At around 10pm on Friday 13 May 1938, the 8pm overnight train from Đà Lạt to Saigon, hauled by SLM HG4/4 0-8-0T cog locomotive No 708, derailed at Kabeu during its ascent from Bellevue. The entire train rolled some way down the mountain, crushing a 4th class carriage and leaving 15 dead and around 20 injured. This is an English translation of an article criticising line safety and management published in L’Effort: journal hebdomadaire, Hanoï, on 17 and 24 June 1938.

“Nice view”

Ar-Lac, Da-Bac, Diêu Trì… so many names, so many catastrophes. And now, Bellevue. This is the fourth fatal rail disaster in less than 10 months. The eighth, if we also count derailments which caused only material damage.

Nice view! I can’t help but think about the cruel irony of these words as I prepare myself for the Dantesque spectacle I will see later. In the meantime, standing on the platform of the rear carriage of a winding train which chugs me towards the scene of the disaster, I admire the long serrated cog rail, located between the running track, which permits trains to climb the mountain.

It’s a beautiful route, admirably constructed. Think about it! This gradient is 12 millimetres per metre!

Of all the railway lines in Indochina, the Tourcham-Đà Lạt cog railway line is the toughest. But also the best built. It was designed and built by German and Swiss engineers, it’s said.

On this track, we have the impression that the risk percentage should not be higher than that of any other railway lines, if the rolling stock was a little less dilapidated. Just a little. But our railroad bigwigs, having been warned about it, rarely risk their own precious persons by travelling on rolling stock which has reached “the limit of its resistance” (Lefèvre dixit), and apparently don’t believe it necessary to safeguard others, namely the passengers. On a gradient of 12mm/m, they put into service eight antediluvian locomotives, all dating from 1932 or before.

We arrive at Krôngpha. From there, we’ll have to travel by road to reach the site of the disaster. The fireman of the train I travelled in, standing on the station platform, smokes a cigarette nervously. His face is pale. He doesn’t seem very communicative. And that’s a shame. Because I’m curious to know his impressions. Fortunately, the driver is more talkative.

“We must say, Mr Journalist, that our nerves have never been more strained. Stretched to breaking point. At the slightest failure of the locomotive, we believe it’s all over. Because this engine is no better than the other.”
The other being the crashed locomotive at Bellevue.

The fireman suddenly emerges from his silence:
“We’re not cowards, sir! We’ve all driven these machines more than once. But to have to carry on in a pile of junk which is in danger of falling apart by the minute, knowing about it and having to do it anyway, is intolerable. When there is “breakage,” we’re the first victims, and if, by some miracle, we come out of it with just a broken arm or a broken shoulder, it’s off to the hospital and then maybe to jail for us.
Our deceased brother at Bellevue was the best of us. Sixteen years of service. Top marks. The pride of our line. Well, sir, that tough guy was also afraid. Yesterday, before going up the mountain, he said to us as he left the station: ‘Brothers, I’m afraid!’ He knew the locomotive wouldn’t stay on the track.
He’d reported problems more than once. But no-one cared!
How can we not be afraid when we know we may be going to our deaths?”

There was still a small miracle. If the convoy, no longer biting the rack rail and cascading down the slope, hadn’t stopped there at the foot of that mountain, rolling instead a little further, the accident would certainly have been far more horrific. There would not have been a single survivor. Everyone, without exception, would have been crushed to death. The convoy could also have plunged down to the left, over the precipice, and that would have left no survivors either.

The wreckage of the train lies in an area of over 100 metres. The locomotive? It’s now just a heap of scrap metal. Two wheels were detached under the violence of the shock. Below, in the precipice, lies the fourth-class carriage, or rather what remains of it. There, for six hours, 19 people were trapped in agony, buried under a layer of scrap iron, woodwork, stones and earth. A large tomb from which 13 corpses and six dying were eventually removed.

Suddenly, a European man with a whip, presumably a supervisor or controller of CFI, charged towards me and shouted: “Get out of here right now!”
Seeing the camera hanging under my armpit, he tore it off violently, and without saying a word, opened it, took out the strip of film, and tore it up angrily.
“No photos, do you hear? And clear off as quickly as possible!”
Such exquisite politeness.

At the scene of the last rail disaster in Da-Bac, it was the same order: Don’t allow journalists to take any photographs. Understanding perfectly why these gentlemen don’t like photos, I retreat under the wrathful gaze of this irascible railway company agent.

On the platform of Bellevue station, I catch sight of a familiar figure: that of Mr Gassier, the big boss of public works, who had travelled down from Hà Nội to contemplate the results of the accident: 17 dead and 23 dying out of 44 passengers.

The causes of the accident are known to everyone. Just three days earlier, the locomotive hauling the ill-fated train had lost its grip on the rack rail and slipped, but at a point on the line where the slope was not steep and the driver was able to stop his machine with the hand brake. The same driver – the “pride of the line” – and the same locomotive. But if this obscure hero had escaped that small accident three days ago, he was still destined for catastrophe.

The expiatory victims

In the depot at Đà Lạt, four steam behemoths are lined up side by side. One of them is still hot and out of breath. Crouching next to it, a man in blue overalls stands in a large patch of grease and oil, wielding a large oil can. On the opposite side, another man hammers heavily.

When we arrive, the man with the oil can looks up. The mechanic who serves as my guide greets him, then we take a little walk around the depot.

“He’s just come in after driving the up express, 300 kilometres in total. Followed by two hours of shunting in the station. Tonight, he’ll drive the down express. Another 300 kilometres and then more shunting in the station.”
I’m stunned.
“And now he has to maintain his machine too? So when will he get some rest?”
My companion shrugs his shoulders,
“He can always rest for an hour or two if his ‘beast’ doesn’t give him too much trouble. Otherwise, he’ll rest tomorrow, after he’s finished taking the down express and done his shunting at the station.
For us, you know, rest, Sundays, holidays, official days off, are all just ideas on paper. We get just four days off per month, and that’s if no colleague gets sick unexpectedly. In the event that one of us gets sick, he must be replaced. So those four days can be reduced to just two.
But let me go and give my colleagues a hand.”

While my companion goes to assist his two comrades, I observe them. All of them have ravaged faces. And they’re all between 25 and 35, that is, all in their prime.
I understand now why they age so early.

They’re now finished. At the station buffet, they throw themselves heavily into the ugly armchairs, wiping large drops the sweat from their brows with the backs of their hands. I offer them a drink.
Both the driver and fireman refuse.
“Well, if you want, you can eat something with us, we haven’t eaten yet.”

Smiling, my companion said to me:
“You could be the unintentional perpetrator of a disaster by offering them something a drink. For us, no alcohol!
But the temptation’s very strong, you know! After 300 kilometres standing on the footplate of our ‘beasts,’ eyes wide open, ears constantly on the alert, we’re running on empty! And with the prospect of having to provide the same amount of exhausting effort that same evening or the next day at the latest, you’ll understand that there’s a strong temptation to pull oneself together with a stiff drink. But woe to anyone who succumbs!” Both the driver and the fireman nod in agreement.

“Already the authorities all too often pour all the fault for any problems onto our heads. We are the expiatory victims, sacrificed to public opinion by our leaders whenever a catastrophe occurs, whether we’re dead or still alive. Whether it’s the Đồng Hới accident or the Da Bac accident or the recent one in Đà Lạt, we’re always the people responsible! No-one ever talks about worn equipment, or poorly made track. Because if there were any question of that, the bigwigs would also have to accept their share of responsibility.”

I try to console them: “Public opinion is not always wrong. For the Diêu Trì accident, for example, the courts sentenced the locomotive driver to one year in prison. But he was then granted a suspended sentence. For public opinion, this was equivalent to condemnation of the senior officials, who they could not dare to drag into the box with the accused.”

But they continue. “Even where we must admit that we caused a catastrophe, who are the real culprits? Ask our masters if they would drive 600 kilometres there and back in less than two days, standing in front of a scorching firebox and constantly staring into the night, not to mention all that exhausting shunting work at the station. Could they do this work? Human resistance has limits.”

Then, after a short pause to calm his indignation:
“I’ll tell you a little story from last year. Once, when my turn came, my eyes hurt. I protested to the chief traction engineer that working in my state would expose passengers to certain disaster. Because everything depends on the power and the visibility of our eyes. But no matter how much I protested, he didn’t give a damn. So I had to harness my “beast” to the express. For the first 100 kilometres, by some superhuman effort, I managed to drive properly.
But then suddenly my eyes became clouded. I couldn’t see anything. I was blinded, and furthermore lost consciousness. The train was about a kilometre from a tunnel preceded by a steep ramp. Catastrophe was certain.
And at first the fireman and the trainee driver who was travelling with us didn’t notice a thing! But then, fortunately, he saw me collapse and caught me as I fell, while the trainee grabbed the brake. This allowed us to slow down when we got to the ramp. Needless to say, for the rest of the journey, the trainee was driving with the fireman’s help.”

His companion smiles:
“That fireman was me. It’s fortunate that on that day, or rather that night, we had a trainee by our side, and one who knew the machines better than I did!”
“So there aren’t always trainees on the locomotives?”
“The presence of driver, fireman and trainee is exceptional. The ongoing shortage of traction personnel is such that these apprentices often have to drive by themselves – sometimes even an express. During the terrible accident in Đồng Hới, it was a trainee who was driving!”
“And today?”
“Nothing’s changed. Today is like yesterday and tomorrow will be like today. There are always trainees driving. It would be understandable if we cost too much, but a fireman earns only 35 piastres and a driver 40 piastres a month. Get rid of just one of those numerous engineers who work in management, and we could recruit a dozen more firemen and drivers …”
I can’t help but shudder. Here, I believe, is one of the main causes of so many railway catastrophes!

On the downward journey the train stops suddenly and hisses for a long time. A station? But I didn’t hear the usual whistle that announces scheduled stops It must be a breakdown. How long will it last?
I clamber out of my carriage and walk forward along the track. We overlook a steep precipice of the kind we often see on the Huè-Tourane route and around the Cap Varella. The locomotive is slipping on the track and can’t climb. Discussion between the fireman and the guard. They ask for some volunteers to slide grit under the wheels. A group of good-natured passengers lend themselves to this little task, which amuses them at first, but which, in the long run, becomes tedious.
Eventually they can move the locomotive forward. Slowly. Soon afterwards, we encounter the same problem again! After a quarter of an hour, again with the help of the passengers, the train can creep up the slope. Now it’s going down again. For the descent, we can be confident. It knows how to freewheel. We just hope the brakes are working well!

The problems

The fireman’s incredible admission. The report to the Grand Council of Economic and Financial Interests in 1936 has since been irrefutably confirmed: The rolling stock is “at the limit of its resistance.” And these disasters are not yet at an end. The Đồng Hới accident which prompted this admission was just the start, and the recent Kabeu accident will not be the last. The abyss is the abyss, catastrophes are catastrophes.
The problem needs to be studied again.

We asked a specialist his opinion in order to close this investigation. This is what he told us:

“The two main issues to be studied are: track and traction.

Track

“The cost of building the Réseaux non concédés [CFI] totalled 2,665 million 1928 gold francs, or roughly 5,400 million Daladier francs. The cost per kilometre was roughly 2 million francs, double the average cost in France, 1.9 times that in Algeria, and 1.5 times that in Tunisia. Yet it was the most poorly constructed of all. Some lines already need completely rebuilding!
First, the cardinal crime. Indochina was underestimated by giving it a track gauge of only 1m instead of the international gauge of 1.435m. The 1m gauge track is fine for ‘lines of penetration’ such as those from Hải Phòng to Yunnan-fu, from Blida to Djelfa (Algeria), etc … with a continuous ramp over several dozen kilometres and minimum slopes of 18mm/m. On such lines, 1m gauge track significantly reduces the cost of earthworks, and is faster and more economical to install.
However, the Transindochinois was not a ‘line of penetration,’ and it was also not very difficult to construct. Only the ineptitude of some engineers made construction so difficult. The hardest sections to build were those of the Col des Nuages ​​and the Cap Varella. However, the altitude in those places is only 400m to 600m, with maximum gradients of just 15mm/m.
The 1m gauge track offers much less safety than that of l.435m. The multiple accidents which bloodied the last semester of 1938 came about partly because of this original design error.
The preliminary studies were also very poorly carried out. A track of only 1m gauge should have very few sharp curves, as few as possible. However, I was able to identify on the Transindochinois alone at least 50 sharp curves which needed to be corrected.”

“These already ill-conceived railway lines were then constructed very poorly. It’s no secret that, thanks to the complicity or ignorance of certain engineers of the Public Works Department, contractors involved in building the line were not properly supervised and gave poor results.

For the Vinh-Đông Hà section, for example, during the first year of operation of the line alone, around 20 bridges had to be repaired, and earthworks had to be re-consolidated over about 50km of track. Additional cost: 77 million francs (Poincaré francs).”

“The safety of the track is further impaired by the use of lightweight rails of 25kg instead of those of 30kg commonly used on all foreign railway lines, and the use of iron instead of ‘lim’ wood for the sleepers (ties). The multiple drawbacks of iron sleepers (excessive expansion and contraction, poor adhesion to ballast, etc) have been pointed out on several occasions. Yet, despite the abundance of ‘lim’ wood in the country, despite its relative cheapness, despite its advantages in adhesion, we continue to buy iron sleepers.

Traction

“Of the approximately 260 locomotives in service on the Réseaux non concédés [CFI], only about 30 are of recent construction. These are the type 300 “Ten-wheels” and 1000 “Pacifics,” which operate the daily fast-track Hà Nội-Saigon services. And they are only capable of hauling 170 tons on the steepest slopes of the Transindochinois such as the Col des Nuages. The rest of the fleet is junk which is between 20 and 40 years old.
But on the Indochina railways, with its fairly steep ramps, very steep curves, and often fanciful layout, the problem of traction is of paramount importance.
The axle load is, as we know, a crucial factor – the grip of a locomotive depends on its axle load, the pressure pushing its driving wheels onto the track. However, our track and other structures are so fragile that the axle load, that is to say the grip of the wheels on the rails, must be kept to a minimum.
Grip also depends on the number of driving wheels. The more wheels there are, the better the grip.
Hence the need to make the locomotive heavier, at the same time as distributing the weight between the greatest number of wheels. But this multiplication of driving wheels lengthens the locomotive, which in turn presents risks when it enters a sharp bend. Hence the need to articulate the locomotive.

The type of locomotive which would best meet the peculiarities of the Indochina railways is the ‘Garratt’ type, a locomotive with a 4-8-2+2-8-4 ‘Double Mountain’ type wheel arrangement, 23m long and articulated into three sections, which is already widely used on more than 70 networks in Africa and in particular in Algeria.
Such a machine is built especially for the colonial networks, that is to say for 1m gauge track. It can haul not only heavy convoys of goods, but also passenger trains at a speeds reaching 100km per hour.
It can make sharp curves of up to 120m radius and carry 12,000kg of coal and 28 cubic meters of water. It’s worth more than two standard locomotives put together: 23% savings in deadweight and an increase in traction power of 38%. It can cover great distances without refuelling, which saves time and makes real savings in operating costs.
This type of machine is fitted with the latest modern improvements: four cylinders, superheating, powerful braking including a compressed air emergency braking system, lighting in the cabin and front and rear headlights.
For shorter trips, the current pile of scrap would need to be replaced by light and fast diesel-electric locomotives, in which a diesel engine drives a group of electric generators.
An 8-cylinder four-stroke diesel develops a continuous power of 920hp at 700 rpm, the electric generator controls three traction motors which are connected to the machine’s three driving axles. In running order, one type, which measures 15.13m in length and weighs 100 tonnes, can haul a 90-tonne train at more than 100km per hour for a journey of about 1,400 kilometres without refuelling. While diesel-electric locomotives have the disadvantage of high weight compared to unit horsepower, on the other hand, they consume diesel oil, which is said to be cheaper than coal.
Of course, if we replace the locomotives, we also have to replace the wagons, or at least strengthen them.”

Too many chiefs and too few workers

Let’s take a look at the budget for the Réseaux non concédés [CFI]. We see in the chapter of expenditure just 373,680 piastres for maintenance, repairs and works, 3,129,900 piastres for rolling stock and 4,025,490 piastres for personnel. The staff of the Réseaux non concédés breaks down as follows: 9,556 locals and 180 Europeans. The 9,556 local employees receive only 2,883,090 piastres, while the 180 European staff earn up to 1,142,400 piastres, that is, an average of 630 piastres per month for each European and an average of 302 piastres per year for each local employee.

Now let’s look at the operating expenses by category since 1932.
While the expenses for personnel fell from 1,819,000 piastres in 1932 to 1,263,000 piastres in 1936 for the Northern Network and from 1,115,000 piastres to 773,000 piastres for the Southern Network, the expenses of the Operations Department increased from 59,000 piastres in 1932 to 162,000 piastres in 1935 and 180,000 piastres in 1936.

“There are too many chiefs!” said a railwayman of my acquaintance. Yes, there really are too many chiefs, and they cost too much. For each Arrondissement in the CFI Directorate there is a Principal Engineer assisted by an Engineer Office Chief and three Deputies, one for Traffic and Movement, another for Track and Buildings and a third for Traction. Each of these Deputies in turn have three Assistants, all engineers who each cost the budget over 400 piastres a month.
On the other hand, lower-grade staff are clearly insufficient. And very poorly paid.
In Traffic, a man working in a team gets 14 piastres per month, a handling manager 18 piastres, a brigadier 30 piastres and a team leader 35 piastres. In Movement, a station postman earns 21 piastres, a substitute station postman 29 piastres, a station master 4 35 piastres and a temporary worker 2 45 piastres. The privileged ones, station masters of major city stations, earn just 60 piastres. For the train service, a guard earns 15 piastres, a student conductor 21 piastres. a chef de train 30 piastres and a route controller 45 piastres. In Traction, a trainee earns 21 piastres. a fireman 35 piastres and a driver 40 piastres.

In return, what a lot of work! I will never be able to forget the words of the mechanic in Đà Lạt: “”For us, you know, rest, Sundays, holidays, official days off, are just ideas on paper.”
An eight-hour day for the telegraph operator who must focus from morning until night on his telegraph machine.
An eight-hour day for the station master who at nighttime must still interrupt his sleep at the sound of the telegraph machine to answer or forward telegrams, to attend the arrival and departure of every express, to update his registers, and to read the tedious circulars which the chiefs throw at him day after day.

What about the 40-hour week? Is the week really just 40 hours for conductors and road inspectors for whom there are no Sundays or public holidays, employees who only have two days off per fortnight? And those two days of rest are often reduced to just one. Because the staff shortage is such that one should expect at any time to be brought in to replace a failing comrade.

And then there’s the conduct of the chiefs, of certain chiefs, of course! Not so long ago, an engineer was witnessed hissing and pouring insults on his subordinates! In important stations, postmen, crew members, etc, are at the same time under the orders of the group leader, the controller, the station master and the deputy station master. Four leaders, four gods who are always angry, who, at the slightest mistake, can issue a written reprimand. However, four reprimands per month result in demotion by one grade. In this way, a station postman was demoted for having received four reprimands in a month: his “crimes” included wearing clogs while selling tickets, forgetting to greet the controller who was on the train platform and whom he had not seen, and smoking when the engineer (who hates tobacco) walked past him!

A station master was also sacked when, with a train waiting in the station, he returned to his room at the calls of his wife, who was being assaulted by a controller!

In this service, where one is under the orders of five fathers and three mothers (năm cha ba mẹ), such cases of injustice are common.

The Đà Lạt accident is simply the last straw which has overflowed the already full cup of public outrage.

Across the country there is a huge cry of anger. Let Governor-General Brévié know this and take the necessary reforms without hesitation!

Xuân Tiếu

POST SCRIPTUM: RECIEPTS FROM THE RAILWAYS

The receipts of the railway lines operated by the colony for the first quarter of 1938 reached 819,449 piastres against 521,075 piastres for the same period of 1937, that is to say an increase of 43,107. All the lines contribute to this increase, with the exception of the Saigon-Bến Đồng Sổ section of the Lộc Ninh line, which shows a loss of 766 piastres.
It should be noted that, of these 43,107 increases in revenue, 20,000 may be accounted for by increases in tariffs decided at the last annual session of the Grand Council of Economic and Financial Interests of Indochina.
The remainder, or 23,107, represents the share of the acceleration of business and economic recovery.
The railway management did not find it useful to specify whether this is gross or net revenue. We believe that these figures represent gross receipts, that is to say without deducting ordinary expenses (personnel, maintenance of equipment etc…) and extraordinary expenses (repair and replacement of locomotives and other rolling stock destroyed in the recent disasters).
For the Réseaux concédés [CIY] from Haiphong to Yunnan-Fu. receipts during the first half of 1938 amounted to 2,381,396 piastres, an increase of 79.50%.
The per kilometre revenue on the CIY line is almost 320% of that of the state network.
Another reason which militates in favour of a radical reorganisation of our railways.

Tim Doling is the author of The Railways and Tramways of Việt Nam (White Lotus Press, 2012) and the guidebooks Exploring Huế (Nhà Xuất Bản Thế Giới, Hà Nội, 2018), Exploring Saigon-Chợ Lớn (Nhà Xuất Bản Thế Giới, Hà Nội, 2019) and Exploring Quảng Nam (Nhà Xuất Bản Thế Giới, Hà Nội, 2020).

A full index of all Tim’s blog articles since November 2013 is now available here.

Translated from the article published in Chemins de fer régionaux et tramways (CFRT), September-October 2014

 Our story begins in Indochina during the French colonial era, when 1m gauge routes stretched for over 3000km, from the north to the south of present-day Việt Nam and as far as Cambodia. Eighty years after their delivery to Indochina, and 60 years after independence, there are still some SACM “Pacific” and “Mikado” locomotives in existence, miraculously spared the decades of destruction.

The administrative framework

In colonial times, the Federation of Indochina comprised Tonkin, Annam and Cochinchina, which today form Việt Nam, as well as Laos and Cambodia.

In Cambodia, the line envisaged at the outset as a route “from Phnom-Penh to Siam” – now Thailand – was granted in 1928 to the Compagnie des Chemins de Fer du Sud de l’Indochine (CFSI). It reached Mongkolborey in 1933, but the border crossing (PK 394) was not installed until 1942. In the meantime, in 1936, the failure of the concessionaire had led the Colony to entrust the line’s operation to the administration of Chemins de fer de l’Indochine (CFI). In 1952, the line was transferred to Chemins de Fer Royaux du Cambodge (CFRC), a creation of the “Associated State,” which then became the independent Royaume du Cambodge.

In Laos, the situation could not be simpler, since no public railway line project was ever to succeed in this country.

Within the perimeter of present-day Viet-Nam, on the other hand, the Doumer Plan of 1898 was to distinguish:

(i) A Hải Phòng-Hà Nội line of penetration extending north towards the Lào Cai border bridge to reach the Chinese city of Yunnan-fu (now Kunming) – this line was conceded to the Compagnie française des chemins de fer de l’Indochine et du Yunnan (CIY) and comprised 851km, of which 485km lay in Chinese territory;

(ii) A coastal axis called the Transindochinois, connecting Hanoi with Saigon and incorporating several branches – this network, not conceded to any private company, was built and operated by the colonial administration, then regulated by CFI, and comprised 2,192km.

While the Yunnan line was inaugurated in 1910, the Transindochinois would not be completed until 1936 (see CFRT No 368). After the fragmentation caused by the Japanese “coup” of March 1945, and then by the triggering of independentist actions, the southern sections (Saigon-Nha Trang and branches) and central sections (around Huế) of the line were reestablished and exploited by CFI, becoming the Régie des Chemins de fer du Viet-Nam in 1952, when, according to the terms of the Hả Long Bay Agreement, it was transferred to the “Associated State.”

In Tonkin, where only the CIY Hải Phòng-Hà Nội section and a very brief portion of the former CFI Transindochinois could be reopened, the technical and human resources of the two entities were united in 1947 in a Réseau provisoire ferroviaire commun CIY et RNC au Tonkin (CIY-RNC Provisional Common Rail Network, or RCT). The operation of the RCT was assured by both CFI and CIY, in both cases on behalf of the Colony and using its budgets. In this context, and of more particular interest here, the distinction between CFI and CIY locomotives was subsequently reduced to their registration; all the materials were then commanded and financed directly by the Ministry of Colonies, and it’s this situation which thus enables us to better understand the “expatriation” of the 231-A CIY Pacific locomotives, which were delivered to Saigon, rather than to Hà Nội.

After the 1954 independence agreements, political division intensified the physical divide between the Đường sắt Việt Nam (ĐSVN) networks north of the infamous 17th parallel and those of Việt Nam Hỏa Xa (VNHX) to the south. Regrouping, under the banner of ĐSVN, took place after Reunification in 1975.

[Note: There is no mention here of the more or less durable reopening, after 1945, of local lines in the Saigon region, on which only old and generally light locomotives were running]

Ten “Pacifics,” then 51 “Super-Pacifics”

After its first successful application on the 1m gauge lines of the Tunisian protectorate, the “Pacific” type locomotive made its appearance in 1933 on Indochina Railways (CFI) as the long Hà Nội-Saigon artery (1730km) drew near to completion. The first delivery was of 10 type 188 locomotives of the Société Alsacienne de Constructions Mécaniques (SACM), numbered CFI 1000 to 1009, and subsequently, 231-301 to 231-310.

Immediately after the completion of the Transindochinois in September 1936, the increase in traffic justified new orders, this time for a significantly strengthened type, designated 194 by SACM, or “Super Pacific” by CFI¹. These locomotives, numbered 231-501 to 231-515, delivered in 1939, distinguished themselves from the outset by hauling long trains of metallic cars at 90 km/h. But the onset of war came to compromise and complicate everything – much more permanently here than anywhere else.

The first four units of the second tranche of “Super Pacifics” rolled out of the Graffenstaden factories from January 1940, and these locomotives, numbered 231-518 to 231-519. could be shipped without a hitch. Shortly after, Indochina was isolated by a maritime blockade. and 231-520, 231-523, 231-524 and 231-525 were reassigned to Dakar-Niger. By cross-checking between incomplete and sometimes contradictory documentary evidence, it can be argued that they would not reach their original destination until 1950². As for 231-521 and 231-522, still under construction at the time of the German invasion of Alsace, they were administratively concealed as “lots of spare parts” in order to escape requisition. In fact, the two locomotives were not completed and shipped until 1947 … and all these setbacks were further aggravated by the absence, for nearly a decade, of six Pacific tenders blocked in metropolitan France and then assigned to the Chemins de fer de Provence (see box below).

“Indochine” Tenders in Provence
The passage of the colonies of French West Africa (AOF) to the camp of Free France in 1942 led the State Secretariat for the Navy and the Colonies to find a use in France for nine 141 type tenders which had been completed at the Corpet-Louvet factories in La Courneuve and destined for the Dakar-Niger Railway, but which had been separated from their locomotives. It was thus decided to assign them provisionally to the Chemins de fer de Provence, whose traffic could justify their use.
Six of them (40.017, 40.018, 40.020, 40.023 and 40.024) were delivered from May to July 1943 to the CP network of general interest of the Alps, in Draguignan. Three others (40.021, 40.022 and 40.030) were put into service from September 1943 on the local CP network of Littoral Varcis, in Saint-Raphael.
The construction of tenders having been subcontracted to another manufacturer, it was urgent to find replacements to couple to large Mikado locomotives. On the network of the Alps, makeshift tenders were fitted on flat wagon frames for the “Dakar” Nos 19 and 20, while No 23, sabotaged on its arrival, was never put into service and served as a spare parts reserve. There were therefore six machines left to be fitted with tenders: Nos 17, 18 and 24 of the network of the Alps, and Nos 21, 22 and 30 of the network of the Littoral.
At that time, six “Indochine” type tenders built by the Compagnie française de matériel de chemin de fer (CFMCF) at Tileul, near Maubeuge, were awaiting loading on the Marseille Arenc quay under the responsibility of the freight forwarder Daher. On examination of the available archives, it seems that they were intended for the Indochinese Pacifics, but could not reach their destination because of the hostilities. As the supplier had not been regulated, the credit insurance company which covered this transaction had become the owner and had relocated them to the Secretariat of State for the Navy and the Colonies, which had taken them into account as part of the AOF. The Ministry therefore undertook to resell them (at least those intended for the Alpine network, no archive relating to those from the Littoral having been found) to the two Provençal networks, with a buy-back clause for after the end of hostilities.
Between the Summer of 1943 and the beginning of 1944, the six tenders were sent to Draguignon and Saint-Raphael, where they were harnessed to the “Dakar” machines. Once the conflict was over, the locomotives were gradually withdrawn from service between 1945 and 1947. After retrocession to the State, still under the title of AOF railways, the six tenders were shipped to Corpet-Louvet by May 1948 to be renovated there before their final destination … which was therefore Indochina. According to Pascal Béjul’s research, they could have been harnessed to 231-521 and 231-522, which had remained blocked at the SACM factory during the war, to 231-520, 231-523, 231-524 and 231-525, sent temporarily to Dakar-Niger, or to 231-516, 231-517, 231-518 and 231-519, shipped to Indochina without their tenders.
José Banaudo

 

During the war years, the Indochina networks suffered very heavy losses and, in March 1946, an ambitious program of rehabilitation motivated new orders, not only for CFI, but also for the private company CIY. However, at that time, the Alsace factory was not in a position to respond to these requests, so these orders were lost to Haine-Saint-Pierre for 231-526 to 231-535 (for CFI), and to Fives-Lille for 231-A-001 to 231-A-006 (for CIY).

This was counting without the harassment of the Việt-Minh, who asserted themselves at the end of 1946 and opposed the resumption of the rail networks by the French: as a result of their actions, hundreds of kilometres of line could never be reopened, so that the needs were drastically reduced on what was recovered. Locomotives Nos 231-529 to 231-533 were “provisionally” allocated to Abidjan-Niger, while in France itself, in 1950, the embarkation of the 231-A locomotives was delayed for several months by the dockers of Dunkirk, who were opposed to the “colonial war.” To add to the confusion, the whole of this small series was assigned to the CFI Saigon-Nha Trang line, so that it was destined never to see “its” CIY Hải Phòng-Hà Nội line.

Even if the renunciation hour had not yet struck, the minimum operation maintained on an amputated and fragmented network could no doubt be served adequately with the existing fleet of locomotives … but despite this, a further, last tranche of 10 units (231-536 to 231-545) was still constructed by SACM in 1947-1948.

In Cambodia, a renumbering muddies the waters

Originally, the fleet of locomotives for the independent line of Cambodia had been entirely supplied by the German company Hanomag. Among these machines, two ex-CFSI “Pacific” locomotives were very quickly transferred to Saigon³. This transfer was compensated for by lighter CFI equipment (in particular: 230-300 Cail, then 131-100 Franco-Belge). Other exchanges followed, quite numerous despite the technical heaviness of transport by barge along the Mekong.

On 1 January 1951, five “Super Pacific” locomotives were in the Phnom Penh fleet. In the following year, during the transfer to the CFRC, these were included in a total of 10 locomotives of this type, which were soon renumbered as 231-500 to 231-510, a renumbering undertaken locally for the whole fleet – which means, confusingly, that that the current Cambodian 231-501 is not the 231-501 CFI locomotive of 1939!

The documents which have come down to us unfortunately do not allow us to determine the original “identities” of these renumbered Cambodian machines, with the two exceptions of 231-505 and 231-509 CFRC, which, in 1988, bore construction plates corresponding respectively to those of the former 231-538 and 231-542 CFI. We can therefore only suppose that the 1952 renumbering aimed to give an appearance of homogeneity to a stock which perhaps mixed ex CFI 231-500 locos together with ex-CIY 231-A locos, and we can also argue that this core of 10 ” Super Pacifics” was then to remain unchanging, because, after decolonisation, the relations between the two states of Việt Nam and the Kingdom of Cambodia would be sufficiently distant (a euphemism!) for any further exchanges to appear highly improbable.

The “Mikados” and their descendants

In mid-1946, SACM put into construction the first “Mikado,” its type 203, which differed from its predecessors only by the configuration of its axles, better adapted to the heavy and slow trains of a network ravaged by war. But the persistence of unrest in Indochina would profoundly affect the history of these locomotives, which was even more confused, if not painful to write about, than that of the “Super Pacifics.” For the orders initially planned, we note a distribution of attributions between CFI (141-501 to 141-528) and CIY (141-A-101 to 141-A-115)… of which the last deliveries would, however, only be partially implemented, when the prospect of a “liquidation” of the networks was no longer in doubt.

Locomotives 141-501 to 141-516 CFI began leaving the factory in 1948 … but, as with the contemporary “Super Pacifics,” their deliveries were disrupted due to the “general situation” in Indochina. Five locomotives of this batch (141-508 to 141-512) had to be “loaned” to Congo-Océan, after provisional conversion to 1.067m gauge. This loan, attested to by several documents, but about which we lack precise details, was extremely brief and ended at the start of the 1950s. In addition, eight other locomotives, inventoried by SACM as type 203-bis, were built specifically for the French Equatorial Africa network⁴.

In any case, in its inventory of 1 January 1951, CFI identified twelve 141-500 locomotives, distributed as follows:

– six in service of their southern network,

– three in service of their central network,

– three “in the course of assembly,” meaning that shipping had been carried out in disassembled elements (chassis, boiler, superstructures).

Units 141-517 to 141-528, ordered in 1947, were delivered in 1950 and 1951. Curiously, neither the manufacturer’s documents, nor those of the Office Central des Chemins de fer d’Outre-mer (OFFERFROM)⁵, seem to “know” the last unit, whose existence is only attested by a photo of it taken after an attack. Moreover, in the Autumn of 1949, CFI had sought to cede, to third-party networks, 16 of their “Mikado” locomotives which were still under construction, and which, in an increasingly difficult context, they no longer required. But no-one was interested in taking these locomotives, and while awaiting a decision, OFFERFROM was forced to take delivery, having them stored for many months in a warehouse located in Albert (Somme), together with a great deal of other rolling stock which had also been ordered for Indochina.

In addition to these measures, taken “for budgetary reasons” though no doubt concealing the loss of all hope of new redeployments, the government added the cancellation of a new order for 15 “Mikados” which had only been placed a few months earlier.

The same causes would produce the same effects in relation to the 141-A locomotives destined for CIY, whose order was reduced from 16 to 10 units in May 1949, then reduced again to only eight units in the following year. These locomotives, 141-A-101 to 141-A-108, were put into service in 1951 and 1952 on the Hải Phòng-Hà Nội line, the last operational line of the CIY concession.

After the 1954 independence agreements, North Việt Nam seems to have inherited about ten “Mikados” including at least seven of the eight 141-A ex-CIY locomotives, and South Việt Nam more than 20, all of them of CFI origin. Moreover, it is now established that 141-515 found refuge in Phnom Penh, where it was renumbered as 141-551 CFRC⁶.

In South Việt Nam, the American takeover, materialised by the contribution of about 50 diesels from the General Electric construction line, marked the death warrant for steam locomotives – most of them were left abandoned by the end of the 1960s. It was quite different in the North, where the Communist regime remained faithful to the machines inherited from the former colonising power. The 141-A ex-CIY locomotives lost their “A,” and the 141-500 ex-CFI locomotives taken over by ĐSVN were renumbered in sequence after them, from 141-109 onwards; after Reunification, some 141-500 locomotives from the centre and/or the south were transferred to the north, where ĐSVN rehabilitated and renumbered them, to at least 141-115⁷.

And they did not stop there: in 1964, the workshops of Gia-Lâm built locomotives Nos 141-121 and 122, replicas of the SACM “Mikados,” which resulted from a subtle mixture of “cannibalisation” of the cancelled machines, use of spare parts, identical machining and, of course, ingenuity. When the liaison between Hà Nội and Saigon was re-established, it was undoubtedly on purpose that No 141-121, which had become an object of national pride, was placed at the head of the inaugural Unification Express, at least for its arrival in Hà Nội on 4 January 1977.

In addition, these two locomotives were followed by Nos 141-151 to 141-216, built by Chinese industry between 1965 and 1974, making in total no less than 67 “copies.” So the plans of SACM had not been “lost to all the world!”⁸

As late as 1992, it was still possible to see SACM “Mikados” hauling expresses from Saigon between Hà Nội and Nam Định (87km). They were then among the very last French steam locomotives in regular service anywhere in the world. After the complete dieselisation of the network in 2002, it seemed that all the steam locomotives had gone, but their Chinese-built “sister,” 141-158, was installed as a monument in front of the new station in Hồ Chí Minh City, ex-Saigon. Faced with this escapee, how can we not recall a very old rule: over time, the success of a plagiarism always ends up constituting the most beautiful tribute that we can pay to the authors of the original!

Forgotten … and nevertheless legendary

In essence, and without taking into account the “copies” of Vietnamese or Chinese execution:

– 61 “Pacifics” and “Super Pacifics” were built for CFI and CIY.

– 36 “Mikados” were built for the two networks, and this type of locomotive was reproduced in a total of 44 examples, including the Congo-Océan lot.

With a total of 105 units, these locomotives have formed a truly exceptional “large family” in the history of French 1m gauge steam traction, even unique if one considers only the decades 1930 to 1950. In addition, through many of their technical features, the Indochina 141 and 231 locomotives presaged the famous 142 and 242 GELSA locomotives of 1951-1953, also of 1m gauge, which were to constitute the definitive design of the French “school” of steam locomotive.

And yet … due to their remoteness, but also perhaps due to the discredit thrown on the 1m gauge track by some and on colonisation by others, the “Pacific” and” Mikado” locomotives of what was French Indochina have never known the notoriety of their counterparts which worked the large metropolitan networks, except of course with those who experienced the epic adventures of the armoured trains and “Rafales,” already more than 60 years ago.

The independentist attacks of the years 1946 to 1954, although innumerable, did not bring about the complete disappearance of all locomotives; most often, the damage could be repaired in the workshops, where every unit of the fleet would experience several successive stays. After 1964, the bombings of the “American War” were more destructive, but several “Pacific” and “Mikado” locomotives survived them all the same, to accomplish their tasks until very recent dates.

In Tonkin, at the end of the 1980s, nine “Pacifics” and “Super Pacifics” (231-302, 231-309, 231-524, 231-525, 231-527, 231-528 to 231-530 and 231-533) could be found parked out of service, but 231-534 was still in operation in Hà Nội, while seven “Mikados” out of a total of nine (141-102, 141-103, 141-105 to 141-110 and 141-115 according to the numbering of ĐSVN) could be found hauling trains to Nam Định, Lạng Sơn and Hải Phòng.

At the same time, in the former South Việt Nam, entirely converted to diesel, Nos 141-507 and 141-510 had just been scrapped, and only the wrecks of the 231-A-003 and 231-A-006 of CIY origin could still be seen dumped in Nha Trang.

In Cambodia, after minimal dieselisation, almost all of the fleet survived the destructive madness of the Khmer Rouge to reach us today. In Phnom Penh, we know that the 231-501 – according to the local numbering here – has just been restored to life; 231-502 to 231-505, as well as 231-508 and 231-509, are stored and seem almost complete, along with 141-551, ex-141-515, last of the SACM “Mikados;” 231-506 and 231-507 are badly degraded; and only 231-510 seems to have disappeared.

Is it necessary to add: Can the hopes of preservation and transfer to France of one of these locomotives, which in 1992 were concentrated on 231-534 in Hà Nội, be considered today as definitively dashed?

Another regret: since the publication of Frédéric Hulot’s book, a quarter of a century ago (see bibliography) some persistent enigmas have certainly been elucidated … but there are still many unanswered questions! Thus, it is feared that the details of the origins of most of Cambodia’s miraculous survivors, as well as the service records of expatriate locomotives in Africa, will remain forever unknown.

Three steam locomotives, or nothing!
Around 1959, when all the colonial networks were engaged in a forced dieselisation, Indochina Railways swore by their “Super Pacifics” and “Mikados.” This position, against the tide of opinion, was justified by the qualities of these locomotives, and it was also explained by Indochina’s abundance of water resources and, in the north, by the relative proximity of the Tonkin coal mines … but that was not the only reason.
Lúc Raynaud reports that an administrator of OFFERFROM, after having masterfully demonstrated to his audience how “a single diesel locomotive replaces three steam locomotives,” heard himself answered by a leader of the CFI: “Well as for me, I keep my three steamers. because, when the Việt-Minh sends one to the bottom of a ravine, I still have two left to pull my trains.” Unstoppable!
It was this mixture of common sense and determination, reinforced by the incredible ingenuity of Vietnamese and Cambodian railway workers, which explains why some of these beautiful locomotives were able to survive through those tragic decades ….

The “Pacifics” of 1933

The 1933 series, later 231-300, was presented by SACM as a reduced version of the “Pacific” that it had built in 1914 and then in 1928 for the Tunisian metropolitan areas. In fact, the lighter rail of the first-completed sections of the Transindochinois limited the axle load to 11 tonnes. The manufacturer succeeded in reducing the empty weight from 51.9 to 48.8 tonnes, while increasing the power of the locomotive, thanks partly to the enlarged grate and heating areas and the increase in stamp boiler pressure from 12 to 14 hpz, but also through the application of some of the principles of André Chapelon: thus did they conduct “an in-depth study of the exhaust steam and smokebox gas ratios” and adopt the Kylchap exhaust.

The main deviation from the designs of that illustrious engineer, a simple expansion, was justified by the narrowness of the space between the side beams, which hardly permitted the accommodation of an interior mechanism, and undoubtedly even more by the imperative of facilitating both driving and maintenance.

It was specified that the length of the grate should not exceed 2.50m, “in order to facilitate correct heating by colonial staff.” The manufacturer also reproduced the provisions which characterised the Tunisian type, with a “special type of frame which accommodated a wide and deep firebox between the side beams” – the main side beams stopped at the front of the firebox, while the rear ones were extended to embrace the firebox, beyond the third coupled axle, so as to achieve a rigid assembly.

The diameter of the coupled wheels, fixed at 1.50m on the Tunisian “Pacifics,” was reduced to 1.40m on their Indochinese descendants, an arrangement which a metropolitan operator would not have classified in the prestigious category of “big wheels.”

The other characteristics were classic ones: thus, the mechanism included numerous parts which were interchangeable with previous types of locomotives – 230 Cail and Fives-Lille in particular – to take into account delays in workshop supplies, and several weeks of navigation from the metropolis.

The “Super Pacifics” of 1939

These locomotives, intended to circulate primarily on the last sections of the Transindochinois to be built, whose heavier equipment permitted a load increased to 13 tonnes per axle, constitute a reinforced version of their predecessors at 54,250 tonnes.

The maximum load was only reached on the three coupled axles, and, by adjusting the suspensions, it was possible to modify the distribution of the mass between coupled and load-bearing wheels and thus allow service on sections fitted with light rail, axle load then being reduced from 13 to 11 tonnes and the adherent weight from 39 to 33 tonnes.

The main advance was an increase from 1480mm to 1586mm in the diameter of the boiler, the length of which, however, remained unchanged. The “Super Pacific” thus had the same length of 4.60m between the tube sheets as their elders, so it could use the same tubes – in this case more numerous. With a stamp pressure of 14-16hpz, the power amounted to 1340hp. The tractive force of 9480kg allowed haulage of loads of up to 350 tonnes on inclines of 10‰ or 280 tonnes on inclines of 17‰.

Among other improvements, we should note the moulded steel cylinders lined with cast iron, the semi-automatic Henricot couplers⁹ and the smoke deflector screens, whose geometry gave these locomotives a typically French appearance.

The “Pacifics” and “Super Pacifics” both received the same type of bogie tender, which weighed 14.4 tonnes and allowed 16m³ of water and 6 tonnes of coal to be carried. The first series of 1933 was intended for an in-service speed limit of 80km/h, but the first locomotive No 1000 in the series reached 94km/h during tests. The “Super Pacifics” were designed to run at 100km/h, but they were limited to 90km/h in current service. However, this only applied to the brief period of peace experienced by the first units to be delivered, as their post-1945 cadets are unlikely to ever have reached such speeds on the battered wartime networks.

The “Mikados” of 1948

Appearing almost a decade after the “Super Pacifics,” the “Mikados” differed from them only by their wheel arrangement – and even the rear bogie was identical for the two types: the SNCF itself would never achieve such a degree of standardisation with its so-called “unified” locomotives!

With their four pairs of coupled wheels, the “Mikado,” however, presented an increased mass of nearly 3 tonnes. The reduction in the diameter of these coupled wheels from 1.40m to 1.20m brought the speed in service down to 60 km/h, but the tensile force was increased to 10,700kg. These locomotives were able to haul trains of 680 tonnes on a 10% incline or 400 tonnes on a 17% incline.

In addition, the brevity of the “Congolese episode” involving Nos 141-508 to 141-512 CFI suggests that the adaptation of this type of locomotive to the 1.067m track had been planned from their design, though no known text confirms this¹⁰.

The first “Mikado” received the same tenders as the “Pacifics” and “Super Pacifics,” but the following ones – the majority – carried “tank tenders” built by Baume-&-Marpent.

The North Vietnamese and Chinese “variations” of the SACM “Mikado” were distinguished from the basic type by their firebox made of steel rather than copper. In addition, they were instantly recognisable by their smoke deflector screens of larger dimensions, which had a very particular shape.

All these locomotives and tenders could subsequently be adapted to coal (North and South Việt Nam) or wood (South Việt Nam and Cambodia) firing.

Finally, in North Việt Nam:

– The Soulerin vacuum brake was replaced by the air brake, which explains the appearance, on all three types of machines, of fairly large air reservoirs between the screens and the smoke box;

– The old draw gear was replaced by automatic couplers.

Main sources:

Anonymous – Notice in Bulletin de la Societe Alsacienne, 1934

Frederic Hulot, Les Chemins de Fe de la France d’Outre-Mer: Indochine – Yunnan, Ed. La Regordane, 1990

Richard Pelham, Some Vietnam Puzzles, in Continental Railway Journey, 1992

David Gurnett, railwaysinvietnam.com website, 2013

Michael Pass, Cambodia Steam Today and Yesteryear, 2014

Acknowledgments: to Didier Delattre, Fabrice Lanoué, Yves Broncard, Luc Reynaud and Bernard Rozé

Footnotes

1. Obviously, these locomotives did not share anything other than this name with the famous 3-1200 of the Compagnie du Nord (231-C SNCF)

2. The total number of “Pacifics” delivered to Dakar (four), then to Abidjan (five), is corroborated by the allocation to the budget of French West Africa (AOF), at the beginning of 1949, of nine “Mikados” ordered for replace them in Indochina … but at least five of the “loaned” locomotives nevertheless made it to their destination, since 231-524 and 231-525 (ex-Dakar-Niger) and 231-529, 231-530 and 231-533 (ex-Abidjan-Niger) were present in Việt Nam in 1988 – restitutions having more than likely pre-dated independence.

3. These “Phnom-Penh Pacifics” (series CFSI Nos 1-7 then CFI 231-401 to 231-407), of German design and execution, have no other point in common with those which constitute the subject of this text other than their axle configuration.

4. The eight “Mikados” built specifically for the Congo-Océan were ceded to Mozambique in 1956; it is in this territory that the complete surveys have been carried out which today make it possible to confirm their “identity” (see table)

5. OFFERFROM: Office Central des Chemins de fer d’Outre-mer, Central Office of the French Overseas Railways; this organisation would become OFFERFOM after the independence of 1960 and would only intervene in the networks of French-speaking Africa and Madagascar

6. The original “identity” of this 141-551, which had quite intrigued us when we only knew of it through a photo of Guy Rannou, is today attested by its SACM number 8042.

7. The concordance between old and new numbers could not be established

8. At the end of the 1980s, discussing No 141-151 and its successors, an official letter from the ĐSVN Directorate, addressed to British tour operator Bill Alborough, did not hesitate to specify: “our Chinese friends faithfully copied the plans for the many locomotives they built for us.”

9. The Henricot coupling, which also fitted the passenger cars delivered to CFI at the same time, could be hidden by a removable buffer plate when the locomotive was attached to a train set comprising only conventional components. In fact, this equipment was not used widely on the non-conceded networks and it does not seem to have been planned for Nos 231-A-001 to 231-A-006 CFI.

10. In this hypothesis, the “universality” of the 203 SACM type would approach that of many more 141 USA-TCs known as MacArthur, which could be adapted to the track gauges of 1m (Cambodia, Tunisia, Franco-Ethiopian etc), 1.055 m (Algeria) and also 1.067m for, in particular, several British possessions.

11. We are unable to specify the construction stages to which each of the two types of tenders has been associated; in addition, exchanges also could be observed later in Việt Nam, where, for example, 231-534 was accompanied by a tank-tender during its last months of activity.

Pascal Béjui

Tim Doling is the author of The Railways and Tramways of Việt Nam (White Lotus Press, 2012) and the guidebooks Exploring Huế (Nhà Xuất Bản Thế Giới, Hà Nội, 2018), Exploring Saigon-Chợ Lớn (Nhà Xuất Bản Thế Giới, Hà Nội, 2019) and Exploring Quảng Nam (Nhà Xuất Bản Thế Giới, Hà Nội, 2020).

A full index of all Tim’s blog articles since November 2013 is now available here.