boiler water gauge glass failure analysis 1

[chumet1]

i also posted this in the boiler section but i just wanted to get the materials experts point of view:
i have a 200psi locomotive boiler with water level gauge glasses breaking every 2-3weeks. i am new to the section (3months) and apparently this problem is a new phenomenon which just started about 6-8mnths back. the problem started when there was a new supplier of these glasses. we have reverted back to the old supplier but we have had 5glass failures in the past 2months. my question is, does water quality affect gauge glass service life?? because from the water analysis, most of the water quality specs. are out of range from the required specs. the following is a summary of water quality i got from the lab over a period of 2weeks, the recommended specs are shown in brackets:
P.H range=12.4 (<10.5); Total dissolved solids = 1460-2280ppm (2000-3500); M.alkalinity = 1460-2550ppm (250-850); P.alkalinity = 730-1490ppm (250-650); C.alkalinity = 325-1035ppm (100-300); total hardness = 0 (0-25); chloride = 2100-3400ppm (0-800); sulphites = 70-80ppm (30-90)

 

[btrueblood]

Yes, glass can corrode. For the high pH and chloride content you show, you should be using a good grade of borosilicate (e.g. Pyrex) glass, preferrably annealed. You might check how the glass is restrained - it should have a somewhat loose axial fit at room temperature, and sealing provided by o-rings along the cylinder o.d. Too tight a fit, and thermal stresses may be causing the breakage.

 

[EdStainless]

Thermal stress is my though as well. Some one may be tightening these too snug, or a seal ring may have been lost in one of the replacements.
The glass expands less then the metal holder will, but the glass does not want to be flexed or bowed. Check clearances, fit, and seals.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[chumet1]

thank you for your swift response. i will have to check on the clearance, fit and seals. broken fragments show a decrease in tickness of the glass wall which i think leads to its eventual failure. then i guess that is caused by the high pH and chloride content like btrueblood put it. do total dissolved solids play a part since these can be controlled by frequency of blowdown because i had to tell the operator to blow down every start of shift, which they were not doing

 

[btrueblood]

If the glass is corroding, it doesn't take much surface loss before it will shatter, typically. You may see signs of surface crazing (the cloudy look that glass dishware gets after years of dishwasher washing is an example), but not much thickness change. The wall thickness variation you see in the broken pieces was likely there to begin with, which is another sign of cheaply made tubing. Order some new tubing, and inspect it very well before installing - good drawn borosilicate sight glass tubing should have a remarkably smooth wall (sight down the length, both i.d. and all around the o.d.), no visible inclusions (typ. bubbles), and the bore diameter should be very uniform and cylindrical - a pin gage that just fits at one end should slide all the way through without hanging up. Reject any tubing that doesn't have these characteristics, as it will likely rupture due to the stress concentrations caused by the non-uniformities.

 

[chumet1]

these glasses come with a 4mm wall thickness when they fail they either explode or just crack. the ones that explode you cant pick up the pieces but the ones that crack i measured wall thicknesses of approximately 1mm for most of them. and from the little research i have just made after you mentioned high ph corodes glass, it shows high ph can also lead to gauge glass thinning. i guess i will have to confirm the glass material composition from the supplier and also maybe research and try the borosilicate glasses

 

[EdStainless]

sounds like they may have sent you the wrong type of glass.
It should take months to get much attack and the surface should look frosted.


= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[btrueblood]

Chumet, are you located in the U.S./North America? Try

http://www.McMaster.com

if so, they sell borosilicate glass tubes in various sizes.

 

[EdStainless]

I am beginning to suspect that the original ones were Vycor, not borosilicate.

Has your pH control improved?
If you were running that high that often you may need to inspect the entire system for signs of caustic cracking in steel.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[rustbuster]

I had a similar issue a few years ago and through investigation determined that the Operators were not warming up the glasses properly when placing them in service. Thermal shock and thermal fatigue was the result. Check to see how they are being commisioned / operated.