Fluid dynamics

[Calculus]

Not a DIY question but this is something I want to understand before I die.

I have a gravity shower fed from the hot water cylinder and the cold water tank. Every other time I turn on the shower the water is only flowing about half as fast as it ought to. Usually I can fix this by holding the hose stretched out and downwards and giving it a shake. There is a gurgle and the water starts flowing at its full force. If this doesn't work I can normally fix the problem by unscrewing the hose (see pic) and letting the water run fast for a couple of seconds. Again this is accompanied by a momentary gurgling and spluttering.

The spluttering suggests air is the problem, but if so, is the air blocking the hot or cold feed, and is so, whereabouts?

But this leads to a couple of mysteries. The first is, do 'airlocks' actually exist? The only mention of them I can think of is in black and white Hammer horrors. The guests are in the old mansion's drawing room when the butler announces the carriages are unable to come because of the stormy night. Then there is a loud clanking. As the women go hysterical one of the men says it's just an airlock. But I can't recall a mention of airlocks in real life. But assuming they are real; according to Wikipedia water is 830 times as dense as air. So why doesn't the water just push the air out of the way? On the same principle that if I tried to impede the progress of Tyson Fury I would not succeed.

Comments welcome from plumbers and physicists.

 

[fixitflav]

Physicist here! (or at least I was many years ago)
I'm sure airlocks can exist, If you have an inverted U and only gravity head available, it may not be enough to push the water over.
I agree strange things can happen. If I drain my system (open gravity with F/E tank, separate feed and vent) it doesn't refill easily, though I can't see why the water can't flow down the fill pipe and expel the air up the vent, or from the manual vent after the pump. I have to leave it 1/2 an hour and then the air is heard leaving from the pump vent. Alternatively, I can fill the F/E tank slowly, holding the ball valve, so the level is 1/2 way up the fill pipe diameter (in the tank side), and that does it.
To get back to your problem - how is it piped? Is the hot teed off the HW cylinder vent pipe in the loft? If you have some ups and downs it might cause your symptoms.

 

[ericmark]

I know had a radiator not working, and simple opening the bleed screw did not cure it, turning first the lock shield off, and bleeding, then turned back on and TRV off and bleed again, so both feed and return bled, and it then worked.

So it seems air lock is possible. But pressure flow to return is very low.

 

[Calculus]

Thanks Fixitflav, you don't meet many physicists on this forum!

I haven't thought of inverted U traps; that could be the answer. Unfortunately my plumbing knowledge is at the same level as my knowledge of embroidery. I don't know anything about vents, but I will attach two old pictures from inside the airing cupboard. The centre of the first picture shows the hot and cold pipes going through into the bathroom. The second pic was taken just below and shows the hot water pipe emerging from the cylinder (that vertical pipe is one with the vertical pipe in the centre of the first pic)

 

[Madrab]

Airlocks are more prevalent in gravity systems as the 'weight' of water sometime struggles to overcome the reciprocal force of the trapped air. They can also be created by badly designed pipework, where high spots have been created and that makes it easier for air to be trapped within those areas. They tend to be areas that rise higher than the horizontal level of the majority of the pipework. That or there is no vertical vent path to air and the trapped air cant escape easily. That air can be created by the HW cylinder when the water is heated.

Airlocks are also more common where taps/outlets drip, as that can introduce air back up through the outlet that again becomes trapped.

Unfortunately that pipework in the pics looks a tad badly designed, it currently allows any air bubbles that are vented out of the cylinder (when the water is heated) to rise up the first vertical pipe it comes across, expelling the water in that pipe and creating an airlock. When the outlets is opened the air has enough pressure to hold back the gravity water and when dropping the shower head or disconnecting it, that is creating enough flow to overcoming the resistance created by the air in the HW pipe and allows it to be expelled (the gurgling), then the HW can flow properly.

 

[fixitflav]

I haven't thought of inverted U traps; that could be the answer. Unfortunately my plumbing knowledge is at the same level as my knowledge of embroidery. I don't know anything about vents, but I will attach two old pictures from inside the airing cupboard. The centre of the first picture shows the hot and cold pipes going through into the bathroom. The second pic was taken just below and shows the hot water pipe emerging from the cylinder (that vertical pipe is one with the vertical pipe in the centre of the first pic)

There aren't many mathematicians on here either!

Is the shower valve below the level of the pipes through the wall? If so, that might not help.
One way to plumb it is to take a dedicated outlet from the CWST, and tee into the HW cylinder vent pipe in the loft. Then both pipes can be run to directly above the shower location, and drop down, minimises the work. I've done it that way in 2 houses and it's worked OK. But probably a big job to alter yours, even if you thought it was worth it.

 

[Calculus]

Thanks Madrab and Fixitflav. The plumbing is indeed less than optimum. The problems started when a builder friend who did plumbing (note: not a plumber friend who did building) did a major job (I can't remember what it was) after which the central heating wasn't working properly. I called in a (proper) plumber who was reluctant to do the work. 'Get the man who did the last job to come back to do it properly', he said, and said something about it not being 'vented'. That wasn't so easy as the builder was a friend, and the plumber did do the fix required at a cost of about 1k to myself. Some time after this I mentioned this to the builder (in a roundabout way) and he said something about colleges failing to teach about venting.

 

[MeldrewsMate]

Airlocks do indeed exist, and not only in Star Trek (other space based sci-fi films are available).
The pressure in a column of liquid is greater as the depth increases (density x force due to gravity x height), sometimes called rho-gee-aitch. The net static pressure at the shower outlet is the sum of the height of the CWS cistern water level above the bottom of the HW cylinder minus the height of liquid column from the cylinder base to the shower outlet. The dynamic pressure is less because the resistance of the water flowing in the pipes produces an effect commonly called back-pressure. This subtracts from the static pressure.
Imagine a hosepipe shaped like a sine wave, completely filled with water. If the left end is higher than the right then water will flow left to right, however if every second half-cycle is filled with air the water trying to get down from right to left will pressurise the air (density x gravity x height of that column), thus the water will settle in all the 'negative' halves of the waveform. The pressure required to blow the first column over the hill will be rho-gee-aitch if that single waveform were open vented; it is not, it is connected to the second waveform and thus their 'discharge' pressures are summed.
This means that 3 successive 'u' tubes of 0.5m height connected sequentially would need 3 x 1m of pressure to clear them.
That was badly explained! Can anyone improve on the clarity of explanation?

Back to your specific problem. @Madrab has already explained that gas bubbles will be encouraged to rise into your shower valve, in lowering your shower hose you reduce the inlet pressure needed to push the (hot) water out, the hot flow increases and drags air bubbles along with it. As more air is ejected the flow increases further.
To prevent those gas bubbles rising into your shower pipework install a 'u' bend into the hot supply pipework by rotating that compression 'tee', so that hot water first flows down before up to the shower valve, a simple job for any embroiderer or plumber.

 

[Calculus]

Thanks MeldrewsMate, that makes perfect sense and takes me back to physics classes at school many decades ago. Considering the various dodgy jobs that have been perpetrated on my house plumbing I would not be surprised to find sinusoidal or even helical piping somewhere. I think I have identified what you call the compression tee (immediately above the immersion heater), in which case it makes sense; the air would travel leftwards to the white pipe on the left, and then upwards. I don't know where that white pipe ends, but could it be the 'vent' others have mentioned?

 

[Madrab]

Yes, pic 2 centre, the 15mm pipe travelling upwards seems to be the hot feed to the shower, that needs to point downwards. That can be achieved by disconnecting the up pipe, loosen and twist the tee so the upwards leg point downwards, then square some pipe so it travels downwards, left, back upwards over the pipe running left then up to join back up to the hot feed to the shower. it's not ideal by a long chalk but would remove the tendency for the air bubbles to migrate into that shower feed pipe.

That white pipe to the left will travel upwards (the vent) into the loft and up and curve over in down into the Cold Water Storage Cistern (CWSC), that's the safety vent pipe. It travels downwards to supply the rest of the HW outlets.

 

[Calculus]

Thanks Madrab, this is all starting to make sense. I'm going to guess there is a head of water in that vent pipe (as per MeldrewMate's post) and the weight of that column of hot water balances the upward force put on the water by the pump, so that hot water doesn't normally flow into the cold water tank.

 

[Madrab]

the upward force put on the water by the pump

You say your HW is gravity so it won't be pumped. The only pumped water will be the CH water that is run through the internal coil in the cylinder that heats the water, they are separate.

The column of water in the vent pipe is leveled out at the same level as the water in the CWSC that feeds it, hence the reason for a bend up and over the CWSC (think of water level pipe that is used sometimes in construction). The bend up and over creates the extra head that the HW does not exceed unless it is over heated and then the extra pressure/expansion pushes the HW up and over the bend and then 'vents' into the CWSC.

 

[Calculus]

Thanks again Madrab. So 'venting' covers both water and trapped air. I presume the water to the hot taps is also not actually pumped. Just one more uncertainty. There is a pump in the airing cupboard ('Flomasta'). Will this be the only pump in the system, or will there be another one in the boiler? (Ideal Classic FF250)

 

[Madrab]

You have an old heat only boiler which means the system pump is external - that'll be the flomasta in the airing cupboard, that serves to pump the CH system water around the rads and through the HW cylinder coil .

The hot water won't be pumped (unless it has a separate whole house/shower pump) and just uses the height of the CWSC above a given outlet to create the pressure and flow to that HW outlet.

 

[Calculus]

Thanks once again Madrab. For the first time I am starting to understand how everything hangs together.