Increasing the number of boilers "on line" was not meant to increase the pressure in the system. Over-pressuring was bound to cause some sort of destructive failure in the mass of tubes, pipes, and valves that comprised the steam plant. So, pressure was maintained within a certain range. The 220 to 225 pounds reported is most likely the "high" end of the safe working pressure on Titanic's system.
Pressure is not the key to the puzzle. One single boiler could have overpressured the system with the engines not turning. But, the steam output from a single boiler could not have operated the engines even at harbor speeds. The reason is that a single boiler doesn't put out the necessary volume of steam.
For higher speeds, the ship needed more steam, not more pressure. Hence the need to add more boilers. Any power-driven vessel will continue increasing in speed under full throttle until the additional drag of greater forward motion becomes equal to the power output. Speed stops increasing when the drive system is "maxed out."
Any marine engine--internalor external combustion--will attain revolutions well beyond it's design limits without a load. It is the load of turning the propeller in the viscous medium of water that limits the revolutions within the "red line." The key is to find the best combination of pitch, diameter, and blade area to allow maximum safe RPM from the engines while achieving maximum speed through the water.
But, no engine will reach its design limits unless enough energy is supplied. With a steam engine, that's translates into the volume of steam available. Anyone who has worked with compressed air tools will understand. A miniature compressor will produce the same pressure as a monster commercial machine. But, the tiny machine can't turn a sander for more than a minute, while the big machine loafs along all day doing the same job.
Turning Titanic's massive engines at higher revolutions consumed massive amounts of energy. That's why the steam pressure dropped from cylinder to cylinder. And, the temperature of the steam dropped as well. Eventually, so much energy was extracted from the steam that it condensed back into water. In order to maintain speed it was necessary to replace heat in the system and that required more square footage of fire grates.
Those who have grown up with internal combustion engines look only at horsepower as an indication of the work a machine can perform. In steam, the limits are set more by the boiler than by the engine. If the engine draws more energy out of the system that the fire grates can replace, the ship will lose power.
So, the addition of more boilers was not to increase pressure, but to increase the square footage of fire grates adding energy (heat) to the system. It is a mistake to think that the additional boilers were meant to increase pressure; or that any high-speed run would have been accomplished by over-pressuring the system. (Remember Sultana!)
Quite obviously, those additional square feet of fire grate required the burning of more coal. As Sam mentioned, the cost in terms of energy of increasing speed goes up by the cube, so a few knots more speed can double the burn rate of coal (or diesel, too). Titanic did not sail with full bunkers, a fact that constrained any plans to set speed records. It is not surprising that whatever high-speed run Ismay had planned was set for later in the voyage when they would have had a clear picture of the coal reserves available with regard to the distance remaining to New York.
As an aside, the total amount of fuel aboard Titanic at departure was equal to the combined bunker capacity of boiler rooms #2, #3, #4, #5, and #6. The "ullage" between the amount of coal loaded and the total capacity of the ship's boiler room bunkers was equal to the bunker capacity of boiler room #1. The boilers in #1 were cold throughout the trip. These facts leads me to believe that no coal was bunkered in #1 and there was no intention to light off those boilers even for Ismay's "speed run." There was no coal aboard to feed them.
How fast did they want to go? That's a question that is never asked, but which must be answered to understand what Ismay had in mind. Titanic could not do the 26 knots of the Cunard boats, that's certain. So, it was really impossible to set any "official" record speed. From a business standpoint there was nothing to be gained from beating Olympic's maiden voyage speed. The public would hardly have been impressed. So, what was intended?
In my view, the alleged "speed run" planned for Monday was not in any sense an attempt to set a record. There was no record to be set. Instead, I believe it was a 22-knot performance test of the entire drive system from firebox to condenser. Titanic's props had been tweeked based on knowledge from Olympic. Hence, the interest in surpassing that ship's performance. A faster speed of Titanic compared to Olympic would have proven that those expensive changes were successful.
Strictly a suspicion, but I get the impression Olympic's speed did not live up to expectations. This led to the different propellers on Titanic. Prop design was a "black art" in 1912 prior to the knowledge gained in airplane design regarding foils. Propeller design was pretty much a "cut-and-try" at the time.
As an operator, I know that no shipping company willingly spends money that does not result in greater ticket sales. Speed sells, but it is a two-headed economic monster. On the one hand, ticket sales go up with speed. On the other, fuel costs can outpace the increasing revenue from fares. Profitability comes from finding the middle ground where income has gone up more than the cost of fuel.
And, something else. Speed is expensive. Ismay knew that if he could force Cunard to run its pair of "ocean greyhounds" at maximum speed, he would create an economic hemorrhage in Cunard's operations. Because the pair of Titanic and Olympic never operated the intended schedule, we will never know.
So, how fast was Titanic going? Boxhall said 22 knots based on the revolutions of the engines. The other surviving officers all testified to somewhat less, between 21 and 21.5 knots. The ship's first CQD position appears based on a 22 knot speed. Boxhall's own work confirms this when the time difference between 11:40 p.m. and midnight is taken into account. However, both the initial and Boxhall's corrected CQD coordinates are well west of where
Carpathia picked up lifeboats and the present-day location of the wreck on the bottom. Since a speed of 22 knots put the dead reckoning west of the actual ship's position, it is unreasonable to assume it was going faster because that assumption only increases the error in navigation.
We are now getting into the twilight zone created by what I call the "Daisy Minahan paradox" and the confusion over crew time versus civil time on Titanic. (Sam is now muttering, "Oh no! not the time question!) Without writing a book about it, my view is that Titanic increased speed to 22 knots at the same time it turned "The Corner." For ease of navigational computation, the dead reckoning was switched into April 15th hours. That's the explanation for the paradox of Daisy Minahan seeing Captain Smith leave the cafe after he arrived back on the bridge per Boxhall and Lightoller.
When everything is taken into account, the true duration of the run from "The Corner" to the accident is 6 hours 19 minutes. If you assume the turn was made 45 to 50 minutes late, this run at 22 knots produces a longitude equivalent to the 50 24 West of the original CQD messages. However, if you "correct" back to the physical location of "The Corner," then the longitude becomes 50 03 West, very close to the actual wreck on the bottom.
From my point of view, Boxhall used 22 knots because that's the speed Titanic's engines were turning for at the time of the accident. Captain Smith used this same 22 knots for his original CQD coordinates.
Boxhall used engine revolutions for his 22 knot speed because that was the standard practice for steamships in 1912. He testified that he did not use the reading on the taffrail ("Cherub") log trailed astern. Navigation texts of the era cautioned steamships against using trailing logs because of their inherent inaccuracies. Sailing ships, however, had to rely on these instruments as they had no propellers and, thus, no shaft revolutions to count. The trailing log of Titanic served primarily as "backup" information and not the primary source of distance for navigation.
The biggest source of inaccuracies in speeds determined by log readings comes from the duration of the run. The shorter the duration, the more critical the clock time of the reading becomes. Hichens's readings are particularly suspect in this regard. He was specific about the distance--45 miles--but we do not know the exact minute of the 8 p.m. reading (taken by someone else) nor the exact time of Hichens' reading. With only two hours, if each were off by 3 minutes, the total error would be on the order of 5%, or more than a full knot of speed.
Rowe's reading is on its face impossible. The reason is that he calls for a higher total average speed from noon greater than the speed being used by Boxhall at the time of the accident. Because the speed had been slower earlier in the afternoon, Rowe's reading should have produced an average of less than the 22 knots Boxhall used. Also, the timing of adding the new boilers indicates that there was an increase in speed late in the day, not during the whole day.
However, Rowe's reading becomes possible if the actual duration of the run is known. It was not 11 hours 40 minutes, but that plus the 24 minutes of "extra" time served by the 8-to-12 watch. This gives a total run of 12 hours 4 minutes, which produces an average speed of 21.5 knots for the day--right on the money with the testimonies of all surviving officers. That is, a speed of about 21+ knots increasing to 22 knots after "The Corner."
Based on my research, I am in total agreement with Boxhall that Titanic's engines were turning for 22 knots at the time of the accident. They had been doing so at least since "The Corner."
Ships are normally designed to operate at about 90% power as "full speed." If that assumption is applied to a 22 knot operational speed for Titanic/Olympic, then the two should have been capable of making about 24.2 knots with everything straining 100%. My view is that the goal for Titanic was to sustain 22 knots, proving the combination of fireboxes, boiler capacity, engine horsepower, and propellers.
Here's why I view it that way. An 18 knot ship requires about 6 and a half days for the open ocean crossing. At 22 knots, that drops to 5 days. You can't get down to a 4 day crossing without achieving a speed exceeding 28 knots. So, at 22 knots the open water crossing on an Olympic Class ship was functionally the same to the traveler (taking about 5 days) as a 26 not crossing on a Cunard high-speed coal burner. The difference was that WSL made the crossing at less cost--meaning more profit.
So, I don't see any reason for exceeding 22 knots with Titanic and every reason for proving that they had come up with the right combination to achieve a sustained 22 knots. This conclusion makes engineering sense but more important, it makes business sense. Supporting this conclusion is the testimony of the man who did the navigating, Boxhall, and the dead reckoning plot reconstructed using his testimony which matches the location of the wreckage on the bottom.
My view is that the ship was turning for 22 knots, but making good about 21.5 against a current setting 104 deg at about 0.5 knots. The longitude of the accident was approximately 50 03 West and the current took Titanic back east of the 50 West meridian as the ship sank.
-- David G. Brown
