Willis Water Heating System

[JohnW2]

In another thread ....

The Irish it seems have cracked it with the Willis system, where it heats from top of cylinder down, so you get hot water fast, and time on controls how much heated, so in that case yes controlled time would help never used a Willis system I am Welsh we have not worked it out yet ....

@ericmark ... it's not just you. I'm not going to say anything about the Irish (apart from anything else, my wife is ¾ Irish!), I have never been able to make any sense of the 'Willis' system, particularly given that even the Irish cannot change the Laws of Physics.

The two 'extremes' of approaches to water heating are (a) to heat water only 'as needed', but that requires a very powerful heater (probably at least 8-9 kW) or (b) to heat, and store, a whole large cylinder full of hot water, and just top-up the heating as HW is used or heat 'lost' - and, starting from cold, it may take 2-4 hours to heat a whole cylinder of water with a standard 3 kW immersion heater.

The Willis system is a sort-of half-way house. One does not heat a whole cylinder of water but, rather, just heats (and temporarily stores) just the amount of water one needs for a particular task, and one does that by turning on the 'Willis' heater for the appropriate amount of time before one needs the hot water. With a 3kW (Willis or anything else) heater, it will heat (to desired temp) just under 1 litre of water per minute. Hence, the period by which one has to 'anticipate' one's need for water is something like 2-5 minutes 'before need' for hand washing/shaving etc., 10 minutes for washing up, 15 minutes for a shower or about 1 hour for a bath etc.

However, the reason I can make little sense of it is that one doesn't need a 'Willis system' to do that. The Laws of Physics being what they are, even in Ireland, mean that "... heats from top of cylinder down, so you get hot water fast" is just as true (and the amount of water heated in a given time by a 3 kW heater is the same) when heating is directly by means of an immersion in a large cylinder as if one heats the water externally (in a 'Willis heater') before transferring it to the cylinder (hot water rises).

Hence, with a conventional in-cylinder immersion (usually containing 'cold' water), if one switches the immersion on for 5, 10, 15 or whatever minutes prior to needing hot water, one will achieve exactly the same as one would with a 'Willis system'. In either case, it sounds like far more hassle than I would personally want (having to always 'anticipate' my need for hot water) but, if that didn't worry me, I really don't see why a Willis system would enable me to do it appreciably better than I could be an ordinary immersion heater and ordinary 'immersion heater switch'!

Maybe the Irish know something that I don't?

Kind Regards, John

 

[bernardgreen]

The Willis system creates hot water externally and feed it into the top of the cylinder while removing cold water from the bottom of the cylinder.

Hence there is no mixing of hot with cold in the cylinder, there is a thermocline between the hot water and the cold water

An immersion heater creates hot water that is surrounded by cold water and this creates convection currents. Hot water and cold water mix above the immersion heater and a thermocline cannot form.

 

[JohnW2]

The Willis system creates hot water externally and feed it into the top of the cylinder while removing cold water from the bottom of the cylinder.

Indeed it does.

Hence there is no mixing of hot with cold in the cylinder, there is a thermocline between the hot water and the cold water

Yes, essentially true.

An immersion heater creates hot water that is surrounded by cold water and this creates convection currents. Hot water and cold water mix above the immersion heater and a thermocline cannot form.

With a 'top' immersion (I perhaps should have mentioned that), the situation is much the same as with a Willis system. Mind you, even with a 'bottom immersion', the heated water only 'passes through' cold water once (until it cools), before becoming part of the mass of 'hot' water at the top of the cylinder (just as if it had 'first appeared there') - so I doubt that, in practice, it makes a lot of difference.

As I said, in general terms the difference is really down to whether or not one heats (and stores) more water than one needs to be heated in the immediate future, the price for that being an awful lot of hassle in terms of having to anticipate one's hot water needs. However, I still think that, in practice, there would not be much difference between the Willis system and a conventional (particularly 'top') immersion, if one 'played with the switch' in the same way - since it's that 'playing with the switch' (with Willis or conventional immersion) which is the main difference between what we are talking about and the 'conventional' system..

Kind Regards, John

 

[ericmark]

Years ago we had twin immersion heaters with a long and short heater, and latter we put the heaters in the side. And we could heat just the water over two set levels, however I know when having a bath water did run out which in theory with bath being around 30 gallons and cylinder being around 40 gallons should not happen, clearly the heaters did not heat from bottom of the tank, as to if due to wrong replacement or poor design I don't know.

I have a feeling it was all OK in North Wales, but in the South of England they had lime in the water, and the heat would leach out this lime, so I guess room was left for this lime to gather without causing the heater to over heat, I would guess Ulster is like North Wales where the water has very little lime in it?

So it seems unlikely the Willis does not work, it has been going for a long time, but maybe not used on the mainland due to the lime in the water.

But like you I would like to know why the rest of UK did not adopt this system, maybe this would be better in the Plumbing section?

 

[JohnW2]

Years ago we had twin immersion heaters with a long and short heater, and latter we put the heaters in the side. And we could heat just the water over two set levels, however I know when having a bath water did run out which in theory with bath being around 30 gallons and cylinder being around 40 gallons should not happen, clearly the heaters did not heat from bottom of the tank, as to if due to wrong replacement or poor design I don't know.

Mine are both 'on the side', and the bottom one is less than 6" from the bottom, so I wouldn't expect it to have any problem in getting the whole cylinder of water 'up to temp'. With the long 'dual' ones which went in from the top, I imagine that the 'bottom element' would be quite a lot higher than that, so the sort of behaviour you describe would not surprise me.

So it seems unlikely the Willis does not work, it has been going for a long time

The concept of thae Will system obviously 'works'. That concept is simply that one leaves the water in the HW cylinder usually 'cold' and then, when one anticipates needing hot water for a certain task in the near future, switches on the heater for the amount of time (5, 10, 15, 60 minutes, or whatever) needed to heat the amount of water one is about to use.

However, as I've said, one can adopt that very same concept by switching an in-cylinder immersian on for short periods 'when required' just as easily as by having an external 'Willis' heater. The only main difference is the initial outlay involved in buying a 'Willis' heater and having it installed.

... but maybe not used on the mainland due to the lime in the water.

I'm not sure what that has got to do with it, unless these Wallis heaters have small bore pipes in them.

But like you I would like to know why the rest of UK did not adopt this system, maybe this would be better in the Plumbing section?

If what I've said above is correct, maybe those in the rest of the UK have a better understanding than do the Irish of the fact that they can, if they so wish, achieve essentially the same without the cost of obtaining a Willis heater and having it installed?

Kind Regards, John

 

[bernardgreen]

the heated water only 'passes through' cold water once

It doesn't do 'passes through'. It mingles with the cold water and cools down, thus the heated water that reaches the top of the cylinder is not as hot as it was when it left the surface of the immersion heater.

 

[JohnW2]

It doesn't do 'passes through'. It mingles with the cold water and cools down, thus the heated water that reaches the top of the cylinder is not as hot as it was when it left the surface of the immersion heater.

Slightly - but, as I said, only once - and, in an event, as I said, I was thinking/talking primarily about 'top' immersions, such that the heated water has minimal 'rising to do..

As always, I might be wrong, but I still strongly suspect that the Willis system offers minimal advantage over what one could achieve by (equally tediously!) switching an in-cylinder immersion (particularly a 'top' one) on and off, for correctly guessed periods of time, immediately before each and every time one had an anticipated need for hot water.

Kind Regards, John

 

[ericmark]

With a short immersion switched on it would heat just first foot of water at top of cylinder, but it would take ½ hour to get warm, the report for Willis system seems to say hot water within 5 minutes, maybe only to 2 inches so would soon run out, but it is claimed hot water in 5 minutes not warm as we would get with short immersion.

 

[JohnW2]

With a short immersion switched on it would heat just first foot of water at top of cylinder, but it would take ½ hour to get warm ....

As I said, the Laws of Physics are the same in Ireland as everywhere else. It might well take something approaching half an hour to heat all the water above the top immersion (be that 'the first foot' or whatever). However, after just a minute or three, the water heated by that element (a foot or whatever below) would have risen to the top (maybe, as you say, just the top couple of inches) and be ready for drawing off for use.

... the report for Willis system seems to say hot water within 5 minutes, maybe only to 2 inches so would soon run out, but it is claimed hot water in 5 minutes not warm as we would get with short immersion.

See above. As I also said (in terms of those Laws of Physics), with a 3 kW heating element one will heat just under 1 litre per minute of water (from ambient to usual HW temp) - which leads to instructions (for Willis heaters) such as ...

A Willis Immersion heats approximately 4.5 litres of water to a high temperature every 5 minutes, so if one can decide before switching on how much water is required the pre-heating time can be calculated as follows:

• 4.5 litres for hand-washing, baby’s bath, shaving, breakfast dishes etc. = Pre-heat for 5 minutes
• 9 litres required for washing-up dinner dishes, medium sized clothes wash = Pre-heat for 10 minutes
• 9–14 litres required for a shower = Pre-heat for 10/15 minutes
• 54 litres required for a bath = Pre-heat for 1 hour

That sounds about right to me. It says that the Willis heater has to be on for 1 hour to get 54 litres of hot water for a bath. To fully heat the 140 litres in my cylinder with a conventional 3 kW immersion (the 'bottom' one this time) takes around 2.5 hours - which would equate to about 58 minutes (remarkably close to "1 hour"!) for 54 litres.

Kind Regards, John

 

[ericmark]

If you heat water with a conventional immersion heater it causes circulation so if in water by 12 inches that 12 inches is all at the same temperature, idea of the Willis is it does not circulate the water in the tank. How it does not circulate is not plain, would seem very dependent on correct pipe work design, plumbers in Wales don't seem to be able to plumb in a standard C Plan, they seemed to think with a flat and house on same boiler having two pumps it would either heat house or flat depending which pump ran, this did not work as with pump stopped you got a reverse flow.

If they can't work that out, what chance is there they could work out how the Willis system is set up?

I am sure also electricians who lack knowledge, I have worked with some, who did not have a clue how star/delta starters worked, they had spent all their live house bashing, never needed to know how a star/delta worked, so in Ireland I would love have the Willis system, but in Wales you get that glazed look in their eyes, and one feels just not worth trying.

It does seem being used a little more now with solar panels, but from what I read the pipe work is critical too little flow and over heats so switching off/on all the time, too much flow and you don't get the hot top, so not worth the chance it is fitted wrong.

 

[bernardgreen]

The Willis system is effectively a cold water tank and a "separate" hot water tank connected by an in line water heater. The volume of hot water in the hot tank increases as hot water from the in line water heater flow into the tank. This is hot water and not a warm mixture of hot and cold water.

In the Willis system the bottom of the hot tank is a thermocline layer on the top of the water in the cold tank. As volume of hot water increase and the volume of cold water decreases the thermocline layer moves down the cylinder.

For proper operation is essential that the hot water entering the top of the hot tank ( the top of the cylinder ) creates no stirring of the hot water in the cylinder as this would disrupt the thermocline.

 

[JohnW2]

If you heat water with a conventional immersion heater it causes circulation so if in water by 12 inches that 12 inches is all at the same temperature ....

Eventually. Initially, the water in the top inch or two will be a lot hotter than that in the remaining 10-11 inches (assuming the heating element is at 12 inches from top) - which, although 'warmed' (by partial mixing during upward passage of heated water through it) will be at an appreciably lower temp.

'Circulation' will obviously be limited to that top 12" (or whatever), since once water has cooled enough to 'fall' below that point, there is nothing to 're-heat it, hence cause it to 'rise' again.

... idea of the Willis is it does not circulate the water in the tank. How it does not circulate is not plain, would seem very dependent on correct pipe work design ....

Once hot water from the external heater has been transferred into the top of the main cylinder (full of essentially 'cold' water), it obviously cannot 'circulate' (by convection), since there is no source of heating in that cylinder to cause any 'heated water' to rise.

Of course, there is conduction as well as convection - which, eventually, would result in all the water in the cylinder (and the cylinder itself) coming to be at more-or-less the same (relatively low) temp, even if one had only introduced a small amount of hot water into the top of the cylinder. However, that would take a very appreciable amount of time, and the whole idea of the Willis system (or emulation of it with a conventional immersion) is that the hot water is used almost immediately after it has been heated, leaving little time for a significant amount of temperature equilibration by conduction.

Kind Regards, John

 

[JohnW2]

The Willis system is effectively a cold water tank and a "separate" hot water tank connected by an in line water heater. The volume of hot water in the hot tank increases as hot water from the in line water heater flow into the tank. This is hot water and not a warm mixture of hot and cold water.

Indeed, but if one has a HW cylinder with an immersion close to the top, then that top part of the cylinder almost becomes the equivalent of the external 'Willis heater'.

I do not disagree with the theoretical points you are making (most/all of which relate to mixing of hot and cold water in part of a cylinder) but, as I have said, I'm far from convinced that they represent a major issue - and remain of the view that the major downside of the Willis system (or an attempt to emulate it with immersions) is that of "inconvenience".

It seems to me that by far the most fundamental advantage of the Willis system derives from the fact that one only switches the heater on for brief periods immediately before each and every time one needs hot water, and never usually stores significant amounts of hot water for more than a few minutes. I think that such a situation would drive me mad (but perhaps the Irish are more tolerant!) but, as I've said, one could achive much the same with a high immersion in a conventional HW cylinder. If one had 3 or more immersions (at different heights) in the cylinder, rather that the usual two, then one could get even closer to the Willis system - but it would still drive me mad!

In fact, having a 3 kW Willis heater is essentially the same (including the amount of 'anticipation time' required) as using a 3 kW kettle to heat as much water as one needed, each time one needed hot water - but, again, that would drive me mad!

Kind Regards, John

 

[ericmark]

The temperature at top of tank must be dependent on speed of flow, looking at the solar panel version the is a thermostat as the water joins the expansion pipe, there needs to be some control one would think to ensure flow rate matched the heat input.

I should think one could have an array of thermostats so you can select how much water, but what we need is an Irish man to say in real life how they work.

 

[JohnW2]

<a diagram>

I don't get that diagram at all - i.e. I cannot see how it could work.

If the pipework were of similar size for the connections to the cylinder and Wallis heater were the same, when you turned on a hot tap, you'd get a roughly 50:50 mixture of (cold) water from the cylinder and heated water from the Wallis.

More to the point, what you depict could not supply more hot water than (the small) volume contained in the Wallis heater (mixed with cold from the cylinder, as above), since there is no way that it would get from the heater into the cylinder.

The idea of the Wallis system is that the externally-heated water (however much one needs) is transferred to the top of the cylinder, but I can't see how what you show could achieve that. In fact, other than a small amount, very slowly, by conduction, I can't see how any heated water would/could get into the cylinder with what you depict.

Kind Regards, John