Home - Dits & Bits - BR3043 - Part Two - Progressive Development of Design and Equipment - Chapter 23

Chapter 23: Pumping and Flooding Arrangements

1. This chapter is concerned with water ballast tanks only. Mention will be made of the blowing arrangements although this service is dealt with in more detail in Chapter 24. Oil fuel tanks are discussed in Chapter 25 and tanks in connection with the operation of the torpedo tubes in Chapter 27.

2. The Holland boats were fitted with the following bilge pumps:

  • A Marburg type main bilge pump, made in America, sited at the after end of the engine and driven by gearing from the main engine. It was capable of discharging against a pressure of 40lb/in². This pump could be unclutched and thehp air compressor driven from the main engine through the same gearing.
  • A motor driven auxiliary bilge pump slung overhead just abaft the conning tower, also of the Marburg type, which could discharge 3500 gallons per hour against a pressure of 100lb/in² it had a 2in sea suction. The makers were E. W. Bliss.
  • A hand pump sited just forward of the engine.

3. All these pumps were connected to a 3½in main line which ran forward from the main pump with suctions to various tanks and bilges. Originally water to the engine cylinder water cooling jackets was supplied from the main line but later on a small pump geared from the main shaft with suction direct from the sea was fitted for this purpose.

A trimming line ran between the forward and after trimming tanks. This line was cross-connected amidships with the main line. Water was blown from one tank to the other; it could not be pumped between the tanks.

4. The various systems to individual tanks were as follows:

  • The main ballast tank was flooded through two 7in. Kingstons (only one was originally specified) which were placed as low as possible in a sump adjacent to the fore end of the tank so that all working parts were accessible. The Kingstons were operated by rod gearing with handwheels at the sides of the torpedo tube. It was vented by a 3in pipe in the crown of the tank each side at the after end, the two vent pipes being led outboard separately through 3 in valves on the hull so that there was no question of cross venting. A 2in suction was led from the main line and the tank was blown from thehp air line through ¾in pipes with reducing valves at 10lb/in2. By-pass valves were fitted at the reducing valves so that the pressure could be increased in an emergency. Although it has been stated that only one blowing line was fitted to the main tank it is assumed that the line was split with leads to the Port and Stbd sides of the tank.
  • The buoyancy tank was flooded through a 9in sea valve or Kingston (specified as 7in. ) operated by rod and crank from just below the control room flat. A hand operated vent was fitted on the top of the tank to vent inboard. It had a l½in suction from the main line and was blown from thehp main in the same way as the main tank.
  • The auxiliary ballast tank had a 3in. Kinston, a 1in. vent pipe led outboard at the top of the pressure hull through a 1in valve, two 2in suctions from the main line and was blown as for the main tank but never above 10lb/in2.
  • The trimming tanks were cross connected by a l½in. trimming line and water was blown from one to the other at 50lb/in2 through ¾in leads from thehp air line reduced. The tanks could be flooded and pumped via the main line through the connection to the trimming line amidships.

5. The pumping and flooding systems in the A Class to C Class are not available in detail although a diagrammatic arrangement of water suction pipes and connections for A5-12 is shown in Fig 23. 1 and is probably typical for these classes. Some changes in principle took place the main points being:

  • Although the large forward main ballast tank in A1 was flooded through two Kingstons as in the Hollands from then on one Kinston only was fitted to each main tank and the size consistently 7in. The Kingstons were often fitted above the tops of the tanks in WT boxes with large tail pipes to the bottom of the tanks. Although working parts of the Kingstons were made more accessible this was at the expense of time to flood the tanks.
  • The buoyancy tank in A1 was fitted with a 3½in . Kingstons, reduced from 9in the Hollands, but in later vessels the Kingstons was deleted although the blowing line was retained. This was undoubtedly done to prevent the tank being used as a quick-diving tank.
  • Kingstons were not fitted to auxiliary ballast tanks as in the Hollands. The blowing line was retained and the main line suction increased from 1½in to 2in.
  • From A2 onwards water jackets were fitted round gasoline tanks.

6. From the Holland boats the pumping capacity increased as the boats got larger and the number of tanks and quantity of water to be dealt with increased. In general in the A Class and B1, two motor driven main bilge pumps were fitted near amidships one port and one starboard. The 25hp motor for the port pump also drove two air compressors and in some boats the 15hp starboard motor also drove a rotary air pump. A 5hp motor driven auxiliary bilge pump was sited at the after end of the engine room and a hand pump was fitted forward. Each pump had a sea suction and was connected to the main line.

From B2 and in the C Class the air compressors were removed from the control room and only one air compressor was fitted in the engine room with its own 30hp motor. The two main bilge pumps were sited amidships as previously each with its own 15hp motor. The auxiliary pump aft was deleted.

7. The D Class was a radical change in design with the introduction of saddle tanks and this had its effect on the pumping and flooding system. D1 had twelve external main tanks, six each side. Each tank had a 9in. Kingston, 8in. in end tanks, hand operated from inboard at main flat level through rods and gearing. The vent valves were hand operated from the corresponding positions of the Kingston handwheels and in some cases this meant extensive rod gearing inside the main tank. Each tank had a suction and flooding pipe with a valve at the pressure hull and these pipes ran to a common main inboard each side connected to the bilge pumps. Each of the four internal main tanks had its own Kingston hand operated locally.

Fig 23. 1

Fig 23. 2

Two bilge pumps driven by a common 30hp motor were sited on the starboard side of the control room, a third bilge pump with its own 30hp motor on the port side of the control room and a fourth pump forward on the port side with a 30hp motor which also drove anhp air compressor. The pumping facilities were therefore good. So far the policy for surfacing was to blow the main tanks withhp air as required to obtain the necessary buoyancy, then close the Kingstons and pump out the main tanks using the bilge pumps.

8. In D2-8 ex D4 an interesting experiment was tried. Only three external main tanks were fitted each side and a flooding trunk of approximately 2 tons capacity was built in the saddle tanks each side in the vicinity of the bulkhead between the two foremost tanks. Each trunk was flooded from the sea through a 20 in. Kingston. Three 11½in flooding valves were fitted inside each trunk each with a tail pipe to one of the three external main tanks. In addition 7in flooding valves were fitted in the trunks with a tail pipe one to each internal main tank. Each trunk could be pumped out through a 5in suction valve. These arrangements are shown in detail in Plate 10 and Plate 11.

With this scheme all the flooding handwheels for all main tanks were concentrated at control positions port and starboard just forward of the control room. In addition the vents to all tanks other than the aftermost external main tanks could be operated from the control positions. Port and starboard ballast pumps were sited at the control positions and these would appear to be concerned exclusively with dealing with main ballast water. Although not clearly shown, it appears that the main ballast water could only be pumped or blown out via the trunks.

The one advantage of this arrangement was the central control. The time taken to flood the tanks and hence the time to dive must have been considerably greater than with Kingstons to individual tanks. The time taken to obtain full buoyancy must also have increased.

The two ballast pumps were 10 in Roturbo ballast pumps in D2 and Gwynnes ballast pumps in D3 and D5-8. In addition each boat had two bilge pumps.

For some reason the flooding trunks were not fitted in D4, which was the odd boat out in a number of ways, which had six external main tanks each flooded through its own Kingston. However 'air blowing pumps were tried in this boat as a substitute for centrifugal pumps for quick emptying of the main tanks. Although not fitted in E1-6, air blowing pumps were fitted in E7-11. This would appear to be the start of what was called later on the LP blower. Another reference states that LP blowers were first fitted in E5 and this may have been the first production model of the development of the pumps in D4.

9. After this experiment, the E Class reverted to a hand operated Kingston and vent to each main ballast tank. Within the next three years we were building a number of double-hull boats with much more main ballast water to be dealt with and much of it high up in the boat. This type of boat was very slow in diving because of the time taken to flood the tanks especially the controlled free flooding spaces above the LWL.

10. In the first double-hull boats of the S Class, V Class, W Class and F Class and Nautilus the Kingstons and vents to main ballast tanks and the flooding valves or scoops and vents to controlled free flooding spaces were all hand operated. In the V Class and Nautilus the Kingstons . for the main tanks were high up in the tanks and fitted with tail pipes to the bottom of the tanks.

For the Swordfish Scotts claim to have 'designed a telemotor system by which valves at a distance could be controlled from a centre position', and they applied this system to the vent and flooding valves of the controlled free flooding superstructure, see Fig 23. 2. The firm also paid considerable attention to the matter of central control. The main vent valves for the main ballast tanks were placed in the control room and the main vents were joined to a master vent valve so that, Kingstons being open which was the normal condition, all tanks which required to be flooded for diving could be vented and filled by opening this one valve. This did have the disadvantage of long leads of piping from the tanks to the vent valves. However, the operation was centralised. In addition the blowing valves for all the main ballast and compensating tanks were in the control room and all these tanks could also be blown in an emergency from positions in the forward and after ends of the boat.

In the G Class the vents and Kingstons in the main tanks were hand operated as in previous classes but the vents and scoops in the controlled free flooding spaces were telemotor operated as being fitted in Swordfish.

In the J Class although designed with large controlled free flooding spaces these spaces were eventually combined with the main tanks and telemotor operated vents to main tanks fitted for the first time. This change would appear to have been made after building had started and the K Class was being designed.

11. As designed in the K Class the external tanks and arrangements for flooding and venting was as in the J Class then building but this meant long leads of piping for venting the tanks below the baling flats and also the need for scoops for the upper tanks. There were forty of these vent valves scattered throughout the hull and although no service experience had been obtained in the Swordfish and the G Class to adopt telemotor control was the obvious thing to do. A Westinghouse hydro-pneumatic system was at the time being fitted in a floating dock building for the Admiralty to operate the flooding of the tanks and a similar system was fitted in the K Class. The baling flats were deleted, so that the upper controlled free flooding spaces disappeared. This made ten main ballast tanks each side, each of which was fitted with a vent at the highest point at the forward and after end of the tank. Each tank was fitted with a Kingston but later on the Kingstons were abolished and holes cut in the bottom of the tanks to get more rapid flooding, except for the two central tanks each side which retained the Kingstons and were arranged to carry emergency oil fuel.

In the M Class scoop valves were fitted to the main ballast tanks just below the surface water line. Additional Kingstons were also fitted. The scoops and vents were telemotor operated. This was to assist in quick diving.

12. The period 1914-16 had therefore seen the introduction of telemotor and centralised control to assist flooding and diving time. In the K Class Kingstons to main tanks had been abolished. The specifications for the L Class in 1916 stated:

  • The main ballast tanks to be flooded through holes in the bottom of the tanks except that hand operated Kingstons to be fitted to external tanks designated to carry emergency oil fuel and to internal main tanks.
  • For venting the external main tanks a pipe is to be led from the highest point in each tank through the inner hull and thence outboard with the upper end of the pipe led through the hull into the superstructure. Valves to be fitted where the pipe passes into and out of the hull.
  • The internal main ballast tanks to be vented outboard, a pipe from each tank being led to a group chest having a large branch connected to the main vent valve which could be operated by hand or telemotor.
  • The vent valves to be telemotor operated and arranged in groups, each group controlled from a change valve.
  • The external main tanks to be blown off the 2500lb/in² system loaded so that the pressure in the tanks cannot rise above 15lb/in². Blowing connections to the internal main tanks to be led from a group chest and the pressure to be 50lb/in².
  • Two motor driven LP blowers to be supplied for the external main ballast tanks with an output of 750ft4of air per minute at 11lb/in². Valves to be arranged so that each tank can be blown independently of the others.

13. The policy at the beginning of 1916 was therefore telemotor operated vent valves for all main tanks, no Kingstons in external main tanks except those used as emergency oil fuel tanks, hand operated Kingstons to internal main tanks, and blowing lines from thehp line to all main tanks reduced so as not to exceed 15lb/in² in external tanks and 50lb/in² for internal tanks. All these operations were controlled from the control room except the Kingstons where fitted. In addition LP blowers had been introduced.

14. There is some doubt whether Kingstons were omitted from the main tanks in the L Class. They had been deleted from the K Class after the first boats had been completed, and this was after the L Class started to build, but it would appear that there was a change of thought after the war.

In March 1924 the question was raised of omitting Kingstons to all main tanks. The reasons why Kingstons 'had been retained in external main tanks' were discussed as follows:

  • It was considered that reliance could not be placed on vent valves alone to keep water out of the main tanks on the surface. RA(S) stated that they could and that throughout the war and subsequently vent valves had been relied upon both at sea and in harbour, DNC's point was safety adversely affected in harbour in case vents were accidentally opened.
  • It was considered that when the submarine was rolling in a seaway, water might enter through the open holes and thus reduce the surface buoyancy. RA(S) stated that heavy rolling had never been shown to reduce to any extent the buoyancy and that in any event a few strokes of the LP compressors would restore full buoyancy. DNC replied that endurance and speed would be affected, to which RA(S) remarked that this had not been experienced on service.
  • It was considered that the open Kingston valves offered a greater resistance to motion than when the valves were closed. RA(S) considered long slots in the bottom of the tank no more resistful than Kingstons either open or shut.
  • It was considered necessary to be able to close the Kingston in order to enter the tank without having to go into dock.

From the discussion the policy was evolved to fit Kingstons to about three tanks each side for safety when in harbour. Coupled with this was the fact that sincehp blows were used on the main tanks on all occasions of surfacing they were to be fitted only to main tanks without Kingstons.

15. In 1928 trials were carried out in L25 and L22 to ascertain the loss of speed due to having Kingstons open in a seaway when the vessels ran on the same course for two hours in a sea 3-4. The results were:

  Kingstons Rev/Min Distance
Per Mile
L25 Open 330 24. 9 190 7. 63
Shut 330 23. 5 180 7. 66
L22 Open 327 26. 5 148 5. 58
Shut 323 26. 7 147 5. 50

From this it appeared that the Kingstons did not make much difference to the consumption. The consumptions were measured using Scotts measuring apparatus and with the vessels rolling must not be taken as absolute. The difference in consumption of the two vessels may be attributed to the fact that L25 had mine tubes in the tanks open at the top. A greater loss of speed will occur through the rise and fall of water in these free venting tubes than would occur in air-locked tanks. From the above results it must be assumed that the tanks were blown periodically, see para 14.

Oberon did some trials and found that steaming into a long head sea at 9 knots the ship lost a knot in speed. The main tanks were not blown and on arrival in harbour she was found to be drawing 1 foot more than usual due to water having entered the tanks.

At this time it was acknowledged that there would be a loss of speed in rough weather due to water entering the tanks if the Kingstons were open. DNC stated that 'it is difficult to estimate exactly what loss of speed is involved' but continued that 'it is improbable that the loss in power would reach 300shp in very adverse weather'. Whatever the loss might be Kingstons must be kept open on war patrol or passage and the loss accepted.

16. So we come to the standard pattern of the 1920 designs as typified by the Odin Class with three external main tanks each side fitted with hand operated Kingstons and the remaining eleven tanks flooded through holes in the bottom, all tanks fitted with 7in. telemotor operated vents except No 1 which had a 12in. vent, those tanks not fitted with Kingstons having anhp blow direct from the 3500lb/in² main and all tanks with an LP blow at 15lb/in2 from the LP blower.

17. Typical of the LP blowers were those in Perseus which had two Reavell blowers, motor driven, one being coupled to the capstan motor and the other to the auxiliary drive motor.

18. The arrangements fitted to other tanks, and last mentioned in Paragraph 5 above, undoubtedly changed in minor detail through the classes. Because it was the first of the double-hull boats built in this country some details of the flooding, pumping and trimming system in the S Class are given as follows:

  • The four main tanks were flooded each through two 8in Kingstons on port and starboard sides of the vertical keel and the compensating tank through one 8in Kingston. The Kingstons were operated through rod gearing from deck plates.
  • The main line ran from the fore end to the after end of the boat in a single line and passed through the double bottom in way of the double-hull. It was of 5in bore in the externals and 3½in bore at the ends. Two 3in. suctions were taken from the main line to each main ballast tank and one 3in suction to the compensating tank each fitted with a 3in suction valve. 3in bilge suctions with valves were taken to each main compartment. A lead from the main line was taken amidships to a 3½in sea flooding valve and two 3½in pump suctions to pumps in the control room. Two sluice valves were fitted in the line at the forward and after ends of the double bottom to isolate the forward and after sections of the main. All valves were operates from deck plates inboard.
  • A single 2¼ in trimming line ran between the forward and after trimming tanks and followed the main line except that it passed from the double bottom through the control room and the only valves on the line were in the control room.
  • The only other piping was in connection with the torpedo tubes and associated tanks. Port and starboard 3½in sea valves were fitted abreast the after end of the tubes with valves chests from which pipes led to the torpedo tubes, torpedo compensating tank and spare torpedo tank.
  • In S1 all the piping was of copper and the valves of gunmetal. In S2-3 the piping in the double bottom was made of galvanised steel. All piping and valves were tested to 120lb/in².

19. Passing to the L Class these boats were specified to have two bilge pumps on a fore and aft flooding and suction main, each pump with a separate sea suction. Branches were taken to the auxiliary ballast tanks and WRT's. The after pump could be used for circulating water to the main engines in case of emergency. A hand operated bilge pump was also fitted for wash deck purposes. Vents from auxiliary ballast tanks led to a group chest and venting was either inboard or outboard.

20. An important development was the introduction in 1926 of the 3-way Kingston for quick flooding and blowing tanks. The overseas patrol boats from Oberon onwards were fitted with two auxiliary tanks A and B forward and two X and Y aft. These tanks in the Odin Class, which is typical, were fitted with 8in telemotor operated Kingstons and 4in telemotor operated vents and could be blown direct from the 3500lbhp line and could be opened to the sea down to maximum diving depth. They were tested only to 75lb/in² because the pressure hull was not designed to withstand internal pressure and when opened submerged the pressure was equalised. The tanks were fitted with a 25lb/in² sea relief valve. With the introduction of these quick flooding and blowing tanks the two trimming tanks were deleted.

Although in the Oberon A and Z tanks were designed for quick blowing and B and Y tanks for quick flooding it was desired in the Odin Class that all four tanks should be both quick flooding and quick blowing. DNC considered the 3-way Kingston desirable for these tanks and RA(S) considered such an arrangement 'almost essential' but suggested a further safeguard of 'so placing the operating handles that air pressure cannot be put in the tank until the Kingston operating gear has been put in the blow position'. This was considered reasonable and a telemeter operated interlock was placed in the blowing line. The Kingston had three positions - 'open', 'shut' and 'blow'. The operation of the valves was intended to be that thehp air control valves were kept open when diving and blowing was effected by pulling the Kingston control valve to the 'blow' position. The system however was not so used by submarine officers because it involved reliance on the interlock which they considered might stick or leak, the tanks were blown by putting the Kingston to 'blow' or 'open' and cracking thehp air control valve, the interlocks having been disconnected permanently by the ship's officers.

A conference was held to discuss these points, and it was decided to abolish the interlocks, retain the 3-way Kingston and fit a differential sea relief valve to the tank. There was a general fear that somehow it would be possible to get an excessive pressure on the tanks. Trials were carried out in Oswald and it was found that on opening wide thehp valve on the panel as quickly as possible from a group at 3300lb/in² pressure, the maximum pressure at the panel was 1900lb/in² and the maximum at the relief valve in B tank was 145lb/in², i. e. this was the maximum it was possible to get on the tank.

As a result of these trials the following conclusions were reached:

  • A sufficiently large relief prevents damage to the tank but it is considered that two should be fitted as in boiler practice.
  • A small pocket fitted at the top of the tank would be an additional safeguard. The reason for this was that if the tank was full of water and excess pressure was put on, water came through the relief valve. To prevent this the top of the pipe to the relief valve was led to a curtain in the top of the tank which formed a small air pocket. When under pressure air was relieved.
  • hp blows must be fitted to tanks that it is desired to blow quickly.
  • In Oberon Class, Odin Class, Parthian Class and R Class, to retain existing blowing arrangements but remove telemotor operated interlocks. To fit 20lb differential sea reliefs to prevent excess internal pressure when blowing at shallow depths.

This was all approved in 1930 and fitted in the Odin and later classes and fitted retrospectively in Oberon.

21. The services fitted to other tanks in the Odin Class are given in detail in Chapter 14 Paragraph 53.

22. In the 1930's oily bilge pumps were fitted in new construction to prevent the main line being fouled by oil fuel.