deisel leak

[MtB]

Awesome! I bet that was caused by a jammed pressure relief valve. Or it was tied down to squeeze a bit more power out! 

[MtB]

56 minutes ago, mross said:

In industry, pressure testing is normally hydraulic, often with water.  This eliminates the risk of an explosive failure.  Certain pressure regulators need a constant, small flow to work correctly.  If connected to a tank and set at, say, 5psi, the pressure continues to rise until it matches the supply pressure.  This is the risk that most concerns me.

A tank seven foot wide and two foot long would experience a force of 13.6 tonnes under 15psi on the top (and bottom).  that's surely enough to cause something to bend or crack.

 

 

Yes but here we are not pressure testing, we are leak finding. Underwater probably. So air is the more appropriate fluid as it can be seen bubbling out. 

And we are testing a little tank built into the structure of a steel NB hull, in the stern hoop, probably 8mm thick, not a 7ft long x 2ft wide rectangular tank. Without doing the calcs, I'd guess the NB tank would be good for several hundred psi, given the tensile strength of mild steel and the thickness. 

[cuthound]

2 hours ago, mross said:

In industry, pressure testing is normally hydraulic, often with water.  This eliminates the risk of an explosive failure.  Certain pressure regulators need a constant, small flow to work correctly.  If connected to a tank and set at, say, 5psi, the pressure continues to rise until it matches the supply pressure.  This is the risk that most concerns me.

 

 

Not in my experience of 44 years in the critical power and cooling industry, always specified air testing to identify leaks. Water testing is too disruptive in case of a leak on a live site. 

Laterly welds would be non-destructively tested first, using either magnetic particle inspection or dye penetrant inspection first.

The largest welded steel I ever saw tested for leaks was circa 50,000 litres capacity. 

Edited July 1, 2017 by cuthound
Spillung

[mross]

40 minutes ago, cuthound said:

 

The largest welded steel I ever saw tested for leaks was circa 50,000 litres capacity. 

What was the design pressure and test pressure of this tank?

I have pressure tested marine boilers rated at 60bar and using air would have been criminal.  60bar is 870psi and the test pressure was higher.

[MtB]

I really do think a sense of proportion is needed here. We are discussing lightly pressurising a 100 litre steel tank created from 8mm steel by partioning off the very back of the elliptical stern of a narrowboat, not a marine boiler rated at 60 bar!!

[mross]

Let no one doubt this; putting 'several hundred' psi of air pressure in your fuel tank is incredibly dangerous.

Edited July 1, 2017 by mross

[cuthound]

1 hour ago, mross said:

What was the design pressure and test pressure of this tank?

I have pressure tested marine boilers rated at 60bar and using air would have been criminal.  60bar is 870psi and the test pressure was higher.

Specification was for a 50,000 litre gas oil tank to be tested by pressurising with air to 5 psi.

That is what the supplier provided and it was successfully tested at 5 psi.

Marine boilers rated at 60 bar will obviously require at different test regime. 

[mross]

I am not worried about people who apply a few psi but I do get upset when they suggest using higher pressures.  

Edited July 1, 2017 by mross

[MtB]

52 minutes ago, mross said:

I am not worried about people who apply a few psi but I do get upset when they suggest using higher pressures.  

The only person in this thread suggesting actual test pressures so far has been you! The original suggestion was just to pump some air into it, without mentioning any pressures. The implication being to use something like a bicycle pump.

 But fundamentally we are in agreement. Much above about 5psi risks damaging the tank, much above 15psi is beginning to approach the realms of danger. I suspect the tank would happily cope with 100psi but I wouldn't want to be closer than about 100yds while that happened. 

[Dinz]

Right then.

Tank height is probably 18-24".

This would give a head of less than1psi with water.

Diesel is at circa 0.7-0.8 Sg. So pressure head even less.

If gas is used instead of liquid then leak is more likely to show up.

So drain tank.

Clean up and get dry.

Turn off engine fuel isolations.

 Cap seal with tape or even a plastic bag round threads.

Via vent apply air pressure. Seal around tube.

A small tyre compressor will suffice. Even human lungs[non smokers] can deliver near 2psi. [dizziness permitting]

Remember that the volume of air needs to be compressed so this will take a fair time.

2-3psi will show leak. Either bubbles if below water line or soapy water on hull if above. [snoop]

This is a slow, controlled way to leak test.

Once leak found carry on with rest of tank as there may be more than 1 hole.

 Once you know where leak is then you can make an informed decision on repair.

I have repaired many motorcycle tanks by Argon filling as an inert and the welded up quite happily.[tig]

Edited July 1, 2017 by Dinz

[Sea Dog]

9 hours ago, mross said:

It's likely that a pinhole leak is due to a minor weld defect.  There is no reason to extrapolate this into a rotten hull.

 

9 hours ago, Mike the Boilerman said:

Yes there is.

We LURVE escalating minor problems like this into major disasters on here!

Gizza break fellas! I was merely suggesting that the thing needs repairing asap and advocating the prudency of further checks whilst it's out of the water. It didn't turn into a devastating explosion on my watch!

[mross]

7 hours ago, Mike the Boilerman said:

I'd guess the NB tank would be good for several hundred psi,

And you say you weren't advocating any pressure 

Edited July 1, 2017 by mross

[mizpah2]

sorry I asked ! it stopped this afternoon , maybe it took up with steam ! off we go again , more pressure stuff !

[rusty69]

2 minutes ago, mizpah2 said:

sorry I asked ! it stopped this afternoon , maybe it took up with steam ! off we go again , more pressure stuff !

Or maybe it is now below the level of the hole. 

[MtB]

On 01/07/2017 at 23:40, mizpah2 said:

sorry I asked ! it stopped this afternoon , maybe it took up with steam ! off we go again , more pressure stuff !

 

I suggest you pump the tank up to 60 Bar, and see if the leak reveals itself again. Should be fine according to mross, if I understood him correctly...  

[mizpah2]

7 hours ago, Mike the Boilerman said:

 

I suggest you pump the tank up to 60 Bar, and see if the leak reveals itself again. Should be fine according to mross, if I understood him correctly...  

i have no method of pumping it up , but am running a pipe from my gas locker from one of the gas cylinders , minus the regulator , I can not read the pressure as it is dark now but my wife has gone inside to get me her lighter - - - - - -

[luggsy]

5 hours ago, mizpah2 said:

i have no method of pumping it up , but am running a pipe from my gas locker from one of the gas cylinders , minus the regulator , I can not read the pressure as it is dark now but my wife has gone inside to get me her lighter - - - - - -

Ha ha nice one