Monday, August 23, 2010

The Real Cost of Chinese Railway Construction in Nigeria

Train in Lagos, Nigeria (photo: Accelerator, Nigeria)
For several years, stories about China Civil Engineering and Construction Corp (CCECC)'s $8.3 billion Lagos-Kano railway modernization contract have circulated in Nigeria and in the international press. The editor of Foreign Policy, Moises Naim, for example, mistakenly claimed in the New York Times that China was giving $9 billion in aid to finance this project. (There was actually no aid offered, although a preferential export credit of $500 million was discussed in connection with the railway).

Nigeria's existing, colonial-era railway between Lagos and Kano is far from adequate for a major transportation artery, antiquated, non-standard (narrow) gauge, and single track. However, I've read that CCECC's price for the new railway was "hugely inflated," that the project was hastily delivered to CCECC without proper tendering, and that there was not an inadequate "front end design" and/or feasibility study before awarding the contract.

I've been doing research in Nigeria off and on since 1987. I know that Nigerians are rightly suspicious of any large infrastructure project proposed for their country. They have bitter experience of money being "eaten", cost overruns, and white elephants. Was the CCECC project, recently scaled back, fated to be another in this long line?

Several weeks ago, Professor Richard Joseph sent me a photocopy of a fascinating full page ad published by CCECC in the Guardian (Lagos) on June 18, 2008, in which CCECC defended their 2006 contract, in particular the claim in the June 15, 2008 Sunday Punch (Lagos) that the contract was "inflated by $5.8bn" and should have cost $2.5bn.

CCECC contends that:
  • An initial feasibility study for the project was produced by the Nigerian branch of the international engineering firm Julius Berger together with an Italian firm, TEAM Consultants, in 2001. CCECC says they also conducted a "detailed feasibility study" before they participated in a limited tender, and an environmental impact assessment after the contract was awarded.
  • They produced 4 possible designs for the railway: (1) Rehabilitation of existing narrow gauge, single track for $1.76 billion; (2) Rehabilitation plus 2nd new narrow gauge, double track for $6.7 billion; (3) New, standard gauge, single track line for $5.81 billion; (4) New standard gauge, double track except Abuja-Kaduna, for $8.3 billion.
  • In April 2006, along with two other (unnamed) "pre-qualified companies", CCECC was invited to submit tender documents, which they did in June 2006 (yes, this does seem hasty).
  • Their costs were very reasonable.
Let's look into the cost issue further.

An important Chatham House report cited an interview with an unnamed World Bank official in Abuja, May 2008, who said that $3.04m per km was "double the cost it should have been." The report noted that a review by the Yar A'dua administration "discovered the costs to be highly inflated" but they didn't question the reports of this review. If the costs were "double" then this implies that the proper cost should have been around $1.52m per km. 

CCECC claimed that their bid for the double-track railway, at US$3.04 million per km, was low compared with the "international average construction cost of US$3.50m per km". (The railway distance is 1315 km. For a double track, this makes 2703 km x $3.04 or $8.22bn.)

As I read all this, I wondered:  what do we know about railway construction costs?  Is there any way to shed light on these various claims? I turned to Google.

I learned that costs per km are affected by a number of things, including materials, the terrain, and so on. But there is actually plenty of information on this online if one wanted to check general ballpark figures for reasonable costs of railway construction.

For example, news reports on a number of projects at the Railway Gazette website give costs, including a project financed in 2009 by the Asian Development Bank in Afghanistan: a 75km project for $170m, or $2.26m per km (single-track).

In 2008, Libya began a railway project, a standard gauge, double-track line awarded to Russian Railways, for a cost of 2.2 bn Euros ($2.8 billion) for 554 km, or $5m per km. Libya later signed a contract in 2009 with China Railway Construction Corporation for $805m for 172 km of railway on the Tripoli-Ras Edjer line ($4.7m per km).

A January 6, 2009 article in the Guardian (Lagos) by Moses Ebosele, "Lack of Policy Trails Nigerian Railway in 2008," quoted from a World Bank study (for Nigeria?) which "estimated capital cost of conversion from the present narrow gauge railway to standard gauge at between $1.5m and $5.0m per route km." Averaging these would give us $3.25m per km. (This doesn't say whether the standard gauge would be single or double-track.)

Here in Washington, DC, I found that our fancy new Dulles Airport Aerotrain cost $1.5 billion for only 3.8 miles (6.12km), or US$245.0m per km! Of course this is an extremely nice train. We're very happy with it.

Clearly without knowing the specifics of the Lagos-Kano line, it's hard to say which estimate is the most reasonable, or even whether the entire project was a good idea or not. But this little survey suggests that the contract costs of $3.04m per km in the CCECC contract do not look out of line. 

Indeed, in November 2009, CCECC and a new Nigerian administration signed a new contract. This contract is for a shorter, 200km stretch of railway between Abuja and Kaduna. This new project is estimated to cost $876 million, or $4.38m per km. The $500 million preferential export credit from China Eximbank will be applied to this project.

Whatever the details of the new contract, it's clear that three years and much negotiation later, the cost per km rose rather than fell. Which is what one would expect, given inflation.
But perhaps Nigerians with access to the inside story will have a different view of all this?


  1. Here is a link of China's railway development, different railway cost could vary significantly:

  2. It might be worth noting that the ADB-funded Afghan project is through empty, government-owned land in more-or-less the middle of nowhere - so there are few if any obstacles in the way, no major earthworks needed, no rivers to cross, no people to resettle, no land acquisition issues, no great crested newts needing a new pond, almost no-one to complain about noise or paperwork. This is perhaps a "baseline" cost for railway building.

    The Libyan project is similar, though the climate may be more harsh and they are starting from scratch.

    Both are freight lines, so more basic standards can be adopted than would be needed for passenger services.

  3. Interesting about the Afghanistan project. I read somewhere that it is also more expensive to construct a railway next to an existing line than to construct it from scratch.

  4. It is not necessarily more expensive to build a new, second line next to an existing track. It can be the case if there is a significant volume of traffic on the existing line, and earthworks and drainage works would cause disruptions and so must be carried out more inefficiently to ensure that the existing track is kept operational.
    On the other hand, an existing track ensures ease of access to the new line. Temporary sidings can be built to accommodate storage of materials and equipment, as well as facilities for the workers etc.
    Often work is scheduled so that anything which interferes with the existing track is carried out during periods where there are no trains or when scheduled closures of the main track can be implemented.
    The basic infrastructure needs for freight and passenger operations - civil works, track, signalling and control equipment, locomotives - are identical. Obviously rolling stock needs are quite different!! There is usually a need for more staff with passenger operations but in the overall context of African trains this is generally not a major cost. consideration

  5. Thanks for your comment, Martin. I can't recall where I read this, but I'm pretty sure they were referring to conditions in Europe, not Africa. So quite possibly it would be cheaper in Nigeria to build next to an existing line.

  6. Am still researching this issue. My interests are
    1. Is this CCECC contract from Lagos to Wherever for the construction of a new rail line or for renovation?
    2. Does the contract provide for the use of disused and rusty iron materials?
    3. I have interesting pictures of this CCECC work attached and would like readers to comment on them on

    The following shows the state of the rails before CCECC statted work

    The finished work which they are handing over this month looks like this:

    So I am interested in finding out what these lines are meant to achieve because from a layman's perspective, this is a repair or renovation that will not last long. What with the kind of materials employed. Larry Happiday

  7. I have tried to update my library with more pictures of the quality of materials being used by CCECC. A glance would show that these materials were either procured from an accident scene but more curiously from used sources. Am wondering: is this what we are paying for?
    Although a conference on Nigerian infrastructure opened two days ago, no one is likely to ask the Transport Minister the hard questions about the quality of CCECC's job because there may have been corrupt compromises But how long would these kind of jobs last?

  8. Yes DB, but dont u think it be wise and cost effective if another rail is constructed away from the old one?(such that the old one still useful for one or two things) at least the later can help to open another town as we know that access to railway is a great factor for social itegration and the likes.

  9. We want democracy and capitlism? thats what we get.I am sure a team of Nigerian Engineers already on the gov't payroll can deliver a better quality rail at 3x lower the cost.ask the chinese gov't how much they pay for their railways.

  10. Just returned from Lagos, Nigeria, where there is one Mohammed Lawal offering "a site with 700,000 MTW of used rails that were decommissioned from the Nigerian rail line..." At 12.5 mt length per rail section (weighing 0.809 kg each), he would have about 10,815 km of used rail!!! I guess he could go a couple of times around the whole country!

  11. To put Nigeria railway in order, we can engaged the indigenous engineers which are far better than foreigners .

  12. I can pride myself as an expert in the railway industry having acquired over 30 years post qualification experience in this field. I can tell you for certain that no two railways are the same even if they are built next to each other. Comparing prices with other railway projects is just an exercise in futility. It does not help anybody in this quest to ascertain if the prices were overblown. I will like to submit that CCECC is not helping issues by stating that its cost is lower than the average as this average is taken across Europe and western world and even so, it does not say anything. The question to ask is this, "what is contained in one kilometer of track?". With this description in place, one can then proceed to cost the constituents and relate that to the local environment. This will get you a price but even so this price cannot by used for the entire length as this price will continually vary as the parameters change.
    To put this thing at rest, the detailed BOQ submitted by CCECC for this works should be made available. From this document, we will be able to see all that went in to arrive at the total sum. Without that, it will all be hot air from various contributors. For example, the jubilee line extension in the UK cost over 300million UKP per kilometer as compared to the kano-abuja line at 3.04mUSD per kilometer. From either figure, one cannot tell anything in terms of value for money. Looking closer in, you will find out that Jubilee line is underground, has wide platform, has secondary doors, etc. hence the huge cost is justified. I am going a long way to say one thing, one needs to see the contract documents to see exactly what is required and then the BOQ to see what was costed.
    With these documents available, then one can argue constructively on what was done and how it was achieved. A friend told me about an experience he had reviewing a contract BOQ some time ago. He found out that the contractor was charging 5000USD for a length of rail. This item sells at about 1000USD on the international market hence he invited them to explain their figure. It turned out that it was a typographical error that was missed. This genuine mistake almost double the contract cost.
    Let us see the documents as listed above and we will be wiser in our deliberations. I will be more than happy to spend my spare time perusing the items and prices contained therein. I will like to let you all know that I know the prices of most railways items without consulting a book or the internet and can also point you to four sources to procure the items. This check will not take too long and it will put this matter to rest promptly.
    If pushed to comment if the CCECC figure is on the high side or not. I would say "SMILES" at least we are beginning to ask the right questions.
    Before I leave you as my wife is fully dressed and ready for the party, kindly ask that who did the tender evaluation on the Nigerian side and if they can be deemed competent to carry out that task. Again, how many organisations were shortlisted.
    You are about to start opening some, "can of worms".

  13. What surprises me is the fact that no one has remembered to quote or mention the largest single ongoing railway construction in the world and use it as a basis for comparison. The Saudi Arabian North-South Railway line is 1418km in length running from Riyadh to Haditha (near the border with Jordan). The project also includes a 576 km segment which detours west from the midway point connecting Az Zabirah (a Bauxite mining town) to the Persian Gulf port city of Ras Al-Khair. Other sections of the project include the Al Nafude junction which detours to Al Baseeta city and the fork which detours to the northern Phosphate mining town of Al Jalamid. All in all, we have a railway length of 2400km and a total contract value of USD$3.5 billion. This works out to a cost/km value of $1.458 million per km. I see no reason why the Nigerian example cannot approach this benchmark in cost seeing as labour is cheap and land acquisition cost is almost negligible (most of the route is rural and underdeveloped).

  14. I apologize. I made a mistake. The 576 km segment detours to the east heading for the persian gulf NOT west (as inferred earlier).