By Artisan
This week we take a look at the design and installation of water gauges. The water gauge is arguably the most important fitting on the backhead of any boiler and must be accurate and reliable. The continuing integrity of any boiler relies on the maintenance of a covering of water over the heated surfaces. In the case of the locomotive type boilers we are usually concerned with this means ensuring that the firebox crown is always covered.
Maintaining this situation is the driver’s responsibility but he inevitably relies on the water gauge to achieve this and it is important that the gauge does not mislead him with erroneous information. Unfortunately the possible faults either designed into a water gauge or arising from its installation result in the gauge indicating that there is more water in the boiler than is actually the case.
The most significant design feature leading to this state of affairs is the choice of gauge glass bore. We are all no doubt familiar with the capillary effect, which causes water to rise inside a glass tube when it is dipped into a vessel of water. The effect is the result of forces occurring between the molecules of the water and glass at the interface and the surface tension of the water. The distance the water will rise in the tube can be calculated for any given tube diameter and is inversely proportional to this diameter. The water in our gauge glass behaves in exactly the same way and in consequence will indicate a higher level of water in the boiler than is actually present. So how significant is the error?
I have calculated the capillary rise for a couple of pieces of gauge glass I had in the workshop and have measured the actual rise for the two tubes.The results are shown in the accompanying table. Fortunately, the surface tension of most liquids decreases with increases in temperature and in the case of water at 150 deg. C (approximately the saturation temperature at 80 pisg) it is about two- thirds of the value at room temperature. Even so, the effect is not insignificant and can easily lead to an error in indicated water level of ¼in. Clearly, the larger the bore of the gauge glass the smaller the effect will be. It must be emphasized that it is the BORE of the tube which is significant.
The first of the samples I used was a length of tube I had had for some time and which I have used for most of my own gauges. The second sample was a length of tube with a red stripe incorporated, purchased at a recent exhibition. The red stripe fired into the glass is magnified by the presence of water in the bore and provides a clearer indication of water level. A good idea in principle but in this case I consider the material quite unsuitable and it is destined for the waste bin. Although 0.5mm larger in outside diameter the bore is 1.5 mm smaller than the tube I usually use and could (would) provide a very misleading indication of water level in the boiler.
My personal opinion is that the minimum bore of a gauge glass should be 4 mm. The current Boiler Test Code makes no reference to the size of gauge glass but the Australian Code referred to earlier specifies a minimum bore of 3 mm throughout the gauge.
Other sources of error arise from installation faults. The most serious of these is the sharing of the top fitting of the gauge with another fitting such as a manifold. This practice is forbidden by the Boiler Test Code. The reason for this is that if steam is drawn off from the manifold by a service (such as a blower of injector) the pressure in the manifold, and therefore the top of the water gauge if it is shared, will fall.
The fall in pressure may well be very small – a few calculations for a typical manifold supplying an injector indicates a figure of 0.02 p.s.i. Not very much, but this translates to 5/8in. rise in the level of water in the gauge glass. Theoretically, the top fitting of a water gauge could be shared with a pressure gauge because no steam flow is involved – I saw such an installation at the Midland Exhibition – but it would be better to avoid the wrath of the boiler tester and stick with a dedicated connection for the water gauge!
The location of the connection on the back head is also of importance. In particular, the clearance between the inner end of the lower fitting and the fire box must be sufficient to allow free flow of water to the gauge without entrained bubbles of steam. Bubbles entering the glass make a nonsense of the gauge reading. The UK Boiler Test Code is silent on this subject but the Australian Code requires a minimum clearance of 5mm. The location of the lower gauge fitting should be such that the minimum level indicated (i.e. when the water is just visible at the bottom of the glass) should still leave an adequate covering of water over the firebox crown. The UK Boiler Test Code states that “when no water is showing in the glass there is still a safe level of water above the crown sheet of the boiler”. While we may understand the intent, this is an unsatisfactory statement since if there is no water showing in the glass there is no way of knowing where the level is – the boiler could be empty! Better to define the requirement as the bottom (visible) end of the glass being an adequate distance above the crown sheet. The Australian Code defines this distance as being 10% of the distance between the crown sheet and the outer firebox wrapper. This seems to be a reasonable rule of thumb but will not suit every boiler design.
While the positioning of the lower gauge fitting in this way may seem an obvious safety requirement some of our famous designers of the past did not consider it. The Martin Evans drawings for the boiler of my current project show the lower gauge fitting positioned such that the crown sheet of the firebox would be bone dry with water still showing in the glass. I raised the position of the bush ¼in. to satisfy the code requirements. If you are building to a published design it is a point worth checking.
The build up of scale around the inner ends of either of the gauge fittings is possible and will result in sluggish movement of the level in the glass. This is readily checked by blowing down the gauge and observing the speed of recovery of the level. This procedure is called
