30 October 2019  

USN	the Wells Brothers' Battleship Index  Debunking the Riachuelo (1914) By David R. Wells Version 1.0
British (RN)
Japanese (IJN)
German (HSF & KM)
Russian & Soviet
French
Italian

Disambiguation: This article is about the 1914 Brazilian battleship project, and not about the 1883-1910 battleship.

Introduction

The Brazilian battleship Riachuelo (1914) was certainly not built. There are correct preliminary drawings and statistics, but there has been a great deal of confusion about her design, owing partly to some incomplete early information.

My purpose here is not to accurately describe the history of the development of the Riachuelo. Ian Sturton and David Topliss have already done that, and probably better than I could. Rather, my intention is to show how so much mistaken information became so widespread, to find the original sources of this bad information, and to point people towards accurate information.

In the early 20th century, Brazil purchased two battleships from Great Britain, the Minas Gerias and the Sao Paulo. Brazil clearly wanted a third battleship, and ordered the Rio de Janeiro. The Brazilians belatedly recognized that her 12" guns would soon be outclassed by Chile's new Almirante Lattore, and sold her to the Turks as the Sultan Osman I before she was completed. This spectacular, seven-turreted ship was taken over by the Royal Navy at the outbreak of World War I, and served as HMS Agincourt.

The Brazilians still wanted a third ship to replace the Rio de Janeiro, and so they began working with the British firm Armstrong, the same firm that had built the Minas Gerias, on designs for a new ship, to be called the Riachuelo. At the start of World War I, she too was canceled, before the ship's keel was even laid. The project was largely forgotten.

The story begins to get a bit more confusing in 1969, when Dr. Alan Vanterpool published an article about the Riachuelo. ¹ Vanterpool mentioned four designs, designated A, B, C and D. He gives specifications that are tantalizing, but incomplete.

Riachuelo designs, according to Vanterpool

	Proposal "A"	Proposal "B"	Proposal "C"	Proposal "D"
Displacement - Normal tons	31,500	32,500	36,000	36,000
Length Overall, feet	685	689	740	740
Beam, feet	96	96	98	98
Moulded Depth, feet	51	52	52.5	52.5
Draft, feet	28	28.5	29	29
Armament
Main	12 - 14"/45	10 - 15/"45	10 - 16"/45	12 - 15"/45
Secondary	16 - 6"/50	20 - 6"/50	20 - 6"/50	20 - 6"/50
Anti-Torpedo	12 - 3"	10 - 3"	10 - 3"	10 - 3"
Anti-Aircraft	4 - 3"	4 - 3"	4 - 3"	4 - 3"
Torpedo Tubes	4 or 6	6	6	6
Armor
Belts	12" (c)	12"	?	?
Turrets	12"	13"	?	?
Decks	2"		?	?
Machinery
Number of Shafts	4	4	?	?
Type	Parsons Turbines	Parsons Turbines	?	?
Boilers	Yarrow Small Tube	Yarrow Small Tube	?	?
Fuel	4000 tons coal (a)	4000 tons coal	?	?
	1000 tons oil	1000 tons oil	?	?
Speed, knots	23	24 (b)	?	?

Vanterpool's Footnotes:

(a) alternatively, 3500 tons of oil only could be carried giving 24 knots speed

(b) raised to 25.5 knots with oil fuel only

(c) 8.5 feet deep

Vanterpool provided simple profile drawings of Proposals A & B.

It is important to note that Vanterpool did not cite any sources.

In 1970, Ian Sturton wrote a letter to the editor of "Warship International "in response to the Vanterpool article. In this letter, he said that he checked primary sources, (Armstrong's archive) and found solid evidence that Brazil had chosen Armstrong Design 781. ² Sturton even provided a remarkably accurate drawing.

Sturton clearly got it right, but it was too late. Vanterpool's data became more widely known with the publication of Siegfried Breyer's classic volume "Battleships and Battlecruisers 1905 - 1970", which became a staple in many libraries. This famously comprehensive and extensively illustrated book included an entry for the never-built Riachuelo, ³ citing the Vanterpool article, and even included a couple of preliminary drawings.

Breyer indicated that Proposals C & D had the same hull and propulsion.

I do not wish to criticize Breyer too harshly. He was doing the best he could with the information available in 1970. He probably had already completed his book when Sturton's letter was published. However, today we have the advantage of considerable research that was not available to Breyer. Breyer's work has become remarkably influential in the Internet age, and seems to be regarded by many gamers and amateurs as a definitive source. Other books, such as Tony Gibbons' "The Complete Encyclopedia of Battleships" (Salamander Books, 1983) also seem to use Breyer's data.

In 1988, David Topliss published "The Brazilian Battleships 1904-1914" ⁴ which shed much new light on the subject of Brazilian battleships. He includes a much detail, and a great many drawings which came from Armstrong's originals. He confirmed Sturton's conclusion, that Brazil had ordered Armstrong Design 781 for the Riachuelo project, but provided vastly more detail on the design process, and other designs that were considered.

So, if the "mystery" of the Riachelo's design was solved by Sturton and Topliss many years ago, then what remains? Their articles could not reveal the sources of information behind Vanterpool's original article. Since Vanterpool's article had no citations, we cannot easily track down the original sources. It would be easy to just dismiss the Vanterpool article as uninformed error, but after looking at many sources, I think I have found some relevant information that Vanterpool might have used.

My hypothesis is that Vanterpool saw some Armstrong designs, at least some of which were "stock" designs not associated with the Brazilian project, and misidentified them as part of the Riachuelo project.

It should also be noted that Armstrong's rival firm Vickers also submitted several designs for the Riachuelo project. I am largely ignoring these because Armstrong seemed to have the inside track all along, and also because Breyer and others indicate that all of the designs came from Armstrong.

Chasing Proposal A

Vanterpool's drawing of Proposal A. Breyer's drawing is similar.

	Proposal "A"	Design 653 (Source: Topliss)	Design 740 (source: Peter Brook)
Displacement - Normal tons	31,500	31,600	31,500
Length Overall, feet	685	690	685
Beam, feet	96	92	96
Moulded Depth, feet	51	?	?
Draft, feet	28	26	31.25
Armament
Main	12 - 14"/45	12 - 14"/45	12 - 14"/45
Secondary	16 - 6"/50	14 - 6"/50	16 - 6"/50
Anti-Torpedo	12 - 3"	14 - 4"	?
Anti-Aircraft	4 - 3"	?	?
Torpedo Tubes	4 or 6	3	4

Data for the two candidates for Proposal A.

Design 653: Brazil ordered this one, but.....

However, the drawing for Design 653 clearly does not match Vanterpool's illustration. The turret arrangement amidships is obviously different. There is also another problem: Design 653 was designed in 1910, as part of the design process that lead to the Rio de Janeiro. The Riachuelo designs date from 1913-14. It is clear then that Design 653 could not be Vanterpool's Riachuelo Proposal A.

Design 740: Ian Sturton's drawing from Peter Brook's article. Looks like a match!

Because the dimensions and the drawings match so well, I conclude that Vanterpool found a copy of Design 740, and mistook it for a part of the Riachuelo project.

Chasing Proposal B

Vanterpool's drawing of Proposal B. Again, Breyer's drawing is similar.

	Proposal "B"	Design 764A (Source: Brook)	Design 782 (Source: Topliss)	Design 786 (Source: Topliss)
Displacement - Normal tons	32,500	32,500	34,500	32,500
Length Overall, feet	689	669	710	700
Beam, feet	96	95	96	95
Moulded Depth, feet	52	?	51.1	51.1
Draft, feet	28.5	28.5	28	28
Armament
Main	10 - 15"/45	10 - 15"/45	10 - 15"/42	10 - 15"/42
Secondary	20 - 6"/50	20 - 6"/50	14 - 6"/50	14 - 6"/50
Anti-Torpedo	10 - 3"	10 - 3"	10 - 4"/50	10 - 4"/50
Anti-Aircraft	4 - 3"	4 - 3"	3 - 3"	3- 3"
Torpedo Tubes	6	4	2	2

Data for three candidates for Proposal B.

Armstrong actually had several battleship designs with the ten 15" guns that Vanterpool describes. These include Designs 764A, 782, and 786. Vanterpool provided a drawing of Proposal B, which only added to the confusion.

Design 764: Ian Sturton's drawing from Peter Brock's article. 764A was similar. Hmm, close, but.....

Given that Vanterpool seems to have assumed that "stock" Design 740 was Design A, it seems logical that he would have mistaken another Armstrong "stock" battleship for Proposal B. Thus, Armstrong Design 764A seems like a likely candidate. Design 764A has some statistics that are relatively close to Vanterpool's numbers for Proposal B. The displacement (32,500 tons) and the secondary armament (20 x 6" casemates) match up very nicely. However, there are some significant mismatches. Design 764a is about 20' (6 meters) shorter than Vanterpool suggests for Design B. Also, the drawings of Design 764 (I couldn't find one for 764a) show casemates in the hull aft. Vanterpool's drawing of Proposal B doesn't. Since 764a would have had more casemates than 764, it is possible that the arrangement was different, but it seems unlikely that they were radically different.

Design 782: Actually proposed to Brazil, February 1914.

Design 782 has the significant historical advantage of being designed for and offered to Brazil in February 1914 as part of the Riachuelo project. As in Vanterpool's drawing, there are no casemate guns in the hull aft. However, the sizes once again do not match. (710' long) Further, Design 782 has one large funnel. Vanterpool's Proposal B has two smaller funnels. Design 782's Q turret faces forward, whereas Vanterpool's Proosal B's Q turret faces aft. Design 786, 10 feet shorter, has similar issues. Apparently there never were separate drawings for designs 782, 784 and 786, so the designs are presumed to be quite similar.

Armstrong's "stock" Design 764a seems like the most likely candidate to be the basis for Vanterpool's "Proposal B", but there are enough problems to cause doubt.

Chasing Proposal C

With Proposal C, the problem becomes still more difficult. Vanterpool provides no drawing, but even worse, his tabular data is even less complete.

	Proposal "C"	Design 685a (Source: Topliss)	Improved Malaya (source: Brook)
Displacement - Normal tons	36,000	32,270	?
Length Overall, feet	740	710	?
Beam, feet	98	92	?
Moulded Depth, feet	52.5	51' 3"	?
Draft, feet	29	29' 11"	?
Armament
Main	10 - 16"/45	10 - 16"/45	10 - 16"/45
Secondary	20 - 6"/50	14 - 6"/50	?
Anti-Torpedo	10 - 3"	4 - "12 pdr"	?
Anti-Aircraft	4 - 3"	8 - "3 pdr"	?
Torpedo Tubes	6	3	?

Armstrong Design 685a is close enough to Vanterpool's description of Proposal C to make me think that we've got a match. It is an Armstrong design, created for Brazil, with ten 16" guns. Some drawings I've seen have stylized, handwritten ones that look enough like fours to make me think that the difference in length (740 vs 710) might be a misreading.

There are three significant problems though. First, the secondary battery on Design 685a is only fourteen 6" guns, instead of the twenty that Vanterpool describes. That's a significant discrepancy. Second, the displacements don't match. (36,000 vs 32,270) Finally, Design 685a was from February 1911, three years before Armstrong submitted their final Riachuelo designs.

Peter Brook briefly mentions in his article a design for an "Improved Malaya" with ten 16" guns in five twin turrets, and a speed of 25.5 knots if oil was used. Somewhat confusingly, he also describes Design 764 as an "Improved Malaya". He wrote that the 16" gun version was one of the last two designs in the Armstrong portfolio, but gives no further details. If this design could be found, it might be an additional candidate. Since we have already shown that Vanterpool took Design 740 for Proposal A, it is possible that this stock design might be the origin for Proposal C.

While Design 685a seems like the most likely candidate so far, there are enough problems to give me significant doubt. With more information, the "Improved Malaya" might prove to be a better candidate.

Chasing Proposal D

Vanterpool's Proposal D seems to be the most elusive of all. I have not found any designs that match his description. This bothers me, because I always felt that Proposal D was the most intriguing.

Many over the years assumed (perhaps from Breyer's lack of a description) that Proposal D would have its twelve guns in four triple turrets. Armstrong did not make a triple 15" turret. Indeed, the Royal Navy never had a main triple turret on any battleship or battlecruiser until after World War I. That said, there was one Armstrong "stock" design that did include a triple 15" turret: Design 768. She would have had eight 15" guns, with triple turrets fore and aft, and a twin turret amidships. This shows that Armstrong was at least thinking about a 15" triple turret. ⁷

It is interesting to note that neither Vanterpool nor Breyer say anything about triple turrets in Proposal D. Breyer does not describe the turret arrangement for Proposal D, but Vanterpool does: twelve 15" guns in six twin turrets. Vanterpool wrote that Proposal D would have had a hull design like Proposal C, but the gun arrangement of Proposal A.

Tony Gibbons' "The Complete Encyclopedia of Battleships" repeats Vanterpool's description of four designs, ⁸ but he includes a profile illustration of "Armstrong's final design". (as well as illustrations of Vanterpool's designs A and B) A casual reader might assume that this is Design D, however it looks very much like Armstrong's Design 781. In this profile illustration, there is no way to tell whether the ship has twin or triple turrets. Between the illustration with four turrets, and the description of a twelve gun ship, readers could be forgiven for assuming a triple 15" turret.

Additionally, there is once again no hull described in the various projects that matches Vanterpool's size description.

Conclusion

There seems to be significant evidence that Vanterpool based his proposals on Armstrong stock designs. There can be little question but that Proposal A is in fact Armstrong stock design 740. Design 764A seems like a good candidate for Proposal B. With additional information (i.e. a drawing) we might have definitive proof. I don't have a really good candidate for Proposal C, (though Design 685A is close) but Brook's minimal description of the 16" version of the "Improved Malaya" stock design suggests that it could be Proposal C. There is too little information on Proposal D to form any meaningful conclusions.

Postscript 1: The Actual Riachuelo, Design 781

Breyer wrote: "By August 1914 four designs (A to D) were prepared; no decision for any of them was, however, reached by the Brazilian client." ⁹

Sturton, and later Topliss, demonstrated otherwise. Sturton's 1970 Letter to the Editor clearly demonstrated that Brazil ordered Armstrong Design 781. Topliss writes in his 1988 article: "Armstrongs meanwhile offered eight designs in February 1914. Design numbers 781 - 788 (Tables XVI, XVII) . There were four variations on two basic designs, the odd numbers having a main armament of eight 15"/42 cal guns, the even numbers ten. In fact, drawings for only the first two designs were prepared, all other design folders referring to these for appearance and layout. The major differences between the designs was in their speeds and fuel, although unfortunately, the design folders give no horsepower figures. One design, 781, was ordered in May. This ship, to be named 'Riachuelo' was given the Elswick yard number 879. Although some work seems to have been started on her by Armstrongs before the outbreak of the First World War, she does not seem to have reached the laying down stage." Norman Friedman says much the same thing. ¹⁰

Because Sturton and Topliss cited reliable primary sources, I conclude that they were right. Armstrong Design 781, dated 25 February 1914, should be considered the definitive Riachuelo design.

Design 781: Sturton got it right in 1970!

	Design 781: the Actual Riachuelo
Displacement - Normal tons	30,500
Length Overall, feet	660
Beam, feet	94
Moulded Depth, feet	51' 1"
Draft, feet	28
Armament
Main	8 - 15"/42
Secondary	14 - 6"/50
Anti-Torpedo	10 - 4"
Anti-Aircraft	3 - 3"
Torpedo Tubes	2
Armor
Belts	13.5"
Turrets	?
Decks	2" - 2.5"
Machinery
Number of Shafts	4
Type	Parsons Turbines
Boilers	Babcock & Wilcox
Fuel	1200 tons coal, 3500 max.
	700 tons oil
Speed, knots	22.5

Postscript 2: Areas for Future Research

I intend to return to this subject if I can find additional data.

1. Additional information on Design 764A. A drawing would be especially helpful, as it might help us prove that it is the source for Design B.
2. Additional information on the 16" version of the "Improved Malaya". I suspect that this might be Vanterpool's Design C. From Peter Brook's description, we know that a drawing exists, and presumably there is basic information with it.
3. Any information from Brazilian archives. Most of the information we have is from the British side.

Thanks To......

The author would like to thank Larry Wells for his help with editing, and his other helpful suggestions. Thanks also to Alexei Sokolov for his help obtaining several rather obscure old articles.

Version History

Version		Date
Draft 1	0.1	19 January 2019
Draft 2	0.2	17 February 2019
Draft 3	0.3	6 September 2019
Version 1	1.0	30 October 2019

Footnotes

Copyright ©2019 David R. Wells. All rights reserved.