, 2009)
2. Introduction
2.1 Strategic needs (project & procurement scheme)
Taiwan is a 400 km long island occupied in its central part by a series of longitudinal mountain
ranges. The land available for urbanization runs along the so called “West corridor” (ref to Fig.1), a
narrow strip of land near the coastline, which has to be served by longitudinal transport modes. HSR
was seen as a good alternative to the existing land and air transport modes, by providing a high
Figure 1: Taiwan High Speed Railway,
General Layout
2
transport capacity with a relatively fast travelling speed. The strategic needs for the government to
initiate the project were to enhance economic growth, to open up more areas for development and to
relieve transport congestion along the western corridor. The government adopted the BOT scheme
with the prospective of achieving a high efficiency with low public funds spending. (Kao, 1999)
2.2 Project Organization and Project Management
The Board of Directors was formed by
shareholder’s representative and was overall
responsible for THSR's performance. Its main
duties included regularly reviewing the
Company’s strategy, planned capital expenditure,
internal restructuring and human resources policy.
It also was authorized to appoint, dismiss and
supervise the Company’s senior management
team, chief auditor and chief accountant.
(THSRC, 2007). The board will be hereinafter referred also as the Project Manager or PM.
2.3 The award process
During the tendering stage, Eurotrain JV (European
Siemens and Alstom) was the preferred CORE
system supplier by THSRC, bringing European
standards for rails, signaling and communication
system into the project. The only other competitor
was the China High Speed Railway Alliance
(CHSRA). The competition turned out to be solely
based on commercial aspects. CHSRA asked the
government for further financing, on top of the original US$ 3.1B (NT$ 108B), for a total value of
US$ 4.6B (NT$156B). THSRC proposed a very favorable plan to the government, not only requiring
zero additional financing, but also offering to return back in full the financed amount from the project
operation revenue. (Ho, 2006). The Construction and Operation Agreement was signed in July 1998.
3. Analysis of the reasons leading to delay and cost overrun
The following chart (figure 3) summarizes the major events leading to delay and costs overrun. The
next paragraphs analyze in detail each contributor (leader) and resultant (lagging).
cost in billions NT$ (US$) – Exchange rate: 1 US$ = 34 NT$
THSRC own equity 65 (1.91)
Sourced by THSRC, by other institutional
investors and public investors 63 (1.85)
Total by equity (30%) 128 (3.76)
Total by debt financing (loans) (70%)
280
(8.24)
Grand Total by private sector (100%) 408 (12)
Table 1: Breakdown of Financial Proposal
Figure 2: HSR Organization Chart
3
Figure 3: Chart of events/leadings/laggings
3.1 Delay in finalizing the agreement with the banks (1998-1999)
The first major delay is related to the initial financial difficulties of THSRC in rising funds. It took
over 1 year from the date of award to finalize the agreement with the financing banks and for THSRC
to be able to gain capital to finance the project.
Two months after award (Sept. 1998), 25 banks formed the finance syndicate but then, since THSR
was the first BOT mega project in Taiwan, and in part due to the 1997 Asian Financial Crisis, the
banks considered too risky to finance the project without a government’s “full” debt guarantees. This
was not foreseen by the government and not reflected in the C-OA. After an initial stage of
negotiation, the government agreed to deposit US$ 6.18B (NT$ 210B) of postal saving into the banks
syndicate to provide credit for the money to be borrowed by THSRC, but asked THSRC to source the
balance 2B US$ by themselves (Lin, 2003). The banks however required additional clauses to the
Tripartite Contract among THSRC, MOTC and banks, with the most significant being the “Forced
buyout and Government takeover” stating that “if the project company defaults and the project is
Tender preparation
& Award 1997 Asian Financial Crisis
1999 Chi Chi Earthquake
Negotiation with Gov. and Banks
Tripartite agreement
Stringent rules by banks
CORE SYSTEM
Redesign part of CORE
system (Hybrid System)
Delay during Testing & commissioning
Redesign due to
Earthquake.
CIVIL WORKS
Ridership estimation by
Government only
No detailed design
Ridership Lower then estimated
Change of CORE
System Contractor
END OF BUILT PHASE
, START OPERATION
Financial
Difficulties
Delay and Costs
Overrun
Delay & Costs
Overrun
DELAY AND COSTS
OVERRUN
TESTING &
COMMISSIONING
Hybrid System (Japanese
trains/European tracks)
No enough funds
4
terminated by the Government, the Government will buy the project back and take the debt
responsibility the company leaves” (in Lin, 2003, p.51). The banks also forced in stringent rules,
limiting the equity to debt ratio. This resulted in a failure to obtain adequate loans as a direct
consequence of THSRC in failing to raise sufficient equities. THSR also failed to raise equity from
public shareholders, not confident in investing in such risky project. The Government supported
THSRC by investing US$ 0.235B in shares. (Lin, 2003). The tripartite (MOTC, THSRC, Banks)
agreement was signed on the 2
nd
February 2000 and THSRC obtained the first credit from the bank
syndicate to finance the construction. (THSR, 2008). This happened 19 months after award.
3.2 Failed attempt to solve the financial difficulties by changing the CORE system
As a direct consequence of the above, THSRC went immediately into financial troubles (as cash
flow). Knowing this, the Japanese Shinkansen offered THSRC, in case of award of the CORE system
contract, a US$ 2.06B (NT$ 70B) export loan, covering 85% of the CORE system contract sum.
Despite the fact of having already designed the system based on the preferred contractor Eurotrain,
THSRC decided to open a bid, later won by the Taiwan Shinkansen Corporation (TSC) for US$ 2.8
B (NT$ 95B). After the award, the Japanese Bank requested stringent measures due to the high risk
nature of the BOT project and THSRC decided not to apply for the export loan. (Note: Eurotrain filed
a law suit against THSRC, and THSRC agreed in 2004 to pay US$65M plus interests to Eurotrain).
This contract was signed in US$, allocating exchange risks to THSRC.
3.3 Redesign of the CORE system due to the use of the German/Japanese hybrid
The standards for rails, turnouts and tunnels were meanwhile already fixed, leading to a hybrid
correctly defined by Lehmann (2010) as “best of breeds”: Japanese trains running for the first time
on European tracks. The Japanese system is based on a single track, while the European one is based
on double tracks. Part of the signaling and automatic train control (ATC) systems had to be
redesigned. This delayed the testing and commissioning, at the completion of the construction.
3.4 Delay and Cost overrun during the Construction Phase
THSRC divided the project into six major contracts components with each one further containing a
number of different contracts, procured under different methods, as summarized in Table 2. Civil
Works (CW) included 39 km of bored tunnels, 8 km of cut-and-cover tunnels, 251 km of viaducts
and bridges and 32 km of cut-and-fill embankments, for a value of 5.5 B US$ (37% of total costs).
5
3.4.1 Changes in Specification
Changes in the Design specifications
to add seismic design criteria were
necessary after the 1999 Chi Chi
earthquake (Overstreet, Dulor, Subrizi,
2006, p.81) The Earthquake, with the
magnitude of 7.3, cause tremendous
disasters including the collapses of
seven bridges due to large fault
displacements (Chang at al, 2004, p.1)
.The contractors did not lose a great deal of time but they were still allowed to claim time extension
because the change was a variation to the works (Rothenburger, in Tee Tang, 2006)
3.4.2 Delay and Costs overrun during Civil Works
CW were constructed by 12 major Joint ventures and despite the large amount of civil infrastructure
and the technical challenges, CW did not had a great impact in the overall time delay and cost
overrun. This is due to an efficient contractual framework which used Design & Built with Lump
Sump price clearly defining risks allocation between the client (THSRC) and the contractors. Only
two conciliations occurred and in both cases the final accounts were agreed after conclusion of the
conciliation. (Tobin, 2005). The major case is relative to the 2002 two major collapses (with only
material damages) occurred during the construction of the HuKou tunnel. The cost impact is
estimated as US$ 30 Millions to re-reconstruction and programme acceleration), which however
Figure 4: The 4.4 km HuKou Tunnel
Table 2: Structure of Project major components
Main Component Procurement Mode Number of contracts
1 Advance work (site preparation,
utility relocation, etc.)
Design, Bid & Built, Lump Sum, Fixed
Price
N/A
2 Civil & structural works Design & Built, Lump Sum, Fixed Price 12, (5.5B US$)
3 Stations Design, Bid & Built, Lump Sum, Fixed
Price
5 (3 at later
stage)
4 Depots and maintenance bases Design, Bid & Built, Lump Sum, Fixed
Price
4
5 Track work Design & Built, Lump Sum, Fixed Price 5
(5.5B US$)
6 Mechanical and electrical work for
core systems
Engineering Procurement, Construction
Lump Sum, Fixed Price, turnkey basis
1: design and procurement;
1: integration, installation, and
commissioning.
(3B US$)
6
enabled the works to be completed with just 2 months delay (Da Via’, Marotta, Peach, 2005). The
CW final costs were finally 5% higher then tender prices, excluding the costs to amend the design
with seismic criteria. (Tobin, 2005). The construction works along the route were completed in May
2004. (Kao, 2006) Track works started in July 2003 and were completed in April 2005. The track
installation proceeded with no major delay in parallel with sectional completion of the Civil Works
along the route.
3.5 Delays and costs overrun during Testing and Commissioning (2004-2006)
In Sept. 2005 THSRC announced the revision
of Target Operation Date from the 31
st
Oct.
2005 to the 31
st
Oct. 2006 but also failed to met
this date due to delay in obtaining the
independent verification and validation
certification for the safety of the train system
from the Lloyds of London. The delays during
are all connected to the use of the
Japanese/European hybrid, and in particular:
3.5.1 Difficulties in training of train drivers
Pressed by the ongoing delay, THSRC decided to use the service of European HSR lines drivers,
faster then inexperienced local ones in learning how to operate the T700 Shinkansen train on
European tracks. The European drivers were required to train the local drivers but, as reported by
Gondo-Leiscallet (2007), in charge of Shinkansen 700T training “How can a German driver
communicate with a Taiwanese train master, a French switchmen or an Australian stationmaster? “
Even writing training material was reported as the most difficult objective (Gondo-Leiscallet, 2007).
3.5.2 Derailments during test drive
During Safety testing and certification, there were three derailments during the tests in early
November 2006. The operational service started on the 7
th
of Jan. 2007, three and half years beyond
the original contractual date.
3.6 Financial Difficulties till completion of Built Phase
Due to the event reported in the previous paragraphs, THSRC kept running in financial difficulties
due to the stringent rules on debt-equity ratio, additional costs due to the redesign of the signaling
system and change in design criteria for civil work, troubles during testing and commissioning, all
Figure 5: Testing of the 700T train in April 2006
7
resulting also in very high financial interest costs The government kept supporting THSRC,
guarantying the debt using in total 10.6B US$ from Postal funds, Public Service pension funds, Labor
Pension fund and Labor insurance fund (Lin, 2003). In March 2005 a controlled nonprofit
organization stepped in to buy securities in Equity investment (0.23B US$). (This was later
determined by the court to be illegal). In December 2005, THSRC announced a final total cost
overrun of $1.7 billion due to one year schedule delay and construction cost overrun.
3.7 Losses during the revenue operation
THSRC started its operation with European
train drivers only, with the first Taiwanese
driver getting its license in June 2007. (Gondo-
Leiscallet, 2007). Original estimates foresaw a
daily ridership of 180,000 after launch, growing
to 400,000 by 2036. The actual ridership as
reported in table 3 shows much lower figures.
4. Analysis of the Project Management responsibilities
4.1 Tendering focus on cost rather than value
The award to THSRC was based an optimistic and unrealistic financial proposal. THSRC (the PM of
the project) failed in underestimating the costs and in using an over optimistic proposal. However, this
would not have been possible without the Government own responsibility in accepting such offer.
4.2 Costs & Revenues estimated before tendering
There was a different of 7.7 B US$ between the cost estimations of the two competitors. We can
confirm today that THSRC underestimated the costs. This one of the recurrent problems in the
Construction Industry, where costs are often estimated before a detailed and final design is
completed. Revenues were over estimated by THSRC based on ridership forecast prepared by the
government without THSRC doing any independent check using an independent consultant to verify
the estimation.
4.3 Discontinued Project Management team (THSRC Board of Directors)
As a consequence of the financial support, the presence of government officials or state owned
company’s representatives in the board increased year by year, starting from the year 2000 with state-
owned Taiwan Sugar Corporation getting 1 representative in the board as result of the financial
Year Ridership seats utilization Ridership/ day
2007 15,555,656 44.72 % 42,618
2008 30,581,261 43.51 % 83,748
2009 32,349,260 46.31 % 88,628
2010 36,939,596 46.96 % 101,204
2011(June) 20,028,108 NA 109,743
Table
3
:
Actual
Ridership 2007
-
2011
(THSRC, 2011)
8
support. In September 2009, THSRC Nita Ing resigned from her post as chairman of THSRC and as
November 2009, the board was having 12 directors, 3 independent directors and 2 supervisors with
the public shareholders accounting for 9 seats plus 1 supervisor and the private shareholders for the
first time in minority with just 6 seats. The impact can be further analyzed in:
4.3.1 Different goals and objective within the PM team.
For a PPP, public and private partner aims to objectives of different nature, with the government
objectives related to public interest, and the private party focusing on gaining a commercial benefit.
It is possible with a PPP to develop common strategy and goals among the different parties. However,
as the government took over the THSRC management over the years, the equilibriums within the
board drastically changed, and the THSRC board was then formed in a growing percentage by
directors representing also the Government. The government was committed to complete this iconic
national project at any cost. The private partner was the major party financially affected by the
decisions taken, de facto, by the public partner.
4.3.2 Political influence and authority
THSRC slowly but constantly lost his “political influence” in proportion of loosing its financial
independency from the Government. This trend went beyond just loosing political influence and
literally resulted in a full loss of authority once the government gained the majority of the seats in the
board of directors (year 2009). All of the above would not have been possible without a very strong
influence of a third party: the banks. The bank syndicate was the one setting stringent financial
conditions and dictating the rules for both THSRC and the Government.
4.4 Managing external Environment
While external events as the Asian Finance Crisis might have contributed and might be out of the
control of the Project Management, it is clear that the management failed in reasonable foresee the
banks’ diffidence in financing the biggest PPP in the world. THSRC failed to predict the low response
by private investors which affected THSRC ability to raise equities. The public invested only at later
stage (at near completion of Civil Works, when risks are lower).
4.5 Strategies alliances
THSRC wrongly attempted to solve the finance problem by changing the CORE system contractor
and deciding to switch to the Japanese System. The change caused delay and costs overrun. THSRC
also didn’t get the financial benefit by deciding not to get the loan from Japan.
4.6 Estimation of life cycle costing (focus on capital costs)
9
THSRC did not properly evaluate and estimate the impact of using a hybrid system in tem of
Operation and Maintenance (O&M) life cycle costs.
4.7 Estimation of life cycle revenues
The estimation of revenues during operation was done by THSRC solely based on studied and
ridership forecast provided by the government. The actual figure shows that the government’s
overestimated the potentiality of the HSR line.
4.8 Original design without earthquake detection system
On the 21
st
September 1999 a powerful earthquake happened in Taiwan. The THSR didn’t originally
propose an Earthquake detection system for the railway line. The “Chi Chi” earthquake triggered
criticism but THSRC used this fact to strengthen his decision to switch to the Japanese system and
cover its initial shortfall in missing this system in its original design. Taiwan is a country with history
of frequent seismic events. Updating the system has contributed in delay and higher costs.
5. Analysis of the actions which could have been taken
5.1 Contract Procurement and Terms of the C-OA
THSRC should have not accepted an agreement not sufficiently addressing critical issues. The
Construction and Operation agreement should have better addressed the following:
1) Finance framework, better defining the financial structure. This might have avoided the
“tripartite” agreement between Governments, THSCR, and banks, which is one of the reasons of
the complexity in the management of this project.
2) Contractual baseline on ridership forecast: this might be limited the financial losses due to the
underestimation (use of a more fair risks allocation)
5.2 Benchmarking
THSRC was having, at the time of tender preparation, a number HSR projects which could have been
used as benchmarking, a useful management tool also when aiming to identify strategies. (Egan,
1998, p.10). The following project were available for benchmarking in the year 2007: (Gupta, 2009)
• Japanese Shinkansen, started in 1986, maintaining a clear cut between financing (by
government) and operation/maintenance (by the private partner, paying a usage fees).
• French HSR, with both operator and developer being two independent state owned entities.
• Korean HSR, built phase ongoing since in 1989, awarded under a BOT mode. As per 1997 the
10
built phase supposed to be close to the target completion, but was actually suffering from
delays and cost overrun (later resulted in an overall in 7 years delay and 206% costs Overrun).
• The Channel Tunnel Rail Link was under the built phase an already in delay. The original
1986 concession scheme in designing, building, financing and operating, was quite similar to
the one adopted in Taiwan. The project later resulted in both delay and cost overrun.
• Italian HSR started in 1991. Since 1997 using a scheme similar to the French one.
As per 1997, there were not cases of successfully implemented HSR project using a scheme similar
too the one adopted in Taiwan HSR. As reported by Chi and Yu, (2000) “Korean HSR project can
give us lots of lessons to learn. Especially political intervention and design change resulted in
schedule delay can and should be analyzed in great detail so similar mistakes will not occur in
Taiwan HSR”.
5.3 Strategic Alliances
THSRC shall have better evaluated the consequences in changing the CORE system contractor,
letting Eurotrain leave the initial alliance. Additionally THSRC shall have better investigated the
terms of the Japanese loan agreement prior to the award of the CORE system contract. The expected
advantageous loan from Japan was later not used due to stringent conditions set by the Japanese bank.
5.4 Standardization & Evaluation of life cycle costs
Changing the CORE system goes in the opposite direction of lean thinking and standardization. Both
capital costs and O&M (life cycle) costs were underestimated or were probably ignored due to the
urgency of THSRC in solving an immediate problem. Using a hybrid system for the first time in the
world could have led to even more catastrophic consequences. Standardization is the base of efficient
construction and project management, and THSRC shall not have compromised this to solve
(partially unsuccessfully) an immediate financial problem.
6. Conclusions
THSRC failed to achieve the targets related to Time objectives (delay), Cost objectives (cost
overrun), Costs in use & Maintenance (O&M costs) and most importantly Commercial Profitability
(High costs, low revenues). Quality Objectives were generally met for the built phase but with the
project characterized by low efficiency and effectiveness (with the use of the hybrid system). HSR
has changed Taiwan’s transport & economy, but this successful “value” probably satisfies only the
Government’s objective, deviating from the original partnership common goals.
11
References
Chang, K.C., Kuo, K.Y. Liu, K.Y. and Lu, C.H. (2004) On Seismic Retrofit Strategies of Highway
Bridges – Experiences Learned from Chi-Chi Earthquake In: 13
th
World Conference on Earthquake
Engineering, Paper No.639, Vancouver, B.C., Canada
Cheng, B. (1999) Government blamed; Company says high-speed rail could face two years delay.
Taiwan Today, 14 May 1999, htttp://www.taiwantoday.tw/fp.asp?xItem=17165&CtNode=122.
Chi, L.S. and Yu, M.Y. (2000) Evaluating High Speed Rail System in BOT scheme by economic
status in Taiwan ROC, In: Proceedings of the 17th ISARC 2000, Taipei, Taiwan, pp 1-5
Chun, L.M. (2006) COVER STORY: Putting BOT back on track, In: American Chamber of
Commerce in Taipei
China Post (2006) Bullet train service not to start at end of this month, In: The China Post, 23 Oct
2006.
Da Vià, A, Marotta, M and Peach, G (2005) High pressure jet grouting in tunnels – a case study, In:
Erdem, Y and Solak, T (Eds.), 31
st
ITA-AITES World Tunnel Congress, 7-12 May 2005, Istanbul.
A.A. Balkema Publishers, Vol. 1, 477-83.
Egan, J. (1998) Rethinking Construction In: The report of the Construction Task Force to the Deputy
Prime Minister, John Prescott, on the scope for improving the quality and efficiency of UK
construction. Department of Trade and Industry.
Gondo-Lescaillet, A. (2007) Staff Training at Taiwan High Speed Railway Corporation (THSRC) In:
Japan Railway & Transport Review, Issue 48, August 2007, pp 47-50
Gupta, M. (2009) International case studies on delivery and financing –a report for HS2, In: Ernst &
Young LLP report to HS2, 18 December 2009.
Ho, S.P. (2006) Government Policy on PPP Financial Issues: Bid Compensation and Financial
Regulation, In: Manuscript on PPP book chapter, Department of Civil Engineering, National Taiwan
University.
Lee, S.K. and Hsieh, W.Y. (2003) A choice-based conjoint analysis on the product line of Taiwan
High Speed Railway, In: (NA) Proceedings of the Eastern Asia Society for Transportation Studies,
Taipei, Taiwan, October 2003, Vol.4, 802-7
Lehmann, O. F. (2010) The Top 12 Causes that Drive Projects into Failure.
Lin, J.S. (2003) The Conflicts and Risks in BOT projects. MSc CE Thesis, Massachusetts Institute of
Technology.
MOTC (2011) Table 2-8 Passenger Traffic of High-Speed Rail In: Monthly Statistics of
Transportation & Communications. http://210.69.99.7/motchypage/monthly_eng/c2080.xls. MOTC
Department of Statistics. Retrieved 2009-02-24
Overstreet, E. Dulor, S.P. and Subrizi, S. (2006) A More Rational Seismic-Resistant Design of
12
Passenger Rail Bridges, In: PB Network, Issue 63, May 2006, Volume XXI, Number 2, pp 80-2
Taiwan High Speed Railway Corporation (2008), annual report 2007, http://www.thsrc.com.tw
Taiwan High Speed Railway Corporation (2010), annual report 2009, http://www.thsrc.com.tw
Taiwan High Speed Railway (2011) Monthly Ridership Report,
http://www.thsrc.com.tw/tc/about/ab_operate_annual.asp. Retrieved on 19 July 2011.
Tee Tang, S. M. (2006) How Front End Factors Affect Project Outcomes: A look at four Infrastructure
project in Asia, In: Working Paper #34, Collaboratory for Research on Global Projects, Stanford
University, September 2006).
Tobin, B. (2005) A design and Construct Success Story, In: Australian Construction Law newsletter,
No. 105 – November-December 2005, pp 50-2.
Wu, Y.C. and Chen, Y.S. (2009) Exclusive Interview: Nita Ing Bows Out, In: CommonWealth
Magazine September 24, 2009
Yu, K.H. and Johannesson, J. (2010) Near-bankruptcy of the Taiwan High Speed Rail Corporation:
What Went Wrong?, In: International Journal of Business and Management, No. 12; December 2010
Vol. 5, 14-22.
