The BRE Garston Ball Valve

In the early 1950’s Building Research Station (now known as BRE) received frequent complaints about ball valve failure. One local council in Hertfordshire wrote that they had to replace nearly ‘a hundred valves every week on their estates’; another in Essex wants to know how to stop the floats dropping off; a plumber in Hayes, Middlesex, wrote that he was losing customers because the new valves he put in did not last six months; and a man in Ventnor wondered if it was sabotage by his plumber that his valve begins to leak a few months after fitting. There is also the sad story of an elderly lady in Golders Green who spent the whole night pulling the chain in the lavatory. This was to stop the flood in a cistern where the ball valve had failed.

The ball valve or to be correct the float valve regulates the flow of water into the lavatory cistern or tank in the loft. It closes when water reaches a predetermined level. Failure was usually due to piston sticking or cuts or channels cut by water flowing into the seating. The rubber washer no longer forms a watertight seal.

For a long time these cuts were thought to be due to chemical action by the water. Experimental testing at BRS showed that they were due to microscopic hammering or blasting called ‘cavitation’. Water under mains pressure flows past sharp edges and corners as a consequence of a rapid pressure drop. Water vapour bubbles form which rapidly collapse increasing the water flow locally to hit the metal with considerable force. As a result of repeated action, small grains of metal are detached and cuts have formed, leading to ultimate failure of the metal. The phenomenon is often heard as sizzling noise.

Having discovered the cause of the problem, BRS then looked for ways of eliminating the problem. One way would be to reduce the water main pressure but the installation of suitable valves would be expensive. Throttling the stop valve could have been beneficial, but at the expense of unacceptable tap flow. BRS redesign reduced cavitation by rounding off all sharp corners and limiting the gap between the seating and the sealing washer. The problem of sticking pistons was overcome by introducing a thick rubber disc with all the moving parts placed on the dry side of the disc.

A local brass founder manufactured 500 valves sent to various Water Authorities for trials. Comments were generally very favourable both about the valve’s performance in controlling the water level and also that it was much quieter than other valve patterns. These reactions encouraged the development of a Mark III valve which improved the flow rate. It also introduced measures to prevent plumbers over tightening the body and deforming the rubber washer. The critical element is the nozzle and trials were undertaken with parts manufactured from stainless steel and nylon. Reports were received of nylon nozzles being damaged when the ball valve water froze solid. Tests at BRS showed no damage to the nozzles; the copper pipe feeding the valve split relieving the pressure.

Original ball valves were manufactured to British Standard BS 1212 and the Garston ball valve is now retitled the diaphragm ball valve. It is manufactured to British Standard BS 1212-2 for brass valves and BS1212-3 for plastics. The main difference from the original design is that requirements to prevent backsiphonage mean that the outlet is at the top of the valve. This is done with a U-shaped pipe diverting water into the cistern.

In terms of the number of valves manufactured and benefits gained, the Garston Ball Valve has achieved the greatest impact of any innovation produced by the Building Research Establishment.

________________________________________

Further reading:
A Sobolev, A New Type of Ball Valve, The Plumbing Trades Journal, 1955, Vol. 3, No.8, pp234 & 236

A Sobolev, Failures of Ball-Valves and their Remedies, Journal of the Institution of Water Engineers, Vol. 9, No. 2, March 1955, pp 208-222.

A Sobolev, Design of an Experimental Ball-Valve, Journal of the Institution of Water Engineers, Vol. 10, No. 7, November 1956, pp 552-557.

A Sobolev, An Experimental Ball Valve, The Plumbing Trades Journal, 1956, 36, (3), 58-60

A Sobolev, BRS Ball-Valve Trials, Journal of the Institution of Water Engineers, Vol. 11, No. 6, October 1957, pp 497-503.

A Sobolev, The Effect of Freezing on Ball-Valve Seatings made of Nylon, Journal of the Institution of Water Engineers, 1958, 12, (7) November, pp 515-518.

BSI BS 1212-2 Float Operated Valves (Excluding Floats) Part 2: Diaphragm Type Float Operated Valves (Copper Alloy Body)

BSI BS 1212-3 Float Operated Valves Part 3: Specification for Diaphragm Type Float Operated Valves (Plastic Bodied) for Cold Water Services Only (Excluding Floats)