Why doesn't boiler collapse?

corkgsxr · 17-01-2014 02:34PM

Thinking about it prv shouldn't leave air in. But they seem to

Wearb · 17-01-2014 02:42PM

shane0007 wrote: »

You are missing the point. There is no vacuum.
There would be a vacuum if you go from positive pressure to negative pressure. You cannot have both in the same pipework.

In the first instance you have brought the pressure to let's say 1.0 bar positive pressure. So let's say its a bungalow with 2.5m head height above the gauge.
This means that when the system is full of water under its own weight the gauge reading is 0.25 bar pressure.
We now overfill the system by raising it to 1.0 bar. We have no expansion & we heat it. The safety valve releases 4 litres water & it is now 96 litres when cold & showing a pressure of 0.8 bar.
The original fill volume would only be about 80 litres when cold if you were to open the highest point & see water with none spilling out.

I have reread the above lots of times and am having difficulty understanding it. I sometimes hit a brick wall like this when trying to understand some things. I will reread it later after thinking about it for a while.

Here is where my line of thought is at present. If you fill a metal can (lets say a jerry can) with hot water and cap it. Then let it cool. Is there not negative pressure in the can when it cools?

Pete67 · 17-01-2014 02:56PM

corkgsxr wrote: »

A normal safety valve will let in air in a vacuum situation.

I didn't know this, it seems counter intuitive to me - the safety valve should not open until there is more than 3 bar (say, depends on the rating) positive pressure difference between the heating system and the atmosphere.

Under the vacuum conditions described the pressure difference would be less than 1 bar, and acting in a direction that would tend to close the PRV rather than open it?

Large industrial systems often have vacuum breakers installed, which open when the internal pressure drops below atmospheric pressure to protect components from collapse. Maybe some domestic PRVs have a built-in vacuum breaker, but the few that I've dismantled did not have any sign of this.

If the expansion vessel has failed, and the PRV opened to release water to lower the pressure due to thermal expansion, then without some sort of vacuum breaker device, the system will go into negative pressure as it cools down, and there will be small pockets at the tops of radiators and other high points that contain no water. They will contain a low pressure mixture of water vapour, and any gases that come out of solution - oxygen, nitrogen, possibly also corrosion by-products such as hydrogen sulphide.

The radiators and piping will not collapse as they are strong enough to withstand the external pressure. Pex will probably compress a little, but being slightly elastic it will recover when the pressure rises.

Most systems are not 100% gas tight, so over time air will slowly be drawn in through valve stems etc and the pressure will slowly equalise. If the situation continues, as more and more water is expelled each time the boiler heats up, the system will gradually reach a point where the water quantity is low enough to trip out the low pressure switch (hopefully) or boiling will take place leading to boiler damage.

shane0007 · 17-01-2014 03:03PM

Wearb wrote: »

I have reread the above lots of times and am having difficulty understanding it. I sometimes hit a brick wall like this when trying to understand some things. I will reread it later after thinking about it for a while.

Here is where my line of thought is at present. If you fill a metal can (lets say a jerry can) with hot water and cap it. Then let it cool. Is there not negative pressure in the can when it cools?

No there is positive pressure of 0.1 bar per meter of head height so if the can was 1.0 meter high regardless of diameter, it would have a positive pressure of 0.1 bar.

Think of the system. The volume is 100 litres. It's the bungalow example. So the gauge pressure is 0.1 bar per meter of head height. The gauge reads 0.25 bar. If you unscrew a cap at the top of the system, no water would come out as this pressure would be 0 at this point but the system is full.
Now we raise the pressure to 1.0 bar. In order to do this we most add more water under force. So we bring the pressure to 1.0 bar. We have just added say 20 litres so the pressure now reads 1.0 bar.
But the system now contains 120 litres.
We heat the system & the prv lets out 4 litres. The system cools & the pressure gauge now reads 0.9 bar as the system now has 116 litres of water in it.

Now this continues every time we heat the system until let's say the gauge reads 0.25 bar. The system volume is now 100 litres. If we opened the prv nothing may come out due to air locking like holding a glad upside down but if we opened the top of the system it would empty. The exact same as emptying a radiator. We need to open the bleed screw to empty it.

Mods: can you add a "Donate" button beside my name. I should be paid for this :-)

Pete67 · 17-01-2014 03:20PM

shane0007 wrote: »

No there is positive pressure of 0.1 bar per meter of head height so if the can was 1.0 meter high regardless of diameter, it would have a positive pressure of 0.1 bar.

Think of the system. The volume is 100 litres. It's the bungalow example. So the gauge pressure is 0.1 bar per meter of head height. The gauge reads 0.25 bar. If you unscrew a cap at the top of the system, no water would come out as this pressure would be 0 at this point but the system is full.
Now we raise the pressure to 1.0 bar. In order to do this we most add more water under force. So we bring the pressure to 1.0 bar. We have just added say 20 litres so the pressure now reads 1.0 bar.
But the system now contains 120 litres.
We heat the system & the prv lets out 4 litres. The system cools & the pressure gauge now reads 0.9 bar as the system now has 116 litres of water in it.

Now this continues every time we heat the system until let's say the gauge reads 0.25 bar. The system volume is now 100 litres. If we opened the prv nothing may come out due to air locking like holding a glad upside down but if we opened the top of the system it would empty. The exact same as emptying a radiator. We need to open the bleed screw to empty it.

Mods: can you add a "Donate" button beside my name. I should be paid for this :-)

Shane, water is virtually incompressible, you will not be able to put 120 litres into a 100 litre system without bursting something.

Back to the can analogy - as you say regardless of the diameter the pressure at the bottom of a 1 meter high container will be 0.1 bar. But it is still zero at the top. If you fill the can, seal it, and heat it up it will quickly burst as the water tries to expand but has nowhere to go. If you fill it with hot water, then seal it, and allow it to cool, the internal pressure will drop below atmospheric, and the can will tend to be crushed slightly. If it is strong enough, it will resist the greater external pressure, and maintain a slight internal vacuum (or negative pressure if you like) if not, it will deform and crumple visibly.

shane0007 · 17-01-2014 03:32PM

Pete67 wrote: »

Shane, water is virtually incompressible, you will not be able to put 120 litres into a 100 litre system without bursting something.

Back to the can analogy - as you say regardless of the diameter the pressure at the bottom of a 1 meter high container will be 0.1 bar. But it is still zero at the top. If you fill the can, seal it, and heat it up it will quickly burst as the water tries to expand but has nowhere to go. If you fill it with hot water, then seal it, and allow it to cool, the internal pressure will drop below atmospheric, and the can will tend to be crushed slightly. If it is strong enough, it will resist the greater external pressure, and maintain a slight internal vacuum (or negative pressure if you like) if not, it will deform and crumple visibly.

Of course if you fill an open top can with water the pressure at the top is zero as it open to atmosphere. The pressure at the bottom would be 0.1 bar. It was an example to explain. I thought that would be a simple analogy to give but I did not know some people would not understand the simple principle.
The last time I checked there was not hot water coming out of my mains so I would not be filling by system with hot water.
If the mains is at 2 bar & left open, the system pressure will increase to 2 bar also as it is under compression. The amount of compression will be determined by the force upon it this trying to force more water into the system than it can hold thus an increase in pressure.
Commercial systems are competely different to domestic as they are far larger & then surface resistance plus pressure loss over the system must be taken into account. I am referring to pressure loss & not pressure drop as the two are very different.

Wearb · 17-01-2014 03:59PM

shane0007 wrote: »

No there is positive pressure of 0.1 bar per meter of head height so if the can was 1.0 meter high regardless of diameter, it would have a positive pressure of 0.1 bar.

Think of the system. The volume is 100 litres. It's the bungalow example. So the gauge pressure is 0.1 bar per meter of head height. The gauge reads 0.25 bar. If you unscrew a cap at the top of the system, no water would come out as this pressure would be 0 at this point but the system is full.
Now we raise the pressure to 1.0 bar. In order to do this we most add more water under force. So we bring the pressure to 1.0 bar. We have just added say 20 litres so the pressure now reads 1.0 bar.
But the system now contains 120 litres.
We heat the system & the prv lets out 4 litres. The system cools & the pressure gauge now reads 0.9 bar as the system now has 116 litres of water in it.

Now this continues every time we heat the system until let's say the gauge reads 0.25 bar. The system volume is now 100 litres. If we opened the prv nothing may come out due to air locking like holding a glad upside down but if we opened the top of the system it would empty. The exact same as emptying a radiator. We need to open the bleed screw to empty it.

Mods: can you add a "Donate" button beside my name. I should be paid for this :-)

Remember that we are talking about a system with no (effectively) EV so if it holds 100 litres cold at .1bar at the bottom and 0 at the top (assuming 1 meter head) you cannot get more than 100 litres (ignoring materials expansion) into it no matter how much pressure is applied. So now if we loose 4 litres as the water heats. What happens when the water cools back to the starting temperature.

And yes, and I do think you and a few others do deserve a donate button for all your help on this forum.
:pac::pac::pac::pac:

shane0007 · 17-01-2014 04:12PM

I have not calculated the compression factors of water as that would be just anal. I used the 120litres as an amount that we can imagine. The punt is most likely far less, but the principle remains the same. There is no vacuum & the water is being compressed to raise the pressure.

I was referring to no exapanion vessel.

Re: donation, I do also accept your bank account number, sort code, your mother's maiden name & of course, the name of your first pet.
PM only for security reasons :-)

Wearb · 17-01-2014 04:41PM

shane0007 wrote: »

I have not calculated the compression factors of water as that would be just anal. I used the 120litres as an amount that we can imagine. The punt is most likely far less, but the principle remains the same. There is no vacuum & the water is being compressed to raise the pressure.

I was referring to no exapanion vessel.

Re: donation, I do also accept your bak account number, sort code, your mother's maiden name & of course, the name of your first pet.
PM only for security reasons :-)

OK, if you are still interested, lets take this back to the can. We have a 1 meter high uncapped (or sealed with a 3 bar PRV) can holding 100 litres (brimful) of cold water. We heat the water to just under 100C which will cause (lets say) 4 litres of water to spill out the top. We then cap it and let it cool down. We now have 96 litres trying to fill a 100 litre space. To my way of thinking this can only happen if gasses are released from the water to occupy the space, or the can collapses a little, or a vacuum forms.

To expand this further If you were to accept the above, the system would opperate next start-up without the need for an expansion vessel, because as the temperature rises, the system would reach the state it was in when we replaced the cap at near boiling point.

All this deserves a donation.
A/C 100-212, Sort code 373.15, Mother Catherine Boyle, Dog Gnasher.

Oppps just noticed the PM note.

shane0007 · 17-01-2014 04:51PM

You can't use that example because too many factors are at play. As soon as you apply heat the water will expand. Once it's heated the volume will increase so as soon as you cap it the volume is the expanded volume & not the cold volume & therefore a weak vacuum would be caused. That is not related to your original question.
Secondly you are capping at atmospheric pressure (open to atmosphere) & not capped at 1.0 bar within.

.jacksparrow. · 17-01-2014 05:06PM

Pete67 wrote: »

Shane, water is virtually incompressible, you will not be able to put 120 litres into a 100 litre system without bursting something.

Back to the can analogy - as you say regardless of the diameter the pressure at the bottom of a 1 meter high container will be 0.1 bar. But it is still zero at the top. If you fill the can, seal it, and heat it up it will quickly burst as the water tries to expand but has nowhere to go. If you fill it with hot water, then seal it, and allow it to cool, the internal pressure will drop below atmospheric, and the can will tend to be crushed slightly. If it is strong enough, it will resist the greater external pressure, and maintain a slight internal vacuum (or negative pressure if you like) if not, it will deform and crumple visibly.

That's not really true, have you never seen a hydraulic test been carried out on a boiler?

Wearb · 17-01-2014 05:20PM

shane0007 wrote: »

You can't use that example because too many factors are at play. As soon as you apply heat the water will expand. Once it's heated the volume will increase so as soon as you cap it the volume is the expanded volume & not the cold volume & therefore a weak vacuum would be caused. That is not related to your original question.
Secondly you are capping at atmospheric pressure (open to atmosphere) & not capped at 1.0 bar within.

I don't see why both examples a not comparable. The materials are the only variables and we can easily assume that the can (for the sake of this discussion we can make it stronger) will react in the same way as the heating system, or boiler in particular.

We seem to be on two different tracks on this one and not understanding each other. At least I am not understanding everything you say. Perhaps I have an erroneous view of the behaviour of water under those different conditions. Must get an E-Book on it for to have on those shopping trips with "er indoors".

Wearb · 17-01-2014 05:26PM

.jacksparrow. wrote: »

That's not really true, have you never seen a hydraulic test been carried out on a boiler?

I assume the water in it is being pressurised and not compressed. I can't see how more water can be gotten in to it than was in there originally at the start of the test, assuming it was full to start with. Only way for more water would be for the boiler to expand before bursting.

Pete67 · 17-01-2014 05:26PM

.jacksparrow. wrote: »

That's not really true, have you never seen a hydraulic test been carried out on a boiler?

I have, many times. Most recently a 5 MW gas fired LPHW boiler. It takes very little water to raise the pressure to test pressure, certainly orders of magnitude less than 20% of the system volume. And some of this water goes to fill the volume increase which arises as the boiler tubes expand slightly under the test pressure. Water is virtually incompressible - I can calculate the reduction in volume per cubic meter at any pressure you want - it is negligible.

Of course, if the expansion vessels are online during the pressure test, then considerably more water is required, as you have to compress the air above the bladders to get to the test pressure.

Anyway, all this is getting away from the the original question - domestic sealed heating systems can end up in negative pressure (vacuum) if the expansion vessel fails and water is discharged via the PRV. Think about it - how else could air be drawn INTO the system?

shane0007 · 17-01-2014 05:39PM

Pete67 wrote: »

I have, many times. Most recently a 5 MW gas fired LPHW boiler. It takes very little water to raise the pressure to test pressure, certainly orders of magnitude less than 20% of the system volume. And some of this water goes to fill the volume increase which arises as the boiler tubes expand slightly under the test pressure. Water is virtually incompressible - I can calculate the reduction in volume per cubic meter at any pressure you want - it is negligible.

Of course, if the expansion vessels are online during the pressure test, then considerably more water is required, as you have to compress the air above the bladders to get to the test pressure.

Anyway, all this is getting away from the the original question - domestic sealed heating systems can end up in negative pressure (vacuum) if the expansion vessel fails and water is discharged via the PRV. Think about it - how else could air be drawn INTO the system?

As I outlined clearly before, I used the 20 litres as a more understandable volume. The compression factor would be more like 4 litres on a 100 litre system.

Show me a domestic system that the pressure is in vacuum, in that it will suck in a quantifiable amount upon opening.

Also hydraulic tests are hydraulic! There is no heat involved. I have had returned boilers to the manufacturer that leaked only when they were hot. When returned & tested by the manufacturer, they passed the hydraulic test as the heat was expanding the metal & opening the crack line which sealed when contracted.

.jacksparrow. · 17-01-2014 05:44PM

Pete67 wrote: »

I have, many times. Most recently a 5 MW gas fired LPHW boiler. It takes very little water to raise the pressure to test pressure, certainly orders of magnitude less than 20% of the system volume. And some of this water goes to fill the volume increase which arises as the boiler tubes expand slightly under the test pressure. Water is virtually incompressible - I can calculate the reduction in volume per cubic meter at any pressure you want - it is negligible.

Of course, if the expansion vessels are online during the pressure test, then considerably more water is required, as you have to compress the air above the bladders to get to the test pressure.

Anyway, all this is getting away from the the original question - domestic sealed heating systems can end up in negative pressure (vacuum) if the expansion vessel fails and water is discharged via the PRV. Think about it - how else could air be drawn INTO the system?

Ah right sorry I misunderstood what you were saying.

Pete67 · 17-01-2014 06:16PM

OK, the compression factor for water is about 1% at 150 bar. So a 100 liter volume of water pressurized to 150 bar would end up at 99 litres. The same 100l pressurized to 1 bar would only would compress by 1/150 litre or 6.7ml so would end up at 99.993 liters.

This is an interesting discussion.

Wearb · 17-01-2014 06:27PM

Pete67 wrote: »

OK, the compression factor for water is about 1% at 150 bar. So a 100 liter volume of water pressurized to 150 bar would end up at 99 litres. The same 100l pressurized to 1 bar would only would compress by 1/150 litre or 6.7ml so would end up at 99.993 liters.

This is an interesting discussion.

Are there gasses forced out from between the water molecules at high pressure or what allows the compression. Maybe you shouldn't answer that yet as it may bring this discussion off on another tangent.

It is fascinating. Great forum. Thanks also for your input.

shane0007 · 17-01-2014 07:01PM

Pete67 wrote: »

OK, the compression factor for water is about 1% at 150 bar. So a 100 liter volume of water pressurized to 150 bar would end up at 99 litres. The same 100l pressurized to 1 bar would only would compress by 1/150 litre or 6.7ml so would end up at 99.993 liters.

This is an interesting discussion.

I agree. The compression is really irrelevant though. It's the principle that is the factor.

shane0007 · 17-01-2014 07:03PM

Wearb wrote: »

It is fascinating. Great forum. Thanks also for your input.

These are the type of threads I enjoy most. It's where I learn most when everything is thrown out there & bashed from all angles.

It's now even better since I now have an income generated from it. Thanks for the details.

jimf · 17-01-2014 07:07PM

shane0007 wrote: »

These are the type of threads I enjoy most. It's where I learn most when everything is thrown out there & bashed from all angles.

It's now even better since I now have an income generated from it. Thanks for the details.

see lads theres no such thing as a free lunch with the Kilkenny boys

;)

scudo2 · 17-01-2014 09:47PM

Negative pressure/vacume is the biggest cause of heated plastic pipe kinking on a bend.

shane0007 · 17-01-2014 09:50PM

scudo2 wrote: »

Negative pressure/vacume is the biggest cause of heated plastic pipe kinking on a bend.

So it's nothing to do with just the heat then?

scudo2 · 17-01-2014 09:57PM

shane0007 wrote: »

So it's nothing to do with just the heat then?

Heat soften the pipe,but will not shrink it, slight vacume affect causes it to shrink internal and kink. In my poor humble common layman opinion. !!

Dtp1979 · 18-01-2014 12:00AM

Lads you're all over complicating things. Just follow this simple basic formula and it'll solve all your questions. It doesn't get simpler than this

jimf · 18-01-2014 12:05AM

Dtp1979 wrote: »

Lads you're all over complicating things. Just follow this simple basic formula and it'll solve all your questions. It doesn't get simpler than this

thank god somebody posted something I could understand

shane0007 · 18-01-2014 12:16AM

Dtp1979 wrote: »

Lads you're all over complicating things. Just follow this simple basic formula and it'll solve all your questions. It doesn't get simpler than this

Sorry but that formula is slightly outdated. It had been found back in 1997, a critical error was discovered. Instead of 1/2p, 7/8p should be used. This allows for the slight thermal bridging caused by the flux capacitor during the allowance factor considered for L + 1.

cerastes · 18-01-2014 12:35AM

Wearb wrote: »

I have reread the above lots of times and am having difficulty understanding it. I sometimes hit a brick wall like this when trying to understand some things. I will reread it later after thinking about it for a while.

Here is where my line of thought is at present. If you fill a metal can (lets say a jerry can) with hot water and cap it. Then let it cool. Is there not negative pressure in the can when it cools?

It seems this has been solved by some science :rolleyes:
but, could it be as simple as this, in the jerry can example, you fill it with cold water, ie start with cold water? then heat it, causing expansion, (not start with hot water, which upon cooling would contract anyway) if the pressure exceeds the PRV set pressure, then the expanded volume is vented, returning to the same or a negligible difference in pressure. Or has this been suggested?

Out of curiosity, are there negative pressure relief valves for plumping that vent or fill, to prevent a negative pressure occuring in other areas, such as tanks which seem to be more susceptible to buckling under certain conditions or maybe in larger systems, Im not a plumber, just curious, although Ive come across Messrs Boyle and Charles.

Also, Is this Pressure Law mentioned not some version of Charles Law?
Ive not really heard of the pressure law, whats proportional to what?
I just looked it up and obviously it must be temp and pressure

Dtp1979 · 18-01-2014 12:46AM

shane0007 wrote: »

Sorry but that formula is slightly outdated. It had been found back in 1997, a critical error was discovered. Instead of 1/2p, 7/8p should be used. This allows for the slight thermal bridging caused by the flux capacitor during the allowance factor considered for L + 1.

The flux capacitor only activates at 88 mph Shane!!

shane0007 · 18-01-2014 12:52AM

Dtp1979 wrote: »

The flux capacitor only activates at 88 mph Shane!!

Not always, 'cos I overtook one in me Transit Connect! Either that or he was low on garbage.
:-)

Why doesn't boiler collapse?

Comments